Alan Wolff • Sally Taylor

ENHANCIN
G
PATIENT
CARE
The Wimmera quality improvement and
risk management model
 Enhancing
Patient Care
A practical guide to improving quality and
safety in hospitals

Alan Wolff • Sally Taylor

DEA

I
T

0
 9

L
I
B
R
A
R
Y

Alan Wolff, MB BS, MD, MBA, DipRACOG. FRACGP, FACHSE,

MRACMA Director of Medical Services and Director of Emergency
Department.' and Adjunct Associate Professor^
Sally Taylor, RN, RM. BN, HDNC, MAppMgt(Health) Clinical Risk
Manager1
1 Wimmera Health Care Group. Horsham. Victoria • 2 School of Medicine. Flinders University
Adelaide Sooth Australia
 First published in 2009 by MJA Books
Australasian Medical Publishing Company Limited (AMPCo) ACN 000 005 854
AMPCo House 277 Clarence Street Sydney NSW 2000 Australia
Locked Bag 3030 Strawberry Hills NSW 2012 Australia
Tel +61 2 9562 6666 • Fax +61 2 9562 6699 • Email medjaust@ampco.com.au
This book is copyright. Apart from any fair dealing for the purposes of research or study, criticism
or review, as permitted under the Copyright Act 1968, no part may be reproduced by any process
without written permission. All enquiries should be addressed to the publisher.

National Library of Australia cataloguing-in-publication data

Enhancing patient care: a practical guide to improving quality and safety in hospitals

Bibliography.

Includes index.

ISBN 978 0 9775786 6 5 (pbk ).

I. Wolff, Alan II. Taylor, Sally

1 Hospital - Safety measures.

2 Hospital care - Quality control.
3 Hospital patients - Care - Quality control.
© Text: Wimmera Health Care
Dewey Number: 362.11
Group © Design and Typography:

AMPCo
 Foreword

As the title indicates, this is a "practical guide" to enhancing patient care through the improvement of
quality and safety in hospitals.

A clear understanding of the key concepts of patient safety, risk management and quality
improvement is critical to providing the care that we, as individuals and as a community, expect from
our health system. Despite well trained, enthusiastic health care professionals trying to deliver the
very best care possible, the system often fails them by a lack of support, encouragement and resources
to colled the data, implement the processes and teach the methods that will help to deliver the very
best care possible. This practical guide to implementation will help managers, and health
professionals at all levels, achieve their goals in safety and quality.

Implementation of processes and methods that are known to work is likely to be the most cost-
effective way of improving care. This guide takes the best information about implementation available
from the literature and adds the authors' practical knowledge, which has been gained from almost 20
years of hands-on experience at their local hospital in rural Victoria. Their successes have been well
publicised and are a credit to them.

Improving quality, addressing risk and providing the safest possible care is not an easy task, with
many barriers and some resistance from busy doctors and nurses. Through their own experiences, the
authors clearly understand the clinician's point of view, which should help reduce resistance to
change. The long-term demonstration and evaluation of successful improvement in an Australian
hospital using established methods from health care and other industries gives the guide considerable
authority. It is easily accessible, well organized, clearly and simply written and is well referenced.

The authors have set out to produce "an integrated set of quality improvement and risk management
modules that can be used individually or in various combinations" and "a practical, step-by-step
approach to implementing and maintaining an effective clinical quality and safety program*. In my
view, they have achieved their goal and have, in so doing, provided great help and a marvellous
resource document for everyone who is actively trying to improve care delivery.

Bruce Barraclough, AO President. Internationa) Society

tor OuaMy in HmM Cam
"tmir Hit- Tnilti 'fifnln' riiniriil TirnHnnr* ..............

Contents

Foreword........................................................................................................................................................................................... ™

Acknowledgements............................................................................................................................................................................. B

Preface.............................................................................................................................................................................................. ™
*
Introduction....................................................................................................................................................................................... ™

1 An overview of clinical governance............................................................................................*

1.1 The roles, structures and processes of health care management.......................................1
1.2 Governance........................................................................................................................2
1.3 Clinical governance............................................................................................................3
1.4 The components of a clinical governance program.............................................................5
1.5 Reporting the results of the clinical governance program...................................................II
1.6 Accountability for the clinical governance program in health services..............................11
1.7 A health service's role with the major components of a clinical governance program.......12
2 Quality improvement — measuring clinical performance and closing gaps in clinical care....17
2.1 Defining quality improvement..........................................................................................17
2.2 Deciding the clinical areas in which quality improvement activities should be undertaken18
2.3 Components of a quality improvement program..............................................................19
2.4 Quality improvement collaborates...................................................................................26
2.5 Whole-of-practice audits..................................................................................................27
2.6 Clinical indicators............................................................................................................29

3 Risk management — why do adverse events happen to patients?..........................................37

3.1 Health care delivery as a system......................................................................................37
3.2 The extent of adverse events............................................................................................38
3.3 The reasons why errors occur..........................................................................................39
3.4 Errors and adverse events...............................................................................................
3.5 The response to errors.....................................................................................................41
3.6 The basic risk management process................................................................................43
3.7 The basic risk management process in health care..........................................................43
3.8 Characteristics of an ideal quality improvement and risk management system...............44
3.9 The Wimmera quality improvement and risk management model....................................44

4 Dctccting adverse events and risk.........................................................................................49

4.1 Methods of detecting adverse events and risk................................................................. 49
4.2 Reactive risk management — detecting adverse events....................................................50
4.2.1 Direct observation.................................................................................................. 50
4.2.2 Legal claims.............................................................................................................50
4.2.3 Medical record review..............................................................................................51
4.2.4 Clinical incident reporting.......................................................................................63
4.2.5 Sentinel event reporting.... .....................................................................................71
 4.2.6 General practitioner feedback..................................................................................74
4.2.7 Clinical pathways and variance analysis..................................................................76
4.2.8 fetieni satisfaction....................................................................................................77
4.3 Proactive risk management — detecting risks.................................................................78
4.3.1 Consultative councils and committees.....................................................................78
4.3.2 Coronial reports.......................................................................................................78
4.3.3 Information from insurers.......................................................................................79
4.3.4 Statewide and national databases of clinical incidents............................................80
4.3.5 Media reports...........................................................................................................83
4.3.6 Clinical journals......................................................................................................84
4.3.7 Pharmaceutical and equipment alerts......................................................................84
4.3.8 Inquiries into individual health services or specific clinical areas in health services....
85
4.4 Additional proactive risk management methods...............................................................86
4.5 Systematic review of external information about risk.......................................................88

5 Analysing adverse events and prioritising action....................................................................95

5.1 Managing a patient who has experienced an adverse event..............................................96
5.2 Managing a stafT member who has been involved in an adverse event.............................97
5.3 Analysis of adverse events................................................................................................98
5.4 Prioritising adverse events for action..............................................................................103
5.5 The surveillance committee............................................................................................105
5.6 Principles for prioritising action to prevent adverse events recurring.............................109

6 Taking action to close gaps in care and prevent adverse events recurring — evidence and theory
119
6.1 Determinants of implementing a change within a health service....................................121
6.2 Theories relevant to the implementation of change and their practical implications for
 change in health care.....................................................................................................126
6.3 Approaches to change....................................................................................................141
6.4 Developing change strategies.........................................................................................141
6.5 The evidence about change strategies............................................................................144
6.6 Change strategies commonly used at the Wimmera Health Care Group.........................145
6.7 Actions commonly taken to prevent adverse events that are generally ineffective............................................149
7 A practical framework to close gaps in patient care and reduce risks to patients.................155
7.1 A practical framework....................................................................................................155
7 2 Motivating doctors to participate in quality and safety programs...................................177
7 3 The framework in action — clinical pathways................................................................179

*
8 Reporting the activities and results of the clinical governance program.................................W

9 A quality improvement and patient safety culture...................................................................I"

9.1 Types of organisational culture......................................................................................'91
9.2 Components of an informed culture.............................................................................192
9.3 Types of safety cultures................................................................................................194
9.4 The safety spectrum in health services — from resilience to vulnerability.....................195
9.5 Measuring the patient safety culture.................... .......................................................198

10Eighteen practical steps to implementing a clinical risk management program in j health service ...
203

11Implementing a clinical risk management program in a small hospital..............................217

11.1 Problems unique to small hospitals that want to undertake a clinical risk
management program.................................................................................................217
11.2 Modifications required for programs in small hospitals...............................................217
 11.3Screening and medical review......................................................................................218
11.4Obtaining the support of small hospitals.....................................................................218
11.5The logistics of screening and medical Teview.............................................................218
11.6 Reference panel recommendations..............................................................................219
11.7Small hospital clinical risk management manual........................................................220
11.8Introduction of the program across rural Victoria.......................................................220
11.9Changes to the screening criteria................................................................................221

12Important messages...........................................................................................................223
12.1Health care as a system...............................................................................................223
12.2Clinical governance.....................................................................................................223
12.3Quality improvement...................................................................................................224
12.4Detection of adverse events and risk...........................................................................224
12.5Analysing adverse events.............................................. .............................................225
12.6 Prioritising gaps and risks...........................................................................................225
12.7Closing gaps and reducing risks..................................................................................225
12.8Conclusion..................................................................................................................227

Appendix 1: Wimmcra Health Care Group Adverse Event - Analysis Form.............................229

Appendix 2: Clinical risk management executive summary....................................................232

Index......................................................................................................................................235

Acknowledgements
Fortunately, in 1987, Duncan Stuart introduced adversfe event screening to Australia after visiting the
United States to find a suitable clinical risk management program for Royal North Shore Hospital. His
work influenced health professionals in other hospitals, and he generously organised seminars to
explain the Medical Management Analysis model. The Wimmera clinical risk management model was
established after one such seminar. Without Duncan's contribution to Australian quality and safety,
the Wimmera model would not have been developed and this book would probably not have been
written.
For a quality improvement and clinical risk management program to flourish jn a health service for
20 years requires the suppon of many dedicated individuals working together as a team. Most
critically; the culture at the Health Care Group has supported the provision of high-quality, safe care,
which has allowed the program to develop, with the benefits to patients always being paramount in
any discussion Wr thank all the staff at the Health Care Group, Medical Staff Group, Executive and
Board of Management for their support of the program over many years.
Over the years, the members of the Patient Care Committee have generously given their time to
participate in the program; in panicular, Ian Campbell, a visiting surgeon, and David Leembruggen, a
visiting general practitioner, enthusiastically panicipated in the program from day one. They have
reviewed a huge number of medical records and contributed enormously to robust discussions in
Patient Care Committee meetings, and supported the program and its recommendations to senior
medical staff. Gram Phelps and Edward Janus, both visiting physicians, were long-term contributors.
Jo Bourke worked tirelessly as the first clinical risk manager in what was largely uncharted
territory. She gained the respect of medical staff, which allowed nursing staff to participate in formal
discussion of adverse events, helping us formulate comprehensive strategies and take action to
prevent these events recurring. After it became clear that many adverse events clustered around
patients with panicular diagnoses, Janeue McCabe became the driving force in establishing the
clinical pathway program, which resulted in dramatic improvements in the quality of care for many
patients.
The success of a comprehensive medical record review program depends on the strong suppon of
Health Information Services. We thank Cathy Dooling, the former manager of the service, for
developing and introducing the medical record screening process in her department, and Briana Farr,
the current manager, and their staff for their continuing suppon. Kieran Loughran and his stafT have
provided excellent information technology suppon over many years.
Rob Irvine very generously provided facilities at the Horsham campus of Ballarat University that
were most conducive for writing. Shirley Mewett, our librarian, found every reference that we have
required, however obscure the journal, which assisted greatly with the book's literature review and
references
James Dunbar, John Christie, Caroline Brand, Les Reti, Karen Dunn, Mane Aitken and Alicia
McGrath most kindly agreed to review the manuscript and made many very helpful suggestions for
improvement. Graeme Exell kindly proofread the final draft. We are grateful for the lime they took and
effort they made to provide valuable feedback about our work.
The generous suppon of the Victorian Department of Human Services, initially with Ginical Risk
Management Pilot programs, and then the Limited Adverse Occurrence Screening program, has been
much appreciated.
The book would not have been possible without the magnificent suppon and expenise of Rivqa
Bcrgcr. our editor at MJA Books, and the guidance of Ruth Armstrong, Deputy Editor of the Mcdical
Journal of Australia, and Glenn Carter, Production Manager, and Peter Humphries, Production
Coordinator
Finally, our thanks go to our spouses Yvonne Cymbalist and Peter Taylor for their suppon during
the lengthy process of writing and editing.
Preface

Small ideas and beginnings can have large ripple effects. Wimmera Health Care Group, where this
clinical quality improvement and risk management model was developed, is 300 km north-west of
Melbourne and has approximately 10 000 admissions a year. Over 19 years, we have developed a tew
basic ideas into an efficient quality improvement and clinical risk management model.
The model has now been adopted by many hospitals, ranging in size from tertiary centres in capital
cities to rural hospitals with one doctor on staff. These hospitals implemented the model because a
person, or group of people, had a commitment to and passion for providing high-quality care, and
continually improving that standard of care. This drive may come from a board member, a senior
executive, a quality or clinical risk manager, a clinical champion for quality or, ideally, a group of such
people.

Strong motivation and dedication are necessary, because, sooner rather than later in the process of
implementing and maintaining a quality improvement and clinical risk management program,
adversity will strike. It is probably the response to things when they go wrong that most determines
whether the program will be successful. Many people will feel threatened (for various reasons) by the
issues raised by such programs and the changes required to implement and maintain them. The
temptation will frequently be to give up and claim that it is just too difficult to achieve the required
change. But health care is complex, and perfect solutions to problems are unlikely to exist. Most
decisions that attempt to change complex systems will result in unintended negative events. What is
required when things go wrong — and they most certainly will — is for you to pick yourself up, dust
yourself off, reaffirm your beliefs about what is required, and return to the battle the wiser for the
experience.

So when applying the principles provided in this book, allow for the complexity of health care delivery
systems and human nature, and continue on through the turbulence. After 19 years of the program at
Wimmera Health Care Group, it has become easier. We are reminded daily of the many obstacles to
achieving change, but also of the significant improvements we have made in the quality and safety of
patient care through the dedication of staff members. Our vision is to continue this pathway of
improvement to achieve the safest and best health care for all in the Wimmera.
We wish you the best for your endeavours during this exciting journey and would be pleased to receive
feedback from readers via our website (www.whcg.org.au).

Alan Wolff and Sallv Tavlor

Introduction

Health care is primarily about improving the quality of patients' lives, but its delivery can be a
complicated process that requires the use of many complex systems. The treatments available tor
providing patient care are becoming increasingly more effective and complex, with an increased risk at
errors occurring. Demand for health services is usually high, and advances in knowledge and
technology are rapidly and continuously being made, while the resources available to provide services
remain limited. As well as "a complex environment with multiple stakeholders, [there are) conflicting
objectives and considerable restraints".'
Much of the care provided to patients is not supported by scientific evidence. There is wide variation
in the treatment being given to patients with the same clinical conditions, and a significant proportion
of the care recommended for the leading causes of death and disability are not being provided. 2
Further, some care that is provided causes preventable harm to patients, 3 and most errors are
thought to be undetected and unreported. 11 Considerable cost is being incurred and valuable
resources expended in providing suboptimal care.
There is an expectation by the community that health care will be of high quality, that people
receiving health care will be safe while it is being delivered, and that errors in providing care are
"neither acceptable nor inevitable". 4 Consequently, if the anticipated outcome of care is not achieved,
retribution is often sought; hence, litigation by patients against clinicians and health services is
increasing.4 At the same time, health professionals and management are being held increasingly
accountable for the quality and safely of the care they provide, not only by patients and their relatives,
but also by governments and regulatory and funding bodies. Despite these conditions, the
management of many health services still predominantly emphasises the service's financial situation
and patient throughput targets, rather than the quality of care and the level of patient safety within
the health service.
In our experience, most health care professionals want to provide the best possible care for their
patients. However, even with the very best of intentions and the greatest vigilance, suboptimal care
and errors in patient management occur frequently. Given the conditions under which health care is
currently provided, how can health services and the professionals working within them ensure that
they consistently provide a high-quality, safe service to their patients? The structures, processes and
systems of health care delivery are created under the influence of leadership. Culture and
performance need to be effectively monitored, and systems appropriately changed in response to the
quality and safety of the care that is being provided.
There is now a great deal written about the quality of care provided to patients and the level of
patient safety in health services. Twenty years ago, the occasional anicle about quality or safety would
appear in major clinical journals. Now, whole journals are devoted to these topics, and the volume of
literature is considerable. The terminology used in these publications has expanded, and can be
confusing. There has been much activity being undenaken aimed at improving the quality and safety
of clinical care, but there are still significant gaps in the quality of care provided to patients, and the
number of adverse events experienced by patients while receiving care is still substantial. Health
services need a simple, clear-cut and practical approach to delivering high-quality and safe health
care to assist with their quality improvement and risk management programs.
This book has been written to assist the many health care professionals and health services who arc
strongly committed to providing high-quality, safe care to their patients, and who arc searching for the
best way to provide such care. It is written from the perspective of a health service that independently
wishes to improve the quality of care it provides for its patients and raise the level of patient safety in
their facility, without these requirements being imposed on the service by external regulatory,
accreditation and funding bodies. We have found that in most health services, there are individual
clinicians and managers with a genuine passion and internal drive to provide the best care possible
for their patients — without having this forced upon them as a requirement of their senior
management or from external bodies.
 Much of what is written about quality of care and patient safety is theoretical, and the evidence —
especially about the effective implementation of quality and safety strategies — is limited. Relatively
tittle is written of practical value to individual health services tTying to satisfactorily address these
important issues in their facilities. The task can be overwhelming for clinicians charged with the
responsibility of developing, implementing and maintaining a quality and safety program for their
individual health services Without clear overall direction from a comprehensive organisation-wide
quality and safety program, individual health services may move in many directions simultaneously,
with little overall integration and coordination. The next practical steps that individual health services
should undertake in their quality and safety programs are often not clearly visible.
To fill this gap, we describe a simple and practical framework that can act as a signpost for health
services wishing to establish a quality and safety program (or enhance an existing program) to
effectively monitor and improve quality and safety in all the clinical areas of their health service. The
framework is:
■ logical
■ fully integrated
■ easy to understand
■ based on relevant theory, evidence, and 19 years of practical experience in designing and
implementing a comprehensive clinical quality improvement and risk management program at
the Wimmera Health Care Group in Horsham, Victoria.
We describe in detail an integrated set of quality improvement and risk management modules that
can be used individually or in various combinations. A practical, step-by-step approach to
implementing and maintaining an effective clinical quality and safety program will also be described,
together with practical examples outlining how each component of the program can be used in
practice.
References
1 Res V. Sutherland K. Organisational change: a review tor health care managers, professionals and researchers.
London: National Co-ordinating Centre for NHS Service Delivery and Organisation R & 0.2001.
2 McGtyrm EA. Asch SM. Adams J. et al. The quality of health care delivered to adults in the United States. N Eng J
Med2003: 348: 2635-2645.
3 Brenoan TA. Leape LL. Laird NM. et al. Incidence of adverse events and negligence in hospitalised patients. N Eng J
Med 1991; 324: 370-376
4 Bion JF. HeffnerJE Challenges in the care of the acutely ill. Lancet 2004; 363; 970-977. □

1 An overview of clinical governance

This chapter wMf:

■ Discuss the organisation of health care management
■ Define clinical governance
■ Describe the major components of clinical governance
■ Discuss accountability for clinical governance
■ Describe the role of health services in establishing and maintaining effective clinical governance.

1.1 The roles, structures and processes of health care management

Management has four main roles in any organisation, including health services — planning,
organising, leading and regulating. Regulating involves monitoring and evaluating the health service's
performance and determining whether activities are being carried out as planned, thereby allowing the
health service to achieve its goals. In essence, regulation involves determining the performance
required within the organisation to achieve a particular goal, monitoring the actual performance of the
organisation towards achieving the goal, and, if there is a significant deviation between the desired and
actual performance, taking appropriate corrective action. The organisation's goals should also be
achieved using the resources the organisation has available — both efficiently (with a minimum of
waste) and effectively (ensuring the right things are being done).
Organisations usually have a distinct purpose for their existence or a main goal they wish to achieve
through their activities. For health services, this goal usually involves providing their communities
with the best range and quality of health care possible with the available resources.
To help meet such goals, health services are provided with resources to construct buildings, employ
skilled staff and purchase appropriate equipment. Most health services have developed strategic plans
that include the range and level of clinical services they will provide. The delivery of these clinical
services is usually supported by a number of structures, including policies and procedures detailing
how work is to be undertaken in the organisation, job descriptions, and lines of authority from the
board of management through to executive staff, department heads and all other staff
A committee structure is usually in place to aid decision making and to review activities occumng
within the health service. For example, a pharmacy committee decides which medications are
approved for use by medical stall. A medical appointments committee appoints all senior medical staff
and determines the range of clinical services, including the operations and procedures each senior
doctor can provide within the service. The standards against which health services are assessed by
national health care accreditation bodies will, among other things, detail the structures that should be
in place in each service and how the structures should be assessed to determine whether they are
functioning effectively.
Using these structures, a number of systems (or sequences of processes) of care are then created to
provide specific clinical services to patients. For example, when an individual presents to a health
service's emergency department, how will the urgency of presenting problem be determined? Which
nurses and doctors will assess the patient's clinical condition and provide clinical care for them? If the
patient requires investigations, how will they be requested, and how will the results be communicated
to medical staff? If the patient requires admission, what administrative and clinical processes will need
to be followed for the patient to be transferred from the emergency department to a ward? Using these
 systems, health care is provided to patients and clinical activity occurs
within the health service. The culture, leadership and education programs
in the health service influence these systems. The process is represented
in Figure 1.1.
From a management perspective, clinical care is usually regarded as the health service's output. As
with other industries, one of management's functions and responsibilities is to regulate the output
of the areas of the organisation for which they are responsible. Such regulation will need to be
applied over the volume and quality of the organisation's output.

1.2 Governance
Another term for the activity of regulation in organisations and the framework used to achieve it is
governance. In business, the principles of corporate governance include the oversight of management,
the appropriate breadth and depth of expenise of organisation's directors, the directors' independence,
ethical and responsible decision making, integrity of financial reponing, monitoring and management
of risk, maintenance of full-time internal audit, encouragement of enhanced performance within the
organisation, and protection of the shareholders' interests. 1-2 Major tenets of corporate governance
include fairness, transparency and ethical business practices. 1
Previously, governance in health care organisations focused primarily on Financial governance.
About 90% of board of management meeting agendas comprise financial items, and only 10% of items
are clinical in nature.4 Discussions at board of management meetings often focus on issues such as
whether:
■ the service can meet its financial obligations adequately and in a timely manner
■ external financial audits are being undertaken regularly, and whether their results are satisfactory
■ areas of financial concern detected by these audits have been properly addressed
■ revenue obtained from private patients utilising the health service has been maximised
■ the health service's expenditure in areas such as pharmaceuticals, diagnostic services or ambulance
transpon are within the allocated budget for these areas.
Although these areas of the health service are clearly important, it is equally important that "the
quality and safety of patient care is not pushed from the |board| agenda by immediate operational
issues", but is placed "at the heart of the board agenda". 4
More recently, there has been a move to extend the regulatory function of management in health
services to include regulation of the health service's clinical output. Clinical output refers not only to
the numbers and types of patients being treated, but also the quality of clinical care that patients
receive and the level of safety they experience while receiving that care. Regulation of the quality of
Rgw* 1.1 Tha structure and procMsas of care that is provided and the level of patient safety in
health car* IMRi^tfMllt clinical areas in health services is referred to as
"clinical governance"
 1 An overview of cinicalgowemeno*

1.3 Clinical governance

Clinical governance has been defined differently by health systems in various countries; there is no
single agreed definition. The term was first used in the Ufiited Kingdom and was defined as "a system
through which [National Health Service] organisations are accountable for continuously improving the
quality of their services and safeguarding high standards of care by creating an environment in which
excellence in clinical care will flourish".5 In Australia, clinical governance has been defined as "the
system by which the governing body, managers and clinicians share responsibility and are held
accountable for patient care, minimising risks to consumers and for continuously monitoring and
improving the quality of clinical care". 6 The World Health Organization has divided clinical governance
into four aspects: professional performance review, resource use, risk management and patient
satisfaction.7 Clinical governance "should consist of an integrated system of different types of activity
all aimed at improving quality of care" 4
In a practical and operational sense, clinical governance has also been described as
doing the right thing, at the right time, by the right person — the application of the best
evidence to a patient's problem, in the way the patient wishes, by an appropriately trained
and resourced individual or team ... working within an organisation that is accountable for
the actions of its staff, values its staff (appraises and develops them), minimises risk and
learns from good practice and indeed mistakes. 8
Clinical governance aims to create a system to continuously improve the quality and safety of
patient care. As corporate governance concentrates on the activities of the board of management and
executives and financial and business issues, clinical governance focuses on what occurs at the
bedside and in wards, specialist units and clinics where health services are provided. 9 The principles
and activities that make up a clinical governance program are not fundamentally new,10 but have
previously been used in "a disparate and fragmented" 11 manner with each component having operated
as a "silo" within the organisational structures of health services over many years. 12 This disjointed
approach often meant that decisions in one area of clinical governance were made in isolation from
key issues in other related areas. 4 What has been missing from these clinical governance activities
has been a framework to systematically incorporate and integrate all these related activities into one
comprehensive program to most effectively improve the quality of patient care and the level of patient
safety.
Put simply, clinical governance has two major components — quality of care provided the patients
and patient safely (Figure 1.2). 13

1.3.1 Quality of care

The Institute of Medicine in the United Slates defines quality of health care as "the degree to which
health services for individuals and populations increase the likelihood of desired health outcomes and
are consistent with current professional knowledge." 14 In basic terms, quality of patient carc can be
defined as ensuring lhai ihe nght things happen lo patients more often 13 The quality of carc
componcni of clinical governance therefore concentrates on the positive things that happen to
patients
 The first step to getting patient care right is, most often, knowing the
right thing to do. The right thing to do for a patient may need to be
determined according to the best available evidence of an intervention's
effectiveness, the appropriateness of that intervention for that patient, the
safety of the intervention, the overall health needs of the health service's
target population, and the resource constraints under which the health
service is functioning. Whether the intervention represents value for
money and whether it provides the greatest benefit for the health service's
population may also need to be determined. To ensure that the right thing
is consistently and increasingly well undertaken in a health service, staff
require appropriate training and adequate resources need to be provided.
Ensuring that the right thing is done at the right time for a patient will
also require attention when health care delivery systems are being
developed.
Evaluation of the quality of care provided to patients requires that a
standard of clinical performance be set. Best practice in a particular
clinical area may be determined from sources external to the health
service, such as scientific evidence from research studies published in
medical journals, national and international clinical guidelines, or clinical
practices at health services that are regarded as exemplars in providing
care in that clinical area. The health service's actual performance in
providing care for patients in this clinical area is then measured. If the
actual performance deviates significantly from what is regarded as best
practice, the deviation is analysed and appropriate corrective action is
taken to close the gap between the actual performance and the desired
standard. Importantly, ongoing monitoring of the health service's
performance should continue to determine whether the action taken has
successfully closed the gap.
The process of closing the gap between the clinical care that is being
provided and best practice is referred to as quality improvement. In some
instances, it may not be possible to achieve completely, and the cycle
perpetuates indefinitely in a process of "continuous quality improvement",
where performance continues to improve and gaps continue to narrow. It
is also important to recognise that what constitutes best practice in a
particular clinical area will change with lime, and therefore should be
regularly redefined to incorporate advances in medical knowledge.

1.3.2 Patient safety

The second major component of clinical governance in a health service is
patient safety.13 The Agency for Healthcare Research and Quality in the
US defines patient safety as "the absence of ihe poiential for, or the
occurrence of, health care-associated injury to patients created by
avoiding medical errors as well as taking action to prevent errors from
causing injury".15 Patient safely can be simply defined as ensuring that the
wrong things happen to patients less frequently. 13 This component of
clinical governance concentrates on the negative things that happen to
patients and taking action to ensure they occur less frequently.
 The level of patient safety in a health service can be assessed reactively
or proactively. Using the reactive approach, clinical activity is
X3nca/gtnwnence ... is continuously
not a method for
managers to govern or
control doctors, nor is it a
method tor doctors to
govern or control
managers or other
clinicians.'

Perkins R.
et al N Z
Med J 2006;
119: U2259
 Enhancing patient care

1 An overview ol clinical
flcwemeno*

monitored, and if an adverse event is detected, it is analysed and appropriate action is taken to
reduce the probability of recurrence. In the proactive approach to patient safety, systems of care are
examined lo identify weak processes that represent an unacceptable level of risk of patients
experiencing an advene event while they are receiving care. Once identified, changes are made to
these processes and their systems are strengthened to reduce the probability of an adverse event
occurring. As with quality improvement, ongoing monitoring of clinical activity should continue to
determine whether the action taken or the process changes made have successfully reduced the
chances of a patient experiencing an adverse event. Reducing the probability of an adverse event
occurring is called risk management.
An important precondition in establishing risk management programs in health services is that
qualified privilege has been obtained. Under the relevant legislation, information obtained from
discussions in quality improvement committees about adverse events and errors oannot be disclosed
or used in evidence in legal actions against hospitals and clinicians. Such privilege helps clinicians to
openly discuss errors during quality improvement committee meetings. 9
Importantly, there is considerable overlap between these two components of clinical governance
(quality improvement and risk management). For example, an adverse event that is experienced by a
patient may be detected by the risk management program, but could have occurred because of a gap
in the quality of care provided. The essence of clinical governance is to ensure that the right things
happen to patients more often (by making it easy to do the right thing) and that wrong things happen
to patients less often (by making it difficult to do the wrong thing). 16

1.4 The components of a clinical governance program

These major objectives of clinical governance are achieved using quality improvement and risk
management. Clinical governance can be simplified into several basic components, comprising
principles and activities that can be described and grouped in different ways. Despite the varying
terminology, the main components of a clinical governance program in a health service include:
■ Evidence-based medicine — providing clinical care for patients in accordance with best practice
determined by scientific evidence.
■ Credentialling and privileging — determining that clinicians working in the health service have the
required qualifications, training and experience to safely undenake the clinical care they are
providing, and that this care will continue to be provided to a high standard over time.
■ Clinical audit — measuring the performance of individual clinicians and the health service's
aggregate clinical performance in providing care to patients, and comparing these performances
with best clinical practice and the performance of similar health services. Measurement should be
across six dimensions of patient care — safety, effectiveness, patient-centeredness, timeliness,
efficiency and equity.17
■ Professional development — providing clinicians with appropriate continuing medical education
and skills training necessary to provide high-quality, safe care.
■ Consumer involvement — determining the needs of individual patients receiving care in the health
service and the needs of the health service's patients as a group, and meeting these needs.
■ Organisational factors — that appropriate resources, strong leadership and commitment, adequate
education and support, and clear, timely and direct communication are provided by the health
 service within a culture that supports quality improvement and risk management. Ensuring these
factors arc in place may require substantial culture change.
Many of these activities are cornerstones of providing high-quality, safe carc. and have been used
for a long time in health services. Some of the activities overlap in their role in the overall clinical
govcmancc
a health service. It is now timely and appropriate to include them in one comprehensive,
integrated ctuucal governance model to continuously and effectively improve quality and safety in
individual health services. We will examine each of these activities of clinical governance to
demonstrate how they fit into the overall model and discuss how to operationalise each activity so
they can be most effectively used by individual health services.

1.4.1 Evidence bawd medicine

To know the right thing to do for a patient requires that the clinical care provided is based on the
best available scientific evidence about which interventions provide the best chance of obtaining the
most favourable outcome for patients (eg, which investigations will provide the most useful
information for medical staff, and which medications will provide the greatest probability of
improving a patient's outcome). Such information is obtained from the results of clinical trials. The
highest quality information is obtained from systematic reviews of multiple randomised controlled
trials (RCTs); single RCTs may also provide high-quality evidence. Observational studies and expert
consensus provide weaker evidence. The stronger the evidence, the more likely the intervention
studied will improve the patient's outcome. 18
Importantly, despite the increasing number of interventions demonstrated to be effective by
RCTs,19 the answers to many questions that commonly arise when providing medical care have not
been determined in such trials. In these situations, information from non-randomised trials or expert
consensus opinion is viewed as representing best practice in those clinical areas, and this
information is used to provide care.
Using information from the best available clinical trials or expert opinion in providing care to
patients is known as practising evidence-based medicine. Evidence-based medicine has been defined
as
the conscientious, explicit, and judicious use of current best evidence in making decisions
about the care of individual patients. The practice of evidence based medicine means
integrating individual clinical expertise with the best available external evidence from
systematic research.20

1.4.1.1 Limitations of evidence-based medicine

Wherever possible, a health service's protocols and guidelines detailing how to manage patients with
particular clinical conditions should be based on the best available scientific evidence. There are,
however, limitations of attempting to practise evidence-based medicine:
■ High-quality evidence from RCTs is not always available to answer the clinical questions that may
arise in everyday clinical practice. Expert clinical judgment will, to some degree, always be
required to develop best-practice guidelines and in clinical practice. 21 Many of these questions
cannot be answered because the necessary trials would not gain the required ethics approval.
■ When high-quality evidence is available, the RCTs on which the recommendations are based have
often excluded patients with comorbidities. Therefore, the findings may not be directly applicable
to individual patients who have comorbidities and the condition studied in the trial.
■ Recommendations from RCTs are based on crude averages or summary results of all the patients
who participated in the trial. These aggregated results do not accurately represent the typically
 Enhancing patient care

substantial and highly skewed vanation in baseline risk of the patients in the trial, and therefore
often cannot be meaningfully and easily applied to a typical patient. Most negative outcomes in
trials occur in participants who are at high risk for such an outcome. However, most patients
participating in these trials are at low risk for an adverse outcome. Further, the differences in
benefit versus harm for different groups in RCTs are not routinely determined and reported.
Importantly, if the therapy evaluated in the trial can benefit patients as well as harm them,
patients with a low baseline risk may be more likely to be harmed by the therapy than benefit from
it. Ideally, in everyday practice, when
1 An overview atdlncMQavtmtnoa

clinicians are determining which

patients should receive a
recommended therapy, they should
stratify their patients according to
their risk of the experiencing the
outcome the therapy has been
demonstrated to prevent .22
■ In the past, researchers have often
not published the results of RCTs in
which no benefit from the
intervention was demonstrated,
making any meta-analysis of trials in
this clinical area less accurate.
However, it is now becoming
obligatory to register trials.
■ Even when high-quality evidence is
available to guide management in a
particular clinical area, an evaluation
of the cost and feasibility of
consistently using this evidence in
providing care to patients has often
not been undertaken. In these
circumstances, several important
questions should then be considered.
Are the resources available to apply
the evidence? Do staff members have
the appropriate skills and equipment
to use the evidence? Will patients
accept the treatment?23
With the large, ever-increasing
amount of clinical research being
undertaken, it is no longer possible for
individual clinicians to read all
published research papers relevant to
their panicular area of clinical practice.
Therefore, clinicians need ready access
to summaries of relevant information
from such research to practise
evidence-based medicine. Such
summaries arc available from
databases of critical reviews of the
scientific evidence or treatment
protocols based on best practice
developed by national or international
bodies. These summaries and
 Enhancing patient
care

guidelines should be readily available

when and where clinicians need to
make decisions about patient
management in their health service.
Unlike the US, the UK and New
Zealand, Australia does not have a
national database of such information,
so Australian health services need to
use commercial sources. A further
challenge for health services aiming to
provide evidence-based care to their
patients is that with the ever-
increasing and changing evidence of
what constitutes best practice in a
particular clinical area, the health
service's clinical policies require
regular and systematic review and
updating, which demands substantial
time and resources. This updating
process can be assisted by maintaining
online policy and procedure manuals
that are readily available to all staff,
rather than hardcopy versions.

1.4.2 Clinical audit

Health services can monitor the degree
of compliance with evidence-bascd
medicine in clinical practice using
clinical audit. Clinical audit ' WV
systematically reviews the care
provided to patients against explicit "Doctor* need to
criteria of the structure, process and accept to every
outcome of care. If gaps in care are cHric* daemon they
detected, appropriate action is taken metfB came*
aneconomc consequence
aimed at closing the gaps and thereby
that they am using up
improving the quality of care. Ideally,
money tor one person that
the audit process will then continue to wt not be available tor
monitor the care to determine whether others ^Managers need to
the actions taken have been successful value quahty as wet as
in achieving the improvement. The financial risk
data obtained must be reliable, and management. They have
audits should therefore be undertaken to accept that doctors
know more about what
using high standards of design,
works for patients than
measurement, data collection and they do. And they have to
analysis. Clinical audits should have a accept that the doctors
multidisciplinary focus rather than just make the decisions Mat
focusing on medical aspects of carc. cost money.'
and Ik continuous rather than
Perkins R. et al
NZ Med J 2006 .
119 U2259
intermittent or project-based. Ideally, clinical audit should be comprehensive and a routine part of
clinical practice in a health service. 10 Clinical audit will be discussed in greater detail in Chapter 2.

1.4.3 Credentialling and privileging

For clinicians to do the right thing for their patients, they need to possess the required professional
qualifications, have undenaken adequate training, and have an appropriate level of relevant
experience. It is therefore essential for health services to ensure that the clinicians providing care in
each clinical area in the service meet these basic requirements that are critical to providing high-
quality, safe patient care.
Most health services have a credentialling and privileging process in which every doctor who
wishes to work in the health service completes a detailed application form, outlining their
professional qualifications and documenting where they have previously worked. They are also
required to provide details of the areas of clinical medicine in which they wish to practise in the
health service and the operations and procedures they wish to perform. Applicants may also be
required to document the numbers of each type of operation or procedure they have performed
during their careers to date and, most importantly, in the recent past.
References are sought by the health service about the clinician's suitability for the clinical
appointment for which they have applied. Advice is then sought from the relevant specialty college or
experts in the field about whether the clinician^ qualifications and experience are sufficient for
practice within the health service, and which operations or procedures they should be allowed to
undertake. The board of management, usually on the advice of its medical appointments
subcommittee, then decide whether to formally appoint the clinician to the health service, and if so,
approves their clinical areas of work, operations and procedures.
Before undertaking the credentialling and privileging process, health services should determine the
range and depth of clinical services they can safely provide given the clinical expertise, facilities and
resources available.24 Even though individual doctors may have appropriate qualifications and
experience to undertake particular clinical procedures, the service may not have the facilities or
support staff to ensure the procedures can be safely performed. The credentialling and privileging
process is a critical pan of an effective clinical governance program. The process in Victoria is
described in detail and the application forms used by Victorian health services are provided in a
policy handbook.25

1.4.4 Professional development

Doing the right thing for patients also requires that clinicians maintain their knowledge and skills in
the clinical areas in which they are practising. Medical knowledge is an essential component of being
able to provide high-quality care. 26 Having given practitioners clinical privileges to practise in
particular clinical areas and to perform specific operations and procedures, the service must then
ensure that they maintain their knowledge in these clinical areas and their skills in performing the
specific operations and procedures. By participating in appropriate continuing medical education, a
clinician's knowledge can be maintained and expanded to incorporate new scientific evidence.
Ongoing education is now mandated by many professional bodies as part of their members'
recenification process. Maintaining up-to-date knowledge in their clinical field is now regarded as
clinicians' professional duty. As well as helping clinicians practise evidence-based medicine, health
services can help them maintain and expand their knowledge by providing them with easy access to
high-quality scientific databases of clinical trials and guidelines to support clinical decision making,
and by encouraging clinicians to keep up-to-date in their clinical area of expertise. Opportunities
should also be -made available to clinicians to receive education in how errors occur while providing
care, and the ways quality and safety can be optimised for patients. 24
 1 An overview at cfcrnca)
gowmgw

The form and content of continuing medical education need to be carefully chosen. Not all tonus
of medical education are effective in changing clinical practice. Passive forms of education, such as
didartir lectures, conferences and dissemination of educational materials, have been shown to be
ineffective11 Education based on individuals' personal learning needs and with active participation,
educational outreach and small group learning can be effective. As continuing medical education
can be costly and occupy considerable clinician time, the potential impact on changing clinical
practice of the methods used should be evaluated. 23
Skills in performing specific operations and procedures can be maintained, firsdy by clinicians
performing appropriate numbers of these procedures over time, and secondly by the health service
auditing the clinician's performance in undertaking these operations and procedures and
benchmarking their patients' outcomes and complication rates against the performance of their peers
and rates published in the scientific literature. Information about individual clinicians' performance
can be provided to the credentialling and privileging process at the time of the clinician's
reappointment, or earlier if clinical audits show there may be serious quality or safety issues.
In our experience, such serious issues involving individual doctors arise infrequently, but they do
occur. As with the general population, doctors may be impaired because of substance misuse or
mental or physical illness; they may also have behavioural problems. Leape estimates that "30-40%
of doctors at one time or another have a problem that poses a threat to safe patient care", and at any
one time in a hospital with 100 doctors, one or two will need help. 27 Leape advocates objectively and
routinely monitoring all doctors' performance against accepted standards (similar to the routine
assessment of airline pilots) to identify those with declining performance early. Constructive, prompt
action can then be taken, including providing these doctors with feedback about their performance
with a view to helping them return to competent and safe practice, thereby preventing patient harm. 27
Health services that implement a clinical governance program should also develop a policy
describing how they will monitor the performance of their doctors and deal with poor individual
clinical performance. Episodes of poor clinical performance need to be carefully and sensitively
managed. Some operation and procedural skills can be maintained with the use of skills training
courses, including those provided in simulator laboratories. Continuing medical education and skills
training are part of clinicians' professional development, which is an imponant pan of any clinical
governance program.

1.4.5 Consumer involvement

To do the right thing for their patients, clinicians must know what their patients' wishes are
regarding their care. Patients need to be involved in discussions about their treatment options and in
making imponant decisions about their care. Patients as a group should also have aggregated input
into how patients are cared for generally in their health service. Without knowing the requirements of
patients — individually and as a group — it is less likely that the care provided will meet patients'
needs. To date, patient involvement in health care processes has been low. Assessing their
involvement can be difficult, as patients' understanding of involvement can vary widely, from feeling
they are being made welcome, to be able to talk about their concerns, to being actively able to assess
the advantages and disadvantages of their treatment options. In the UK, patients report good
communication with doctors and continuity of care, but rate choice, involvement and information as
poor.28
Importantly, expectations about the degree of their involvement in decision making about their
care will vary considerably between patients. Further research is required to determine how and
when patient involvement should occur, and the feasibility and cost of their involvement. 21 Patients
can also be part of the safely and quality improvement process by making complaints lo health
scrviccs and by health services effectively dealing with the concerns raised, and actively by services
 through seeking their feedback. Other important areas of patient involvement arc open disclosure by
Enhancing
clinicians patient careservices
and health after adverse events have occurred and obtaining informed patient
consent before procedures are undertaken. 9

1.4.6 Organisational faction

Many organisational factors are significant components of an effective clinical governance program.
They include:5-9 29 30
■ Strong leadership provided by the governing body and executives to set the agenda.
■ Quality of care and patient safety are made a high priority by the governing body and executive
staff. Clinical governance ranks equally with corporate governance.
■ High vigilance to ensure the health service is effectively providing high-quality, safe care.
■ Systems are in place to recognise gaps or risks, and appropriate action is taken to coned these.
■ Organisational structures and processes are created to undertake the program.
■ Adequate resources are provided to conduct the program.
■ The health service's clinicians show strong, effective leadership and commitment.
■ Full engagement of clinicians and their acceptance of a central role in the program.
■ Comprehensive education for all staff about the clinical governance process.
■ All staff have clearly defined roles and responsibilities in the program.
■ Extensive use of multidisciplinary teams.
■ Effective monitoring, data collection and reporting systems are in place to aid accurate,
comprehensive assessment of clinical performance.
■ Ready access to clinical evidence via specialist clinical databases and high-quality data to support
effective clinical decision making.
■ A culture that supports quality of care and patient safety by being open and participative, valuing
education, and using blame rarely.
■ A strong, healthy working relationship between senior management and health professionals in
the organisation. This relationship should be built around "responsible autonomy" — the balance
between clinicians' professional autonomy and accountability.
■ Participation in accreditation processes wherever the organisation's performance is measured
against predefined standards by external bodies.
If these conditions are met, they should "empower highly motivated, highly skilled frontline staff to
deliver a high quality service". 31 Importantly, although the concept of clinical governance makes
intuitive sense and much work has been done to create appropriate structures for it to be
undertaken, it has not been critically evaluated, and there have been few reports of improved patient
outcomes following its implementation.32
There is significant overlap and interdependence between the components of an effective clinical
governance program. One component may use another to help solve a problem. For example, the
risk management component may detect an adverse event that a patient has experienced. After
analysis of the event, it may be decided that a clinical audit is required to help develop a strategy to
reduce the probability of the adverse event recurring. An audit may also allow an estimate of the
frequency of the adverse event. The standards used in the audit against which clinical performance
will be measured should reflect best practice, as determined by the available scientific evidence. If
the audit shows a significant gap between the care being provided and best practice, continuing
medical education as part of the clinician professional development program may be used to try to
close the gap Alternatively, if the clinical audit shows that the type of adverse event found by risk
management has only occurred in patients under the care of one clinician and the event is thought
to be preventable, it may be appropriate to report the results of the audit to the service's
credentialling and privileging committee.

10
1 An overview ot rtimcaJ Qoverrwnoa

In this way, the various clinical

governance activities will be
integrated, allowing the program to be
maximally effective in improving the
quality and safety of patient care.
Previously, integration of clinical
governance activities did not occur,
and opportunities to improve the
quality and safety of patient care were
not fully utilised. Such integration
requires good communication between
the personnel involved in each of the
program^ activities and appropriate
reporting of the activities and results
achieved to relevant staff and to the
peak quality committee within the
health service. Reporting clinical
governance activities and the results
achieved will be discussed funher in
Chapter 8.

1.5 Reporting
the results of
the clinical
governance
program
The results of the quality improvement
and risk management components of
the program should be regularly
reponed in two directions within the
health service — up through the
committee and management structure
of the organisation to the board of
management^ and down through the
organisation to all clinical staff Clinical
governance activities and the
improvements they have achieved in
the quality of patient care and level of
patient safety are also increasingly
being reponed publicly as pan of
health service annual reporting
requirements.

1.6 Accountability for

the clinical
governance program
in health services
Clinical governance in health services
should be rigorous, organisation-wide
and accountable.5 Responsibility and
final accountability for the way in
which the clinical governance
framework is developed and
implemented and its effectiveness,
together with the quality and safety of
the clinical services delivered within
the health service, rests with the board
of management and chief executive. In
Victoria, the responsibilities of the
board of management of each health
service regarding quality and risk
management are defined in legislation
in the Health Services Act 1988 (Vic)."
It has been suggested that the
accountability for and reporting of
clinical governance activities within a
health service should rank equally in
importance with accountability for the
organisation's finances. 5 Although the The reason for
tweeting in cUnictt
board of management and executive
governance is lo improve
stafT have significant responsibility for
quality of cere AM else is
quality and safety in clinical areas, subservient to this.
other stafT in the health service, However this principle in
particularly clinicians, also have itself causes a dash of
professional "daily responsibility" 5 for ideologies between those
the effective implementation of and who believe that * each
compliance with the components of the patient deserves the best
care (typically doctors
organisation's clinical governance
with a sense of 'duty'
program. Such responsibility requires towards their patients)
clinicians lo accept transparent and those who have to
accountability, to work in teams in distribute funding and
providing care, and to share relate to a population
(characteristically board
directors and managers
within an ethic of
'utHHariarrism' —the
greatest good for the
greatest number).'

Pertuns R. et
al A/Z Med J
2006: 119
U22S9

12
power in clinical areas. 32 However, clinicians should also be given a high level of autonomy in areas of
health service delivery in which they have a high level of expertise. 9 Responsibility and accountability
for quality and safety should be clearly defined at all levels of the health service from "board to
bedside".30

1.7 A health service's role with the major components of a clinical

governance program
Although all the activities of a clinical governance program discussed in the previous section are
important, this book will deal predominantly with the main components of quality improvement and
risk management. The other activities will be discussed as they apply to both these areas. However,
even the above brief discussion of each of the activities of a clinical governance program highlights
many of the critical processes that are required be in place for such a program to be fully effective.
Each health service must have a comprehensive, effective credentialling and privileging process in
place for the appointment of its clinicians as one of the cornerstones of effective clinical governance.
Each doctor wishing to work in the service must complete a detailed application form and proceed
successfully through the appointment process before they commence work. Just as important as the
initial assessment of the application and the candidate, the credentialling and privileging committee
has an ongoing role with the clinicians it appoints to seek regular information, from clinical audits
and the risk management program, regarding the performance of individual clinicians.
If these activities detect areas of concern regarding the care provided by individual clinicians, these
must be fully investigated. Initially, the clinician involved should be given the opportunity to respond
to the fmdings. However, in some instances, action may need to be taken, including curtailing or
withdrawing the clinician's clinical privileges, requiring that specific operations or procedures are no
longer be performed by that clinician, or that they only be performed under the direct supervision of
a nominated clinician, or that the clinician be required to undertake specific medical education or
skills training before providing further care to patients. Clearly, if such actions are determined to be
required, they should be promptly taken. Although significant performance issues involving
individual clinicians are relatively rare, they do occur, and the effectiveness of any clinical governance
program will potentially be undermined if examples of individual poor clinical performance are not
dealt with promptly and effectively. It is paramount that the health service deals with individual
performance issues professionally, sensitively and in a timely manner.
Senior management and the health service can demonstrate their genuine commitment to clinical
governance by fully supporting the professional development of their clinicians. Appropriate
opportunities should be provided to clinicians to maintain their professional knowledge and skills.
Such up-to- date knowledge and skills are important to practising evidence-based medicine and, as
will be discussed later, can help effectively change clinical practice.
Ready access to the current knowledge base of evidence-based medicine is central to providing
high- quality, safe care. As previously mentioned, to support the clinical governance program, health
services need to provide clinicians with ready access to up-to-date clinical databases of critical
reviews of the scientific evidence. Computers with links to such databases should be readily available
in the health service's wards and various clinical departments.
As will be discussed further in the chapters on risk management, the conditions under which
clinical work is undertaken in a health service have a significant impact on the level of patient safety.
Work conditions should be made as safe as possible within the limitations of the resources available.
Clinical
 1 An overview of cfcmcaf
jovemenoe

risk managemem continuously balances the level of patient safety that it is possible to achieve and
com of achieving this level of safety. The level of safety may be increased by reducing workloads and
shift lengths, providing adequate supervision and appropriate equipment, and by co ntinuously
monitoring patient safety using a comprehensive risk management program.
Clinical audit is a core part of quality improvement and, as discussed earlier, is essential for the
effective ongoing credentialling and privileging of medical staff. Senior management in health services
should therefore actively support clinical audit. This support includes: the availability of appropriate
data collection systems; using information technology where possible; and adequate protected time
for clinicians to design, undertake and analyse clinical audits-, and, most importantly, take
appropriate action in response to the results of the audits. The methodology and place of clinical
audit in quality improvement and patient safety will be discussed further in the chapters on detecting
gaps in patient care and implementing change.
In the past, consumers' views about the quality of the care they have received in health services
have been inconsistently sought. In an effective clinical governance program, the health care provided
is patient-centred, so consumers' views are regularly and appropriately sought. Consumers' views
can be obtained from surveys, focus groups and consumer advisory panels. Health services should
aeate systems to ensure consumers' views about the care they have received are obtained and acted
upon.
The culture of the health service is an important factor in achieving effective clinical governance.
Support of staff, communication throughout the organisation and how negative events are handled in
the service are all important issues and will be discussed in detail in Chapter 9.
*»
■flnMipMnl
■ Regulation is a core role of management that ensures an organisation's activities will achieve its
goals.
■ Governance is about regulation and the framework used to achieve it.
■ Previously, governance in health services has concentrated on corporate governance (ie,
financial and business outcomes).
■ Clinical governance is the regulation of quality and safety in clinical care.
■ Clinical governance has two main components — quality improvement (doing the right thing to
patients more frequently) and risk management (doing the wrong thing to patients less
frequently).
■ The core principles and activities of clinical governance are:
>■ Evidence-based medicine — providing clinical care for patients in accordance with best
practice determined by the available scientific evidence.
> Credentialling and privileging — determining that clinicians working in the health service
have the required qualifications, training and experience to provide care in a designated
clinical area and they are providing this care to a high standard.
> Clinical audit — measuring individual clinicians' performance and the health service's
aggregate performance, and comparing them with best practice.
> Professional development — providing clinicians with appropriate continuing medical
education and skills training.
> Consumer involvement — determining the needs of individual patients and patients as a
group, and meeting these needs.
> Organisational factors — that appropriate resources, strong leadership, adequate support
and education and clear, timely and direct communication are provided by the health
service to support the above activities.
 Enhancing patient care

■ Final accountability for clinical governance rests with the health service's board of
management and chief executive, but clinicians have daily responsibility for complying with
clinical governance requirements.
■ A key factor in establishing and maintaining an effective clinical governance program is the
support of the health service and senior management for the program. Such support can be
provided by:
> Allocating protected time for medical staff to participate in activities that support effective
clinical governance (eg, clinical audits, continuing medical education and skills updates).
> Ensuring the conditions of work under which clinical care is provided are conducive to
high-quality, safe care.
> Supporting the clinical audit process (eg, by providing appropriate stalling, information
technology and data collection).
> Dealing with poor clinical performance by individual clinicians sensitively, lairly and in a
timely manner
> Providing ready access to current clinical databases of high-quality clinical research and
guidelines in all clinical areas to support clinical decision making.
1 An overview of
clinical go*emanoe

References
1 Lapides PO, Hermanson DR, Beasley JV, et al. Corporate Governance Center. Kennesaw State
Unweraity, 2002
2 ASX Corporate Governance Council. Corporate governance principles and recommendations.
2nd ed. 9f&ey Australian Securities Exchange. 2007.
www.asx.com.au/about/governance/index.htm (accessed Oct 2006)
3 Encycogov.com. What is corporate governance? 2006.
www.encycogov.com/WhatlsCorpGov.asp (accessed Oct 2008).
4 Deighan M, Bullivant J. Re-energising clinical governance through integrated governance. Clin
Chem Lab Med 2006; 44: 692-693.
5 Scally G, Donaldson LJ. The NHS's 50 anniversary. Clinical governance and the drive for
quality mprcvomor* in the new NHS in England. BMJ1998: 317: 61-65.
6 Australian Council on Healthcare Standards. ACHS News, Issue 12. Sydney. ACHS, 20M.
7 World Health Organization. The principles of quality assurance. Copenhagen: WHO. 1983
(Report on a WHO meeting).
8 Gray C. What is clinical governance? BMJ Career Focus 2005; 330: 254-261.
9 Braithwaite J, Travaglia JF. An overview of clinical governance policies, practices and
initiatives. Aust Health Rev 2008; 32: 10-22.
10 Spark Jl, RoweS. Clinical governance: its effect on surgery and the surgeon. ANZ J Surg 2004; 74:
167-170.
11 Halligan A, Donaldson L. Implementing clinical governance: turning vision into reality. 8M/2001;
322:1413-1417.
12 Degeling P. Realising the developmental potential of clinical governance. Clin Chem Lab Med 2006;
44:688-G91.
13 Niselle P. Managing medical indemnity: must we choose between quality assurance and risk
management [letter]? MedJAust2DOA. 181: 520.
14 Lohr KN, Schroeder SA. A strategy for quality assurance in Medicare. N Engl J Med 1990; 322:
707-712.
 15 Agency for Health Care Research and Quality. AHRQ's Patient Safety Initiative: building
foundations, reducing risk. Interim report to the Senate Committee on Appropriations.
Rockville. Md: AHRQ Publications. 2003.
16 Kohn LT. Corrigan JM. Donaldson MS. To error is human: building a safer health system.
Washington. DC: National Academy Press. 1999.
17 Institute of Medicine. Crossing the quality chasm: a new health system for the twenty-first
century. Washington. DC: National Academies Press. 2001.
18 Berenholz SM. Pronovost PJ. Monitoring patient safety. Crit Care Clin 2007:23:659-673.
19 Cook DJ. Montori VM. McMullin JP. et al. Improving patients' safety locally: changing
clinician behaviour. Lancet 2004; 363: 1224-1230.
20 Sackett DL. William MC. Rosenberg WMC. et al. Evidence based medicine: what it is and
what it isnl. BMJ 1996: 312: 71-72.
21 Chassin MR. Quality of health care. Part 3: improving the quality of care. N Engl J Med
1996:335:1060-1063.
22 Kent DM. Hayward RA. Limitations of applying summary results of clinical trials to
individual patients: the need tor risk stratilication. JAMA 2007, 298: 1209-1212.
23 Grol R. Improving the quality of medical care. Building bridges among professional pride,
payer profit, and patient satisfaction. JAMA 2001: 284: 2578-2585.
24 Scott IA, Poole PJ. Jayathissa S. Improving quality and safety of hospital care: a reappraisal
and an agenda for clinically relevant relorm. Intern Med J 2007; 38: 44-55.
25 Victorian Government Department ol Human Services. Credentialling and defining the scope
of clinical practice for medical practitioners in Victorian health services - a policy handbook.
Melbourne: Victorian Government Department of Human Services, 2007.
26 Holmboe ES. Lipner R. Greiner A. Assessing quality of care: knowledge matters. JAMA 2008; 299:
338-340
27 Leape LL. When good doctors go bad. Ann Surg 2006; 244: 649-652.
28 Elwyn G. Buetlow S. Hibbard J, Wensing M. Respecting the subjective: quality measurement
from the patients perspective. BMJ 2007; 335: 1021-1022.
29 Degeling PJ, Maxwell S, ledema R. Hunter DJ. Making clinical governance work. 6MJ2004. 329:
679-681
30 Balding C. Clinical governance: another name for quality - or improving on improvement?
HIM J 2004. 33 137- 139.
31 HaKgan A. Right here, right now. NHS Magazine 2001; Jul 6.
32 Peftons R. Pefcowitz A, Seddon M. Quality improvement in New Zealand healthcare. Part 7.
clinical governance — an attempt to bring quality into reality. N Z Med J 2006; 119: U2259.
33 Victorian Health Services Act 1988. Division 9 B - Public Health Services, Section 65 S - Board of
Directors, (iv) - (vii). □
 Enhancing patient care

2 Quality improvement — measuring clinical

performance and closing gaps in clinical care
This chapter will:
■ Define quality improvement
■ Discuss what best practice is and how to locate it
■ Discuss how to determine the clinical areas in which quality improvement should be undertaken
■ Outline how to conduct a clinical audit — how to define criteria that represent best
practice: data entry, collection and analysis: benchmarking; how to report and discuss
results: how to take action: the audit cycle: and the resources required
■ Review whole-of-practice audits
■ Define clinical indicators, and discuss their role and the many different indicators available
■ Present a comprehensive audit checklist.

2.1 Defining quality improvement

Quality improvement is one of the two major components of clinical governance. The second
component, risk management, is discussed in Chapter 3. There are two main approaches to
quality improvement used in health care. One approach, clinical audit, begins by defining a
desired standard of performance or care to be provided to patients. The
care provided is then measured, and if it meets the desired standard,
no action is usually required. However, if there is a significant
Figure 2.1 Bask quality
difference between the desired and actual standards of care, the
improvement cyd*
difference is analysed and action taken, aimed at closing the gap. The
care provided continues to be monitored to determine if the action has
been successful in closing the gap (Figure 2.1). In practice, several
cycles of measurement and action are often required to close the gap.
An alternative approach to quality improvement is system redesign. A
system of health care delivery that clearly needs improvement is chosen,
and the processes that comprise the system are identified. Each process
is then examined to see whether it can be modified to improve the overall
performance of the system and reduce waste. The two approaches
overlap, as system redesign is commonly used to close gaps detected in
performance.
Unfortunately, there are many significant gaps in the care provided to
patients compared with best practice. Only 50% of patients receive
recommended preventive care, and 70% receive recommended acute
care.1 These gaps are so extreme that "the greatest opportunities to
improve patient outcomes will probably nol come from discovering new
treatments bui from effective deliver)' of existing therapies". 2 A systematic
strategy is required to routinely monitor and report the quality of carc
being provided.1
 2.2 Deciding the clinical areas in which
quality improvement activities should be
undertaken
Despite extensive summarisation of research findings and best-practice
recommendations into formats of manageable size for clinicians to use,
the number of opportunities available to develop standards of care in
different clinical areas is far greater than can be fully utilised by
individual health services. Each service will therefore need to determine
the clinical areas in which they wish to develop standards of care and
measure their performance.
Several approaches are available. Often, a useful place to begin is with
the health serviced most frequently coded principal diagnosis-related
groups by admission. Best practice could be defined for several of these
high-volume diagnoses, and clinical performance could then be
measured. Alternatively, some organisations, such as the National
Institute of Clinical Studies, have published lists of major gaps in
provision of clinical care across the full continuum of health. 3,4 Some gaps
are in primary care (eg, inappropriately prescribing antibiotics for the
common cold and acute bronchitis; advising on smoking cessation) and
others in acute hospital care (eg, preventing venous thromboembolism,
managing acute and cancer pain). The best available evidence is
presented for each gap, together with current clinical practice.
Further, medical journals such as Evidence-Based Medicine
(www.ebm.bmj.com) and Evidence-Based Nursing (www.ebn.bmj.com)
alert clinicians about important advances by summarising and reviewing
the best evidence-based research from the large number of articles
published in medical journals; these summaries are published every 2
months. The contents page of each issue of these journals could be
scanned by individual health services to identify areas where there may
be gaps in the care they provide, thereby providing direction for quality
improvement activities. Based on research about medical errors, lists of
clinical interventions whose effectiveness is supported by high levels of
scientific evidence have also been published, 5-6 and could also provide
direction for local quality improvement programs.
There are many clinical areas in which quality improvement initiatives
can be chosen for implementation, making selection of initiatives
potentially difficult. One selection criterion could be that the improvement
initiative should have the maximum potential positive impact on patient
mortality and morbidity. The Pareto principle, where the 20% of initiatives
that could potentially achieve 80% of the total possible positive impact on
patient care, could be chosen for implementation.
Alternatively, interventions could be chosen according to the resources
that their introduction would save the health service. Such interventions
' there isms systematic could be implemented on the understanding that the savings made from
evidence about qualify
of care... More information
is available on the quality
of airlines, restaurants,
cars and VCRs than on
the quality of health

Schuster MA.
et al MilbankO
2005; 83:843-
895
their introduction would be used to make further 2 Quality mpfowemen
improvements in the
quality of care provided, perhaps by implementing interventions that
 significantly reduce patient mortality and morbidity,'but either require additional mouiw to
implement, or where implementation will not result in savings being made. Another important
selection criterion is that the clinicians in the health service feel the intervention chosen is
clinically tmpottara This criterion has significant implications for successfully achieving change in
clinical practice, and wil be discussed further in Chapter 6.
In essence, the clinical areas chosen to undertake quality improvement initiatives depend on two
major factors:
■ The importance of the clinical problem in which intervention is being considered. Is the clinical
area one of high volume, significant mortality and morbidity, substantial cost, unexplained
variation in care, or suboptimal care being provided to patients?
■ The opportunity the clinical area provides for effective clinical intervention Gaps in care provided
should have been identified, and evidence-based guidelines that would improve patient outcomes if
implemented should be available. 2,7,8

2.3 Components of a quality improvement program

2.3.1 Finding best practice
The first task in quality improvement in health care is to define the desired standard of care to be
provided. Quality is a multidimensional-construct, meaning it has multiple attributes, each of which
may have a different measure. 9 No single measure can comprehensively reflect all the attributes. The
dimensions of care include clinical factors, cost, patient satisfaction and functional health status. A
"balanced scorecard" or "clinical value compass" can be developed that includes measures of each
dimension.10 Health care should also be safe, effective, patient-centred, timely, efficient and
equitable." Patients may place greater weight on some dimensions of quality (such as waiting times)
than clinicians.12 Measures of best practice for each of these dimensions should be sought. Such
measures will still not provide a complete assessment of quality, but an increasingly more accurate
one can be drawn from multiple perspectives.9
The desired standard may be determined by the scientific evidence that is available, benchmarking
against similar organisations, or using standards set by external organisations, such as accreditation
or funding bodies. The desired standard of clinical care is usually best practice according to the
highest- quality scientific evidence available at the time. The major sources of information about
what constitutes best practice in clinical care are systematic reviews or meta-analyses of clinical
research publications in medical journals. There are many medical journals, and the number of
individual articles reporting the results of clinical research is huge, with over 30 000 new
publications added to the MEDLINE database each month. The volume of clinical research that is
continuously published is far greater than could be regularly reviewed by clinicians in an individual
health service.
To provide high-quality, safe care, clinicians will be increasingly reliant on "clinical pathways,
clinical practice guidelines and decision support tools". 13 Fortunately, a number of organisations
systematically review the scientific literature and produce critical reviews and summaries of the
research published over a wide range of clinical areas. Clinical practice guidelines, or "systematically
developed statement(s) to assist decisions for practitioner and patient about appropriate health care
for specific clinical circumstances", are also developed. 14 These guidelines are developed by many
national and international groups, and describe best practice for treating patients with many clinical
conditions. Sources of information about best practice include:
Enhancing patient care

■ Systematic reviews undenaken by the Cochrane Collaboration (www.cochrane.org) and in Clinical

Evidence, which is published by the BMJ biannually in prim and updated monthly online
(clinicalevidence.bmj.com/ceweb/index.jsp).
■ Individual clinical reviews, meta-analyses or the results of randomised controlled trials published
in medical journals.
Clinical guidelines and the level of scientific evidence supporting each component are available for a
wide range of clinical conditions, from sources including:
■ The National Guideline Clearinghouse (with almost 2000 guidelines currently listed), which is an
initiative of the Agency for Healthcare Research and Quality (AHRQ) and the Department of Health
and Human Services in the United States (www.guideline.gov).
■ The National Health Service's National Institute for Health and Clinical Excellence in the United
Kingdom (www.nice.org.uk).
■ The Scottish Intercollegiate Guidelines Network (SIGN) (www.sign.ac.uk).
If high-quality scientific evidence for best practice in managing patients with a particular medical
condition is not available, expert consensus may be used. 15 SIGN has formalised expert consensus
using a three-stage process of considered judgment. The level of evidence and the impact of that
evidence (including potential harm and practicality of implementation) are judged, and key
recommendations are identified.
One problem with systematic reviews and guidelines is that best practice in a particular clinical
area will change over time because of the large amount of clinical research being continuously
undertaken. An analysis of 100 systematic reviews found that in 23%, signals for updating occurred
within 2 years.16 Therefore, reviews and guidelines should be regularly updated. Finally, guidelines
are documents that indicate what should be done in patient care in ideal circumstances. In
individual health services, these guidelines should be used to develop protocols that reflect local
needs and conditions and detail what should be done, by whom, when and how.

2.3.2 Defining criteria that represent best practice

As discussed previously, there are several quite different dimensions of patient care, and the quality
of each is often assessed using different methods. For example, the level of patient satisfaction may
be assessed using patient surveys, interviews or focus groups. By contrast, the compliance of an
individual health service with the standards set by accreditation bodies is usually assessed by health
services completing a presurvey questionnaire before surveyors observe how the health service
functions during a site visit conducted over several days. Traditionally, quality of care has been
divided into three components:
■ structure (how care is organised)
■ process (what is done to patients)
■ outcome (the results of what is done to patients). 17
It has been suggested that culture, or the "collective attitudes and beliefs of caregivers", should be
added as a fourth element of quality.18 New structural measures have also been suggested, including:
institutional variables (eg, credentialling of clinical staff and the degree of senior management
involvement in safety and quality programs); task variables (eg, the presence of protocols); or team
variables (eg, the degree to which frontline staff feel comfortable alerting management about areas of
concerns).19
Several structural components of health services (such as staffing levels, equipment, human
resources processes, and methods of reimbursement) have been shown to be associated with clinical

20
Enhancing patient care

outcome, although caution should be used in making inferences about cause and effect. Often,
structural changes require substantial investment, beyond the resources available to individual
health services.20 Therefore,

21
 2 Quality
improvwnani

we will confine our discussion to the structural elements that are within
the control of local senior management, and particularly the assessment
of the clinical component of patient care and how clinical audit is
undertaken.
Once best practice in the management of patients with ? particular
clinical condition has been determined, clinical audit can be used to
measure and compare the performance of the health service. Firsdy, the
criteria that reflect best practice according to the scientific evidence need
to be determined. A criterion can be defined as "a discrete, clearly
definable and measurable phenomenon ... Relevant to definition of quality
... [and] so clearly defined that we can say whether it is present or not". 21
The choice of criteria is important, as the frequency with which criteria
are attained determines the level of performance by clinicians and the
standard of care provided to patients.22
Criteria should be developed by a multidisciplinary group of medical,
nursing and allied health clinicians, representing the health professionals
who deliver care to patients in the clinical area under review. Such
membership should increase the probability that the criteria chosen accu-
rately reflect the quality of care being provided, and facilitate local
ownership of the measures and acceptance of the process. The criteria
should reflect best practice in medicine, nursing and allied health, and
the important aspects of effectiveness, safety and efficiency as well as
professional, organisational (such as wailing times and coordination and
continuity of care) and patient- oriented dimensions. With all these
considerations, a large number of criteria may be developed, and it may
be necessary to rank them according to their importance and
measurability, and reduce them to a workable number. 8
Information about best practice can be obtained from a review of the
scientific literature, clinical practice guidelines or expert consensus. 23
Best-practice criteria can then be developed from these sources that
reflect the highest level of scientific evidence available, and that level can
be stated for each criterion.7
Although guidelines are usually comprehensive and of high quality,
some are lengthy and require transformation into criteria against which
performance can be measured. For example, a literature review of
mechanical ventilation produced a 100-page evidence-based summary,
from which five actions were identified that could be taken daily and
measured for patients receiving ventilation9
Alternatively, criteria that have already been developed in particular
clinical areas can be obtained from the literature, databases and sources
of international indicators, such as the RAND Corporation
(www.rand.org), the AHRQ (www.ahrq.gov/) and the National Primary
Care Research and Development Centre (www.npcrdc.ac.uk). Indicators
chosen from these resources require adjustment for local needs and
conditions.8
Whether criteria are developed locally or taken directly from external
sources, they should be clinically important, clearly definable, objective,
•QuaWyl is...
apmrnaMngtHQ,.
time-consvning business
that depends for its
success st leest as much
on our s&tty to modify the
behavior of patients,
purchasers, and
providers of care as it
does on the Collection of
good data about
performance '
Blimenthal 0
N Engl J Med
1996.
335 1328-1331
 Enhancing patient care

specific, measurable and theoretically sound. 23 Precise definitions of the

numerator and denominator of each criterion should be available from
external sources and modified appropriately, or developed locally. 8
Without such definitions, the validity and reliability of the data that are
collected will be compromised.9
 Enhancing patient care

If appropriate, patient inclusion and exclusion criteria should also be determined. 7 Ideally, few
patients will need to be excluded from the audit, allowing a comprehensive assessment of care to be
made. The criteria should be valid (ie, measure what it is intended to measure) and reliable (ie,
repeated measures obtain similar results).7 The criteria should also be clinically important and have
been demonstrated to improve patient outcome when provided in their care. 7
As well defining the criteria to be used in assessing performance, standards for acceptable levels of
compliance with each criterion should also be developed. For example, a criterion could be that all
patients with acute coronary syndrome should receive aspirin. It is unlikely that a health service with
a large number of such patients will achieve 100% compliance with this criterion, and some patients
may not be able to safely take aspirin. In acknowledging this reality, an appropriate standard could
be that 85% of patients with acute coronary syndrome receive aspirin.

2.3.2.1 The advantages and disadvantages of process and outcome measures

Criteria represent standards of care, and can be process or outcome measures. 715 Process measures
examine what is being done to provide care (eg, if patients have undergone investigations, been
referred to specialists or received medication), and outcome measures examine the health status of
the patient after treatment (eg, functional health status, quality of life, morbidity or mortality). 715 Both
measures require specific definitions if the results are to be meaningful, especially for comparison.
Importantly, process measures vary with the behaviour of the health care provider. 22
Process measures are generally easier to interpret, and are more sensitive in detecting changes in
clinical practice in small samples of patients over short periods of lime, than outcome measures.
Outcomes often require considerable time to become evident, and much larger numbers of patients
are needed to detect significant differences between samples. Outcomes measures vary not only
because of changes in the quality of care that has been provided, but also because of definitions, data
quality, casemix, and chance. 20-24 Outcome measures may require adjustment for casemix, including
patient characteristics (eg, age, sex, functional status, comorbidities and illness severity), and such
adjustment may still be incomplete because of unknown prognostic factors. 20
Importantly, apart from a few surgical procedures, "outcome is neither a sensitive nor a specific
marker for quality of care". 20 However, some outcome measures can be used as criteria, especially if
the outcomes occur in a short period of time or if a high volume of patients are being audited. 15-25
Although process measures are more frequently used to assess quality of care, outcome measures
may demonstrate differences in the quality of care (such as a surgeon's operating skills) that are not
assessable by using a limited number of process measures. 26
If process criteria are used, they should be scientifically demonstrated (preferably by randomised
controlled trials) to be clearly linked to patient outcomes. In other words, the process measures
chosen should clearly affect outcomes, so that they will be a measure of ihe quality of care that has
been provided.7 15 In practice, a combination of process and outcomes measures are commonly used
to measure quality, costs and patient satisfaction. 10
Clinicians prefer process measures, to which they can influence adherence, over outcome
measures, which can be influenced by factors beyond their control. Process measures can be easily
incorporated in routine data collection, and can act as a reminder to clinicians about processes of
care they should be providing. However, these measures require updating as research influences best
practice. Patients and funders often prefer outcome measures, which do noi usually require revision.
However, patient outcomes may be infrequent and delayed, thereby hindering timely feedback lo
clinicians and requiring more resources for long-term follow-up of patients and risk adjustment of
data. If possible, a balanced set of process and outcome measures should be used. 2 Alternatively, one
or two major outcome measures (eg, mortality and morbidity) and several related process measures
could be chosen "
 2
Qua
My
impr
ove
ment

Using process measures has a significant advantage — if the measurement is below the target
levd and timely feedback is provided, management and clinicians can quickly identify the problem
and determine appropriate action to address it. If an outcome measure (such as mortality) is poor,
clinicians may not know where to direct their attention to improve it. Was the result due to poor
quality of care or other factors, such as poor data quality, inadequate adjustmenL for casemix,
structural factors, chance, or unknown factors? If outcome measures are used to penalise or
stigmatise individuals or health services, then dysfunctional behaviour, such as gaming of data,
patients or care, may occur to diston the measure.
The criteria chosen may depend on their clinical importance to patient outcome, the strength of
evidence supporting the effects of the criteria on patient outcome, their validity and reliability, and
how easy it is to collect data on the criteria.7,8 The criteria should also be diverse, and should
comprehensively cover the treatment of the clinical condition. As a final check of the criteria 's
validity, the clinicians who will be receiving data on their clinical performance need to be asked
whethlt they feel the criteria are measurable, accurately evaluate imponant aspects of quality of care,
are meaningful to them in trying to improve the care they provide, and are measures they have the
control to alter.2-918
One potential problem with conducting clinical audits is choosing too many criteria against which
to measure clinical performance. A balance is required, as large number of criteria can make data
extraction and collection more difficult, whereas a small number of criteria may limit the components
of patient care that can be assessed. A good balance can be achieved using about 12 criteria to
measure a care process.8
At Wimmera Base Hospital, a medium-sized hospital, we prefer to measure a small number of key
criteria, which makes data collection simpler and allows the results of the audit to be concisely
reported. For example, interventions that have been shown to decrease mortality in patients with
acute myocardial infarction include administration of aspirin, thrombolytic therapy, ^-biockers and
ACE (angiotensin converting enzyme) inhibitors. Therefore, appropriate criteria would be
administration of aspirin in the emergency department, administration of ^-blockers in the first 24
hours of the patient!; admission, the lime between the patient arriving at the emergency department
and commencement of thrombolytic therapy and administration of ACE inhibitors on discharge.
Outcomes measures, such as 30-day mortality, may not be useful in a hospital of this size because
significant trends may take some time to become apparent.
Whichever criteria are chosen, they will require updating over time as the results of further
research in the clinical area becomes available. 26

2.3.3 Data entry

The time clinicians must spend entering clinical audit data is a significant factor in gaining their
suppon for the activity. Participating in clinical audit should not be a significant time burden for
senior clinicians, and data entry should be made as efficient as possible. Designing the audit to allow
data to be entered by non-clinical personnel or junior medical staff can save considerable time for
senior clinicians. Data entered by clinicians can also be significantly reduced if patient identification
data from the hospital patient information technology system can be automatically entered into the
database created for the audit, rather than being re-entered by clinical staff. Data entry can be
assisted by clinical data being rapidly entered by clinicians using personal digital assistants when
they are close to the patient (eg, in the operating theatre or at the bedside). The data from personal
digital assistants can then be uploaded to the database when it is convenient. 27

23
 Enhancing patient care

Where possible, prospective audit is preferred to retrospective audit, as the data collected is more
likely to be complete.28 The accuracy of the data is critical. A sample of the data should be checked
for completeness and accuracy.29 The method chosen for data entry and the data collection
spreadsheets or forms to be used, whether paper-based or electronic, should be reviewed for any
ambiguities by clinicians not associated with the audit before it commencing. 28 They should also be
tested on a small scale to determine whether the criteria chosen are clear and the data can be readily
collected,30 and to identify and reduce problems when the audit is commenced on a large scale.
Whichever method is used to enter or collect data, the audit should be embedded as firmly as
possible as pan of routine, everyday practice of the health service, and, wherever possible, included
in existing clinical and administrative patient care documentation. Such actions should reduce costs
and missing data, and support the sustainability of the audit. 7

2.3.4 Data collection

The audit should be as simple as possible. Wherever possible, electronic data collection is preferred,
but if paper-based audits are used, data should be easy to collect. Developing and using data
abstraction tools that are simple, shon and easy to use will assist data collection. Wherever possible,
data that are already being routinely collected in the health service should be used, 25 and obtained
using automated processes and existing data files. 818 The data should be readily available and
inexpensive to collect.7
Sometimes, it may be necessary to balance the scientific soundness of measures that are chosen
against the feasibility of their collection. 18 The "who, what, when, where and how" of the data
collection should be defined, as should the burden of data collection on staff. 2 Ideally, a database
should be created that optimises the quality of data to be analysed, aids identification of missing
data, and assists with the creation of reports. 31 Data that identify individual patients must be kept
confidential, and all data must be securely stored.

2.3.5 Measure baseline performance — small-scale testing

The measurability of processes of care can be difficult to predict. It is therefore important to test the
audit tools, data entry and collection systems on a small scale before conducting a comprehensive
audit 8 Testing may show that criteria definitions require more detail, the data sought are not
available, or the resources required to collect them are too great. 2 A final decision should be made
regarding the feasibility of each component of the audit: criteria measurement, data entry, data
collection, analysis, and timely, actionable feedback before proceeding to a large-scale audit. The best
test of the feasibility of the data collection tools and process is to measure the baseline performance
against which the effectiveness of the intervention will be assessed. 31

2.3.6 Analysis and review of audit results

Following data collection, the proportion of patients whose care meets each criterion representing
best practice is determined. The rate of compliance with best practice can be compared with rates in
the scientific literature. National or international rates may be available for some criteria. For
example, national and state averages and individual hospital data are available for a number of
process measures of the management of acute myocardial infarction, cardiac failure, pneumonia and
surgical infection prevention in the US from the Department of Health and Human Services
(www.hospitalcom- pare.hhs.gov).
Local comparisons can be made with similar-sized health services in the same region. Individual
health services can also review their rates over time to identify trends. When comparing data external
to the health service, it is essential that the definitions of the criteria being measured are rigorously
specified, and that the populations measured and the data collection methods are standardised.
Without standardisation, results may vary substantially because of differences in the methods used,
 2
Oufl
My
impr
ove
men
t

and differences' in the measures obtained may not represent differences in quality of care, making
comparisons meaningless. 9
If a significant difference is found between best practice and the care provided, an analysts of
the difference should be undertaken and appropriate action taken to try to close the gap. The
process of analysing the gap in care and taking action to close it is discussed funher in Chapter 6.
The audit results should be analysed for statistical and clinical significance. As these audits are
not randomised controlled trials, there is risk of bias from confounding variables. In particular,
organisational factors may greatly influence treatment effects, and it may therefore be difficult to
irplicate the results achieved in one setting in others. Also, organisational factors change in
individual organisations over time, altering the effects of interventions. 2
Many quality improvement programs continually measure performance using small samples of
data collected over time. Such measurement can be incorporated into the daily clinical routine. The
results can then be graphed (rather than using tables) on run or process control charts^o look for
variation in performance. Such charts plot the criterion measurement on the y axis and time on the
x axis. If the variation in performance exceeds three standard deviations (the control limit) above or
below the trend line, this is referred to as special-cause variation, and the reason for it should be
quickly sought. Variation within the control limits is regarded as common-cause variation, and is a
property of the system delivering care. Using graphs to depict the data allows variation in
performance to be easily seen and interpreted without requiring statistical expertise. Such time-
series data and the resultant process control charts provide continuous feedback to clinicians. 2
Often, measurements of criteria will be made using samples of patients, rather than all patients, to
reduce costs. However, in some clinical areas, activity may be low and the sample size will therefore
be small, even if all patients are included in the audit. Whatever the circumstances, when presenting
d2ta to clinicians, estimates of the precision of measurements (confidence intervals) should be
included.18

2.3.7 Taking action

It is essential that areas of concern raised by an audit are thoroughly analysed, and that
recommendations for changes in how clinical care is provided are made and acted upon.
Unfortunately, recommendations are often not made and action not taken following an audit, 28
making the process ineffective in improving quality of care, and probably demotivating staff to
continue with the method. Without appropriate action, an audit becomes an observational study
about clinical practice at one point in time, and is of limited value. Some actions may require
additional resources, which may test the senior management's support for clinical audit.
Importantly, once action is taken, the audit should continue to ensure that the gap detected has
been closed; this may take several audit cycles. A common error in clinical audit is to assume that
once action is taken, the gap will be closed. Such effective action is uncommon, and repeated audit
cycles are usually needed to successfully close a gap. Also, changes made in response to an audit
may have unexpected positive or negative consequences, which may also need to be managed. 23
The overall effect of audit and feedback in changing clinical behaviour was assessed in five
randomised controlled trials. A median absolute improvement of 7.0% (95% CI, 1.3%-16%) in the
process measures used was found. 32 These results will be discussed further in Chapter 6.

2.3.8 Reporting audit results

The results of an audit, and the resultant recommendations for changes in how patient care is
provided, should be reported to the relevant clinicians and their respective groups and throughout
the health service using its management and committee structure. Reporting clinical governance
activities, including clinical audit, will be discussed further in Chapter 7. Although such reports need
 Enhancing patient care

to be detailed, ii is imperative ihey maintain patients' privacy and confidentiality The reports should
contain appropriate statistical information (eg. means and confidencc intervals) together with a
simple
 2 OuflMy improvement

interpretation of the results for staff members

without training in statistics.8 Colourful tables
and graphs that clearly demonstrate trends
and significant results should also be used.

2-3.9 Resources
Health services' senior management must pro-
vide adequate resources to support clinical
audit so that it can be undertaken to a high
standard. Resources may include appropriate
dedicated staff, such as project officers and
secretarial assistance, and computers and
information technology support. Some of staff
involved in audit may need additional training
in measurement and analysis of data. 18 Most
importantly, clinicians cite a lack of protected
time to participate in audits as the most fre-
quent deficiency in such programs, 25 and this
critical issue should be addressed with the
relevant clinicians when designing the audit
program.

2.3.10 Education, communication and

leadership
All clinical staff should be provided with edu-
cation about clinical audit, which should help
engage them in the program. When an audit is
being considered, potential participants
should be informed of the proposal. Their
involvement and feedback should be actively
sought for the selection of the clinical area to
be audited, the audit criteria, design of data
collection tools, data analysis and review the Figure 2.2 Wlmmera model of quality
results. A senior clinician working in the improvement
clinical area under review should also have
overall responsibility for each of the major
components of the audit
30
process.
The quality improvement arm of the Wimmei model is represented diagramatically in Figure
2.2.

2.4 Quality improvement

collaboratives33
Another approach to quality improvement is for
a number of three- to four-person teams from
 Enhancing patient care

different health services (typically between 20

and 40) to collaborate and focus on quality
improvement in one specific are» of clinical care
over 1 to 2 years. Strategies for improvement
are designed and implemented, and learning
and insights are shared in multiple meetings
between teams over several months, with the
aim of achieving dramatic improvements in
short periods. New knowledge is applied
 2 Quality tmprovtnrmn

locally by teams using multiple plan-do-study-act cycles. A set of core indicators are determined tor
the collaborative, and reported progressively by each team.
Competition is promoted between teams, and the results of innovations are shared with a view to
rapidly developing effective strategies. An initial collaborative involving a particular area in a health
service may then be expanded to involve the whole* service and other health services. The
collaborative approach has been developed by the US Institute for Healthcare Improvement, and has
been used on a large scale in the US and the UK. There are similarities between the collaborative
approach and the usual audit cycle. Although collaboratives are conducted over multiple sites, the
many teams involved are thought to enhance learning *about, and development of, effective
strategies.
Collaboratives require substantial time and money. Unfortunately, acceptance qf the methodology
is largely the result of subjective impressions, such as shared beliefs and anecdotal evidence, rather
than rigorous, objective evaluation of its effectiveness and appropriate use. The studies are
uncontrolled and have design limitations. Publications of collaboratives consist largely of subjective
evaluations or self- assessments of the effects and lessons leamt during the project, without objective
evaluations of outcome changes.
Therefore, "the overall effectiveness [of the quality improvement collaborative method] remains
highly uncertain but is probably modest". 33 The improvements achieved by individual health services
range from remarkable to only modest, and success may be due to other factors within the
organisation. More research is required into the determinants of successful use of collaboratives to
achieve quality improvement.

2.5 Whole-of-practice audits

As well as conducting clinical audits for specific conditions, whole-of-practice clinical audits, in
which all clinical performance in a clinical unit or area is continuously assessed, should also be
undertaken in individual health services. Whole-of-practice audits are most developed in surgery. An
outstanding example is the Lothian Surgical Audit in Scotland, which commenced in 1946. This
audit has demonstrated a reduction in patient complications and mortality, has been instrumental
in changing surgical practice in that region, and has provided a valuable source of education and
research opportunities for surgeons.34
The model has been extended into all surgical specialties across 51 hospitals in Scotland as the
Scottish Audit of Surgical Mortality, and has been replicated in Western Australia as the Wfestem
Australian Audit of Surgical Mortality. Participation in the program by surgeons is voluntary and
involves "systematic, independent and external peer review" of "all patients who have died while
under the care of a surgeon".35 The Royal Australasian College of Surgeons intends to extend the WA
program throughout Australia and New Zealand. 36
For ongoing professional re-certification, the Royal Australasian College of Surgeons requires that
surgeons participate in peer-reviewed surgical audit, undergo credentialling at an approved hospital,
and maintain knowledge and skills. 37 A survey of Australian surgical units in public hospitals in
1998 found that, although most units conducted audits, the audits did not met the then
requirements of the College in over two-thirds of cases, because of insufficient resources. 38

2.5.I Data to be collected

Currently, the College suggests that surgical audits collcct data comprising a minimum and
expanded dataset; additional data can be added for each specialty group as required, or to meet an
individual surgeon^ requirements. Patient complications can be graded according to their severity

27
 2 Quality tmprovtnrmn

and recorded as pan of the audit. Using generic databases, individual health services can create
databases to meet their clinicians' needs. Importantly, the creation and maintenance of these
databases will require information technology suppon from the health service. 27
The data collected in whole-of-practice audits can provide useful information about clinical
workload and patient access (eg, patient throughput, waiting times, and operation postponements),
utilisation and efficiency (eg, use of investigations and adequacy of pain relief), and safety and
efficacy (eg, mortality, complications of surgery, re-operation and re-admission rates). 27
Whole-of-practice audits of medical services are more difficult to undertake than in surgery, as
medical patients often have several comorbidities, making meaningful comparisons more difficult.
However, despite this limitation, medical units should still be able to undertake meaningful whole-of-
practice audits in a similar manner to their surgical colleagues. For example, the National Sentinel
Audit of Stroke, conducted biannually by the Royal College of Physicians' Clinical Effectiveness and
Evaluation Unit since 1998, reviews all patients admitted with stroke across all hospitals in England,
Wales and Northern Ireland.39 In 2006, detailed data were compared with evidence-based standards
for the investigation and management of 13 625 patients admitted with stroke over 3 months,
demonstrating improvement in clinical practice, highlighting areas requiring attention, and allowing
health services to compare their performance with best practice and national data. 39

2.5.2 Review of audit results

In larger health services, whole-of-practice audit data should be formally reviewed on a regular basis
by each clinical or specially unit. The frequency of these reviews will depend on the size of the
hospital and its clinical throughput. A regular clinical or specialty unit review meeting attended by
panicipating clinicians may discuss the audit results, "near misses", incident reports and patient
complaints that have been made for the period under review, and may present specific cases. 27 The
discussion in these meetings should be "open, non-confrontational, non-judgmental, friendly and
confidential".30 In smaller hospitals, the health service's peak quality committee can review the
whole-of-practice audit results with its clinicians. The results should be reviewed over time to
identify trends within the health service and be benchmarked with similar clinical units in like
health services.

2.5.3 Implementing audit recommendations

Implementing some of the recommendations made by the audit review will require the assistance of
various groups (eg, safety committee, pharmacy committee) and individuals (eg, clinical risk
manager, infection control manager, supervisor of intern training) within the health service. For
some major recommendations, especially those involving several departments or clinical areas,
specific working parties may need to be established to analyse the problems and determine and
implement appropriate solutions. The process of implementing major change in how patient care is
provided will be discussed in detail in Chapter 6.
The health service's peak quality committee, hospital management and the clinical teams who
participated in the audits should monitor progress made towards implementing major audit
recommendations. It is important that audit recommendations are implemented in a timely manner.
Such action sends an important message to clinicians that the health service values clinical audit
and their participation in it, and is also genuinely committed to providing high-quality, safe care.
Without appropriate, timely action to close the gaps in care and reduce the risks to patients detected
by audits, the process will most likely wither.
2.5.4 Reporting audit results to the aedentlalllng and privileging committee
Whole-of-practice audit data detailing the performance of individual clinicians should also be
regularly forwarded to the health service's credentialling and privileging committee. These data

28
 2 Quality tmprovtnrmn

allow the conmnnrr to ensure that the clinical privileges granted to individual clinicians remain
appropriate. The comraioee should use the data to determine whether individual clinicians are
undertaking sufficient numbers of each operation or procedure they have been granted privileges to
perform to maintain their technical skills, and whether these operations and procedures are being
performed safely. If the committee is concerned with the performance of an individual clinician, it is
essential that data from the whole-of-practice audit (especially outcome measures) are reviewed to
ensure that they do not refleci casemix variation, aspects of the structure of the organisation, or the
processes of care used in managing such patients before the clinical performance of the individual
clinician is deemed primarily responsibleJor the outcomes.20

2.5.5 Feedback to clinicians

Whole-of-practice data should also be provided to individual clinicians who can then compare their
data over time, with clinicians undertaking similar work in the health service or in like health
services, and against standards set by professional organisations, best practice and national and
international data.8 Such feedback can lead to improvement in performance of 10% to 15%.*°

2.6 Clinical indicators

2.6.1 Definition
The Australian Council on Healthcare Standards defines a clinical indicator as:
a measure of the clinical management and outcome of patient care. They are not exact standards
against which hospitals must measure their clinical performance, but rather are designed as flags
that "indicate" possible problems or opportunities to improve patient care. They are a tool to assist in
assessing whether or not a standard in patient care is being met. 41 A clinical indicator has also been
defined as "a measurable aspect of care provided for which there is evidence that it represents
quality on the grounds of scientific research or consensus among experts". 22 42 More recently,
indicators have been used less as flags for potential improvement, and more as measures of
performance or quality.

2.6.2 Types of indicators

Clinical indicators are measures related to a clinical condition. They may be: 43
■ Structural indicators (what is needed). These provide qualitative information about the health
services environment (eg, physical resources, policies and procedures) and are usually present or
absent
■ Process indicators (what is done in providing care). These provide quantitative data on the
effectiveness of policies and procedures.
■ Outcome measures (what is achieved). These provide quantitative data on the outcomes of
systems. Process measures are favoured because they are often more sensitive to changes in how
clinical care is
being provided. Desirable characteristics of indicators include being:
■ measurable
■ clcarly definable
■ meaningful as a measure of care
■ valid and reliable
■ objective
■ able to be benchmarkcd over time or with like organisations, or national or international standards

29
 2 Quality tmprovtnrmn

The quality of many clinical activities can be measured using clinical indicators and benchmarking
the results obtained. However, determining the clinical area in which to measure clinical performance
can present problems because of the many different options that are available. These options include
measuring clinical performance in the management of patients who present with diagnoses in the
most frequent diagnostic-related groups in the health service, where measures are available for
benchmarking, what is measurable (although care must be taken, as what is measurable may not be
clinically important), and where the absence of a panicular component of care has a negative effect
on mortality and morbidity.
Importantly, what matters clinically may be difficult to accurately measure, and indicators may be
of limited use in some clinical areas. Also, different indicators may be important in different levels of
the health system. What is regarded as imponant and useful to measure by individual clinicians,
clinical teams, senior management, patients and health departments may be quite different.

2.6-3 Indicator programs

The major Australian program for acute health care quality indicators is the collaborative
development of clinical indicators by the Australian Council on Healthcare Standards and the
relevant speciality colleges. These indicators reflect the provider's perspective and emphasise
technical proficiency. They are available to all organisations, and are included in the Council's
accreditation requirements.
Reporting indicators to the Council is not mandatory for accreditation, but health services are
required to use data to demonstrate improvement in performance. Health services participating in the
clinical indicator program report their results to the Council for each indicator they have chosen to
measure. Aggregated data is provided by the Council twice yearly to allow similar-sized hospitals to
compare their results for particular indicators. 43 There are many indicator programs, including:
■ The Australian Council on Healthcare Standards Clinical Indicator Program — 23 clinical indicator
sets, covering general and specialist surgery, obstetrics, pathology, radiology, internal medicine,
intensive care, anaesthetics, mental health, adverse drug reactions, dermatology, emergency
medicine, and infection control (www.achs.org.au/ClinicalIndicators). Many of these indicator sets
have been updated multiple times.
■ The US Joint Commission on Accreditation of Healthcare Organizations National Hospital Quality
Measures — indicator sets for acute myocardial infarction, heart failure, community acquired
pneumonia, surgical care, pregnancy and related conditions, and paediatric asthma
(www.jointcom- mission.org/PerformanceMeasurement/PerformanceMeasurement)
■ The UK National Health Service (NHS) Better Care, Better Value Indicators — 15 indicators ol
efficiency including length of stay, day-case rate, preoperative bed-days (www.productivity.nhs.uk/
default.aspx).
■ The NHS Scotland Clinical Indicators — survival following hip fracture, stroke, acute myocardial
infarction; medical and surgical re-admissions; and mortality after surgery (www.indica-
tors.scot.nhs.uk).
■ AHRQ Quality Indicators are grouped in four areas: prevention (hospital admissions and disease
complications that could have been avoided with good outpatient care); inpatient (32 indicators
covering volume, utilisation and mortality); patient safety (indicators screen for adverse events);
and paediatrics (www.qualityindicators.ahrq.gov).
■ Quality Use of Medicines in Australian Hospitals — 30 indicators grouped into six sets: antithrom -
botic therapy; antibiotic therapy; medication ordering; pain management; continuity of care; and
hospital-wide medication management policies
(www.ciap.health.nsw.gov.au/nswiag/indicalors.html).

30
 2 Quality tmprovtnrmn

■ Victorian Maternity Services Performance Indicators — 10 key maternity services "evidence-

informed" indicators (www.health.vic.gov.au/maiernitycare).

31
 Clinical indicators that are valid, reliable measures of the quality of care being provided in
panirular clinical situations should be chosen and collected. An extensive review of the literature has
found very few indicators that meet these requirements. 44
At Wimmera Health Care Group, clinical indicators have been used as (lags of potential
deficiencies in service delivery systems. Clinical indicators are not used extensively, but on some
occasions, they have alened the hospital's clinical quality committee that there may be a problem in
a particular clinical area. In response to the level of a particular indicator, the committee has
requested that a focus audit and analysis be undertaken to determine the strengths and weaknesses
of the health care delivery systems that are involved.
The nature and content of individual audits will vary. Although the points listed in Checklist 2.1
are not relevant to every audit, the various issues they raise should be actively consideiyd when
designing and conducting an audit, rather than passively overlooked.
Owckfest 2.1 Conducting a dinkal audit715 "'27

________________________________________________Yas No WIPN/A
lAldMoristtRhamiaoaivad comprehensive education about the audit process. 1
2 Relevant staff have received training in how to undertake an audit.
3 Ihe aucft process has management support demonstrated by:
■ Adequate resources are available to conduct the audit (eg, a project officer if required, secretarial
assistance, protected clinician time to participate in the audit activities).
■ A senior executive staff member is closely involved with the audit process.
■ This staff member is respected by medical and nursing staff in the hospital.
4 The audit topic is selected after the consideration of the:
■ size of the gap in care being provided;
■ frequency, harm and preventability of a particular adverse event; or
■ benefits, costs and feasibility of an intervention.
5 The audt topic is regarded as important by clinicians.
6 There is a clinical leader of the audit.
7 Al potential participants in the audit are informed well in advance of the proposal to undertake the
audt They are invited to be involved with and give feedback regarding the development of the area to
be audited, the selection of audit criteria, the design of data collection tools, data analysis, and the
review of the audits findings.
8 The criteria against which performance will be assessed are developed by a
multidisciplinary team of relevant health professionals, reflect best practice and are:
■ balanced across the dimensions of care
■ evidence-based
■ measurable
■ collectable
■ key measures that affect patient outcome (about 12)
■ process measures, or outcome measures with short timeframes
■ able to be readily benchmarked
9 The suitability of the criteria are checked in a small-scale trial.
10 The care provided to all patients, or an appropriate sample of patients is audited.
11 Data entry is electronic where possible, and designed to minimise clinicians' time commitment.
12 Data collection is prospective where possible.
13 AB data collected are securely stored.
14 A sample of the data is routinely checked for accuracy and completeness.
15 If a paper-based audit is conducted, data extraction tools have been developed and
successfully tested on a small scale before being fully implemented.
16 Where possible, some data that are already being routinely collected are used in the audit.
17 Data that identify individual patients are kept confidential.
 2 Quality improvement

18 The results of the audit are reviewed constructively in meetings of relevant

clinicians, and appropriate recommendations to improve the quality and safety of
patient care are made.
19 The results of audits and the recommendations made are forwarded to all participating
dinidans. relevant staff and senior management at the health service.
20 Benchmarking over time, between clinical units and against like health services and
national and international data are undertaken.
21 Recommendations for action are analysed and acted upon in a timely manner.
22 The audit cycle continues until the identified gap in the care being provided is closed.
23 The criteria against which performance is assessed are regularly reviewed to consider changes in
best practice over time.
24 Comprehensive whole-of-practice audits are continuously conducted in all major clinical
areas of the healtfi service.
25 The results of whole-of-practice audits are regularly forwarded to relevant clinicians, senior management and the
credentialling and privileging committee of the health service._________________________________________________
j viflP = *xv m progress N/A = not applicable
m Quality improvement is a major component of clinical governance.
■ The quality improvement process identifies gaps between the actual performance in providing care
and (he desired standard of care, then takes action to close the gaps.
■ The usual desired standard of care is best practice according to the highest-quality scientific
evidenoe available at the time.
■ Multiple sources of information about what constitutes best practice in many clinical areas are
available in written and electronic form.
■ Choosing clinical areas and interventions for improvement may be difficult because of the large
number of opportunities available. ♦
■ Clinical audit is the method used to detect gaps in care.
■ The clinical area to be audited should be regarded as important by clinicians working in that area.
■ Adequate resources are required to undertake clinical audit.
■ Participating clinicians need to be kept informed and involved with each stage of the audit process.
■ The criteria used to measure performance require careful selection. Where possible, a small number
of key evidence-based process measures or outcome measures with short time frames should be
used.
■ The amount of time required by clinicians to fully participate in clinical audit is critical to its success.
Data entry should be designed to require minimal clinician time, and clinicians should be given
protected time to fully participate in the audit process. '
■ The results of audits should be reviewed constructively in meetings of relevant clinicians, and
appropriate recommendations about patient care should be made.
■ The results of audits and the recommendations made should be forwarded to all the clinicians
involved in providing care in the clinical area, relevant staff and senior management.
■ Recommendations for action arising from the analyses of the results of audits should be acted upon
in a timely manner.
■ Multiple audit cycles in which performance is measured and action is taken are almost invariably
required to close a gap in care.
■ The results of audits should be benchmarked within the health service and externally with like
organisations.
■ As well as specific audits, comprehensive whole-of-practice audits should be conducted in all major
clinical areas of a health service, and the results should be used in the credentialling and privileging
process.
■ Clinical indicators are flags of possible problems or opportunities to improve care for particular
clinical conditions. A variety of indicators in many clinical areas are available from programs
developed in several countries.
References
t Schuster MA. McGtynn EA, Brook RH. How good is the quality of health care in the United States?
Milbank Q 2005; 83:843-896.
2 Pronomst pj. Nolan T, Zeger S, et al. How can clinicians measure safety and quality in acute care?
Lancet 2004; 363:1061-1067.
3 National Institute of Clinical Studies. Evidence-practice gaps report. Vol 1. Melbourne: NICS. 2003.
www.nhmrc.gov.au/nics/asp/irKlex.asp (accessed Oct 2008).
4 National Institute of Clinical Studies. Evidence-practice gaps report. Vol 2. Melbourne: NICS. 2005.
www.nhmrc.gov.au/nics/asp/index.asp (accessed Oct 2008).
5 Shojania KG. Duncan BW. McDonald KM, Wachter RM. editors/Making health care safer: a critical
analysis of patient safety practices. Rockville, Md: Agency for Healthcare Research and Quality, 2001.
(Evidence Reporl/ Technology Assessment No. 43; AHRQ publication 01-E058.)
6 Shojania KG. Duncan BW, McDonald KM, Wachter RM. Safe but sound. Patient safety meets
evidence-based medicine. JAMA 2002; 288: 508-513.
7 Mainz J. Developing evidence-based clinical indicators: a state of the art methods primer. Int J Qua!
Health Care 2003; 15:i5-i11.
8 Wollersheifn H. Hermens R. Hulscher M, et al. Clinical indicators: development and applications.
Neth J Med 2007; 65: 15-22.
9 Pionovost PJ. Thompson DA. Holzrnueller CG. et al. Defining and measuring patient safety. Crit Care
Clin 2005; 21: 1-19.
10 Nelson EC, Mohr JJ, Batalden PB, Plume SK. Improving health care, part 1: the clinical value
compass. Jt Comm J dual Improv 1996; 22: 243-258.
11 Institute of Medicine. Crossing the quality chasm: a new health system for the 21st century.
Washington. DC: National Academy Press. 2001.
12 Blumenthal D, Epstein A. The role of physicians in the future of quality management. N Eng J Med
1996; 335: 1328-1331.
13 Bion JF. Heffner JE. Challenges in the care of the acutely ill. Lancet 2004; 363:970-77.
14 Institute of Medicine; Field MJ. Lohr KN. editors. Guidelines for clinical practice: from development
to use. Washington. DC: National Academy Press. 1992.
15 Mainz J. Defining and classifying clinical indicators for quality improvement. Int J Oual Health Care
2003; 15:523- 530.
16 Shoiania KG. Sampson M. Ansari MT, et al. How quickly do systematic reviews go out of dale? A
survival analysis. Ann Intern Med 2007; 147: 224-233.
17 Donabedian A. Evaluating the quality of medical care. Milbank Mem Fund O 1966; 44: 166-206.
18 Berenhotlz SM, Pronovost PJ. Monitoring patient safety. Crit Care Clin 2007; 23: 659-673.
19 Pronovost PJ. Angus DC, Dorman T. et al. Physician staffing patterns and clinical outcomes in
critically ill patients: a systematic review. JAMA 2002; 288: 2151-2162.
20 Litford R. Mohammed MA, Spiegelhalter D. Thomson R. Use and misuse ol process and outcome
data in managing performance of acute medical care: avoiding institutional stigma. Lancet 2004; 363:
1147-1154.
 2 Quality improvement

21 Donabedian A. Explorations in quality assessment and monitoring. Vol II. The criteria and
standards of quality. Ann Arbor. Mich: Health Administration Press, 1982.
22 Lawrence M. Olesen F. Indicators of quality in health care. Eur J Gen Pract 1997; 3: 103-108.
23 Seddon M. Buchanan J. Quality improvement in New Zealand healthcare. Part 3: achieving care
through clinical audit. NZMedJ2006; 119: U2108.
24 Mant J, Hicks N. Detecting differences in quality of care: the sensitivity of measures of process
and outcome in treating acute myocardial infarction. BMJ1995:311: 793-796.
25. National Institute for Health and Clinical Excellence. Principles for best practice in clinical audit.
Oxford: Radcliffe Medical Press. 2002. www.nice.org.uk/media/796/23/BestPracticeClinicalAudit.pdf
(accessed Oct 2008).
28 Brook RH. McGlynn EA. Cleary PD. Measuring quality of care. N Eng J Med 1996; 335: 966-970.
2
Quali
ty
impr
ovem
ent

27 Watters DA, Green AJ, van Rij A. Guidelines for surgical audit in Australia and New Zealand. ANZ
J Surg 2006; 76 78-83.
28 Brown PM. The audit cycle: a guide to success. Br J Hosp Med 2006; 67:116-117.
29 Collopy BT. Ensuring accuracy of clinical data is only part of the audit process. J QuaI Clin Pract
2001; 21: 74-75.
30 Gibbons AJ, Dhariwal DK. Audit for doctors: how to do it. BMJ2Q03, 327: S1-S2.
31 Pronovost PJ, Berenholtz SM, Goeschel CA, et al. Creating high reliability in health care
organizations. Health Sarv Res 2006; 41: 1599-1617.
32 Grimshaw JM, Thomas RE. MacLennan G, et al. Effectiveness and efficiency of guideline
dissemination and implementation strategies. Health Technol Assess 2004; 8:1-84.
33 Mittman BS. Creating the evidence base for quality improvement collaboratives. Ann Intern Med
2004; 140:897- 901.
34 Aitken RJ, Nixon SJ, Ruckley CV. Lothian Surgical Audit: a 15-year experience of improvemeat
in surgical practice through regional computerised audit. Lancet 1997; 350: 800-804.
35 Semmens JB, Aitken J, Sanfilippo FM, et al. The Western Australian Audit of Surgical Mortality:
advancing surgical accountability. Med J Aust 2005; 183: 504-508.
36 Semmens JB, Mountain JA, Sanfilippo FM, et al. Providers and consumers support the Western
Australian Audit of Surgical Mortality. ANZ J Surg 2006, 76: 442-447.
37 Royal Australasian College of Surgeons. Continuing professional development program
information manual 2007- 2009. Melbourne: Royal Australasian College of Surgeons, 2006.
www.surgeons.org/Content/NavigationMenu/
FellowshipandStandards/CPDRecertification/default.htm (accesssed Oct 2008).
38 Eno LM, Spigelman AD. A survey of surgical audit in Australia: whither clinical governance? J
Qua! CUn Pract 2000; 20: 2-4.
39 Royal College of Physicians. National Sentinel Audit 2006. Prepared on behalf of the Intercollegiate
Stroke Working Party. London: RCP, 2007.
 40 Van der Weijden T, Grol R. Feedback and reminders. In: Grol R, Wensing M. Eccles M, editors.
Improving patient care: the implementation of change in clinical practice. Edinburgh: Elsevier,
2005:158-172.
41 Australian Council on Healthcare Standards. Clinical indicators — user manuals. Sydney: ACHS.
1998.
42 Australian Council on Healthcare Standards. ACHS Clinical Indicator Program Information 2008.
Sydney: ACHS. 2007.
43 Campbell SM, Braspenning J. Hutchison A, Marshall M. Research methods used in developing
and applying quality indicators in primary care. Oual Sal Health Care 2002; 11: 358-364.
44 McNeil JJ Report on patient safety monitoring systems. Melbourne: Victorian Department of Health
and Aged
Care. 1999. □
Enhancing patient care

3 Risk management — why do adverse events

happen to patients?

This chapter will:

■ Discuss health care delivery systems and their properties
■ Examine the frequency of adverse events in health care and their preventatiblity
■ Examine reasons why errors occur and the role of latent factors, active errors and barriers
■ Compare the punitive and systems responses to errors
■ Characterise an ideal quality improvement and risk management system in health care *
■ Provide a flow chart of the Wimmera Model of Clinical Quality Improvement and Risk Management.

3.1 Health care delivery as a system

The second major component of clinical governance is risk management. Unless adequate barriers
are in place, accidents will occur.' In health care, an accident that causes harm to a patient is called
an adverse patient event. When looking at why adverse patient events happen, it is useful to view
health care delivery as a system. This systems approach helps develop models to explain how adverse
events occur. The factors contributing to the event are determined, and this information helps
develop strategies to make the system safer by reducing the probability of adverse events recurring.

3.1.1 Defining o system

Industries deliver their products and services using systems. Health care is no exception, although it
is somewhat behind other industries in embracing a systems approach to improving the quality of
the
services it provides.
A system comprises many inter-related processes designed to achieve an objective. For example, a
patient attending an emergency department with chest pain will be entering a system designed to
deal with that clinical problem. The processes in this system may include: triage of the patient by
nursing staff; medical assessment of the patient by hospital medical officers; an intravenous line
being inserted: an electrocardiogram being performed and interpreted; a blood sample being taken,
sent to the pathology laboratory, analysed and the results reported; and oral and intravenous
medication being administered.
Importantly, health care delivery systems differ from other complex industries, such as aviation,
where service delivery systems comprise a greater proportion of mechanical, rather than human,
 2 Quality improvement

components Unfortunately, the greater human component in health care, and the sometimes high
level of uncertainty in the diagnosis and treatment of patients, may result in error and gaps in
patient care occurring more frequently than in other, more mechanical, industries.

3.1.2 Properties of systems

The outcome produced by a system is a property of that system. Improving the outcome requires a
change in the processes that make up the system, not merely a requirement that health professionals
work harder within the existing system.
Often, the relationships between individual processes in health care delivery arc not simple causc
and effect, bui are complex networks of relationships that can change with time 2 In the above
clinical
 Enhancing patient care

scenario, a patient may be unable to give a clear history of their chest

pain, which may result in inappropriate triage and delayed medical
assessment. The process of patient assessment outside of normal working
hours may require a doctor to be paged to the emergency department,
thereby introducing an additional process into the system. After-hours
laboratory staff may need to be called into the hospital to perform blood
tests, which may delay analysis and reporting of results.
Health care is provided using simple and complex delivery systems.
Complexity can arise from the number of processes in the system and the
number of options available to clinicians at each decision point in the
system. Simple systems can have predictable outcomes. Complex
systems, however, are usually not linear, and can have multiple
interactions between its components and with components external to the
system.3 Outcomes produced by complex systems can rarely be predicted
with certainty. Unexpected and unintended negative outcomes occur
frequently, and are more likely to occur as systems build in complexity. 4
When a change is made to a system, factors not directly related to the
change may affect the outcome. Many components of complex systems
are inter-related, and a change in one component may unexpectedly affect
another. In particular, social interactions in systems are rarely linear and
are often unpredictable. 3 System participants may view the system's
function and purpose differently. 5 Unfortunately, it is not possible to keep
all the other components of the system constant while a change to one
component is made.3
These general properties of complex systems are important to consider
when determining how health care delivery systems should be redesigned
to reduce adverse events. Because of the complexity of many health care
delivery systems, a systematic, detailed analysis — of the characteristics
ol any proposed change, the people who will be required to change their
behaviour, and the setting in which the change is proposed — will need to
be made, and a change implementation strategy tailored to these
characteristics should be developed. Relevant staff members should help
identify interventions with the greatest probability of success and
important barriers and facilitators of the proposed change or
intervention.3 The process of implementing change will be discussed in
detail in Chapter 6.

3.2 The extent of adverse events

Previously, iatrogenic patient injury in health care was thought to be rare,
unpreventable and related to the performance of individual clinicians.
"Acute care hosptals However, the Harvard Medical Practice Study, which reviewed a random
are complex, dynamic sample of 30 000 patients in acute hospitals in New York state in 1984,
environments (hat
kjnctcn (or
showed that iatrogenic injury occurred in 3.7% of hospitalised patients. 6
maftjx dun) as a result The study concluded that iatrogenic complications affect over 1.3 million
Of tie interplay of
tfwQTiO departments;
loams and
individuate.'
Cook OJ. et
al Lancet
2004; 363:
1224-1230
 3 Riek management

hospitalised patients annually in the United States, and that almost 70%
ol these events are preventable. 6
 Enhancing patient care

The Quality in Australian Health Care Study showed that 16.6% of patient admissions in
Australia were associated with an adverse event. 7 Over half of the adverse events were considered to
be preventable. Extrapolating these findings to all Australian hospitals indicated that about 470 000
patient admissions and 3.3 million bed-days annually were attributable to adverse events.
There has been much debate about why the rate of adverse events in Australia was so much
higher than in the US. Further studies were undertaken, comparing the adverse event rate in Utah
and Colorado with the results of the Quality in Australian Health Care Study. After taking into
account the differences in how the studies were conducted and the how the data were treated, the
adverse event rale in Australia was adjusted to 10.6%, compared with 3.2% in "the US. This
threefold difference was thought to be due to dilTerent thresholds for admission and discharge in
Australian hospitals, and underreporting of some adverse events by reviewers in the US. 8 A further
comparison was undertaken, and no differences in quality of care were identified between the
countries.9
The Harvard Medical Practice Study methodology was replicated in New Zealand, the United
Kingdom, Denmark, France, the Netherlands and Canada. Although these studies found different
rates of adverse events, it is thought that about 10% of hospital patients experience an adverse
event, and at least 50% of these are preventable. 10

3.3 The reasons why errors occur

Organisational accidents in other complex industries, such as aviation and nuclear power, have
been analysed extensively.11 The model developed by Professor James Reason helps explain why
accidents occur (Figure 3.1). 12

3.3.1 Latent factors

Every organisation has a number of latent factors that provide the background against which an
error can occur. The number and magnitude of these latent factors are frequently determined by
decisions made a long way from the coalface. Usually, senior management makes decisions about
how resources are allocated to each area in an organisation. Inevitably, because of the finite
resources that are available, there is a trade-ofT between the quantity of resources allocated for
running the organisation and clinical care and the resources directed toward safety. These resource
allocation decisions affect many aspects of the organisation's work environment, including:

Figure 3.1 Organisational accident model 12

Contributory factors influencing clinical Defenc practice Accident/

TaskError- e incident
producing barriers Errors
Managemen
t decisions conditions
and
organisatio Violation
nal - Violations
processes producin
g
condition
s
TTT
 Enhancing patient care

■ the number, seniority, knowledge and skill level of the staff employed to
work in each area
■ the workload each individual is expected to complete
■ the length and frequency of work shifts
■ the degree of supervision of staff members
■ the quantity and quality of equipment provided to undenake the
required tasks
■ the frequency of equipment maintenance.
How senior management allocate the resources available to the
organisation will affect the probability of particular types of adverse events
occuring. For example, a decision may be made by senior management to
reduce the frequency of equipment maintenance. This reduction may
increase the probability of a serious equipment failure occurring.
Although reducing maintenance will save costs in the short term, the
resultant equipment failure could contribute to an accident or downtime
occurring and reduced production in the long term.
However, attempting to address individual latent factors is not always
straightforward. For example, efforts have been made in many countries
to reduce clinician shift lengths to reduce fatigue in an attempt to reduce
the frequency with which clinicians make errors. However, reducing shift
lengths increases patient handovers between shifts, which may result in
increases in communication failure and discontinuity of care, leading to
adverse events. 13 Latent factors are "accidents waiting to happen". 14 Often,
latent factors are not addressed proactively because they contribute to a
background level of risk that is often not recognised until a serious event
occurs. Funher, a decision by management that increases the magnitude
of a latent factor may not have an immediate effect. The adverse events
resulting from this management decision may take months or years to
occur. Therefore, one of the most productive things an organisation can
do is proactively examine all its potential latent factors, and take steps to
reduce their number and size, thereby strengthening the systems of care
before a serious event occurs. "Latent failures represent a tremendous
opponunity to improve patient safety." 13

3.3.2 Active errors

On the other hand, active errors occur closer to the interface where health
professionals provide clinical care to patients. The Agency for Healthcare
^Heaflhoate Research and Quality in the US defines medical errors as "mistakes made
systems are in the process of care that result in, or have the potential to result in,
compter adaptive
harm to patients".15 Active errors can be classified according to the
systems—the parts
have toe freedom and appropriateness (or otherwise) of the management plan formulated for the
ab&ty to respond to patient's care. The management plan may be appropriate for the clinical
Stimuli in many different situation, but not be correctly executed because errors such as slips (eg,
arid fundamentally due to inattention) or lapses (eg, due to failure of memory) occur.
unpredictable ways.'
Alternatively, the action may be performed correctly, but the plan may be
Ptsek PE. Green inappropriate because of mistakes such as not applying an appropriate
halgh T 8MJ2001;
323 625-628 rule or a lack of knowledge about how to deal with a particular clinical
situation. Unlike latent errors, active errors are more likely to have an
immediate effect.
 Enhancing patient care

Errors can be further classified as errors of omission (failure to provide a component of

appropriate care) and errors of commission (where a component of care is provided but it is
inappropriate). Errors of omission are more common, more difficult to identify, and are thought to
have a greater effect on patient
safety.16,17
Latent factors and active errors frequently act together, often through a unique combination of
circumstances. Fortunately, barriers are often in place to prevent adverse events, even when an
active error has occurred and latent factors are present. The strongest barrier against adverse events
may be patients' resistance to potentially harmful events.
For example, a patient may be given incorrect medication or the wrong dose of a medication but
suffer no ill effects; or an active error may delay the diagnosis of a serious condition being made, but
often the patient will make a full recovery, perhaps delayed, despite the error having occurred.
Further, there are often double and triple checks built into health care delivery systems that art as
barriers to adverse events occurring. For example, a nurse or ward pharmacist may detect an
incorrect dose of a drug written by a doctor on a medication chart. Similarly, a mechanical barrier
such as an oximeter alarm may be in place to warn an anaesthetist that a patient's oxygen
saturation is falling.

3.4 Errors and adverse events

Although some errors can be fatal or cause'significant and permanent harm to patients, it is
important to realise that not all errors result in harm, and that errors are far more common than
adverse events. 18 Most errors occur with medications, and the vast majority of these are minor and
do not harm patients. Such errors may include patients being administered medication at the wrong
time or being given medication that was prescribed for another patient. Some of these errors are
detected, and intervention occurs before they reach the patient. Others do reach the patient, but
have little or no effect. Adverse events can occur as the result of error or despite appropriate care
having been given. For example, a procedure that is correctly indicated for a patient may be
performed correctly, but still result in a complication (eg, infection or haemorrhage).
There may be a bias in the errors that are detected. It is more likely that errors related to treatment
will be detected than errors related to diagnosis or prevention. Often, treatment errors are highly
visible or recognised more quickly because the patient has experienced obvious harm. Failure to
diagnose or adequately prevent a medical condition may not be realised for months or years. 19

3.5 The response to errors

Previously, when adverse events have occurred in the delivery of medical care, there was a tendency
to blame individual health care professionals, punish them, demand perfection in their performance
and insist that they try harder. Importantly, such punitive responses, rather than positively
changing clinical behaviour, have frequently led to adverse events and near misses not being
reported for fear of punitive action being taken. 20 Punitive responses result in under-reporting, the
continued presence of hazards to patient safety, and valuable opportunities being missed to examine
and strengthen the systems that allow adverse events to occur. Viewing health care deliver)' as
comprising a number of systems focuscs the analysis of adverse events on the components of the
systems involved, and away from apportioning blame to individuals.
 Enhancing patient care

Why is the natural tendency to blame individuals when an error occurs?

At the individual level, it is a human tendency to blame others when bad
outcomes occur. When people perform a task poorly, it is often attributed
to their personality or ability, but invariably, on more detailed
examination, the circumstances in which the error occurred are also
found to have affected the individual's actions. Further, many people
erroneously believe that only incompetent clinicians make errors, whereas
in reality, highly competent doctors can also make mistakes. Individuals
are also blamed for making errors with the bias of hindsight — others
examining the incident often believe that errors were more foreseeable
than they were in real time, because their knowledge of the outcome alters
their perceptions of the incident. 21
At the organisational level, it is easy to find errors made by individuals
at the staff-patient interface or "sharp end" after an incident has occurred.
Therefore, investigation of the incident may stop at the interface, without
an examination of the conditions of work or latent factors that were
present at the lime the incident occurred, which conveniently limits any
attribution of blame at the higher management levels in the
organisation.21
The modem industrial approach to error reduction is to avoid blaming
individuals unless they have committed an intentionally unsafe act, and
instead analyse the components of the systems that contributed to the
error occurring. Then, the latent factors that created the environment that
allowed the error to occur can be dealt with, improving the system as a
whole. Delivery systems can be redesigned to reduce the probability of
errors occurring by taking appropriate action, including:
■ simplifying the systems involved
■ standardising procedures
■ introducing constraints
■ using reminders and checklists.
These actions will be discussed in detail in Chapter 6.
It is important to acknowledge that, although some authors believe that
health care can learn from other industries, they have also questioned the
complete applicability of such industrial approaches to health care, citing
PRMisioniNotosihe the highly complex nature of health care delivery, the very individual
defray of mKons of
needs of each patient (as opposed to producing a relatively small number
sendees by
thousands ct heath of different types of items on a production line), the uncertainty associated
professionals to hunteds with many o( the decisions that are made in providing care to a patient
of thousands of (eg, when the diagnosis is uncertain or a patient has several associated
inpatients using comorbidities), the incomplete evidence base in the scientific literature
complicated equipment,
upon which to base rational decision making, and the unpredictable,
procedures and
processes. A major sometimes extreme workloads. It has been suggested that a military model
difficulty is that a zero
error rate... is not
seriously possible to get
close to. let alone
achieve, in health care.
People are living longer,
with more serious
illnesses and multiple
comorbidities, and
treatment is complex,
sometimes risky, even
dangerous.'
Braithwaite J,
Travaglia JF Aust
Health Rev 2008:
32: 10-22
may be more applicable in managing acutely ill patients "with the
uncertainties of warfare, friendly fire and 'collateral damage'". 13
 3
RM<
man
age
mtra

3.6 The basic risk management process

A basic risk management program should detect adverse events that have occurred in an organisation;
analyse them and ascertain their level of risk; determine and take appropriate action to reduce the
probability of the adverse event recurring; and continue to monitor activity to assess whether the
actions taken have been successful (Figure 3.2).
Also, risks present in an organisation can be detected by analysing adverse events that occurred in
similar organisations, and the associated risk reduced in a similar manner as if the event had occurred
within the organisation.

3.7 The basic risk management process in health care

The basic risk management process requires expansion for use in health care. Clinical activity should
be screened for potential adverse events and
clinical incidents should be reported. Clinical
Figure 3.3 The Wimmera clinical risk review is then required to assess if an adverse
management model event has occurred. If so, its risk is determined
according to its probability and consequence.
Risks to the health service can also be detected
by analysing events that have occurred in other
health services and can be reduced in a similar
manner than if the event had occurred in the
organisation.
Once identified, risks require prioritisation for
action. Such action many involve minor
changes, but more often major changes to

health care delivery systems are required.

Most importantly, after action has been taken, clinical activity should continue to be monitored and
evaluated to determine whether the action has been successful in reducing the nsk and preventing
the panicu- lar adverse event from recurring (Figure 3.3)
 3 Riek management

3.8 Characteristics of an ideal quality improvement and

risk management system
What characteristics of best-practice quality improvement systems should we strive to incorporate when
developing a quality improvement and clinical risk management program? Although all the following
criteria are unlikely to be met by one quality improvement approach or system, they provide valuable
signposts to the direction that quality improvement and risk management projects should take in
individual organisations. The criteria are:
■ positive impact on the standard of patient care
■ continuous
■ misses few (if any) major negative outcomes
■ provides, direct, timely feedback and action
■ automatic, continuous monitoring of risk and the effects of action
■ uses multiple sources of information
■ relatively inexpensive
■ confidential
■ objective (uses external performance standards where possible)
■ provides a comprehensive, continuous flow of information to hospital staff members, senior
executives and the board of management
■ logical and scientific
■ involves all clinical staff members and functions across the continuum of care in the hospital
■ provides valid, useful information
■ simple to understand and apply
■ takes a minimum of time.

3.9 The Wimmera quality improvement and risk

management model
The Wimmera quality improvement and clinical risk management model attempts to address most of
the above criteria, and provides a systematic approach to quality improvement and clinical risk
management that comprises five basic, sequential and coordinated components:
■ detecting gaps in patient care, adverse events or risk to patients using multiple methods
■ analysing these gaps, adverse events and risks
■ determining the importance of each gap and the risk each event poses to patients and the hospital,
and prioritising them
■ determining and taking appropriate actions to close the gaps or reduce the risks
■ monitoring the effects of these actions.
The components of the model are outlined in Figure 3.4.
Over the next four chapters, we will examine each of these components in detail, and describe the steps
that can be followed to implement such a program in a hospital.
 Enhancing patient care

Figure 3.4 The Wlmmere quality tmprovement end risk management model

■ Health care is delivered using complex systems consisting of many inter-related processes and a
large human component.
■ The outcomes produced by complex systems are uncertain.
 3 Riek management

■ Gaps in care, and unexpected and. unintended negative outcomes, occur frequently. Further, a
large proportion are preventable.
■ Adverse events occur when active errors (eg, slips, lapses, rule violations or inadequate
knowledge) interact with latent factors (eg, inadequate supervision, lack of skills, high workloads
and inadequate equipment) and barriers are ineffective in preventing the adverse event from
occurring.
■ The usual response to errors in heath care is punitive — to blame individuals, expect perfection
and demand that individuals try harder.
■ Punitive responses do not change clinical practice, and may result in under-reporting of adverse
events to avoid punishment.
■ A systems approach to gap and adverse event analysis can identify weaknesses in health care
delivery systems. These systems can then be redesigned to close gaps and reduce the probability of
adverse events recurring.
■ The characteristics of an ideal quality improvement and risk management program can be defined.
■ The Wimmera quality improvement and clinical risk management program:
> detects gaps in care, adverse events and risks using multiple methods;
> analyses them;
> assesses the importance of each gap and the risks associated with each adverse event and
prioritises them;
> determines and takes appropriate actions to close the gaps or reduce the risks; and
> monitors the effects of these actions.
References
1 Reason J. Human error. Cambridge. Mass: Cambridge University Press. 1992.
2 Senge P. The fifth discipline: the art and practice of the learning organization. London:
Doubleday/Cenlury Business, 1990.
3 Litaker D. Tomolo A. Liberatore V, et al. Using complexity theory to build interventions that improve
health care delivery in primary care. J Gen Intern Med 2006:21: S30-S34.
4 Nolan TW. Understanding medical systems. Ann Intern Med 1996; 128:293-298.
5 lies V. Sutherland K. Organisational change: a review for health care managers, professionals and
researchers London: National Co-ordinating Centre for NHS Service Delivery and Organisation R &
0.2001.
6 Brennan TA. Leape LL. Laird NM. et al. Incidence of adverse events and negligence in hospitalised
patients: results of the Harvard Medical Practice Study 1. N Engl J Med 1991; 324:370-376.
7 Wilson RM. Runciman WB. Gibberd RW, et al. The Quality in Australian Health Care Stufly. MedJAust
1995:163: 458-71.
8 Thomas EJ. Studdert OM. Runciman WB. et al. A comparison of iatrogenic injury studies in
Australia and the USA I: context, methods, casemix. population, patient and hospital characteristics.
Int J Oual Health Care 2000; 12: 371-378.
9 Runciman WB. Webb RK, Helps SC. et al. A comparison of iatrogenic injury studies in Australia and
the USA II: reviewer behaviour and quality of care. Int J Qual Health Care 2000:12:379-388.
10 Leape LL. Scope of problem and history of patient safety. Obstet Gynecol Clin N Am 2008; 35:1-10.
11 Leape LL. Error in medicine. JAMA 1994; 272:1351-1857.
12 Reason J. Understanding adverse events: the human factor. In: Vincent C. editor. Clinical risk
management enhancing patient safety. London: BMJ Book's. 2001.
13 Bion JF, Heffner JE. Challenges in the care of the acutely ill. Lancet 2004; 363: 970-977.
14 Shojania KG, Wald H, Gross R. Understanding medical error and improving patient safety in the
inpatient setting. Med Clin N Am 2002: 86: 847-867.
 Enhancing patient care

15 Agency for Health Care Research and Quality. AHRQ's Patient Safety Initiative: building
foundations, reducing risk. Interim report to the Senate Committee on Appropriations. Rockville. Md:
AHRQ. (AHRQ Publication No. 04- RG005, 2003.)
16 Angus DC, Black N. Improving care of the critically ill: institutional and health-care system
approaches. Lancet 2004; 363: 1314-1320.
17 Cook DJ, Montori VM, McMullin JP, et al. Improving patients' safety locally: changing clinician
behaviour Lancet 2004; 363: 1224-1230.
18 Resar RK, Rozich JD, Classen D. Methodology and rationale for the measurement of harm with
trigger tools. Oual Sal Health Care 2003:12 (Suppl II): ii39-ii45.
19 La Pietra L. Calligaris L. Molendini L. et al. Medical errors and clinical risk management: state of
the art. ACTA Otorhinolaryngol Ital 2005; 25: 339-346.
20 Leape LL. Why should we report adverse events? J Eval Clin Pract 1999; 5:1-4.
21 Reason JT. Carthey J. de Leval MR. Diagnosing "vulnerable system syndrome": an essential
prerequisite to effective risk management. Oual Saf Health Care 2001; 10: ii2l-ii25. □
Enhancing patient care

4 Detecting adverse events and risk

This chapter will:
■ Discuss the methods available to detect adverse events and risk in health services
■ Describe the major reactive methods of detecting adverse events and proactive methods of JMucMng rirtc. and their
advantages and disadvantages.

4.1 Methods of detecting adverse events and risk

o
As has been discussed previously, clinical risk management is about ensuring that adverse patient
events occur less frequently. The first step in reducing the frequency of adverse events is to identify
when such events have occurred. Such information is available to health services from two imponant
sources:
■ Adverse events that have occurred to patients within the health service. This method is referred to as
reactive risk management. No single method has been developed to detect all adverse events that
occur, so multiple, diverse methods are used.
■ Adverse events that have occuned in other health services should be obtained from many sources
and used to examine the health service's relevant delivery systems to identify areas of risk within the
service. This method is referred to as proactive risk management.
The benefits of this second approach are numerous. Lessons can be leamt and applied from a much
larger pool of adverse events than those that might occur in an individual health service. Information
about adverse events that occur rarely in individual health services can also be utilised. Using
information about adverse events that have occurred elsewhere, weaknesses in health care delivery
systems can be proactively identified without patients being harmed or staff in the health service
having to experience the psychological trauma of a serious adverse event.
There are four basic methods that can be used to detect the adverse events that have occuned to
patients in a health service:
■ directly observing an adverse event
■ receiving a legal claim made by a patient concerning an adverse event
■ reading the details of an adverse event in the patient medical record
■ being told the details of an adverse event by a clinician or patient.
Information about adverse events at other health services is available
from:
■ legislative requirements to report specific clinical events (eg, consultative committees in specific
clinical areas, coroners' inquests)
 3 Riek management

■ voluntary and mandator)' reporting of clinical incidents and specific events (eg, national databases of
clinical incidents and sentinel events)
■ commercial requirements to report adverse events or risks to the health care field (eg. drug and
equipment alerts from pharmaceutical companies and medical equipment manufacturers)
■ media reports of adverse events
■ clinical journal articles in which adverse events are described
■ the findings of inquiries into individual health services or specific clinical areas in those health
services.
As well as health services learning from adverse events that occurred elsewhere, different t\pcs of
proactive risk management methods are available to reduce risk and prevent adverse events, including
■ examining the work conditions in the health service when designing health care delivery svstcms.
and taking action to strengthen any weaknesses that arc dctcctcd
 4 Detecting advert* events and rta*

Despite these limitations, claims can still 3 Previous failure of medical management or
provide useful information about rare events unfavourable results
that occur in health services, and often provide 4 Trauma incurred in hospital
a very detailed (albeit legally based) analysis of 5 Unfavourable drug reaction4n hospital
an individual adverse event. Claims can detect
6 Transfer from general care to a special care unit
latent factors that were present when the
7 Transfer to another acute care hospital
adverse event occurred and deficiencies in
8 Return to operating theatre during current period of
processes of care that were delivered. With the
hospitalisation
analysis of large samples of claims usually
9 Treatment for organ damage after an invasive
conducted by medical defence organisations
procedure
rather than individual health services,
10 Acute myocardial infarction, cerebrovascular
important clinical themes in specific clinical
accident, or pulmonary embolus during or after an
specialities become apparent. Key lessons from
invasive procedure
this information can be used to direct quality
11 Neurological deficit at discharge
and safety programs in health services and
12 Death
prevent recurrence of similar events. At the 13 Temperature higher than 38.3°C on day before or
individual health service level, all claims made day of discharge
against the service should be analysed as part 14 Cardiac or respiratory arrest
of the service's quality improvement and 15 Five-minute Apgar score < 6. or complication of
patient safety program, but in most services, abortion or labour and delivery
such analysis will only be a very small pan of 16 Other undesirable outcome
their overall program.5 17 Indication of litigation in the medical record
Because of the limitations of direct 18 Length of hospital stay above 90th percentile for
diagnosis-related group in patients aged under 70
observation and legal claims, most adverse
years, and 95th percentile in those aged 70 years or
event detection programs in health services use older.
medical record review or clinical incident
reporting, or both, to detect adverse events.
Box 4.1 Harvard Medical Practice Study screening
criteria for advene events'
1 Hospitalisation within previous year for patients
younger than 65 years old, and previous 6 months for
older patients
2 Admission to any hospital after current discharge

4.2.3 Medical record review

It is not practical to continuously review the
medical records of all the patients admitted to
hospital. The volume of records requiring
review would be high, and most records would
not contain any evidence of adverse events.
Therefore, a more efficient approach is
required. How can most significant adverse
events recorded in patient medical records be
found without having to review all or a large
number of medical
records? Medical records with a higher-than-average probability of containing an adverse event need
to be identified.
 4 Detecting advert* events and rta*

4.2.3.1 Inpatient screening criteria

Fonunately, several studies have developed
screening criteria to help identify these "high-
risk" records The largest study, the Harvard
Medical Practice Study,7 developed and used 18
screening criteria (Box 4.1) Using these criteria
to screen medical records in a sample group of
hospitals, the Harvard Medical Practice Study
found most adverse events that occurred
during the period under review Most
importantly, it is not necessary to find all
adverse events documented in a hospitals
medical records for the clinical risk
management program in an individual health
service to be effective Finding all
Enhancing patient care

■ assessing the risk of individual patients experiencing common adverse events and taking
appropriate preventive action
■ encouraging patients to take an active role in the care they are receiving at the health service
■ using preventive health strategies to reduce the prevalence of common and preventable medical
conditions, thereby reducing the need for hospital admissions.
Wc will first discuss each of the methods available to individual health services to detect adverse
events that have occurred in their facilities. We will then outline how individual health services can
detect areas of risk in their organisation using information about adverse events that have occurred
outside the service, and discuss other proactive methods of risk reduction.

4.2 Reactive risk management — detecting adverse events

4.2.1 Direct observation
Directly observing an adverse event has the advantage of the event being detected prospectively, and
studies suggest that most adverse events, especially medication errors, are detected using this
method.1 Direct observation has been used for relatively short periods of time in operating theatres,
and has detected process variations that had the potential to result in adverse events. However, most
of these variations were recognised by theatre staff and did not progress to adverse events. Direct
observation in operating theatres identified communication breakdown, information loss, multiple
competing tasks and high workload as areas for patient safety initiatives. 2 Observing clinicians
discussing adverse events during regularly scheduled daytime education and patient care meetings
found these discussions to be a useful source of information about adverse events. 1 Directly
observing nurses found two types of failures in the care delivery process: problems (a disruption to
being able to execute a task because something that was needed, such as supplies, medication or
information, was unavailable), and errors. Of the failures, 86% were problems. 3
Although short observation periods in clinical areas may yield useful information to improve
patient safety, complete 24-hour direct observation of patients and staff would be very costly. The
routine use of direct observation to detect adverse events is not financially or logistically feasible, nor
would it be desirable to have all hospital staff members and patients observed constantly.

4.2.2 Legal claims

Legal claims related to adverse patient events occur infrequently compared with the number of
admissions and outpatient attendances in an individual health service, and are a small proportion of
the adverse events that occur in acute hospitals. In the Harvard Medical Practice Study, more than
two- thirds of adverse events detected were judged to be preventable, and half of these were due to
negligence. However, less than 2% of patients with injuries due to negligence took legal action. 4
Therefore, legal claims are not a large source of information about adverse events, and when such
claims are made, they often concern an adverse event already detected by the health service using
other methods
Claims are a small, unrepresentative sample of the adverse events thai occur in an individual
health service. The claims made can be biased to adverse events that led to severe injuries, and there
are often long delays between the event occurring and the claim being made. Sometimes the claims
relate to injuries that were not caused by medical management. 5 Claims are more likely to be made if
patients feel they have been treated with disrespect or if clinicians have not been open and
forthcoming about what has happened, whether or not negligence has occurred. Overall, medical
errors are common, but legal claims are rare. 6

50
 4 Detecting adverse event* ana rm

Despite these limitations, claims can still 8 Return to operating theatre during current period of
provide useful information about rare events hospitalisation
that occur in health services, and often provide 9 Treatment for organ damage after an invasive
a very detailed (albeit legally based) analysis of procedure
an individual adverse event. Claims can detect 10 Acute myocardial infarction, cerebrovascular
latent factors that were present when the accident, or pulmonary embolus during or after an
adverse event occurred and deficiencies in invasive procedure

processes of care that were delivered. With the 11 Neurological deficit at discharge
12 Death
analysis of large samples of claims usually
13 Temperature higher than 38.3°C on day before or
conducted by medical defence organisations
day of discharge
rather than individual health services,
14 Cardiac or respiratory arrest

important clinical themes in specific clinical 15 Five-minute Apgar score < 6, or complication of
specialities become apparent. Key lessons from abortion or labour and delivery
this information can be used to direct quality 16 Other undesirable outcome
and safety programs in health services and 17 Indication of litigation in the medical record
18 Length of hospital stay above 90th percentile for
prevent recurrence of similar events. At the
diagnosis-related group in patients aged under 70
individual health service level, all claims made
years, and 95th percentile in those aged 70 years or
against the service should be analysed as part older.
of the service's quality improvement and
patient safety program, but in most services, 4.2.3.1 Inpatient screening criteria
such analysis will only be a very small part of Fortunately, several studies have developed
their overall program.5 screening criteria to help identify these "high-
Because of the limitations of direct risk" records. The largest study, the Harvard
observation and legal claims, most adverse Medical Practice Study,7 developed and used
event detection programs in health services use 18 screening criteria (Box 4.1).
medical record review or clinical incident Using these criteria to screen medical records
reporting, or both, to delect adverse events. in a sample group of hospitals, the Harvard
Box 4.1 Harvard Medical Practice Study screening Medical Practice Study found most adverse
criteria for advene event* 7
events that occurred during the period under
1 Hospitalisation within previous year for patients review Most importantly, it is not necessary to
younger than 65 years old, and previous 6 months tor
find aII adverse events documented in a
older patients
hospitals medical records for the clinical risk
2 Admission to any hospital after current discharge management program in an individual health
3 Previous failure of medical management or
service to be effective Finding all
unfavourable results
4 Trauma incurred in hospital
5 Unfavourable drug reaction in hospital
6 Transfer from general care to a special care unit
7 Transfer to another acute^are hospital

4.2.3 Medical record review Si

It is not practical to continuously review the
medical records of all the patients admitted to
hospital. The volume of records requiring
review would be high, and most records would
not contain any evidence of adverse events.
Therefore, a more efficient approach is
required. How can most significant adverse
events recorded in patient medical records be
found without having 10 review all or a large
number of medical
records? Medical records with a higher-than-average probability of containing an adverse event need
to be identified.
Enhancing patient care

documented adverse events is important for The criteria chosen are broad, and some
epidemiological studies to determine overall hospitals, especially those providing
adverse event rates in health services in a predominantly specialised services such as
region or country. However, at the individual obstetrics or specialised surgery, will benefit
health service level, continuously detecting all from modifying these screening criteria and
adverse events, and analysing and responding developing and using additional screening
to them all, would be an overwhelming, criteria specific to their clinical areas. For
demotivating task. At Wimmera Health Care example, the rate of adverse events in
Group, our aim was to find most major adverse obstetrics is lower than in other clinical areas
events that occurred, and to provide a in hospitals, although important quality issues
continuous flow of information about these are still present. Medical record review using
events to give direction and a realistic, general patient outcome screening criteria will
practical workload to our clinical quality be ineffective in monitoring and guiding quality
improvement and risk management activities. and safety programs in obstetrics. However, a
Intuitively, if a serious adverse event occurs large number of specific obstetric screening
to a patient in a hospital they may: criteria have been developed to overcome
■ die this.10
■ have an increased length of stay in hospital In addition to adverse events detected using
■ be transferred to the hospitals intensive care these nine general patient outcome criteria,
unit there are clerical diagnostic codes that can be
■ be re-admitted to the hospital after a recent used to detect adverse events. All admissions
discharge to Australian hospitals have the patient's
■ be taken back to the operating theatre after diagnoses coded using the International
surgery classification of diseases, 10th revision,
■ be transferred to a larger hospital for Australian Modification (1CD-10-AM) after the
management of complications arising patient is discharged from the health service.
from the adverse event. Other patient Adverse events, such as injuries and
outcomes may result from the adverse complications that have arisen from care
event having occurred, but these general provided to patients, can be coded using this
patient classification.
outcomes are an effective place to begin looking Unfortunately, these codes do not
for adverse events. differentiate between adverse events that
T occurred outside the hospital and required the
he nine general patient outcome criteria used patient's admission to hospital for treatment
at Wimmera Health Care Group (Box 4.2) were and adverse events that occurred during the
developed from a more comprehensive list of patient's admission. Therefore, an additional
screening criteria initially used in the "C" (complication) prefix is added as an
Californian Insurance Feasibility Study 8 and incidence flag to the code in Victoria and
modified by Medical Management Analysis 9 Queensland to indicate that the complication
(Box 4.3). occurred and was treated during the
admission. However, this prefix does not
capture complications that are detected
Box 4.2 Inpatient screening criteria used at
Wimmera Health Care Group after the'patient has been discharged from
hospital.
1 Death
1 Admission or adverse results of
2 Return to operating theatre within 7 days for complications outpatient management
3 Transfer from general ward to intensive care unit
4 Unplanned re-admission within 28 days of discharge
5 Cardiac arrest
6 Transfer to another acute care facility
7 Length of stay greater than 21 days 0 Booked for
theatre and cancelled
9 Any medical record recommended for review.
 Box 4.3 Medical Management Analysis general outcome screening
criteria*

2 Admission for ic not addressed by

complications or ti physician
incomplete 15 Neurological deficit
s
management of present at discharge
problems on previous s that was not present
hospitalisation u on admission
3 Operative consent: e 16 Transfer to another
acute care facility
a Incomplete c 17 Death
b Missing before N 18 Subsequent visit to
procedure c Different o emergency
ti department or
procedure d Different outpatient department
surgeon e Not signed s
for complication or
I Risks not explained s adverse results
g Other u related to this
4 Unplanned e hospitalisation
19 Length of stay
removal, injury, or re greater than___________________
repair of organ or m or allotted
structure during days:
o
surgery, invasive
procedure or v a Complication
vaginal delivery e b Diagnostic or
5 Unplanned return to d treatment problem c
operating theatre on d Placement problem
this admission O 20 Medical records
6 Invasive procedure or
pathology: th review — physician
a Pathology report er 21 Medical records
does not match pre- 7 Transfusion: review — nursing
operative 22 Departmental or
a Iatrogenic bleeding or
d other problem(s) (eg.
anaemia b Not clinically
i admitting, emergency
indicated c Transfusion department,
a
reaction laboratory, x-ray)
g
8 Nosocomial infection 23 Patient or family
n dissatisfaction.
9 Antibiotic or drug
o
utilisation
s
10 Cardiac or respiratory
i
arrest: low Apgar score
s
11 Transfer from general
b care to special care unit:
a Complication
N J> Utilisation problem
o 12 Other patient
n complication(s)
- 13 Hospital-incurred
d patient incident a Fall
i b Intravenous problem c
a Medication error d Skin
g problem e Other
n
o «
14 Abnormal laboratory,
s
x-ray, or other test results
t
Enhancing patient care

Despite this limitation, these "external cause

codes" can be used as a clerical screen to
detect adverse events in addition to the nine
general patient outcome criteria listed above.
Importantly, the successful use of ICD-10-AM
codes to detect adverse events is dependent on
the codes being correctly assigned, and the
pre-existing diagnoses and the diagnoses made
during the admission being correctly identified
by medical record department coding staff."
The ninth screening criterion, "any medical
record recommended for review", although not
a clerical screening criterion, provides a useful
and important method to detect additional
adverse events. Using this criterion, any
member of the clinical staff (medical, nursing
or allied health) can refer the medical record of
any patient about whom they have concerns
about their care to the risk management
program for review in the same manner as
medical records screened positive for any of the
other eight criteria. Using this criterion allows
medical records to undergo medical review
even if they do not meet any of the first eight
general patient outcome criteria. Not all
adverse events will be detected using only
these eight general criteria. Therefore, having a
broad additional criterion that facilitates
medical review of patient records not meeting
any of the limited general patient outcome
screening criteria allows additional adverse
events to be detected.
 4 Detecting advert* events and rta*

4212 The screening process

Importantly, clerical rather than clinical staff can use these criteria to screen medical records. This
reduces screening costs, so that clinical staff can concentrate on reviewing medical records that have
been screened positive, rather than reviewing the medical records of all patients admitted to the
hospital or a randomly selected sample of these records. Screening and review of medical records can
occur retrospectively or concurrently with the patient^ admission. However, in practice it is easier to
undertake medical record screening after the patient has been discharged.
At Wimmera Health Care Group, health information services staff screen all inpatient medical
records after patient discharge as part of their normal duties. Medical records are screened at three
points in the health information services department:
■ at assembly of the record after patient discharge
■ when the discharge summary is typed
■ during coding of the admission.
This triple check results in few records being screened incorrectly, either as falsely positive or
falsely negative. In our experience, about 10% of the inpatient medical records screened in a medium-
sized general hospital will be positive for one or more of the nine general patient outcome screening
criteria.12
At Wimmera Health Care Group from 2000 to 2007, 9.4% of admissions screened positive for one
or more criterion. Of these, 40% were unplanned re-admission within 28 days and 40% were transfer
to another acute care facility.
However, the proportion of medical records that screen positive will depend on the size and nature
of hospital using the methodology and the number of screening criteria used. For example, the Royal
Children's Hospital (a quaternary paediatric hospital in Melbourne, Victoria) used three screening
criteria (death, unexpected admission to the intensive care unit and clinician referral for record
review), and found that 1.6% of all admissions to the hospital over a 6-year period screened positive
for one or more of these criteria.13

4.2.3.3 Which screening criteria to use

Health services considering introducing a medical record review program often ask us which of the
nine screening criteria we recommend. We have used the same screening criteria since the
commencement of our program in 1989 to observe trends over time. We strongly believe that the
medical records of all patient deaths that occur in a hospital should be reviewed. Some hospitals
using our program have changed the "patient death" screening criterion to "unexpected patient death"
to reduce the number of records screened positive but are found on medical review not to contain an
adverse event. In our experience, many patients who die in hospital are receiving palliative care.
Therefore, changing the criterion to unexpected deaths may be useful to reduce the number of patient
deaths requiring medical review, but the records of all patient deaths would still need to be screened
by a clinician to eliminate deaths that were expected. Voluntary, confidential review of 90% of deaths
of surgical patients over 9 years in Scotland, followed by feedback to the surgeons and anaesthetists
involved and collective annual reporting has been associated with a reduction in surgical mortality
and associated adverse events. 14
The criterion of admission to the intensive care unit from a general ward is the most effective and
efficient screening criterion, and we recommend its use in any medical record review program. In a
review of 100 consecutive adult emergency admissions to two intensive care units in England, 54
were assessed as having received suboptimal care before their admission to the unit, and two-thirds
of these were transferred late.15 It is often the case that seriously ill patients who would benefit from
intensive treatment are in general wards 16 and although early warning programs have been developed
to detect ward patients who are deteriorating, 1718 the criterion of transfer to the intensive care unit
Enhancing patient care

from the general ward still yields useful information about the quality and safety of the care provided
on general wards.
 4 Detecting advert* events and rta*

We also recommend using the criterion of clinical staff being able to refer any records for review
about which they have concerns. In our experience, this criterion is used sparingly but raises
important usue* The criterion of cardiac arrest has not been very useful. Inpatients who have a
cardiac arrest either die and are reviewed under patient deaths, or survive and are transferred from
the general ward to intensive care (unless they are already in intensive care), so their medical
record is reviewed under this criterion.
Patients booked for theatre and cancelled was a very useful screening criterion when we
commenced our program. However, since the establishment of a preadmission clinic, most patient
records that screened positive under this criterion have been found on medical review not to
contain an adverse event. Most of these cancelled theatre cases are patients booked for elective
surgery who develop respiratory tract infections or whose clinical condition resolves.
We have reduced the length-of-stay criterion from 28 days to 21 days, in line with the general
reduction in hospital length of stay in Victoria and elsewhere. Transfer to another acute care facility
is a useful criterion for small and medium-sized hospitals that often transfer patients with complex
clinical problems to larger centres. However, this screening criterion is not useful for tertiary centres
that infrequently transfer patients to other health services because of the complexity of their
management.

4.2.3.4 Who should screen medical records?

Instead of using medical records department stafT, nursing stall can screen medical records. Nurse
screening has two advantages. Firstly, additional screening criteria, such as those used in the
Harvard Medical Practice Study, can be used — including those that require some clinical judgement.
Using more screening criteria increases the number of adverse events detected. Secondly, nursing
staff can screen medical records while the patient is in hospital, rather than after discharge. This
prospective screening can result in adverse events being detected and acted upon more quickly and,
in some cases, can thereby limit the harm that occurs to patients from adverse events. A major
disadvantage of having nurses screen medical records is the increased cost. Nurses working on a
ward are unlikely to be able to screen the medical records of all patients on the ward in addition to
their normal duties. Employment of additional nurses would be necessary to screen medical records,
and would therefore increase screening costs.

4.2.3.5 Emergency department screening criteria

Adverse events do not only occur to patients admitted to hospital, and the two-step screening and
medical review process to detect adverse events can be used in other clinical areas of a health service,
such as the emergency department. At Wimmera Health Care Group, emergency department medical
records are screened using slightly different criteria to those used for inpatient screening (Box 4.4). A
specific report within our administrative database details those records that meet these criteria.
At Wimmera Health Care Group, 3.3% of emergency department attendances screened positive for
one or more of these criteria between 2000 and 2006. Of these, 36% were attendances within 28 days
of discharge and 21% were unplanned representation within 48 hours.

4.2.3.6 Medical review

Regardless of the method used to screen medical records, when a record is screened positive, it
requires medical review to determine whether an adverse event occurred Designated senior medical
staff usually
Enhancing patient care

perform this review. In a general hospital, a cross-section of medical staff, including at least
surgeons, physicians and general practitioners, are required to review medical records. These
medical reviewers should preferably be nominated by the hospital's medical staff and have a strong
commitment to quality improvement and risk management. Medical records that screen positive can
be allocated randomly to these nominated doctors for review, or each doctor can be allocated all the
records that screen positive for one or two of the screening criteria. Obviously, doctors should not
review a patient's record if they admitted the patient into hospital, or if they were significantly
involved with the patient^ treatment.
At medical review, the doctor reviews an allocated medical record and completes an adverse event
analysis form. The Wimmera Health Care Group adverse event analysis form is shown in Appendix 1.
This form contains patient details and scales to help determine whether an adverse event has
occurred, and if so, to assess its severity, preventability, and the factors that have contributed to it
occurring. Hospitals should design an adverse event analysis form to meet their own needs.
1. Defining adverse events. First, a definition of what constitutes an adverse patient event should be
chosen. Examples of definitions used in two large studies are:
■ An untoward patient event, which under optimal conditions is not a natural consequence of the
patient's disease or treatment (from Medical Management Analysis). 9
■ An unintended injury that results in temporary or permanent disability, including increased length
of stay and/or financial loss, which is caused by health care management rather than the disease
process (from the Quality in Australian Health Care Study).19
We found the Medical Management Analysis definition more useful as it allows a broad range of
patient events to be classified as adverse.
2. Determining whether an adverse event has occurred. Next, a scale to help medical reviewers
determine using implicit judgement whether an adverse event has occurred should be developed or
chosen. The six-point scale (Box 4.5) was used in the Harvard Medical Practice Study 4 and the
Quality in Australian Health Care Study 19 to help the medical reviewers determine whether medical
management, rather than the disease process, caused an adverse event.
When assessing the quality of care provided to a patient, implicit or explicit judgement can be used.
When using implicit judgement, prior standards of care that reflect good quality care are not set.
Medical reviewers using implicit judgement attempt to answer the questions: was the process of care
provided to the patient adequate?; could better care have improved the patient's outcome?; and, was
the overall quality of care acceptable? Alternatively, when using explicit judgement, process measures
thai represent good-quality care are determined before the judgement being made by the medical
reviewer.20 As discussed previously, explicit judgement is used in clinical audits of care received by
patients with particular clinical conditions. When medical record review is used to detect adverse
events, records of patients with a wide range of medical conditions are reviewed. It is therefore not
feasible to determine explicit criteria for good-quality care for all clinical conditions. To overcome this
problem, medical reviewers use implicit judgement to determine whether an adverse event was
present in the medical record.
In the Harvard Medical Practice Study and the Quality in Australian Health Care Study, a score of
four or more on the management causation scale was considered to be an adverse patient event. 419
This scale is easy to'use and was shown in the Harvard Medical Practice Study to have a moderate
level of reliability between reviewers 4

Box 4.5 Management causation scale

1 = Little or no evidence
2 = Slight evidence
3 = Not likely (less than 50: 50 odds but a close call)
4 = More likely than not (greater than 50: 50 odds but a
close call)
5 = Strong evidence
6 = Virtually certain.
 4 Detecting advert* events and rta*

Recent assessments of the use of implicit judgements to determine whether a medical record
contains an adverse event have shown moderate to poor inter-rater reliability (ie, the degree u>
which independent medical reviewers agree that a medical record contains an adverse event).
However, medical reviewers disagreed about errors and preventability more than about recognising
an advene event.21 Medical reviewers' judgements may also be affected by hindsight bias, 22 and by
their attempts to reconstruct what had occurred during a patient's admission while having access, on
occasion, to incomplete information.2
Increasing the number of doctors who review each medical record can increase the reliability
between medical reviewers, but if more doctors were routinely involved in their health serviced risk
management programs, this would add considerably to the cost of the program. 23 Inter-rater reliability
is imponant in research to determine the rates of adverse events. However, it is less important for risk
management programs in individual health services. *
In our experience, most major adverse events are easily recognised by medical stafT. In the program
at Wimmera Base Hospital, there has been occasional disagreement between medical reviewers about
whether some minor events were adverse, but in practice, these differences in opinion were usually
not significant. The main aim of the screening and review process is not to find every adverse event,
but to detect most major events and, where appropriate, to take action to prevent their recurrence.
Significantly, when using a limited number of general patient outcome screening criteria to detect
adverse events in medical records, there is a high false-positive rate (ie, records-screening positive for
one or more of the criteria but not containing an adverse event). In our program in a medium-sized
general hospital, about 10% of medical records screen positive for one or more of the nine screening
criteria, but only one in 10 of these records has been found to contain an adverse event on medical
review. 3. Rating the severity and preventability of an adverse event. The adverse event analysis form
used at Wimmera Health Care Group also contains scales for the medical reviewers to rate the
severity and preveniability of the adverse events they delect, and checklists to help document the
factors they consider contributed to the event. This information aids the analysis of each event, and
collecting these data over time allows trends to be observed. At Wimmera Health Care Group we used
the severity scale developed by Medical Management Analysis 9 (Box 4.6).
Another useful severity scale that categorises patient harm is adapted from the National Coordinating
Council for Medication Error Reporting and Prevention taxonomy of adverse medication events
(www.nccmerp.org/aboutMedErrors.html), which has been applied to adverse events (Box 4.7). 24

Minor adverse severity score 0-2

event > Mapr severity score 3-6
adverse event ■

Box 4.6 Medical Management Analysts severity

scale9
1 = Minor severity (no disability; no significant resultant
discomfort; no cosmetic or functional impairment; and
no increased length of stay as a result of adveree
event).
2 = Minor temporary (minimal to moderate clinical
effect requiring no or minimal clinical intervention: or no
increased length of stay or rehospitalisation for same or
related problem).
3 = Minor permanent (minimal to moderate clinical
effect with permanent residual effect, without significant
functional or cosmetic impairment).
4 = Major temporary (moderate to severe clinical effect
with no significant functional or cosmetic residual effect.
This usually results in increased length of stay or
rehospitalisation and requires moderate to maior
clinical intervention).
5 = Major permanent (moderate to severe clinical effect
with significant functional or cosmetic residual)
6 = Potential major or major contributing (when doubt
exists as to the outcome but the probability is that a
major impairment or repeated hospitalisation will be
necessary. The outcome may result in major
impairment).
$ = Death.
Enhancing patient care

The Agency for Healthcare Research and Quality in the United States defines a preventable adverse
event as "harm that could be avoided through reasonable planning or proper execution of an
action".25 The preventability of adverse events can be assessed using a scale used in the Harvard
Medical Practice Study4 (Box 4.8).
When the medical review and adverse event analysis form is completed, the medical reviewer
returns the medical record and form to health information services. Data from the analysis form are
entered into a database, and if an adverse event has been detected, the form is sent to the clinical
risk manager. The medical reviewer keeps the details of the adverse event for discussion at the next
surveillance committee meeting. This committee reviews each significant adverse event and
determines what actions are necessary to reduce the probability of a particular event recur ring. The
operations of the surveillance committee will be discussed in detail in Chapter 5.
The two-step screening and medical review process should be undertaken in a timely manner in
relation to the patient's discharge, as an urgent response to an adverse event is sometimes required.
If detailed analysis of an adverse event is required, it is preferable that this analysis be undertaken
promptly, while rotating junior medical staff are still working at the hospital and staff involved in
providing care to the patient can clearly remember the patient and the circumstances surrounding
the event. Ideally, the screening of a patient's medical record should be completed 1-2 days after
discharge and medical review completed 2-4 weeks after screening the record.

4.2.3.7 Medical record review form modifications

Although the medical record review form used in the Harvard study was the model used for
developing the Wimmera medical record review, and has been used in epidemiological studies of
adverse events in hospitals in several countries, it has limitations. Modifications have therefore been
made to make the medical review process and the form easier to complete: 26
■ The form has been divided into five modules — patient information and background, the disability
caused by the adverse event, when during the hospital admission the event occurred, the principal
problem (ie, the most important problem in the delivery of care) and the contributory factors and
preventability of the event.
■ The definitions of significant terms have been included to save reviewers referring to other
documents.
■ A shdn accompanying instruction document has been developed.
Box 4.7 National Coordinating Coundl for
Mediation Error Reporting and Prevention
severity scale
CMagory E: contributed to temporary harm to the
patient and required intervention
Crtayuiy F: contributed to temporary harm to the patient
and required initial or prolonged hospitalisation
Category G: contributed to permanent patient harm
Category H: required intervention to sustain life
Category I: contributed to the patient's death.

Box 4.8 Preventability scale

1 = Little or no evidence for preventability
2 = Slight or modest evidence for preventability
3 s Preventability not quite likely — less than 50: 50
but a close call
4 = Preventability not quite likely — more than 50: 50
but a close call
5 = Strong evidence for preventability
6 = Virtually certain evidence for preventability.
The score can be grouped into three categories: No

preventability (score = 1)
Low preventability (score = 2-3) High preventability
(score = 4-6).
 4 Detecting advert* events and rta*

a A systematic approach to determining the contributory factors of the adverse event has been
included. These contributory factors will be discussed in Chapter 5. Often, the contributory
factors are best determined by interviewing the clinicians who provided care for the patient,
although it was fek that if the reviewer was familiar with the work environment in which the
adverse event occurred, he or she would be able to determine the major contributory factors
without conducting interviews. The authors note that training reviewers including the review of
trial cases was essential. They also commented that medical record review had not been used to its
full potential, and it could have an important role in routine review at the local level. The modified
form comprises 29 pages, although tick boxes are used extensively and not all pages need to be
completed for each case review. The form is an excellent resource that allows individual health
services to determine which modules they wish to use and modify to suit their local needs and
conditions and in particular the resources they have available.26 The Modular Review Form 2 is
available at the Quality and Safety in Health Care website
(www.qshc.com/cgi/data/12/6/411/DC2/1).

4.2.3.8 The limitations of medical record review Medical record review has three main limitations.
1. It is dependent on the quality of the contents of the patient's medical record 27 In many hospitals,
most documentation in the medical record is handwritten, and some parts of the record may be
difficult to read. Entries may be of poor quality with incomplete, inaccurate, confusing and conflicting
data.28 Sometimes the details of the nature and effect of an adverse event may not be recorded in the
medical record.26
Importantly, the effective screening of discharge summaries requires that a summary is available.
In one study, a third of patients discharged from hospital had not had a discharge summary
generated. Also, the summary's contents need to be of high quality. 29
Each patient admission to Wimmera Health Care Group has a discharge summary dictated and
typed shortly after discharge. These typed summaries greatly assist the medical staff when they
review medical records that screen positive. A prompt, accurate and typed discharge summary also
improves communication between hospitals and referring doctors.
To help junior medical staff prepare high-quality patient discharge summaries, we developed an
electronic discharge summary template. The template comprises headings of the clinical information
required in the summary and unlimited space for free text under each heading. The template can be
accessed when the patient is admitted to the hospital and added to as required during the admission
by junior medical stalf using computers located in most clinical areas. Alternatively, junior staff may
dictate or handwrite discharge summaries if they prefer; these summaries are then promptly typed by
health information service staff. To assist the medical record review process, some hospitals request
that their junior doctors indicate on the discharge summary documentation whether an adverse
event has occurred during the patient's admission.
A comprehensive, typed discharge summary is an essential clinical risk management strategy and
it also helps the medical review process detect adverse events. We encourage hospitals that do not
already have a typed discharge summary in place to consider its benefits for clinical communication
and detection of adverse events.
2. It requires additional resources to implement a continuous and ongoing program. Medical record
review has been described as "a cumbersome and costly method of identifying adverse events" 22
However, in the program used at Wimmera Health Care Group, using a limited number of general
patient outcome screening criteria that can be identified by clcrical rather than clinical staff, and
having typed discharge summaries, the overall cost of the program was modest. 12 Using fewer
screening critcna can reduce the cost of implementing and maintaining a medical record review
program With fewer
Enhancing patient care

criteria, fewer medical records will screen positive and require medical review, which is the most
expensive part in the program. Individual health services need to determine whether they wish to
introduce such a program, and if so, the scale of the program.
3. It has a higK false-positive rate. In other words, many medical records are screened positive for one
or more of the screening criteria but on medical review are found to not contain an adverse event.
Attempts have been made to address this limitation using a combination of electronic and manual
review processes. Some health services have undertaken electronic screening of multiple databases,
including administrative data, ICD codes, keyword searches of discharge summaries and medical
records searching for patient complications, laboratory results and medication prescribing. These
databases have been screened independently or in combination using searching algorithms and
natural language processors to extract information to help detect adverse events, and have met with
varying success.
However, the main problem with electronic screening, as with manual medical record review, is the
high false-positive rate. The rate has been reduced by requiring multiple screening criteria to be met
before a medical record review is required. 30 Such electronic screening will be most effective if a health
service is already using an electronic medical record.
Although medical record review results in review of many medical records that do not contain an
adverse event, some of these records do contain evidence of errors that either caused no harm, or
were intercepted before they caused harm. It has been our experience that when the review of patient
records has not delected adverse events, other valuable information has frequently been found,
highlighting system weaknesses in the hospital. Information about errors and near misses, as
opposed 10 information solely about adverse events, provides valuable additional information to help
make health care systems safer.

4.2.3.9 Trigger tools

In further attempts to reduce the high false-positive rate of medical record review, some health
services use triggers to detect adverse events. A trigger is an "occurrence, prompt or flag found on
review of the medical record that 'triggers' further investigation to determine the presence or absence
of an adverse event". 28 Rather than focusing on errors, triggers target events that could be linked to
patient harm.28
Medical record review has been enhanced by using "active surveillance" 31 in which 21 electronic
triggers of specified abnormal laboratory results (above and below predetermined thresholds) or
changes of predetermined magnitude, prescrip-
tion of high-risk medications (eg, warfarin) and medications abruptly stopped or used
as antidotes known to be associated with adverse events (eg, naloxone) are used as
screening criteria to prompt a rapid, focused and real-time medical record review to
detect possible adverse events28 (Box 4.9).
Using electronic triggers, patient medical records triggered as positive can be reviewed
while the patient is still in hospital. Such review allows more information about the
patient's care that may not be recorded in the patient's medical record to be obtained,
and for rapid intervention to potentially occur while the patient is in hospital to
prevent or reduce harm from the adverse event. Using triggers, electronic lists of
patients
 4 Detecting advert* events and rta*

who meet one or more triggers are generated each day and an appropriately trained nurse and
pharmacist review each medical record.
In one study, the triggers of elevated international normalised ratio (INR) and positive blood
cultures had the highest yield for detecting adverse events. Importantly, reviewing the medical
records prompted by a trigger also detects adverse events beyond those related to the trigger. Of the
327 medical records prompted by a trigger for review, 74% contained air adverse event. 51
To reduce costs, trigger tools can also be used regularly with small samples of medical records
audited at set time intervals, rather than continuously screening all patient admissions. 20 Although
much of the work done with trigger tools has used electronic screening, the costs can be reduced by
using a manual screening process and reviewing a small sample of medical records. 28
An advantage of trigger tools is that they can be customised to specific clinical environments. 28 For
example, trigger tools have been developed that are patient age- and medical department-specific,
which is logical, given that the numbers and types of adverse events that occur in hospitals vary with
patient age and the area in the hospital in which they are receiving care.
A trigger tool with 23 triggers has been developed to detect adverse events in intensive care units.
Of 1294 medical records from 13 intensive care units in the US that were triggered and subsequently
medically reviewed, 55% were found to contain one or more adverse events. The most frequent
triggers that screened positive were procedure-associated events, haemoglobin level drop, intubation
or reintubation, pulmonary embolism tests, pneumonia not present on admission, and positive blood
culture.24
A trigger tool with 17 triggers has also been developed for use in neonatal intensive care units. Of
749 medical records from 15 neonatal intensive care units, 554 adverse events were detected, of
which 56% were preventable. The most frequent adverse events detected in neonatal intensive care
units were nosocomial infection, catheter infiltration, abnormal cranial imaging, and unplanned
extubation requiring re-intubation. 32 Trigger tools can also be customised for use in ambulatory care
settings using triggers such as emergency department attendance, admission to hospital, and
prescription of more than five medications. 28 Trigger tools have been used with the standardisation of
protocols for the use of high- risk drugs and a medication reconciliation program, to reduce adverse
drug events by 60% over 4 years in a community hospital in the US.33
The Institute for Healthcare Improvement in the US developed a comprehensive trigger tool in
which 53 triggers are used across six modules — cares, medication, surgical, intensive care, perinatal
and the emergency department. The Institute recommends fonnightly review of a sample of 10
medical records, randomly chosen from all patients discharged from the health service over the
previous 2 weeks, to determine whether a trigger is present. All records are reviewed for triggers in
the care and medication modules, and in the other modules if applicable. A 20-minute limit is set for
the review of each record. Data are collected on the number of adverse events delected per 1000 bed-
days, per 100 admissions or as a percentage of admissions. 34 Using this methodology, 40-50
incidents of harm have been found per 100 admissions. 35
Importantly, trigger tools do not measure the true incidence of adverse events in health services,
but they do improve the detection of such events, thereby providing valuable information to health
services about how to improve the quality and safety of the patient care they provide. 32
A significant limitation of trigger tools is that they may miss adverse events that are not apparent
or that have not occurred by the day of the review. Such events may be detected by retrospective
medical record review. Therefore, using trigger tools and retrospective medical record review together
may be more effective than using one method alone. Further limitations of the trigger tool
methodology are the cost of the staff undertaking the surveillance of medical records, and that the
triggers currently bring used are likely to overestimate adverse drug events and underestimate
adverse events not related in
Enhancing patient care

medications. Balancing these limitations, however, is the significant advantage that interventions
to prevent patient harm were made in 14% of patients whose medical records were triggered and
required review. These interventions mostly related to incorrect prescription of medications and
incorrect documentation of clinical information.

4.2-1/0 Medkol record review — the bottom line

Medical record review detects adverse events not found by other methods, and provides a unique
perspective of the adverse events that occur in hospitals. It ensures that clinical activity in an
individual health service is automatically and continuously reviewed, without the need for prompts
or reminders from management to clinicians to undenake quality improvement and risk management
activities.
In our opinion, health services should use medical record review as an important component of
their clinical quality and safety programs. However, it is important that each health service tailors
the process to their needs and conditions. Such programs should be commenced on a small scale,
perhaps in one unit or clinical area in the service using a limited number of screening criteria, before
being implemented organisation-wide. It is less costly to use medical records staff to screen records
using clerical screening criteria than it is to use nursing staff, even if clerical screening is conducted
after the patient's discharge and limits the criteria that can be used.
Wimmera Health Care Group has run medical review programs for inpatients for 19 years and in
the emergency department for 11 years. Not all the adverse events that have occurred in the hospital
over these periods have been delected by the programs, as only a limited number of screening criteria
have been used. However, the number and types of screening criteria and three of the four reviewers
have remained constant, allowing the rate of adverse events to be meaningfully compared over a long
period, and the effects of the actions assessed. A reduction in the rate of adverse events has achieved
by both programs (Figure 4.1 and Figure 4.2).
We believe that the medical records of all patients who die in hospital should be reviewed, although
most will probably not contain an adverse event. The criterion of admission to the intensive care unit

Figure 4.1 Wimmera Health Care Group Inpatient medical record review adverse event rate (percentage of
separations)
16%
14% ■

cfO- ^ jfi ^ J? ^ # ^^&

from a
general yyyyyyyyyyyyyyyy
ward is the most effective and efficient screening criterion that we use, and we
 Figure 4.2 Wimmera Health Care Group emergency medical record review adverse event rate (percentage
of attendances)
1.6* -|
1.4% - 1.2* -

§ io%-

1 0.8% -
j 0.6% -
o

<

0.4% - 0.2% -

0% -I---------1----------1---------1---------1----------1---------1---------r----------1----------1
^^^^ J* ^ &
^#r& ^#
Year
recommend its use in any medical record review program. We also recommend using the criterion of
clinical staff being able to refer the records of any patient about whom they have concerns about their
care for review.
Once established, the medical record review program can be supplemented with a few relevant
triggers that can be easily screened manually (eg, screening 20 medical records each month, which may
detect eight to 10 adverse events) or complication codes that are created when coding records. The
doctors performing the medical reviews should use a simple analysis form consisting of several simple
scales. Ensuring that the health service's discharge summaries are high-quality also supports the
review process.

4.2.4 Clinical incident reporting

The details of many adverse events and most (if not all) near misses that occur in a hospital are not
recorded in patients' medical records, and therefore require other methods of detection. Another
important source of information about adverse events is for clinical incidents to be voluntarily reponed
by health service stafT.

4.2.4.1 Definition of a clinical incident and near miss

The Australian Incident Monitoring System (AIMS) defines a clinical incident as "any event that has
caused harm or has the potential to harm a patient, visitor or stafT member, or any event which
involves malfunction, damage or loss of equipment or property and any event which might lead to a
complaint".1'1 A near miss is "an occurrence of an error that did not result in harm". 17
The definition of a clinical incident is much wider than the definitions of adverse events discussed in
the section on medical record review, which require patient harm to have occurred as the result of
medical management. Clinical incidents include events where there is potential of harm occumng, such
as near misses — where an error occurs, but is delected before the patient is afTectcd. Many hospitals
already have an incident reporting system in place to capture information about incidents (clinical and
non-clinical) and report the details to their insurers as a requirement of their insurance policies
 Relationship of dMcof incidents to adverse events
Chnkat incidents in which a patient is harmed while receiving care are regarded as adverse events.
Chnical incident reporting is the best method of detecting potential adverse events and near misses, 30
and provides valuable information about the risks experienced by patients while they are receiving
care. A study of industrial accidents in the 1940s estimated that for every major adverse event, there
were 29 minor events and 300 near misses. 38 Therefore, waiting for adverse events to occur, rather
than reporting and acting upon near misses, overlooks many opportunities to improve patient safety
and prevent adverse events that could have been avoided if action had been taken in response to a
related near miss.
In our experience, adverse events detected by clinical incident reporting have little overlap with
adverse events detected by screening and reviewing medical records; 39 this finding has been supported
by other studies.40 Further, the proportion of adverse events occurring in a health service that are
detected using clinical incident reporting alone is low. 30,41

4.2.4.3 What incidents are reported and who reports them

The incidents reported in hospital-based reporting systems fall into three broad categories: unexpected
adverse outcomes, procedural breakdowns and catastrophic events. 21 The most common clinical
incidents reported by stafT are patient falls, medication errors and equipment failure. Unfortunately,
there is limited reporting by stafT of issues arising from clinical treatment such as misdiagnosis or
delayed therapy.13 ''0
In our experience, most incident reports in hospitals are written by nursing staff; medical staff
complete few forms. The low reporting of incidents by medical staff has been found in other studies, 42
although in the United Kingdom, doctors provide about 10% of reports of clinical incidents. 40 The AIMS
found that only 2% of clinical incidents reported to it by health services between 1998 and 2002 were
submitted by doctors. 43 However, the incidents reported by doctors tended to have more serious patient
outcomes.44
A survey of 12 AIMS users found that clinical incident reporting had resulted in significant changes
being made in several areas, including how equipment was used, medication was prescribed and
administered and patients' risk of falling while they were hospital was assessed, although changes in
patient outcomes as a result of such changes were difficult to measure. 44 But some authors feel that
"the potential of incident reporting has not been realised in health care, primarily because incident
reporting tends to be punitive and focused on people rather than systems" 45

4.2.4.4 The characteristics of an ideal incident reporting system

Ideally, an incident reporting system should clearly detail the types of incidents to be reported, and
staff should be educated about which incidents they should report and how to report them. The system
should facilitate easy, prompt reporting by staff, whether harm has occurred to a patient or staff
member, and the incident should be investigated immediately. Information from the investigation
should be used to provide prompt feedback to the staff member who reported the incident, and
appropriate action (eg, system changes rather than punitive action) should be promptly taken to reduce
the chances of the incident recurring. 41 "Systems that result in no feedback or improvements will soon
produce only a trickle of reports." 21
The rationale and operational procedures (ie, what types of incidents should be reported and how
they should be reported) of the incident reporting program should be a major part of staff orientation
and education programs.41 Information from individual and aggregate clinical incidents, and especially
the actions that have been taken to improve safety in response to reports, should be continuously
incorporated into staff education programs and regularly reported to clinicians and hospital manage-
ment. Clinical incidents should also be reported to a national database to identify trends in rlimcal
incidents and low-frequency incidents and their associated risks that may only be able to be identified
on a larger scale than within an individual health service.

4.2.4.5 Limitations of clinical incident reporting

The major limitation of clinical incident reporting is under-reporting, especially by medical staff. Other
limitations include potential bias in the events that ate reported, incomplete details being provided in
reports, and an ambiguous, or not readily available, denominator to determine valid incident or adverse
event rates.46
Under-reporting of clinical incidents. Although clinical incident reporting is the most common method
used by health services to detect adverse events, not alPclinical incidents that occur are reported. 47,46
Extremely large differences have been found between the number of medication errors that were
directly observed in studies and the number of these incidents reported. "Major events may be more
reliably reported, but near misses are likely to be ignored, deferred or forgotten in the pressured
environment of clinical work." 49
Under-reporting is thought to occur because of work demands creating time pressures on staff,
beliefs by staff that the reporting of incidents may result in punitive repercussions (such as legal
action, or negative effects on careers), failure to recognise that a clinical incident has occurred, not
understanding what types of incidents should be reported, 41 cumbersome reporting formats, difficulty
identifying errors of omission, a long lead time between the errors and the negative outcome, and
health services not providing feedback or taking appropriate remedial action. 49
The very low rate of reporting by doctors is related to "fear of blame, retribution, disciplinary and
legal action" in response to mistakes that they or their colleagues have made. 30 Doctors need to be fully
engaged in reporting incidents for the program to be most effective. Therefore, any legitimate concerns
that doctors have about the incident reporting process should be adequately addressed. 51 Other issues
raised by doctors regarding incident reporting include not knowing when or how to report an incident,
reporting being viewed by doctors as a nursing duty, time constraints and lack of feedback after
incidents are reported.
In a study in six hospitals in South Australia, in which an incident reporting system had been in
place for at least 5 years in each hospital, over half the doctors responding to a survey did not know
how to locale an incident reporting form in their hospital or what to do with a completed incident form.
In addition, 40% of consultants or registrars working in these hospitals had never completed an
incident form.30 This study also found that the most frequent barrier to the reporting of clinical
incidents by both doctors and nurses was lack of feedback after incidents were reported.
In another study of 65 surgical trainees working in a National Health Service (NHS) Trust in the UK,
52% were unclear about which incidents should be reported, only 12% of those who had reported an
incident received feedback, and none were informed of changes that were made as a result of the repon
they had made. When asked to give reasons for not reporting incidents, 55% of trainees stated that it
was not part of the doctors' culture. 52 Some organisations believe that not reporting and acting on
safety problems is "an unacceptable breach of duty'' and in the long term, addressing safety problems
in a health service honestly and transparently significantly outweighs any potential risk of legal action
or negative media reports.55
Part of the large difference in the frequency of clinical incidents reported by doctors and nurses is
related to the different cultures between these two groups.
Much of the emphasis in nursing education is on learning practices and following rules.
Physicians arc taught analytic thinking and individual responsibility, rules are much less
important and sometimes regarded as subject to individual veto. 54
In not reporting clinical incidents, doctors were concerned about the consequences from forces external
to the health service such as the coroner and litigation.
 The medical profession has a culture that values privacy, professional autonomy, collegiality, self-
regulation, organisationally opaque systems of accountability and relatively few directives. However,
nurses reponing clinical incidents were more concerned about internal forces within the health service
and how these forces would respond to the error having been made. Nursing practice is heavily
protocol-driven, with strong role models reporting incidents and concern by nurses about covering
themselves by reporting an incident. 43
Reporting of clinical incidents by doctors has been enhanced in anaesthesia using handheld
computers. Anaesthetists electronically record routine data on the cases they undertake, their
successes and failures in undertaking specific procedures and clinical incidents that occur. Incidents,
including near misses, are recorded where and when they occur. All data are transmitted to a secure
database, and data analysis is automatically sent back to the clinician in the form of process control
charts. Information about major adverse events is automatically emailed to the hospital quality
manager.55 An incident was reported for 3.5% of 4441 anaesthetic procedures reported to the database.
In a comparison with medical records review, only one incident was noted in the medical records of a
sample of 208 of these patients. 56 The authors believe similar programs can be developed and
implemented in all medical specialties.
Incomplete and biased reporting. A further limitation of incident reporting is that in those clinical
incidents that are reported the information provided is frequently incomplete, especially regarding
details of the factors that contributed to the incident occurring. 57 Also, the incidents chosen to be
reported may be subject to bias, and are most likely to be reported shortly after they occur — before the
patient's outcome is completely known. 58 With this bias in identifying and reporting incidents, only a
small, non-random proportion of incidents being reported, and the lack of a denominator, meaningful
error rates cannot be determined from incident reporting. 45,46,59 Importantly, unlike medical record
review, incident reporting does not have a readily available denominator and is therefore not useful in
determining the rates of incidents it detects.
Difficulties interpreting data. Interpreting the reasons for trends in the incident reporting rate is also
difficult. Does an increase in the number of incidents reported during a period mean that incidents are
occurring more frequently, or that staff are reporting events more frequently? Does a decrease in the
number of incidents reported mean that fewer incidents are occurring and the health service is
therefore safer, or that staff are reporting less frequently because they are too busy, or the reporting of
incidents rarely results in appropriate change occurring, or punitive action has been taken against a
staff member following the reporting of an incident? 46,59 Unfortunately, "trending of incident reports is a
biased measure of whether safety is improving". 46

4.2.4.6 Addressing the limitations of clinical incident reporting

The limitations of incident reporting are significant, and many steps have been taken to address them.
Previously, reports of clinical incidents were entirely paper-based, but online reporting systems are now
used in many health services. To facilitate incident reporting, the forms or electronic formats used
should be readily accessible and provide a balance between being easy to complete and providing
sufficient information for the clinical risk manager to begin investigating the incident. The more specific
the questions on the incident report form are, the easier it is to complete, but the more limited the
information provided will be. The more general the questions are, the more detail is provided, but the
more time is required to analyse the report. 60
A trial-and-error process may be necessary in individual health services to develop the most useful
form for staff who submit and analyse reports. Some incident reporting programs use several reporting
options, including a one-page incident tepoit form, online reporting and 24 hours a day, 7 days a week
telephone reporting facility to reduce the burden of reporting. 42 Interestingly, despite these repotting
options being available, staff still preferred a paper-based reporting method over online reporting; one-
 fifth of reports were made by telephone 42 The voluntary reporting of clinical incidents, especially near
misses, have also been increased by continuously prompting doctors, nurses and pharmacists to
report incidents.10
Successful implementation and maintenance of an effective incident reporting program require
substantial education of all staff about the incident reporting program (what types of incidents should
be reported and how they can be reported). They also require appropriate changes in the culture of
the health service in how it deals with the information provided in these reports and the staff who
have made reports. Some incident reporting programs have developed manuals explaining its rationale
and the program operation. These manuals have been distributed to clinical staff to increase their
knowledge and address their concerns about incident reporting^ These programs have also organised
education sessions as part of existing departmental meetings and strategically placed posters about
the program in clinical and staff areas to promote reporting. 42 In one program, feedback was enhanced
by providing individual feedback to staff who reported major incidents, aggregate data were presented
at regular department meetings, and a regular newsletter containing statistics, summaries of
investigations and recommendations arising from the analysis of incidents was distributed to all
clinical staff. 42
The clinical incidents reported by staff should be investigated and appropriate action taken in a
timely manner. In one program in the US, all the clinical incidents that were reported were immediately
assessed for the frequency of their occurrence and the severity of their consequences. On the basis of
this assessment, each incident was given one of three ratings. According to this rating, the health
service's patient safety alert policy decreed that investigation of the incident and its resolution was
required to occur within 24 hours, 72 hours or 1 week of the incident being reported. Significantly, as
the staff in the health service in the study realised that the issues contributing to the incidents they
were reporting were being properly addressed, the number of incidents reported increased
substantially.33
Reporting of clinical incidents should be confidential, but should health services allow anonymous
reporting of incidents? The availability of anonymous reporting should, in theory, increase the number
of incidents reported by staff, as there is little chance of blame or retribution occurring after the report
is made, unless the clinician involved in the incident can be identified from the details in the repon. A
disadvantage of anonymous reporting is that the staff who receive the incident repon and are trying to
analyse the event cannot contact the person who submitted the report to obtain more details about the
circumstances surrounding the incident. Further, the information in the repon may be unreliable.
Anonymous reporting also limits the degree to which the important issues of accountability and
transparency can be addressed when dealing with incidents. 61 A further danger of anonymous reporting
is that managers may assume that such reports are being made by disaffected staff, and may therefore
not lake ihe details of such reports seriously.
The small size of some health services and the very small number of staff in highly specialised clinical
areas may make it impossible to anonymously report some incidents. 60 However, given that the details
of a significant incident may not be reported unless it can be done anonymously, without the repon
being made, the health service may remain unaware of a serious system weakness. If the option to
repon an incident anonymously — with the reporter not having to identify themselves and not allowed
to identify others involved in the incident — is available, medical staff may be more motivated to report
a problem.51 On balance, having anonymous incident reporting available is preferable to significant
clinical incidents not being reported.
424,7 The Wimmera Health Can Group approach to clinical incident reporting We have used the AIMS incident
reporting system developed by the Australian Patient Safety Foundation. The incident reponing forms
and computer software program have recently undeigone considerable improvement. In addition, a
web-based system will be available in the near future.
The AIMS program has developed incident reponing forms that are individualised for users, and a
reponing system suitable for use by all clinical and non-clinical areas in a health service. The form
 provides prompts for the collection of information about the incident. A free narrative section provides
the most useful information to assist with the analysis of the incident and significant space is allocated
for this information.62
Ginical incidents are reported locally, either with identifying details or anonymously. Data are
entered locally by staff who have undertaken accredited coder training with the Australian Patient
Safety Foundation. All information entered into the local database is de-identified. Public hospitals in
most Australian states enter clinical incident data into statewide databases. These databases can
identify trends in low-frequency adverse events that may not be obvious to individual hospitals.
The AIMS allows individual health services to generate their own reports of the numbers and types of
incidents reported throughout the health service or in individual clinical areas, and to compare their
data with other like, but de-identified, health services. 62 Although the AIMS is the most frequently used
incident reporting system used by public hospitals in Australia, other reporting systems are available.
To determine the severity of each clinical incident, the Victorian Department of Human Service's
incident severity rating classification is used at Wimmera Health Care Group. This method of
classification is used instead of risk rating for a number of reasons: the person reporting the incident is
able to determine the level of severity, initiating the appropriate response by management at the time of
the incident; incidents are not necessarily risks — risks are detected as a result of the incident
investigation; and to be able to rate risks, knowledge is required about consequence and probability
and this is generally not known at the time of the incident. 63 The Victorian Department of Human
Service's Incident Severity Rating Classifications are:
■ Incident Severity Rating 1. Relatively infrequent, clear-cut events that occur independently of a
patient's condition; commonly reflect hospital system and process deficiencies; and results in or have
the realistic potential to result in, an unexpected death or a permanent disabling injury or
psychological harm to a person and includes reportable sentinel events.
■ Incident Severity Rating 2. Events that result in a temporary loss of function (sensory, motor,
physiological or intellectual) that are unrelated to the natural course of the patient's illness and differ
from the expected outcome of person's management.
■ Incident Severity Rating 3. Events that result in a person requiring increased treatment, but not
hospitalisation or an increased length of stay.
■ Incident Severity Rating 4. Events that result in minor injury requiring only first aid treatment or no
injury.
For clinical incident reporting to be effective, clinical staff require education about:
■ the types of incidents that should be reported
■ the availability of incident reporting forms or electronic formats and how they should be completed
■ how the information that is reponed is used to improve the quality and safety of patient care
provided in the hospital.
Most importantly, stafT members require prompt feedback that the incidents they have reported have
been fully investigated and appropriate changes made to the systems that allowed the events to occur.
If it can be demonstrated to staff members who report incidents that this information is being used to
improve patient care in their hospital, this provides positive reinforcement of the reporting behaviour
and will act as a powerful motivator for staff members to continue to report incidents.
At Wimmera Health Care Group, the low reporting of incidents by doctors is compounded by the
many junior doctors who rotate from Melbourne to Horsham for 10-week periods. In this short time
frame, it is challenging to educate these doctors about which incidents they should report and how
they can be reported. It is especially difficult to reassure them that the culture of the health service is
such that incidents reported will be positively received, and the information used constructively to
improve the quality and safety of patient care. Many of these doctors feel that they will be blamed for
errors if they report incidents. These junior doctors are educated about the incident reporting program
shortly after they arrive at the hospital. However, this education on its own is usually not enough to
encourage them to report incidents.
Additional strategies are required to motivate junior doctors to report the incidents they have
observed or have been involved in. The director of medical services meets with these doctors each week,
and among the administrative matters that are discussed, the director asks them whether there have
been any instances where the care provided to patients has not gone well, or whether they have
experienced or observed any problems in providing care to their patients in the previous week. The
incidents raised are discussed and recorded, promptly investigated and appropriate action taken to
reduce the probability of the incidents recurring. The director then provides feedback to the junior
doctors about the investigations and the actions taken in response at the next meeting with them. Once
an initial reticence with discussing incidents with the medical director (who has line responsibility for
these doctors in the hospital) is overcome and the doctors see they can influence how clinical care is
provided in the hospital, a steady stream of incidents usually flows over the remaining weeks of their
rotation.
Using such an intensive approach to obtaining information about clinical incidents from junior
medical staff is supported by a study that surveyed staff in two tertiary hospitals in the US about their
experiences with adverse events. The study found that junior medical staff commonly encountered and
were associated with errors. Many of these adverse events were not recorded in medical records,
perhaps because of fear of punitive action and litigation. Therefore, obtaining information about clinical
incidents from junior medical staff may provide information that other methods will not find, and which
is complementary to information obtained from other sources. Junior medical staff should, therefore, be
closely involved in quality improvement and safety programs in health services. Such involvement may
also assist in educating these stalf about errors. 64
Engaging senior medical staff to report clinical incidents can be even more challenging. The hospital
has made it clear to senior doctors that it would be pleased to receive information about incidents in
any form they wish to use: by letter, telephone call or stopping the director of medical services in the
corridor to discuss an incident. Medical staff are an important source of information about clinical
incidents that occur in a hospital, and we have tried to make reporting clinical incidents as easy as
possible for them.
Clinical incident reporting is a major component of Wimmera Health Care Group!; clinical risk
management program. We consistently provide staff education about the importance of incident
reporting, feedback to them about the findings of investigations of incidents they have reported, and
timely action to correct weaknesses that are identified. Although we cannot be certain that these
actions have contributed to the increase in the number of incidents reported, we are pleased that staff
have substantially increased the number of reports they are making (Figure 4.3).
All adverse events delected by clinical incident reporting are discussed at surveillance committee
meetings. These adverse events are dealt with in the same manner as events detected using medical
record review. The events are analysed and appropriate action is determined and taken to reduce the
chances of the events recurring. Some adverse events detected by clinical incident reporting may
require immediate action by the clinical risk manager and, if appropriate, analysis and action are
expedited and
 Rgure 4.3 Annual number of Incidents reported at Wimmera Health Care Group

Year

the details discussed at the next surveillance committee meeting. The methods used in analysing
adverse events will be discussed in more detail in Chapter 5.

4.2.4.8 Clinical incident reporting — the bottom line

Each health service should have a clinical incident reporting system as one of the cornerstones of its
quality improvement and risk management program. However, used in isolation, incident reporting
does not provide a comprehensive view of the risks present in a health service. Clinical incident
reporting detects different types of adverse events from those recorded in the patient medical record. It
an important source of information about near misses, which are rarely recorded in the medical
record. Without such reporting, many risks present in the health service will most likely remain
undetected, perhaps until a major adverse event occurs. Incident reported should therefore be used in
conjunction with other more systematic and powerful methods of adverse event detection such as
structured medical record review. 40
The core components of an effective clinical incident reporting program are:
■ Provide extensive initial and ongoing education about the program for all staff.
■ Ensure all staff are clearly aware of what constitutes an incident (especially the importance of near
misses), which incidents should be to reported, and the process to be followed in reporting incidents.
■ Make it easy for staff to repon incidents by providing multiple, simple, readily available reporting
methods.
■ Ensure that reporting is confidential, and allow anonymous incident reporting.
■ Rapidly acknowledge to staff who have reported an incident that it has been received by the program
and is being investigated.
■ Investigate all incidents in a timely manner.
■ Provide timely feedback of the results of the investigation to the staff who reported the incident and
those who were involved in the incident.
■ Take timely and system-based actions to reduce the risks that have been identified by the
investigation of clinical incidents.
■ Do not take punitive actions in response to reported clinical incidents unless an "intentionally
unsafe act* has occurred.
 4 Detecting advert* events and rta*

a Ensure that some of the actions taken in response to reported incidents are highly visible to all staff
a Demonstrate to stafT that incident reponing can change their work environment, and increase the
quality of care they provide and the level of patient safety, a Regularly report the results of the clinical
incident reporting program both up and down organisation a Use national incident reporting databases
to detect areas of risk in the health service and to provide resources and tools to improve patient safety.

4.2.5 Sentinel event reporting

Reporting of major adverse events, and their subsequent investigations externally to the health service
in which they occurred, allows the lessons learnt by an 'individual health service to be used by other
organisations to avoid similar events from occurring in their facilities. The external organisation can
aggregate the results of multiple reports, identify trends and risks and djpseminate this information
widely. Important characteristics of successful external reporting systems are that those reporting the
events are safe Trom disciplinary action and that the process of reporting is not onerous. Also, reponing
of such events must be perceived as worthwhile, with expert analysis of these reports being
undertaken, timely feedback being provided to reporting health services and appropriate dissemination
of useful findings to the health care field. The process should also be confidential, independent,
systems-oriented and responsive to recommendations. 65

4.2.5.1 Sentinel event reporting in the United States

In the US, the Joint Commission on Accreditation of Healthcare Organizations reviews the response of
individual health services to major or sentinel events that have occurred at their facilities as pan of its
accreditation process. These events are referred to as sentinel because they require an immediate
investigation and response. An appropriate response includes "a timely, thorough and credible root-
cause analysis; developing an action plan designed to implement improvements to reduce risk;
implementing the improvements and monitoring the effectiveness of those improvements". 66
Root-cause analysis is a process of identifying the causal factors of the sentinel event and making
improvements to reduce the probability of the event recurring. An action plan is developed, outlining
the strategies to be implemented by the health service, who is responsible within the health service for
their implementation, and how strategies' effectiveness will be measured. Each health service's sentinel
event policy and examples of root-cause analyses the service has conducted during the previous year
are examined during the service's accreditation survey to determine if they comply with the
Commission^ standards.
In the US's health service accreditation system, sentinel events include surgery on the wrong patient
or body part, inpatient suicide, and incompatible blood transfusions. Each health service is encouraged
to report any sentinel event that occurs in the service to the Commission. Although most sentinel event
reports to the Commission are made by health services, they may also be informed by complaints made
to them by patients and their relatives, media reports and events identified by surveyors during health
service accreditation surveys.
Once the event is reported to the Commission, the root-cause analysis performed by the health
service is reviewed to ensure that it is "thorough and credible" and that the action plan is acceptable. If
these documents meet the required standards, the Commission will assign a follow-up activity to the
health service. This activity is usually an audit comprising measures to ensure that the actions taken
in response to the sentinel event were effective and sustained. The report of this audit is required to be
submitted to the Commission within 4 months. Failure by the health service to meet any of these
standards may result in a change in their accreditation status.
Since January 1995, over 4000 sentinel events have been reported to the Commission, most
commonly for wrong-site surgery, inpatient suicide and complications associated with operations.
Health services in the US have reported about 350 sentinel events each year for the past 3 years.
 Enhancing patient care

Comparing reporting to the Commission with mandatory reporting to some state authorities in the US,
sentinel events may be under-reported to the Commission by a factor of 20. 21
The Commission enters the details of each sentinel event, root-cause analysis and action plan into a
database for analysis. If panicular sentinel events are found to have occurred with significant
frequency, Sentinel Event Alerts are written and made available on the Commission's website. The
Alerts provide de-identified descriptions of the sentinel events, the number reported to them, the
causes found in root- cause analyses, the risk reduction strategies that were implemented by the
reporting health services, the Commission's recommendations regarding the sentinel event, and links
to further resources. Since February 1998, 37 Sentinel Event Alerts have been issued by the
Commission on a many topics, including wrong-site surgery, restraint deaths, blood-transfusion errors,
needlestick injuries and anaesthetic awareness.
Unfortunately, the Commission sentinel event program is viewed as cumbersome and time-
consuming, and may result in health services losing their accreditation status. Health services are also
worried about the public disclosure of details of individual sentinel events leading to possible litigation
and the potential loss of reputation. These factors may inhibit the reporting of sentinel events to the
Commission.
Sentinel event reporting and analysis requires a balance between the requirements for public
evidence of accountability with the possibility of sanctions for serious violations, and the potential
problems for health services if full disclosure of all sentinel events occurs. Some legislative protection
for health services may be necessary for some documents created during analysis and response to
sentinel events, and such legislation may increase reporting of significant adverse events. 65 Although
the hurdles surrounding the reporting of sentinel events are substantial, the benefits of aggregated
information about adverse events provided by the Sentinel Event Alerts for the development of safer
systems across the health care field are substantial.

4.2 J.2 Sentinel event reporting in Australia

In April 2004, Australian Health Ministers decided that "all public hospitals [are | to report sentinel
events, either to the state department or to an agreed third party ... [and] that all states and territories
will contribute to a national report on sentinel events . . ." 67
In addition to hospitals reporting clinical incidents to their insurers, some health care bureaucracies
in Australia request that hospitals report to them the details of specific major adverse or sentinel
events when they occur in the hospital.68
The Victorian Government Department of Human Services has developed a sentinel event program as
part of its Clinical Risk Management Strategy. Its reporting criteria are similar to those of the US Joint
Commission. The Department has defined sentinel events as relatively infrequent, clear-cut events that
occur independently of a patient's condition; commonly reflect hospital system and process
deficiencies; and result in negative outcomes for patients. 68 The Department wishes to gather as much
information as possible by accumulating reports of similar events from different health services, and
examining them for common underlying contributing causes. This information is used to aid collective
preventive efforts to improve patient safety. The current list of sentinel events required to be reported to
the Department is:
■ procedures involving the wrong patient or body part
■ suicide in an inpatient
■ retained instruments or other material after surgery requiring re-operation or further tuigcai
procedure
■ intravascular gas embolism resulting in serious neurological damage or death
■ haemolytic blood transfusion reaction resulting from blood-group incompatibility
 4 Detecting adveree even* and flefc

■ medication error leading to the death of a patient reasonably believed to be due to incorrect
administration of drugs
■ maternal death or serious morbidity associated with labour or delivery
■ infant discharge to wrong family
■ other catastrophic event.
In addition to reporting the details of these events <o the Department when they occur, Victorian
public hospitals are required to conduct a root-cause analysis and report the results of these
investigations and a risk reduction plan of the actions they have taken, orjjlan to take, to prevent the
event recurring. The risk reduction plan describes the actions that have will be taken to reduce the
probability of the event recurring, who within the health service is responsible for taking each action,
when the action is to be completed, which senior executive is accountable for minimising the future
risk of the event and how change will be measured and outcomes evaluated after action has been
taken.
In 2005-2006, 91 sentinel events that occurred in Victorian public hospitals were reported to the
Department; 49 were classified as "other catastrophic events". Annual reports detailing the number
and type of sentinel events, and factors contributing to the events, are reported to state departments
of health and published annually in Victoria, New South Wales, South Australia, Western Australia
and Queensland.
In Victoria, a Department-appointed reference group reviews all sentinel events reponed by
Victorian health services and the detailed analysis of the event conducted by the health service and
actions they have taken in response to the event. Feedback is provided to the health service on the
root-cause analysis they have conducted and, if necessary, recommendations are made regarding
funher action to be taken by the organisation. Information from these reviews thought to be of value
to all health services in the state is provided to them via a newsletter, issuing alens for significant
events and publishing an annual report including several case studies and risk reduction strategies.
In addition, the reference group makes a number of recommendations to the Department regarding
areas of clinical risk detected in the sentinel event reports and thought to be applicable to other
health services in Victoria.
In July 2007, the Australian Institute of Health and Welfare and the Australian Commission on
Safety and Quality in Health Care published a report on sentinel events that occurred in Australian
public hospitals in 2004-05. 69 The report documented the number of sentinel events reponed by
health services to their state bureaucracies across Australia under each type of event together with
contributory factors and several case histories.

4.2.5.3 Sentinel event reporting — the bottom line

Sentinel events are serious adverse events and usually only occur infrequently in individual
hospitals. It is likely that medical record review or clinical incident reponing will detect most sentinel
events Individual health services will probably gain more from reviewing the case histories in the
annual reports of sentinel event programs produced by each state health department and the
national sentinel event report for potential areas of high clinical risk than from only reviewing the
probably infrequent sentinel events that will occur within the health service itself each year.
Making use of the alerts issued by organisations such as the Joint Commission can assist health
services by indicating areas of potential risk that may not be evident at the local level, and providing
resources to reduce risk developed from reviewing the scientific literature and consulting clinical
experts. These resources can be appropriately modified by individual health services to meet local
 Enhancing patient care

needs and conditions. Such resources can save individual health services much time in not having to
sun the nsk reduction process in a clinical area with a blank sheet of paper, undertake their own
detailed literature review and "reinvent the wheel" when developing appropriate policies and tools for
staff to use.
Health services should regularly scan the alerts prepared by organisations such as the Joint
Commission to see whether any of the alerts issued are relevant to their own organisation. Also, if risks
art detected in particular clinical areas using the health serviced medical record review or clinical
incident reporting programs, the alerts can be scanned again to see if resources have already been
developed elsewhere to effectively deal with these risks.

4.2.6 General practitioner feedback

Patients are spending less lime in hospital. Therefore, the effects of adverse events that have occurred
dunng hospital admissions may not become evident until after patients have been discharged. Some of
these adverse events may be serious enough to require the patientls re-admission to the hospital. If so,
the adverse event will probably be detected by the medical record review process, as the admission will
screen positive for the criterion "unplanned re-admission within 21 days of discharge". More commonly,
the adverse event will not require re-admission to hospital. To capture details of such adverse events
not requiring re-admission at Wimmera Health Care Group, each patient's general practitioner is sent a
feedback form attached to the patient!; inpatient discharge summary (Figure 4.4). The form is open-
ended and simple to complete.
The patient's general practitioner is requested to keep the feedback form in their patient's clinic
medical record for 28 days, and to send the form to the hospital if an adverse event related to medical
care received at the hospital is evident when the general practitioner reviews the patient. The form is
returned to the director of medical services, who forwards the information to the clinical risk manager.
Support for this component of the clinical risk management model was obtained from the local
Division of General Practice (whose membership comprises most general practitioners in the region),
who assisted the hospital in its attempts to encourage doctors in the region to complete the feedback
form when adverse events were found.
When the hospital receives a completed feedback form from a general practitioner, the patient's
medical record is promptly reviewed and the details of the event are discussed with the relevant staff
members. If an adverse event is thought to have occurred, an adverse event analysis form is completed
and the event discussed at the next surveillance committee meeting. Only a small number of feedback
forms are returned to the hospital, but adverse events detected by using general practitioner feedback
have been different from those detected using medical record review and clinical incident reporting. As
wuh other adverse event detection methods, general practitioner feedback appears to detect events
from a different perspective to other detection strategies.
After the clinical risk manager has conducted a thorough investigation of the event, the general
practitioner who provided the feedback is sent a letter thanking them for taking the time to inform the
hospital of the event. The letter also provides a summary of the findings of the investigation into the
event and details of the actions taken to reduce the probability of the event recurring. It is hoped that
by providing this feedback, it will demonstrate to general practitioners that their action in providing
this information about their patients' care can improve the system of health care delivery, thereby
encouraging them to report further events.
The nsk of an adverse event increases when there arc weaknesses in health care delivery systems.
Weaknesses can become apparent with a change in provider and location of care, which occurs after a
patient is discharged from hospital In a Canadian siudy, more than one in five patients discharged
from a teaching hospital medical unit experienced a post-discharge adverse event related lo the
medical care

74
 4 Detecting advert* events and rta*

Figure 4.4 Wimmera Health Care Group general practMener feedback form

Wimmera Health Care Group General Practitioner Feedback

Form

Feedbedt IflfMiiiilkin
This card it to be completed for thw patiem following transfer or diKhaige torn the Wimmera
Health Care Group if an advene event hat occurred. If no atone event please retain in patient's record for 28 days
3
J
c
«
• Local CP
C
I
V Advene event definition
An untoward patient event that, under optimal conditions, it not a natural consequence of
the patient's disease or treatment.

Date adverse event recognised

Summary of adverse event

received in hospital. The most common adverse events were adverse drug events (72%). therapeutic
errors (16%), and nosocomial infections (11%). Half of these adverse events were preventable or
ameliorable.70 Obtaining feedback from general practitioners after patients are discharged from hospital
is an important source of information about adverse events related to hospital care, and provides useful
information to improve the quality and safety of care. The authors of the study suggested that patients
required closer follow-up after discharge from hospital, and that communication with primary care
providers and home care services should be enhanced. 70
Adverse events detected by general practitioners after the patientls discharge from hospital can result
in action being taken by the hospital to improve the transfer of care between the hospital and primary
care, allowing patients to move more safely between these two providers of their care.
 Enhancing patient care

4L2.7 OWcol pathways and variance analysis

Clinical pathways provide an evidence-based framework for high-quality, coordinated care. They aim to
optimise outcomes for patients by adapting best practice in a particular clinical area to local conditions
with prudent use of resources. Ideally, clinical pathways should be developed by a multidisciplinary
team using evidence from relevant high-quality clinical trials or, where this is not possible, from
consensus of expert opinion. Clinical practice guidelines issued by national medical bodies are useful
resources, but they may require modification to suit local conditions.
Ginical pathways enable standardisation of care where appropriate, which improves quality by
preventing omissions of essential elements in care and reducing practice variation. When used within a
multidisciplinary team approach, a clinical pathway can be incorporated into a medical record and
provide details of the care to be provided by all members of the health care team within a time frame. It
reduces documentation time for standard care, as only the initials of the health care professionals
involved are required to indicate the components and details of the care that has been provided.
When using clinical pathways, staff should consider whether the recommended management is
appropriate for their patient. If a particular treatment or investigation is not required, or a different
approach is necessary, a variance should be recorded. This recognises that an individual approach to
care is essential to allow for comorbidities, patient variation and process difficulties. Similarly, if
expected outcomes are not met, complications arise or patient recovery time exceeds expectations then
documentation of these outcomes as variances is required.
Variation in care is treated as an exception, and written documentation is required, detailing the
reason for the variation if known, the action taken and outcome achieved. This detail ei\ables analysis
of variances and allows system and clinical improvements to be made as trends are identified.
Variances may be analysed for each diagnosis-related pathway using a specific software program, or an
audit of relevant clinical indicators may be conducted. A combination of variance analysis and audit of
clinical indicators provides an opportunity to measure targeted improvements and detect adverse
events. Regular review (by the multidisciplinary team that designed and uses the pathway) of the
results

Figure 4.5 Proportion of patients with ST elevation acute myocardial infarction who received key interventions before and
after the introduction of the clinical pathway
irrw. a-----------------------------------—-----------■ - - -■-----------------------------------m-----------^
80% -i □ ■ - - - a A Patients who received aspirin in the
//
c .3 emergency department

o o 0-Blocker prescribed to eligible patients on discharge

| 40%- q/ 0 Lipid therapy commenced (or eligible patients
0£ L P-Blocker received by eligible patients within 24 hours of admission

20% - X Fasting lipid levels measured

Enhancing patient care

I I Pre-pathway Post- 2002 (n-43) I 2003 ' 2004 2005 2006 2007 (n-25) (n-22) (n-24) (n-17) (n-21)
pathway Jarv-Oec Oct 2000-
1999 Dec 2001 (n-42) (n —
31)
 4 Detecting adverse events and

achieved using individual pathways enables process and clinical improvements to be made using the
clinical pathway as a guideline for care and as a measurement tool.
The stroke pathway and ST elevation myocardial infarction pathway, examples of clinical pathways
developed at Wimmera Health Care Group and incorporated into patient medical records, are
available on the Wimmera Health Care Group website (www.whcg.org.au/Quality/index.aspx).
Figures 4.5 and 4.6 show the improvements in process resulting from the introduction of these
pathways. The development and use of clinical pathways will be discussed further in Chapter 7.

4.2.8 Patient satisfaction

Research has shown that patients are unable to accurately judge the quality of the technical aspects
of the clinical care that they receive. 71 They are, however, able to accurately assess the quality of their
nonclinical care, which is a significant component of the quality of care and service provided by
hospitals.
Despite the limitations of comments received from patients about the care they have received, such
information still provides valuable feedback to hospitals about how to improve aspects of their health
care delivery. This information can be obtained using patient satisfaction surveys, focus groups and
patient complaints.
We have used information from these three sources to provide information from yet another
perspective about where to direct our efforts to improve our health care delivery systems. Previously,
most information obtained from these sources has been about the non-clinical aspects of patient care
(eg, noise, privacy, temperature and the quality of food). More recently, treatment and inadequate
communication between clinical staff and patients are issues raised most commonly as an area of
patient dissatisfaction. Although patients may not be able to appreciate whether an error has
occurred while they are receiving medical care, they are usually well aware if they have been harmed
during the process.24
We have found focus groups extremely valuable for obtaining feedback from patients. At Wimmera
Health Care Group, focus group meetings, comprising patients who have received treatment in a
particular clinical area of the hospital, are conducted. These patients are invited to attend a group
meeting with a facilitator. A consistent format for these group meetings has been developed.
Information

Figure 4.6 Proportion of patients with stroke who received key interventions before and after the introduction
of the clinical pathway
100%
-,

80% -

60% -
Dysphagia screen performed within 24 hours of
admission
£ 40% - Aspirin or clopidogrel administered within 24
hours of admission of patients with ischaemic
20% - stroke
Computed tomography brain scan within 24 hours of
admission
Regular neurological observations
performed for the first 48 hours following the
stroke
 4 Detecting advert* events and rta*

Pre-pathway Post-pathway Jul

1999- Jun- Apr 2000 Dec 2000 (n-
27) (n - 37)
2006 2007 (n-35) (n-31!

2001 2002 2003 2004 2005

(n-8) (n- (n- (n-44) (n-
21) 36) 49)
obtained from focus group sessions is then discussed with the staff members who cared for the
patients in the hospital. This information can be positive and negative. Hospital stafT members in the
area under review are requested to review this information and develop and implement plans to
address the issues before a follow-up focus group session.
One limitation of focus groups is that, given the small number of patients involved in each group
session, an individual patient^ views may dominate discussion in the group and therefore may bias
the results obtained. To overcome this limitation, some issues raised by participants in focus group
sessions are investigated further by drafting appropriate questions and including them in a patient
satisfaction survey mailed to a larger sample of patients who have received treatment from the same
clinical area of the hospital.

4.3 Proactive risk management — detecting risks

4.3.1 Consultative councils and committees
The Victorian Government has established a number of consultative councils, comprising health
professionals working in specialised clinical areas. These councils are given the task of reviewing
clinical care in their area of expertise, and making recommendations about how this care should be
provided. They are also advisory bodies to the Victorian Minister for Health. The Consultative Council
on Obstetric and Paediatric Mortality and Morbidity (established in 1962 under the Health Act 1958
[VicD, the Victorian Consultative Council on Anaesthetics Mortality and Morbidity, and the Victorian
Surgical Consultative Council all analyse information sent to them voluntarily by clinicians and
hospitals on patient deaths and adverse events. Under legislation, the information provided to
Councils cannot be accessed by third parties.
After undertaking this analysis, the Councils make recommendations about how care can be
provided with increased safety. This information is included in their annual reports, which are
circulated to relevant clinicians, hospitals, the Department of Health and the Minister for Health.
These recommendations can then be adapted by individual hospitals for local conditions and
incorporated into their service delivery systems. 72

4.3.2 Coronial reports

Coronial reports about patient deaths in hospitals often provide very detailed analysis of the failures
of some health care delivery systems. In 1 year, individual hospitals may be involved with a small
number of coronial inquests into deaths associated with an adverse event. Fortunately, few adverse
events that occur in hospitals result in patient death. In our experience, by studying coronial reports
about adverse events that have occurred in other hospitals, it has been possible to strengthen the
related health care delivery systems within our own hospital.
A coronial report into the unexpected death of a patient at another hospital is analysed by the
surveillance committee in the same way as an adverse event that has occurred at Wimmera Health
Care Group. Staff members ask: could the adverse event described in the coronial report occur in this
hospital? Invariably, the answer is yes. Action can then be taken to strengthen the relevant systems
to reduce the probability of a similar event occurring at Wimmera Health Care Group. As a
consequence, the resilience of the delivery systems in question and the subsequent quality of care are
improved without staff members having to experience the negative emotions attached to dealing with
the consequences of a serious adverse event that has occurred in their own hospital.
On" occasion, the Victorian Department of Human Services has sent each health service in the
state a summary of a coronial inquest into a patient's death in a hospital, copies of the actions taken
 by the health service in response to the death and the coroner's findings. The Department has also
provided health services with copies of any forms or policies that the health service has developed
in response w the death, and contact details to obtain further information about the actions the
health servioe has taken. Such information has been well received by health services and has
saved them considenMe work in not having to "reinvent the wheel" to strengthen system
weaknesses related to the advene tw* that has been the subject of the coronial inquest.
The Coronial Liaison Service in Victoria (an initiative of the State Coroner's Office of Victoria and
the Victorian Institute of Forensic Medicine) electronically distributes two quarterly newsleuen: the
Coronal Communiqut (www. vifm. org/communique html) and the Residential Aged Care Coronial
CommunufU (www.vifm.org/n963.html). These newsletters discuss cases reported to the State
Coroner and thought to be of interest to health care professionals and the aged care community,
with an emphasis on system failures. In our experience, Coronial recommendations can be an
effectivejnethod of changing clinical behaviour.

4.3.3 Information from Insurers

Fortunately, only a very small proportion of the adverse events that occur in hospitals result in
litigation. Through their imponant role in providing indemnity cover, hospital insurers and medical
defence organisations have an overview of the types of events that may result in litigation, and they
regularly provide advice about how hospitals and clinical staff members can prevent such events. For
example, some medical defence organisations give advice about many issues, including
documentation in medical records, communication with patients and preventive measures that
should be taken in high-risk clinical situations to prevent adverse events occurring. Using this
information as a guide, the relevant hospital systems can be checked to ensure that these
recommended practices are in place.
Similarly, some hospital insurers recommend processes that hospitals should have in place, such
as procedures for obtaining informed consent and for undertaking medical research. Again, it is
relatively easy to check the relevant hospital systems to see whether these measures are in place.
Insurers and medical defence organisations send their members regular reports and bulletins,
highlighting significant adverse events that have occurred in hospitals and private medical practices,
and which recommend appropriate preventive actions that should be taken to reduce the probability
of such events occurring. This information can be dealt with in a similar manner to coronial reports.
Some medical defence organisations encourage their members to develop clinical risk management
programs in their practices and offer reduced medical indemnity insurance premiums to doctors who
develop such programs to an appropriate standard. These organisations also offer their members
education programs about clinical risk management, which should help health services educate their
medical staff about risk management and in developing hospital wide programs.
Each hospital has information available about adverse events that have occurred at the hospital
and have resulted in litigation. The surveillance committee can review these adverse events, but most
will already have been detected by medical record review or clinical incident reporting. In our
experience, very few adverse events in individual hospitals result in litigation. Reviewing aggregate
data and the details of events and claims that have occurred at other hospitals is a more useful way
to improve delivery systems in one's own hospital with claims information.
In the US, the 10 most common causes of medical malpractice lawsuits are: medication errors,
incorrect diagnosis, resulting in incorrect or delayed treatment; inadequate supervision of junior
medical and nursing staff; delayed treatment; failure to provide adequate information about the rtsks
of treatment for the patient to be able to give informed consent; doctors performing procedures for
which they have inadequate skill or experience; unexpected patient death due to inadequate clinical
history being taken, iatrogenic injury, such as wound infections and fractures obtained while in
hospital; pain, suffering and emotional distress following procedures where patients have unrealistic
expectations of the outcomes; and lack of teamwork and communication by hospital staff members
about the patient's clinical condition. Numerous strategies are available to reduce the probability of
many of these adverse events occurring. 71 In our experience, a consistent theme in adverse events
that are referred to insurers is poor communication between doctors and with their patients.

4*3.4 StotewMt and national databases of clinical Incidents

As discussed previously, clinical incident reporting can be expanded beyond incident reporting at the
individual health service level to reporting on a state or national basis. Such reporting creates a
large, unique and valuable database of incidents that can occur in a health service. These national
databases, where data from incidents in many health services have been aggregated, can identify
risks associated with low-frequency events and higher-level system factors whose significance may
not be realised at the local level. 6''74 National databases can also focus analysis on clinical incidents
reponed in specialised clinical areas, such as intensive care units. 45-65
The Australian Patient Safety Foundation in Australia has developed statewide databases. In the
UK, the NHS has created the National Patient Safety Agency to develop a national database of clinical
incidents. The National Center for Patient Safety of the Veterans Health Administration, a 173-facility
system in the US, has also developed a database of investigations into major incidents.

4.3.4.1 The Australian Patient Safety Foundation

The Australian Patient Safety Foundation is a non-profit, independent organisation originating from
an incident monitoring study in anaesthesia in 1989. The Foundation provides a patient incident
reporting and monitoring system (the AIMS software) through a commercial subsidiary for all public
hospitals in New South Wales, Western Australia and South Australia, and some health services in
the Northern Territory, Victoria, Queensland and New Zealand. The Foundation aggregates and
analyses deidentified data to determine "contributing, minimising and preventive factors which can
then be used in devising corrective strategies and action plans for incidents affecting patient
safety."75
For example, incidents involving patients and physical violence or violent verbal exchange that
were reported over a 2-year period were analysed and preventive strategies proposed and published, 75
as have data about bums patients have received in hospitals. 76 Although hospitals from many pans of
Australia send incident data to the Foundation, this information is entered into statewide databases,
not an overall national database.

4.3.4.2 The National Reporting and Learning System in the United Kingdom
The National Patient Safety Agency, which reports annually to the British parliament, has
responsibility for the National Reporting and Learning System, through which clinical incidents are
voluntarily reported by staff in all NHS organisations in England and Wales. In 2005, the Agency
received almost half a million reports of patient safety incidents. These incidents are analysed "to
inform patient safety learning, action and priority setting across the NHS". 58 Individual health
services are also able to compare their incident reporting data with like services.
After analysis of clinical incidents at a national level and examination of other sources of relevant
information such as published literature, national audits, litigation claims and product regulatory
agencies and consultation with relevant expert clinicians and patient groups, information may be
sent to all NHS organisations. This information may be provided in one of three forms: patient safety
alerts that require prompt action by health services to address high-risk areas; patient safety notices
that strongly advise health services to take action in particular areas; or patient safety information,
which suggests actions that organisations should consider taking. Lay versions of this information
are also developed.
 4 Detecting adverse
everts and «#

Patient safety alerts. These alerts detail the number of clinical incidents that have been reported in
the area of concern and relevant litigation data. Details of actions to be taken by health services to
reduce risks in these areas and deadlines for their completion are also given. Appropriate templates
and exemplar documents (including patient education) to address the risk are created and sent to
health services to help them take action. These documents are also available on the Agency's
website (www.npsa.nhs.uk/alerts-and-directives), with the .reminder that the documents will require
local adaptation and should be ratified for use by local clinicians. Where possible, attempts are made
to redesign systems to eliminate risks rather than issue further guidelines. 74 Over 30 alerts and
notices have been issued by the Agency over the past 3 years. They cover a wide range of topics,
including standardisation of patient wristbands, administration of intravenous infusions to children,
and early identification of failure to act on radiology imaging reports.
With the issuing of alerts and patient safety notices and information, much valuable and useful
information has been sent to health services arising from the analysis of clihical incidents at a
national level. However, there have been concerns expressed about some of the alerts issued. Some
clinicians fed some alerts detailing mandatory requirements for changes in clinical practice are being
issued before strong scientific evidence supporting such changes is available. Such alerts are viewed
by some clinicians as "knee-jerk" reactions to a small number of adverse events, and that issuing
alerts in such a manner may increase doctors' resistance to changing their practice in a particular
clinical area. One such alert concerned the positioning of nasogastric tubes and the type of paper and
supplementary tests used to test for the position of the tube. The changes suggested in the alert had
not been tested in clinical practice. A survey of neonatal units in the UK showed over half the units
had not made the changes in the alert 9 months after it had been issued.
These events highlight the importance of good evidence in the scientific literature supporting the
proposed changes in clinical practice described in alerts, of recommending changes that are practical
to implement, and of a perceived need by clinicians for change, and of strong support for the
recommendation by relevant professional bodies. Changing practice without strong evidence of the
change being safe may result in replacing one risk with another. 77 The difficulty arises when the
Agency receives a series of incident reports in a clinical area that clearly indicates a problem, but
there is no high-quality evidence in the scientific literature to support recommendations to change
clinical practice. It has been argued, however, that although the guidelines regarding testing of the
position of nasogastric tubes were not ideal, they were still a step in the correct direction, as issuing
an alen should increase awareness of the problem and the associated uncenainties and confusion,
and also, hopefully, encourage reporting of further incidents in this area. 78
Detailed reports are also published by the Agency. One such repon reviewed incidents involving
107 patients whose deaths in hospital in 2005 were reponed because of concerns about the safety 7 of
their care.79 The report also reviewed relevant claims data and the scientific literature.
Recommendations were made about clinical deterioration in patients not being recognised or not
being acted upon, and problems associated with resuscitation. The report not only highlighted the
problems, but also provided examples of good practice regarding these issues already being
undenaken in some health services in the NHS with a view to individual health services reviewing
their systems and performance in these important areas.
There has been criticism of the Agency, as there is still significant under-reporting of serious
incidents and medication errors, and the potential learning from over one million reponed incidents
into improvements at the coalface across the NHS has not been maximised, with serious delays in
developing and sharing of effective solutions to incidents. 80 Pan of problem may be lack of a minimum
dataset to classify patient safety incidents 81 43.43 The feterans Health Administration incident reporting
system
The Veterans Health Administration incident reponing system in the US is run under the direction of
its National Center for Patient Safety, which provides tools, training and expert consultants for all the
 4 Detecting advert* event* «nd tm

Administration^ facilities. The system was developed to meet several criteria essential in creating an
effective incident reponing program:
■ The program was to be seen as non-punitive by Administration staff reporting incidents.
■ Analyses of the incidents were to be undertaken by multidisciplinary teams drawn from the
patient- staff coalface of the health service, to generate the most innovative and effective solutions
to the system weaknesses that were detected.
■ Timely feedback was to be provided to all staff who reported incidents to indicate that their report
had been appropriately dealt with by the facility and not lost in the system.
■ Near misses or close calls, as well as adverse events, were identified and reported. In the
Administration^ experience, near misses occurred more frequently than adverse events, with high-
priority close calls making up 90% of reported incidents. Importantly, these near misses identified
the same weaknesses in systems as adverse events, and were therefore equally valuable in
improving patient safety. The major difference between a near miss and an adverse event is that
with the near miss, a barrier in the organisation stopped the error affecting the patient and
causing an adverse event. Focusing only on events that cause harm to patients and not capturing
data from near misses would significantly detract from the potential benefits of a health service's
quality improvement and patient safety program. 61 Also, the analysis of near misses is less likely to
be affected by the reviewers displaying hindsight bias (where the assessment of an event is affected
by the reviewer knowing the patient's outcome). 21 Importantly, the Administration felt it may be
easier for staff to report near misses than adverse events because there was no attached fear of
punitive action, liability or embarrassment.
■ Reporting clinical incidents did not require significant additional work to be done by staff.
■ The major objective of the program of improving the level of patient safety was clear to staff.
The Veterans Health Administration also recognised that implementation of such an extensive
patient safety program needed to be incremental and would require significant staff training. The
Administration placed the program highly in its organisational structure, centrally and in each
facility. At the local level, such positioning of the program clearly indicated to staff that patient safety
was a high priority in the organisation and facilitated communication between the program safety
manager and the chief executive at each facility. Centrally, the Director of the National Center for
Patient Safety, which oversees the incident reporting system, reports directly to the Under Secretary
of Health. This process may assist if the program is not being fully supported by the management at
a local facility.82
The Administration made it clear to its staff that the program was not an incident-counting
exercise, and that it would not measure the success of the program by the number of incidents that
were reported. Rather, if more incidents were reported, it could indicate that an appropriate culture,
in which stafT felt able to freely report incidents without personal repercussions, had been developed
at a facility. The Administration regarded incident reporting as very important, because "you can't fix
what you don't know about".83
The Administration's program created a prioritising scoring method related to the severity and
frequency of each incident to allow a Safety Assessment Code score to be calculated and action to be
taken in response to the priority of the incident. This scoring method resulted in consistent handling
of clinical incidents throughout the many Administration facilities. This assessment and scoring of
incidents will be discussed in more detail in Chapter 5.
An intentionally unsafe act. Importantly, although the Veterans Health Administration incident
reporting system aimed to be non-punitive, it did not provide those reporting incidents with total
immunity from disciplinary action. The Administration gave an undertaking to its staff that as long
as the incident reported did not involve an "intentionally unsafe act", punitive action would not be
Enhancing patient care

taken because of findings of an investigation into an incident. The Administration defined an

"intentionally unsafe act" as "a criminal act, a purposefully unsafe act, an act related to alcohol or
substance abuse of an impaired provider or staff member, or events involving alleged or suspected
patient abuse of any kind".83 Incident reporting in the Administration's program involving such acts
could result in disciplinary action being taken by the facility. 83 In other industries such as aviation,
immunity for reponing non-criminal offences has been an incentive to reponing incidents. 61
Alerts and advisories. As a result of the reponing and analysis of clinical incidents, over 70 alerts and
advisories have been sent to Veterans Health Administration facilities and posted on the
Administration's website (www.va.gov/ncps/alerts.html) since 2000. Many of these relate to
equipment warnings and provide details of incidents, actions that facilities should take to reduce or
eliminate the associated risk, information from the manufacturers of the equipment and contact
detail? for further information. A bimonthly newsletter with details of findings and ideas from the
analyses of reponed incidents is also published (www.va.gov/ncps/TIPS/tips.html).
In 2000, the Administration extended their mandatory internal reporting system to include a
voluntary external reporting system developed in partnership with the National Aeronautic and Space
Administration (NASA), which manages an aviation safety reporting system. NASA was seen as an
impartial third party with the skills and track record to run the program. The program is called the
VA- NASA Patient Safety Reporting System, and its primary aim was learning from incidents, rather
than accountability. Administration staff mail incident reports send directly to NASA. Importantly,
these reports are not made anonymously. Identification of the reporter allows contact to be made by
the person analysing the report to obtain additional information to enhance understanding of the
incident. After this contact, the report is de-identified, ensuring that no punitive consequences result
from the incident being reported. Reports are made in narrative form to provide more information
about the context of the incident.84
Other smaller-scale incident reporting databases have been established in the US, including a web-
based incident reporting system in Pennsylvania. The Patient Safety Authority, an independent state
agency, was established under legislation aimed at reducing medical error and all Pennsylvania-
licenced hospitals are required to report serious events and incidents. The Authority publishes
quarterly patient safety advisories (www.psa.state.pa.us), in which it reviews trends in reported
events and advises agencies "of immediate changes that can be instituted to reduce Serious Events
and Incidents".

4.3.5 Media reports

Reports appear frequently in the media about adverse events that have occurred to patients in
hospitals These reports are often useful in identifying weaknesses in health care delivery systems.
Unfortunately, unlike coronial reports, media reports are often sensationalised and do not contain all
the details of how and why particular adverse events occurred; they also often contain inaccuracies.
Despite these shortcomings, we have still found it worthwhile to analyse media reports. We use them
as prompts to examine the strength of relevant delivery systems in our own hospital that relate to the
system deficiencies discussed in individual media reports of adverse events.
Some adverse events may occur only rarely in an individual hospital, but even those occurring as
infrequently as once every 10 years may have significant negative consequences. It is worthwhile,
therefore, to take action to prevent such an event occurring in one's own hospital by learning about
the system weaknesses that have contributed to such an event occurring elsewhere. This proactive
method of preventing adverse events requires a degree of discipline in following up such media
reports, and a systematic approach to analysing and acting on the adverse events described in them.
As with coronial
 reports, although this proactive approach requires additional work, it
requires considerably less emotional energy than having to deal with the
same system issues in an emotionally charged environment after a major
adverse event has occurred in one's own hospital.

4.3.6 Clinical Journals

The volume of clinical information available to health professionals is
expanding rapidly. It is not possible to read about and implement all
advances in clinical knowledge and quality activities described in medical
journals. We often find useful information in relevant journals, which we
regularly scan for details of innovative quality improvement programs and
clinical improvement initiatives in specific areas.
Useful articles fall into two board categories; they describe:
■ innovations which can enhance the components of our quality improve-
ment and clinical risk management system
■ clinical advances that can guide future modifications for how we provide
clinical care.
Invariably, we find far more ideas in these articles than we can
effectively and efficiently implement. Copies of relevant articles are sent to
the surveillance committee for discussion or to multidisciplinary groups in
the hospital who are developing clinical pathways.
In the Wimmera program, information from multiple external sources
about adverse events and system improvements is used to reduce the
probability of adverse events occurring in our hospital. We have gained
valuable insights and benefited from the observations made of other
hospitals' experience with adverse events and their system improvements.

4.3.7 Pharmaceutical and equipment alerts

Health services periodically receive product alerts from pharmaceutical
companies and equipment manufacturers about the safety of their
products. These alens may require withdrawal of medication from the
hospital pharmacy, or urgent withdrawal and review of equipment used in
delivering care. These products may represent an unacceptable risk to
patient safety until they are appropriately dealt with. In our experience,
companies send such alerts to several people within each health service in
the hope that by widely distributing this information, it is more likely that
appropriate action will be taken to eliminate the risk. Unfortunately, this
wide distribution may result in an uncoordinated approach in dealing with
the risk, or worse, that appropriate action is not taken as staff sent the
alert believe someone else in the health service will take the required
action.
As with other areas of risk management, a systematic and integrated
approach is required to reduce or eliminate the risk associated with each
Tenant safety inquiries product alert. The products that represent a risk to patients and the alert
acnsstoeworid should be recorded on the health service's risk register when they are
consistently identity toe received by the service, and remain on the register until the risk has been
same recurring problems
eliminated or reduced to an acceptable level. In the NHS, numerous alerts
as the cause of iatrogenia:
health care below
promulgated standards:
lack of quality-monitoring
processes, patients,
family members and
concerned staff being
ignored and excluded;
wtiisttebtowers being
vilified, and persistent
deficiencies in teamwork,
systems and
communication.'

Travaglia JF. et al
Med J Aust
2008: 188: 437-
438
from the Medicine and Healthcare Products Regulatory Agency are sent to
all
 4 Detecting advert* events and rta*

health services via the Central Alening System (previously known as the Safety Alert Broadcast).
These alerts provide details of the problem with panicular items of equipment and designate those
NHS personnel and organisations that need to take action. In Australia, alerts and advisories
regarding medicines and medical devices are displayed on the Therapeutic Goods Administration
website (www.tga.gov.au) and sent to individual doctors and health services.

4.3.8 Inquiries into individual health services or specific clinical areas In health service*
After high-profile adverse events or a series of serious adverse events in one health service,
governments may establish inquiries to determine the extent of the adverse events and the reasons
why they have occurred, and to make recommendations to correct deficiencies in clinical governance
and systems in the health service. Such inquiries have been held in the UK into complex paediatric
cardiac surgery undertaken the Bristol Royal Infirmary 83 and in Australia into patient deaths at
Bundabeig Base Hospital in Queensland, 86 Campbelltown and Camden Hospital in Sydney, 87
obstetrics and gynaecological services at King Edward Memorial Hospital in Penh 88 and neurosurgical
services at the Canberra Hospital. 89 These inquiries were established after stafT, often at great
personal cost, became whisdeblow- ers after repeated attempts to have concerns addressed internally
did not result in appropriate action being taken by senior management and clinicians in these
services.
The reports are salutary, useTul reminders to management and clinicians in any health service
about what can go wrong in the delivery oT health services without adequate and effective clinical
governance procedures. The recommendations often describe what should be included in a health
service's clinical governance framework to ensure that effective quality improvement and patient
safety activities are undertaken. Faunce and Bolsin provide a brief summary of the major deficiencies
found by the inquiries in the Australian health services listed above (except Bundaberg) in a
discussion on whistleblowers. 90 A review of hospital inquiries in Australia looked at the impact on
staff and safety and found that: 91 a Concerns about clinical standards were raised by staff with
management, but these concerns were
mishandled, resulting in whistleblowing in four hospitals, a After each whistleblowing episode, an
official inquiry was undertaken and made recommendations
about how to improve patient safety in the hospital concerned, a Following these inquiries, there
was generally a loss of trust in management and among clinicians.
Patients and the community also lost trust in the hospitals, a Such trust needed to be rebuilt
before stafT will report mistakes or other concerns about safety, a To successfully implement patient
safety procedures requires the development of policies detailing the professional duty of staff to report
concerns about colleagues if they believe these concerns pose a risk to patients.
A review of numerous inquiries conducted in the NHS over many years found although the reports
were published (and therefore accessible), they were lengthy and hard to read, and that probably only
the executive summaries and press reports would be widely read. Most inquiries into failures in
quality of care found similar causes: organisational and geographic isolation (lack of transfer of
innovation and limited peer review); inadequate leadership (unwilling to address known problems);
system and process failure; poor communication (problems not detected); and disempowerment of
staff and patients (staff discouraged from raising concerns). The authors fell that the consistency of
the causes of these failures could reflect that the recommendations of these inquiries were not
properly implemented across the health care field, and that lessons from these failures were not
always leamt. They believed that appropriate communication and dissemination of the reports of such
inquiries was very important 01
Enhancing patient care

A more recent review of 13 major investigations undertaken in the NHS since 2004 found that "the
themes that emerge are depressingly familiar": 93 services with weak leadership; lacking effective
direction; tolerating quality failures; trying to meet conflicting targets; inadequately using data, and
lacking teamwork. The review also warned of the negative effect of continuous organisational
restructuring, which diverts management^ attention from the quality of patient care. Continuity of
leadership is an important factor in providing high-quality care. One health service that was
investigated had seven chief executives in 10 years. 93,94

4.4 Additional proactive risk management methods

As well as health services learning from adverse events that occurred elsewhere, different types of
proactive risk management methods are available to reduce risk and prevent adverse events. These
methods include examining the conditions of work in the health service when designing health care
delivery systems and taking action to strengthen any weaknesses that are detected; assessing the
risk of individual patients experiencing common adverse events and taking appropriate preventive
action; encouraging patients to take an active role in their care; and using preventive health
strategies to reduce the prevalence of common and preventable medical conditions, thereby reducing
the need for hospital admissions.

4.4.1 Examining latent factors or conditions of work

It is important to examine the latent factors or conditions of work present in the health service when
designing new systems or assessing existing ones. When examining how they provide care, health
services should ask: do staff have the appropriate skills and training for the clinical seryices they are
providing; are stafT members' workloads and shift lengths appropriate, and is the appropriate
equipment readily available them to safely provide the required care? Proactively examining such
latent factors or conditions of work present in delivery systems and addressing the deficiencies can
reduce the probability of errors and adverse events.
Such a proactive approach to risk management has been formalised by the Veterans Health
Administration in its Health Care Failure Mode and Effect Analysis. The program is consistent with
the Joint Commission's accreditation standard that health services in the US have a proactive
program to identify risks to patient safety. The standard requires that the program identify failure
modes and determine the causes of the failure and how critical the effect of such failure is to the
safety of that system of care. Processes should then be redesigned to address these failures and
tested to ensure that the changes made have resulted in the desired outcomes being achieved. Health
services accredited by the Commission are required to undertake at least one Health Care Failure
Mode and Effect Analysis annually.
The Veterans Health Administration program begins with the health service deciding which area
should be studied. The area chosen is usually high-risk because of the considerable resources
required to undertake the analysis. For example, an area identified by the Joint Commission as being
frequently reported as sentinel events may be chosen for study. A multidisciplinary team including
subject experts is then assembled. The processes and subprocesses of the selected area are then
assembled into a flow diagram.
Processes in health care can be very complex, and often a small, manageable pan of the process
will be selected and studied in detail. All failure modes (how the system may fail) in the processes in
the selected area and their causes are identified. A hazard score is calculated using the same
definitions for severity and probability as the Safety Assessment Code (see Section 4.2.4). A decision
 4 Detecting adveree everts and

tree algorithm is then used to determine which failure modes require action to be taken. The action
aims to reduce the probability of the failure occurring, or to reduce its severity.
Some failure modes will not require action because there is an effective control measure in place
to decrease the probability of the failure occurring, or the area identified is not assessed as critical.
Where action is required, the support of management is obtained and outcome measures are
determined. After action is taken, the functioning of the system is tested to ensure that the changes
have not created new vulnerabilities in the system.95
As mentioned previously, such analysis is resource-intensive and time-consuming, and its
effectiveness in improving safety has not been determined. 96 Health Care Failure Mode and Effect
Analysis should be used selectively and in carefully chosen areas to be cost-effective. For
organisations with limited resources, a simple examination of the latent factors present in a
particular system of health care delivery could be used more frequently and should also bs effective
in reducing risk.

4.4.2 Assessing individual patients' risks of experiencing common advene events

It is imponant to assess each inpatient's risk of experiencing adverse events 1hat are known to occur
commonly in hospitals. For example, when patients are admitted to hospital, they can be assessed for
their risk of falling or developing a pressure ulcer or venous thromboembolus. If the risk of
experiencing such adverse events is found to be high, appropriate preventive measures can be taken
to reduce the probability that these events will occur during the admission. In our experience, such
assessments are often completed, but when a patient is determined to have a medium or high risk of
a panicular adverse event, appropriate preventive action is not consistently taken.
Criteria have also been developed to alen staff of patients on the ward who may be deteriorating
clinically and at risk of cardiac arrest. When one or more of these criteria or clinical markers (such as
the patient's pulse rate, respiratory rate, blot>d pressure, oxygen saturation and urine output) are
outside predetermined limits, a medical emergency team is called and immediately assesses the
patient to prevent their clinical condition deteriorating further. This proactive approach has resulted
in earlier intervention for patients who are deteriorating clinically 97 These early warning systems 49
empower staff to obtain assistance for seriously ill patients to prevent an adverse event or limit the
harm it causes.

4.4.3 Patients taking an active role in their care

Encouraging patients (or parents, if the patient is a child) to lake an active role in monitoring their
safety while receiving care can make it more difficult for adverse events to occur. Patients can be the
ultimate barriers in preventing errors by clinicians from causing harm. Suitable roles for patients that
will increase their level of safely include helping clinicians reach the correct diagnosis, ensuring they
are well informed about the advantages and disadvantages of their treatment options, and playing an
active role with their clinicians in deciding which treatment option is the most appropriate to meet
their needs.
Patients should choose an appropriately trained clinician who is experienced in treating patients
with their medical condition, or performing procedures the patient requires, and who practises
medicine safely. Patients should also ensure that the treatment they receive is the treatment that is
regarded as best practice for their condition and is consistent with the treatment that has been
ordered for them to have. Patients can check that the treatment is being correctly administered,
identify side effects and adverse events related to their treatment, and rapidly notify nursing and
medical staff about these effects.
Enhancing patient care

To date, the contribution patients can make to increase their own safety while receiving care has
been under-utilised. Often, patients are passive recipients of medical care, but their potential to play
a positive, active role is considerable. There will be considerable variation between patients and
different clinical circumstances with the same patient as to where on the passive-active spectrum they
wish he when receiving carc.qH Variation in patient involvement in their safety may be related to
demographics, the severity of their illness, the selling in which health carc is being provided and
whether the safctv behaviour challenges the clinicians' professionalism. However, whatever safety role
patients choosc to lake, they can only be a safety "buffer" in addition to those in place in the hearth
service — the responsibility for providing safe care still resides with the clinician. 99

4.4.4 Preventive medicine

Patients can also prevent adverse events from occurring to them by taking good care of their health.
One latent factor that contributes to errors occurring in providing health care to patients is the
workload of the clinicians providing the care. Many clinical conditions, such as ischaemic heart
disease and diabetes, are significantly related to patients' lifestyles, and are therefore highly
preventable. If these conditions can be prevented by lifestyle modification, the demand on health
services and individual clinician workloads will be lowered and the probability of adverse events
occurring in hospitals will be reduced.

4.5 Systematic review of external information about risk

There are many sources of information about adverse events that have occurred in other health
services and several other proactive methods of risk management. Individual health services require
a coordinated approach to systematically and regularly review each of these sources and methods,
and to direct appropriate information to the service's quality improvement and risk management
program. A comprehensive list of sources of information relevant to the health service should be
created, and a staff member assigned to regularly review each source and present relevant
information to appropriate staff members and to the quality and safety program. Alternatively,
specific clinical areas (eg, pharmacy) could review the external sources appropriate to that area and
direct this information to the program. Responses to this information throughout the health service
could then be tracked. 100 In our experience, much information will be available from these sources,
and will require extensive filtering so that information relevant to the needs and conditions of the
individual health service is located, distributed and appropriate action taken. The major challenge
will be to deal effectively with the volume of information available, and to harness it to achieve the
maximum positive effect on quality and safety.
Reactive risk nwnnflwnwtf—delecting adweiae evwils
In the Wimmera model, adverse events are detected using diverse methods:
■ Medical record review — health information services staff screen all inpatient medical records after
discharge using nine general patient outcome criteria; and records screening positive undergo medical
review. Record review provides an automatic, continuous flow of information about adverse events.
The method detects adverse events not found by other methods and important issues not related to
adverse events. Medical record review has a high false-positive rate and can be costly and time -
consuming. Three screening criteria are recommended for initial use: patient death, transfer from the
general ward to the intensive care unit, and any patient record referred for review by a clinician. The
screening criteria used should be tailored to local needs and conditions and using trigger tools can
enhance record review
 4 Detecting adveree everts and

■ Clinical incident reporting — staff members report adverse events and near misses locally and the
health service reports the details to a national database. Incident reporting detects adverse events
not detected by other methods. The effectiveness of incident reporting is limited by under-reporting,
especially by doctors. Staff should be educated about what incidents to report and how to report
them. It should be easy for staff to report incidents, and timely feedback should be provided to them
about the results of the investigation and the actions taken in response to reports they have made.
Where possible, actions taken in response to reports should be made visible to staff. It should be
demonstrated to staff that they can improve the care they provide and their work environment by
reporting incidents.
■ Sentinel event reporting — details of a prescribed list of serious adverse events are reported by
health services to the Victorian Department of Human Services shortly after they occur, with
subsequent reporting of the results obtained from their detailed analysis of the event and action
taken by the hospital in response to this analysis.
■ General practitioner feedback — feedback forms are sent to all referring general practitioners after
their patients are discharged from hospital, requesting details of any adverse events they have
detected that may be related to their patient's hospital care.
■ Clinical pathway variance analysis — variation from guidelines and/or outcomes within a clinical
pathway are reviewed to detect adverse events and process difficulties.
■ Patient satisfaction — details of adverse events are collected from patient satisfaction surveys,
focus groups and patient complaints.
Proactive risk management — detecting risks
Risks are proactively detected in the Wimmera Clinical Risk Management program:
■ Using external sources — adverse events that have occurred at other hospitals are reviewed to
determine if they could occur at Wimmera Health Care Group. Details of adverse events occurring at
other health services obtained from media reports, coronial investigations, clinical journals,
consultative committees, state and national databases and health service inquiries, are
systematically reviewed to detect system weaknesses and risk locally.
■ Reviewing latent factors when developing new delivery systems or reviewing existing systems.
■ Assessing individual patients' risks of common adverse events (eg, falls, thromboembolisms) and
taking action to prevent them occurring.
References
1 Arx*ews LB, Stocking C, Krizek T. et al. An alternative strategy (or studying adverse events in
medical care. Lancet 1997; 349:309-313.
2 Christian CK, Gustafson ML. Roth EM, et al. A prospective study of patient safety in the operating
room. Surgery 2006; 139.159-173.
3 Tucker AL. Edmondson AC. Why hospitals don't learn from failures; organisational and
psychological dynamics that inhibit system change. Calif Manage Review2003; 45: 55-72.
4 Brennan TA, Leape LL, Laird NM, et al. Incidence of adverse events and negligence in hospitalised
patients: results of the Harvard Medical Practice Study 1. N Engl J Med 1991; 324:370-376.
5 Vincent CA. Davy C, Esmail A, et al. Learning from litigation. The role of claims analysis in patient
safety. J Eval Clin Pract 2005; 12: 665-674.
6 Wu AW. Handling hospital errors: is disclosure the best defence? Ann Intern Med 1999; 131: 970-972.
7 Hiatt HH. Barnes BA, Brennan TA, et al. A study of medical injury and medical malpractice: an
overview. N Engl J Med 1989; 321: 480-484.
Enhancing patient care

8 California Medical Association and California Hospital Association. Report on the Medical Insurance
Feasibility Study. San Francisco: Sutter Publications, 1977.
9 Craddick JW, Bader B. Medical management analysis: a systematic approach to quality assurance
and risk management. Auburn, Calif: Joyce W Craddick, 1983.
10 Forster AJ, Fung I, Caughey S. et al. Adverse events detected by clinical surveillance on an
obstetric service. Otostef Gynecol 2006; 108: 1073-1083.
11 Jackson T. Duckett S. Shepheard J, Baxter K. Measurement of adverse events using incidence
flagged' diagnosis codes. J Health ServRes Policy 2006; 11: 21-26.
12 Wolff AM. Limited adverse occurrence screening: using medical record review to reduce adverse
patient events. MedJAust 1996; 164: 458-461
13 Dunn KL. Reddy P, Moulden A, Bowes G. Medical record review of deaths, unexpected intensive
care unit admissions, and clinician referrals: detection of adverse events and insight into the system.
Arch Dis Child 2006; 91: 169-172.
14 Thompson AM, Ashraf Z, Burton H, Stonebridge PA. Mapping changes in surgical mortality over 9
years by peer review audit. Br J Surg 2005.92: 1449-1452.
15 McQuillan P, Pilkington S. Allan A, et al. Confidential inquiry into quality of care before admission
to intensive care. BMJ 1998; 316: 1853-1858.
16 Goldhill DR. Worthington L. Mulcahy A. et al. The patient-at-risk team: identifying and managing
seriously ill ward patients. Anaesthesia 1999; 54: 853-860.
17 Stenhouse C. Coates S. Tivey M, Allsop P. Prospective evaluation of a modified Early Warning
Score to aid earlier detection of patients developing critical illness on a surgical ward. Br J Anaesth
2000: 84: 663.
18 Buist MD. Moore GE. Bernard SA. et al. Effects of a medical emergency team on the reduction of
incidence of and mortality form unexpected cardiac arrests in hospital: preliminary study. BMJ 2002:
324: 387-390.
19 Wilson RM. Runciman WB. Gibberd RW. et al. The Quality in Australian Health Care Study
MedJAust 1995:163: 458-471.
20 Brook RH, McGlynn EA. Cleary PD. Measuring quality of care. N Engl J Med 1996; 335: 966-970.
21 Shojania KG. Wald H. Gross R. Understanding medical error and improving patient safety in the
inpatient setting. Med Clin N Am 2002; 86: 847-867.
22 Neale G. Woloshynowych M. Retrospective case record review: a blunt instrument that needs
sharpening Qual Saf Health Care 2003; 12: 2-3.
23 Thomas EJ. Studdert DM. Brennan TA. The reliability of medical record review for estimating
adverse event rates Ann Intern Med 2002; 136: 812-816.
24 Resar RK. Rozich JD. Simmonds T. Haraden CR A trigger tool to indentify adverse events in the
intensive care Lr»t Jt Comm J Qua! Saf2006; 32: 585-590.
25 Agency for Health Care Research and Quality. AHRQ's Patient Safety Initiative: building foundation*,
(educing risk. Interim report to the Senate Committee on Appropriations. Rockville. Md: AHRQ. (AHRQ
Publication No 04- RG005, 2003.)
26 Woloshynowych M, Neale G, Vincent C. Case record review of adverse events: a new approach.
Qual SafHetfh Care 2003; 12:411-415.
27 Michel P, Quenon JL. de Sarasqueta AM, Scemama O. Comparison of three methods tor estimating rata
of adverse events and rates of preventable adverse events in acute care hospitals. BMJ2004,328:19®-
202
 4 Detecting adveree everts and

28 Resar RK, Rozich JD, Classen D. Methodology and rationale for the measurement of harm with
trigger tools. Qua/ Saf Health Care 2003; 12 Suppl II: ii39-ii45.
29 Forster AJ, Anrade J. van Walraven C. Validation of a discharge summary term search method to
detect adverse events. J Am Med Inform Assoc 2005; 12: 200-206.
30 Murff HJ, Patel VL. Hripcsak G. Bates DW. Detecting adverse events for patient safety research: a
review of current methodologies. J Biomed Inform 2003; 36:131-143
31 Szekendi MK, Sullivan C, Bobb A, et al. Active surveillance using electronic triggers to detect adverse
events in hospitalised patients Qual Saf Health Care 2006; 15:184-190.
32 Sharek PJ, Horbar JD, Mason W. et al. Adverse events in the neonatal intensive care unit:
development, testing, and findings of an NICU-focused trigger tool to indentify harm in North American
NICUs. Pediatrics 2006: 118 1332-1340.
33 Rozich JD, Resar RK. Medication safety: one organisation's approach to the challenge. JCOM2001;
8:27-34.
34 Institute for Healthcare Improvement. Global Trigger Tool 2007.
www.ihi.org/IHI/Topics/PatientSafety/SafetyGen- eral/Tools/lntrotoTriggerToolsforldentifyingAEs.htm
(accessed Oct 2008).
35 McCannon CJ, Hackbarth AD. Griffin FA. Miles to go: an introduction to the 5 Million Lives
Campaign. Jt Comm J Qual Patient Sat 2007; 33: 477-484.
36 Australian Patient Safety Foundation. What is incident reporting? The Australian Incident
Monitoring Study. Adelaide: APSF, 1997.
37 Kohn LT. Corrigan JM. Donaldson MS. To error is human: building a safer health system.
Washington, DC: National
Academy Press, 1999.
38 Heinrich HW. Industrial accident prevention: a scientific approach. New York: McGraw-Hill. 1941.
39 Wolff AM, Bourke J. Campbell IA. Leembruggen DW. Detecting and reducing hospital adverse
events: outcomes of the Wimmera clinical risk management program. MedJAust200'\: 174:621-625.
40 Olsen S. Neale G, Schwab K. et al. Hospital staff should use more than one method to detect
adverse events and potential adverse events: incident reporting, pharmacist surveillance and local real
time record review may all have a place. Qual Saf Health Care 2007; 16:40-44.
41 Cullen DJ, Bates DW, Small SD. et al. The incident reponing system does not detect adverse
events: a problem for quality improvement. Jt Comm J Qual Improv 1995; 21:541-548.
42 Evans SM, Smith BJ, Esterman A. Runciman WB. Evaluation of an intervention aimed at improving
voluntary incident reporting in hospitals. Qual Saf Health Care 2007; 16: 169-175.
43 Kingston MJ, Evans SM. Smith BJ. Berry JG. Altitudes of doctors and nurses towards incident
reporting: a qualitative analysis. Med J Aust 2004; 181:36-39.
44 Spigelman AD. Swan J. Review of the Australian Incident Monitoring System. ANZJSurg 2005;
75:657-661.
45 Pronovost PJ, Thompson DA. Holzmueller CG, et al. Defining and measuring patient safety. Crit
Cere Chn 2006: 21: 1-19.
46 Berenholtz SM, Pronovost PJ. Monitoring patient safety. Crit Care Clin 2007; 23: 659-673.
47 Stanhope N, Crowley-Murphy M. Vincent C, et al. An evaluation of adverse incident reporting. J Eval
Chn Pract 1999; 5: 5-12.
48 O'Dowd A. Adverse incidents in NHS are still under-reported. BMJ2006; 333: 59.
49 Bion JF, Heffner JE. Challenges in the care of the acutely ill. Lancet 2004; 363: 970-977
50 Evans SM, Berry JG. Smith BJ. et al. Attitudes and barriers to incident reporting: a collaborative
hospital study Qual Saf Health Care 2006; 15: 39-43.
51 Runciman W, Merry A, McCall Smith A Improving patients' safety by gathering information. BMJ
2001; 323 298
Enhancing
52 Sharmapatient P. Ray B, et al. Incident reporting in surgical trainees. Ann R Coll Surg Engl 2004; 86 Suppl: 158- 159.
A. Jain care
53 Furman C. Captan R. Applying the Toyota production system: using a patient safety alert system to reduce error. Jt
Comm J Oual Irr^xov 2007; 33:376-386.
54 Leape L. Interview Is hospital patient care becoming safer? A conversation with Lucian Leape. Health Aff (km*ood) 2007;
26: w687-w696
55 Bolsin SN. Faunce T. Colson M. Using portable digital technology for clinical care and critical incidents: a new model.
Aust Health Rev 2005; 29: 297-305.
56 Freestone L. Bolsin SN, Colson M. et al. Voluntary incident reporting by anaesthetic trainees in an Australian hospital. Int
J Oual Health Care 2006; 18: 452-457.
57 Tighe CM, Woloshynowych M, Brown R, et al. Incident reporting in one UK accident and emergency department. Ac&d
Emerg Nurs 2006; 14: 27-37.
58 National Patient Safety Agency. Annual report 2006107. London: NPSA, 2007.
59 Pronovost PJ. Miller MR. Wachter RM. Tracking progress in patient safety: an elusive target. JAMA 2006:296:696- 699.
60 01.eary M, Chappell SL. Confidential incident reporting systems create vital awareness of safety problems. ICAOJ 1996;
51: 11-13.
61-Barach P, Small SO. Reporting and preventing medical mishaps: lessons from non-medical near miss reporting systems.
BMJ 2000; 320: 759-763.
62 Runciman WB. Lessons from the Australian Patient Safety Foundation: setting up a national patient safety surveillance
system — is this the right model? Oual Saf Health Care 2002:11:246-251.
63 Victorian Department of Human Services. Root Cause Analysts Education Module 1 and 2 What's in it for you?
www.health.vic.gov.au/clinrisk/downloads/rca_education_mod12.ppt (accessed Oct 2008).
64 Jagsi R, Kitch BT, Weinstein DF, et al. Residents report on adverse events and their causes. Ar.ch Intern Med2005; 165:
2607-2613.
65 Leape L. Reporting of adverse events. N Engl J Med 2002; 347:1633-1638.
66 Joint Commission on Accreditation of Healthcare Organizations. Comprehensive Accreditation Manual for Hospitals.
Oakbrook Terrace, III: Joint Commission Resources. 2005.
67 Australian Health Ministers' Conference. Health Ministers agree to reform agenda. Joint communique 2004; 23 Apr.
Canberra: Australian Government Department of Health and Ageing. 2004. www.health.gov.au/internetymain/
publishing.nsf/Content/health-mediarel-yr2004-jointcom-jc001.htm (accessed Jan 2009).
68 Victorian Department of Human Services. Acute Health Division Clinical Risk Management Strategy 2001. Melbourne:
Victorian Department of Human Services, 2001.
69 Australian Institute of Health and Welfare and the Australian Commission on Safety and Quality in Health Care. Sentinel
events in Australian public hospitals 2004-05. Canberra: AIHW, July 2007.
70 Forster AJ, Clark HD, Menard A, et al. Adverse events among medical patients after discharge from hospital. CMAJ200A.
170: 345-349.
71 Fitzpatrick R. Surveys of patient satisfaction: 1 — important general considerations. BMJ 1991; 302: 887-889.
72 Dunn KL. Moulden A. McDougall P, Bowes G. Patient safety: a view from down under. Pediatr Clin N Am 2006. 56: 1217-
1230.
73 Glabman M. The top ten malpractice claims. Hosp Health Netw2004; 78: 60-66.
74 Williams SK. Osborn SS. The development of the National Reporting and Learning System in England and Wales. Med J
Aust200B: 184 (10 Suppl): S65-S68.
75 Benveniste KA. Hibbert PD. Runciman WB. Violence in health care: the contribution of the Australian Patient Safety
Foundation to incident monitoring and analysis. Med J Aust 2005: 183: 348-351.
76 Hibbert P. Benveniste K. Burns received in hospitals. Insights from the Australian Incident Monitoring System Adelaide:
Australian Patient Safety Foundation. 2003.
77 Freer Y. Lyon A. Risk management, or just a different risk? Arch Dis Child Fetal Neonatal Ed 2006: 91: F327-F329.
78 Vincent CA, Lee ACH. Hanna GB. Patient safety alerts: a balance between evidence and action Arch Dis Child Fetal Neonatal
Ed 2006:91: F314-F315.
79 Thomson R. Luettel D. Healey F, Scobie S. Safer care for acutely ill patient: learning from terioue nodinK London: National
Patient Safety Agency. 2007. www.npsanhs.uk/nrls/alerls-and^ective6/dtrect^^ acutely-ill-patient/ (accessed Jan 2009).
80 Committee of Public Accounts. Fifty-first report: a safer place for patients: learning to improve patient aMy London:
United Kingdom Parliament, 2006: 6 Jul. wvw.parliament.uk/pariiamenlafy _committoos/oorrwTwBaejtif_
public_accounts/pac060706_pn5l elm (accessed Oct 2008).
61 Emslie S. 2001-2006: a patient safety odyssey where the vision remains unfulfilled. Health Care Risk Bap 2006: Sep: 14-15.
82 Bagian JP. Gosbee J. Lee CZ. et al. The Veterans Affairs root cause analysis system in action. Jt Comm J Que!
Improv 2002; 28:531-545.
83 Bagian JP. Lee C. Gosbee J. et al. Developing and deploying a patient safety program in a large health care delivery
system: you cant fix what you don't know about. Jt Comm J QuaI Improv 2001:27:522-532.
84 Bagian JP. Gosbee JW. Lee CZ. The VA-NASA Patient Safety Reporting System. Fed Pract 2001; Mar: 10-15
85 The Inquiry into the management ot care of children receiving complex heart surgery at the Bristol Royal Infirmary
Bristol: Bristol Royal Infirmary Inquiry. 2001. www.Bristol-inquiry.org.uk/ (accessed Oct 2008).
86 Final report of the Queensland Public Hospitals Commission of Inquiry. (The Davies report). Brisbane: Queensland
Government, Nov 2005. www.qphci.qld.gov.au (accessed Oct 2006).

92
 87 Special Commission of Inquiry into Campbelltown and Cambden hospitals. Final report. Sydney: NSW Department of
Health, 2004.
88 Douglas N, Robinson J, Fahey K. Inquiry into obstetric and gynaecological services at King Edward Memorial Hospital
1990-2000. Final report. Perth: Government of Western Australia. 2001.
89 Community and Health Services Complaints Commissioner of the ACT. A final report of the investigation into adverse
patient outcomes of neurosurgical services provided by the Canberra Hospital. Canberra: ACT Government. 2003.
90 Faunce TA, Bolsin SNC. Three Australian whistleblowing sagas: lessons for internal and external regulation. Med
JAust2004, 181: 44-47.
91 Dunbar JA. Reddy P, Beresford B, et al; In the wake of hospital inquiries: impact on staff and safety. MedJAust
2007; 187: 476-477.
92 Walshe K. Higgins J. The use and impact of inquiries in the NHS. BMJ2002.325:895-900.
93 Ham C. Quality failures in the NHS. SH/2008; 336: 340-341.
94 Healthcare Commission. Learning from investigations. London: HC. 2008.
95 DeRosier J, Stalhandske E. Bagian J. Nudell T. Using Health Care Failure Mode and Effect Analysis™: the VA National
Center lor Patient Safety 's prospective risk analysis system. Jt Comm J Qual Improv2002: 28: 248-267.
96 Wollersheim H. Clinical incidents and risk prevention. Neth J Med 2007; 65: 49-54.
97 Green AL, Williams A. An evaluation of an early warning clinical marker referral tool. Intensive Crit Cam Nurs 2006; 22:
274-282.
98 Vincent CA. Coulter A. Patient safety: what about the patient? Qual Saf Health Care 2002:11: 76-80.
99 Davis RE, Jacklin R. Sevdalis N. Vincent CA. Patient involvement in patient safety: what factors influence patient
participation and engagement? Health Expect 2007:10: 259-267.
100 Stevens P. Mallow A. Laxer RM. Blueprint for patient safety. Pediatr Clin North Am 2006: 53:1253-1267. □
Enhancing patient care

5 Analysing adverse events and prioritising action

This chapter will:
■ Discuss management of patients who have experienced an adverse event and staff who have bean involved in an adverse event
■ Examine the analysis of adverse events using systems and root-cause analysis
■ Discuss the determination of the level of risk of an adverse event
■ Describe the development of a risk register
■ List the criteria against which adverse events can be prioritised for action
■ Examine the composition and activities of a surveillance committee that reviews adverse events.

After an adverse event is detected, the usual next stage in the risk management process is to analyse the event. This
helps the staff involved in the incident, and the health service, leam from the event and take appropriate action to reduce
the probability of it recurring. Before analysis is commenced, two critical checks are required:
■ Have the patient and their relatives been appropriately managed following the incident?
■ Have the clinical staff involved in the incident been appropriately managed?
Many adverse events have considerable negative effects on the patient, their relatives, and the staff who provided the care.
It is imponant to ensure all three groups have been appropriately managed. If so, the event can be analysed in detail and
action to prevent a recurrence prioritised.
It is now believed that "open, honest and timely disclosure should be the only approach to medical error". 1 However, the
issue is "multidimensional, emergent and complex". 2 Previously, it was thought that open disclosure (including admitting
mistakes to peers and talking openly and apologising to patients) was too difficult and would increase the nsk of litigation
and tarnish the reputation and financial viability of individual clinicians and health services. 2 The decision whether to
disclose information about errors to patients was individual and did not occur frequently. When it did occur, it was not
done systematically or with organisational suppon, and was often not handled satisfactorily.
Much of the anger expressed by patients and their relatives about errors related to how they were treated after the
event, rather than the event itself. They understood that mistakes happen, but expected an honest, human response from
their clinicians after they occurred. The fear experienced by clinicians and health services after an error in patient care
was often worse than what followed in reality.
Although open disclosure has been practised in very few health services, there are limited high-quality data and some
anecdotal evidence that open, honest and timely disclosure may decrease the risk of litigation and result in better
outcomes for patients, clinicians and health services. Open disclosure is now a standard of Australian Council on
Healthcare Standards and the Joint Commission on the Accreditation of Healthcare Organizations in the United States. 1
The Australian Council for the Safety and Quality in Health Care developed the National Standard on Open Disclosure,
which was endorsed by the Australian Health Ministers Conference in July 2003. The Standard includes an apology, a
factual explanation of what occurred, the consequences of the event, how it was managed and the action taken to prevent
a recurrence. Currently, 42 Australian health services are participating in a national evaluation of open disclosure. An
evaluation by the 12 participating Victorian health services found, among other things:
■ there was no standardised approach to open disclosure within the sites
■ if there was a strong culture of quality and safety in an individual health service, open disclosure was more readily
adopted
Enhancing patient care

■ open disclosure complements existing risk management strategies

■ overall the documentation of open disclosure discussions was poor
■ the term "open disclosure" was felt to be negative.
The recommendations of the evaluation included that a standardised
education program and toolkit be developed, open disclosure be integrated
into each health service's clinical risk management policy, and a more
meaningful name for the process be developed. The Victorian Depanment
of Human Services plans to roll out the open disclosure program to all
health services in the state.3

5.1 Managing a patient who has experienced an

adverse event
Patients who are unwell or are in hospital to undergo procedures may feel
vulnerable and traumatised even after successfully undergoing straightfor-
ward procedures. If harm occurs to a patient receiving medical care, it
may cause a severe reaction because of the event or how clinical staff
managed it. If the event is handled insensitively, the patient may
experience more trauma than if the staff acknowledged the event
occurred, were open about what had happened and why, and explained
the action that was taken or planned to be taken to prevent other patients
experiencing a similar event. It is particularly difficult for patients if they
must continue to be treated by the staff responsible for the event. 4
Patients and their relatives may fear further harm or retribution from
those caring for them, and clinicians may turn away from them. 5
After patients experience an adverse event, they may feel guilty, fearful,
distrusting and isolated. 4,5 Their experience is similar to patients with
posttraumatic stress disorder. They may experience flashbacks of the
event, anxiety, and depression, all of which may affect their relationships
and work. Their relatives may also have difficulties coping, especially if the
patient dies as a result of the event, and the death is thought to be
avoidable.4 Relatives often experience similar, or even stronger feelings of
guilt (about not watching closely enough), than the clinicians treating the
patient. They also fear that confronting clinicians may result in further
injury.5
If patients believe, correctly or incorrectly, that they have experienced
T^opte bate*) that you an adverse event as a result of medical management, the clinician
let a patient about am* providing the care should take their concerns seriously. It may add to the
and they make tao
patient's distress if these concerns are ignored. 4 Relatives of patients who
phone cats. One is to
fie press, the other have been harmed by a medical error have said "what we needed was for
totoeir lawyer. It does not someone to reach out and connect with us in human terms ". 5 When
work that way. * responding to such concerns, clinicians should be sympathetic, open and
Conway J. et al Qual honest. They should provide a full explanation to the patient of what has
Sal Health Care 2004; occurred, apologise if appropriate, and inform them of the steps being
13: 3-5
taken to prevent similar events occurring to others. If clinicians ignore
 5 Analysing adverse events and prioritising action

these actions, they may add to the patient's distress, and the patient or
relatives may make a formal complaint to the
health service or the relevant professional registration body, or commence litigation,4 as "silence
and evasion breed distrust".5
Although clinicians may understandably feel uncomfortable about their relationship with these
patients and wish to distance themselves, such patients require more, not less, suppon. Clinicians
should ask their patients how the adverse event has affected them, physically and psychologically. If
the patient is severely affected, formal psychological treatment may be needed, and the patient
should be offered the option of referral to another service.4
If medical mismanagement has clearly occurred, the health service should inform their insurers
and discuss with them the possibility of providing immediate financial suppon for the patient and
their dependants. Such suppon may initially pay the patient's medical and hospital costs for any
additional treatment and hospital stay required as a result of the fiarm experienced. These
responses to adverse events should be strongly and explicitly supported by written policies and by
detailing procedures that should be followed when a patient has been harmed while receiving care in
the health service.

5.2 Managing a staff member who has been involved in an

adverse event
The staff involved in providing the care that resulted in an adverse event and significant harm also
need support. Many physicians experience significant distress and job-related stress following a
serious error or near miss.4
Clinicians ... often suffer alone after making mistakes, agonising over the harm they have
caused, the loss of their patients' trust, the loss of rheir colleagues' respect, their diminished
self-confidence, and
the potential effects of the error on their careers. 5
Such feelings often remain unaddressed, and these physicians feel inadequately supported by their
health service.6
As well as fearing sanctions, complaints and litigation affecting their reputation, these staff may
be feeling guilt, shame, anxiety, depression and isolation. Some may feel like giving up their clinical
practice or leaving the profession. These effects can be so severe that this group has been described
as the "second victim" of an error.7 Clinicians are often self-critical and set high personal standards
in providing care. Consequently, when they make a mistake, they can be extremely hard on
themselves. Some may respond dysfunctionally with anger and projection of blame onto other staff
or patients; this may become more extreme with litigation. 7
The culture in clinical practice expects perfection in providing care. Mistakes are unacceptable'
and. if made, reflect a weakness in the clinician's character. These views are unrealistic and weigh
heavily on some clinicians. After a significant adverse event has occurred, the clinicians involved
often require support from their colleagues, and psychological intervention if their reaction to the
event is severe. Such support and intervention are rarely forthcoming 7 and, even if provided, may be
difficult for some doctors to accept. 4 Colleagues should encourage the clinician to talk with them
about what has occurred and inquire about the psychological effect of the error and how they are
coping.'
Given the significant impact major errors can have on clinicians, health services should consider
providing formal and informal methods of supporting clinicians after their involvement in serious
adverse events or near misses. 6 Given the increasing complexity and demands of modem medicine,
clinicians will continue to make errors. How the health service and its clinicians deals with patients,
relatives and professional staff after an error has occurred can be decisive in its outcome."
5.3 Analysis of adverse events
Analysis of adverse events is well established in other complex industries,
such as aviation and nuclear power. In health care, the analysis of
adverse events can take several forms. In the US, the Joint Commission
uses root- cause analysis (described later in this chapter) to analyse
sentinel events; the Veterans Health Administration has a similar process
using triage questions to analyse frequent and catastrophic events in its
facilities. These methods are detailed, resource intensive and time-
consuming.
In the case of adverse events detected by medical record review,
sufficient information can be often obtained from the review form
(discussed in Chapter 4). If required, this information can be
supplemented by interviews to clarify aspects of care. 8

5.3. I Systems analysis

If a detailed analysis is required, systems analysis — a less intensive
method of analysing adverse events than root-cause analysis — has been
developed.4 This approach combines Reason's model of organisational
accidents (described in Chapter 3), which was developed for use in
complex industrial systems, with the factors affecting how clinical care is
provided. In the systems analysis model, accidents result from an often
complex series of events and departures from safe practice, and are
associated with a wide range of contributory factors. System analysis aims
to find system weaknesses to explain why an event occurred.
When adverse events occur, the media often concentrate on the actions
taken by the doctors and nurses involved in the event. However, research
has shown that the probability of an error occurring in the workplace
increases if the conditions of work are adverse or suboptimal. Reason's
model of organisational accidents aims to identify and focus on pre-
existing organisational factors that have contributed to the adverse event
or near miss. Reason believes that adverse events and near misses are
preceded by an "unsafe act", where a staff member at the coalface makes
an error or mistake.
Importantly, these errors are preceded by error-producing conditions
and latent factors, which are together referred to as contributory factors.
These factors are the conditions under which staff were working when the
adverse event occurred. Many contributory factors are the result of
decisions made by management that have negatively affected the
conditions of work in the organisation and increased the likelihood that
errors will be made. In systems analysis, error-producing and
organisational factors have been combined into a comprehensive
framework of contributory factors that affect how care is provided. By
determining and understanding these contributory factors, appropriate
action can be planned and taken to improve the conditions under which
care is provided, thereby improving the quality of care and level of patient
safety.
 The systems analysis of an adverse event commences with a summary
YMierlhansarrpV of the event being made in chronological order. Three steps are then
ass&nring biame, undertaken:
patients and tamiies
want both to
understand their
s^jeHon fully and to
know what the event
has taught caregivers
and their institutions."
Delbanco T. Bell SK
N Engl J Med 2007;
357: 1682-1683
1. The patient's care management problems are identified. These problems are actions or
omissions made by staff such as mistakes, lapses of judgement or deliberate departures
from the policies, procedure* and standards set by health services, which led to the
adverse event. Example include: (ailing to monitor a patient, not obtaining assistance
when needed, delay in making a diagnosis, failing to adequately supervise junior staff, and
providing the wrong treatment to a patient. There may be more than one care
management problem associated with an adverse event; each is identified and individually
analysed.
2. Relevant clinical events and the patient's condition are recorded to provide the clinical context
of the event. Examples include; the patient was extremely short of breath, bleeding
profusely, his or her conscious state was deteriorating, or blood pressure was falling.
3. The factors that resulted in the care management problem are identified. These contributory
factors are grouped under seven headings — institutional (eg, the medicolegal environment
prevents open disclosure of errors to patients and their relatives); organisational and
management (eg, financial constraints delay replacement of ageing equipment); work
environment (eg, staff have excessive patient workloads, there are insufficient staff); team
(eg, inadequate supervision of junior doctors); individual staff member (eg, a doctor lacks
certain procedural skills); task (eg, lack of a protocol to treat patients with a particular
condition, investigation results not available in a timely manner); and patient (eg, multiple
and complex clinical problems, communication problems). A combination of contributory
factors will frequently be identified when analysing a single adverse event. The boors
identified may be specific for that adverse event or recurrent and longstanding, reflecting a
systemic problem that affects patient safety in a clinical area or the whole service.
Depending on the complexity and seriousness of the adverse event or near miss, the analysis
can be undertaken by the relevant clinicians, the clinical risk manager or a multidisciplinary
team. A two-page summary (available from Department of Psychology at University College
London at wwwpatient- safety.ucl.ac.uk/Caseanalysisl.pdf and
www.patientsafety.ucl.ac.uk/Caseanalysis2.pdO and a forma) protocol (second edition available
from the Clinical Safety Research Unit of Imperial College London at
www.csru.org.uk/downloads/SACI.pdO have been developed so the investigation can be
systematically and comprehensively undertaken, and to reduce the chance that simplistic
explanations are developed that blame individuals for the event. Staff require training and
practice to effectively use the protocol.
The information required by the protocol can be obtained by reviewing the patient^ medical
record and discussing the event with key staff caring for the patient. A structured interview
using a checklist of psychological and organisational factors is used, and staff are asked their
opinions about the reasons for each care management problem. This information helps
establish the often complex sequence of events resulting in the adverse event, and helps
identify the main care management problem and contributing factors. Staff are also asked
about the impact of the adverse event on them, the patient and their relatives, and other staff.
The analysis aims to identify the most important contributory factor and to use this
information to improve the safety of the relevant delivery system. Four questions should be
answered by the analysis:
■ What happened?
■ How did it happen?
■ Why did it happen?
■ What can we learn from this and what changes should we make, if any?
It has also been suggested that adverse events should be discussed with the patients
involved and. if appropriate, their relatives. Both are potentially important sources of useful
information that have not previously been routinely used. Frequently, open discussion with
patients who have expenenced an adverse event and their relatives may not occur because the
clinician and health service fear that punitive action and litigation may result. The patient
should be informed of the results of the analysis of the adverse event, particularly what action
will be taken to prevent similar adverse events from occurring
 Analysis of adverse events and near misses may repeatedly find the same contributory
factors, indicating that certain delivery systems require strengthening. If analysis shows that
substantial system changes may be required to prevent an adverse event recurring, it may be
necessary to obtain additional information about the system under review from analyses of
previous similar adverse events and audits. Some analyses may reveal problems requiring
immediate action to reduce risks to patients.
After the analysis is completed, a formal repon should be written, detailing the care
management problems and contributory factors identified, the actions required to prevent the
event recurring, who is responsible for the implementation of these actions, and how the
outcomes of these actions will be monitored. A brief version of systems analysis can be used for
the immediate investigation of an adverse event. 9 In our experience, this may often be sufficient
to adequately deal with many of the adverse events that occur in a medium-sized general
hospital. If, after immediate investigation, it is found that the event is serious or has important
educational benefit, a full systems analysis should be undertaken.

5.3.2 Root-cause analysis

In the Veterans Health Administration, incident reporting system staff voluntarily report clinical
incidents to their facility's Patient Safety Manager. The Manager assesses the real and potential
frequency and consequence of the incident. Often, perhaps because of good fortune, the actual
consequence of an incident is minor, but it could have been catastrophic. For clinical incidents
scored as 3 or higher (ie, catastrophic, or major and frequent) using the Safety Assessment
Code, whether real or potential, the Administration requires a detailed root-cause analysis to be
undertaken. The analysis provides a structure to investigate the incident and answer three
questions:
■ What happened?
■ Why did it happen?
■ What can we do to prevent it happening again?
A multidisciplinary team of frontline staff undertakes the analysis. The team uses 78 triage
questions, available in card format on the Administration's website, and follow five rules of
causation to ensure that a sufficiently deep analysis of the causative factors is undertaken. The
questions cover: human factors, communication, training and fatigue or scheduling; environ -
ment and equipment; rules, policies and procedures; and barriers or safeguards. 10
After the multidisciplinary team has completed the root-cause analysis, it is electronically
ipnMQVSOiOR submitted to the Administration's National Center for Patient
•Hr awn wtfurea Safety. The expert staff at the Center review the analysis and
patients and ther provide rapid, detailed feedback to the reporting facility. This
kniK it may be
feedback includes comments on whether the elements of each
because of their own
taatngsolgum. fear, criterion have been met for the Safety Assessment Code scores and
ant isolation, whether they accurately reflect the description of the event, and
compounded by legal the adequacy of the flowchart, contributing factors, action
or institutional advice. statements and outcome measures. Reporting all root-cause
Paralyzed by shame or
analyses to the Center allows it to monitor the quality and
lacking their own
iMtderstanding of why completeness of all analyses
the error occurred,
physicians may find a
bedside conversation
too awkward.'
■ Detbanco T. Bell
SK N Engl J Med
2007;
357: 1682-1683
undertaken in the Administration's facilities. Further, the Center can rapidly alert other facilities of
hazards that have been identified if they require prompt action. The Centert staff can also provide a wide
range of technical expertise to reponing facilities regarding incident analysis, and provide reference
materials or solutions for similar incidents that have been developed at other Administration sites. 10
The findings of the root-cause analysis are communicated to the staff member who reported the incident
to demonstrate that the incident has been appropriately investigated. This staff member is able to comment
on the findings of the investigation and the proposed corrective actions. To ensure that appropriate action is
taken in response to the incident, the chief executive of the facility must approve the actions proposed by
the analysis team. If the chief executive wishes to modify the actions, he or she needs to provide the team
with reasons for the changes. This process may go through several cycles until there is final agreement
between the chief executive and the team. The chief executive then endorses the action plan, arranges
funding for its implementation and accepts responsibility that it will be implemented within 45 days. The
team then determines how the effectiveness of these actions and any unintended consequences can be
measured. Team members may be rewarded for their participation in the program with continuing
education credits."
The details of root-cause analyses are entered into a database that can be accessed by other
Administration facilities. In response to the incidents reported centrally, system-wide alerts and advisories
are sent by the Center to all Administration facilities in the form of shon safety summaries and are posted
on the their website. Sometimes, several months after an alert has been sent to all facilities, a follow-up is
conducted by email to determine whether the alert has successfully reached appropriate staff at the
frontline of each facility and whether operational changes have been made as a result of the facility
receiving the alert.

5.3.2.1 Limitations of root<ause analysis

Root-cause analysis is an important tool in the assessment of major accidents in complex industries such
as aviation and nuclear power. Major accidents occur infrequently in these industries; however,
epidemiological studies of adverse events in health services in several countries show that major adverse
events occur more frequently in health care. 1213 However, the effectiveness of root-cause analysis is
limited,14 because:
a Analyses are often incompletely or incorrectly undertaken, and their quality varies widely between health
services.
a Undertaking a comprehensive analysis to determine the root causes of an adverse event is time-
consuming, requiring 20-90 hours of staff time, a The commonest actions taken in response to analyses are
re-education and writing a policy; these are unlikely to reduce the risk and have little effect outside the
individual health service. The causes of incidents are often common across health services and high-
probability actions often require system or equipment redesign at a higher level than individual health
services, a After action is taken in response to the risks identified by analyses, follow-up on the outcomes of
actions taken is rare, so it is not known whether the risk has been reduced, a It is difficult lo establish a
root cause, as adverse events occur because of "varied, diffuse and overlapping failures". 15
a As with other methods of analysis of adverse events, hindsight bias may influence their assessment . 11,
Berwick believes that "the very idea of a 'root cause' is misleading, because most system failures result from
complex interactions between latent failures ... and specific actions; conclusions about root causes are often
illusions created by hindsight bias".1'
Comprehensive root-cause analysis programs have been established in most Australian states and
territories. These programs are supported by detailed education and documentation packages Information
is available online about programs in Victoria (www.hcallhAic.gov.au/clinrisk/sminclAca.himV South
Australia (.www.safctyandquaHty.sa.gov.au/safetyandquality), and Queensland (www.health.qld.gov.au/
patientsafetyAm/webpages/whatisrca.asp).

SJ.3 The Learning from Defects tool

Referred to as a "lighter" version of root-cause analysis, the Learning From Defects tool uses the
contributing factors identified by systems analysis. 18 An important addition in the Tool is documentation of
factors that prevented or reduced harm from the error. The Tool is divided into three sections:
1. An explanation of what happened.
2. A review and check of all the factors that caused or increased patient harm and reduced or eliminated
harm.
3. Specific actions to reduce the probability of
the defect recurring, the designation of a ___________ _________
person to lead the actions being taken, specification of dates for the assessment of the progress of the
actions, their completion date, and details of how an evaluation will be undertaken to determine if the risk
has been reduced.
All stafT involved in providing care related to the defect are present when the defect is analysed and a case
summary learning tool is prepared for distribution to relevant staff. A one-page user guide for the tool has
been developed.18
Whichever method is chosen to investigate adverse events, health services should develop appropriate
policies and procedures that include first and foremost how the needs of the patient, their relatives and
the staff involved in the incident will be addressed. The policy should also detail how the incident will be
investigated, recommendations will be made, appropriate action will be taken, and ongoing monitoring of
clinical activity will be undertaken. A policy on disclosure of details of adverse events, to help clinicians
talk with patients and their relatives, should also be developed. 19

5.3.4 Analysis of adverse events detected by medical record review at Wimmera Health Care Group
At Wimmera Health Care Group, we use a relatively brief technique to analyse adverse events detected by
medical review. After a medical reviewer detects an adverse event, the details are recorded on an adverse
event analysis form (Appendix 1) The medical reviewer determines the sevepty and preventability of the
adverse

Box 5.1 Quality in Australian Health Care Study error

cause scale20
Treatment
T1 = Complication/failure in technical performance
T2 = Failure to follow policy, protocol or guideline when
indicated
T3 = Acting on insufficient information
T4 = Slip/lapse in activities in which operator is skilled
T5 = Inadequate knowledge/experience
T6 = Inappropriate or not indicated care
T7 = No/wrong/delayed treatment
Diagnosis
D1 = Failure synthesise/decide/act on available
information
D2 = Failure request/arrange procedure or consultation

D3 = Missed diagnosis System

51 = Lack of care or attention/failure to attend when
called
52 = Failure in communication
53 = Equipment failure/unavailability
54 = Inadequate staff
55 = Staff latigue/distress/sickness
56 = Poor discharge planning
Prevention
P1 = Inadequate (ollow-up ol treatment/monitoring
P2 = Failure to provide indicated prophylactic
treatment.
 Box 5.3 Use deAnMom
Underuse: Failure to provide a health care servioe when it would have produced a favourable outcome for a patient.
Overuse: A health care service is provided under circumstances in which its potential for harm exceeds the possible benefit
Misuse: An appropriate health service is selected, but a preventable complication occurs and the patient does not receive the full
potential benefits
of the service.

event using the scales outlined in Chapter 4. The reviewer then documents the latent factors and active
errors that were present and that may have contributed to the adverse event occurring. It can be difficult to
classify precisely all the latent factors and active errors associated with every adverse event that is detected.
We found the framework used by the Quality in Australian Health Care Study for analysing errors enabled
us to classify the factors associated with most events (Box 5.1). 20 Another classification of errors was used in
the Harvard Medical Practice Study21 (Box 5.2).
A funher classification of adverse events into those associated with underuse, overuse and misuse is then
made. These terms are defined in Box 5.3.
This classification is useful because actions taken to reduce adverse events associated with overuse and
misuse if successful, will result in cost savings. A reduction in adverse events associated with underuse will,
however, increase costs. For example, the use of low molecular weight heparin perioperatively will reduce
the incidence of deep vein thrombosis and pulmonary embolus. These adverse events may occur more
frequently if low molecular weight heparin is not used consistently (ie, it is being underused), but its
appropriate use will increase pharmaceutical costs.

5.4 Prioritising adverse events for action

Analysis of adverse events is an important part of the clinical risk management process. In our experience,
analysis of adverse events and the systems in which they occur, together with the planning of appropriate
action to prevent recurrences, has the potential to take some lime, delaying prompt action to prevent the
recurrence of particular events. A practical balance is required between these two factors

5.4.I Why adverse events should be prioritised

The number of adverse events detected by clinical risk
management programs will depend on the number of
Box 5.2 Harvard Medical Practice Study types of
errors21 detection methods used. In our experience, using all
detection methods outlined in the
Diagnostic
Error or delay in diagnosis
Failure to employ indicated tests
Use of outmoded tests or therapy
Failure to act on results of monitoring or testing
Treatment
Error in performance of an operation, procedure, or test
Error in administering treatment
Error in the dose or method of using a drug
Avoidable delay in treatment or in responding to
an abnormal test
Inappropriate (not indicated) care
Preventive
Failure to provide prophylactic treatment Inadequate

monitoring or follow-up of treatment Other

Failure of communication Equipment failure Other
system failure.
 lhbleS.1 Measures of probability Uml Daecriplor Description

5 Almost certain Is already occurring or is very likely to occur several times in the next 12 months
4 Likely Will probably occur several times in the next 1 to 2 years
3 Possible Might occur once during the next 1 to 2 years
2 Unlikely Could occur at some time over the next 5 years
1 Rare May occur in exception circumstances, sometime in 5 to 30 years

Itt lie 5.2 Measures of consequence

U mel Descriptor Description

5 Catastrophic Death, huge financial loss or threat to goodwill

4 Major Major permanent injury, increased length of stay or re-admission, major financial loss

3 Moderate Major temporary injury, increased length of stay or re-admission, medium financial loss

2 Minor Minor treatment required, no increased length of stay or re-admission, minor financial loss

1 Insignificant No injuries, low financial loss

Tabic 5.3 Wimmera Health Care Group risk assessment matrix

Probability

8 Almost certain (5) Ukely (4) Possible (3) Unlikely (2) Rare(1)
c• Catastrophic (5) Extreme Extreme Extreme Extreme High

5| Major (4) Extreme Extreme Extreme High Moderate

co Extreme High Moderate Moderate Low
Moderate (3)
o Minor (2) High Moderate Moderate Low Low

Insignificant (1) Moderate Moderate Low Low Low

Wimmera model will detect more adverse events than can be effectively dealt with at any one lime. Even
with sufficient resources, it may not be possible to simultaneously make all the system changes required to
address the subsequent issues for change and improvement raised by the analysis of an adverse event.
Therefore, the adverse events detected must be prioritised systematically. Different methods of
prioritisation are discussed later in this chapter.

5.4.2 Assessing the level of risk

The level of risk associated with each adverse event that is detected is assessed. The probability or
likelihood of the adverse event occurring and the consequence of the adverse event are estimated using
quantitative and qualitative measurement scales. At Wimmera Health Care Group, we use measures
modified, from AS/NZS 4360: 200422 (Tables 5 1, 5.2 and 5.3).
 5 Analysing adverse events and
prioritising action

A level of risk for each adverse event is then determined by multiplying the likelihood of the event by its
consequence. This calculation provides a score for the level of risk associated with each event, and allows
adverse events to be categorised as low, medium, high, or extreme risk, and then to be ranked.

5.4.3 The risk register

Ranking adverse events allows a risk register to be created. The register enables adverse events with the
greatest level of risk to patients or the health service to be seen easily, and allows the systems in a hospital
with the greatest potential for improvement to be identified. The register also provides a useful method of
reporting the risks facing the organisation to senior executives, clinical stafT, the hospital's quality
committee and board of management.
Examples of the types of adverse events that may appear on a risk register report are shown in Table 5.4.
As more adverse events and risks are detected, the register will become longer. Maintaining the register in
an electronic format helps keep track of the increasing number of A'ents and their associated
recommendations.

5.5 The surveillance committee

At Wimmera Health Care Group, a surveillance committee reviews the information contained in the risk
register. When the project began in 1989, this committee consisted entirely of medical reviewers, and the
committee's recommendations regarding patient care were sent to the medical staff for discussion and
ratification. Coordination of the medical record review process and the implementation of the committee's
recommendations were undertaken by the director of medical services.
At this time, it was felt that if a full and frank discussion of adverse events was to occur, only medical
staff could be present at the surveillance committee meetings. Medical stafT believed that the presence of
other clinical staff at these meetings would inhibit an open discussion of errors. Occasionally, senior
resident medical staff were invited to the surveillance committee's meetings to provide them with some
experience in clinical quality improvement and risk management. From time to time, other specialists and
general practitioners were invited to assist with particular clinical issues facing the committee, and to
demonstrate how the medical record review process and committee's activities were conducted.
It is difficult to change clinical practice across professional boundaries without the participation of key-
clinical staff from a cross section of different professions. By restricting the membership of the surveillance
committee to medical staff, other members of stafT, especially nurses, were not pan of the discussions that
analysed and determined the appropriate actions to take in response to individual adverse events that were
detected. This membership restriction occurred despite nurses being an essential part of the hospital's
health care delivery systems. Consequently, the restriction limited the effectiveness of surveillance
committee by reducing the range and types of issues that could be thoroughly and comprehensively
discussed at meetings and the possible changes that could be successfully made to improve delivery
systems.
In 1997, the Department of Human Services in Victoria chose Wimmera Health Care Group to be one of
four hospital clinical risk management program pilot sites in Victoria. This decision and the associated
additional funding enabled the hospital to appoint a full-time clinical risk manager. A senior nurse with
intensive care and nurse unit manager experience was appointed to the position. The clinical risk manager
became a member of the surveillance committee and a nurse unit manager of an acute ward was also
invited to attend committee meetings.
After these membership changes, the range of issues discussed by the committee increased
Recommendations for changes to clinical care made by the surveillance committee were then taken

106
 Ifcbfe 5.4 Rkk register
9smple risk register/action plan

Rlsk/hazardfexposure to Risk ranking

loss; opportunity lor -----------------------
Source Improvement P C S R

Poor communication of abnormal

radiology results by external service
DM risk provider
Wmttfte
Feb Lack of security for N20 5 3 15 E
d
200
8

Current falls risk

Sep ED assessment scoring 5 3 15 E
200 medica system does not score patients
7 l presenting post- fall as "high risk" for
record falls
review
Oct Inadequate assessment of 2 3 10 H
200 Extern pain
7 al
report

Abdominal
pain incorrectly diagnosed

Inadequate clinical handover

between covering medical units and
home medical units
Inadequate analgesia for patients 428M
discharged from day procedure
Risk pereotV Expected Dele: May 2000
reduction departm end date Measurable outcome
strategy; ent
continuous Improvement

Contract r ient
negotiatio i s
ns to s adm
include k itted
process follo
for a win
notificatio s ga
n of s fall
abnormal e are
results s auto
Purpose- s mati
built m call
cupboard e y
placed in n ass
obstetrics t ess
unit. ed
Mobile f as
N20 o "hig
cylinders r h
kept in m risk"
locked for
drug s falls
room. o Impl
Security eme
increased t ntati
in h on
medical a of
gas store t Vict
area oria
p n
Modify
a Qua
the falls
t lity
 5 Analysing adverse
events and
prioritising action

Council's ' r
Acute H i
Pain o n
Manage w g
ment ;
Measure t
ment o N
Toolkit u
m r
Clinical a s
pathway n e
develope a
d for the g U
diagnosis e n
and i
manage p t
ment of a
abdomin i
M
al pain n
a
Develop n
a
and a
t
implemen g
ta e
h
process r
o
for
m
clinical O
e
handover &
"
between Ex G
medical ec ;
staff after uti
hours ve N
Patient u
informati r
on s
brochure En e
develope gin
d on ee U
 nit P Serv
ice Jun
Man h
prov 200
ager a
Enhancing patient care ider 8
ED r cont
m ract
Falls a incl
c Dec
Risk ude
y 200
Com s
/ 8
mitte proc
D ess
e a for
y notif
icati
P on Dec
Clini r 200
of
cal o 8
abn
Risk c
orm
Man e
ager
al
d
radi
u Jun
olog
r 200
y
e
resu 8
Clini
lts
cal U
Risk n NO
Man i stor
t age Sep
ager
s 200
are 8
Dire secu
ctor re
Aug
of
200
Medi
8
cal
Serv
ices
Jun
200
8
 5 Analysing adverse
events and
prioritising action

The C es
number o (clin
of m ical
patients p path
admitted l way
post-fall i audi
who are a t)
assessed n
as "high c Staff
risk" for e com
falls plian
(medical w ce
record i and
audit) t satis
h facti
Number
of on
p with
patients a
whose han
t dov
pain is h er
assessed w proc
according a ess
to pain y (staf
measure
f
ment p surv
policy and r ey)
protocol o
(medical Pati
c
record ent
e
audit) satis
s
facti
s
on
m with
e anal
a gesi
s a
u
r
 informatio
n
(nalip
Enhancing patient care nl
 5 Analysing adverse events
and prioritising action

directly to the hospital's senior medical staff group and nursing practice committee.
Recommendations resulting from high and extreme organisational risks were taken to the
organisation's peak quality committee and board of management. Surveillance committee members
gave the reasons for each of the recommendations made in person to medical and nursing staff. In
addition, the clinical risk manager attended regular allied health staff meetings to provide details
about adverse events relevant to their clinical areas, and explaining the background behind recent
recommendations made by the surveillance committee. The clinical risk manager also obtained
details from allied health staff about adverse events that may have occurred recently in their
departments and any areas where they thought patients or the hospital may have been at risk. After
these changes to the membership of the surveillance com miner were made, the recommendations
for change made in response to adverse events covered the aspects of the hospital's health care
delivery systems more broadly.
More recently, with the increasing use of clinical pathways to implement changes to clinical caie in
the hospital, the hospital's clinical pathway coordinator was invited to be ^member of the
surveillance committee. This has helped the coordinator to design clinical pathways that reduce the
probability of these adverse events occurring.

5.5.1 Composition of the committee

The current membership of the surveillance committee
is: a visiting surgeon (chairperson) a director of medical
services a visiting physician (also director of intensive
care)
a visiting general practitioner (also director of postgraduate medical
education) a nurse unit manager (acute medical/surgical ward) a clinical
risk manager a clinical pathway coordinator.
Because of the large number of adverse events related to medication errors, some health services
have added a pharmacist to their surveillance committees, 8 and suggestions have been made that
consumers should also be part of the committee.
If a surveillance committee is established in a health service, it is important to select an
appropriate name for the committee that appears neither punitive nor exclusive. Our surveillance
committee is called the "Patient Care Committee", as this focuses on what we are trying to achieve —
high-quality, safe patient care. A more specific name such as the "Medical Audit Committee" may
project an image of a medical committee that judges medical practice and has no input from other
disciplines, such as nursing.

5.5.2 Statutory Immunity

For the surveillance committee to be effective, clinicians need to participate in full, frank discussions
of quality issues. Such discussion is promoted if participating clinicians are confident that the
content of such discussion will not be made available outside the committee. In individual Victorian
health services, statutory immunity is available from the Minister for Health to services under
Section 139 of the Health Services Act 1988 (Vic). Such immunity aims to ensure that confidential
information generated by approved quality assurance bodies is not disclosed to people other than
committee members or admitted in court proceedings. To provide the surveillance committee with
statutory immunity, the committee at Wimmera Health Care Group is constituted as a subcommittee
of the board of management's quality committee.
The surveillance committee meets cver>' 2 months. At these meetings, cach medical reviewer
presents a report summarising details of the medical records they have reviewed and the adverse
events they have delected. The clinical risk manager presents a summary report of the risk register,
identifying issues lor
discussion by the committee and recommendations for clinical practice changes
in the hospital. The committee also reviews information about adverse events
that have occurred outside the hospital and that have been included in coronial,
media and consultative council and committee reports, and the relevant
recommendations that have been made.

15J Committee reports

Box 5.4 contains examples of recommendations made by the Wimmera Health

Care Group Patient Care Committee. These recommendations are presented to
the medical staff group — the peak medical group in the hospital. Care is taken
that the recommendations are positively framed and clearly define actions to be
considered, rather than detailing errors that have been made and allowing the
responsible clinicians to possibly be identified.

Box 5.4 Wimmera Health Care Group medical staff group report
1. Management of snake bite
The committee recommends that antivenom stocks in Wimmera hospitals be documented and that this
information, together with guidelines regarding the management of snake bite, be disseminated to all
doctors and hospitals in the region.

2. Intravenous lines
The committee recommends that intravenous lines be dated on a sticker on the entry site dressing.

3. Caidtopulmonary resuscitation
The committee recommends that the physician on call be contacted in the event of cardiopulmonary
resuscitation continuing for more than 30 minutes or if the patient requiring cardiopulmonary resuscitation is
transferred to the intensive care unit.

4. Emergency temporary pacing

The committee recommends that emergency temporary pacing wires be inserted in the operating theatre.

5. Invasive procedures
The committee recommends that before any invasive procedure is undertaken by the resident
medical staff (including central line and chest tube insertion), the proposed procedure be discussed
with the appropriate consultant.

o. renem repons
The committee recommends that when a hospital medical officer is called to the ward in response to an
incident, they should make an entry in the medical record in addition to completing the incident report.

7. Patient admission procedure

 5 Analysing adverse events
and prioritising action

■ All non-elective patients being admitted to the hospital (except obstetric patients) must come through
the accident and emergency department. Patients in a stable condition may only require a rapid medical
and nursing assessment. These patients can then be quickly transferred to the ward, where a lull
admission can be done at a later time.
■ The appropriate registrar (during business hours) or the second on-call hospital medical officer (after
hours) shall review all patients who are admitted via the accident and emergency department. The
appropriate registrar or the second on-call hospital medical officer shall be informed of the admission by the
intern on duty in
the accident and emergency department. On some occasions, the review can take place by telephone.
■ The intern in the accident and emergency department must inform the appropriate visiting medical officer
and obtain their agreement to admit the patient under that visiting medical officer s bed card. The one
exception is wtien patients are admitted for short-term observation (eg. patients having a 4-hour neurological
observation after a minor head injury)
5.6 Principles for prioritising action to prevent adverse
events recurring
The risk register ranks risks according to the multiple of their probability of occuning and their
consequence when they occur. The relatively large number of events detected creates a problem, as
the resources available for quality improvement and ris*k management programs in individual health
services are limited.
Which risks should be addressed by the health service, and how can they be prioritised so
appropriate action is taken in a timely manner? Ideally, the highest risks should be eliminated or
minimised first. However, usually because of resource limitations, other approaches are used for
prioritisation. These methods include addressing adverse events that occur frequendy, cause serious
harm, use substantial resources after they occur, and where there is strong evidence that available
in^rventions are effective in preventing adverse events, feasible, safe, generaliseable and sustainable.
The advantages and limitations of these alternative approaches to prioritisation for action will now be
discussed.

5.6.1 Using frequency and harm to prioritise action

After the publication of the large epidemiological studies of adverse events such as the Harvard
Medical Practice Study and the Quality in Australian Health Care Study, many clinical risk
management programs have emphasised taking action to prevent adverse events that cause the most
harm to patients. Although the consequences of some of these adverse events are extremely serious,
many events, occur infrequently. Therefore, prioritising actiQn to prevent adverse events according to
the harm they cause will reduce significant harm to a small number of patients.
Another approach in determining which adverse events to address is to concentrate on events that
occur frequently. However, many frequently occurring adverse events do not cause major harm to
patients. Therefore, prioritising action according to the frequency of adverse events will prevent many
events, but only reduce the overall harm caused to patients by a small amount. As demonstrated in
the risk register, multiplying the probability of the adverse event occurring by the consequence or
harm that results when it does occur provides a measure of the risk of the event and helps prioritise
action.

5.6.2 Using the cost of adverse events to prioritise action

Another approach is to focus on adverse events that use the most resources. Using data from the
Quality in Australian Health Care Study and diagnosis-related group cost weights "crude... back of
the envelope" estimates of the annual direct cost to hospitals of adverse events in 1999 was made.
This assessment was based on the additional bed-days required to treat patients who had
experienced an adverse event.21 The most costly adverse events are listed in Box 5.5.
 Enhancing patient care

In a further study, a sample of adverse events that would typically occur annually in a notional 250-
bed hospital was examined. The resource utilisation of each adverse event was determined according
to the additional patient length of hospital stay required. The adverse events that would have
occurred were determined to require 11 500 additional bed-days, or 10% of the hospi tal's total
annual bed-days. Also, 60% of the additional resources were used after "mundane"

110
 Enhancing patient care

adverse events occurred that led to minor disabilities and only 40% were used after "serious" adverse
events that led to major disability or death. Mundane adverse events included postoperative pain,
nausea and vomiting, pressure sores and catheter related infections. Reducing the number of
mundane adverse events would benefit a greater number of patients than addressing serious events.
Clearly, mundane and serious adverse events need to be addressed by individual health services in a
balanced way."
Although mundane adverse events have less dramatic consequences than serious events, it may be
appropriate to take action to reduce the rate of mundane events because of the considerable
additional resources they consume. Addressing resource-intensive adverse events could result in
considerable savings to a health service. These savings could then be used to reduce the number of
less resource- intensive adverse events.

5.4.3 Using the evidence of the effectiveness of Interventions to prioritise action

Another approach is to focus on those patient safety practices for which there is strong evidence of
effectiveness in preventing adverse events. A patient safety practice is "a type of process or structure
whose application reduces the probability of adverse events resulting from exposure to the health
care system across a range of diseases and procedures". 25
In an ambitious, challenging project, potential practices for patient safety improvement were
comprehensively evaluated and a list of 83 practices were rated using general scores (low, medium,
high) according to: their potential impact (based on the prevalence and severity of the patient safety
target and the current utilisation of the practice); strength of evidence supporting the practice (only
practices supported by randomised controlled trials reporting clinical outcomes were evaluated);
effect size (unclear, negligible, modest, robust); the need for vigilance to reduce potential negative
collateral effects of the practice; cost; and complexity (logistic barriers). In an attempt to answer the
following question for the leader of a health service "if I wanted to improve patient safety at my
institution over the next 3 years and resources were not a significant consideration, how would 1
grade this practice?", a rating from zero to 10 was given as a combined assessment of the impact of
the patient safety practice and the strength of supporting evidence. A table of ratings for each
practice is available at www.ahrq.gov/clinic/ptsafety/chap59a.htm.
The categories of patient safety practices include adverse drug events, infectious complica tions of
hospitalisation, complications of anaesthesia, surgery and other invasive procedures, complications
in hospitalised older patients, system issues and human factors engineering, approaches to
education and training and informed consent. The greatest strength of evi dence ratings (impact and
strength) were found for the practices listed in Box 5.6. 2526
If the three criteria of importance, frequency and strength of evidence are used the three highest
priorities were:

Box 5.6 Interventions with the greatest

sbength of evidence
■ Appropriate venous thromboembolism prophylaxis.
■ Use of perioperative ^-blockers.
■ Use of maximum sterile barriers during catheter
insertion
■ Appropriate use of antibiotic prophylaxis to
prevent surgical site infection.
■ Asking that patients recall and restate what
they have been told during the process of
obtaining informed consent.
■ Continuous aspiration of subglottic secretions to
prevent ventilator-associated pneumonia.
■ Use of pressure-relieving bedding materials.
■ Use of real-time ultrasound guidance during
central line insertion.
■ Patient self-management using home
monitoring devices to prevent adverse events
related to warfarin therapy
■ Various nutritional strategies to reduce mortality
and morbidity in postsurgical and critically ill
patients.
■ Antibiotic-impregnated catheters to prevent
central venous catheter-related bloodstream
infections.
111
 5 Analysing adverse events
and prioritising action

a anticoagulation therapy to prevent deep vein thrombosis a

antibiotic prophylaxis to prevent surgical infections a use of
pressure relieving materials to prevent pressure ulcers.
Importantly, "the evidence base in patient safety remains young and has not yet stabilized", 27 and
subsequent studies can reveal contradictory evidence about benefits and previously unrecognised
problems. Meta-analysis including subsequently published randomised controlled trials of the use of
perioperative p-blockers has shown "questionable benefit and definite harm". 27 Therefore, more than
just evidence should be considered in deciding which interventions should be implemented. 27

5.6.3.1 Limitations of using effectiveness to prioritise action

Prioritisation of safety practices according to their supporting evidence and effectiveness has been
keenly debated. Leape and colleagues regarded the list of 83 patient safety practices as neither
complete nor appropriate, with most being quality improvement initiatives rather than safety
practices.20 They also felt many of the practices reflected technical advances in science that prevent
complications, rather than reduce errors.28
The evaluation of the effectiveness of each safety practice placed emphasis on the existence of
randomised controlled trials. However, many key, simple, intuitively attractive, commonly used and
well accepted safety practices lack evidence about their effectiveness and only a few of these practices
were on the list. Some of the trials required to demonstrate the effectiveness of many of these
practices may not be able to be done, and those that could be done may show that some of these
practices are ineffective, wasteful or even harmful.26
Because of the requirement for high-quality evidence of the effectiveness of a safety practice for it
to be included on the list, the listed practices are biased toward areas where considerable research
has been undertaken such as drugs, devices and procedures. However, there has been little research
in health care on error prevention; this is why few error prevention practices are listed. Further, even
if there is strong evidence that an intervention is effective, it should not necessarily be a high priority
for implementation.28
Leape and colleagues state that to provide safe care to patients requires three
steps: a identifying that the care works
a ensuring patients who would benefit from the care receive
it a delivering the care without error.28
The list of safety practices only addresses the first step. Successful implementation of the second
and third steps requires the development of safe systems of health care delivery that are based on
human- factor principles and that attempt to reduce errors due to human cognition by using
practices such as simplifying and standardising systems and using checklists. This approach results
in many small changes being made to areas such as equipment, organisation, supervision and
training that haw a significant aggregate effect.
Making many small changes based on sound principles and experience has been used successfully
for many years in aviation, 28 although not with the level of evidence of effectiveness required in
clinical medicine.26 The equivalent scenario to aviation in medicine is anaesthesia, where the
approach of incremental and multiple small changes, rather than the use of controlled trials, has
been associated with a reduction in anaesthesia-related mortality. 28 The counter-argument is that
appropriate standards need to be used to evaluate even "obviously beneficial" practices. Some
practices that appear to make obvious sense may be later shown to cause more harm than benefit.
Also, small changes to one aspect of a complex system can increase the complexity of the system and
result in unexpected and sometimes negative effects elsewhere in the system. Therefore, ongoing
vigilance is required once change? have been made. 26
 Enhancing patient care

A further difficulty in conducting controlled trials in patient safety research is that although the
number of adverse events occurring in hospitals is large, the frequency of individual adverse events is
low. Therefore, trials aimed at demonstrating reductions in already small numbers of specific adverse
events will be difficult to conduct. 28 Surrogate markers of adverse events may sometimes need to be
used, so a reduction in actual adverse events will not be demonstrable. As with drug trials, there may
be difficulties in implementing patient safety interventions that are effective in clinical studies into
routine clinical practice, with potentially significant changes in skill levels, patients and settings. 26
When deciding which patient safety practices to implement in a health service, reasonable
judgement based on the best available evidence, experience and cost needs to be exercised. Decisions
need to be made about which practices will most likely result in the greatest reduction in adverse
events. Unfonunately, the availability and quantity of high-quality evidence about the effectiveness of
individual practices will probably not reflect the frequency with which particular types of adverse
events occur. Waiting for trials to be undertaken because the relevant evidence about the
effectiveness of a particular patient safety practice is not available would result in inaction, and may
not always be an option. Ideally, the practices implemented should be those that will reduce the most
adverse events.
"An unquestioning embrace of dozens of promising practices from other fields is likely to be
wasteful, distracting and potentially dangerous." 25 Overall, "the general insistence on evidence does
not prevent implementation of practical, low risk but understudied interventions that seem likely to
work." Even with a balanced and considered approach errors in the selection of practices for
implementation will be made. However, on balance, such an approach will increase the level of
patient safety.26

5.6.4 Using comprehensive programs developed by external organisations to '

prioritise action
Clearly, the direction of quality improvement and patient safety programs in individual health
services should be tailored to the local needs and conditions of each organisation and the community
it serves. Where possible, gaps in health care delivery and risks that are detected should be
addressed. However, there is probably much commonality in the gaps and risks in different hospitals
that provide acute care. Another approach to prioritising action is to look at the initiatives being
taken by large quality improvement and safety programs elsewhere and see if they are applicable
locally. These programs are called quality improvement collaboratives, and were described in Chapter
2.
An example is the 100000 Lives Campaign in the US, which was expanded to the 5 Million Lives
Campaign in 2006. This campaign aimed to reduce five million incidents of harm during a 2-year
period across many hospitals in the US. The initial program recommended hospitals in the campaign
utilise six well known interventions to reduce mortality:
■ deploy rapid-response teams to patients at risk of cardiac or respiratory arrest
■ deliver reliable, evidence-based care for acute myocardial infarction
■ prevent adverse drug events through drug reconciliation
■ prevent central line infections
■ prevent surgical site infections
■ prevent ventilator-associated pneumonia.
Each participating health service developed systems of care to ensure that each eligible patient
received each evidence-based element of care for the above clinical situations. An additional six
interventions were added for the 5 Million Lives Campaign:
■ prevent pressure ulcers
■ reduce methicillin-resistant Staphylococcus aureus infection
■ prevent harm from high-alert medications (initially anticoagulants, sedatives, narcotics and
insulin)
■ reduce surgical complications (eg, venous thromboembolus prophylaxis)

112
 5 Analysing adverse events
and prioritising action

■ deliver reliable, evidence-based care for congestive hean failure to reduce readmission
■ involve boards of directors to make hospitals safer.
Participating hospitals were encouraged to add their own harm-reduction innovations and interven -
tions to the list. The campaigns provided resources to participating hospitals. These resources, which
include how-to guides and tools for managing and measuring change are available on the Institute lor
Healthcare Improvement's website (www.ihi.org/ihi/programs/campaign/). These interventions will
be assessed using trigger tools to determine the rate of harm in a random sample of patient medical
records The initial 100000 Lives Campaign is currently being evaluated, 29 and preliminary results are
positive with significant reductions in infection rates. 30
The Joint Commission in the US requires hospitals to implement 13 high impact, evidence-based
and measurable safe practices, referred to as National Patient Safety Goals. Goals include improving
patient identification, communication and medication safety, reducing health care associated
infections, reducing harm from falls, minimising risks of influenza, encouraging patient involvement
and preventing pressure ulcers. For example, to improve the accuracy of patient identification, the
two requirements are: use at least two patient identifiers when providing care, and conduct a final
verification process, such as "time out", before starting any procedure. The details and requirements
for each goal are available from the Joint Commission website
(www.jointcommission.oi^PatientSafety/ NationalPatientSafetyGoals).
Some authors believe that safety risks and problems in individual health services are heterogeneous
and local. Therefore, safety problems identified in large programs may not be problems in a local
setting. It has been suggested that individual health services adopt a "local epidemiological approach"
— assess their environment and identify local problem patterns and allocate their limited resources to
local priorities to use local solutions rather than "a one-size-fits-all strategy" from a national
program.31 Such an approach is similar to the risk register developed from adverse events and risks
that are detected locally that was discussed earlier and has been used in delecting adverse events in
neonatal medicine.32
When prioritising gaps and risks for action, the criteria in Box 5.7 should be considered. 19-27,33
Recently, two further approaches to prioritising interventions have been proposed.

5.6.5 A simple, straightforward approach to deciding which adverse events to address and how to
address them
Given the preceding discussion about the many criteria that can be used to determine which adverse
events to address, it is important to acknowledge that simpler methods of setting intervention
priorities can be used. One such method is described below.

Box 5.7 Criteria to help prioritise gaps and

risks for action
■ The frequency of the adverse event, size of the
gap between the care provided and best practice,
and the prevalence of the clinical condition
■ The harm patients are experiencing from
suboptimal care and the adverse event
■ Preventability — the strength of the evidence
o( the effectiveness of the intervention in
improving patient outcome: the size of the impact
the intervention has on patient outcome
■ Does the intervention have any unintended or
undesirable consequences that require
monitoring?
■ The complexity of the intervention — is it a
major change in how care is delivered?
■ Cost of implementing the intervention and
change in practice — are adequate resources
available?
■ Benefits to the health service — the resources
saved (by preventing adverse events) by
implementing the intervention
■ Feasibility — what barriers and facilitators are
there to implementing the intervention?
■ Does implementation of the intervention have
the potential to create momentum for additional
quality and safety initiatives?
■ External imperatives (eg. mandatory
accreditation standards)
■ How many initiatives can be realistically
undertaken at any one time given the resources
and expertise availability and the magnitude of
initiative7
■ If higher-risk adverse events occur before
implementation of the intervention, priorities may
change with some initiatives suspended or
delayed
Enhancing patient care

Each clinical area in the health service selects five topics based on errors made in that area
described in the literature. Topics are chosen according to the frequency with which the errors are
made, the harm that patients experience following the error, and the degree of preventability of the
error. Two process indicators and one outcome indicator are selected or developed for each topic. The
indicators are measured and the results are peer reviewed and benchmarked against the best
performance locally, nationally and internationally. Comparative feedback is also provided to the
clinicians in the clinical area. If the results are more than two standard deviations below the average,
a plan for improvement with indicator targets and a timeframe is developed with the input of
clinicians and implemented. The indicators continue to be measured and feedback provided until the
indicator targets are reached.34

5.6.6 A "balanced diet" approach

Another approach is to choose a "balanced diet" of interventions in the following categories: 27
■ "Low hanging fruit" — interventions that address common patient problems, have strong
supporting evidence, high impact and are easy to implement and have low potential for harm (eg,
ultrasound- guided central venous catheter insertion, prevention of catheter-related bloodstream
infection, removal of urinary catheters after 48-72 hours to reduce catheter-associated urinary
tract infections). Interventions with lower-quality supporting evidence and high impact (eg,
structured discharge summaries including changes in medication since admission and pending
investigations, postdis- charge follow-up phone calls).
■ Momentum-generating projects — interventions with modest or little supporting evidence, but that
promote interprofessional collaboration which has the potential to generate further quality and
safely initiatives (eg, rapid response teams, executive "walk rounds" and teamwork training).
■ System-wide interventions that affect day-to-day practice (eg, computerised provider order entry
and electronic medical records).
The authors did not recommend some commonly used interventions thai provided less benefit lo
patients compared with the effort required lo implement ihem (eg, medication reconciliation, incident
reporting and root-cause analysis). 27 These interventions may be of benefit to some patients, but
other inventions may be more beneficial and resources available are limited.
m Analysis of adverse events comprises three stages: ensuring that the patient, their relatives and the
clinicians who provided the care have been appropriately managed: analysis of the event: and
prioritisalion of action to prevent a recurrence.
■ Adverse events can be analysed using systems analysis, root-cause analysis or a simplified versions of
these methods. A balance is required between taking adequate time for appropriate analysis of adverse
events and taking action in a timely manner to prevent the recurrence of events.
a Analysis determines the probability, consequence, preventability. and latent factors and active errors
that contributed to the adverse event occurring.
a The level of risk associated with an adverse event is equal to the probability of the event multiplied by the
consequences of the event.
a A hospital clinical risk management program will detect more adverse events than can be effectively
dealt with. Therefore, adverse events need to be prioritised for action. 4

a Criteria to consider when prioritising events include: the frequency of the event, the harm it causes, its
preventability and the cost, feasibility, complexity and unintended consequences of the required
intervention. Comprehensive external risk reduction programs may also be used to prioritise actions.
a Maintaining a risk register of adverse events ranked by their level of risk helps focus attention to where
action is most required to reduce risk.
 Important points
5 Analysing adverse events and pnoribeng actoor

a The composition of the surveillance committee that reviews adverse events should alkm meaningful
discussion of a wide variety of events.
a Statutory immunity should be obtained for the surveillance committee's activities.
References
1 Lamb R. Open disclosure: the only approach to medical error. Qual Saf Health Care 2004; 13:3-5.
2 Australian Commission on Safety and Quality in Health Care. Open disclosure: a review of the literature.
Canberra- ACSQHC. 2008.
3 Victorian Government Department of Human Services. Open disclosure statewide pilot project evaluation
report. Metoourne: DHS. 2007.
4 Vincent C. Understanding and responding to adverse events. N Engl J Med 2003; 348:1051-1056.
5 Deibanco T. Bell SK. Guilty, afraid, and alone — struggling with medical error. N Engl J Med 2007; 357:
1682- 1683.
6 Waterman AD. Garbutt J, Hazel E, et al. The emotional impact of medical errors on practicing physicians in
the United States and Canada. Jt Comm J Qual Improv 2007; 33: 467-476.
7 WU A. Medical error: the second victim. BMJ 2000; 320: 726-727.
8 Dunn KL. Moulden A, McDougall P. Bowes G. Patient safety: a view Irom down under. Pediatr Clin N Am
2006; 53: 1217-1230.
9 Vincent C. Taylor-Adams S. Stanhope N. Framework for analysing risk and safety in clinical medicine. BMJ
1998: 316: 1154-1157.
10 Bagian JP. Gosbee J. Lee CZ. et al. The Veterans Affairs root cause analysis system in action. Jt Comm J
Qual Improv 2002: 28: 531-545.
11 Bagian JP, Lee C, Gosbee J, et al. Developing and deploying a patient safety program in a large health
care delivery system: you cant fix what you don't know about. Jt Comm J Qual Improv 2001; 27: 522-532.
12 Brennan TA. Leape LL, Laird NM, et al. Incidence of adverse events and negligence in hospitalised
patients: results of the Harvard Medical Practice Study 1. N Engl J Med 1991; 324:370-376.
13 Wilson RM, Runciman WB, Gibberd RW, et al. The Quality in Australian Health Care Study. MedJAust
1995; 163: 458-471.
14 Wu AW. Lipshutz AKM, Pronovost PJ. Effectiveness and efficiency of root cause analysis in medicine.
JAMA 2008: 299:685-687.
15 8km JF. Heffner JE. Challenges in the care of the acutely ill. Lancet 2004:363:970-977.
16 Shojania KG. Wald H. Gross R. Understanding medical error and improving patient safety in the inpatient
setting. Med Clin N Am 2002: 86: 847-867.
17 Berwick DM. Errors today and errors tomorrow. N Engl J Med 2003; 348: 2570-2572.
18 Pronovost PJ. Holzmueller CG. Martinez E. et al. A practical tool to learn from defects in patient care. Jt
Comm J Qual Patient Saf 2006:32: 102-108.
19 Stevens P. Matlow A. Laxer RM. Blueprint for patient safety. Pediatr Clin North Am 2006; 53: 1253-1267.
20 Wilson RM. Harrison BT, Gibberd RW. et al. An analysis of the causes of adverse events from the Quality
in Australian Health Care Study. MedJAust 1999: 170: 411-415.
21 Leape L. Lawthers AG. Brennan TA. et al. Preventing medical injury. Qual Rev Bull 1993; 19: 144-149.
22 Standards Australia/Standards New Zealand. Risk management guidelines companion to AS/NZS 4360:
2004. Sydney/Wellington: SA/SNZ. 2004.
23 Rigby K. Clark RB. Runciman WB. Adverse events in health care: setting priorities based on economic
evaluation. J Qual Clin Practice 1999; 19: 7-12.
24 Runciman WB. Edmonds MJ. Pradhan M. Setting priorities for patient safety. Qual Saf Health Care 2002;
11: 224- 229
25 Shojania KG. Duncan BW. McDonald KM. Wachter RM. eds. Making health care safer: a critical analysis
of patient safety practices. Rockville. Md: Agency for Healthcare Research and Quality. 2001. (Evidence
Report/Technology Assessment 43. AHRQ Publication No. 01-E058.)
Enhancing patient care

28 Shojania KG. Duncan BW. McDonald KM. Wachter RM. Safe but sound. Patient safety meets evidence-
based medicine JAMA 2002; 288: 508-513
27 Ranji SR. Shojania KG. Implementing patient safety interventions in your hospital: what lo try and what to
avoid. Med Clin N Am 2008; 92: 275-293.

flfl
 5 Analysing adverse events and
pnontiamg aovon

28 Leape LL. Berwick DM. Bates DW. What practices will most improve safety? Evidence-baaed medicine meets patient safety.
JAMA 2002: 288: 501: 507.
29 McCannon CJ, Hackbarth AD. Griffin FA. Miles to go: an introduction to the 5 Million Lives Campaign Jt Comm J Oual Patient
Sat2007; 33:477-484.
30 Berwick DM. Will is the way to win the patient safety war. Health ServJ2007; 117. 18-19.
31 Larsen G. Parker H. Patients' safety: think and act locally. Lancet 2008; 371:364-365.
32 Ligi I. Arnaud F. Jouve E. et al. Iatrogenic events in admitted neonates: a prospective cohort study. Lancet 2000: 371:404-410.
33 Grimshaw J. Eccles M. Thomas R. MacLennan G. Toward evidence-based quality improvement: evidence (and its limitations) of
the effectiveness of guideline dissemination and implementation strategies 1966-1996. J Gen Intern
Med 2006,21: s14-s20.
34 Wollersheim H. Clinical incidents and risk prevention. Neth J Med2007; 65: 49-54. □
Enhancing patient care

6 Taking action to close gaps in care and prevent adverse

events recurring — evidence and theory
This chapter will:
m Examine the factors that facilitate and impede the adoption, implementation and maintenance oI change and innovation in
health services
■ Discuss theories from many disciplines relevant to implementing change and their practical implications lor achieving change in
health care
■ Describe different approaches to change based on how people behave individually and in groups and organisations
■ Outline how to develop change strategies *
■ Discuss the evidence supporting various change strategies
■ Describe a change model using the strategies of quality improvement and system redesign.

Determining the gaps in care provided to patients and the risks to which they are exposed is demanding, but it is easier
than determining and taking effective action to close the gaps and reduce the risks. Health care delivery systems are
complex and the components can interact unpredictably effecting interventions.1 Also, closing gaps and reducing risks
requires changing clinical behaviour, which is often difficult.
Most health care research is about understanding disease processes and finding effective treatments, rather than
determining the best ways of delivering safe care. 2 Clinicians seek high-quality scientific evidence that treatments are
beneficial and safe, but such evidence is often not sought before implementing quality improvement and risk reduction
initiatives. Health services often implement quality improvement initiatives without first evaluating their effectiveness and
whether they have unintended negative consequences (which is well recognised when making changes in complex systems)
or cause harm to patients, or whether they are the best use of the limited resources available. 3
Evidence for the effectiveness of implementation strategies is often not available or is limited. 3 The effects of
implementation strategies are modest, and they may be effective in one setting and not another. Although there are "still no
magic bullets",4 some strategies, such as computerised ordering and prescribing involving "information technology |are)
often oversold as a definitive quality and safety solution". 5 Future improvements are likely to be incremental. 4 Strategies
currently used are based on intuition, anecdotes of strategies used in other health services, and high-profile, poorly
evaluated quality improvement techniques from other industries, rather than on high-quality scientific evidence. 4
Quality improvement implementation strategies in health care began with passive diffusion. It was generally thought that
clinicians would automatically change their behaviour when research providing strong evidence of significant treatment
effects was published or presented at conferences. Clinicians found it difficult to accurately evaluate the evidence, and
reading journals or attending lectures had little effect on clinical practice. 4 6 It was then thought that creating guidelines and
reviews summarising research and presenting "bottom-line" recommendations would change clinician behaviour But the
effect was not as great as initially thought. 4 6
Next, health services tried to implement industrial-style quality improvement strategies, such as total quality
management and continuous quality improvement, whose effectiveness was only superficially evaluated in the other
industries4 They were used with audit and feedback, reminders and educational research, consumed considerable resources
and only produced modest benefits in patient care Doctors
 felt that these measures reduced their autonomy, leading to clinician
resistance.6
Systems engineering, involving system redesign and often requiring
major information technology input, was then used. Decision suppoYt
tools, together with automated prompts and reminders, were made
available at the bedside, but the results were often disappointing. 4,6
Clearly, further study of the barriers to effective implementation of
change is required.4 Currently, theories from other disciplines in which
changing behaviour is also important have been used to implement
change in health care6 and will be presented later in this chapter.
Systems of health care delivery are often complex, unstable and non-
linear, and many proposed changes are complex forms of social change
with multiple components. The effectiveness of such change is sensitive to
a broad range of contextual factors. 7 To be successful, programs should
"introduce the appropriate ideas and opportunities ('mechanisms') to
groups in the appropriate social and cultural conditions ('contexts')". 8
Attention to the mechanism of change and the context in which it is being
made, which does not usually occur in experimental designs (such as
randomised controlled trials), may help generalise the change to other
contexts.7
Important contextual factors in determining the effectiveness of quality
improvement implementation strategies are: the nature of the
intervention; the target group of health service staff; the patients who
are .to receive the intervention; features of the organisation (eg, its size,
structure, leadership, governance, culture, communication and
resources); and sociopolitical and other factors (eg, regulations and
legislation) that affect the health service. The paucity of high-quality
research adds to the uncertainty of each factor's role. Often, factors such
as culture and leadership within organisations are not robustly studied
and the interventions used are only broadly described.
There is a pressing need to determine the "active ingredients" of each
intervention. This information would help us understand why
interventions work and what aids or interferes with their effectiveness in
particular settings. An adjustment in the expectations of senior
management in health services, government and other funding bodies of
what the implementation strategies can realistically achieve is also
required. As with advances in clinical medicine, the gains achieved in
quality improvement are likely to be incremental and modest 4
Effective implementation of quality improvement and patient safely
strategies is difficult. How, then, can health services implement initiatives
that will close the gaps and reduce the risks they have identified? The
success of an intervention depends on the interaction of many factors at
multiple levels of ihe health service.
With the shortage of good scientific evidence to suppon strategies, it is
appropriate to look at theories about how individuals leam, behave and
Thedsfictetolhe change. A theory can be defined as "a system of ideas or statements held
performance of the as an explanation or account of a group of facts or phenomena" 9 Theories
heath care system an
enormous and the rate
el change appears

McGlynn EA
Med Care 2007;
45: Ml
are made up of many theoretical constructs or component parts 9 Theories
aid our understanding of phenomena, and can be used to predict
behaviour in varying circumstances. By understanding theories of how
these factors are influenced and interact, appro-
 6 Taking action to close gaps in care and prevent
adverse everts recurring

priately tailored strategies can be developed. This approach should also encourage the systematic
analyse at factors influencing the success of an intervention. 10 The following actions are required:
1. Systematically determine the factors that will influence (facilitate or impede) the implementation
at the proposed change. Section 6.1 provides a comprehensive checklist of factors developed from
the literature.
2. Review the relevant theories about individual, group and organisational behaviour. Section 6.2
gives a brief description of theories from a wide range of disciplines thought to be relevant to
quality improvement and implementing change in health care.10
3. Develop practical approaches and strategies from these theories to address the identified (actors
The main aim of this book is to provide a practical guide to quality improvement and patient
safety for those at the coalface. Therefore, assumptions about human behaviour will be distilled
from the theories presented and listed in a practical format (Chapter 7). *

6.1 Determinants of implementing a change within a health service

A review of implementation of innovations in health care organisations identified 50 determinants or
influencing factors, and categorised them into five groups: characteristics of the sociopolitical
context; characteristics of the organisation; characteristics of the user; characteristics of the
innovation; and characteristics of the patients (Figure 6.1)." The authors recommended that
determinants relevant to the successful implementation of a proposed change be systematically
sought; this was rarely done in the studies reviewed."

Figure 6.1 Determinants of innovation within health care organisations 11

121
 Enhancing patient care

The target group of the strategy and relevant others should be asked, using interviews or
questionnaires, to identify the determinants they perceive as influencing the success of the
implementation of the change. They should classify them as facilitating, impeding or neutral to the
implementation and grade their importance. In a review, determinants were reported as impeding 2.5
times more frequently than facilitating, and were infrequently reported as neutral. 11 When the
determinants are identified, an appropriate theory can be selected as a starting point to influence the
determinant and a strategy developed to organise ■ or operationalise the theory. A change strategy
tailored to the determinants can then be developed.
Undertaking a systematic determinant analysis ensures that the strategy developed focuses on the
determinants that the target group believes are relevant. The importance of each of the 50
determinants identified in the review will vary with the change or innovation being proposed. However,
the authors suggest that this list of potential determinants provides a checklist to help develop effective
change strategies. 11
A review of the diffusion of innovations in service industries, including health care, reviewed
literature in sociology, psychology, anthropology, epidemiology, marketing, organisation and
management and information and communications technology. 1 The authors developed a conceptual
model for considering the determinants of diffusion, dissemination and implementation in health
service delivery and organisation.1 The review enabled a useful, but not totally exhaustive, lists of
attributes and determinants to be developed, which should be used by health services considering
adopting an innovation. However, these factors should be examined with the understanding that —
given the complexity of health care delivery systems, the multiple and often unpredictable interactions
of the interdependent components of the system, and the great variety of contexts and settings in
which health care is provided — the presence of these factors will not guarantee the successful
adoption of the innovation.
Another study reviewed change in clinical practice from a slightly different perspective to determine
why doctors do not follow guidelines. 12 The authors found a number of determinants that impeded the
uptake of guidelines. They identified potential barriers to a proposed change and categorised them as
follows:
■ knowledge — lack of awareness or familiarity
■ attitudes — lack of agreement, self-efficacy, outcome expectancy, inertia of previous practice
■ external barriers — patient, guideline and environmental factors. 12
Although concentrating on barriers, this study found many determinants also elicited by the other
reviews. We have integrated and modified the results of these three reviews to produce a list of the
determinants of the success of an innovation (Box 6.1). These determinants are supported by variable
levels of evidence and expert opinion. The presence of a determinant may facilitate the adoption,
implementation and maintenance of an innovation, whereas its absence may impede the innovation.
Clearly, not all determinants are relevant to each innovation nor are all determinants required to be
present before an innovation is introduced.
Systematic determinant analysis to identify key determinants and design and implement strategies to
address them is critical to successfully changing how health care is delivered. In a study that examined
use of ^-blockers in patients with myocardial infarction, cardiologists, emergency physicians, nurses,
quality managers and senior administrators in eight hospitals in the United States were interviewed
and asked to describe the major initiatives their hospital had undertaken to improve the care of
patients with acute myocardial infarction and the difficulties and successes with each initiative. Six
broad factors described the improvement efforts of hospitals:
■ setting goals (goal content, goal specificity, goal challenge, the degree to which goals were shared)

122
 Enhancing patient care

■ administrative support (organisational philosophy regarding quality improvement and resources

made available — care coordinators, chart abstractors, computer and analytical support, quality
improvement training)
Box 6.1 Determinants of adoption, Implementation and maintenance of Imovetlon1'"'"

Characteristics ot Innovation
a Clearly effective
a Low complexity
a Compatible with values, norms and needs of users
a Can be trialled on a small scale
a Benefits clearly visible to users
a Can be modified to meet individual needs
a Relevant to users' work; improves task performance
■ Knowledge required to use the innovation can be transferred across settings
a Supported — training, help desk
e High relative advantage in effectiveness or cost-effectiveness
a Supported by high-quality scientific evidence *
a Observable benefits
a Appealing to use
a Low associated risks
a Frequent use
a Clear procedures
a Low cost
a Little time required
e Implications and effects of the innovation fully assessed and anticipated
a Innovation adapted to local needs and conditions

Characteristics ol users — target groups/subgroups and their members

a Sufficient knowledge level to implement innovation
a Sufficient skill level to implement innovation
a Sufficient experience to implement innovation
a Motivated to use innovation
a Meets an identified need for users
a Aware of innovation, what it does, how to use it, and how it will affect them personally
a Individuals considering adoption have similar backgrounds
a Champions in social network support innovation
a Colleagues are implementing (modelling) required behaviour
a Opinion leaders (respected, trusted professionals who influence [positively or negatively] the beliefs and
actions of their colleagues) are available
a High support of colleagues to implement
a High support of other professionals to implement

123
 Enhancing patient care

B High support ol supervisors to implement

a High self-eflicacy
a Perceived ownership of innovation
a Belief that patients will cooperate
a Belief that patients will be satisfied
B No ethical problems with innovation
a Users' goals match aims of innovation
a Low level of work-related stress
a Few potential users j

flov 6 1 (continued
or> •"><=»,' as j*"

124
Enhwdng patient care

So* 6.1 (contfciued)

■ Patients willing to cooperate

■ Patients are aware of benefits
■ Adds value for patient
■ Lo* risks for patient
■ Low cost to patient
■ Low level of patient discomfort
■ Patient preferences match innovation
■ High level of patient compliance

Characteristics of the organisation

Size
■ Size — may be easier to implement an innovation in a medium-sized organisation with fewer staff to engage than in a large
organisation. However, it may also be difficult to introduce change in a small health service if there are only one or two
doctors who may not support the innovation.
■ Manageable number of users to be reached — it may be easier to engage 5-10 clinicians than 50-100
Structure
■ Divided into semi-autonomous departments or units
■ Departments or units contain foci of specialised professional knowledge
■ Semi-autonomous multidisciplinary project teams established
■ Flexible structure for decision making

Leadership
■ Strong leadership
■ Clear vision
■ Senior management actively involved and frequently consulted
■ Decentralised decision-making process and procedures
■ Operational decision making devolved to teams "on the ground"
■ Senior management supports implementation
■ Innovation designed to appeal to key opinion leaders

Culture
■ Current situation viewed as intolerable — clear evidence change is required
■ Learning organisation culture — support for shared knowledge
■ Opportunities for interprofessional teamwork, and involvement of clinicians in management networks and vice versa
■ Accurate and timely feedback provided to users
■ Climate conducive to experimentation and risk taking
■ Good collaboration between involved departments
■ Users involved in development
■ Supporters of the innovation outnumber, and are more strategically placed in the organisation than, opponents of the
innovation
Governance
■ Close and effective monitoring and evaluation process

Communication
■ Effective communication across departmental boundaries

Box 6 1 (continued on next page)

Box 6.1 {continued)

Resources
■ Adequate financial resources available
a Dedicated and ongoing funding available

124
Enhwdng patient care

■ Reimbursement or replacement staff available for involved staff

■ Administrative support available
■ Effective data collection systems available
■ Adequate time available
■ Coordinator available
■ Low staff turnover
■ Full staffing
■ Much relevant expertise available Education
■ On-the-job training provided with high-quality materials on how to use innovation
■ Support provided to integrate in individual's daily work
Administrative
■ Well arranged relevant logistical procedures
■ Organisational climate — secure financial situation, low level of threats, role and viability of service secure
Characteristics ot sociopolitical context and external factors
a Critical mass of comparable organisations adopting innovation
a Opportunity for lormal networking initiatives (eg. collaborations) of comparable organisations working
together to implement the innovation a Complies with relevant regulations
and legislation
a People who have significant ties inside and outside the organisation and link the organisation externally can
capture and assimilate ideas for innovations a Adequate funding for health services, appropriate health
legislation and policies.

a clinician support (presence of leaders, degree of engagement in improvement effort, ability to lead change, supportive,
single leader — a clinical champion who is highly respected as an expert clinician, committed to using the intervention,
and has consensus-building skills to resolve conflicting views)
a design (improving adherence to an existing system or redesigning an existing system) and
implementation styles of improvement strategies a use of data (availability and acceptance of current evidence-based
research on benefits of intervention; using valid and credible data benchmarked against a reasonable comparison
group, to provide timely feedback to clinicians' on their use of the intervention) a contextual factors (hospital size,
affiliation and organisational turbulence — turnover of senior
management and clinicians, financial stress). 13 Four characteristics were present in hospitals in which more than 65%
of patients with acute myocardial infarction received (i-blockers on discharge:
a widespread sharing of and agreement with the goal of improvement (including substantial buy-in by clinical staff)

125
 ■ substantial administrative suppon (board of management and senior
management requesting performance data, senior management
participating in quality improvement project teams, and adequate
resources being provided to undertake the intervention)
■ strong clinician leadership (clinicians committed to using the
intervention) rather than just clinician participation
■ high-quality data feedback (that is credible to clinicians). Information
given to clinicians about their individual performance in
prescribing ^-blockers (described by one clinician as "objective enlighten -
ment") powerfully influenced their behaviour. Surprisingly, given the
considerable efforts being made to use and evaluate different strategies to
improve quality of care, the high- and low-use hospitals did not differ in
the type or style of initiatives they used to try to increase p-blocker usage.
This suggests that factors such as those discussed above are equally — if
not more — important in determining the success of implementing
interventions, and emphasises the importance of analysing the proposed
intervention, target groups and settings when designing implementation
strategies.
Importantly, staff in both hospital groups described change occurring
slowly, and difficulty in sustaining improvement A commonly described
scenario was "a cycle of change, characterised by inertia in the beginning,
a learning curve, accelerated improvement, a plateau, and either decline
or maintenance". 13 A subsequent quantitative study found that "hospitals
without strong administrative support and physician leadership for
quality improvement efforts are unlikely to effect desired performance
improvement".14
A list of determinants has been presented for individual health services
to identify the major determinants influencing the success of the
implementation of change. We will now describe theories of human
behaviour from a variety of disciplines from which practical approaches
and strategies can be developed to implement change.

6.2 Theories relevant to the implementation of change

and their practical implications for change in health
care
Attention should be given to theories that may be relevant in health care
and attempt to describe and explain human behaviour and how changes
take place in individuals and organisations. Theories should address the
questions: "Why do people or organizational entities behave as they do?
Given the way they behave, what would motivate them to change
behaviour?The theories presented here are from many fields, often outside
health care. No single theory addresses all the factors that influence
change, and the factors that they consider overlap considerably. 9
These theories allow hypotheses to be extracted about factors that may
influence the success of implementing change. 10 Using these hypotheses,
nhetacfco/sciantff
ic work underpinning
even some of the most
popular models for
change in health care is
striking'

Groi FtP. et al
Milbank Q 2007;
85 93-138
 6 Taking action to
close gape in care
and prevent
adverse events
recurring

appropriate interventions can be designed to overcome barriers and use

facilitators ("why do people behave as observed in this setting, and what
intervention could effect desirable change?") as part of an overall strategy
to implement the required change." Importantly, few theories have been
thoroughly tested, and there is only limited evidence that using theoretical
approaches in health care is effective in achieving change. 10
Interventions in clinical medicine have often been evaluated in ran-
domised controlled trials, meta-analyses and systematic reviews. However,
the evaluation of implementation strategies is mostly descriptive and
observational, rather than experimental. This lower level of evidence only
allows "educated guesses" or "speculative rules" to be made about how to
best support the implementation of interventions. 16
Many theories could be relevant to implementing change in health care.
As this book aims to provide a practical guide to change, the theories are
briefly described and the practical implications of these theories
summarised after the description. The theory descriptions have been
summarised from two detailed sources. 1017 Theories used extensively, or
thought to have significant potential in implementing change, are
described in more detail using additional sources. The aim is to use
theories as a source of practical suggestions about potential strategies for
achieving change. Many of these suggestions are incorporated into a
practical step-by-step framework for change, described in Chapter 7.
These theories can be grouped in two'ways. Firstly, they can be
categorised according to different approaches to change, including
educational, epidemiological, marketing, external influence, social
interaction, managerial and control and compulsion, 18 which will be
discussed later in this chapter. Some theories have contributed to more
than one of the approaches. Relevant theories can also be classified as
theories:10
■ of change in health care
■ focused on individuals
■ related to social interaction
■ related organisational context
■ related to political and economic context
■ related to patients.
The theories will be discussed using this classification.

6.2.I Theories of change in health care

6.2.1.1 Process theories

Process theories relate to implementation of change, including planning
the change and influencing the target group. Practical elements common
to these theories and relevant to health care include: 10
■ Use a systematic, well planned approach that considers all relevant
factors.
■ Involve the target group in developing and implementing the plan.
■ Carefully examine the characteristics of the innovation (eg, the evidence
base supporting the innovation, its complexity, compatibility with norms
and values, cost-bencfit analysis, associated risks, 'ttatao became dmr that
the involvement environmental factor*
which rrtbOed cOnidan
access to knowledge or
its application to overlay
practice were mote
important than clinician
ignorance or
pigheadedness in
retardmg4xogress to
best practice'

Scon i ACP J Club

2007 146 A6-A11
 6 Taking action to
close gape in care
and prevent
adverse events
recurring

required of the target group) looking for factors thai will facilitate or
impede its implementation.
■ Use a sequential approach to implementation, which solves different
problems at each stage.
■ Use diagnostic or problem analysis to determine factors that will assist
or hinder implementation.
■ Develop and use strategies that are closely linked to the results of the
diagnostic or problem analysis.
■ Monitor progress to determine whether the intended changes are being
achieved, and adjust the plan accordingly.
■ Use a cyclical approach of continuous learning and improving.

6.2.1.2 Stages of Change Theory

The stages of change or "readiness to change" model describes a sequence
of six stages of change within a cycle through which individuals progress
to successfully achieve change. 19-20 The model was developed to success-
fully change behaviour in individuals with addictions to nicotine, alcohol
and other drugs, and is now being used in changing other types of
behaviour. Each stage in the sequence requires specific strategies to be
used to facilitate progression to the next stage. The same model is used
whether the impetus for change is an external influence, an internal
commitment by the individual, or both. Most importantly, different
subgroups within the target group in a health care organisation will be at
different stages of the change cycle, and will require different strategies to
help them move to the next stage.
Individuals move through each of the stages of change at least once,
and frequently several times, while trying to change their behaviour. The
stages of change are as follows:20-21
1. Precontemplation — the individual has not yet thought about making
changes to his or her behaviour in this area, perhaps because of a lack
of basic knowledge, and may even avoid reading, talking or thinking
about such a change. The individual has no intention of changing in
the foreseeable future.
2. Contemplation — the individual is aware of the problem and has
started to seriously think about making a change, but their level of
ambivalence is high. On one hand, the individual wants to change this
behaviour; on the other hand, the individual wants things to stay the
same. The individual is weighing up the positive and negative aspects
of making the change and the personal effort needed to make the
change. The individual has not made any commitment to change and
has not done anything to make the change. Individuals may stay in
this stage for long periods of time.
3. Preparation — the individual has made a decision to change his or her
behaviour in the near future. The individual is unsure whether this is
the right decision, and is concerned about how to deal with potential
*.., knowtadgemustbe
problems that could arise from making the change. The individual has
emctedandmade
not yet taken effective action.
sooM entering into the
stock ot knowledge
constructed and stand by
other individuals.
Knowledge depends for
its circulation on
interpersonal networks
and will spread only if
these social features are
taken into account and
barriers overcome.'

"Greenhalgh T,
etal MilbankQ
2004;
82: 581-629
4. Action — the individual takes active steps to make a real change to his or her behaviour The
actions are visible, and may require considerable commitment and effort.
5. Maintenance — the individual continues successfully with this change in behaviour, which has
become consolidated and fully integrated.
6. Relapse — the individual had started to make changes in behaviour, but has slipped back into
original behaviour and an earlier stage in the change cycle.
At any one lime, individuals in groups will be* at different stages of the cycle of change. Most
individuals will be in the precontemplation and contemplation stages of the cycle. Few individuals
will be in the action stage. Each stage requires specific action to be taken by an individual to
successfully move to the next stage. The strategy used to facilitate change should be matched to the
stage of the change cycle individuals are in. Those in the:
■ Precontemplation stage may benefit from having their consciousness raised by creating
awareness and thinking about the consequences of not changing. ^
■ Contemplation stage may benefit from more information about the issue (especially evidence
about the benefits of the change) as they continue to weigh up the positives and negatives of
making the change.
■ Preparation stage may benefit from a discussion reinforcing their ability to make the change and
to consider undertaking a trial of the change.
■ Action stage may benefit from reinforcement of their changed behaviour using incentives, support
and resources to facilitate the change.
■ Maintenance stage made need support to fully stabilise and maintain the change to avoid
relapsing. Regular positive feedback may also assist in maintaining the change.
a Relapse stage may feel demoralised and need support to leam from their experience and use
different strategies for change the future.
If individuals are in the early stages ofthe cycle, forcing change on them (eg, by senior
management) is unlikely to be successful, and it may be detrimental. Individuals in the
precontemplation stage ofthe cycle who are coerced into changing by others may change their
behaviour while pressure is being applied, but their behaviour will most likely revert back when the
pressure is not longer being applied.
Individuals will probably experience some ambivalence throughout the change cycle, and the
degree of ambivalence will vary according to which stage in the cycle they are in. Ambivalence
usually reduces as an individual moves successfully through the cycle. Individuals are most likely
to give up making the change in the relapse stage. Relapse is a normal part of the change process,
and not a problem unless individuals despair about not being able to change and give up.
The stages of change model has been extended to behaviour in organisations, and in particular to
target groups within a health service, with the following strategies for the organisation attempting to
have the target group move to the next stage in the cycle 19,22 (Table 6.1).
The main implication of the stages of change model for health services is that different members
of the target group or subgroups will almost certainly be at different stages of the model regarding
any- proposed intervention. They will require the different strategies to help them move to the next
stage The model has been expanded and synthesised with other change models into a 10-siep
model invoking the stages of orientation, insight, acceptance, change and maintenance. 23
a Determine which stage each subgroup and key individuals in the target group is at, and use
appropriate change strategies to help them progress to the next stage.

6.2.2 Theories related to individuals

These theories discuss determinants of change in individuals — how they make decisions, how thev
learn and what influences their motivation to make changes.
 Enhancing patient care

IMUe 6.1 Stages of Change model

Stage Target group action Organisation's strategy

Piecortempiati Not aware improvement Raise awareness (eg, opinion leaders, academic
on possible detailing, workshops, conferences small group
discussions)
Contemplation Weighs up advantages and Increase understanding of change (eg, opinion leaders,
disadvantages of change academic detailing, workshops, conferences small
group discussions)
Preparation Makes plans made for change Reduce barriers and increase facilitators to change
(eg, modify practice environment, participatory skills
development, patient education, involvement in
development and adaption of guideline)
Action Changes behaviour Change behaviour (eg. modify practice environment,
participatory skills development, patient education,
involvement in development and adaption of guideline)

Maintenance Decides to continue change, or Maintain behaviour (eg, audit and feedback, reminder
relapses to previous behaviour systems)

6.22. / Cognitive theories

Cognitive theories examine how individuals think and make decisions. They assume that professionals
balance the advantages and disadvantages of the available treatment alternatives with the patient to
provide the best and most appropriate care. For this process to be accurately undertaken, high-
quality, convincing information should be available to professionals detailing the advantages and
disadvantages of the available treatment alternatives. Professionals sometimes use their previous
experience with patients with similar clinical problems to make decisions — in some cases, this
information may be inadequate or out of date.
■ Provide current, timely, relevant, convincing and high-quality (scientific) information to
professionals to help them make clinical decisions (eg, guidelines, decision aids and evidence-
based clinical pathways and protocols).

6.2.2.3 Education theories

For learning to occur, new knowledge should be actively used and linked to previous knowledge.
Motivation to change is increased if professionals are working on problems they have experienced in
practice. Motivation to leam is driven more by internal motivation (especially to improve personal
competence) than external pressure. Individual motives to learn and personal learning styles can vary
greatly. For example, some professionals need new experiences and accept the need for change
quickly. Others prefer to carefully examine all the options before making a change. Some professionals
require rigorous, logical arguments for change; others decide to change because of personal
experience.24
■ Seek to achieve change in clinical areas where professionals have experienced problems and
had to deal with their effects and involve them in finding solutions.
m Determine the learning needs and styles of individual professionals and tailor strategies for
change accordingly.
6.2.2.3 Attitude theories — Theory of Planned Behaviour
The Theory of Planned Behaviour states that individuals' behaviour is influenced by their intention to
perform that behaviour. Intention is influenced by their attitudes towards the behaviour (determined by

130
 6 Taking action to close gaps in care and prevent
adverse everts recurring

the expected outcome of the behaviour and whether these outcomes are positive or negative), by the
perceived subjective norm (Do important others behave this way?), and by perceived behavioural
control (Can 1 perform the behaviour in this setting?; Can I resist the social pressure?). 25
■ In the education program for the proposed change, opinion leaders should closely link the
change in behaviour to the desired clinical outcome and view the outcome positively.
■ Opinion leaders should model the behaviour to demonstrate that it can be done and that they
believe it is important.

6.2.2.4 Motivation theories

Changing behaviour within organisations requires motivation by individual to change. Unfortunately,
the commonest feeling towards change is ambivalence. The current situation is most familiar and
within individuals' comfort zone, whereas the required change will often take them outside this zone.
Therefore, changing individual behaviour can be difficult, particularly in organisations where many
individuals work.
An individual's drives and desires that make up motivation are complex. The large number of
theories developed to explain motivation probably reflects the complexity and limited understanding of
the subject, as well as the difficulty of incorporating all aspects of motivation into one cohesive theory.
Despite these difficulties, motivation remains a key determinant of whether a change can be
successfully implemented in an organisation. Therefore, we describe several motivation theories
relevant to individuals working in health care, and provide practical strategies arising from these
theories. Herzberg's Motivation-Hygiene Theory.26 This theory proposes that intrinsic factors (eg,
achievement, recognition and responsibility) are associated with job satisfaction and motivation,
whereas extrinsic factors (eg, salary, supervision, working conditions and relationships with work
colleagues) are associated with job dissatisfaction Hygiene factors are factors that reduce job
dissatisfaction. According to Herzberg, reducing job dissatisfaction does not increase motivation.
a Concentrate on achievement, recognition and responsibility to increase motivation.
Three Needs Theory.26 In this theory of motivation, individuals have three acquired needs: achieve ment,
power and affiliation. The need for achievement refers to personal (rather than organisational)
achievement, and is the desire to do something better than it has been done before. Individuals with a
high need for achievement like to be stretched. They are willing to accept personal responsibility for
finding solutions to problems, want rapid feedback about their performance and like to be set
challenging (but not extreme) goals. Easy goals do not motivate them.
a Provide responsibility for solving problems, challenging goals and rapid feedback.
Goal-setting Theory.26 This theory states that developing specific goals increases individuals' perform -
ance, and that difficult goals — when individuals accept them as appropriate — will increase their
performance. Individuals' participation in setting their goals increases their acceptance of these goals
and their commitment to achieving them.
a Allow individuals to set their own goals, and ensure they arc specific.
Reinforcement Theory.26 This theory states that behaviour is the result of conscqucnces The
consequences of a particular behaviour reinforce that behaviour, thereby increasing the likelihood that
the behaviour will continue or recur. Therefore, employees should be rewarded if they arc behaving in a

131
Enhancing patient care

manner thai is in accordance with the organisation's wishes and the

organisation desires that this behaviour continue. The sooner a reward is
provided for desired behaviour, the more effectively it will reinforce it.
Importantly, punishment in response to an undesired behaviour is only
temporarily effective in changing behaviour, and may also result in
dysfunctional behaviour. If undesirable behaviour is exhibited, it is often
best dealt with by non-reinforcement, such as ignoring the behaviour or
not rewarding it.
■ If possible, reward appropriate behaviour shortly after it occurs.
Jobs Characteristics Model. 26 Designing motivating jobs can increase moti-
vation. The Jobs Characteristics Model seeks to increase work's
meaningful- ness by increasing the number and variety of skills that an
employee is required to possess to complete the task; increasing task
identity by allowing the employee to complete whole pieces of work, rather
than small pieces of a larger work, and increasing the significance of the
tasks to complete; providing increased responsibility to complete the task
through increased autonomy; and providing information to the employee
on the their performance through feedback. Using this model allows
employees to leam about their performance on a task they care about.
This approach is most suitable for employees with a high growth need,
bui is not suitable for low growth need individuals. This difference in the
suitability of the jobs characteristics model between individuals
emphasises that, despite the large number of models of motivation thai
have been developed, on some occasions an individual approach will be
required to motivate employees.
■ Allow individuals to work on whole tasks rather than parts.
a Provide appropriate training where required to increase skills,
m Increase autonomy and responsibility, m Provide feedback on
performance.
Expectancy Theory.26 This is the most comprehensive theory about motiva-
tion. According to this theory, an individual has an expectation lhat if they
exert themselves at a particular level the effort will be followed by an
outcome, and the outcome will have a degree of attractiveness. The theory
attempts to link motivation or individual effort to individual performance,
the rewards offered by the organisation and the level of satisfaction
provided towards achieving the individual's goals. The theory strongly
relies on the individual's perceptions of performance, rewards and goals.
This perception will vary according to individuals' personalities and needs,
and in some instances, the individual's perceptions will noi be based in
reality.
Expectancy theory considers individuals' perceptions of the rewards ihe
activity will provide, how attractive the rewards are 10 individuals, the
effort individuals must make to achieve ihe rewards, and individuals'
perceptions of their abilities and skills to successfully undertake ihe
organisational change activity. To maintain individuals' motivation, there needs to be a strong
found in the literature
link between the amount of effort lhat they make and the performance
gma the impression
that change is (or
can be) a rational,
controlled and orderly
process. In practice,
however, organisational
change is chaotic, often
involving shifting goals,
discontinuous activities,
surprising events and
unexpected
combinations of changes
and outcomes'

lies V. Sutherland K
Organisational
change: a review tor
health care
managers,
professionals and
researchers
 6 Taking action to
close gape in care
and prevent
adverse events
recurring

they achieve. The reward provided by the organisation should be attractive

to
them and consistent with their goals. Using the expectancy model, motivation will decrease if
individuals perceive that, despite much effort, they cannot achieve a high level of performance, that
the reward provided by the organisation is not appropriate; and the reward is not consistent with
their goals.
a Provide adequate support so that effort by an individual will result
in performance. a Tailor rewards to the individual's
personality and needs.
6.2.2.5 Motivating professionals
The factors that motivate employees vary across different groups within organisations. As well as
reflecting individual personality and skill differences, these differing motivating factors may also
reflect different needs and goals between individuals of different groups in an organisation. Improving
quality and safety in health services involves working predominantly with the professienals who
deliver clinical care. Research has found that some motivating factors are specific to professionals in
the workforce.26'27 Professionals are strongly committed to their profession, which they may view as a
lifefc work. They are prepared to work variable hours and have a strong need to keep up-to-date in
their fields. They like to work on important areas in their field, be challenged by problems and find
solutions. They also like to be supponed by the organisation they work for. Professionals are
motivated by challenging problems, and like autonomy to work on what interests them and flexibility
in solving problems and implementing solutions.
Appropriate rewards for health care professionals include: providing opportunities, supponed
financially and with locum cover, to attend education programs to maintain knowledge in their area
of expertise; public and private recognition of their efforts in solving important problems in the
organisation; acknowledgement of their value to the organisation and real interest from management
in the areas and problems they are working on. A major difference between other groups and
professionals in an organisation is that professionals are well paid, and therefore may not rank
financial reward highly as a motivating factor.
Ensure:
a The proposed change is in an area clinicians consider important. a
The problem to be solved is challenging. a Adequate support (eg,
resources) is provided.
a Autonomy and flexibility are given to solve the problem and implement the solution, m
Rewards are provided to increase the knowledge of clinicians, a Recognition is given of the
efforts of clinicians to solve important problems. a Senior management has a genuine
interest in the problems clinicians are trying to solve.
How hard and long individuals will persist in trying to carry out a task to achieve a goal depends
on their beliefs about their ability to successfully complete the task. If individuals have previously
successfully completed a task, or if they believe they can successfully complete the task, they will be
motivated to try Therefore, it is important that organisations realistically assess their employees'
abilities. Appropriate training should be provided to aid motivation if organisations believe their
employees abilities are below that required to successfully undertake a specific task.
a Assess whether clinicians have the knowledge and skills to solve the problem, and if not,
provide the required training.
6JL2.6 kmombon Diffusion Theory
Diffusion of innovations in health care is often slow, and provides a major challenge to clinicians and
health services. There are three "clusters of influence" that correlate with the rate of spread of an
innovation:16-20 the willingness of different groups to adopt the innovation, characteristics of the
innovation, and the degree to which changes can be made to the innovation locally.
Innovation Diffusion Theory recognises different groups with regard to their willingness to adopt
innovations.28 The population to whom the innovation may spread are divided into five statistically
 6 Taking action to close gape in care and
prevent adverse events recurring

defined, but somewhat artificial, categories relative to the average time for adoption of the
innovation:16'28
■ Innovators (venturesome; 2.5% of those to whom the innovation may spread) are the fastest
adopters. Innovators are adventurous, cross boundaries and are able to tolerate risk. Clinical
innovators are focused on new ideas and are personally heavily invested in a specialised topic. The
search for innovations is best undertaken by the innovators in the service. Importantly, innovators
influence early adopters.
■ Early adopters (respectable; 13.5%) have high status in the organisation and are regarded as opinion
leaders. They speak with innovators and are influenced by them. Early adopters speak with each
other and select what they would like to try with the aim of reducing uncertainty. The early majority
adopters watch the behaviour of the early adopters.
■ Early majority adopters (deliberate; 34%) watch the early adopters and are influenced by ihem. The
early majority will try what meets their immediate needs rather than just interesting ideas.
■ Late majority adopters (sceptical; 34%) are sceptical about change, but the opinions of their
colleagues are important to them. They watch the early majority for local proof (rather than proof
from external sources) that it is safe to adopt an innovation. They adopt an innovation when it
becomes standard practice.
■ Laggards (traditional; 16%) resist change. They are traditionalists who use what is tried and true.
When 15% to 20% of the potential adopters have adopted the innovation, sufficient momentum has
been generated and it is difficult to stop further dissemination. 28
Opinion leaders are respected, trusted professionals who represent the social norms in the selling
and are role models for other professionals in the network. Although not innovators, they are
potentially key "facilitators, supporters and problem solvers" 10 in any significant change. Others in the
network trust them to compare the proposed innovations with the current norms and provide a valued
opinion. The opinion leaders in the health service should be identified to use "the social and
communication networks that are naturally developed by ... practitioners". 29 Randomised controlled
trials have shown that using opinion leaders to promote evidence-based practice enhances clinicians'
adherence to such practice by an average of 10% in absolute terms.30
Initially, clinicians may be sceptical about the relevance and transferability of research findings to
iheir patients. Therefore, they look to clinicians they trusi and regard as leaders to use ihe research
findings or innovations, evaluate them and then provide information about their effectiveness and
relevance to local needs and conditions. Opinion leaders remove the risk of using the findings or
innovations for other clinicians. 29 Most contact between opinion leaders and other clinicians occurs
informally in hospital wards, telephone consultations and during hospital sponsored events. Formal
communication channels, such as ground rounds and pharmaceutical sponsored events, are far less
commonly used.29
However, it may be difficult to define and identify opinion leaders. There is little knowledge about
what opinion leaders do or how they do it, and whether they have particular traits that affect their
success. The attributes of an effective opinion leader cannot be confidently defined.' 0 Opinion leaders
can be identified by asking clinicians who are the opinion leaders, or asking them to judge others from
a preselected list of possible opinion leaders. 30 The former method may result in incomplete
identification of opinion leaders if survey response rales are low.
Different opinion leaders may need to be identified to promote changes in different or complex
clinical conditions and to different target groups. 30 The mean number of "technical advice" opinion
leaders in each health care organisation was five (range, 3-9). 30 In a survey of all clinicians in 17 health
organisations in the US, some clinicians identified as opinions leaders were not recognised as such by
directors of medical service. 29 Many opinion leaders were proud of being viewed as such by their
peers.29 Opinion leaders are important messengers, and should be chosen to actively participate in
projects to implement change.
■ Health services should identify innovators in the organisation and provide support for them to
systematically search for suitable innovations supported by scientific evidence by formally searching
the scientific literature, attending scientific conferences and visiting exemplar health services or
services with innovative programs.16
m Once innovators have identified potential innovations, health services should facilitate and support
meetings between innovators and early adopters, so that innovators can report on their findings, m
Health services should then provide the time and resources required for early adopters who wish to test
the changes on a small scale. These early adopters should be asked to make the changes highly- visible
so that the early majority (especially those in the target group) can observe the process.' 6 m Following
successful small scale-testing of the change, the health service should encourage social interaction
between early adopters and the early majority. Early adopters could speak to the early majority as a
whole or individually. Face-to-face communication is most effective in providing information about the
effects of innovations. Importantly, the early majority should receive information about the change from
someone in their local network who they believe is credible.' 6 m Throughout these processes, it is critical
that health services ensure there are frequent opportunities, both formal and informal, for extensive
communication between these various groups.
Perceptions about an innovation have a significant influence on the rate it spreads, and five
perceptions are most influential:16,11
a Benefit of the change ("is it a better idea or practice than that which exists now?") — if knowledge
about the innovation can be provided to individuals and reduces their uncertainty, they will be more
likely to adopt it. The more unusual or unfamiliar the innovation, the greater the uncertainty,
increasing the need for information about the innovation for adopters to accept it. a Compatibility with
values, beliefs, past history, structures, procedures and current needs ("does it fit in with the existing
circumstances?") — if current practices are acceptable to clinicians, it will be difficult for them to adopt
changes, a Complexity of the innovation ("is it too complicated for our needs?") — degree of difficulty in
implementing innovation. Simpler innovations spread faster, so sometimes innovations are simplified or
re-invented to aid diffusion. If possible, adopt the high leverage (responsible for most of the result of the
change) components of the change (eg, guideline), a Trialability ("can it be piloted and modified to fit
our purposes?") — belief that the change can be
tested on a small scale without requiring extensive lime and resources, a Observability ("can we see it
working well in our environment?") — can first waich others implement change.
Changes to interventions or "reinvention" at the local level are a normal part of ihc successful
dissemination of innovations.
■ W/h tc possible, the innovation should be perceived to be benrficial, com/Hit iblc. simple, Itialahlc antl
obsc noble.
a Health services should cncoura^c intcncntions to be adapted to local needs and conditions.
 Enhancing patient care

Finally, motivation for behavioural change has a number of properties that should be kept in mind
when trying to implement change in an organisation:21'32
■ An individual's level of motivation to change is not constant over time.
■ Motivation is task-specific.
■ Taking action creates motivation.
■ Although goals are important to achieve behavioural change because they provide a direction for
actions to take place, goals alone do not create motivation.
■ Mixed feelings or ambivalence commonly accompany change. Total commitment to change is not
required to commence change. If an individual feels more positive than negative about change, it can
commence.
■ Change is not the result of a simple decision.
■ Change is an ongoing, complex process, not a destination.
■ Change can move individuals out of their comfort zones and the natural response is to resist the
change. Change may be an opportunity for an individual to leam and develop, to use creativity and
initiative and offers challenges and stimulation, but it may often be perceived as threatening. However,
the discomfort individuals feel during the change process may indicate they are moving forward.
■ Individuals have different needs, personalities, skills and attitudes, and therefore different
motivators. A uniform approach to motivating a group may not be successful — some individuals will
need a personal approach to increase their motivation.

6.2.3 Theories related to social interaction

6.2.3.1 Communication theories

Theories about communication aim to change individual attitudes and behaviours using effective
communication. The characteristics of the source of the communication (eg, credibility), the recipient
(eg, prior knowledge) and the communication itself (eg, repetition, perceived validity, personal
relevance and functionality) are all important.
■ Adapt the message to the target group's knowledge, skills and motivation.
■ Use convincing messages presented by credible individuals, and use them repeatedly.
■ Allow time for the target group to absorb, understand and accept the message.

6.23.2 Social Learning Theory

Social Learning Theory assumes a continuous interaction between the professional, his or her
performance and the social environment, reinforcing each other in changing behaviour. Factors that
explain changing behaviour include the individual's ability to learn by doing, observing others
modelling the behaviour (demonstrating that it can be done and leads to expected results), and
reinforcement by positive feedback from important others in the setting.
■ Professionals need to observe each others' behaviour.
m Professionals need to observe the behaviour of respected peers and opinion leaders as role models, m
Respected peers need to provide positive reinforcement of desired behaviour to professionals.

6.23.3 Social network and influence theories

136
 6 Taking action to close gape in care and
prevent adverse events recurring

The Social Network Theory descnbes how change is influenced by the structure of the social network.
Individuals do not behave in isolation, and uptake of change may depend on the strength of lies
between individuals in the network, the proportion of people who have already taken up the change
and the individuals threshold for taking up change.
■ Enhance interaction between professionals who have and have not taken up the change.
The Social Influence Theory describes how an individual's behaviour is influenced by the behaviour
they observe in others, the social norms in the network and the formal or informal exchange of
opinions with peers and important others in a network. Change may occur after a local consensus is
reached. Interactions in the social network, views of opinion leaders, expectations of significant peers
and availability of education are important factors in facilitating or preventing change.
Opinion leaders play a key role in these theories. Surprisingly, clinicians rarely direcdy access research
or clinical guidelines, but rely on "mindlines", or internalised, tacit guidelines that are developed by
brief reading, but mainly by experience and through formal and informal interactions within
professional networks with opinion leaders, and interactions with patients and pharmaceutical
representatives (who are often regarded with considerable scepticism). 33 Opinion leaders play a key role
in influencing the clinical decisions made by other doctors who respect them. Importantly, opinion
leaders' knowledge should be based on the best available evidence.
a Study interactions in the network and identify the opinion leaders.
m Seek the support of opinion leaders for the proposed change.
m Emphasise to opinion leaders that they have a key role in modelling the behaviour required for the
proposed change, changing the social norm in the target group and disseminating information.
a Ensure that opinion leaders have ready access to high-quality evidence about best practice.
a Ensure that adequate opportunities are available for networking formally and informally between
clinicians and those they regard as opinion leaders.

6.2.3.4 Theories related to team effectiveness

Teams' success relies on many factors, including: having clear, common goals; the presence of a team
champion; the sharing of information; trust; participative decision making; encouragement for new-
ideas; few conflicts; a commitment by team members to perform to the best of their ability; support for
the proposed change; and the size and mix of the membership of the team.
a Develop common goals and targets.
m Encourage participative decision making.

6.2.3.5 Theories on professional development

Health professionals have considerable autonomy in making decisions about clinical care. They have a
strong loyalty to their profession, which is probably greater than their loyalty to their hospital
Professional standards set by professional bodies influence the behaviour of individual professionals.
Therefore, where possible, it is useful to propose changes that are consistent with professional
standards, as such changes should be easier to implement.
a Where possible, seek a standard for the area of change being proposed developed by the relevant
professional body and use it to develop a local standard.
a Use professional pride, professional standards and loyalties to help implement change.
AJ.3.6 Theories on leadership
Leaders can be influential in facilitating or obstructing change. The leadership required by an
organisation may vary with the type of change being proposed. Leaders may have formal authority or be
informal leaders by possessing valued expertise or skills or belonging to a key group in the hospital.
Enhancing patient care

Leaders can support change by achieving goals or changing the organisation's culture. The involvement
and commitment of senior hospital management is a critical to the success of any proposed major
changes to patient care — it can increase productivity and staff satisfaction. 34,35 Management staff can
demonstrate their commitment to major change projects by making them pan of the hospital's strategic
plan.
Strategies that consistently improve quality include: engaging physicians, training personnel, building
systems, effective delegation and accountability, personal involvement and modelling of values, a
flexible strategy with resources, and creating a vision. 36
■ Identify the formal and informal leaders in the relevant area.
m Inform leaders about how they can use their power and influence to continuously support the
proposed change (eg, initiating activities, modelling the desired behaviour).

■ Ensure the support of senior management for the proposed change. 6.2.4

Theories related to organisational context

6.2.4.1 Theories of innovative organisations
The characteristics of organisations that adopt change more quickly and easily include: highly
specialised individual roles, a high level professionalism, decentralised decision making, available
technical knowledge, good internal and external communication and a positive attitude to change
among leaders and managers.37
■ In health care organisations, specific individuals, more than the organisation as a whole, have been
shown to be responsible for and influential in achieving change
m Engage the support of the professional who is most influential regarding the proposed change.
m Promote participation in the change process.
m Delegate decision making about the change to multidisciplinary teams.
■ Increase the level of responsibilities at the clinician-patient interface.

6.2.4.2 Theory of Quality Management

Total quality management or continuous quality improvement "refers to a management process
directed at establishing organised continuous improvement activities, involving everyone in an
organisation in a totally integrated effon toward improving performance at every level". 39 Quality
management continuously examines the processes in a system in an iterative approach using a plan-
do-study-act cycle. According to this theory, change can only be successfully achieved by changing the
system. The process is organisation-wide, uses multidisciplinary teams and is patient-centred.
Clinical activity is continuously monitored, data collected and variation in care is analysed statistically
and if appropriate reduced. Uncontrolled variation in processes is regarded as the cause of quality
problems and requires control by those working at the front line to improve quality. 40 There is a
continuous cycle of monitoring, analysing, making appropriate changes and then continuing to monitor
their effects.
Total quality management requires strong leadership to create an appropriate culture in which staff are
a key element. Physician involvement in the process and concentration on patient care, rather than
cost saving, is crucial 41 However, quality management has not had a dramatic impact in health care,
as it has in other industries. Importantly, the evidence of its effectiveness elsewhere was anecdotal and
usually comprises case descriptions of its implementation in individual companies written by their staff
40

Senior management staff in administrative areas of hospitals have used quality management, but
there has been little evidence in the scientific literature of its effectiveness in improving clinical
outcomes. Many clinicians are sceptical of the process and view it as a passing fad. The decentralised
control lhat is a major feature of quality management can be threatening to senior management.
 6 Taking action to close gaps in care and prevent
adverse events <*cumng

Successful innovation in organisations requires a customer focus (particularly understanding their

needs), suppon from leaders, the organisation keeping to its mission, and obtaining information about
best practices from other organisations.42
Newer methods of quality management have met some of the requirements discussed here, but their
effectiveness in terms of cost, time demands and applicability to clinical practice has not been
studied.43
■ Examine processes of care in the system for which change is proposed and change them if necessary.
a Use an ongoing cycle of continuously monitoring activity, analysing the results, planning and
mahing changes and monitoring their effects.
6.2.4.3 Process Redesign Theory and theories of integrated care
These theories aim to improve the organisation and management of care processes so that optimal
patient-centred care is provided. They involve analysing and redesigning multidisciplinary care
processes and using clinical pathways. The redesign may be radical or gradual, and can involve task
reallocation, information transfer and greater coordination of care between health professionals. A case
manager may be appointed for areas where clinical care is complex. The redesign may involve the
delegation by medical staff of some duties to nursing, allied health and non-clinical staff. Reminders to
healih professionals and feedback about their performance are imponant elements of the process.
a Analyse, reconsider and redesign related processes of care that make up a system so that they cm be
optimally delivered and integrated into normal routine patient care.
6.2.4.4 Complexity Theory
Health care is a complex system with many interconnected components that interact continuously and
can behave unpredictably. These interactions are more imponant than actions on the components of
system. Small influences in one part of system can have large effects on other pans of system, so
targets for parts of systems may not be met. The system should be viewed as a whole, rather than
concentrating on its parts, and simple goals should be set for the whole system.
The conditions under which strategies are developed to implement a change in practice will change
over time. Therefore, the strategies initially developed may require change as implementation
progresses. Dealing with changed conditions may be assisted by an external facilitator, who maintains
regular contact with opinion leaders and other relevant clinicians participating in the change project to
help with problem solving and deal with unanticipated challenges as they arise. 15
6.2.4.5 Theory of Organisational Learning
A learning organisation can change its behaviour when new knowledge is acquired by individuals in the
organisation. Barriers to organisational learning include problem solving in pans without considering
interactions, changing by reacting rather than continuously trying to improve and focusing on
competition rather than cooperation leading to an emphasis on short term results rather than long-
term solutions
A learning organisation:44""1 a uses the collective learning
experience of all employees a knows what is happening
outside the organisation

1»
 Enhancing patient care

■ allows its employees to learn through open communication, making all information (positive and
negative) readily accessible and allowing debate and experimentation while tolerating failure and
supponing ongoing education at all levels
■ uses measurement as a source of learning
■ has leadership that is involved, interactive and suppons staff and change
■ has a common picture of the gap between actual and desired performance
■ supports new ideas at all levels
■ demonstrates its learning by changing routines and embedding the change.
The components of a learning Organisation will be discussed further in Chapter 9.
■ Ensure the collective knowledge in the organisation is used to change behaviour.
m Encourage continuous information sharing and experimentation in problem solving.
■ Find out how other organisations are dealing with this problem.
m Ensure leaders support the change project and are actively involved, m
Offer opportunities for continuous learning.

6.2.4.6 Theories on organisational culture

Organisational culture is difficult to define. Is culture "something an organisation possesses" or does
culture itself define "the whole character and experience of organisational life"? 47 Is the culture of an
organisation linked to its performance? If so, can a change in an organisation's culture change its
performance? 47 It is postulated that there are four types of cultures depending on the position of the
organisation along two continuums: flexibility-control and internal-external orientation. Changes in
teamwork, flexibility and external orientation can stimulate changes in performance. However, cultures
emphasising formal structures and regulations are detrimental to quality improvement activities.
■ Aim to create teamwork, flexibility and an external orientation.

6.2.5 Theories related to the political and economic context

6.2.5.1 Economic (reimbursement) theories

Reimbursement may influence individual and organisational performance. The nature and liming of
reimbursements and incentives for clinical services influence how care is provided. Fee-for-service
payments result in additional actions being undertaking by clinicians, which can be targeted toward
implementation of quality improvement activities.
■ Use material and non-material rewards and incentives (eg, increased budget or staff, support for
educational activities) to encourage and reward appropriate behaviour and good results.

6.2.5.2 Theory of Contracting

Quality standards and targets could be incorporated into clinicians' contracts. The contractual
specifications could include the use of specific clinical practice guidelines and targets for process
criteria for recommendations in these guidelines.

6.2.6 Theories related to patients

6.2.6.1 Theories on factors related to patients

Patient expectations and behaviour and their care providers' perceptions about these expectations can
influence care. Patients could be informed about major programs to change clinical care. They could be
informed of large change programs via news media, and local quality improvement initiatives when
they

140
 6 Taking action to close gape in care and
prevent adverse events recurring

are admitted to hospital. Caution would be needed to ensure they correctly understood the aims of
the program and how it affects their care during their admission. Patients could remind
professionals about this part of their care during their stay in hospital.
a Consider whether patients could act as facilitators of change in any change program.

6.3 Approaches to change

Numerous theories of change in human behaviour have been described and practical strategies
distilled from each. Many of these theories and strategies have been grouped together to develop
several general approaches to change. These approaches will be briefly described and various
strategies described in detail.
Different disciplines have developed different approaches to implementing change. These
approaches are based on assumptions about how people behave individually, in groups and in
organisations. These assumptions are, in turn, based on many theories of behavioural change and
have been used in different approaches to assist in understanding and implementing change in
health services. These approaches can be divided in two groups of influence on behaviour — internal
and external to the individual. 18
Internal influences include a desire by some individuals to achieve competence in a particular area
(small group interactive education sessions are used); others have a desire to make rational
decisions based on evidence (evidence-based guidelines are used) and marketing influences
(products are designed to meet an individual's needs and influence their behaviour). In these
approaches, change comes from within the individual and satisfies an inner need, and the individual
takes personal responsibility for making the change. The emphasis is on processes internal to the
individual and is found in the educational, epidemiological and marketing approaches. 18
External influences include the conditioning of particular behaviour (using feedback, reminders
and incentives); social influences (the opinion of experts are important); a managerial approach (the
system is redesigned to influence behaviour); and the use of external power (using legislation,
regulation and discipline). The change comes from external influences and is found in the social
interaction, managerial and control and compulsion approaches. 18
These approaches are listed below together with associated interventions and strategies (adapted
from Grol18) (Table 6.2). There is no evidence that any one of these approaches is more effective than
another.

6.4 Developing change strategies

Multiple theories describing individual and organisational behavioural change have been developed
However, there is no single coherent and unified theory to explain the factors that facilitate and
impede change. The lack of a theory and the absence of robust, generalisable evidence for
implementation makes it difficult for health professionals and senior managers to make rational
decisions about which change strategies to use. 48 "We generally have very limited understanding of
the factors that determine the success of a quality improvement intervention." 4
A theoretical framework for interventions is needed to explain and predict the behaviour of
individual health professionals, clinical teams, health organisations and health systems.
Implementation strategies should be evaluated through the stages of theory, modelling, exploratory
tnal, definitive randomised controlled trials and long-term implementation — similar to the process
of evaluating new medications
Enhancing patient care

Ifcble U Approaches to Implementing change

Owaoach Focus Intervention; strategies

Education* Internal motivation to Interactive small group learning

achieve
professional competence Problem-based learning using problems faced each day
Epidemiologica Rational decision making Provide convincing and credible evidence (eg. guidelines)
l
based on evidence
Moftoting Attractive proposal for Conduct a needs analysis — adapt change to meet needs
change
based on target group needs Adjust change to local conditions
Disseminate change widely — personally and through
media
External Behaviour directed by Feedback about performance — compared with others
influence external influences before Reminders
or after a specific action — Rewards
conditioning Sanctions
Social Influenced by other people Using opinion leaders to influence others
interaction considered important Outreach visits from experts — academic detailing
Peer review
Patient pressure to use an innovation
Managerial Creating conditions for Changing system
change Redesigning care processes
Poor quality is equated with
system failure
Control and Use external power to Legislation
change
compulsion behaviour—will want to avoid Regulation (eg. accreditation)
negative consequences Contracts
Disciplinary measures
Budgeting

Without such evaluation, current implementation research is similar to introducing an

antihypertensive drug without understanding its pharmacology, the physiology of blood pressure, the
pathophysiology of blood pressure control, and conducting trials in animals. 49
Where possible, strategies for implementing change should be considered that match the determi -
nants (barriers and facilitators) for the change, the systematic determinant analysis of the proposed
change, and the target group, patients, organisation, and sociopolitical context. 22 Each part of the
change may need its own implementation strategy. Further, more barriers and facilitators will
probably be identified than can be appropriately addressed, so barriers and facilitators should be
ranked according to their imponance and changeability. 22
Often, interventions for overcoming barriers will be developed using "evidence of effectiveness when
available, as well as experience, commonsense and creativity", 50 theory, personal preference, intuition,
familiarity22 51 and personal and potentially biased beliefs about human behaviour and change, 10 rather
than formal and objective methods.
When people are planning changes they often adopt a naive and opportunistic attitude. A
strategy is usually chosen quickly and often does nor produce the expected result ... Until we
have gained a better understanding, the most practical advice to individuals responsible for
 6 Taking action to close gape in care and
prevent adverse events recurring

changing and improving practice is to be aware of their own assumptions about human
behaviour and change.10
 Importantly, the barriers to effective implementation may be due more
to organisational problems than to individual doctors who, after
superficial analysis, are often cited as the barrier.
The rationale for choosing particular implementation interventions for
research is usually not given. Analysis of barriers often influences the
content of the implementation intervention, rather than the selection of
the type of intervention. Design of implementation interventions is still in
its infancy and "the translation of identified barriers to tailor made
implementation interventions is still a black box". 51 A review of the use of
guideline dissemination and implementation strategies found few studies
that explicitly stated the rationale or theory behind why particular
interventions were selected, and selection could have been based on an
assessment of causal mechanisms or a "kitchen sink" approach. 48 "It is
plausible that multifaceted interventions built upon a careful assessment
of barriers and coherent theoretical base may be more effective than
single interventions." 48
Strategies to promote professional behavioural change that could be
used to implement research findings are: (adapted from Grimshaw and
colleagues52):
a educational materials — distribution (by mail or personal delivery) of
recommendations for clinical care, including clinical practice guidelines
a participation by health care providers in conferences, lectures and
workshops
a local consensus process to develop an appropriate agreed guideline and
intervention process a
educational outreach visits
a local opinion leaders
a patient-mediated interventions — specific information sought from or
given to patients a
audit and feedback a
reminders
a marketing — personal interviewing, focus groups, surveys of targeted
providers to identify barriers and the design of interventions to address
these barriers
a multifaceted interventions — any intervention lhat includes two or more
of ihe above.
Some barriers, such as lack of knowledge, have a logical strategy to
overcome ihe barrier, but may require additional resources. However, for (Many concept*
some barriers, such as changing the organisational culture, there is no described here] can ba
logical, directly applicable strategy. Other barriers, such as clinician implemented in many
workload and lime availability, require senior management input to different ways, not at of
them observing the core
address, which should be a key role of ihe executive champion on the
principles although they
change project team.50 As with best practice in clinical medicine, where
may sport the label.
possible, ihe strategies chosen to implemeni change should be supported
lies V. Sutherland K
Organisational
cnange a review lor
health care
managers, protessio-
ials and researchers
 6 Taking action to
close gape in care
and prevent
adverse events
recurring

by high-quality scientific evidence of their effectiveness. Although such

evidence is often limited, a literature search should still be conducted as
part of the planning process to determine
 Enhancing patient care

6J IwptoxarwnU achieved with Implementation interventions*4

No. of
Msdlsn absolute improvement randomlaad (range)
controlled trials

Dissemination of educational materials Mufti faceted interventions

Audit and feedback incorporating education
Reminders outreach Opinion leaders30
+8.1% (+3.6% to+17%) 4
+7.0% (+1.3% to+16%) 5
+14.1% (-1.0% 10+34%) 14
+6.0% (-4.0% to+17.4%) 11
+10% (-6.0% to +25%) 12

which strategies, if any, have been previously used to successfully

implement similar changes in similar settings.
Adding further complexity to the selection of strategies is that dilferent
strategies may be needed to implement change in each of the subgroups
of the target group. The effectiveness of individual implementation
strategies varies with many of the determinants discussed earlier, and a
strategy effective in one setting may be ineffective in another. Using a
combination of strategies may increase or decrease the overall
effectiveness of the plan, be more difficult and require more effort to
implement, and may increase costs. Also, with multiple strategies, there
will be a limit to the amount of energy within the organisation available to
implement change at one time.

6.5 The evidence about change strategies

The first systematic review of trials of interventions to improve
professional practice (aptly entitled No magic bullets) examined 102
studies published between 1970 and 1993, and classified strategies
according to their effectiveness as:
■ Generally ineffective — lectures, publication of research and mailing of
information.
■ Variably effective — audit and feedback, local opinion leaders, local
consensus and patient mediated intervention.
■ Generally effective — reminders to perform actions (manual or
computerised), small group interactive problem based educational

144
 6 Taking action to close
gaps in care and prevent
adverse everts recurring

meetings, academic detailing and multiple strategies chosen after an

analysis of potential barriers to change.33
A review of 41 systematic reviews of interventions 5 years later found
modest improvements with active (eg, reminders and educational
outreach) and multifaceted interventions, but passive interventions (eg,
mailing guidelines to targeted clinicians) when used alone were largely
ineffective in changing behaviour. Audit and feedback and the use of local
opinion leaders were variably effective approaches. Reminders and
educational outreach aimed at changing prescribing behaviour were
found to be generally effective. No strategies were effective under all
circumstances. Multifaceted interventions aimed at addressing potential
barriers to change were more effective than single interventions, 52
although it was difficult to separate the effective components. A further
review in 2004 showed modest but consistent median absolute
improvements of approximately 10% in process measures, with wide
variations in effect size54 (Table 6.3).
There were two very different conclusions in this review from earlier
ones — there was no significant difference in the effect sizes of multiple
and single interventions, and passive

145
 Enhancing P*twnt care

interventions showed modest but consistent improvements. 54 For exam-

ple, the use of printed educational material resulted in improvement and
has the advantages of low cost and feasibility in individual health
services.48 Further, the effects of multiple interventions did not increase
with the number of interventions used. 48 In conclusion, "uncertainty
remains as to the likely effectiveness of different quality improvement
strategies",48 and "general conclusions about what works are still tenta-
tive."4
Funher complicating the issue are the results of a review by Shojania
and colleagues of the effects of 11 quality improvement strategies on
glycaemic control in patients with type 2 diabetes. 55 They reviewed 66
trials (including 50 randomised controlled trials). As with previous reviews
of quality improvement strategies, most strategies produced small to
modest improvements in glycaemic control. However, two strategies, team
changes and case management, were associated with the largest pooled
reductions in glycated haemoglobin.
Case management was defined as "any system for coordinating
diagnosis, treatment or ongoing patient management ... by a person or
multidiscipli- nary team in collaboration with or supplementary to the
primary care clinician". Team changes were defined as "changes to the
primary care team including adding a team member (eg, nurse specialist
in diabetes care, pharmacists, nutritionists, podiatrists) or using
multidisciplinary teams ... in ongoing management, or expansion ... of
professional roles (nurse ... adjusting medication)". 55 An evaluation of the
resources required to use each strategy was not made, but it is likely that
case management and team changes are resource-intensive when
compared with other less effective strategies.

6.6 Change strategies commonly used at the

Wimmera Health Care Group
The effectiveness of most change strategies is disappointing. Health
services have therefore looked to other industries for guidance. Quality
management and analysis and redesign of service delivery processes and
systems (described earlier in this chapter) have been used extensively to
reduce errors in other complex industries. Briefly, quality management
(also called continuous quality improvement or total quality management)
involves a continuous cycle of analysing quality problems, determining
solutions, taking appropriate action, monitoring the effects of these
actions and then determining whether further action is required. In
process redesign (also called re- engineering), the processes that make up
a health care delivery system arc analysed and redesigned with the aim of
optimising patient care.
We have found these strategies effective in improving the quality and
safety of health care delivery at the Wimmera Health Care Group,
•.wh*n(ta*9,wmf
management] hm been
tried in hoapt^t eo far
doctor* are not
effective on quality
improvement teams.
They arrive lata or not at
all to the meeting*, they
dominate when they are
present; and they
sometimes leap to
solutions before the team
has done its « proper
diagnostic work on the
process *

Berwick D etal
BMJ1992: 304 304
306

146
 6 Taking action to close
gaps in care and prevent
adverse everts recurring

although the results we have obtained may be specific to the contcxt in

which we work

147
 Enhancing P*twnt care

Some authors have criticised the use of these strategies because they
have not been evaluated to the standard of clinical research, and their
effectiveness is only supported by observational and anecdotal evidence
rather than high- quality experimental studies. 4 However, the lack of
strong evidence may reflect the lack of high-quality studies, rather than
the ineffectiveness of the strategies.56
At the Wimmera Health Care Group, quality management is used and
processes are redesigned to improve systems using three main strategies,
as described below.

6.6.I Simplifying systems and tasks (to reduce complexity)

By reducing the number of processes in a system, the probability of
unintended and unpredictable outcomes is reduced. For example, if a
system has eight processes and the probability of any single process
failing is 10%, the probability of a system failure due to one or more
process failures is l-0.9 8 = 0.57. If the complexity of the system is
decreased to five processes of care, the probability of the system failing
due to one or more process failures drops to 1 -0.9 5 = 0.41.

6.6.2 Standardising procedures (to reduce unwarranted variation)

If procedures are consistently performed in the same way, the probability
of unintended negative events occurring will be reduced. In hospitals,
variations in the way medications are prescribed or different types of
equipment are used for similar purposes are common. These variations
usually add little to the quality of care provided, but increase the
probability of adverse events occurring. For example, there are several
equally effective protocols that can be used for commencing patients on
anticoagulation therapy. If more than one protocol is used commonly in a
hospital, it is possible that two protocols will be confused and an error
may result. If junior doctors and nursing staff are trained to use only one
anticoagulation protocol, this will reduce variation in care delivery and
thereby reduce the probability of associated adverse events.
Standardisation is important for achieving quality improvement, but
there is a lack of standardisation in providing evidence-based or guideline
adherent health care. Some doctors are resistant to standardisation and
label the approach "cookbook medicine". However, "they fail to
acknowledge that if you have a good cookbook and follow the instructions,
you can expect to achieve the intended outcome".57

'At the organisational

iM( fwnuw from
consktamgan innovation
to successful routinamg H
is generally a nonlinear
process characterized by
multiple shocks, setbacks
and unanticipated events.'

Greenhaigh. et al
2004. citing Van de
Ven AH. et al.
The innovation journey

148
 6 Taking action to
close gape in care
and prevent
adverse events
recurring

In a study of vulnerable patients aged over 65 years old and living in

the community, those provided with higher-quality care in accordance
with 21 quality indicators (eg, pneumococcal vaccine, falls assessment,
weight measurement, cognitive screen) lived longer — 28% of patients who
received 44% of the recommended care died in the following 3 years,
compared with 18% of patients who received 62% of the recommended
care. The authors concluded "better performance on process quality
measures is strongly associated with better survival among community-
dwelling vulnerable older
Enhancing patient care

adults".38 Sometimes there can be a significant delay between the

provision of some processes of clinical care and the resultant positive
patient outcome. Such delay can limit the motivation of doctors to comply
with these processes of care. Although achieving standardisation in the
delivery of care requires substantial work, it should be a key component
of quality improvement programs.57

6.6.3 Using reminders end checklists (to decrease dependence on

memory)
Clinical trials provide information about an increasing number of
interventions that, if used, benefit the health of patients with panicular
clinical conditions. As a result, medical care is becoming increasingly
complicated. Frequently, the most beneficial medical care comprises a
number of investigations, medications and other treatments. Often,
medical staff are required tc^ recall detailed treatment plans for medical
conditions from memory, without other information sources being readily
available. Health professionals are often expected to function at an
unrealistically high cognitive state, requiring heavy reliance on accurate
recall from memory.59
If there are 10 important steps to be followed in the management of a
patient with a particular medical condition, is it reasonable to expect
clinicians to be able to recall perfectly all 10 steps from memory for each
of the next 100 patients that they treat with that condition? A simple way
to reduce this heavy reliance on human memory, with its inherent
unreliability, is to use comprehensive checklists and reminders. In their
most complex form, these checklists and reminders form a detailed
patient pathway that maps a complete treatment plan for patients with
particular medical conditions for each day that they are in hospital. Often,
a simple checklist requiring each treatment step to be initialled is
sufficient to ensure clinical staff are reminded of all the steps they need to
take when treating a patient in a particular clinical situation.
Checklists have been a key instrument in reducing errors in aviation,
aeronautics and product manufacturing. A checklist is "a list of action
items or criteria arranged in a systematic manner, allowing the user to
record the presence/absence of the individual items listed to ensure that
all are considered or completed" 60 "Typically, each item is checked off as it
is completed, verified, identified or answered, by placing a mark in a
designated space" 61 Checklists aim to reduce error and increase
adherence to best practice. They improve the recall of items and
standardise processes. Checklists may be paper-based or electronic. In
health care, it can be difficult to standardise all care processes because of
patient variation. Resistance to checklists can arise because they could be
perceived as an admission of weakness in not being able to recall of the
treatment processes required, or limiting clinical judgement and the
autonomy of decision making 60
Expert panels can design effective checklists in a particular clinical area
using prepublished guidelines of best practice in that area and conducting
"... one ct ihe moat
important issues in
dosing the loop on safety
initiatives * taking a
scientific approach to
assessment of t
behavioural change
strategies are indeed
being applied, and
whether they are
achieving the desired
patient outcome This
feedback is only possible
with rigorous
surveillance, audits, and
other observational
studies, followed by
reinforcing strategies
when needed.'

Cook DJ. et al
Lancet 2004:
363 1224-1230
 6 Taking action to
close gape in care
and prevent
adverse events
recurring

small-scale tests of draft checklists until appropriate checklists arc

developed
Checklists should not be developed for all decisions required to be made by clinicians, as there is a risk
of 'checklist fatigue* developing.61
Ultimately, checklists can only act as guides and prompts for clinicians to remind them of all the
steps required to treat patients with particular conditions. Checklists cannot provide advice that is
totally appropriate for every patient with a specific clinical condition. The variation in care given to
patients while clinicians are using guidelines can be as high as 25%, due to important differences
between individual patients. For example, patients with the same primary condition, which would
benefit from treatment given in accordance with the guidelines, may have different comorbidities that
could affect which treatment would provide them with the most benefit.
When using clinical guidelines, it is appropriate to provide treatment that deviates from sections of
the pathway guidelines when recommendations are not appropriate for an individual patient.
Deviations from guidelines, and the reasons why the deviations were necessary, should be documented
in the medical record. Analysis of these deviations can be used measure compliance with processes and
to detect adverse events, which further improve the guidelines.
The effectiveness of using reminders and checklists is demonstrated in a study of 100 intensive care
units in Michigan in the US. The study aimed, among other things, to reduce the number of patients
with catheter-related bloodstream infections following insertion of central lines. Five evidence-based
behaviour-specific interventions were introduced: wash your hands, use full barrier precautions,
prepare the insenion site with chlorhexidine antiseptic, avoid the femoral site for insertion, and remove
unnecessary lines. A checklist of these actions was created for doctors. The nurses assisting them were
empowered to perform an independent check and ensure that the doctor followed each for the actions
under non-emergency conditions. The proportion of intensive care units observed with no catheter-
related bloodstream infections increased from 59% at baseline to 80% 7-9 months after implementation
of the strategy.62

6.6.4 Introducing constraints (to make performing en error more difficult)

Constraints are incorporated into systems to make it more difficult to take an action that will increase
the probability of an adverse event. For example, the design of an item of equipment can be a
constraint. Two items of equipment, if inadvertently joined, may result in an adverse patient event. If
the two pieces of equipment cannot be physically joined because of the way they are designed, the
adverse event cannot occur. A junior resident being unable to directly discharge a patient from an
emergency department without the patient being reviewed by a senior doctor is another example of a
constraint.

6.6.5 Providing timely delivery of adequate and accurate information (so that decisions can be
made with appropriate data)
Designing a system appropriately should result in the right clinical information being available in the
right place, to the right person, and at the right lime. Examples include a patient's biochemistry and
haematology results being available in the emergency departmeni immediately after their blood samples
have been analysed in the laboratory; a copy of the patient's discharge summary for their previous
admission to the hospital being available to the resident medical staff when the patient re-presents to
the hospital's emergency department; and a typed list of the patient's discharge medication following a
recent hospital admission being available lo the patient's general practitioner when the patient presents
to them for follow-up.
Other system changes include:
■ building buffers into systems so that if errors occur, their effects are absorbed before causing harm to
patients
■ designing processes so that taking the correct action is ihe easiest action to take
 6 Taking action to close gape in care and
prevent adverse events recurring

■ designing processes that are consistent with human abilities to manage stress, time pressures and
workloads (given the limits of human memory and the limitations on human performance due to
circadian rhythm)
■ carefully automating systems so that staff know in which situations to over-ride the automation.

6.7 Actions commonly taken to prevent adverse events that are

generally ineffective
It is important to comment on those strategies aimed at changing clinical behaviour that are
ineffective, because in our experience they are still commonly used in health care. Firstly, medicine
has traditionally used a punitive model to attempt to change clinical behaviour. Individuals are
blamed for adverse events that occur while care is being provided. There is an expectation that
health fare professionals should deliver care perfectly. If an adverse event occurs, it is regarded as
the fault of the health care professionals providing care to the patient. The clinicians involved are
frequently punished, told to try harder and to be more careful.
Although punishment has not been shown to prevent adverse events and may be
counterproductive, it is still commonly used. Some weaknesses in complex systems may not be
evident until a serious adverse event occurs, several minor events or near misses may have
previously occurred. Staff are less likely to report these events if they fear punitive action being
taken. The system may therefore remain weak until a serious adverse event occurs.
Secondly, attempts are frequently made to change clinical behaviour by creating or changing
clinical policies and issuing memoranda. After an adverse event, a clinical policy may be issued as a
memorandum to prevent the event recurring. However, with constantly changing clinical staff in
many hospitals it is difficult to ensure that a policy detailed as a memorandum is followed. In our
experience, prompts, reminders and extensive training are also required.
■ Implementing and sustaining change is the most difficult part of quality improvement and risk management
■ There are many determinants (both facilitating and impeding) of the success of implementing change. They
in>«tve the change itself, the target group, the organisation, patients and the sociopolitical context. The
effectiveness of implementation strategies can be sensitive to differences in the characteristics of each of these
groups.
■ A single unifying theory of changing human behaviour has not been developed. There are many theories (often
overlapping) from a variety of disciplines that can be used to understand various aspects of human behaviour.
They provide a range of potential strategies for implementing change. There is, however, little scientific
evidence to support many of these theories.
■ Due to the complexity of health care delivery systems and the large number of determinants of implementing a
change, a comprehensive and systematic approach is required to identify major determinants, their effect on
the change and their changeability. Strategies developed from theories of behavioural change often need to be
selected and implemented to address the major determinants.
■ Many strategies used by health services to implement quality improvements and reduce risk are not
evidence-based. There is limited evidence about what is effective in implementing change to improve
quality. Most quality improvement strategies have a modest effect (10% in absolute terms), with a wide
• effect range. Single quality improvement implementation strategies may be as effective as multiple strategies.
■ Although not supported by high-quality evidence, our experience (which may be specific to the Wimmera Health
Care Group) of using a combination of quality management and system redesign to: simplify systems and
tasks: standardise procedures: use reminders and checklists: and introduce constraints and provide timely
delivery of adequate, accurate information has been effective in implementing and sustaining change to
improve quality and safety.
References
1 Greenhalgh T, Robert G. MacFarlane F, et al, Diffusion of innovations in service organisations systematic renew
and recommendations. Milbank O 2004; 82:581-629.
2 Pronovost PJ, Thompson DA, Holzmueller CG, et al. Defining and measuring patient safety. Crit Care CUn
2005 21: 1-19.
3 Auerbach AD, Landefeld CS. Shojania KG. The tension between needing to Improve care and knowing how to
do it. N Engl J Med 2007; 357: 608-613.
4 Shojania KG, Grimshaw JM. Evidence-based quality improvement: the state of the science. Health Aff
(MUtwood) 2005; 24: 138-150.
5 Scott IA, Poole PJ, Jayathissa S. Improving quality and safety of hospital care: a reappraisal and an agenda
tor clinically relevant reform. Intern Med J 2007; 38:44-55.
6 Scott I. The evolving science of translating research evidence into clinical practice. ACP J Chtb 2007: 146: A8-
A11.
7 Berwick DM. The science of improvement. JAMA 2008; 299: 1182-1184. •
8 Pawson R, Tilley N. Realistic evaluation. London: Sage, 1997.
9 Michie S, Johnston M, Abraham C, et al. Making psychological theory useful for implementing evidence based
practice: a consensus approach. Oual Sal Health Care 2005; 14:26-33.
10 Grol RP, Bosch MC, Hulscher ME, Eccles MP. Planning and studying improvement in patient care: the use of
theoretical perspectives. Milbank O 2007; 85:93-138.
11 Fleuren M, Wiefferink K, Paulussen T. Determinants of innovation within health care organizations. Literature
review and Delphi study. Int J Qual Health Care 2004; 16: 107-123.
12 Cabana MD, Rand CS, Powe NR, et al. Why don't physicians follow clinical practice guidelines? A framework
for improvement. JAMA 1999; 282: 1458-1465.
13 Bradley EH, Holmbee ES, Mattera JA. 'et al. A qualitative study of increasing beta-blocker use after
myocardial infarction: why do some hospitals succeed? JAMA 2001:285:2604-2611.
14 Bradley EH. Herrin J. Mattera JA. Holmbee ES. Quality improvement efforts and hospital performance. Rates
of beta-blocker prescription after acute myocardial infarction. Med Care 2005; 43: 282-292.
15 Sales A. Smith J. Curran G. Kochevar L. Models, strategies and tools. Theory in implementing evidence-
based findings into health care practice. J Gen Intern Med 2006; 21: S43-S49.
16 Berwick DM. Disseminating innovations in health care. JAMA 2003; 289:1969-1975.
17 Grol R. Wensing M. Eccles M. editors. Improving patient care: the implementation of change in clinical
practice. Oxford: Elsevier. 2005.
18 Grol R. Beliefs and evidence in changing clinical practice. BMJ 1997; 315: 418-421.
19 Moulding NT. Silagy CA. Weller DP. A framework for effective management of change in clinical practioe:
dissemination and implementation of clinical practice guidelines. Oual Sal Health Care 1999; 8: 177-183
20 Prochaska JO. DiClemente CC. Norcross JC. In search of how people change: applications to addictive
behaviours. Am Psychol 1992; 47: 1102-1114.
21 Grant A. Greene J. It's your life. What are you going to do with it? 2nd ed. Harlow: Pearson Education. 2004
22 van Bokhoven MA. Kok G. van der Weijden T. Designing a quality improvement intervention: a systematic
approach. Qual Sat Health Care 2003; 12: 215-220.
23 Grol R. Implementing guidelines in general practice care Qual Health Care 1992: 1: 184-191.
24 Lewis AP. Bolden KJ. General practitioners and their learning styles. J R Coll Gen Pract 1989: 39: 187-189
25 Ajzen I. The theory of planned behaviour. Organ Behav Hum 1991: 50: 179-211
26 Robbins S. Bergman R. Stagg I. Coulter M Management. 4th ed Sydney: Pearson Education. 2006
27 Alpert M. The care and feeding of engineers Fortune 1992; Sep 21: 86-95
28 Rogers EM. Diffusion of innovations. 4th ed New York: Free Press. 1995
 6 Taking action to close gape in care and
prevent adverse events recurring

29 Collins BA. Hawks JW. Davis R. From theory to practice: identifying authentic opinion leaders to improve
care Managed Care 2000: 56-62.

1S1
Enhancing patient care

30 Ooumt G, Gatteilari M, Grimshaw J, O'Brien MA. Local opinion leaders: effects on professional practice and health care outcomes.
Cochrane Database Syst Rev 2007; (1): 1-28.
31 Brafthwafte J. Travaglia JF. An overview of clinical governance policies, practices and initiatives. Aust Health Rev 2008:32: 10-22.
32 Grant A. Greene J. Coach yourself O work. Get yourself a (working) life. Sydney: ABC Books, 2005.
33 Gabbay J, Ie May A. Evidence based guidelines or collectively constructed "mindlines?" Ethnographic study of knowledge
management in primary care. BMJ 2004; 329:1013-1016.
34 Weiner B, Shortell S, Alexander J. Promoting clinical involvement in hospital quality improvement efforts: the effects of top
management, board and physician leadership. Health Serv Rev 1997; 32:491-510.
35 Rodgers R, Hunter JE, Rogers DL. Influence of top management commitment on management program success. J Appl Psychol
1993; 78: 151-155.
36 Ovretveit J. The leader's role in quality and safety improvement; a review of research and guidance; the 'Improving Action
Evaluation Project." Fourth report. Stockholm: Association of County Councils (Lanstingsfor- bundet), 2004.
37 Damanpour F. Organizational innovation: a meta-analysis of effects of determinants and moderators. Acad Manage J1991; 34:
555-590.
38 Scott WR. Innovation in medical care organizations. Med Care Rev 1990; 47: 165-192.
39 AJmaraz J. Quality management and the process of change. J Organ Change Manag 1994; 7: 6-15.
40 lies V, Sutherland K. Organisational change: a review for health care managers, professionals and researchers. London: National
Co-ordinating Centre for NHS Service Delivery and Organisation R & D , 2001.
41 Blumenthal D, Kilo CM. A report card on continuous quality improvement. Milbank Q 1998; 76: 625-648.
42 Gustafson DH, Hundt AS. Findings of innovation research applied to quality management principles for health care. Health Care
Manage Rev 1995; 20:16-33.
43 Grol R. Improving the quality of medical care: building bridges among professional pride, payer profit, and patient satisfaction.
JAMA 2001: 20: 2578-2585.
44 DiBella AJ. Nevis EC, Gould JM. Understanding organizational learning capability. J Manage Stud 1996; 33:361- 379.
45 Nevis EC, DiBella AJ, Gould JM. Understanding organizations as learning systems. Sloan Manage Rev 1995:36: 73-85.
46 Lahteenmaki S, Toivonen J, Mattila M. Critical aspects of organizational learning research and proposals for its measurement. Br J
Manage 2001; 12: 113-129.
47 Scott T, Mannion R, Marshall M, Davies H. Does organisational culture influence health care performance? A review of the
evidence. J Health Serv Res Policy 2003; 8:105-117.
48 Grimshaw J, Eccles M. Thomas R, MacLennan G. Toward evidence-based quality improvement: evidence (and its limitations) of
the effectiveness of guideline dissemination and implementation strategies 1966-1998. J Gen Intern Med 2006; 21: S14-S20.
49 Eccles M, Grimshaw J, Walker A, Johnston M. Changing the behaviour of healthcare professionals: the use of theory in promoting
the uptake of research findings. J Clin Epidemiol 2005: 58: 107-112.
50 Schouten JA, Hulscher ME, Natsch S, et al. Barriers to optimal antibiotic use for community-acquired pneumonia at hospitals: a
qualitative study. Oual Saf Health Care 2007; 16:143-149.
51 Bosch M, van der Weijden T, Wensing M. Grol R. Tailoring quality improvement interventions to identified barriers: a multiple case
analysis. J Eval Clin Prac 2007; 13: 161-168.
52 Grimshaw JM. Shirran L, Thomas R, et al. Changing provider behaviour: an overview of systematic reviews ol interventions. Med
Care 2001; 39: ii2-ii45.
53 Oxman AD, Thomson MA, Davis DA, Hayes RB. No magic bullets: a systematic review of 102 trials of interventions to improve
professional practice. CMAJ1995; 153: 1423-1431.
54 Grimshaw JM, Thomas RE. MacLennan G, et al. Effectiveness and efficiency of guideline dissemination and implementation
strategies. Health Technol Assess 2004; 8: 1-84.
55 Shojania KG. Ranji SR, McDonald KM, et al. Effects of quality improvement strategies lor type 2 diabetes on glycemic control
JAMA 2006: 296: 427-440.
56 Ovretveit J. What are the best strategies for bensuring quality in hospitals? Copenhagen: World Health Organization Regional
Office for Europe's Health Evidence Network, 2003.
57 McGlynn EA. Intended and unintended consequences. What should we really worry about? Med Care 2007, 45 3-5.
58 Higashi T, Shekelle PG, Adams J, et al. Vulnerable older patients receiving higher quality care survive longer Ann Intern Med
2005; 148: 274-281.
59 Leape LL. Error in medicine. JAMA 1994; 272: 1351-1857.
60 Hales BM, Pronovost PJ. The checklist — a tool for error management and performance improvement. J Crit Can
2006; 21:231-235.
61 Hales B, Terblanche M, Fowler R, Sibbald W. Development of medical checklists for improved quality of patient care. Int J Qual
Health Care 2007; 1-9.
62 Pronovost PJ. Berenholtz SM, Goeschel CA, et al. Creating high reliability in health care organizations. Health Serv fles2006; 41:
1599-1617. □
Enhancing patient care
 6 Taking action to close gape in care and prevent adverse
events recurring

7 A practical framework to close gaps in patient care and reduce

risks to patients
This chapter will:
m Present a practical, step-by-step framework to close gaps and reduce risks
■ Discuss why it can be difficult to motivate doctors to participate in quality and safety programs
■ Present strategies to help engage doctors in quality and safety programs
■ Describe how the framework was used to develop clinical pathways and improve the management of patents with stroke
■ Present a checklist to close gaps and reduce risks.
«

It is easy to understand why improving the quality and safely of health care delivery is so difficult — health care delivery
systems are complex and comprise many parts that often interact unpredictably. A single unifying theory of change in
human behaviour on which to base change strategies is not available. Instead, many theories of human behaviour, often
with substantial overlap, have been developed in an attempt to explain and predict behaviour (described in detail in
Chapter 6). These theories have resulted in the development and use of many strategies (often of modest or unknown
effectiveness) to address the multiple determinants of implementing change, which are sensitive to characteristics of the
intervention, target group, patients and organisation.
How, then, should individual health services try to close gaps and reduce risks? Several broad principles are useful:
■ A comprehensive, systematic approach is required to identify the major facilitating and impeding determinants of
success of the intervention, the target group and the organisation.
■ Strategies should be developed to address the imponant and modifiable determinants.
■ Once action has been taken to implement these strategies, very close monitoring of its effects is required. Timely
feedback of results should be given to the target group and changes in strategies and objectives made as required.

7.1 A practical framework

We have developed a detailed framework for successfully implementing change in health services by combining the
available evidence 1"' with our experience in quality improvement and patient safety over 19 years. The framework is a
series of practical steps to help health serv ices achieve change to closc gaps and reduce risks. A series of questions and
actions function as signposts for designing and implementing a comprehensive, well planned change program, and
motivating clinicians to change their behaviour and practice.
Not all the steps in the framework need to be addressed for even- change in practicc that is contemplated. Many
changes required to close gaps and reducc risks arc simple, and appropnate action can be quickly taken. Some changes
arc small and will not require teams to design, implement and evaluate them in practice. But, however small the proposed
change, each step in the framework should be considered by the health service's clinical risk manager, who should make
an active decision about whether an action needs to be taken, rather than being passively omitted
In essence, change involves identifying gaps or risks (usually as deviations from a standard or best practice"*,
systematically analysing the situation, determining and taking appropriate corrective action to close the gap or reduce the
risk (by providing service that meets the predetermined standard), and closely monitoring the results of the actions taken.
These steps can be expanded as follows:
■ Identify a gap in care or risk to patients and determine its causes.
■ Identify an appropriate intervention, and ensure it is supported by scientific evidence.
■ Identify the target group.
■ Establish whether key clinical groups regard the intervention as important — ask opinion leaders.
■ Undertake a risk-benefit analysis of implementing the intervention.
■ Determine whether the target group supports the intervention.
■ Determine whether the intervention has management support. Is there an executive champion? Will adequate
resources be made available?
■ Undertake a systematic determinant analysis at all levels — identify the major facilitators and impediments.
■ Design a detailed implementation and maintenance strategy and plan.
■ Test the intervention on a small scale.
■ Implement the intervention systematically on a larger scale.
■ Closely monitor the results of the intervention and modify the plan accordingly.
■ Communicate the results widely.
■ Evaluate the program.
The practical aspects of each step will be discussed below.

7.1.1 Step I: identify a gap in the quality of patient care or detect an adverse event with significant risk to patients
Clinical governance comprises two parts, quality and safety. The quality improvement program will detect gaps in the care
being provided and best practice as supported by scientific evidence. The risk management program will delect adverse
events and risks. To close a gap or reduce a risk, a change may be required in how care is delivered. More gaps and risk
will be detected than can be dealt with at one time and they should be prioritised for action, as discussed in Chapter 5.

7.1.2 Step 2: analyse the gap or risk to determine its causes and whether any change is required in how health
care is delivered in the area concerned
The gap or risk should be analysed to determine its causes and whether any change is required in how care is delivered.
The analysis need not be exhaustive, but must determine whether change is required and if so, identify the main
changes. A comprehensive analysis will be undertaken later.

7.1.3 Step 3: determine the quality of the evidence supporting the benefits of the required change
Clinicians require convincing evidence of the effectiveness of a proposed change for them to modify the way they care for
their patients.4 High-quality evidence of the benefits to patients should come from reputable sources, including Cochrane
reviews, professional colleges, or guidelines from national expert bodies Detailed evaluations of many quality and safety
initiatives have been done by these groups and are readily available.

7.1.4 Step 4: identify the target group, major stakeholders, opinion leaders and clinical champions
The target group (staff members who need to change their behaviour to successfully implement the required change) and
its subgroups should be identified so that change strategies can be tailored to meet
 6 Taking action to close gape in care and prevent adverse
events recurring

their needs. The main clinical groups in acute hospitals are medical, nursing and allied health staff. However, there
are important subgroups within each group. For example, the medical group may comprise junior medical staff,
registrars, specialists and general practitioners. The needs of each subgroup will be different, and strategies to change
their clinical practice should be developed to account for these differences.
Three other key groups should be identified:
■ Stakeholders — individuals or groups affected by the change, and whose cooperation is needed to implement it.
■ Opinion leaders — respected and influential clinicians whose opinion target group members will seek about the
proposed change.
a Clinical champions — early adopters of the change, passionate about making the change, will encourage other
clinicians to make the change , and support the change at clinical meetings. Some authors suggest patients and their
relatives should also be involved in considering changes in how their care will be provided. 5 Clearly, patients should be
active participants in decisions about their care. Many health services now have consumer representatives on their
quality improvement committees, but not many involve them in designing new systems of health care delivery or
changing existing systems.

7.1.5 Step 5: determine whether the gap or risk Is important for the target group members and whether they
acknowledge ownership of the issue
To improve care, health services need to motivate health professionals to change their behaviour. Professionals are
likely to be motivated to work on problems they consider important. The impact of the proposed change should be
identified "in ways that resonate with decision makers and frontline providers". 6 Health care professionals are likely to
support changes that directly and positively affect their patients' outcomes. By contrast, clinicians may consider
compliance with paperwork unimportant. Therefore, they are more likely to be motivated to change their prescribing of
aspirin, P-blockers and ACE (angiotensin-converting enzyme) inhibitors for patients with acute myocardial infarction,
which will increase their survival rates. They are less likely to be motivated to ensure patients' medical records are
complete so that the coding, which determines the level of funding the health service receives to provide clinical care,
is accurately completed.
In a postal survey in the United Slates, a third of physicians reported thai although they had experienced errors
when their family members had received care, they believed errors occurred
infrequently.7 These doctors did not see medical errors as one of the important Success is likely to
problems in health carc, and did not share the urgency of some national depend as much on the
quality of implementation,
organisations about the issue. They also believed that a substantial number
on the sensitivity to
of deaths associated with medical errors were not preventable, and only two different points of view
actions would be effective in reducing errors — requiring hospitals to develop and the degree & support
systems to reduce errors and increasing the number of nurses.' Therefore, in from influential
individual health services, some doctors will not see error prevention as organisation members as
important The the soundness of the
change approach
adopted... Much of the
evidence demonstrates
that top management
involvement is critical to
success but we must
[also] remember the
importance 6f opinion-
formers within the
professions.'

lies V. Sutherland K
Organisational
change: a review for
heath cam managers,
prdassunala and
researchers
 Enhancing patient care

authors conclude, "perhaps the most critical issue will be to provide sceptical physicians with
scientific pnxrf that proposed strategies will, in fact, reduce preventable medical errors and the harm
they cause".7
At any one time in health services, there are many more behaviours that can be changed with a
positive effect on patient outcomes than there are resources and organisational energy to successfully
make these changes. As behavioural change is more likely to be successful if clinicians feel the
proposed behaviours are important, health services may best concentrate on these issues.
A determination should be made about who "owns" the problems associated with the gaps and
risks being reviewed. It can be difficult unsuccessfully analyse and solve system problems that are
viewed by clinicians as belonging to the health service's administration. Often, clinicians do not
accept ownership of any ponion of the problems associated with particular risks, even though they
are major stakeholders in the delivery system under review.
It will be difficult for health services to change clinician behaviour if the relevant clinicians believe
that the proposed change is clinically unimportant, and that their behaviour is not responsible for
any part of the gjip or risk. These clinicians are in the precontemplation and contemplation stage of
the change cycle. Rather than forcing change on them, health services should continue to raise
awareness and understanding about the issue.

7.1.6 Step 6: describe the present state of health cere delivery regarding the gap or risk to the
target group
The target group should be provided with an accurate, well documented assessment of the present
state of care delivery in the area under review. Although major studies have shown large gaps in care
provided to patients and significant rates of adverse events, individual clinicians at the coalface are
often unaware of these problems. From their personal experience, they believe harm occurs
infrequently to their patients.8 The assessment should make the gaps and risks visible, and raise
clinicians' awareness. The data presented may include details of admissions with poor patient
outcomes, clinical audits showing high complication rates, variation in the care provided, or
unfavourable benchmarking against similar health services. Presentation of the absolute numbers of
adverse events or poor patient outcomes in the health service or the community may have more
impact than percentages or rates, as may comparative data from similar health services. 9
The data presented should be concise, clear, and of a high standard. Diagrams and graphs in
which trends can be easily seen are useful. Clinicians often have limited time for reading proposals,
and do not appreciate reading inaccurate information and poorly presented documents. It is difficult
to motivate clinicians to change their behaviour if the data outlining the problem is inaccurate or
flawed. A supportive clinician should review the validity and accuracy of the data to be presented to
the target group.

7.1.7 Step 7: determine whether a number of clinicians are dissatisfied with the
present situation
Dissatisfaction with the present situation is an excellent starting point to create motivation for
change.12 Therefore, it is useful to assess whether clinicians are dissatisfied with how carc is
currently provided in the area under review.

7.1.8 Step 8: present a clear, realistic vision for the future to the target group
After the present situation has been presented, a clear, realistic vision of what can be achieved by
making appropriate changes should be presented to the target group. The vision should be a sharp
contrast to the present situation. The greater the difference between the present situation with which
clinicians are unhappy and the vision, the greater the motivation to change. 12 Clinicians also need to
feel the vision provides something worthwhile for their patients.

158
 7 A practical framework to close gaps in patient care and reduoe risks to ps*ients

The vision should be accurate, aiid presented concisely and clearly. With the volume of research
published each year, it is impossible for doctors to read all research relevant to their clinical practice
and develop their own evidence-based guides for delivering care. 4 If there are significant gaps in
patient cane in their area of practice in the hospital, many doctors may appreciate being informed of
potentially significant improvements that can be made to the care they provide.
The way this information is presented is important. Many doctors will complain about increasing
demands being placed on them and their time. They'will, therefore, appreciate concise, easy-to-
understand and high-quality presentations in areas of significant potential benefit to their patients.
As well as concise summaries of the evidence and slide presentations of the literature, the complete
literature review and the publications from which the review was derived should be available for
clinicians who wish to read them.10
The vision presented to the target group must be realistic. Otherwise, it will be difficult to gain the
support of relevant clinicians. The vision should be based on valid information from reliable, well
regarded sources. If the change has been achieved at a comparable health service, details about their
program should be presented. A respected clinician from that health service could be invited to speak
to the target group about the health service's experiences. A respected expert in the clinical area, from
outside the hospital if necessary, could be asked to present an up-to-date review of research in the
area of concern.

7.1.9 Step 9: weigh up the costs and benefits of the change being considered
After a presentation has been made, many clinicians will assess its validity and how realistic the
proposed changes are for the health service. They will also evaluate how the proposed change fits into
their existing practice. Medical staff may be influenced by their personal clinical experience when
assessing proposed changes to clinical practice. For example, individual doctors may not have had a
patient under their care who did not receive appropriate thromboembolism prophylaxis and went on
to develop a pulmonary embolus. However, they may have had several patients who were given
appropriate prophylaxis and developed wound haematomas or more serious bleeding. If clinicians
have recently had patients develop severe negative outcomes when using the proposed change in their
practice, these experiences may weigh heavily in their assessment of the change.
If the benefits of making the change are greater than the personal cost of making the change and
clinicians see the change as feasible, it will be attractive to them and they will be more likely to
change their practice. It is therefore imponant to make the costs of the change to the target group as
small as possible. Wherever possible, the proposed change should not impose an additional workload
on clinicians and especially not increase demands on their lime. Removing or reducing barriers to the
change, creating a more efficient delivery system and making it as easy as possible for clinicians to
undertake the proposed change, can reduce the costs of changing behaviour. A clear, concise written
list of the benefits and costs of making the proposed changes may help clinicians determine whether
they wish to be part of the project.
In obtaining clinician support, the proposed changes should be a genuine attempt by the health
service to improve the quality of care being provided, and not primarily a cost-cutting exercise by
management (although some quality improvement activities can reduce the use of inappropriate care
and thereby simultaneously improve the quality of care and reduce costs for the health service) 4 Cost
containment or reduction and quality improvement can be compatible with health care, 11 but it can
be difficult to convince clinicians of this relationship, as "previous efforts to improve quality often
were perceived as cost-cutting efforts that were disguised as quality",' with the main aim or reducing
costs and increasing profits. 12
"In addition to deciding whether and when evidence is sufficient in quality and quantity to be used
in practice, issues such as minimal harm, feasibility, and low cost might shift clinicians over the
Enhancing patient care

trratmcnt threshold.* 11 Importantly, the cost-benefit ratio will change with time. Clinicians will
continually assess the benefits and costs to them as they change their practice to determine whether
the benefits continue to outweigh the costs, continuing to make the change worthwhile. Ideally, once
the changes are in place, the benefits will increase without an increase in costs, and clinicians will
continue to suppon the change.

7.1.10 Step 10: determine whether the hospital's senior management will support the propon
d change
Consistent suppon for the proposed change from senior management stafT is critical. They should
allocate an appropriate budget, appoint a project officer to coordinate the change if the project is
large enough to require such a role, provide secretarial suppon, and, if warranted, approve the
provision of extra stall to cover the normal duties oT key people in the target group while they are
participating in the project. The proposed changes may also require structural changes, including job
redesign, new equipment and additional training. These changes require additional resources in the
shon term beTore any medium- to long-term savings may be made, funher stressing already-
stretched health service budgets. In shon, "improving safety costs money in the short term", 8 and for
some changes, the financial costs may always exceed the benefits.
Senior management should also nominate an executive champion or sponsor to directly support
the change project. The executive sponsor should be an active, hands-on member of the change
project team, and should guide the project through the hospital's bureaucracy such as obtaining
approval for new or changed hospital policies. The executive champion can also obtain resources for
the project and remove some barriers. 10 The chief executive and other senior management staff often
have limited knowledge about key clinical issues. Having an executive champion on the project team
will keep them informed about important issues that arise and the progress the team is making. 14
An ideal executive sponsor for a significant clinical change project is the director of medical
services. Directors of medical services are medically trained, can appreciate the clinical aspects of the
proposed change, and have detailed knowledge of the hospital's bureaucratic and administrative
structures and processes. Directors of medical services also have an appropriate level of authority to
rapidly and effectively solve problems, especially those involving several clinical areas or
departments. They can also ensure funding and support for the project from senior management.
Frequently, quality improvement projects are given funding for a limited time. After a change in
clinical behaviour has been achieved, the funding ceases and the health service is expected to
continue to maintain the improvements made, which often requires ongoing funding. The director of
medical services, having been an active participant in the project, is well positioned in the health
service to help secure ongoing funding for the project.
If the change project is large, the chief executive should formally approve the support that senior
management and the health service will provide. Such support signals to the target group and other
hospital staff that senior management has a genuine interest in quality and safety in the hospital,
regards the project as important, and will provide the necessary resources and expertise for the
change in order to be successfully implemented.

7.1.11 Step 11: assess the stage of change of individual clinicians in the target group
A critical issue is the stage in the change cycle that individual clinicians in the target group are at
when trying to initiate change. After the presentations of the current state and future vision, some
clinicians will be ready to change, while others will be contemplating the change and not ready to
take action. Clinicians at different stages of the change cycle require different approaches to help
them move to the next stage Importantly, clinicians who are not ready to take action should not be
Enhancing patient care

forced or coerced to change, as this is unlikely to achieve sustained change. Not all clinicians need to
be ready for change in order for the change process to commence
 7 A practical framework to close gaps in patient care and
reduce riak* to ptfiantt

The health service should begin working with clinicians in the target group who are early adopters
of change and are in the preparation and action stages of the change cycle. An assessment should
be made as to whether there is a commitment to change from a cross-section of the target group and
whether there are enough clinicians to form a project team to design and implement the change.
Working with clinicians in the preparation and action stages of the cycle may result in the change
initially being implemented in one clinical area, rather than throughout the service. However, making
the change in one area allows it to be trialled on a small scale and improvements to be made and
used later on a larger scale.
Clinicians in the precontemplation and contemplation stages of the change cycle may need
individual approaches to help them progress to the preparation and action stages. They may benefit
from receiving more information to increase their understanding of the proposed change. Even
though they are not ready to make the change, they are now aware of the issue and have high-
quality information with which to make their assessment. Also, after they see the changes made and
the success achieved by the "early adopters", they may progress to the preparation and action stages.
Eventually, a critical point will be reached at which many of the clinicians in the target group have
made the required changes to their practice, acceptance of the change has substantial momentum,
and hopefully most of the remaining clinicians will follow their colleagues and make the change.

7.1.12 Step 12: form a project team

When extensive gaps in care are identified or adverse events associated with a significant level of risk
are detected and complex delivery systems are involved, a multidisciplinary team should be formed to
close the gap and reduce the risk. The team should not be regarded as another hospital committee,
but as a problem-solving group with a specific task and a short-to-medium-term time horizon. The
selection of appropriate team members is critical as "the effort, skill and even personalities of key
personnel can substantially affect the success or failure of a given quality improvement
intervention".15

TAMA Project team composition

The team should comprise:
■ clinicians from the target group who are innovators or early adopters of the proposed change
■ if the change project is large, a project officer to coordinate and organise the project (including
undertaking literature searches and sourcing information about what occurs at other hospitals
regarding the proposed change)
■ an executive sponsor and staff with technical expertise in areas relevant to the change (eg, clinical
specialists, pharmacists, health information managers and information technology staff)
■ secretarial assistance to keep minutes, prepare agendas and policy documents, and design
implementation tools.
Membership should also include major stakeholders from areas of the hospital affected by the
delivery systems to be reviewed. It is essential that the team includes clinicians with a strong
motivation to solve the problems at hand and clinicians who are regarded as clinical leaders by their
peers. "The selection ... of clinical champions is a fundamental step for the success of the strategy." 4
As previously discussed, senior management should be represented with sufficient authority to
provide the resources to analyse the gaps and adverse events, review the systems involved and guide
the team's proposed system changes through the hospital's bureaucracy. A group facilitator with
problem- solving and risk-management skills should be considered to assist the team. A health
information manager should be pan of the team, as such managers have skills in clinical data
extraction and measurement, which are important for analysing adverse events and monitoring the
cffects of actions

161
Enhancing patient care

Hospitals should also consider patient or consumer representation on the

team, and should be able to co-opt other staff as required.
When deciding if the composition of the team is appropriate, an assess-
ment should be made whether the team has the technical expertise
required to solve the problems before it and a sufficiently diverse
membership to promote ownership of the changes in the relevant areas of
the organisation and assist the implementation and maintenance of the
changes.
Where possible, members of the team who are not employees of the
hospital should be paid for the time they spend at team meetings and
preparing information for the team. Additional resources to provide
payment and cover for the usual duties of project team members should
be made available for hospital employees if, as is frequently the case,
significant lime is required to close the gaps and reduce the risks that
have been identified. To burden hospital staff who already have heavy
workloads with considerable extra duties will reduce the effectiveness of
the team.

7.1.12.2 Role of the project teom

The team's role is to analyse the gaps identified or the risks detected and
determine and take appropriate action to close the gaps or reduce the
probability of the risk. This role requires the development of goals and a
detailed implementation plan. The team should also have responsibility
for implementing the plan. The team requires terms of reference detailing
its objectives, membership, responsibilities and reporting requirements.
The team should appoint a leader, usually a clinical champion who is
passionate about quality and safety generally or the proposed change, or
an opinion leader who is an influential, respected clinician in the hospital.
Appropriate responsibility and autonomy should be given 10 ihe team to
allow it to undertake an analysis of ihe referred gap in care or risk and to
take appropriate action. The team members will have diverse needs and
will want to adapt the change to their work setting, which should be
encouraged. Team members should be brought together early in the
change process 10 support each other, to allow local ownership of the
project and the tools 10 developed, and to share the project's goals. This
involvement also ensures that the representatives of stakeholder
clinicians have an interest in the project's goal and are therefore more
likely to accept it.
Involvement in planning and implementing the change gives clinicians
increased control over their work environment. Control can be a powerful
motivator for clinicians to take appropriate action to achieve the goals of
the group. Importantly, goals imposed by senior management onto
clinicians without their input are likely to decrease motivation, even if the
goals are appropriate. A collective decision that the change is worthwhile
is required, and the team seeking to make the changes should share this
vision.

7.1.12.3 Advantages of teams

 7 A practical framework to close gaps in patient care and reduoe
risks to ps*ients

Using teams to achieve goals in organisations has several advantages. 16

The creation of teams to plan and implement change allows staff from a
V. organ'salio*iu< diverse range of clinical and other backgrounds in the hospital to come
change is a process that
together.
can be fac&tated by
perceptive and insightful
planning and analysis and
well crafted, sensitive
implementation phases,
while acknowledging that
it can be never fully
isolated from the elfects of
serendipidity, uncertainty
and chance.'

lies V. Sutherland K
Organisational change:
a review for health care
managers,
professionals and
researchers
Diversity of membership brings staff with a wide range of skills and experience to the discussion,
allowing different perspectives to be presented on how goals can be achieved, and stimulating the
development of innovative ideas.
A significant proponion of the team should comprise clinicians who directly provide patient care in
the area where change is proposed. These clinicians are part of the target group and will know the
setting for the change well. They will bring imponant knowledge to the discussion about what occurs
at the coalface when providing care. Such informed discussion'will often allow anticipation of
problems that may occur in implementing the change. One or more of these clinicians should be the
clinical champions of the project, and should speak in support of it at clinical meetings. The team
approach also promotes cooperation between clinicians and other staff from the various disciplines
involved in the change process and allows better decisions to be made.

7.1.13 Step 13: analyse the gap or risk and Identify barriers and facilitators of the proposed
change or intervention
The initial task of the project team is to:
■ Analyse the gap identified or risk detected.
■ Identify the determinants (or factors) of the proposed change or intervention, target group, patients,
organisation and sociopolitical context and their facilitators and barriers.
■ Determine the resources available in the health service.

7.1.13.1 Analysis of the gap in care or risk

The project team reviews the referred gap or risk, determines the service delivery systems involved,
and identifies the latent factors (eg, equipment maintenance, work design, workload, skills,
supervision, lengths of shifts and allocation of resources) and active errors (eg, lapses, mistakes and
rule violations) lhat contributed to the gap or risk occurring.
The team should draw a (low diagram of the existing delivery system and identify key processes in the
system. The team should determine where ihe gaps and latent factors or errors leading to adverse
events occur in the system. 10 The team should then determine if any further information is required.
On some occasions, insufficient information will be available from analysing a gap or single adverse
event for the team to accurately identify the main issues. They may then request that a focused audit
be undertaken lo assess the frequency and severity of the gap or adverse event and to identify the
contributory factors. The analysis may also involve reviewing relevant case notes and scientific
literature, and conducting interviews, surveys or focus groups with staff and patients. Analysis of
adverse events is discussed in detail in Chapter 5.

7.1.13.2 Analysis of the determinants of the proposed change or intervention, target group, patients,
organisation, sociopolitical context and their facilitators and barriers
The determinants of the proposed change or intervention should be identified — this is similar to a
diagnostic process or needs assessment in clinical medicine. The influence of the context (with its
barriers and facilitators) in which a change is being proposed is being recognised as increasingly
important to the successful implementation of change. 1' Factors to consider include
■ Is the proposed change simple or complex?
■ Does it comprise a single component or multiple components?
■ How do these components interact?
■ Is the change dependent on the skill or effort of key people?
■ Is the change supported by high-quality scientific evidence? 18
 Enhancing patient care

The complexity of the change is also affected by the number of processes or subprqcesses in the
relevant delivery system, and the breadth of the process or the number of choices presented at
decision points in the system.17
The components of the change or intervention may be quite different. For example, the proposed
change may be that a clinical practice guideline be used in the health service. Such guidelines
usually comprise a number of quite different recommendations. Individual recommendations may
relate to different aspects of patient care, such as monitoring, investigations or medication. Each
recommendation may be quite different in nature, require adoption by different clinical groups, and
present different barriers to implementation, therefore requiring different implementation strategies.
The determinants of the target group and each of its subgroups should be identified. This analysis
requires:
■ Examination of individual factors of target group members, such as their knowledge, experience
and skills in the relevant clinical area, attitudes to change generally and the proposed change
specifically, working arrangements (eg, full-time or pan-time, salaried or fee-for-service) and
personality and their needs regarding the proposed change.
■ Identification of innovators in the target group who could influence the practice of early adopters
regarding the proposed change.
■ Examination of the social context in which the gap in care or adverse event has occurred,
including the culture of the hospital, the social network and the behaviour of other clinical
colleagues.
■ Consideration of patient factors, such as their knowledge, attitude, needs, socioeconomic status,
and severity of illness.
■ Identification of economic, administrative and organisational factors, including whether there are
adequate budgets, numbers of suitably qualified personnel, and equipment; relevant hospital
policies; the organisational structure; and how change is usually achieved in the organisation.
■ Identification of major stakeholders in the change.
The checklist of potential determinants that influence adoption and implementation of an innovation
presented in Chapter 6 should be used to establish the determinants in each major category
(innovation, target group, patients, organisation and sociopolitical context) for the change being
proposed. The determinants should be assessed for their effect (facilitating, impeding or neutral),
importance and changeability and ranked according to their perceived importance.
With small innovations or changes, it may be sufficient to ask project teams members to identify the
determinants and their effect, importance and changeability. However, it is often necessary to go
outside the project team and seek information about the determinants from a larger sample of the
target group, such as all junior doctors in the hospital. This information may be sought using in-
depth interviews, small group sessions, such as focus groups, or surveys. 19
To determine the perspectives of intensive care unit staff about the potential determinants and
consequences of semirecumbent position in patients receiving mechanical ventilation, interviews
were undertaking using the following structure and prompts: 13
■ Current practice
> How do you usually manage patients with this condition?
■ Current knowledge
> What is the optimal management of patients with this condition?
> What are the risks of not managing the patient optimally?
■ Considering research in this area
> How should care be provided for patients with this condition?
■ Barriers to using the intervention
> How do you decide whether to use this intervention?

164
 7 A practical framework to close gaps in patient care and
reduce risks to paeans

> What makes it easy to use?

> What makes it difficult to use?
> Why would you or other clinicians not use this intervention?
> What are the most important factors in determining whether to use the
intervention? a Implementing the intervention
> What strategies should be used to implement the intervention?
> What can be done to make it easier for the intervention to be used?
> What can other health professionals do to make it easier to use the intervention?
> What can the hospital administration do to make it easier to use the intervention?
> Do you have any other suggestions?
As the interviews progressed, emerging themes were identified and modified in an iterative process
after presentation for validation or refutation to different professionals, focus groups and groups
representative of intensive care staff. The study found that knowledge of the intervention among staff
was variable. After being presented with the scientific evidence, staff endorsed acceptance of the
intervention. Barriers to implementation of the intervention were raised under the general headings
of: alternatives to the intervention, contraindications to the intervention, potential risk of harm,
safety, and lack of resources. Staff suggested multiple implementation strategies. 13
The authors believed the wide range of stafT interviewed and participating in the process led to a
broader view being obtained of the determinants of the intervention's implementation. They also felt
more trials supporting the intervention would have led to staff citing fewer barriers. Although not
undertaken in their study, the authors suggested that the perspective of patients and family members
regarding interventions should also be obtained and considered. 13 The study found that under-
utilisation of the intervention was due to "insufficient awareness of its potential benefits, teal and
perceived deterrents, lack of agreement about who is responsible for its implementation and lack of
enabling and reinforcing strategies". 13 These reasons are probably common to why many interventions
are not fully utilised.
All major determinants should be identified. Clinicians might not identify some high-level organisa -
tional determinants; this information may need to be objectively, widely and actively sought from
other stakeholders. 20 In a study of barriers to the use of key recommendations for management of
community- acquired pneumonia, a large variety of barriers were reported. Different barriers were
identified for each recommendation by different clinical groups and subgroups. Doctors did not
perceive the same barriers perceived by nurses, and junior doctors did not perceive the same barriers
perceived by senior doctors. 19 Eliciting opinions from only one group is likely to result in missing
significant barriers that could be identified by other groups. Each recommendation has its own
pattern of barriers that must be determined and addressed. Barriers may also vary widely between
health services, and each service should identify its barriers and develop implementation strategies to
address these, rather than simply implementing a strategy that was successful elsewhere. 19

7.1.13.3 Analysis of resources and skills

Two determinants of successful adoption and implementation of change require further discussion. It
is unlikely that individual hospitals will have sufficient resources to make appropriate changes to
their delivery systems in response to every gap or adverse event. If the level of risk associated with an
adverse event is sufficient to require action to be taken, an assessment should be made as to whether
the systems review and changes required are feasible. A complete review of a large delivery system
may require more resources than an individual hospital has available. If resources are insufficient to
undertake a complete review, senior management, together with the project team, should decide
whether to initially undertake a partial review of the relevant systems and make some of the changes

1«S
Enhancing patient care

required (especially those with the greatest leverage), or, if the system is large, to provide sufficient
additional resources to complete a comprehensive review and make all the required changes. Some
changes required to close gaps or reduce risks are costly, especially if additional staff or new
equipment are required; however, some simple strategies can also be effective. 9
The project team should also assess whether the members of the target group have the knowledge,
skills and experience to make the proposed changes, or whether they need further education and
training. The team should review previous successful change programs in the health service that
achieved similar goals, and identify strengths in the organisation that can be used for the current
project.
The project team should look to other health services that have successfully achieved the changes
being proposed. Have these health services published the methods they used, the results they
obtained, and the problems they encountered and solved when achieving this change? A literature
search will often locate such publications. It can be useful for project team members to speak with
people involved in these projects. Health services are often proud of what they have achieved in
improving patient quality and safety, and are pleased to talk about their experiences, receive visits
from other health services and share any relevant resources they have developed during their own
change projects. Seeing and learning about such projects not only provides valuable information for
the project team, but also increases motivation.

7.1.13.4 Determining best practice

The evidence of best practice in the delivery system under review should be sought. It is important
noi to spend excessive time undertaking an exhaustive literature review, as this is likely to delay
making ihe proposed changes. The most recent evidence-based review of the relevant clinical area is a
useful place to begin. However, high-quality information of best practice based on the results of
randomised clinical trials is frequently not available to answer many common questions in clinical
practice. Therefore, treatment recommendations are often based on less rigorous trials or on ihe
consensus of experts in the field. It is likely that other health services, professional bodies and
national health care organisations have developed treatment guidelines and clinical pathways for the
management of patients with the clinical conditions being reviewed. Copies of such documents are
readily available and will expedite the change process. They should be developed into local protocols
by the project team to meet local needs and conditions.

7.1.14 Step 14: the project team set shared goals

Although the health service has already outlined a vision for ihe proposed change, this should be
developed into a goal or goals. Goals provide a focus and direction for subsequent action. Having
goals makes it more likely thai they will be achieved. Working towards goals can be enjoyable, and
achieving them can be a significant source of job satisfaction. The goals should be set by ihe project
team and shared by its members and senior management.
A useful acronym in developing goals is thai they should be SMART (specific, measurable,
attractive, realistic and time-framed): 1,2
■ Specific — goals should clearly specify the actions required to achieve them. The specificity of the
goal helps determine whether the actions taken are achieving the goal.
■ Measurable — meaningful, valid measures allow objective evaluation of progress being made
towards the goal.
■ Attractive — goals should reflect what clinicians in ihe target group believe is important and want
to achieve
 7 A practical framework to close gaps in patient care and reduoe risks to ps*ients

■ Realistic — the goals set must be realistically, within the health service's and the team's
capabilities. The goals should reflect the resources available to achieve them, the abilities of the
participants to undertake the actions required to achieve them, and, if available, the health
service's past experience in achieving similar goals. The members of the team should feel
confident that they can perform at the level require^ to achieve the goals. If the goals are set
unrealistically high, success will be unlikely and motivation for the project and similar future
projects may decrease. Commitment to the goal will be high if clinicians in the target group feel
the goal is both important and attainable.
■ Time-framed — a date for achieving each goal should be set. Major goals often take time to
achieve, and it may be difficult to consistently maintain motivation for the change over a long
period. Therefore, the goals set • should be achievable in a reasonably shon time frame. If the
date for achieving the goal is not clearly visible on the short-term horizon, intermediate goals
with closer completion dates should be set. This action will allow smaller goals to be regularly
achieved aiding motivation for the long-term major goal. 3
An example of a goal that meets these criteria is:
to decrease our current 2% incidence of catheter related bacteraemia to 1% within 1 year by
implementing the following: insenion under full barrier sterile techniques, ultrasound'
guidance for difficult insertion, avoidance of the femoral insertion site, use of antibiotic
coated catheters for high risk patients or long term use, ho scheduled catheter changes, new
site replacement and avoidance of guidewire exchange, and dry gauze rather than
transparent dressing 9

7.1.15 Step 15: construct a list of strategies available to close the gap or reduce the risk
After a systematic analysis of the determinants and their efTects, importance and changeability
have been made, a list of the major and changeable determinants can be compiled. Some
members of the target group will already have been asked which strategies they believe would be
effective in implementing the proposed change. The list of determinants and strategies should be
verified with other members of the target group.
If the project team is reviewing an adverse event and one or more errors have been identified
there are four broad actions that can be considered. Ranked from strongest to weakest, they are:
1. reducing or eliminating the errors from occurring
2. making the errors visible
3. reducing harm to the patient after the error has occurred
4. creating a policy regarding behaviour in the clinical area and re-educating staff. 21
Unfortunately, often the weakest interventions arc used and only superficial attempts arc made
to improve patient safely. If the weakest intervention.
'I eanl metre ite
effort that* needed
to bring about
ethcthm change if
I'm not truly convinced it
is necessary'

lies V. Sutherland K
Organisational change:
a review for heath can
managers, profession's
ana researchers
creating a policy, was chosen, the effectiveness of this intervention can be measured in the following
ways:
■ the presence of the policy
■ the staffs knowledge of the policy
■ the use of the policy.
Previously, the presence of a policy was used extensively in health service accreditation programs as
a measure of effectiveness. Measuring the appropriate use of a policy is the most reliable method of
assessing the effectiveness of the intervention, and is the preferred method of measurement. 21

7.1.16 Step 16: choose the best strategy or strategies

After constructing a list of strategies available to close the gap or reduce the risk, the team should
decide which strategy or strategies are the most appropriate. After this decision has been made, the
team should determine whether there are sufficient resources available to make the proposed changes
and to educate staiT about them. If resources are not adequate, the team should assess whether there
is sufficient lime lo obtain the required resources and to act on the problem in a timely manner.
Once the issue of resources has been addressed, the project team should anticipate the potential
barriers to implementing the change. Strategies should then be developed to overcome these barriers.
Unintended negative consequences resulting from the strategy should also be anticipated, and action
planned to prevent them. Where possible, these implementation strategies should address the latent
factors identified in the analysis and improve defences.
Before finishing the development of the strategy, the team should assess whether the overall strategy
is too complex. A highly complex strategy is less likely to be successful.

7.1.17 Step 17: plan the Implementation of the change strategy carefully but efficiently
Careful planning is an essential pan of any change. The extent of the plan will vary with the size of the
project and the setting. As discussed previously, it is paramount to the success of a change project that
the project team have autonomy and responsibility for analysing and solving the problem and
implementing the solution.

7.1.17.1 Creating a plan

A plan is a series of concrete steps or actions designed to achieve a goal. Setting goals stimulates
planning. However, effective change requires a balance between planning and commencing action.
Time is required to analyse the problem, evaluate possible solutions, and carefully plan the necessary
actions to design and implement the change. Careful planning increases the likelihood that the
project's goal will be achieved. However, over-analysis of the situation may lead to procrastination.
Although detailed planning will reduce the chances of things going wrong, such planning is unlikely
to totally eliminate negative events. Some events will not be foreseeable, however painstaking the
planning. Also, if planning takes an excessively long time, this may delay action being taken; in turn,
inaction may reduce the target groups motivation for change. A balance between the lime spent
planning and commencing action is needed. Planning should be comprehensive but not exhaustive.
Attempts should be made to anticipate potential problems. Action can then be taken to prevent these
problems occurring, or, if they cannot be eliminated, preparations to deal with them can be made.

7. I. 17.2 Components of a plan

A written plan for the implementation of the change should be developed. The plan should include a
sequence of concrete, actionable steps that need to be taken to successfully achieve the proposed
change The plan should also include details of the individuals or groups responsible for each action
and
the target date for completion of the action. The resources required to achieve the change,
including people, materials and budget, should be determined and documented.
The timetable to complete the plan needs to be both realistic and feasible. In our experience,
hospitals make two common errors in planning:
■ Significantly underestimating the time to complete each of the actions requited to successfully
implement the change.
■ Trying to achieve too much change too quickly, which can lead to the failure of the project.
In our experience, it is best to begin with small steps towards achieving the overall goal and to plan
for some early successes. By testing the effectiveness of the various implementation strategies on a
small scale first, the effects of the strategies can be observed and if required, the strategies and
sometimes the goals can be adjusted. Using this small-scale iterative process, the change program
can be gradually expanded throughout the hospital and serious large-scale implementation
problems can be avoided.
The plan should include details of how the implementation will be organi^d, performance will be
measured, how tools for change and data collection will be developed and the project will be
evaluated.

7.1.17.3 Developing took to implement change strategies

Often, the strategies for implementing a proposed change require the development of a tool to
facilitate the change. Some of these tools can form pan of a clinical decision suppon system *in
which characteristics of individual patients are used to generate patient specific assessments or
recommendations that are then presented to clinicians for consideration''. 22 For example, to improve
the use of thromboembolism prophylaxis among inpatients, one strategy would be to develop a tool
to assess the risk of each patient developing a thromboembolus. Tools can be computer-based or
non-electronic.
The development of effective tools is critical to the success of a change project. Kawamoto and
colleagues reviewed 11 randomised controlled trials, and found four features of clinical suppon
systems that were associated with improved clinical practice: suppon was provided automatically as
part of the clinician workflow, delivered at the time and place of decision making, provided
actionable recommendations, and was computer-generated. Importantly, these features reduce the
amount of work and time required by clinicians to receive and act on the recommendations
generated by support systems.22 Tools should also be:
■ based on good-quality evidence
■ adapted to meet the needs of the target group, its subgroups and the characteristics of the setting
■ easy to understand
■ attractively designed
■ of high quality
■ difficult to miss in the medical record
■ readily available when required at the clinician-patient interface
■ incorporated into the daily routine of patient care
■ updated as new evidence becomes available.
Often, there is no need for health services to "reinvent the wheel" when developing tools. Sharing
information and tools for quality improvement and patient safety between health services can save
many hours of valuable staff time. Health services can also save considerable resources by not going
down the same "blind alleys" as other health services.
However, whenever tools have been developed in one health service, they arc unlikely to fit perfectly
into the delivery systems of another. Therefore, it is important to modify such tools to the target
group's needs and the health service's local conditions. Target group members will need input into
modifying
 Enhancing patient care

these tools, as local ownership is imponant. Finally, before changing health care delivery systems
radically, it is important to determine whether some systems and resources already in place may assist
in the change process.

7J.I 7.4 Develop a balanced set of outcome and process measures

The plan should include how performance will be measured during the project, including what
measurements will be made, when they will be made and how data will be collected. Where possible,
existing hospital data collection .systems and databases should be used. Valid, reliable measures of
performance toward the overall goal of the project should be developed.
The project team should devise a balanced set of system performance measures, preferably including
process and outcome measures. To be balanced, measures should be chosen from each of four
categories:23
■ clinical (eg, patient outcomes, compliance with treatment processes)
■ cost (eg, patient length of stay)
■ patient satisfaction
■ functional health status — in this category, a patient's health is measured in several areas using
standardised assessment tools before and after they receive medical care.
Of the four categories, functional health status requires the most resources and is used the least.
Where possible, measures should address the six dimensions of patient care identified by the US
Institute of Medicine: safety, effectiveness, patient-centeredness, timeliness, efficiency and equity. 2''
Where possible, we prefer to use process measures, as they are more sensitive than outcome measures
in assessing changes in system performance. For example, administering aspirin to patients with an
acute myocardial infarction is a treatment process that has been shown in high-quality clinical trials to
significantly reduce patient mortality, a clinical outcome measure. The proportion of patients with an
acute myocardial infarction who are given aspirin is a sensitive process measure, and significant
changes in the rate of compliance with this treatment can be rapidly demonstrated. However,
demonstrating a significant improvement in the related outcome measure of subsequent patient
monality requires a larger sample of patients, a longer time period and risk adjustment of the
outcomes.
A funher example is preventing venous thromboembolus in hospital inpatients. Using a process
measure, the proportion of inpatients who have venous thromboembolism risk assessment completed
and appropriate prophylaxis given, provides a meaningful measure of the success of the intervention
within a few days. However, using the outcome measure of the proportion of inpatients who develop a
deep vein thrombosis may take some time to show a statistically significant result. The performance
measures chosen should be:
■ Limited in number — too many measures increases the complexity of data collection.
■ Of high leverage — measures highly correlated with patient outcomes are most important.
■ Able to be clearly and unambiguously defined — measures that are poorly defined make analysis
less meaningful.
■ Easily identified and collected — if measures can be collected as part of the daily routine of existing
hospital staff, data will be easier to collect and more likely to be collected.
■ Low cost to collect — resources for quality improvement and risk management activities in health
care are limited. If these resources are not spent unnecessarily on complicated data collection, they
can be directed to other quality and safely activities.
These criteria are the same as those used for measuremeni of performance in conducting clinical audit
as outlined in Chapter 2.
The continuous monitoring and measuremeni of progress towards achieving ihe proposed change
should be part of the plan.

170
 7 A practical framework to close gaps in patient care and
reduoe risks lo patterns

7.1.17.5 Create data collection tools

Data collection tools that reflect the key performance measures chosen by the team should be created.
An assessment should also be made about whether the required data can be simply collected. Project
teams aim to improve the safety and quality of health care delivery systems in a timely manner, no*
conduct formal clinical research. Although the data collected should be valid and reliable, they should
be able to be collected efficiently as part of hospital staff's daily activities, and should not be a separate
task that requires substantial additional resources.

7.1.17.6 Education and training

The implementation plan should contain a program to educate staff about the proposed change and
training them in making the proposed changes. Effective education and training programs involve small
groups, are personal (with outreach visits), provide the opportunity for social interaction and use
opinion leaders. Education and training programs should develop and use high-quality materials

7.1.17.7 Sustainability
The changes should be fully integrated into the existing structures and processes in place in the
hospital and made part of the daily routine of providing patient care. Such integration will sustain the
changes.

7.1.17.8 Project name

Developing an appealing name and symbol for the project can help remind and motivate staff about the
change. A recent project to increase the use of thromboembolism prophylaxis in hospital inpatients was
entitled "stop the clot" and eye-catching stickers were used on medication charts to remind staff to
assess their patients' risk of thromboembolism.
As discussed earlier, if effective action can' be safely taken, it should be allowed to commence as
soon as possible with close monitoring of its effectiveness. Prolonged delays in commencing action may
lessen the target group's motivation.

7.1.18 Step IS: disseminate draft versions of the plan widely for comment and Input
A major mistake in developing implementation plans with potentially far-reaching effects is lo develop
ihe plan in isolation from those whose work will be affected by it. At a minimum, key stakeholders in
the change project should be provided with drafts of the implementation plan and their input actively
sought. Drafts of the plan should be placed on the agenda of the regular meetings of the peak medical,
nursing and quality committees or groups of the hospital, again seeking input and feedback.
As the drafts are revised, the details should be discussed regularly at key clinical and education
meetings at the hospital. Ideally, drafts of the plan should be widely disseminated throughout the
hospital. Comments on the plan should be discussed by the project team and changes and
improvements made accordingly. Such wide review before the implementation of the plan may flag
significant problems that were not originally foreseen by the project team.
The final version of the plan should be available to all staff, especially members of the target group,
stakeholders and opinion leaders. Copies should be sent to all major clinical groups in the hospital and
a copy placed on the hospital intranet. The plan should contain the waiver that it may be modified in
response to the results obtained after implementation.
Implementation of the plan should provide no surprises to the various stakeholders and other
hospital staff. It is clearly detrimental to the success of changc projects if major stakeholders and
senior members of the target group first learn of the change being implemented from their junior
mcdical staff or bv a memorandum from senior management to all staff.

171
 Enhancing patient care

7.1.19 Step I9: toke baseline measurements

A baseline measurement of the key performance measures selected should be taken before any change
strategies are implemented to avoid the Hawthorne effect (where observation of staff influences their
performance), and so thai improvement can be demonstrated after the changes have been made. Taking
buseline measurements also tests the suitability of the data collection tools and systems.

7.1.20 Step 20: implement the chosen strategies using small-scale rapid tests
Because health care delivery systems are complex and their outcomes are often unpredictable, the
strategy chosen should initially be implemented as a small-scale, low-cost and easily reversible test.
This allows knowledge to be gained about the properties of the redesigned system. It is then possible to
determine whether the new system and the strategies chosen for its implementation can achieve the
intended outcome and what unintended consequences may occur.
Small-scale tests also allow ihe suitability of materials and interventions lo be tested with the target
group, and the coherence and effectiveness of the change process and the accuracy of the planned lime
frames to be assessed. 25 If the outcome of the system changes is suboptimal, ihe effects of any negative
outcomes will be limited because of its small scale, and of little risk to the health service. It may also be
easier to convince staff who have doubts about the proposed change lo participate in a small-scale trial.
Despite a lack of high-quality scientific evidence of its effectiveness, the plan-do-study-aci cycle is a
well established quality improvement process. The cycle includes identifying system design or redesign
opportunities, setting priorities for improvement and implementing an improvement strategy. The model
is based on a trial-and-learning approach that requires setting aims, determining measures, making
changes, assessing results and taking further action (Figure 7.1). 26

7.1.21 Step 21: analyse the results of the test

A rapid assessment should be made after the firsi small-scale test. The results should be analysed and
a decision made about whether the system changes and implementation strategies should be adopted,
modified or abandoned. The most likely outcome is that the changes and strategies will require
modification, and further small- scale and rapid tests will need to be performed. The changes will be
usually ready for larger-scale implementation after several cycles of modifications and small-scale tests.

7.1.22 Step 22: take deliberate and continuous action to implement the plan
An obvious key to successfully achieving goals is taking action. However,
emergencies and distractions abound in busy modern health services,
and it can be easy to remain inactive about quality and safety issues,
Artao^ed frryn Leape et al' which often appear less urgent than

Figure 7.1 The plan-do-study-act approach to

performance improvement

What are we trying

to accomplish?

How will we know that a change is an

improvement7

What changes can we make that

will result in improvement?

t
172
 7 A practical framework to close gaps in patient care and
reduoe risks to patterns

other problems. Procrastination can also seem a comfortable solution to many problems, especially
in the shon term. However, there is usually a cost associated with procrastination. In the medium and
long term, inaction through procrastination will leave the problem unresolved. With time, the problem
may become larger, more complex, and more difficult to solve.
Purposeful action needs to be taken to increase motivation and achieve goals. The first steps are often
the most difficult — ambivalence will be high and the comfort of the status quo will appear attractive
However, by taking appropriate action toward achieving the goal, motivation is increased, ambivalence
reduced and further action results. A positive spiral of action, motivation and further action is created. 1
*

7.1.23 Step 23: monitor the progress being made to/ward the goal and provide feedback about progress
to the participants
The implementation should be closely monitored to ensure that the actions in the plan are completed
as scheduled. Stetler and colleagues suggested asking the following questions: A^e there discrepancies
between the plan and how it is being operationalised? Are there any barriers or facilitators that were
not foreseen in the planning stage? Is the plan being implemented to schedule? Do any pans of the
implementation strategy require revision? 27 The effects on patient care of the changes should be
continuously monitored using the performance measures detailed in the plan.
Continuous monitoring of progress towards the project's goal helps provide timely, accurate, tailored
and meaningful feedback to the target group and other staff. Continuous monitoring is more resource-
intensive than monitoring at set intervals; however, it reduces the chance that problems and poor
performance that occur during a non-monitoring period will be missed. Existing communication
avenues (eg, committees, education meetings and staff newsletters) can be used to provide regular
updates on the progress of the project, aggregate feedback on performance in providing clinical care
and the opportunity to discuss the results.
The effect of feedback on behaviour is variable, which may be explained by how it is given. In one
study, health services lhat provided care thai was highly adherent to clinical practice guidelines,
provided feedback to individual clinicians that was: timely enough to be useful and actionable (monthly
or more frequently was regarded as timely); about their individual clinical performance; given in a non-
punitive manner (assessed by the tone in which feedback was given), making it less likely to be
resisted; and, if possible, feedback could be customised by individuals according to their needs
(allowing them to put data into a form that was meaningful to them). Providing such feedback is
consistent with the feedback intervention theory and is giving "actionable feedback". 28
In interviews about data feedback conducted with 45 clinical and administrative staff in eight US
hospitals, major themes elicited were:29
a the data must be perceived as valid and credible by doctors (this may take time lo develop and is
partly dependent on the source and timelines of the data) a benchmarking (against other like health
services and over time within an individual health service)
improves the meaningfulness of the data a feedback is enhanced by having respected doctors (clinical
champions) review the data and present them to other doctors
a feedback to doctors about their individual performance may be effective, but may also be perceived as
punitive
a data feedback must be ongoing to maintain improved performance.
The perceived validity and meaningfulness of the data were central to the effectiveness of data
feedback and the process should be embedded in the delivery of care to patients. 20
Feedback about performance should be given as soon as possible after care has been provided Such
feedback can be a strong motivator for staff to continue their participation in a change project
Fccdhack to clinicians about their individual performance and their performance compared with* other
clinicians can be a powerful motivator, and may appeal to some clinicians' competitive nature. In a
Enhancing patient care

randomised controlled trial of physician performance in the management of patients with diabetes in an
ambulatory setting, physicians given feedback about their own performance together with achievable
benchmark feedback (the average performance of the top 10% of the physicians being assessed)
performed significantly belter than physicians only given feedback about their own performance.
Achievable benchmark feedback "represents a realistic standard of excellence attained by the top
performers in that group* and providing this information enhanced the effect of individual feedback. 30
Individual clinician feedback should be given privately and not made available to the project team —
feedback to the team should be given in aggregate form. Data comparing clinicians' performance
should not identify the performance of individual clinicians. As with other pans of the change process,
feedback can be commenced with the small-scale change trials, and its form and frequency can be
modified over time.
In addition to monitoring progress toward the shared goal or goals, the benefits and costs of making
the change need to be continually assessed. If the costs to individual clinicians begin to rise compared
with the benefits, their motivation may be reduced. If this reduction in motivation occurs, the plan and
goal may require adjustment.

7.1.23.1 Monitoring to deal with relapse

If the implementation of the project is positive, the clinicians involved will see that they are moving
closer to their goal. They will experience the benefits of making the change and this will help maintain
their motivation. However, relapse to the original behaviour and unexpected setbacks are often an
inevitable pan of change. Relapse needs to be expected, prepared for, and when it occuVs, corrective
action taken. Relapses may be temporary and due to external factors beyond the project team's control.
When a relapse occurs, the project team should relleci on what has happened, determine what went
wrong and take corrective action. The team and hospital can learn from relapses about what does and
does not work in the change process, especially at the local level. When a relapse occurs, team
members should be reminded that despite this outcome, they have done their best.
Perseverance, in spite of slow progress towards the goal or relapses in behaviour, is important in
ultimately achieving the goal. However, perseverance should only be sought from the team if the goal
remains reasonable. Therefore, after a relapse the team should consider whether the goal is still
realistic and whether the strategies being used remain appropriate. If not, it may be appropriate for the
team to revise their goal and the plan and strategies they are using. In retrospect, it may be fell that
the initial goal was set unrealistically high.

7.1.24 Step 24: communicate proposed changes and results to staff

Hospital staff should regularly receive updates about a project's progress. A balance in the quantity of
information provided is required between overload, where staff may ignore important information, and
insufficient information, which may make implementation at the coalface difficult. The project team
should determine the appropriate mode and frequency of communication to staff. Options include:
posters, emails answering frequently asked questions, messages in communication books, newsletters,
staff meetings and in-service training. It is important to engage key medical and nursing staff in the
project. This can be assisted by sending drafts of new policies or changes to existing policies, processes
and forms for their information and comment before implementation of any changes.
Once the changes have been implemented, the results obtained should be reported regularly to all
clinical staff members, senior management through the committee structure of the hospital and
ultimately to the board of management. Hospitals in Victoria are now required to publicly report the
 7 A practical framework to close
gaps in patient care and reduoe
risks to patterns

improvements in health care delivery that they have achieved in annual

"quality of care" reports.

7.1.25 Step 25: assess whether the changes made are sustainable
After implementing the changes and having them in place for some time,
the team should assess whether the changes are sustainable and have
become routine practice. If sustainability has not been achieved, further
work is required. The change process is not about achieving short-term
results, but aims for sustainable, long-term improvements. Such
improvements require discipline lo remain focused on the changes being
implemented, and vigilance to ensure the project is achieving its goals.
In the Wimmera Health Care Group, we have been able to achieve
significant, measurable improvements in the quality of care provided in a
particular clinical area on several occasions. We have then concentrated
on another clinical area requiring improvement, only to find that the
improvement achieved in the first area has been significantly eroded some
time later. Unfonunately, a reduction in clinical improvement often only
comes to our attention when an adverse event occurs.
Frequently, an adverse event recurs because the initial changes made
to the relevant delivery system were not maintained. The recurrence may
also be due to factors that were not detected in the initial analysis, or not
present when the initial adverse event occurred. Once an improvement
has been achieved in an area, sample measurements of key processes
related to the redesigned system should continue to be made and
reported. The project team could then meet once or twice a year to review
the ongoing results of the change. Sustainability is hard to achieve, and,
in our experience, constant vigilance is required to maintain the positive
changes that have been made.

7.1.26 Step 26: when the project's goal Is achieved, celebrate the success
and reward the participants
When the project's goal has been achieved, the accomplishment should
The quality at erne
be celebrated within the health service and publicly recognised. The cannot improve i we
important role played by each of the project team members should be harness the Knowledge
acknowledged. Consideration should also be given to appropriately and creative energy at
rewarding the team members. Appropriate rewards for clinicians include physicians and other
formal, public recognition of their contribution to the project and the health professionals tor
significance of their work in improving the quality and safely of patient the purpose of
redesigning the intricate,
care, and supporting them to participate in activities to maintain their
interlocking processes
knowledge and skills in their area of clinical expertise. Rewards may also
that constitute modern *
need lo be tailored to reflect the personality differences of some project healthcare.... physicians
team members. do have a number of
The efforts of the project team members could also be recognised by skills and attributes that
helping them publish a paper in a professional journal describing the no other participants in
methods used and the results achieved in the project. Publishing a paper the healthcare
marketplace have in quite
serves two purposes. Firstly, it allows the staff who participated in the
the same combination
project to be publicly acknowledged to their peers. Secondly, others in the
scientific training and
health care field can learn about how behaviour was successfully understanding of medical
diagnostics, and
therapeutics, an
understanding of
patients'individual
circumstances, an
ongoing personal
relationship with patients,
and perhaps most
important, an ethical and
professional
commitment to placing
patients 'welfare first'

Blumenthai D.
Epstein A N Engl
J Med 1996.
335 1328-1331
changed in a clinical area in a health scrvicc and can consider the
appropriateness of applying these lessons in their own scrvicc.
 7 A practical framework to close gaps in patient care and
reduoe risks to patterns

7.1*27 Step 27: evaluate the project

The project should be evaluated and the project team should reflect on the process in which they
have participated. Were the goals of the project achieved? Were the steps of the implementation plan
undertaken as proposed? What did and did not work at each stage of the project? Did the overall
benefits of the change in practice outweigh the costs? What were the positives and negatives
associated with the change process? How could some of the lessons leamt about health care delivery
systems, problem solving and quality improvement be transferred to other groups in the hospital who
are planning projects to change clinical practice, and to the health care field generally?
Stakeholders can also be asked about the value of each intervention, their level of satisfaction with
the intervention and change process, additional barriers or facilitators they became aware of as the
change was implemented, and any refinements that would make in undertaking future change
projects.27

7.1.28 Step 28: continue to monitor performance and resource availability

When the change in clinical practice has been successfully implemented, the health service may
consider the project complete and stop monitoring results in that clinical area. The health service
may also reduce or withdraw funding for the activity. If the clinical changes are now embedded in
routine patient care, these actions may be appropriate. However, in our experience, such action may
result in clinical practice relapsing to levels before the project commencing. Some change processes
require ongoing support and vigilance for the changes to be sustained.
Clinical practice should continue to be monitored even though the change project has formally
concluded, and feedback about clinical performance in this area should continue to be provided to
relevant groups as occurred during the project. If the measures monitored fall over time, action will
be required to deal with these relapses.
If ongoing resources are required for the changes and clinical improvements to be maintained,
steps should be taken to ensure they continue to be provided. The project's executive champion is in
an ideal position in the hospital to ensure lhat, if required, such resources continue to be made
available to maintain the improvements achieved. The executive champion has extensive knowledge
of the change project having been an active member of the team and possessing the detailed
knowledge of the processes required by the hospital to be followed to obtain ongoing resources.

7.1.29 Step 29: regularly update and review the scientific evidence for best practice in the clinical ana
With the large amount of research being undertaken, what constitutes best practice in a clinical area
may change rapidly. It is unlikely that clinical practice guidelines will remain current for long.
Therefore, when the project is completed, a date should be set to review the evidence regarding
clinical practice and what constitutes best practice in the area of the change. This review should
occur earlier if research substantially altering best practice is published.
It is difficult to specify a time period in which a review of best practice should occur. However, as a
guide, in an analysis of 100 quantitative systemic reviews, the median time before substantive new
evidence signalled that updating the review was required was 5.5 years. However, in 23% of reviews,
a signal for updating occurred within 2 years of the publication of the review. 31
7.2 Motivating doctors to participate in quality
and safety programs
With their key role in delivery of patient care, doctors should be an
integral part of the quality and safety programs in the health services in
which they work. For a variety of reasons, such involvement does not
always occur. Therefore, it is important to review doctors' attitudes
towards quality and safety programs and to discuss factors that may
increase their engagement.
Most doctors care greatly about the care they provide to their patients.
However, some believe that it is only their skills that protect their
patients from harm from the system. Further, most doctors believe that
they provide good-quality care that does not need to improve. Therefore,
many doctors are not motivated to change. They would be motivated to
change if they fell more discomfort with the current level of quality being
provided. For this reason, quality of care needs to be consistently
measured, although some doctors resist being measured. 32
Although many doctors are passionate about the quality of care they
provide to their patients, doctors as a group are often viewed as obstacles
to quality improvement and patient safety. Some doctors view such
programs as interfering with the delivery of care to their patients and a
waste of lime, although others are clearly leaders in this area. 32
Difficulty in fully engaging doctors in quality and safety programs has
been a longstanding problem in health care, and is a significant issue.
Doctors regard clinical governance as primarily financially and patient-
throughput driven. They view audits as time-consuming and resulting in
little change in practice. They report few clinical incidents. Why do
doctors behave this way toward quality and safety programs? Are doctors
just too busy, or do they need to remain in control to participate in these
programs and feel this control is threatened? 33
Doctors and other clinical groups need education about safety and
quality, and this should start at university (even though the curriculum
is already crowded). However, medical, nursing and pharmacy students
receive little education about patient safety in their undergraduate
courses.34 Therefore, their education in health services about quality and
safety after they graduate is important. In addition, ongoing teaching
about errors and adverse events as occurs for clinical advances may be
required.
This education should begin at orientation, with staff told of the
programs in the service and also what incidents to report and how to
report them. Regular education sessions could cover many of the topics "TnapHtofevmyHnt
discussed in this book, including how errors occur, the importance of that hat occurred. IHe
latent factors, how adverse events are detected and acted upon, and majority of doctors today
lessons learnt from external sources. Individual and group feedback are Stat not very Involved
about incidents that have been reported and adverse events detected can in the promotion of
patient safety. There is
be given, and quality and safety initiatives currently being undertaken
leas resistance... but
there is not the active
participation that is
needed. Safety is not a
daily concern of most
doctors, and it should be.
Nursing is way ahead of
us on this, but even there
things can be improved.
Physician apathy
continues to be
area/drag.'

Leape L
Health Alt
(HMwood)2007: 26.
w687-w696
 7 A practical framework to close
gaps in patient care and reduoe
risks to patterns

discussed. Disclosure of information to patients and relatives about

adverse events that have occurred should also be discussed. Using
lectures and monthly conferences has resulted in no
overall change to junior doctors' attitudes and behaviour to quality and safety issues, with major
barriers cited being lack of time, extra paperwork and career and reputation concerns. 35
Given the increasing complexity of medicine, with patients having more comorbidities and the
availability of more diagnostic and treatment options, doctors need to work as pan of a team to solve
difficult, complex problems. However, doctors are given almost complete autonomy to provide health
care. Therefore, influencing their behaviour can be difficult. Their training instils fierce professional
pride, individualism and resistance to attempts to reduce their independence. Some doctors find it
difficult to work in teams. 36 Doctors are sceptical of and resistant to quality improvement programs
because of a reluctance to change, misunderstanding the aims of the programs, and a dislike of how
programs are promoted. It takes time to influence doctors. Their initial support of programs may be
fragile, and they will need to see tangible and ongoing evidence of the benefit of the program. 37 They
have seen many local and national initiatives in the past, and are sceptical that new programs will
make any difference to the quality of care or their work environment. Some reasons why doctors do
not follow clinical practice guidelines are given in Box 7.1. 38 Engaging doctors in quality and safety is
complex, and there will not be one simple solution. As discussed in Chapter 6, change strategies
should be tailored to the characteristics of the target group (often substantially comprising doctors)
as well as the characteristics of the intervention, the setting, the patients and the context. In each
situation, the important barriers and facilitators will need to be identified and strategies developed to
overcome them to change clinical behaviour.
Sometimes, the link between appropriate action and its effect is delayed, influencing clinical
behaviour. Doctors may not use some evidence-based preventive interventions because they do not
receive direct, immediate feedback — and by consequence, positive reinforcement — about the
intervention's effectiveness in preventing a negative outcome in an individual patient. The benefit to
the patient is real and often significant, but it is delayed. This is in direct contrast to some evidenced-
based therapeutic interventions, where the benefit to the patient is obvious and the doctor receives
direct, immediate feedback that the intervention has had a positive effect on the patient's outcome. 9
For example, ensuring that patients who have had an acute myocardial infraction are taking P-
blockers will reduce the probability of them having further infarcts, but the effect is not immediately
obvious to the medical staff providing the medication. In contrast, administering intramuscular
adrenalin to a patient with acute anaphylaxis
will have immediate effect, reinforcing its
Bo* 7.1 Reason* why doctors do not follow dMcal benefit to medical staff.
Many of theses factors have been discussed
practice guidelines *
3

■ Unaware of guideline in earlier chapters. In any situation, the

■ Difficult to access guideline important factors impeding clinicians taking
■ Lack of time to become familiar with guideline appropriate action should be identified. Many
■ Do not agree with content of guideline of these barriers can be modified or eliminated,
■ Guideline not applicable to the patient and implementation strategies can be tailored
■ Guideline not cost-beneficial to overcome them and improve care (Checklist
■ Source of guideline not credible 7.1).
■ Guideline too general
■ Guideline too rigid
■ Guideline challenges autonomy
■ Guideline not practical
■ Believe cannot perform guideline
recommendation (self-efficacy)
■ Believe following guideline will not lead to
desired patient outcome
■ Habit and routine of previous practices
■ Guideline inconsistent with patient preferences
■ Presence of contradictory guidelines
■ Lack of time or resources to use guideline
■ Organisational restraints
■ Lack of reimbursement for using guideline
■ Perceived increase in malpractice liability if use
guideline.
 7 A practical framework to close gaps in patient care and
reduoe risks to patterns

Checklist 7.1 Engaging doctors in quality Improvement and patient safety programs Actions_____tts NetWF *A
1 The effectiveness of the proposed change is supported by High-qualHyaueiiUlK.evKleiiusfoni v
a credible source.
2 The proposed change is in an area of clinical importance to physicians and improves their patients' outcomes.

3 The change is not primarily a cost-cutting exercise.

4 The change does not impose significant work and time commitments on already-busy doctors.

5 Doctors are consulted in the early stages when the change proposal is being developed and
asked for their input. Any legitimate concerns they express are adequately addressed. I l l
6 The effectiveness of the implementation strategy for the change process of change is scientifically based and is not just the latest management fad.
«

7 Adequate resources are provided to implement the change.

8 Doctors are given protected time to participate in the program.

9 It is easy for doctors to participate in the change process.

10 Management is consistent in its approach to quality improvement and safety initiatives.

11 There is little turnover in management staffing. 34

12 Doctors are provided with confidential, timely and meaningful feedback about their performance in implementing the change.
13 The participation of doctors in the program and the results achieved are publidy acknowledged.
N/A = not applicable. WIP = work in progress.

7.3 The framework in action — clinical pathways39

Many of the strategies to close gaps and reduce risks discussed in this book have been incorporated
in the development of clinical pathways at Wimmera Health Care Group.
Wherever possible, the care provided to patients should be consistent with best practice according to
the highest-quality scientific evidence. To help health professionals, evidence-based clinical practice
guidelines have been developed to manage many clinical conditions. These guidelines can be
complex and lengthy, and multiple guidelines can exist for the management of the same condition.
Further, these guidelines may not be readily available or easy to use in acute care situations. Their
effectiveness may be limited by resource constraints such as staff skills and equipment availability.
Clinical guidelines need to be simplified and incorporated into everyday hospital practice.
Many gaps identified by quality improvement programs, and adverse events detected by risk
management programs were clustered around the care provided to patients with panicular clinical
conditions. There was high variability in the care provided. With the shon rotations and high
turnover of junior medical stalT, high compliance with evidence-based clinical guidelines would be
unlikely to be
achieved.
Using system redesign, it was decided to:
■ determine best practice in the management of high-frequency clinical conditions:
■ modify best practice for local needs and conditions;
■ break down best practice into a series of actions; and
■ incorporate these actions into a daily checklist of reminders that would form the patient's mcdical
record
Such clinical pathways were developed for 30 clinical conditions, and the following steps were
followed:
1 A senior nurse was appointed as the clinical pathway coordinator.
2. There was an established clinical risk management culture in the hospital. Clinical and
executive champions were identified and invited to participate in the pathway program.
3. Clinical conditions were chosen for pathway development because
> Gaps and adverse events were consistently delected in management of patients with the
conditions.
Suboptimal treatment of the conditions resulted in considerable patient
harm. >• Some clinical conditions were treated frequently in the health
service.
4. A multidisciplinary team was established to develop the pathway. If individual members were
required to undertake considerable additional work in developing the pathway, they were paid for
their time or they were replaced in their normal duties.
5. Information was collated on what constituted best practice for the condition under review.
Sources of information included a literature search, clinical guidelines developed by other
organisations and information about how the condition was managed at other health services.
6. The pathway was developed. It contained information for clinical decision making, evidence-
based guidelines, and checklists and reminders for patient care in the emergency department
and each day in hospital.
7. Draft pathways were circulated for comments to all medical staff and other relevant staff.
8. Small-scale trials of the pathway were conducted and appropriate modifications made.
9. Education sessions about how pathways were to be used were conducted for staff.
10.Audits measuring compliance with process measures known to have considerable positive
impact on patient outcome were undertaken. If a gap was detected, corrective action was taken
and monitoring continued. Several cycles of action, monitoring, analysis and modification were
required to close the
gap
11. Medical staff members were provided with feedback about their individual performance.
12. The results of the program were reported using a top-down and bottom-up approach — up
organisation to board of management and down the organisation to all staff.
The limitations of clinical pathways include:
■ It is not feasible to develop clinical pathways for the management of every clinical condition. In
our experience, using 30 clinical pathways in several specialties, about 46% of the patients on the
wards at any one time are on a clinical pathway.
■ Each clinical step in a pathway is not necessarily appropriate in the management of every patient
on that pathway. Clinical judgment must still be used to determine an individual patient's care.
■ Management of patients with some medical conditions, such as stroke, is complex, but such
pathways can still be developed, despite being difficult to write. Often, management of patients for
elective surgery can be much simpler, and pathways for these clinical situations can be developed
more quickly.
■ A patient is placed on a clinical pathway for management of their major clinical condition.
However, some patients have more than one significant clinical condition. In our experience, care is
required to ensure that their other significant comorbidities continue to be actively managed and
not neglected.
■ Pathways are not suitable for the management of patients with two significant conditions for
which clinical pathways are available. It is confusing and may be detrimental to place a patient on
two pathways simultaneously.
■ Effective use of clinical pathways is resource-intensive. Each new clinician requires orientation
and ongoing education about the method With many junior medical staff rotating to Wimmera Base
 7 A practical framework to close
gaps in patient care and reduoe
risks to patterns

Hospital every 10 weeks, such education takes significant time. Also,

the pathways used at Wimmera Base Hospital are paper-based, and
audits of their use are also resource-intensive. In addition, pathways
should be kept current for changes in best practice, and this requires
ongoing scanning of the scientific literature.

7.3.1 Case study — the stroke clinical pathway •

Using medical record review, adverse events were detected in the manage-
ment of patients admitted with stroke. These events included aspiration
and prolonged length of stay.
A multidisciplinary team was formed, with the aim -of reducing the
frequency of these events and establish a plan of care for acute stroke
management. The team had the support of senior executive and clinical
stafT to help implement change strategies. The team aimed to produce a
clinical pathway for management of patients with acute stroke, and was
given adequate resources and infrastructure support to undertake the
task. Membership of the team included clinical stafT strongly committed
to providing high-quality care and comprised: a specialist physician, a
resident medical officer, a pharmacist, a physiotherapist, a speech
therapist, a social worker, an occupational therapist, nursing staff from
the medical ward, emergency department and rehabilitation area, and a
clinical pathway coordinator.
A literature search was undertaken to identify evidence-based best
practice for each stage in the management of patients with stroke. The
pathway was designed to be the medical record of the patient's admission.
The pathway contained: a summary of evidence-based clinical guidelines
for easy reference; a flow diagram of major diagnostic and management
decision points; detailed pro-forma medical admission and allied health
assessment forms; a nursing risk assessment form; a checklist for
medical, nursing and allied health staff regarding management,
investigation and referrals for each day of the admission; and sections for
medical and allied health clinical notes.
All clinical areas involved in the care of these patients were sent draft
copies of the pathway requesting feedback. This feedback resulted in
significant modifications. A final pathway was then produced, and exten-
sive education sessions were held with nursing, resident medical officers
and visiting medical staff regarding its implementation.
An evaluation of the pathway implementation was undertaken, which
included stroke patients admitted between 31 July 1999 and 30 April -HyouwoMBmm start
2000 (prepathway) and between 1 June 2000 and 31 December 2002 tomorrow to change
(postpath- way). This evaluation measured the degree of compliance with practice and implement
evidence, prepare well:
four key process measures before and after the introduction of the
involve the relevant
pathway. The four measures were chosen to be followed were the people; develop a
proportion of patients with stroke who: proposal for change that
a had a computed tomography brain scan done within 24 hours of is evidence-based,
admission (100% prepathway v 99.0% postpathway); a had a dysphagia feasible, and attractive:
screen performed within 24 hours of admission (51.8% v 92.5%); Study the mam difficulties
in achieving the change,
and select a set of
strategies and measures
at different levels linked to
that problem; of course,
within your budget and
possibilities. Define
indicators for
measurement of success
and monitor progress
continuously or a regular
intervals. And. finally,
enjoy working on making
patients' care more
effective, efficient, safe,
and friendly.'

Grol R. Grimshaw
J Lancet 2003: 362
122S-1230
 Enhancing patient care

■ if the stroke was ischaemic stroke, were given aspirin or clopidogrel within 24 hours of admission
(36.8% v 92.2%). and
■ had regular neurological observations performed during first 48 hours after stroke (25.9% v
78.3%).
These improvements have now been sustained for a period of 8 years (see Figure 4.4). The pre- and
postpathway results were reviewed by the team and reported in graphical form to all clinical groups
in the hospital. The progress of the team in developing and implementing the pathway was reported
each month to the hospital's peak quality committee, which is a subcommittee of the board of
management.
Following the discussion in this chapter, a checklist of the practical steps to be considered when
planning to implement change is given in Checklist 7.2.

182
 Enhancing patient care

Checklist 7.2 Implementing change

Actions Yes No WIP N/A

11dentify a gap in the quality of patient care, or detect a significant risk to patients.
2 Analyse the gap or risk to determine its causes and whether any change is required in the way health care is
delivered in the clinical area.
3 Determine the quality of the evidence supporting the benefits of the change.
4 Identify the target group, major stakeholders, opinion leaders and clinical champions.
■ Target group — staff members who need to change their behaviour to successfully implement the change.
■ Stakeholders — individuals or groups who will be affected by the change, and whose cooperation is needed to
implement the change.
■ Opinion leaders — respected and influential clinicians whose opinion members of the target group will seek
about the change.
■ Clinical champions — clinicians with a strong motivation to change behaviour in the area of the change.
5 Determine whether the gap or adverse event in which the health service would like change to occur is an
acknowledge any ownership of the issue.
important issue for the target group, and whether they
6 Describe the present state of health care delivery regarding the gap or adverse event to the target group.
■ Ensure data are accurate, clear and concise.
7 Determine whether clinicians are dissatisfied with the quality of the care presently provided.
8 Present a clear, realistic vision for the future to the target group.
9 Weigh up the benefits and the costs of making the change.
10 Determine whether the management of the hospital is prepared to support the proposed change.
■ Adequate resources need to be provided.
■ A senior manager should be actively involved on the change project team.
11 toaem the stage in the change cycle that individual clinicians in the target group are in.

Checklist 72 (continued on next page)

Checklist 7.2 [continued)

Actions Mo ^

183
 Enhancing patient care

12 Form a project team: «

■ that is short-term
■ that is multidisciplinary ^
■ whose membership comprises representatives of target group (and its subgroups),
clinical champions, major stakeholders, individuals with required expertise, project
coordinator, secretarial support
■ that has terms of reference.
13 Undertake a comprehensive, systematic analysis of the gap or adverse event
Identify the determinants of the proposed change or intervention, target group, patients,
organisation and sociopolitical context, and their barriers and facilitators.
■ identify latent factors and active errors
■ widely seek and validate information from target group and relevant others to
identify important determinants

■ identify the major determinants that can be changed

■ ensure resources are adequate to make the change
■ determine best practice in the clinical area under review.
14 The project team should set shared goals.
■ Goals should be are specific, measurable, attractive, realistic and time-framed.
15 Construct a list of strategies available to close the gap or reduce the risk.
■ Where possible, appropriate strategies are matched against major determinants.
■ Evidence-based strategies are considered if appropriate — dissemination of
educational materials, audit and feedback, reminders, multifaceted interventions
incorporating education outreach, and opinion leaders.
■ System redesign is considered — simplifying systems, standardising procedures,
using reminders and checklists, introducing constraints, providing timely and accurate
communication.
16 Choose the best strategy or strategies.
■ Is the strategy simple?
17 Plan the implementation of the chosen change strategy carefully but efficiently.
■ Put plans in writing and detail the actions to be taken to achieve the proposed
change, the individuals responsible for completing each action and the date when the
action will be completed.
■ Identify barriers to implementation and design strategies to overcome them.
■ Develop data collection tools that are easy and efficient to use.
■ A balanced set of outcome and process measures are developed:
> At least one measure from each category: clinical, cost, patient satisfaction and
functional health status.
Measures are limited in number, highly correlated with patient outcomes, of high
leverage, clearly defined, easily identified and collected and low cost to collect.
18 Widely disseminate draft versions of the plan for comment and input.
19 Take baseline measurements.
■ Could the required data be obtained and were adequate data collection systems in
place?

184
 Enhancing patient care

('Wis! ~ (continued ,->-

nape1
Checkfict 7.2 (continued)

OIMIJIO Yas No WIP N/A

20 Implement the chosen strategies using small-scale rapid tests.

21 Analyse the results of the fkst test.

■ Do the chosen strategies require change? Abandon or modify the strategy if required.

■ Do the project goals require change?

■ It is most likely that several cycles of testing will be required.
■ When test is successful, implement changes more widely in the organisation.
22 Take deliberate and continuous action to implement the plan.

23 Monitor the progress being made towards the goals, and provide feedback about progress to the participants.

24 Communicate the proposed changes and results to staff.

25 Make assessments about whether the changes made are sustainable.

■ Are the changes embedded into the daily routine of patient care?

26 When the goal is achieved, celebrate success and reward the participants.

27 Evaluate the project.

■ Reflect on the lessons leamt during this change process.
28 Continue to monitor performance and resource availability.

29 Regularly review the scientific evidence for best practice in the clinical area and update practice in the
hospital.

W1P = work in progress. N/A = not applicable.

Important points
■ Strategies to successfully change behaviour must involve the target group (those in the health sanrioe
required to change their behaviour in order to successfully implement the change).
■ Given the complexity and unpredictability of health care systems, results of change strategies must be
closely monitored. Timely feedback of results should occur, and modification of strategies or goate made if
required.
Key characteristics ot successful change projects
■ High-quality scientific evidence supports the benefits of the proposed change.
■ The proposed change is important to the members of the target group.
■ Senior management support Ihe change by providing adequate resources and a senior manager to actively
participate in the project team responsible for implementing the change.
■ There is a large gap between the present and desired states of health care deliver in the clinical area in
which the change is proposed.
■ The benefits to the target group of the proposed change outweigh the costs.
■ A project team comprising target group representatives, those with required expertise and support staff is
created.
■ The project team and senior management develop shared goals to achieve the proposed change.
■ The project team is responsible for analysing the gap or risk, creating a detailed plan to close the gap or
reduce the risk, and implementing the plan.

185
 Enhancing patient care

■ Comprehensive, systematic analysis of the determinants influencing the success of the change (the
innovation, target group, organisation and sociopolitical context) is undertaken.
■ The plan addresses each of the major determinants (facilitating and impeding) to achieving the change and
comprises a sequence of actions with timelines designed to achieve the change.
■ Deliberate, continuous action is taken to implement the plan, commencing with small-scale rapid tests.
■ Individual, timely feedback is given to target group members about their performance.
■ Project team members are recognised for their contribution when the change is achieved.
■ The results of the change are monitored and reported after the change project is completed.
Engaging doctors In quality and safety programs
■ Doctors should be an integral part of a health service's quality and safety program.
■ Most doctors are committed to providing high-quality, safe care to their patients.
■ Doctors are more likely to be engaged in quality and safety programs if making changes does not require a
significant amount of their time, the proposed change is supported by high-quality scientific evidence and
will improve their patients' outcomes, they are consulted early and consistently about the proposed changes
and are provided with adequate resources and timely, meaningful feedback about their performance.
References
1 Grant A. Greene J. It* your life. What are you going to do with it? 2nd ed. Harlow, UK:
Pearson Education, 2004.
2 Grant A. Greene J. Coach yourself 9 work. Get yourself a (working) life. Sydney: ABC

Books. 2005.
3 Montgomery R. The truth about success and motivation. Melbourne: Lothian, 1987.
4 Chassin MR. Quality of health care part 3: improving the quality of care. N Eng J Med
1996; 335:1060-1063.
5 Acton JD. Kotogal U. Improvements in healthcare: how can we change the outcome? J
Pediatr 2005; 147: 279- 281.
6 Pronovost PJ. Thompson DA, Holzmueller CG, et al. Defining and measuring patient
salety. Crit Care Clin 2005; 21: 1-19.
7 Blendon RJ. DesRoches CM, Brodie M, et al. Views of practicing physicians and the
public on medical errors. N Engl J Med 2002; 347: 1933-1940.
8 Berwick DM. Errors today and errors tomorrow. N Engl J Med 2003; 348: 2570-2572.
9 Cook DJ, Montori VM, McMullin JP. et al. Improving patients' safety locally: changing
clinician behaviour. Lancet 2004: 363: 1224-1230.
10 Pronovost PJ, Berenholtz SM. Goeschel CA, et al. Creating high reliability in health
care organizations. Health Sen/ Res 2006; 41: 1599-1617.
11 Berwick DM. Continuous improvement as an ideal in health care. N Engl J Med 1989;
320: 53-56.
12 Blumenthal D, Epstein A. The role of physicians in the future of quality management. N
Engl J Med 1996: 335: 1328-1331.
13 Cook DJ, Meade MO, Hand L, McMullin JP. Toward understanding evidence uptake:

semirecumbency for pneumonia prevention. Crit Care Med 2002; 30:1472-1477.

14 Angus DC, Black N. Improving care of the critically ill: institutional and heallh-care

system approaches. Lancet 2004; 363: 1314-1320.

15 Shojania KG, Grimshaw JM. Still no magic bullets: pursuing more rigorous research in

quality improvement. Am J Med 2004; 116: 778-780.

16 Robbins S, Bergman R, Stagg I. Coulter M. Management. 2nd ed. Sydney: Prentice Hall.
2000.

186
 Enhancing patient care

17 Kochevar LK, Yano EM. Understanding health care organization needs and context.
Beyond performance gaps. J Gen Intern Med 2006; 21: S25-S29.
18 Shojania KG, Grimshaw JM. Evidence-based quality improvement: Ihe state of the

science. Health All (Millwood) 2005:24: 138-150.

19 Schouten JA, Hulscher ME, Natsch S, et al. Barriers to optimal antibiotic use for

community-acquired pneumonia at hospitals: a qualitative study. Qual Sal Health Care

2007; 16: 143-149.
20 Bosch M, van der Weijden T, Wensing M, Grol R. Tailoring quality improvement

interventions to identified barriers: a multiple case analysis. J Eval Clin Prac 2007; 13:
161-168.
21 Berenhoitz SM. Pronovost PJ. Monitoring patient safety. Crit Care Clin 2007; 23: 659-

673.
22 Kawamoto K. Houlihan CA, Balas EA, Lobach DF. Improving clinical practice using

clinical decision support systems: a systematic review of trials to identify features critical
to success. BMJ 2005; 330: 765-768.
23 Nelson EC, Mohr JJ, Batalden PB, Plume SK. Improving health care, pari 1: the clinical

value compass. Jt Comm J Qual Improv 1996; 22: 243-258.

24 Institute of Medicine. Crossing the quality chasm: a new health system for the twenty-

first century. Washington, DC: National Academies Press. 2001.

25 van Bokhoven MA, Kok G. van der Weijden T. Designing a quality improvement

intervention: a systematic approach. Qual Saf Health Care 2003: 12: 215-220
26 Leape LL. Kabcenell A, Berwick DM. Roessner J. Reducing adverse drug events.

Boston: Institute for Healthcare Improvement. 1998.

27 Stetler CB. Legro MW. Wallace CM, et al. The role ol formative evaluation in

implementation research and the OUERI experience. J Gen Intern Med 2006; 21: S1-S8
28 Hysong SJ. Best RG, Pugh JA. Audit and feedback and clinical practice guideline

adherence: making leedback actionable Implement Sci 2006; 19

29 Bradley EH, Holmbee ES, Matters JA. et al. Data feedback efforts in quality
improvement losaont teamed from US hospitals. Qual Saf Health Care 2004:13:26-31.
30 Kiefe CI. Allison JJ. Williams OD. Person SD. Improving quality improvement using

achievable benchmarks for physician feedback: a randomised controlled trial. JAMA

2001; 285:2871-2879.
31 Shojania KG. Sampson M, Ansari MT. et al. How quickly do systemic reviews go out
of dale? A survival analyse Ann Intern Med 2007; 147: 224-233.
32 Goode LD. Clancy CM. Kimball HR. et al. When is "good epough"? The role and

responsibility of physicians to improve patient safety. Acad Med2002; 77. 947-952.

33 Neale G. Vincent C. Darzi A. The problem of engaging hospital doctors in promoting

safety and quality in clinical care. JRSoc Health 2007; 127: 87-94.
34 Leape L. Is hospital patient care becoming safer? A conversation with Lucian Leape.
Interview by Peter I Buerhaus. Health Alt (Millwood) 2007; 26: w687-w696.
35 Coyle YM, Mercer SQ, Murphy-Cullen CL, et al. Effectiveness of a graduate medical
education program for improving medical event reporting attitude and behaviour. Qual
Saf Health Care 2005; 14:383-388.
36 Blumenthal D, Kilo CM. A report card on continuous quality improvement. Milbank Q
1998; 76: 625-647.

187
 Enhancing patient care

37 Gollop R, Whitby E, Buchanan D. Ketley D. Influencing skeptical staff to become

supporters of service improvement: a qualitative study of doctors' and managers'
views. Qual Saf Health Care 2004; 13. 108-114.
38 Cabana MD, Rand CS, Powe NR, et al. Why don't physicians follow clinical practice

guidelines? A framework for improvement. JAMA 1999: 282: 1458-1465.

39 Wolff AM, Taylor SA, McCabe JF. Using checklists and reminders in clinical

pathways to improve hospital inpatient care Med J Aust 2004; 181: 428-431. □
Enhancing patient care

in

188
 6 Reporting the activities and results of the clinical governance popem
8 Reporting the activities and results of the clinical
governance program
This chapter will:
a Outline the activities of the quality improvement and risk management programs that should be reported
■ Discuss what form they should be reported in
■ Describe how and to whom they should be reported.

The activities of the clinical governance program and the results achieved should be reported both up the
organisation through the committee structure and to the main clinical groups and senior management to
the board of management, and down the organisation to all clinical staff "One of the dHfictAiee
in the health service. in addressing quality
The board of management should be confident that the program's management issues
with the same interest
activities are being undertaken as planned, that clinical care is provided
and rigour as financial
in accordance with best practice, and that the risk for patients receiving management is the lack
care is appropriate. of robust universal
Some health services have developed scorecards to evaluate and report systems for collecting
their performance in quality and safety. These scorecards are usually and reporting on key
areas of quality and risk
brief, and provide a rapid way for board members, senior executives, staff
. . . All health services
members and the public to obtain an overview of the service's should agree upon a
performance. Much thought and effort is required to determine the minimum dataset of
measures to be used in the scorecard. The measures should be valid high-risk issues to be
(including the evidence supporting the measure and that improving regularly reported
throughout the
performance will improve patient outcomes), reliable, useful, feasible,
organization.'
clinically significant, and important to those required to improve quality
Balding C Health Inf
and safely at the frontline of the service. Measures will often require
Manag 2004: 33 137-139
prioritisation because of the many available. 1
Adequate resources should be allocated by the health service to collect,
analyse and report the data. Using multiple measures to assess
performance often requires the development of standardised data
collection tools, training staff to use them, regular review of the quality of
data collection, and reporting and minimisation of missing data. 1 Many
aspects of best practice and data collection and analysis (discussed in
Chapter 2) are also relevant to measures used in quality and safety
scorecards.
The quality improvement program should repon the number of medical
staff who have been appointed over the period of the report, the clinical
areas in which they have been granted privileges to practice, and any
changes in clinical privileges made for existing medical staff. The results of
clinical audits, both routine and focused, that have been undertaken
should also be reported. Any gaps between the clinical care provided and
best practice should be highlighted, together with the actions taken, or
planned to be taken, to close the gaps. Routine clinical audits may be
undertaken in high- volume clinical areas, and a balanced set of measures
should be used to concisely report on the health service's performance.
Enhancing patient care

The risk management program should report the number of medical records that screened positive for
each screening criterion and required medical review, and the number of adverse events detected. The
number of clinical incident reports made by clinical stafT and the number containing adverse events
should also be included. A risk register detailing the extreme and high clinical risks currently present in
the service should be pan of the report. The report should also include the actions proposed to reduce or
eliminate these risks, the staff responsible Tor these actions being taken, and their completion dates.
According to the "rule of rescue", humans respond to stories more than they respond to statistics.
Therefore, the statistics presented should also be supported by carefully chosen case histories, which
help focus the health service and keep patient safety "front and center". 2
Equally imponant to reporting up the organisation is reporting the activities and results of the clinical
governance program to all the clinical staff in the health service. In our experience, producing an
attractive one-page newsletter each month can best disseminate this information. The contents of the
newsletter cover the major activities of the program and the actions taken that to improve quality and
safety. The layout is colourful and graphs are used to display results and trends. To ensure all clinical
staff receive the newsletter, it is attached to their pay slips and is displayed on all hospital noticeboards.

Important points
m Quality improvement and risk management activities should be reported up the organisation to the governing body and down the
organisation to those at the clinician-patient interface.
■ A balanced set of measures should be reported, and could include: the results of focus and whole of practice audits and the gaps identified and
actions taken; clinical indicators that have been measured and benchmarked; the risk register detailing significant adverse events and risks detected
by the risk management program and the actions taken; medical staff appointments and changes in clinical privileges.
■ Education about the program can be undertaken at staff orientation and by distributing regular newsletters.

References
1 Pronovost PJ. Berenholtz SM. Needham DM. A Iramework for health care organizations to develop and evaluate a safety scorecard. JAMA 2007; 298:
2063-2065.
2 James B. Brent James on reducing harm to patients and improving quality. Interview by Susan V White. J Healthc Qua! 2007; 29: 35-44. °

9 A quality improvement and patient safety culture

This chapter will:
m Discuss the components of an informed culture
■ Outline the types of safety cultures
■ Describe the spectrum of safety cultures, from vulnerable Jo resilient
■ Present a checklist of the characteristics of a high-reliability health service
■ Discuss how an organisation's patient safety culture can be measured, including the instruments available.

Hospital services are facing ever-increasing demands, and financial pressure limits the services
provided. Each day, many pressing "bushfires" occupy senior management. Quality and safety
 9 A quality improvement and patent safety a«ure
issues may be pushed into the background while "more pressing" issues are dealt with. Safety
issues may therefore not receive appropriate attention. Eventually, action is taken, often only after a
"disaster" occurs and an external agency, such as the coroner, obtains the attention of relevant
individuals in the organisation. 1 The development of a "safety culture" may detect and reduce risks
in a timely manner and prevent adverse events occurring before action needs to be taken.
Culture is "the way we do things around here". 2 An organisation's culture influences patient
safety.3 For example, can concerns about safely be raised with senior management by frontline
stafT? Are these concerns adequately investigated, and is appropriate action taken in a timely
manner? Safety culture is one component of the wider construct of organisational culture. Culture
is something an organisation is (the beliefs, altitudes and values of its members) and has (the
structures, practices, controls and policies). 4 5 Culture evolves over time according to local
conditions, past events and leadership. It is easier for management to change structures and
policies than beliefs and attitudes.
The concept of a safety culture, although often poorly understood, is strongly desired by many
organisations. There is increasing recognition that differences in safety cultures is a significant
factor in the variation in accident rates between organisations, and that changing organisational
culture is important in improving patient safety. 6 From 2008, the Joint Commission on
Accreditation of Healthcare Organizations in the United States will require health services to
undertake an annual assessment of their organisational culture as part of the leadership standard
for accreditation of services. A safely culture can be "socially engineered" by gradually and
persistently bringing together and applying the essential components a safety culture, which are
ways of doing, thinking and managing.1
A safely culture reflects the "ability of individuals or organizations to deal with risks and hazards
so as to avoid damage or losses and yet still achieve their goals". 7 The Insiiiuie of Medicine in ihe US
stated that "the biggest challenge to moving toward a safer health system is changing the culture
from one of blaming individuals for errors to one in which errors are treated not as personal failures
but as opportunities to improve the system and prevent harm".8
In a strong culture, goals and values are shared across ihe entire organisation guiding people in
how they should act in particular situations. The safely culture drives the organisation to the
maximal level of safety

9.1 Types of organisational culture

Westrum proposed three organisational cultures according lo how an organisation processed
information, including how much information (lows, its relevance, timeliness and appropriateness °
He

1»1
 Enhancing patient care

Boa 9.1 lypes of organisational culture*

■ Preoccupation with personal power and needs.

■ Information is a personal resource to be withheld or provided to advance particular groups
within the organisation.
■ Failure is covered up and individuals blamed.
■ New ideas are actively crushed.
■ Sharing and learning with others is discouraged.
■ Most immature state of cultural maturity.

■ Preoccupation with rules, positions and departmental turf.

■ Communication may stop at departmental boundaries.
■ Systems have been developed for communication.
■ Information is collected but may be ignored.
■ New ideas are seen as creating problems.
■ Learning and sharing are tolerated but not encouraged.
Generative
■ Preoccupation with the organisation's mission.
■ Information is actively sought.
■ Good information flow assists creativity, problem solving and innovation, and builds bridges and
cooperation between departments.
■ New ideas are encouraged and often implemented.
■ Failure prompts inquiry.
■ Most advanced state of cultural maturity.

described three responses (thoughts, emotions and actions) to opportunities for innovation, problems
and danger signals. These responses were shaped by management's preoccupations, which are
absorbed by the workforce and shape the organisation's culture (Box 9.1). 9
Although organisational culture is much wider than safety case studies and anecdotal evidence suggest,
good information flow in organisations is strongly related to safety. 9 In generative cultures, safety
problems are more likely to be raised with inquiry into the problem and repair of latent sources of error.
What has been leamt about the problem and its causes and solutions are then communicated to other
parts of the organisation.

9.2 Components of an informed culture

The most important component of a safety culture is that it is an informed culture (ie, there is an
"effective safety information system").1 The system collects data about incidents and near misses,
analyses the data, and regularly communicates appropriate information throughout the organisation.
Management is therefore aware of the hazards in the organisation and the level of safety.
Four subcomponents interact to create an informed culture: reporting culture, just culture, flexible
culture and learning culture. Importantly, the presence of these four elements in an individual
organisation does not ensure a safety culture exists. Putting them into place is only the first step. These
practices need to function effectively to move to the next stage of a safety culture — being something an

192
 9A
quali
ty
impr
ove
men
t
and
patie
nt
safet
y
cult
ure

organisation is, such as its beliefs and values. Developing a safety culture
is a continuous struggle and joumey, rather than a destination and
endpoint.1

9.2.1 Reporting culture

Those in contact with the hazards facing the organisation at the coalface
repon incidents and near misses. Such reponing requires staff trust,
which is developed by taking non-punitive action rather than blaming
individuals for incidents (unless "unsafe acts" have occurred), and
maintaining confidentiality. Reponing is easy, rapid feedback from
investigations and actions taken are given to the reporter, and aggregate
information about trends and actions are reponed to all staff. 1 There is a
record of appropriate action being taken in response to reports.

9.2.2 Just culture

Most adverse events are the result of system deficiencies. However, at ti
actions taken by staff are clearly unsafe and unacceptable, and
appropriate sanctions need to be applied. A framework could be
developed to help determine culpability, and could consider:
were the actions and consequences intended, was there
unauthorized substance abuse, did the individual knowingly
violate safe operating procedures and were the details of these
procedures available, workable, intelligible and correct, in light of
how events unfolded and were perceived by those in real time, is it
likely that another individual with comparable qualifications and
experience, would have behaved any differently, does the
individual involved have a history of unsafe acts?1
Staff usually know which individuals continually break rules. If
management and staff are aware of such behaviour and appropriate
action is not taken, morale may be affected and the effectiveness of the
organisations disciplinary system put in doubt. If appropriate disciplinary
action is taken, it demonstrates to staff that the system works, and
reinforces the boundary between acceptable and unacceptable behaviour.
Such action makes the organisation safer and increases the belief by stafl
that they are working in a just culture.1

9.2.3 Flexible culture

A flexible culture adapts effectively when demands change. If a dangerous
situation occurs, the organisation has the flexibility lo give control to
experts al ihe coalface (thereby respecting their skills and experience) and
revert to the normal line of reporting after the emergency has passed.
Such flexibility requires a disciplined approach to working and
frontline leaders who have been carefully selected, extensively trained,
and who are able lo work safely and effectively in infrequent and
dangerous situations for which standard operating procedures cannot or organisations mm I
thatleedtoreHstMyle foe
may not be developed. Such staff require detailed knowledge of the
suppression and
activities undertaken ai the frontline and the conditions thai may be containment, in short, 0te
management of fatures...
those organizations
committed to safety at foe
highest level — high
reliability organizations —
adopt « a special
approach to its pursuit.'
ScHukna
n PR Qual Sal
Health Cm* 2004;
13: *3»*44e
 9A
quali
ty
impr
ove
men
t
and
patie
nt
safet
y
cult
ure

encountered there, have a high degree of professionalism, training and

experience, and the respect of management and frontline staff. 1
 Enhancing patient care

•J.4 Learning cufture

A Warning culture analyses safety information appropriately and takes action to make* the required
changes. Learning organisations maintain flexibility in an environment of rapid change and
uncertainty, and innovate to be able to compete. 1 Key cultural features of a learning organisation
are:6,1011
■ Open systems thinking — to see the interconnection of activities through knowledge sharing,
feedback and communication.
■ Improving individual capabilities — for an organisation to achieve excellence, individuals
constantly improve their proficiencies by recognising their knowledge, experience and results.
Members of the organisation are encouraged, and personal and professional development are valued.
■ Team teaming — develop team, rather than professional or individual, learning.
■ Updating mental models — finding new ways of doing things.
■ A cohesive vision — a shared understanding of a coherent vision.
■ Tolerance of mistakes — to encourage innovation and taking measured risks.
■ Trust between managers, leaders and staff.
Key structural features of learning organisations are: 6
■ "Flauer"organisational management structures (fewer levels of management).
■ Teamwork structures.
■ Incentives and rewards for learning.
■ Belter information and communication networks.
■ Research and development programs and budgets. There are three levels of learning: 11
■ Single loop — detection and correction of error (eg, clinical audit).
■ Double loop — redefining the organisation's goals, norms, policies and procedures (eg, system
redesign).
■ Meta-leaming — building on experiences of learning to develop new learning strategies. "Implicitly,
a practice that leams is also going to be safer and be able to improve continuously." 6
However, most publications about organisational learning in health care discuss this theory and its
application and barriers. Little has been published about its full implementation in a health service
or evidence of its effectiveness.6

9.3 Types of safety cultures

Additional frameworks have been developed to assess the maturity of the safety culture within an
organisation. Parker and Hudson extended Westrum's model of organisational cultures to five levels
and applied it to safety culture (Box 9.2). 12

Bex 9.2 Levels of organisational safety culture12

Level 1: pathological — why do we need to waste our time on risk management and salety issues?
Level 2: reactive — we take risk seriously and do something every lime we have an incident.
Level 3: caiculative — we have systems in place to manage all likely risks.
Level 4: proactive — we are always on the alert, thinking about risks that might emerge.
Level 5: generative — risk management is an integral part of everything we do.

194
 9 A quality improvement and
patient safety cufc*e

These levels can be applied across organisational dimensions, such as commitment to patient
safety, perceptions of causes of incidents and their reponing, investigating incidents, learning
following an incident, communication within the workplace, staff management, staff education and
training about risk management, and teamwork. Organisations can develop criteria for each
dimension that are appropriate for their local needs and conditions. The service can then compare
safety behaviour in the workplace with the criteria for the five levels of safety across each dimension.
This comparison should highlight deficiencies, stimulate discussion, identify areas for improvement,
allow interventions to be evaluated and provide appropriate benchmarks to assess safety
performance in the health service. Safety culture assessment can be a springboard for directing
action to raise the safety culture to the next level.13

9.4 The safety spectrum in health services — from resilience to

vulnerability *
There is a spectrum of safety in health services — some are resilient to adverse events, whereas
others are quite vulnerable. 14 Most lie in the middle of the spectrum; however, they can move in
either direction over lime. Organisations can use "navigational aids" to determine their position on
the spectrum. These aids are reactive (information about adverse events and near misses) and
proactive (safety indicators to identify latent conditions that increase the probability of an adverse
event or near miss occurring), and provide information about the level of a health service's resilience
to error.14

9.4. I Why some health services are more vulnerable to operational hazards than othen
Previously, in most health services, clinical activity was reviewed in mortality and morbidity meetings
and grand rounds, which focused on the medical care patients received, and the performance of
individual doctors and the errors they made. Such review was made with the benefit of hindsight, and
there was a marked lack of systems, multidisciplinary or human factors analysis. 15 Systems with
weaknesses resulting in adverse events remained weak.
In vulnerable health services, where the systems created and used are more likely to result in an
adverse event or near miss, "a recurrent cluster" of three "organisational pathologies" are present:
blaming those at the frontline of the organisation (blame), denying that the errors occurring are the
result of weaknesses in workplace systems (denial) and pursuit of financial and production goals
without attention to safety issues (pursuit of the wrong kind of excellence). No organisation in a
complex industry is completely free of these characteristics, and they need to be recognised and
remedial action taken.16
Until recently, there was little impetus for the culture in vulnerable health services to change.
Unlike aviation and nuclear power, where a single error can have catastrophic effects, errors in
health care, although they can cause the death of a patient, usually only afTect one patient at a time.
However, in contrast with other complex industries, errors are frequent in health care.

9.4.2 The important characteristics of resilient or high-reliability health services

More conditions occur than can be anticipated or planned for when designing health care delivery
systems, creating a large variety of novel problems. The components of systems can interact
unpredictably, resulting in unexpected events that place a high premium on the experience and skill
of clinicians to provide a wide range of services. Demand for acute health care is often unpredictable
and uncontrollable, further challenging clinicians to be able to respond appropriately in real time The
need for improvisation may be high at times, cutting across departments and hierarchies and
requiring "bending* of procedures to provide appropriate care in a timely manner. There is also a high

1«S
Enhancing patient care

potential for serious adverse events to occur. Professionals have a critical role in making decisions in
high-reliability organisations, which focus on real-time challenges using formal principles,
experiential knowledge, pattern recognition, intuition and improvisation. 17
A safety culture can provide a defence against inherent weaknesses in the organisation's systems. A
poor safety culture may mean that workers do not understand or fear the operational hazards. Such
ignorance may encourage non-compliance with safe operating procedures and the perpetuation of
longstanding, recognised gaps in defences, which may be worked around rather than closed, as
concern about them is not great enough for corrective action to be taken. Recurrent patterns of error
may then occur, although major adverse events may only occur infrequently, resulting in complacency
and making a change in culture difficult. As the resources available for quality improvement are
limited, a balance between safety and cost is necessary to keep risks as low as practical while
ensuring that the organisation remains in business. The practical components of an informed culture
are known and can be engineered and a safe culture created (Checklist 9.1). 18

Oieddbt 9.1 Characteristics of a high-raliabllity haalth service

Actions Yes No W1P N/A

1 Mian safety is placed above other short-term (usually financial) priorities. 19

2 High-reliability professionals, who can combine principles and experience, are recruited and trained.' 7

3 Staff education about quality and safety is important, commences during their orientation, and is ongoing. They
responsfcility of all staff, not just management. Because of the high
are taught that quality and safety are the
turnover of junior medcal staff between hospitals and clinical areas within hospitals, education about quality and
patient safety is a standing component of their education program. Each staff member regularly receives easily
understandable information, including results of the quality improvement and patient safety activities undertaken in
the health service.
4 Adequate resources are provided by senior management for effective quality improvement and risk management
activities. These resources may include the employment of a clinical risk manager and other staff, secretarial
assistance, computers, audit tools, electronic incident reporting systems and trigger tools.
5 The dinical risk manager and quality manager report directly to the chief executive or a clinical (^visional head,
are appropriately trained in risk management, and are given appropriate status and remuneration.
6 Staff are genuinely encouraged to raise quality and safety issues in real time. They can raise issues directly with
the chief executive, executive staff, the clinical risk manager, the quality manager, the occupational health and safety
officer, or the appropriate department head.
7 Quality or safety issues raised by staff are responded to quickly.
8 Board of management and senior management meetings routinely discuss quality and safety issues together with
financial and business issues. Safety is put above finance on board meeting agendas. 30 Board of management
members and senior executives are aware of all major adverse events or near misses that occur and the major risks
faced by the health service. They are regularly updated about the progress being made to reduce these risks. Some
board members are allocated specific responsibility for quality and safety. 1

Checklist 9 1 (continued on next page)

 9 A quality improvement and
patient safety culture

Checklist 9.1 (continued)

ACTIONS YM NO IMP MM
9 Quality improvement and pstiant safety are standng agenda Heme of meetings of lay professional groups within the health
service. The dinical risk manager, the quality manager or the executive responsible for risk management attend these meetings and
report on the
activities of these programs. *
10 Responsibility for safety cannot be delegated from the chief executive. 20
11 Patients have an integral role in safety efforts, and patient representatives are involved in quality improvement and risk management project team
meetings.20
12 Multidisciplinary teams are empowered to analyse gaps, adverse events and risks, and plan and implement changes to close gaps and reduce risks. 20
13 Quality improvement and patient safety are key components of the health serviced « strategic plan, together with finance, the range of dinical services
provided and the organisation's major staffing, building and equipment requirements. The health service displays its strategic plan throughout its facilities
and features it prominently in its annual report. The quality improvement and safety objectives and activities of each department and division are a key
component of annual business plans and reports. The indusion of quality and safety in the health service's strategic plan sends a strong message to staff
and the community of the strong commitment to these areas at senior levels in the organisation. Regular reports of the health service's progress towards
the strategic plan objectives are made to the board of management, key professional groups and staff.
14 Adverse events and near misses are openly discussed in regular multidisciplinary meetings to learn from them and prevent their recurrence.
15 In analysing adverse events and near misses, major stakeholders and frontline staff are actively involved in analysing the event and determining
appropriate action to reduce the probability of recurrence. Senior management is actively involved, and assists if the event crosses departmental
boundaries. Senior management approves proposed actions to reduce risks, the timetable for implementation, delegation of tasks, and any other substantial
requirements, such as additional staff, building alterations or new equipment. Where safe to do so, senior management allows frontline staff to determine
and make changes they feel are appropriate to adequately address the event
16 The health service uses proactive and reactive methods to increase resilience to error. The organisation proactively
looks for weaknesses in the systems it uses to provide care. Latent factors or conditions of work, such as staff workloads,
shift length, skill levels and the degree of supervision of junior staff members, are assessed when designing new health
care delivery systems or assessing existing systems. The health service uses multiple methods to detect adverse events
and near misses (eg, medical record review, clinical incident reporting) to obtain different perspectives of events
occurring in the organisation.
17 There is a positive incident reporting culture in the organisation. Staff dearly understand which incidents they should report and appreciate that reporting
can increase quality and safety Reporting rates are high for doctors and nurses. Staff receive timely feedback of investigation results of incidents they have
reported. Recommendations for action are made in response to incident reports, and the timelines for completing these actions are consistently met
18 Clinical audits are undertaken These include continuous whole-of-practice audits and, where appropriate, focused audits following adverse
events and near misses
19 Actions taken in the health service to improve quality and safety are system improvements, and are not punitive against individual staff members unless
there is evidence that an "unsafe act" has occurred. Assessments are made of unsafe acts, and management and appropriate peers determine the
disciplinary action to be taken. I A process is available for individuals if they
wish to appeal against the judgment.1 I
N/A = not applicable WIP = work in piogress
Enhancing patient care

9.5 Measuring the patient safety culture

It is important to understand the patient safety culture in an individual organisation. This can be
achieved using individual interviews or focus groups, but they are time-consuming and resource-
intensive. Self-administered surveys using standardised questions designed to measure the
components of a safety culture that can be completed anonymously by health service staff are more
efficient.21 An assessment of nine surveys showed considerable variation. 22 The surveys can measure
the patient safety climate throughout the health service or in specific areas. Seven of the surveys are
completed by individuals and two by teams. All the surveys use a 5-point Likert scale and the number
of items to be completed varies from 19 to 194. Two surveys measure the degree to which actions
have taken place in response to safety concerns that have been raised. Five common dimensions in
these surveys are: leadership, policies and procedures, staffing, communication and reporting. The
quantity and quality of psychometric testing varies considerably between survey tools. The survey
results are used to make comparisons within and between health services, and to compare health
services with other complex industries. 22
Each survey has different domains, limited validity and reliability, and variable response rates. 22
The measurement of culture is evolving, with disagreement about measurement (which domains are
important), scoring and how to improve culture. Also, the causes of variation in culture are not
understood.2 Two examples of surveys are given below.

9.5.I Safety Attitudes Questionnaire

"The most commonly used and rigorously validated is the Safety Attitudes Questionnaire." 23 The
questionnaire uses a double-sided single page, has 60 items and "measures six domains: safety
climate, teamwork climate, perceptions of management, stress recognition, job satisfaction, and
working conditions". The survey takes about 10 to 15 minutes to complete. There is also an open-
ended question: "What are your top three recommendations for improving patient care in this clinical
area?" Versions of the questionnaire, together with a user's manual and benchmarking data, can be
downloaded at the University of Texas Center of Excellence for Patient Safety Research and Practice
website (www.uth.tmc.edu/schools/med/imed/patient_safety/survey6ztools.htm).24
The Safety Attitude Questionnaire is the only survey that has been used to explore the relationship
between the safety climate score and patient safety outcomes. 22 Interventions have been shown to
improve the safety and teamwork climate, although this may be associated with deterioration in
another domain.2 Such tools can measure teamwork and communication, and there is evidence that
these measures relate to safety outcomes. 23 Valid comparisons can be undertaken in individual
clinical areas over time, between clinical areas in an individual health service, between health
services, and with other complex industries. 24
Some authors have recommended that health services use the Safety Attitudes Questionnaire
annually to measure the safety culture across all work units throughout their organisation. 2 The
Questionnaire has received a greater than 80% response rate, which aids interpretation. After the
survey, the results should be presented to senior management and all staff. Specific interventions,
focusing on individual work units, should be undertaken to improve culture, although there is
currently limited evidence to support the effectiveness of such interventions. Most importantly, the
variability of culture is greater within than between health services. 2 Although caution is being
recommended in the interpretation of safety climate surveys, the greatest challenge will be changing
the organisational climate after it has been measured 22
 9 A quality improvement and
patient safety culture

More recently, Pronovost and colleagues developed a safety framework that incorporates the
Solely Attitudes Questionnaire and is used in over 100 intensive care units in Michigan in the US.
It compiurt four domains:3
1. How often do we harm patients? — measurements of specific adverse event rates (eg, catheter-
related bacteraemia).
2. How often do we provide the interventions that patients should receive? — measurement of the
proportion of patients who receive a specific evidence-based intervention (eg, proportion of
ventilated patients receiving elevation of the head of bed and prophylaxis for peptic ulcers and
deep vein thrombosis).
3. How do we know we leamt from defects? — proportion of months each patient care area leamt from
mistakes (eg, proportion of months in which at least one sentinel event was reviewed and a policy
was created/revised and/or staff awareness or use of that policy was measured).
4. How well have we created a culture of safety? — annual assessment of safety culture in each unit
using the Safety Attitudes Questionnnaire (eg, percentage of areas in health service where 80% of
staff report a positive safety and teamwork culture).
The first two domains can be measured as valid rates; however, the second two, although essential
for safety, cannot.3,25

9.5.2 Manchester Patient Safety Framework26

The Manchester Patient Safety Framework has also been recently developed using the earlier work of
Westrum and Reason. Acknowledging the complexity of and difficulty in understanding the concept of
a safety culture, a framework was developed that aimed to deconstruct safety culture and make it
meaningful to managers and frontline staff by increasing their understanding and stimulating
discussion for improvement. The Framework uses nine dimensions to express safety culture (Box
9.3), which are described in a straightforward manner at each of the five levels of organisational
maturity.
The instrument is completed individually by
staff, and the level of maturity at which the
Box 9.3 Dimensions of the Manchester
health service is functioning is determined. Patient Safety Framework — acuta
Teams within the health service can then use the ■ Commitment to overall continuous improvement
results in self- reflection and education activities ■ Priority given to safety
to help determine the appropriate action to take ■ System errors and individual responsibility a
to improve the safely culture. The instrument Recording incidents and best practice
can then be used lo measure progress. a Evaluating incidents and best practice a
Frameworks have been developed for primary Learning and effecting change
care, acute care, ambulance and mental health. ■ Communication about safety issues
A facilitator's guide and the frameworks and a Personal management and safety issues a
evaluation sheets are available at Staff education and training.
www.npsa.nhs.uk/nrls/improvingpatientsafety/
humanfactors/mapsaf/.
■ Cuftixe is something an organisation is (eg, beliefs, attitudes) and has (eg, structures, practices).
Management has greater control to change structures and practices than culture.
■ The most important component of a safety culture is that it is informed, which has four
subcomponents — reporting, just, flexible and learning culture.
■ The presence of these components does not ensure a safety culture — they must also function
effectively.
■ There is a spectrum of safety in health services — from resilient to vulnerable. Health services can
determine their position on the spectrum using reactive aids (information about adverse events
and near misses) and proactive aids (indicators to identify latent factors).
■ The characteristics of high-reliability health services have been described, and a checklist is available.
■ The patient safety culture can be measured. The most commonly used and best evaluated tool is
the Safety Attitudes Questionnaire. New frameworks such as the Manchester Patient Safety
Framework have been developed, but they have not extensively evaluated.
 9 A quality improvement and patient safety
culture

References
1 Reason J. Managing the risks of organizational accidents. Aidershcrt. UK: Ashgate. 1997.
2 Pronovost P. Sexton B. Assessing safety culture: guidelines and recommendations. QuaI Saf Health Care
2006. 14: 231-233.
3 Pronovost PJ, Berenholtz SM. Goeschel CA, et al. Creating high reliability in health care
organizations Health Sen Res 2006; 41: 1599-1617.
4 Bate P. The impact of organisational culture on approadhes to organizational problem-solving In:
Salaman G. editor. Human resource strategies. London: Sage, 1992.
5 Thompson N. Stradling S, Murphy M. O'Neill P. Stress and organisational structure. BrJSoc Work 1996;
28:647-
665.
6 Dunbar J, Reddy P. McAvoy B, et al. The contribution of approaches to organisational change in
optimising tfw primary care workforce. Canberra: Australian Primary Health Care Research Institute,
2007.
7 Reason J. Safety paradoxes and safety culture. Injury Control Saf Prevent 2000; 7: 3-14.
8 Institute of Medicine. Crossing the quality chasm: a new health system for the 2fst century.
Washington. DC: National Academy Press, 2001.
9 Westrum R. A typology of organizational cultures. Oual Saf Health Care 2004; 13 (Suppl II): 022-027.
10 Senge PM. The fifth discipline: the art and practice of the learning organisation. New Yortc
Currency Doubleday. 1994.
11 Davies HTO, Nutley SM. Developing learning organisations in the new NHS. BMJ 2000: 320: 998-1001.
12 Parker D, Hudson PT. HSE: understanding your culture. Shell International Exploration and
Production. EP 2001-
5124,2001.
13 Nieva VF, Sorra J. Safety culture assessment: a tool for improving patient safety in healthcare
organizations. Oual Saf Health Care 2003; 12: ii17-ii23.
14 Carthey J. de Leval MR. Reason JT. Institutional resilience in healthcare systems. Oual Health Care
2001:10:29-
32.
15 Barach P. Small SD. Reporting and preventing medical mishaps: lessons from non-medical near
miss leporting systems. BMJ 2000; 320: 759-763.
16 Reason JT, Carthey J, de Leval MR. Diagnosing "vulnerable system syndrome*: an essential
prerequisite to effective risk management. Oual Saf Health Care 2001; 10: ii21-ii25.
 9 A quality improvement and patient safety
culture

17 Schulman PR. General attributes of safe organisations. Oual Saf Health Care 2004; 13: ii39-ii44.
18 Reason J. Achieving a safe culture: theory and practice. Work Stress 1998; 12:293-306.
19 Stevens P, Matlow A. Laxer RM. Blueprint for patient safety. Pediatr Clin N Am 2006:53:1253-1267.
20 Hastings G. Eight steps to better patient safety. Health Serv J 2006; 116:28-29
21 Kho ME. Carbone JM, Lucas J. Cook DJ. Safety Climate Survey: reliability of results from a
multicenter ICU survey Qual Sat Health Care 2005; 14:273-278.
22 Colla JB, Bracken AC, Kinney LM. Weeks WB. Measuring patient safety climate: a review of
surveys Oual Saf Health Care 2005; 14. 364-366.
23 Berenholtz SM, Pronovost PJ. Monitoring patient safety. Crit Care Clin 2007; 23: 659-673.
24 Sexton JB, Helmreich RL, Neilands TB, et al. The Safety Attitudes Questionnaire: pyschometric
properties, benchmarking data and emerging research. BMC Health Serv Res 2006; 6: 44.
25 Pronovost P. Holzmueller CG, Needham DM, et al. How will we know patients are safer? An
organization-wide approach to measuring and improving safety. Crit Care Med 2006; 34: 1988-1995.
26 Kirk S. Parker D. Claridge T. et al. Patient safety culture in primary care: developing a theoretical
framework tor practical use. Oual Saf Health Care 2007; 16: 313-320. □
Enhancing patient care
10 Eighteen practical steps to implementing a clinical
risk management program in a health service
This chapter will:
m Present a comprehensive and practical step-by-step approach and a checklist for imptemendng a clinical risk
management program in a health service.

10.1 Where to start

The most difficult task in commencing a clinical risk management program is to determine precisely how and where
to begin. In our experience, with many quality improvement activities, there is the potential for much activity to take
place. Undertaking such activity may make us feel good, as if we are achieving something worthwhile. However, when
 9 A quality improvement and patient safety
culture

we examine the effects of this activity on the quality of patient care provided, it is often difficult to demonstrate
measurable improvement.
The primary aim of implementing and maintaining a clinical risk management program in a health service is to
reduce the risk to the patients who are receiving care. It is important, therefore, not to fall into the "activity trap" of
undertaking quality improvement and risk management activities to comply superficially with external accreditatiori
and funding requirements. All risk management initiatives undertaken must aim to demonstrate measurable
improvement in the safety of clinical care provided to patients.
Modifying this step-by-step guide to meet local needs and conditions will significantly increase the likelihood of
successfully implementing a clinical risk management program. Although this guide may- look daunting, it is a series
of signposts. It is not necessary for individual health services to undertake all the steps outlined in this chapter in
their entirety or at the same time. As with the implementation cycle discussed previously, implementation of a
clinical risk management program in an individual health service should be undertaken in small, manageable steps
using signposts that are appropriate for the local needs and conditions unique to each organisation.

10.1.1 Step I: develop an Information sheet to explain the purpose and methods used ki a clinical risk management
program
Explaining the concepts of a clinical risk management program to health service staff and board of management
members with little knowledge in this area can be difficult. We have prepared a summary (Appendix 2), which can be
modified to suit local conditions. The summary can be used to help clinicians and management understand what
clinical risk management is, and provide an overview of how such a program is developed, implemented, run and
maintained.

10.1.2 Step 2: find clinical champions

Medical staff support is critical to the success of any clinical risk management program Some doctors may be
apprehensive about an activity that closely examines the mcdical care they provide Despite an initial negative
response from some doctors, mcdical staff should be closely involved with the development ofthe program, and all
doctors working in the health service should be kept fully informed of the program's progress.
 A cnttcal early step in the development of any clinical risk management is to find a clinical
champion. In our experience, every health service will have some medical stafT who are passionate
Enhancing patient care
about providing high-quality, safe medical care to their patients. Clinician champions should
preferably have some of the following characteristics:
■ a high profile in the health service
■ a track record of introducing innovations in their clinical areas
■ have helped write clinical policies and protocols
■ have the respect of medical and nursing staff.
Some potential clinical champions may be finding it difficult to introduce changes to clinical
practices in their areas of clinical expenise. A clinical risk management program that is effective in
introducing changes in clinical behaviour may help such clinical champions in their quality
improvement and risk management endeavours in their own clinical area. Instead of working in
isolation to achieve change, these clinical champions could use the resources of a more effective
clinical risk management program to achieve quality and safety improvements in clinical care.

10.1.3 Step 3: find an executive champion

For clinical champions to effectively change clinical behaviour, they will frequently require the
assistance of an executive champion with the authority to make decisions about how resources are
allocated and how changes in service delivery systems are to be made in the health service. An
executive champion will know how major decisions are made in the health service and will be able to
guide system changes through the health service bureaucracy. The more "hands on" the executive
champion is, the more effective the project will be in successfully reducing clinical risk. Ideally, the
executive champion should be an active member of the surveillance committee that reviews adverse
events, and of some of the multidisciplinary teams attempting to close gaps in care and prevent
adverse events recurring. Such committee participation will allow the executive champion to facilitate
rapid changes in delivery systems when they are required.
At Wimmera Health Care Group, the executive champion is the director of medical services. In most
health services, directors of medical services deal daily with quality and risk management issues, and
frequendy with coronial and claims management matters. They are appropriately placed to be
significant drivers of clinical risk management programs. Also, medical record screening and review
(and the important role of the health information service in these processes) are cornerstones of the
clinical risk management process, and directors of medical services are usually responsible for the
health information service in most health services' organisational structures. Obviously, it assists the
implementation process of a clinical risk management program if the director of medical services and
the director of nursing services are both strongly committed to the health service providing high-
quality clinical care.

10.1.4 Step 4: obtain chief executive officer support and find champions in the board of
management
Strong support for the clinical risk management program will also be required from the health
service's chief executive officer and board of management members. Most chief executive officers are
extremely busy and. in the current environment, predominantly occupied with ensuring the financial
survival of their health services. Increasingly demanding clinical governance requirements indicate
that chief executive officers are also responsible for the standard of clinical care that is provided
(together with clinical staff and board of management members).
In most health services, the peak quality committee is usually a subcommittee of the board of
management Some board members will be members of the quality committee and are logical choices
as clinical risk management champions at board of management level. Having such champions
should facilitate discussion about issues raised by the clinical risk management program at the
highest levels d the organisation, and aid appropriate and effective action being taken by the health
service to address these issues.
The dedication of senior staff and the board of management to the development and maintenance of
a clinical risk management program can be strengthened by the health service committing itself to
the attainment of effective clinical risk management as one of its major objectives. This objective can
be incorporated into the health service's strategic plan and be included in its quality plan and annual
quality of care report. Also, individual departments can appropriately modify and incorporate this
objective into their annual department business plans and daily work practices.

10.1.5 Step 5: allocate adequate resources

When clinical and executive champions have been identified, sufficient resources should be allocated
to effectively design, implement and run the clinical risk management prograi®. Although staff in all
clinical areas of a health service should participate in quality improvement and risk management
10 Eighteen
programs as pan of their practical
usual activities, stepsbusy
already to implementing a clinical
clinical staff should riskexpected
not be management program
to develop
and maintain extensive clinical risk management programs as pan of their existing duties. Additional
resources are required for clinical risk management activities to be effectively undertaken. The extent
of the additional resources required will depend on the size of the clinical risk management program.
Even small programs will require additional resources to detect and review adverse events and
coordinate the actions designed to prevent the recurrence of particular adverse events. Redesigning
complex health care systems and successfully implementing and maintaining such system changes is
labour-intensive. To burden existing staff with yet another significant duty without providing
adequate resources will limit the effectiveness of the program and may threaten its viability.

10.1.6 Step 6: appoint a clinical risk manager

Medium and large health services should consider appointing a full-time clinical risk manager to be
responsible for coordinating the development, introduction and maintenance of the clinical risk
management program. At Wimmera Health Care Group, we appointed an experienced senior nursing
staff member to the position of clinical risk manager. Ideally, the clinical risk manager should have
considerable clinical experience and the respect of medical and nursing staff. The clinical risk
manager will require a high degree of self-motivation, have good written and verbal communication
skills and be able to work effectively as pan of a multidisciplinary team. Some administrative
experience may also be an advantage.
At the Wimmera Health Care Group, the clinical risk manager collects information about adverse
events delected using the various detection methods described in Chapter 4. Serious events are
reported immediately lo the director of medical services and/or the director of nursing services for
analysis and action. All events, whether serious or less serious, are analysed and investigated, and
the clinical risk manager coordinates implementation of appropriate clinical risk reduction strategies.
The clinical risk manager compiles data from all events delected and writes monthly aggregate reports
for the director of medical services, director of nursing services, the surveillance committee and the
quality committee The clinical risk manager should be an active member of both these imponant
committees. The clinical risk manager works closely with:
■ the health service's quality manager — to ensure that the clinical risk management program is fully
integrated into the health service's quality improvement and risk management activities: and
■ the complaints liaison officer — to allow information about adverse events described in patient
complaints to be analysed and acted upon in the same manner as adverse events detected using
other methods.
Al Wimmera Health Care Group, the clinical risk manager reports directly to the director of medical
services. The clinical risk manager liaises with medical, nursing, paramedical and administrative staff
to promote organisation-wide awareness and ownership of clinical risk management activities and
programs.
Although the board of management and senior executives are responsible for providing overall
direction to the clinical risk management program and promoting cultural change, the clinical risk
manager is responsible for the daily operation of the program.

19.1.7 Step 7: ensure the health service has adequate clinical Information systems
Adequate resources are also required to ensure that appropriate clinical information systems are in
place and operating effectively. A high level of commitment to the clinical risk management program
is required from the organisation's health information services and information technology
department, whose role is to establish and maintain information systems that can consistently
provide high-quality clinical information.
Health information services staff screen inpatient medical records lo determine which ones require
medical review, locate records to enable clinical audits lo be undertaken, and enter data from adverse
event analysis forms. Using software specifically written by the information technology department,
health information services staff compile reports of ihe medical record screening process and adverse
event detection rates.
Health information managers are also responsible for extracting data from administrative dataseis
to assist the performance and analysis of specific clinical reviews. Furthermore, these managers help
all clinicians maintain high-quality documentation in the medical record. Medical records are usually
a valuable source of clinical information. If the medical record documentation of clinical care is poor,
opportunities to review and improve this care will be diminished.

10.1.8 Step 8: obtain statutory immunity for the activity of the clinical risk management
program
 Before implementing a clinical risk management program, health services should obtain statutory
immunity for their board of management quality committee and its subcommittees, including the
Enhancing patient care
clinical risk management program surveillance committee. Every health service should ensure that all
quality improvement and risk management activities involving confidential discussions about patient
medical records and documentation identifying the health service and its staff are covered by
statutory immunity. All participants in these activities need to be aware that the program is covered
by statutory immunity, and the implications of this cover.
Granting statutory immunity prevents program participants from being compelled to divulge any
information derived from any confidential discussions that were part of approved quality activities,
and that identifies any individual, in a legal court or tribunal against a medical practitioner or health
service.
In Victoria, statutory immunity is granted under Section 139 of the Health Services Act 1988 (Vic). The
Australian Government grants statutory immunity under Part VC of the Health Insurance Act 1973
(Cwlth). There are some differences between the two acts, so health care organisations should contact
the relevant government service to decide which is the most appropriate act for their needs.

10.1.9 Step 9: educate staff about the principles of clinical risk management and the activities of
the health service's clinical risk management program
It is essential to educate health service staff about the principles of clinical risk management and the
details of the program operating in their health service. All staff members should participate in the
program. It should not be viewed as yet another quality activity that the organisation has to
undertake to meet external quality improvement requirements, and that is someone else's
responsibility in the health service to undenake and complete.
As well as education about the program, there must also be active participation by clinical staff
and management in the activities of the program. Actions speak louder than words, and staff need
to see that their participation in the program results in improved patient care and an enhanced
work environment, and that senior management actively support the program. Detection of adverse
events without appropriate action and improvement will limit the effectiveness of any clinical risk
management program.

10.1.10 Step 10: arrange a program launch to signify the organisation's commitment to the process
When the essential components of the clinical risk management program are in place, the health
service should officially launch the program. This could be a specific event involving a wide cross-
section of staff and board of management members. Such an event could be used to publicly
announce the health service's commitment to the program from the highest levels within the
organisltion.

10.1.11 Step 11: determine the extent of the clinical risk management program
What form should the clinical risk management program take in your health service? Fortunately, the
components of the Wimmera clinical risk management model can be implemented gradually. Each
adverse event detection method can be added individually and sequentially as the clinical risk
management program expands.
In our experience, commencing clinical risk management with a form of medical record review is
the most rational approach. By starting the medical record review component of the program on a
small scale, perhaps using only one or two medical record screening criteria, the detection, analysis
and action components of the program can be correctly and firmly established. We recommend
commencing screening with three criteria: patient death,- transfer from general ward to intensive care
unit and any record recommended for review. Once these processes are in place, the number of
screening criteria being used can be increased.
We use nine criteria lhat can be screened by clerical stafT. We chose these criteria from a larger list
used in the Medical Management Analysis' and the Harvard Medical Practice Study. 2 Some oT the
other criteria require some form of clinical assessment of the medical record to determine whether it
screens positive.
Either health information services or nursing staff can undertake the screening process. We
decided to use health information services staff, as they could screen the medical records as part of
their usual activities. This screening could be done at three points in the department:
■ when the medical record was compiled in the department after discharge
■ when the discharge summary was being typed
■ during coding of the admission.
This triple check resulted in few records being screened incorrectly.
 When the clinical risk management program began and medical records were sent for medical
review, the first major10finding
Eighteen
waspractical
the poorsteps toofimplementing
quality documentation a clinical
in somerisk management
pans of the medicalprogram
record. Medical reviewers commented that on some occasions they were unable to determine whether
an admission contained an adverse event because significant parts of the medical record had not
been completed. The poor quality of some sections of the medical record has been a consistent initial
finding among other health services that have implemented the medical record review component of
the clinical risk management program.
At Wimmera Health Care Group, one of the first major system changes made as a result of regularlv
rcviewing medical records was to replace the usual blank medical record admission sheets with a
structured admission history and examination sheet. The use of this sheet quickly improved the
quality of information recorded in patients' medical record. More recently, multidisciplinary
clinical pathways, including detailed reminders and checklists, have been introduced as the
medical record sheets for patients admitted for one of a list of specific conditions. The use of clinical
pathway medical records has further improved the standard of documentation in the medical
records.
Once a medical record screening and medical review program is established, the surveillance
committee will receive a constant stream of information about adverse events and suboptimal
processes in the health service. In our experience, these events raise more issues requiring review
than can be dealt with at one time. A clinical risk management program can attempt to deal
simultaneously with all the issues raised by the adverse events it detects. If this approach is taken,
however, clinical staff are more likely to become confused and overwhelmed by the number of
changes occurring simultaneously to the already complex systems they are expected to use when
caring for patients. This highlights the need for a risk register to be developed and adverse events to
be prioritised for action.
We wanted to establish a clinical risk management program that was viewed by staff as being an
effective way to improve patient care. As the medical reviewers begin to detect adverse events that
have occurred in a health service, we recommend that the surveillance committee concentrates on
several smaller and somewhat easier health care delivery system weaknesses that these events
identify. Hopefully, by attempting to deal with a few easier and more manageable issues first, a small
number of improvements in the quality of patient care can be achieved rapidly. These improvements
can then be used to demonstrate the effectiveness of the clinical risk management process to staff.
Building on these early successful interventions, more complex problems can be tackled. We have
consistently found that a useful approach to successfully changing complex systems is to "think big
but start small".3

10.1.12 Step 12: promptly address areas and Individuals providing consistently poor-
quality care
After the clinical risk management program has been established and the surveillance committee has
reviewed the details of a number of adverse events, evidence of consistently poor-quality care
provided in some clinical areas of the health service and by some individual clinicians may become
evident. Poor- quality care concentrated in one clinical area of a health service or involving one or a
small number of clinicians can be one of the most difficult issues that a clinical risk management
program and health service have to address.
In our experience, once such issues are detected, if a detailed analysis and appropriate action are
not taken, the adverse events are likely to continue to occur and be detected, and therefore require
discussion at subsequent surveillance committee meetings. One of the strengths of the Wimmera
model is that if appropriate action is not taken and these types of adverse events recur, they will
continue to be found by the detection processes that are in place.
If these problems are ignored and left to recur because the surveillance committee feels they are
too difficult and unpleasant to deal with, the integrity of the whole clinical risk management program
may be called into question by health service staff. If the surveillance committee does not take
appropriate action, eventually the magnitude of the problems will expand and the issues will become
confronting. The first step in examining the issues may be for the surveillance committee to request a
focused audit of clinical activity involving the area or clinician in question. In our experience, as
difficult as these types of situations may be, the earlier the issues are faced directly, the quicker
appropriate corrective action can be taken. Ultimately, clinical risk management programs are about
reducing risk and improving the safety and quality of care provided to patients, not about avoiding
difficult problems because they are unpleasant to tackle.
Fortunately, evidence of repeated poor-quality work by individual clinicians is seen rarely, but
when it does occur, the risk management program usually detects it. On becoming aware of the
problem through the program, the clinical risk manager will refer the matter to the director of
medical services
Enhancing patientIfcare
the problem cannot be resolved appropriately at that level in the organisation,
the matter will need to be referred to the health service's appointments committee.

10.1.13 Step 13: prepare for hard work — the road to quality Improvement b not an earn/ and steady
path
Changing clinical behaviour in a health service can be hard work. The health care delivery systems
within which we provide care to our patients are complex, and positive changes are often difficult to
achieve. Unintended negative events will occur even with the best intentions, and with the most
meticulous planning and implementation of change processes. Clinicians and senior health service
management should acknowledge the limitations the current environment in health services places
on their performance in providing high-quality patient care. ■*
Successfully improving health care delivery systems frequently requires taking two steps forward
and one step back. Some health care professionals will be resistant to the change process. For some
clinicians, the prospect of a significant change in their work environment may engender fear about
losing territory in the health service and control over their workplace. Individual stafT members'
responses to change will vary greatly. If possible, when initiating change to the processes within
systems, concentrate efforts on staff members who are naturally early acceptors of change. Change
may take some lime to be accepted by all the staff who are involved in the system under review —
allow adequate time for it to occur. Persistence is a key factor in successfully changing behaviour.

10.1.14 Step 14: do not expect clinical risk management activities to routinely generate
revenue or reduce costs
As discussed earlier, adverse events can be associated with the underuse, overuse or misuse of
resources. Quality improvement activities that aim to reduce underuse in a health service can
increase costs in the short term. Activities that aim to reduce overuse and misuse can reduce costs.
Although many quality improvement activities in health services will reduce health-related costs to
the community, these savings will not necessarily be translated to improvements in health services'
balance sheets. Further, the money saved by some quality improvement activities can be directed lo
other improvement activities that are needed lo deliver besi-praciice clinical care, but require
additional resources and do not save money.

10.1.15 Step 15: acknowledge and reward high-quality care

Some health care financing systems reward health services financially for providing medical care that
complies with evidence-based processes that have been shown in clinical trials to improve patient
outcomes. Appropriate compliance by health service staff with treatment guidelines based on the best
evidence, and the resultant improvement in the quality of care provided to patients, should be
acknowledged and rewarded, both internally by the health service and externally by the health
service's funding body.

10.1.16 Step 16: stay focused on taking and completing actions that will really make a difference to
the quality of patient care
As discussed earlier, the surveillance committee will receive information about more adverse events
and system weaknesses than it can immediately address. Members of the surveillance committor and
the clinical risk manager may wish to deal with all the issues raised and not prioritise the adverse
events ihev haw cktected for more detailed analysis and action. Health services can be very busy
places, with patient demand for services exceeding those available, resulting in intense financial
pressure. It can be easy for the attention of senior executives and clinicians to be directed daily to the
urgent major service or financial issues facing their health service. They can be sidetracked putting
out "bushfires" and be unable to concentrate fully on ensuring that the health service's clinical risk
management program is effectively reducing the frequency of adverse patient events. Although latent
factors will be present and will contribute to weaknesses in the health service's delivery systems,
there is often a considerable time delay before these factors cause an adverse event. It is easy for
senior staff to defer correcting these latent factors, as it is unlikely that inattention to such factors
will result in a serious adverse event occurring in the short term.
Senior executives in the health service and the clinical risk manager should regularly take time to
reflect on where the health service's resources are best directed to have the most positive impact on
the standard of patient care. To maintain an appropriate focus, they should regularly review the risk
register to ensure that appropriate action is being consistently taken to reduce the frequency of
adverse events with the highest associated level of risk. In addition, the executive champion and
clinical risk manager should regularly review the comprehensive clinical risk management program
checklist provided at the end of this chapter (Checklist 10.1). The checklist details the main signposts
10 Eighteen
to improving the chances practicaldeveloping,
of successfully steps to implementing
implementing aand
clinical risk management
maintaining program
a clinical risk
management program in a health service.
It is not envisaged that individual health services would fulfil all the criteria on the checklist;
however, the more criteria that are modified by health services and programs to meet local needs and
conditions, the greater the probability that individual clinical risk management programs will be
successful. By taking time to refer regularly to the checklist, health services will be able to see where
their existing clinical risk management program may be strengthened.

10.1.17 Step I7: continue to look for Ideos to enhance your health service's clinical risk management
program
Ideas for clinical quality improvement abound, and can be easily found by regularly looking inside
and outside of your own health service. Many quality improvement initiatives and programs are
reported in medical journals. Two excellent websites are the United Slates Agency for Healthcare
Research and Quality's Morbidity and Mortality Rounds on the Web (www.webmm.ahrq.gov), which
presents cases in many clinical specialties with commentary, take-home points and references; and
their Patient Safety Network (www.psnet.ahrq.gov), which provides a comprehensive list of journal,
newspaper and magazine articles. In our experience, more innovative ideas about clinical risk
management and quality improvement are available than can be efficiently and effectively
implemented into an individual health service's clinical risk management program.
Developing and maintaining a clinical risk management program is a continual process of
planning, action and evaluation. Small changes are frequently made as the result of practical
experience gained in implementing and running a clinical risk management program and ideas
described in journals, at conference presentations, and in discussions wiih colleagues.
Informal and formal links for the exchange of ideas can be established with other health services
with clinical risk management programs, university departments with research interests in quality
improvement, and leaders in the quality improvement field. Many health service and medical
conferences now have quality streams, and conferences with quality themes are regularly held
around the world. These conferences give organisations the opportunity to present their work to the
field and discuss ideas with colleagues grappling with the same problems of how to reduce risk and
improve the quality of clinical care they provide.
 10 Eighteen practical steps to implementing a clinical risk
management program

I0.1.I8 Step 18: take a constructive, long-term view to changing culture

The Wimmera model presents a solid framework upon which an effective clinical risk management
program can be built in individual health services. The modules that make up the model can be
selected, modified and implemented at a rate appropriate to the resources and quality improvement
requirements of individual health services.
Whatever quality improvement model a health service chooses to use, it will need to function within
the culture of the organisation. It is difficult to define a health service's culture and positively
influence the culture to increase the probability of successfully implementing a clinical risk
management program. What practical advice can be given to health services about how to create a
culture that encourages improvement in the quality of care it provides?
The following core practices (and those discussed in the previous chapter) can help create a culture
in a health service that supports quality improvement.

10.1.18.1 Communicate openly and honestly

Communication between staff in the health service should be open and honest. Staff at all levels in
the organisation should feel comfortable to express their true feelings about the factors that may have
contributed to adverse events that were detected by the health service's clinical risk management
program, and about any proposed changes to delivery systems in the health service following analysis
of these events. If staff feel that their concerns will not be entirely heard and considered when
decisions about changes to delivery systems are made, or that they will be singled out for saying what
they honestly feel, open and honest communication will be stifled and consequently, some system
weaknesses in the health service's delivery systems will remain unaddressed.

10.1.18.2 Respect decisions made at the coalface

Staff at the coalface often know accurately the changes required to heath care delivery systems in
their work area to produce high-quality patient care. Although health service senior executives and
boards of management have overall responsibility for the quality of care provided by the organisation,
they should be prepared to delegate an appropriate level of responsibility and autonomy for changing
delivery systems to staff at the coalface. The opinions of staff should be routinely sought, valued, and
taken into account to allow decisions about delivery systems to be made regularly from the bottom up
in the health service, rather than the more frequent top-down approach.
Using the Wimmera model, staff at the coalface can undertake appropriate analysis of an adverse
event and the factors in the system's design that have contributed to the event. They can design and
implement subsequent system changes to prevent recurrences. These system changes should not be
overturned by senior management without detailed discussion with the staff involved, and without
careful consideration of the message such action sends to workers about their true level of autonomy
and responsibility in the organisation. Such negative messages from senior management severely
affect the success of a clinical risk management program.

10.1.18.3 Use a just culture approach in dealing with errors

A just culture approach to dealing with errors is a cornerstone in the successful establishment of a
clinical risk management program — staff must feel safe to report errors. A just culture approach,
however, requires a major shift from the current practices in many health services.
If a punitive rather than just culture approach to adverse events is in place, health service staff are
unlikely to inform senior staff that they have made an error. If such communication does not occur,
senior management may not be aware of serious weaknesses in their health serviced service delivery
systems until a serious adverse event occurs that cannot go unreported Without staff feeling
comfortable
to communicate the details of adverse events that occur, weaknesses in delivery systems.will remain
in place and more adverse events will occur.

10.1.18.4 Senior management should encourage the reporting of adverse events

How can this situation be addressed so that adverse events are consistently reported? Again,
actions speak louder than words. Senior management should encourage reporting of adverse events
by all staff, and state clearly that punitive action will not be taken against staff members who report
adverse events lhat occur as a result of delivery system failures. Undoubtedly, after such statements
staff members will be watching carefully to see what actions senior management take when an
adverse event is reported, to determine whether management is genuine about what they are
saying.
Clearly, in some extreme situations (eg, if a staff member comes to work intoxicated or under the
influence of drugs and an adverse event occurs as a result), strong action is appropriate and should
be taken. However, if a system failure contributes to an adverse event occurring (the far more

211
common situation) and this adverse event is reported, the stafT member should be thanked for
reporting the event. The importance of reporting such events in the subsequent improvement of
delivery systems should be emphasised again to the staff member, and no punitive action should
taken.

Checklist 10.1 The implementation and maintenance of a clinical risk management program
in a health service
Actions Yes No W1P N/A
1 Has an information sheet been developed to help explain the objectives of a
dinical risk management program, and the methods such a program uses?
2 Have clinical champions been found for the program?
■ Do they have high profiles in the health service?
■ Do they have a record of introducing innovative practices in their areas of activity in the health service?
■ Have they had experience writing clinical policies and protocols?
■ Are they respected by medical and nursing staff in the health service?
3 Has an executive champion been found for the program?
■ Does the executive champion have authority to allocate some resources in the health service?
■ Does the executive champion have knowledge of how major decisions are made in the health service?
■ Is the executive champion prepared to be an active member of the surveillance committee and some of the
multidisciplinary problem-solving teams?
■ Is the executive champion strongly committed to the health service providing high-quality clinical care?
4 Do the chief executive officer and board of management support the program?
Have board of management clinical risk management champions been found?
■ Is attainment of an effective clinical risk management program a major objective m the health service's
strategic plan?
| 5 Have adequate resources been allocated to effectively run the clinical risk ; management
program?

Checklist tO 1 (continued on next page)

212
 10 Eighteen practical steps to implementing a clinical
risk management program

Checklist 10.1 (continued)

6 Has a clinical risk manager been appointed?

■ Does the clinical risk manager have:
> Extensive, varied clinical experience?
> Some administrative experience?
> Good written and verbal communication skills?
> The respect of medical and nursing staff members in the health service?
> The ability to work well in a multidisciplinary team?
■ Is the clinical risk manager a member of the health service's main quality committee; the
occupational health and safety committee; and other clinical review committees in the medical,
nursing and allied health areas of the health service?
7 Does the health service have adequate clinical information systems?
■ Do the heads of health information services and the information technology department
support the clinical risk management program?
■ Are health information services staff able to screen medical records?
8 Has statutory immunity been obtained for all activities of the program?
9 Have staff members been educated about the principles of clinical risk management and the
activities of the health service's clinical risk management program?
■ Is education about the principles of clinical risk management and the health service's
program part of the health service's staff orientation program?
■ Are sessions about the principles of clinical risk management and issues raised by the health
service's program held as a regular part of medical, nursing and allied health postgraduate
education programs?
■ Are some changes made in systems of health care delivery as a result of the clinical risk
management program highly visible to health service staff?
10 Has a program launch that will signify the organisation's commitment to the process been arranged?
11 Has the extent of the clinical risk management program that the health service wishes to
establish been determined?
■ How will adverse events be detected?
■ Medical record review:
> Which general patient outcome screening criteria will be used?
> Should any unique screening criteria be developed (eg, in obstetrics or paediatrics)?
> Will clerical or nursing staff members undertake the screening?
> Will screening be undertaken while patients are in health service or after they have been
discharged?
> Will screening be undertaken on inpatient or emergency department medical records, or both?
> If inpatient records are to be screened, is there a typed discharge summary available for each
inpatient admission?
■ Clinical incident reporting:
>• Has a clinical incident reporting system been chosen?
> Have staff members received education about which incidents to report and how to report
them?
> Are incident report forms readily accessible?
rtinklit 10.1 (continued)

_________________________________________________________________Yea No WIP
N/A
I Sentinel MWA reporting:
Am staff aware of the current list of sentinel events that are to be reported?
■ General practitioner feedback:
> Are general practitioners consistently reminded about reporting adverse events
experienced by recent inpatients?
> bit easy for general practitioners to report adverse events?
>• Is reporting of adverse events by general practitioners supported by the local division of
general practice?
> Are general practitioners rapidly acknowledged and given feedback regarding adverse
events they have reported?

231
 10 Eighteen practical steps to implementing a clinical
risk management program

■ Clinical pathways:
> Have clinical areas been chosen in which clinical pathways will be developed?
> Have key process measures been developed to monitor compliance with pathways?
> Has a method for analysing variance from pathways been developed?
■ Clinical indicators:
> Have clinical indicators for measurement been chosen?
> Are the indicators chosen valid and reliable?
■ Patient satisfaction:
>■ Which methods will be used to measure the level of
patient satisfaction (eg, patient satisfaction surveys, focus
groups and/or patient complaints)?
■ External sources of risk (eg, coronial reports, media reports, information from insurers,
statewide and national databases, consultative councils and committees and clinical
journals):
> Which newspapers, medical journals and websites will be regularly screened and by
whom?
■ Has a definition of an adverse event been chosen?
■ Has an adverse event analysis form been devised?
■ How will adverse events be analysed?
■ Has a surveillance committee been established?
■ Is the membership of the surveillance committee appropriately balanced?
■ How will appropriate action be determined in response to an adverse event being detected?
■ Are adverse events prioritised for detailed analysis and action according to their level of
risk?
■ Have adverse events that may provide relatively easy
and early "wins" been targeted for early action to provide
the program with an early positive image of being
responsive and effective?
■ Are health service staff members who report incidents and
adverse events acknowledged? Are they thanked for
reporting adverse events? Are they informed in a timely
manner of the results of the investigations of the events
they have reported and the actions that have been taken to
prevent the events recurring?
■ How will the effectiveness of actions taken to prevent the
recurrence of particular adverse events be determined?

Checklist 10 1
(continued on next page)
Checklist 10.1 (continued)

Actions Vss No wr MM

232
 10 Eighteen practical steps to implementing a clinical
risk management program

■ How are surveillance committee recommendations followed up to ensure they have 4

been implemented?
■ Is the overall response time from an adverse event being detected to action being taken
appropriate? .
12 Are areas and individuals who are providing consistently poor-quality care
promptly addressed?
■ Have adverse events been detected but not addressed because the issues raised are
too difficult or unpleasant to deal with?
13 Where possible, have clinicians who are early acceptors of change been involved
with attempts to implement changes to delivery systems in the health service?
14 Have senior management and clinicians been educated that clinical risk
management activities do not routinely deliver cost savings in the short term?
15 Has high-quality patient care been acknowledged and rewarded?

■ Are staff informed regularly about the quality of care they provide? If so, which of
the following methods are being used?
> posters
> emails answering frequently asked questions
> messages in communication books >■ newsletters
> staff meetings
> education sessions.
■ Have staff members been rewarded for providing high-quality care?

16 Are staff and senior management focused on taking and completing actions that
will make a significant difference to the quality of patient care the health service
provides?
■ Is the risk register, and the actions decided upon to reduce the probability of specific
adverse events recurring, reviewed regularly to ensure progress is being made to
address the issues associated with the highest level of risk?
17 Is the health service's clinical risk management program being enhanced
continually using ideas gained from sources external to the health service?
■ Are clinical journals, websites and other sources scanned continuously for
innovative ideas to improve components of the program?
■ Does the health service have links with other health services operating dinical risk
management programs, so that information can be shared?
18 Are constructive actions being taken to positively change the culture in the health
service in the long term?
■ Do senior management and clinicians encourage open, honest communication?

■ Are staff members delegated an appropriate level of responsibility and autonomy to

be able to effectively change service delivery systems?
■ Are the system changes in response to adverse events developed from the bottom up
in the health service?
■ Have senior management stated clearly that a non-punitive approach will be taken
throughout the health service when dealing with errors that are reported?
■ Is there clear evidence that a non-punitive approach is being taken with staff who
report adverse events?

N/A = not applicable WIP = work in progress

Thanwiarslapsnirnplemertrqadini&riskmanageme^
■ Dovotap an information sheet to explain the purpose and methods used in the program.
■ find clinical champions.
■ find an executive champion.

233
 10 Eighteen practical steps to implementing a clinical
risk management program

■ Obtain chief executive officer support and find board of management champions.
■ Allocate adequate resources.
■ Appoint a clinical risk manager.
■ Ensure the health service has adequate clinical information systems.
■ Obtain statutory immunity for the activity of the program.
■ Educate staff about the principles of clinical risk management and the activities
of the health service's program.
■ Arrange a launch of the program to signify the organisation's commitment.
■ Determine the extent of the program the health service wishes to establish.
■ Promptly address areas and individuals providing consistently poor-quality care.
■ Prepare for hard work.
■ Do not expect clinical risk management activities to routinely generate revenue or
reduce costs.
■ Acknowledge and reward high-quality care.
■ Stay focused on taking and completing actions that will make a difference to the
quality of patient care.
■ Continue to look for ideas to enhance your health service's program.
■ Take a constructive, long-term view to changing culture.

References
1 Craddick JW. Bader B. Medical management analysis: a systematic approach to quality
assurance and risk management. Auburn. Calif: Joyce W Craddick. 1983.
2 Hiatt HH. Barnes BA. Brennan TA. et al. A study of medical injury and medical
malpractice: an overview. N Engl J Med 1989: 321: 480-484.
3 Nelson EC. Splaine ME. Batalden PB. et al. Building measurement and data collection
into medical practice. Ann Intern Med 1998: 128: 460-466. □

11 Implementing a clinical risk management

program in a small hospital

Thh chapter will:

■ Discuss practical steps for implementing a clinical risk management program in a smal hospital. wMch nay have only one or
two doctors on staff.

11.1 Problems unique to small hospitals that want^o undertake a

clinical risk management program
Clinical risk management programs can be implemented in hospitals of any size. The nature and size
of the programs will vary, however, not only because of different local needs and conditions, but also
with the complexity and size of the hospitals involved. Many components of the clinical risk
management program described so far are designed to be implemented in medium-sized and large

234
 10 Eighteen practical steps to implementing a clinical
risk management program

hospitals. However, some features of clinical risk management programs in hospitals of this size
cannot be translated easily into small hospitals, which may have only one or two doctors on staff.
One significant problem for doctors who work together closely is that it is difficult to objectively
review the medical records of patients treated by their colleagues. Similarly, it is difficult for doctors
working alone in small hospitals to objectively review their own work. A different approach to clinical
risk management is clearly required in these situations.

11.2 Modifications required for programs in small hospitals

To attempt to overcome these difficulties, the medical records that meet screening criteria could be
sent to a larger centre for review. Such an approach would need to be handled sensitively, as it would
be easy for doctors in small hospitals to feel that medical reviewers in the larger centres were sitting in
judgment of the medical care they provide.
Alternatively, several small hospitals could combine resources to run a clinical risk management
program for all the small hospitals in a region. Doctors in small hospitals could review the medical
records of patients who were admitted to other small hospitals in the region. This approach has the
advantage that the doctors reviewing medical records are working in a similar environment to the
doctors whose work is being reviewed, and would therefore be sensitive to some of the issues faced by
doctors working in isolation in small hospitals.
Such an approach has been developed in the Wimmera region of Victoria. Building on the expenence
of the clinical risk management program developed at Wimmera Health Care Group, the Monash
University Centre for Rural Health engaged the West Vic Division of General Practice to develop a
clinical risk management program for small rural hospitals. Initially, the Division set about gaining
general practitioner support for the program. With general practitioner support, small rural hospitals
in the area were then approached to join together to conduct a region-wide clinical risk management
program that was coordinated by the West Vic Division of General Practice.
11.3 Screening and medical review
In essence, the small hospital clinical risk management program operates in a similar manner to the
medical record review component of the program at Wimmera Health Care Group (described in
Chapter 4 and other earlier chapters). Medical records of patients discharged at each participating
hospital are screened by health information managers at that hospital, using seven general outcome
criteria:
■ patient death
■ return to theatre within 7 days
■ unplanned re-admission within 28 days
■ transfer to another acute care facility
■ patient length of stay greater than 21 days
■ unexpected re-admission following discharge from another hospital
■ any record that has been recommended by a doctor for peer review.
If a record screens positive for at least one of these criteria, it is photocopied and sent to the
program's project officer, who sends it to a doctor from an adjacent geographical area for medical
review. At medical review, the doctor determines whether an adverse event has occurred, and if so,
assesses the severity and preventability of the event. The adverse event is discussed at a regular
meeting of the medical reviewers. This group of reviewers makes recommendations about clinical
care, and details of these recommendations are sent to all participating doctors and hospitals.

11.4 Obtaining the support of small hospitals

By participating in this program, small hospitals gain the benefit of:
■ an independent review being undertaken of some of their medical records
■ access to the details of recommendations for dealing with adverse events that have occurred within
their hospital and at others of a similar size
■ general practitioner involvement with hospital quality improvement and risk management
programs. As some of the participating hospitals are independent organisations and can be separated
by significant distances, the program is conducted with some important differences from programs
conducted in larger hospitals. Considerable effort is required to provide information to general
practitioners and hospitals about how a region-wide clinical risk management program would be

235
 10 Eighteen practical steps to implementing a clinical
risk management program

conducted and the benefits of participating. It is necessary to gain the approval of these general
practitioners and hospitals. In the Wimmera small hospitals project, the West Vic Division of General
Practice and the Monash University Centre for Rural Health undertook this task. To help gain
acceptance among general practitioners and small rural hospitals, the program was coordinated by
an independent body, even though expertise in clinical risk management was available at the main
referral hospital in the region.

11.5 The logistics of screening and medical review

After approval has been obtained, an extensive education program about the clinical risk
management program should be provided to key hospital staff. The Division seeks doctors to act as
medical reviewers, and conducts training sessions on how to review records and complete the adverse
event analysis form.
The hospital's health information manager screens the medical records of all patients at each
participating hospital on discharge. Because of the distances between hospitals, any record that
screens positive is photocopied. The original medical record remains at the participating hospital
and the copy is sent to the program project officer. The project officer then forwards the photocopied
medical record and an adverse event analysis form to a medical reviewer from an adjacent
geographical area. Because of the confidential nature of the medical record, a system is developed
that maintains the security of the copied medical record; this includes posting all copied medical
records by registered mail and coding data entry.
General practitioners working in difTerent geographical areas within the same region are grouped
together for the purposes of the program, to allow anonymous "arms-length" medical review.
General practitioners do not review their own medical records. The medical reviewer reviews the
medical record and determines whether an adverse event has occurred. If an event is detected, the
medical reviewer summarises the clinical details and assesses the severity'and preventability of the
event. Once the medical review has been completed, the photocopy of the record and the completed
adverse event analysis form are returned to the project officer. If no adverse event is identified, the
copy of the record is shredded.
If, in the opinion of the medical reviewer, an adverse event has occurred, the reviewer's comments
are sent to the treating general practitioner for comment. The treating general practitioner may be
able to add further information about why panicular action was taken, and this information may
assist further assessment of the admission. The treating general practitioner is also given the
opportunity to comment on the assessment of the medical record made by the medical reviewer.

11.6 Reference panel recommendations

The general practitioners undertaking the medical reviews meet regularly as a reference panel. The
panel discusses the clinical details of each of the adverse events detected. Issues are identified,
and, if appropriate, recommendations are made to participating doctors and hospitals about clinical
practice and administration changes they should consider. In addition, specialists' comments about
management issues are sometimes sought. Following discussion of the adverse event at the reference
panel, the copied medical record is shredded.
The summary of the clinical details of each case and the recommendations of the reference panel
are sent to each member of the medical staff and the chief executive officers of the participating
hospitals. To complete the clinical risk management cycle at the local level, the doctors discuss the
appropriateness of the reference panel recommendations in their hospital at the main quality forum
in the hospital, and modify recommendations to be implemented locally. Following these discussions,
a representative from each hospital provides feedback to the reference panel on action their hospital
has taken in response to each case and the recommendations.
As with implementing clinical risk management programs in medium and large hospitals, we
recommend that divisions of general practice, or equivalent organisations starting such a region-wide
program, organise an extensive education program involving key hospital stafT — including hospital
chief executive officers, general practitioners and health information managers. Incentives to general
practitioners to participate include gaining credits towards maintaining vocational registration
Clinical and executive champions need also be identified, and their support for the program should be
gained A project officer should be appointed to conduct and coordinate the components of the
program, and a steering committee formed to oversee the running of the program.

236
 10 Eighteen practical steps to implementing a clinical
risk management program

The Wimmera small hospitals clinical risk management program commenced in 1994 The program
has promoted discussion of adverse events, provided meaningful peer review for general practitioners
and small rural hospitals, and strengthened professional relationships in isolated communities There
has been a cultural shift away from an attitude of individual blame to one that acknowledges
systemic causes of adverse events.
Knowledge and experience have been shared in a non-threatening way and recommendations have
head k> quality improvement in small rural hospitals. As the coordinating body, the West Vic Division
of General Practice is external to the hospitals in the program, and is not perceived as a threat to
them. There has also been acceptance of the program by participating hospitals because it provides
noncompetitive, anonymous peer review between neighbouring hospitals. The general practitioners
have ownership of the program. The benefits of participating has become self-evident to them, and
they therefore remain committed to the process.
The small rural hospitals and the Division have been able to establish interventions aimed at
improving the quality of patient care. The program has identified clinical and system issues, and
implementing recommendations and providing general practitioner education has led to improved
patient care in participating hospitals. The program has also identified trends that individual general
practitioners and small rural hospitals would have been unable to recognise on their own. There has
also been an improvement in the standard of medical records. Documentation improvement and the
availability of discharge summaries, copies of antenatal results, and copies of death certificates have
improved the continuity of patient care.

11.7 Small hospital clinical risk management manual

The West Vic Division of General Practice's clinical risk management program for small rural
hospitals builds on the model developed at Wimmera Health Care Group. But, as already discussed,
there are important differences in the two programs because of the diverse geographical location and
independent functioning of participating general practitioners and hospitals. It is recognised that
most small hospitals and some divisions will not have the infrastructure to establish a
comprehensive, effective clinical risk management program without some guidance. Therefore, a
detailed guide for each component of the program is provided in a manual for small hospitals wishing
to establish such a program. This manual is available at the West Vic Division of General Practice
website (www.westvicdiv.asn.au/resources/ Resources%20Folder/CRM%20Manual.pdO and covers
the following areas in detail:
■ preplanning
■ information meeting
■ governance
■ the roles of medical reviewers, chief executive officers, health information managers and project
officers
■ statutory immunity
■ data collection tools
■ postal arrangements
■ flow of medical records
■ reference panel meetings
■ sample documents.

11.8 Introduction of the program across rural Victoria

In 2001. the West Vic Division of General Practice began implementing the program through the six
rural divisions of general practice into the 75 small hospitals in rural Victoria. The first annual report
of the statewide program (for 2006-07) reported that 1681 medical records were reviewed across the
state, and 163 recommendations to improve the care patients receive were made. Through the first
annual report, some of the recommendations are now available to all doctors and health services
throughout the state.1

11.9 Changes to the screening criteria

237
 10 Eighteen practical steps to implementing a clinical
risk management program

The screening criteria used by small hospitals have been reviewed and modified to reduce the
number of records that screen positive and are sent for medical review, but in which an adverse is
not found. The modified criteria are as follows:
■ unexpected patient death
■ unplanned return to theatre within 7 days
■ unplanned re-admission within 28 days of discharge
■ unexpected transfer to another health service
■ patient length of stay greater than 35 days
■ any record that has been recommended by a doctor or other health professional for review.

Important points
■ The effectiveness of standalone clinical risk management programs in small hospitals is limited
because it is difficult to independently and objectively review clinical care locally, as occurs in larger
hospitals.
■ Grouping small hospitals together in a regional clinical risk management program allows inpatient care
provided in a small hospital in one geographical area of the region to be reviewed by a doctor
working in another area.
■ Having a neutral body, such as the regional division of general practice, rather than the major referral
hospital in the region, coordinate the program helps engage doctors in small hospitals with the
program.
■ The basic steps of a clinical risk management program of screening of medical records, medical
review, analysis and recommendations are followed.
> A photocopy of medical records screening positive lor one or more criteria is sent for medical
review to a participating doctor in another hospital in the region.
> If an adverse event is detected at medical review, the treating general practitioner is sent the
reviewer's comments and invited to respond.
>- A reference group of general practitioners in the region examines the medical review and the treating
doctor's response and makes recommendations about patient care. >• These recommendations
are considered by doctors in the region and the quality committee in each participating small
hospital.

Reference
1 Victorian Government Department of Human Services Limited Adverse Occurrence Screening (LAOS)
program annual report for 2006-07. Melbourne: Victorian Government Department of Human Services. 2007
www.health.vic.gov.au/clinrisk/publications/laos_report.htm (accessed Oct 2008) □
12 Important messages
This chapter wM:
■ Summarise the important messages of this book.

We have read many books and attended many workshops and conferences on clinical quality and
safety Despite much being written, many people attending such programs, and much activity,
improvements »n quality and safety have occurred slowly. There may be many reasons for this,
including: the difficulty of changing health care delivery systems and clinical behaviour, information
overload, insufficient resources being provided for such programs, lack of motivation to change,
unsupponive health service cultures, or just too many other important issues diverting the attention
of management and clinicians Given this depressing scenario, what important messages would we
like clinicians and management to retain and apply after reading this book?

238
 10 Eighteen practical steps to implementing a clinical
risk management program

12.1 Health care as a system

Health care is delivered using complex systems with many components that can interact
unpredictably. Outcomes are often uncertain. Despite detailed planning, expect unintended negative
outcomes and setbacks. Do not be disheartened and do not abandon quality and safety initiatives
when such events occur. Every health service that is trying to improve will experience adversity —
usually sooner rather than later. Try to leam from these experiences and apply the lessons leamt in
the future.

12.2 Clinical governance

Clinical governance is a difficult term to explain to clinicians, and doctors are likely to interpret
governance as a form of control by management. Also, the principles and activities of clinical
governance programs may be packaged in different ways and vary between programs, adding further
complexity. The simplest explanation of clinical governance that we have found is that it is the
regulation of the health care delivery systems to enhance patient care by ensuring:
■ quality improvement — doing the right thing for patients more often
■ risk management — doing the wrong thing for patients less often.
Clinical governance is usually a much lower priority in health services than corporate governance
(or the regulation of financial or business outcomes). The financial performance of health services
receives much more attention than their clinical performance. However, both components of
performance are equally important. To raise the level of importance of clinical governance, it should
be made a core component of the service's strategic plan, the first item on the agenda of the meetings
of its board of management, and a standing agenda item at meetings of key clinical groups in the
health servicc Adequate, ongoing resources (including staffing, protected time for clinician
participation and ready- access to clinical databases) should be provided for the activities of the
clinical governance program, or it will languish and be ineffective.
12.3 Quality improvement
Audits have often been undertaken as a "compliance exercise" to meet the requirements of clinicians'
continuing education or health service accreditation. Frequently, audit results have not resulted in
action being taken to close gaps in clinical care. Audit should be a continuous, integral part of a
serviced quality and safety program. Audits should be conducted with early and ongoing clinician
involvement. Data collection and analysis should be of high quality. Above all, the recommendations
for changes in health delivery systems arising from the analysis of audits should result in timely,
appropriate and concrete action. After action has been taken, the audit cycle should continue until
the gap in care has been closed.

12.4 Detection of adverse events and risk

There are two major ways to detect adverse events and risk — reactive methods, which detect adverse
events, and proactive methods, which detect risk at a health service following the study of adverse
events in other health services or the detection of weaknesses in health care delivery systems by
actively looking for them. A clinical risk management program should use both methods, but we
strongly recommend that health services use proactive methods extensively, because they avoid
physical and psychological trauma to patients, their relatives, and staff, by addressing system
weakness and preventing adverse event occurring.

12.4.I Reactive risk management — detecting adverse events and near misses
We recommend using a diverse range of methods to detect adverse events, as there is little overlap of
events detected by different methods. Three methods are most useful:
■ Medical record review
>■ A useful way to start is to use three screening criteria — patient deaths, transfers from the
general ward to the intensive care unit, and any records referred for review by a clinician, then
expanding the criteria as the program develops.
> Ensuring there is comprehensive, typed discharge summary for each patient admission greatly
assists with medical record review.
> Be aware that many records will screen positive for a criterion but will not contain an adverse
event.

239
 10 Eighteen practical steps to implementing a clinical
risk management program

■ Clinical incident reporting. Many incident reporting programs languish because of poor education
of stafT about the process and non-existent feedback to those reporting incidents. Important
strategies for a successful program include
> Educate staff about what incidents should be reported and how to report them.
>• Provide timely feedback about the results of investigations and the actions taken in response
to reports
> Make many of the actions taken in response to reports visible to staff. Show staff thai they can
improve the care they provide and their work environment by reporting incidents.
■ Feedback from general practitioners
> Systematically obtain feedback from general practitioners about adverse events they detect
after a patient's discharge from hospital.
> Many adverse events obvious to general practitioners will not require re-admission to hospital,
so without much feedback, health services would not be aware of ihem.
/ 2.4.2 Proactive risk management — detecting risks
m Review latent factors or conditions of work when developing new delivery systems or reviewing
existing systems. If delivery systems are created with inadequate staffing levels, supervision,
training and equipment, then adverse events will occur.
■ Assess each patient's risk of common adverse events (eg, falls, thromboembolus) when they are
admitted to the health service, and take action to reduce the probability of these events occurring.
■ Systematically review multiple sources of information about adverse events (national databases,
consultative committees, coronial and media reports, pharmaceutical alerts and hospital inquiries)
occurring at other health services to detect system weaknesses and local risks. Although this requires
additional resources and work, it is well worth the elTort if it avoids the trauma of a serious and
preventable adverse event within the health service.

12.5 Analysing adverse events

The first steps when analysing an adverse event are assessing whether:
■ ihe patient and their relatives have been a managed appropriately
■ the staff involved have been managed appropriately.
In our experience, these steps are often not adequately taken. In panicular, the effects of an adverse
event on the staff involved are often overlooked. A full root-cause analysis of an adverse event is
resource-intensive, and should be reserved for the mosi
serious adverse events. Systems analysis, which determines the care management problems, the
relevant
clinical events and the contnbulory factors, is a more efficient analysis and should be most commonly
used.

12.6 Prioritising gaps and risks

In health services with a strong safety culture, many more gaps, adverse events and risks will be
detected than can be appropriately managed at any one time. Gaps, adverse events and risks can be
prioritised for
action according to:
■ their frequency
■ the harm they cause
■ their preventability
■ their cost
■ their benefits
■ their feasibility lo change.
Each health service will probably prioritise similar adverse events in a different order, depending on
local needs and conditions.

12.7 Closing gaps and reducing risks

Many opportunities will be available to health services to close clinical gaps or reduce risks. It is
important valuable resources are not wasted on interventions that are going to be ineffective The

240
 10 Eighteen practical steps to implementing a clinical
risk management program

effectiveness of an intervention will be extremely sensitive to the context in which it is introduced,

and an intervention effective in one health service may not be effective in another.

241
 Enhancing patient care

12.7.1 Three Important questions when considering an Intervention

When considering an intervention to close a gap or reduce a risk, the following conditions should be
met before developing and implementing the change:
■ Is the change scientifically based? There are many interventions available that have been
demonstrated to improve quality and safety, so it is important not to waste valuable resources on
unproven interventions.
■ Do medical stafT believe the gap or risk is important? The support of medical staff is crucial to the
successful implementation of many changes in health care. Without their support, many
interventions will not be successful. There are so many interventions available — why waste
resources on interventions not supported by them?
■ Will management provide adequate resources and hands-on involvement? Many interventions
require substantial resources for implementation and maintenance. Under-resourced interventions
are likely to fail. Senior management involvement in project committees assists implementation,
especially across departments, and creates ownership of the intervention at executive level in the
health service. Executives are less likely to stop a successful quality and safety program if they have
been closely involved in each stage of the program^ development.

12.7.2 The importance of context

The context in which the intervention is planned is critical to its successful implementation. It is
important to systematically determine (by interview or other forms of feedback from the target group
and others) the characteristics of the
■ change or innovation
■ the users or target group
■ patients
■ organisation (size, role, structure, administration, governance, communication, culture, leadership and resources)
■ sociopolitical context and external factors
that facilitate or impede the adoption and implementation of the change. In our experience, the characteristics of the users or target
group are the most critical elements in determining the success of the intervention.
After the determinants have been identified, they should be ranked according to their importance and changeability. Tailor the
change strategy to address the major determinants using theories about behavioural change and the judgement of target group
members. Remember, some determinants (eg, strong clinical leadership, adequate resources, and senior management participating
involvement in project teams) may be more important than the number and type of implementation strategies used. Although this
approach is subjective, it is valuable, given the complex nature of the health care environment and the limited scientific evidence
about successful change strategies.
12.7.3 Implementation strategies
Implementation strategies (eg, dissemination of educational materials, audit and feedback, reminders, multifaceted interventions
incorporating education outreach, and using opinion leaders) only have a moderate effect, which is sensitive to many determinants.
However, in some contexts they can be very effective.
Although not supported by scientific evidence, we have found it most effective to use a combination of quality improvement and
system redesign strategies to implement and sustain change to improve quality and safety. Individual strategies that we have used
include: simplifying systems, standardising procedures, using checklists and reminders, introducing constraints, and timely
delivery of adequate and accurate information. We acknowledge that our success in using these
strategies may be speafu. to the context in which we work, but we feel they may be also be effective
in other health services

12.7.4 Use resources developed elsewhere

Avoid reinventing the wheel — use resources (eg, literature and systematic reviews, policies,
procedures, key process and outcome measures, educational material, and strategies) developed
externally to the health service and tailor them to local needs and conditions. The resources available

242
 12
Important
kimnqh

for quality and safety programs are usually very limited. Why waste them by re-creating a tool that
has already been developed by another health service?

12.7.5 Small-scale testing of strategies

Test the change strategy on a small scale first, and make modifications before implementing it
through the health service. As mentioned previously, as well as the change strategy itself, the success
of quality and safety interventions are highly sensitive to the context in which they are introduced.
Negative surprises can and do occur in complex health care delivery systems. Small-scale tests allow
the effects of strategies in complex systems to be observed and changes made accordingly made
without major disasters occurring.

12.7.6 Be vigilant
It is imponant to be vigilant. Monitor the effects of any change closely and modify the change strategy
if required. We have frequently observed a change being successfully made, but attenuating over
time. Surveillance of the performance of systems should be ongoing.

12.7.7 Provide feedback to clinicians about their performance

Provide confidential, timely, credible and non-punitive feedback about individual performance and
achievable benchmarks to clinicians. People like feedback about their performance, and clinicians are
no exception. If they are performing poorly compared with their peers, they usually do not want
others to see their results. However, many will wish to improve their results. Do not provide clinicians
with poor- quality data, as this could undermine the credibility of the quality and safety program.

12.8 Conclusion
Finally, quality improvement and risk management in health care is a joumey rather than a
destination The terrain may be difficult, the process slow and improvements difficult to maintain.
Much evidence and opinion about how to proceed has been presented here. Perhaps the biggest
obstacle to change in complex organisations is inertia. Taking effective action is up to the individuals
within organisations As with any strategy in life: keep it simple, do your best, do not give up and
success is likely to follow.1

References
1 Buchanan J. Learning Irom legends: Australian cricket Sydney: Fairfax Media Publications. 2008 □

2S7
Enhancing patient eve

Appendix 1: Wimmera Health Care Group Adverse Event - Analysis

Form
 Episode No URNo: Admission Date | Discharge Date
ED Inpatient

Criteria (Circle all criteria met during admission)

1. Death «
2. Transfer to Intensive Care Unit
3. Cardiac arrest
4. Length of stay greater than 21 days
5. Returning to theatre within 7 days
6. Unplanned re-admission within 28 days
7. Transfer to another acute care facility
8. Booked theatre cases cancelled
9. External cause code (Y40-Y84)

10. Accident &r Emergency length of stay greater than 6 hours

11.Unplanned A& E re-presentation within 48 hours
12.Unplanned A & E attendance within 28 days of hospital discharge

Clinical Review
An adverse event is an untoward patient event which under optimal conditions is not a natural
consequence of the patient's disease or treatment.
Please evaluate the clinical care given by rating the evidence that an adverse patient occurrence was causcd bv medical
management (circle the appropriate number)
1. Little or no evidence of an adverse event caused by management
2. Slight evidence
3. Not quite likely (less than 50:50 odds but a close call)
4. More likely than not (greater than 50:50 odds, but a close call)
5. Strong evidence
6. Virtually certain evidence

If an adverse event has occurred where did it occur:

1. Within this hospital
2. Another hospital
3. Outside hospital including general practitioner care
If yes, please give relevant clinical details:________________________________________________
 12
Important
kimnqh

No disability no significant resultant discomfort of functional impairment; and no increased length of stay as a result
of the adverse event.
Minimal to moderate clinical effect requiring no or minimal clinical intervention, or no increased
length of stay or re-presentation for the same or related problem. Minimal to moderate dinical effect with
permanent residual and without significant functional or cosmetic impairment.
code the severity of the adverse event
Moderate to severe clinical effect with no significant functional or
1 a MinorsMi^f
5 Major permanent cosmetic residual effect. This usually results in increased length of
26 Potential
Minor major or stay or re-hospitalisation and requires moderate to major clinical
temporary
major continuing intervention.
Moderate to severe clinical effect with significant functional or cosmetic
3 Mtoor permanent
Death residual. When doubts exist as to the outcome but the probability is
that a major impairment or repeated presentations or hospitalisation
4 Major temporary
will be necessary. The outcome may result in major impairment.

Error cause scale

Tl Complications/failure in technical performance
T2 Failure to follow policy, protocol or guideline when indicated
T3 Acting on insufficient information
T4 Slip/lapse in activities in which operator is skilled
T5 Inadequate knowledge/experience
T6 Inappropriate or not indicated care
T7 No/wrong/delayed treatment
D1 Failure synthesise/decide/act on available information
D2 Failure request/arrange procedure or consultation
D3 Missed diagnosis
SI Lade of care or attention/failure to attend when called
S2 Failure in communication
S3 Equipment failure/not available
S4 Inadequate staff
S5 Staff fatigue/distress/sickness
56 Poor discharge planning
PI Inadequate follow up of treatment/monitoring
P2 Failure to provide indicated prophylactic treatment
Use Scales
1 Mhinr Appropriate health service selected but poorly provided
2 Underuse Failure to provide health service when benefits exceed the risk
3 Overuse Provision of a health service when risks exceed the benefits
 12
Important
kimnqh

1 Little or no evidence for preventability

2 Slight or modest evidence for preventability

3 Preventability not quite likely: less than 50: 50 but dose call

4 Preventability not quite likely: more than 50: 50 but close call

5 Strong evidence for preventability

6 Virtually certain evidence for preventability

What further action would you recommend

1 None

2 Change in relevant hospital policy

3 Presentation of case at a postgraduate meeting

4 A detailed quality assurance program to investigate these type of occurrences in detail

5 Discussion with the doctor involved

6 Informing the hospital insurers of the case

7 Other, please specify:

Reviewer's signature. Date_______/_______/

Appendix 2: Clinical risk management executive summary
This summary explains the concepts of a clinical risk management program to health service staff and board of
management members with little knowledge in this area. It can be modified to suit local conditions and is available
at www.whcg.org.au/Quality/index.aspx
Health care delivery is associated with a high error rate compared with other complex industries.
Traditional medical quality assurance has resulted in much activity, often lacking integration and
resulting in disappointingly small, inconsistent improvements in the quality of patient care (with a
few exceptions in specific areas). This summary describes a comprehensive clinical risk management
program designed to integrate detection and reduction of adverse patient events into the everyday
activity of an acute general hospital.
Clinical risk management is about reducing the probability of negative patient events. This involves
detecting adverse events, analysing their risk severity, and taking appropriate action to prevent their
recurrence. The Wimmera Clinical Risk Management Project began this process by developing several
methods for detecting adverse events. The initial detection method was medical record review using a
two-step process of screening all inpatient medical records using general medical outcome criteria
(eg, return to the operating theatre, re-admission within 21 days of discharge) and medically
reviewing records that screened positive. This method detects many adverse events in an effective,
efficient manner. Analysis of these adverse events and the taking of appropriate action has resulted
in a significant reduction in the rate of inpatient adverse events that has been sustained over 16
years.
The two-step process has been modified for use in reviewing emergency department medical
records. These modified screening criteria have been in use for over 11 years, and have allowed
efficient detection of adverse events in the emergency department. Appropriate action has then been
taken, which has reduced the rate of adverse events.
To improve the detection of adverse events, a coordinated method of clinical incidents being
reported by staff members was introduced. Clinical incidents and near misses are reported by staff
both to the hospitals involved in this program and to a national database of the Australian Patient
Safety Foundation. The most common incident reported has been patient falls (47.7% of all incidents
reported). A significant finding was the low overlap of adverse events detected by both the medical
record review process and clinical incident reporting. Each detection method appears to find adverse
events from a different pool.
Despite using these comprehensive methods to detect adverse events, some events are still not
found. An adverse event may occur after a patient is discharged from hospital, but the event may be
due to their treatment while they were in hospital. For example, a wound infection after surgery may
develop several days after discharge. To capture this important information, each hospital inpatient
discharge summary that is sent to the patients general practitioner also contains an adverse patient
event reporting form. If the patient's general practitioner detects an adverse event after discharge, it
can easily be reported to the clinical risk management program. This process detects adverse events
that are often not recorded in the hospital medical records and have not been reported by hospital
staff.
 Some adverse events occur infrequently, and may only be detected once every few years in
individual hospital clinical risk management programs. To maximise the benefit of this pool of
valuable information about serious but infrequent events, information about adverse events that
have occurred at other hospitals has been used as if the event has actually occurred at Wimmera
Health Care Group. Information about such adverse events is available from coronial reports,
consultative committee annual reports (eg, the Consultative Council on Obstetric and Paediatric
Mortality and Morbidity, the Victorian Consultative Council on Anaesthetic Mortality and Morbidity,
and the Victorian Surgical Consultative Council), hospital insurers, medical indemnity organisations
and clinical journals. After an adverse event is repotted by one of these sources, the question is
asked: Could this adverse event occur here? If the answer is yes, analysis and action are taken to
reduce the probability of such an event occurring It is not necessary for a hospital to wait to
actually experience a particular low-frequency advene event to benefit from it having occurred
elsewhere.
In a similar preventive vein, undertaking a risk assessment for each patient admitted to the
hospital can prevent common adverse events. Fot* example, each patient aged over 65 years who is
admitted lo the hospital undergoes a falls risk assessment. Patients assessed as high risk for having
a fall while in hospital are managed on adjustable-height beds and in rooms close to the nurses'
stations so that assistance is close at hand if required.
There is sometimes a delay in adverse events being detected by reviewing medical records or by
staff members reporting clinical incidents. To capture the details of more adverse events shortly after
they occur, the use of evidence-based clinical pathways for particular filinical situations will be
expanded for both inpatients and emergency department patients. Variances from pathways can be
reported at the time they are recorded in the medical record and analysis of the variance will tie
undertaken prompdy.
Although research has shown patients to be poor judges of the quality of the clinical care they
receive, they are able to accurately assess the quality of non-clinical care, which is a significant
component of the quality of care and service provided by hospitals. Information from patients
provides valuable data for improvement of systems in hospitals. Adverse events that have occurred
from the patient's perspective are also delected. This is achieved using information from patient
satisfaction surveys, focus groups and patient complaints.
In this clinical risk management program, it has become apparent that it is not possible to manage
the risk associated with all the adverse events detected. Therefore, as each adverse event is detected,
the risk associated with the event is determined. The probability and consequences of the risk are
assessed, and a risk severity rating is given to the adverse event in accordance with the Standards
Australia/Standards New Zealand Risk Management Standard. The adverse events are then ranked
according to their risk severity rating. Those with the most severe rating are given priority for
analysis and action aimed at eliminating or reducing their risk. In the case of adverse events with low
risk severity, a decision is made as to whether to take action or to accept the risk and continue
monitoring the particular adverse event.
The actions taken to manage clinical risk have been a large component of this program to date.
After an adverse event has been detected, the risk is analysed, and if the risk severity is high, a
multidisciplinary team is assembled. An aim for quality improvement is established by the team,
together with a balanced set of process and outcome measures. Options to improve systems involved
in the adverse event are derived empirically or from research (evidence-based, if possible) or
observation, and are examined in detail. Small-scale tests are undertaken to implement chosen
improvement strategies after baseline measurements are taken. Measured results of the tests are
analysed and improvement strategies are adopted, modified or abandoned. Further small tests are
then undertaken if required, following the same sequence. This strategy has underpinned the
significant improvements that have been achieved in the adverse event rates in this clinical risk
management program.
 Enhancing patient care

A major component of the system improvements program has been the implementation of the
results of research into error prevention. Adverse events are analysed to identify latent factors (eg,
workload, level of supervision, communication and equipment) and active factors (eg, memory lapses
and slips) that contributed to the error occurring. Actions taken have concentrated on improving the
work environment and removing total reliance on the weakest components of human cognitive
function, especially shon-term memory and the ability to be distracted from the task at hand.
Strategies used to address these issues include simplification of systems, standardisation of
procedures, use of reminders and checklists, and the timely delivery of information.
Information from the program is regularly reported to participating doctors, clinical staff and
hospital quality committees and boards of management.
Reducing the probability of adverse events occurring in hospitals requires a comprehensive,
integrated approach to detecting adverse events, analysing the events and taking appropriate action
to successfully reduce the chances of the events recurring. The process requires support from senior
clinicians and executive staff, board of management members and all health care professionals
working at a hospital. The work is labour-intensive and requires additional resources. However, the
rewards of improving the level of patient safety are attainable and significant. The establishment and
maintenance of an effective clinical risk management program is a worthy objective for all hospitals.
□

234
Index
A u
a a
a l c
cc i c
id t r
e y e
nt di
s, o t
3 f a
7 ti
a s o
cc e n
o r b
u v o
nt i di
a c e
bi e s,
lit , 1,
y, 2
6 3 0,
6, 3
7 a 0,
2 c 6
cl c 8,
in r 7
ic e 1
al d a
g i d
o t m
ve a i
r t n
n i is
a o tr
n n a
ce , ti
p o
ro 1 n
gr 0 ,
a U 1
m n 2
, i 5
1 t b
1- e a
1 d d
2, m
1 S is
4 t si
d a o
ef t n
in e s,
in s s
g, . e
1 e
2 8 h
q 6 o
s i 1
pi n
ta t c
l s o
a , s
d t
m 1 .
is 1
si 5 1
o 0
n a 9
s u b
a d ,
d i
ve t 1
rs i 0
e n 9
ev g -
e . 1
nt 1
s, 1 0
5 0
a , d
n e
al 4 f
y 1 i
si n
s, A i
5 u n
6- s g
5 t ,
8, r
7 a 5
4- l 6
7 i ,
5, a 6
9 , 4
8- ,
1 3
0 9 1
3, 0
2 c 3
1 o ,
8, n 1
2 s 0
2 e 3
5 q b
i u
m e d
p n e
o c l
n e e
a s c
n , t
t i
1 n
p 0 g
o 4 ,
 - k
4 2 ,
9 2
, 5 7
1 4
5 f
6 r H
, e a
1 q r
7 u v
9 e a
n r
b c d
a y
s , M
i e
c 3 d
8 i
m - c
e 3 a
t 9 l
h ,
o 1 P
d 0 r
s 9 a
. , c
2 t
4 1 i
9 0 c
- e
5 g
0 e S
n t
i e u
m r d
p a y
o l ,
r
t p 5
a r 1
n a
t c i
t n
p i f
o t o
i i r
n o m
t n a
s e t
, r i
o
8 f n
9 e
, e s
d o
2 b u
2 a r
4 c c
e e n
s n t
, a
d b
4 i l
9 s e
, c .
8 l
8 o 3
, s 9
8 u .
9 r
, e 5
2 , 0
2 ,
5 1
0 5
i . 7
n 9 -
v 5 5
e 8
s a
t n p
i d r
g e
a p v
t a e
i t n
n i t
g e i
, n o
t n
4 .
2 e
8 v
m 7 i
u . d
n 9 e
d 5 n
a , c
n 9 e
e 6
. - a
9 n
1 7 d
1 ,
0 9 t
, 9 h
1 . e
1 2 o
2 2 r
- 5 y
1 ,
1 p
3 r 1
e 1
o v 9
p e
i - f
n 1 a
e 1 c
f 4 t
f o
e p r
c r s
t o ,
i b
v a 3
e b 9
i -
m l 4
e i 0
a t ,
s y 4
u , 2
r
e I r
s 0 e
, 4 c
( u
1 r
4 r r
9 e i
a n
p s g
r o ,
i n
o s 1
r 0
i f 9
t o -
i r 1
s 1
i o 4
n c .
g c 1
u 7
a r 5
c r .
t i 1
i n 9
o g 6
n , .
, 2
3 0
1 9 8
0 -
3 4 a
- 1 n
1 l d
0 a
5 t r
. e e
1 n l
0 t a
9 t
i . af
v f.
e 3 9
s 7 5.
. - 9
4 7,
9 7 9
5 , 9,
. 7 2
9 8 2
6 5
- s s
9 c y
7 r st
. e e
2 e m
2 n s
5 i a
n n
r g al
e y
p c si
o r s,
n i 9
i t 8
n e -
g r 9
, i 9
a v
1 , u
7 l
7 5 n
, 1 e
2 , r
1 5 a
2 2 bi
, li
r 5 ty
i 2 ,
s b 1
k 9
. s 5
e s
1 r e
6 i e
5 o al
ri u s
s s o
k . r
m e
a 5 p
n 2 o
a a rt
g n i
e d n
m g
e s A
nt t H
R n rts
Q g se
s S e
e y pa
e s tie
U t nt
ni e sa
te t fet
d Q y,
St A al
at I ert
es H s
. W all
A ie
g s d
e e he
n e alt
c h
y A st
fo u aff
r s ,
H t 15
e r 7
al a an
th li ae
c a st
ar n he
e si
R I a
es n ha
e s nd
ar t he
c i ld
h t co
a u m
n t pu
d e ter
Q o s,
u f 66
al H m
it e ort
y a ali
AI l ty,
M t 11
S h 1
se an
e a nu
Au n al
str d re
ali W po
an e rts
In lf ,
cid a 11
en r ,7
t e 3,
Mo a 78
nit l .1
ori e 75
,2 A n
20 u He
+ s alt
an t h
on r Ca
ym a re
ity li St
, a ud
67 , y,
.2 3 39
19 0 ,5
,2 , 6.
20 8 10
att 0 2b
itu , ,
de 1 10
th 0 9
eo 1 se
rie h nti
s, o ne
13 s l
1 p ev
au i en
dit t t
. a re
9, l po
14 i ni
4. n ng
16 q ,
3, u 72
17 i -
7. r 73
18 i Th
1. e er
20 s ap
8 , eu
pr 8 tic
os 5 Go
pe - od
cti 8 s
ve, 6 Ad
23 Q mi
ret u ni
ro a str
sp li ati
ect t on
ive y ,
, i 85
23 n se
se e
e A al
al u so
so s Ne
cli t w
nic r So
al a ut
au li h
dit a W
ale A ca
s; u tor
No s Pr
nh t og
er r ra
n a m,
Te li 30
rri a -
tor n 31
y; Au
Qu C str
ee o ali
nsl u an
an n Co
d; c un
So il cil
ut o for
h n th
Au e
str H Sa
ali e fet
a; a y
Vi l an
cto t d
ria h Q
; c ua
We a lit
ste r y
rn e in
Au S He
str t alt
ali a h
a n Ca
Au d rt.
str a Na
ali r tio
an d na
Co s l
m , St
mi 2 an
ssi 9 da
on , rd
on 3 on
Sa 0 O
fet , pe
y 9 n
an 5 Di
d C scl
Qu li os
ali n ur
ty i e,
in c 95
He a Au
alt l str
h I ali
Ca n an
re, d He
73 i alt
h et on
Mi y s.
nis F 11
ter o 4
s, u "b
72 n al
,9 d an
5 a ce
Au ti d
str o sc
ali n or
an , ec

In this index to subjects, the following symbols have been used: b—box; c—checklist; f—

In 6 ar
cid 8 d",
en , 19
t 8 ba
Mo 0 rri
nit a er
ori v s
ng e to
Sy r ch
ste a an
m. g ge,
63 e 12
. s. 7,
64 6 14
,6 3-
8, B 14
80 4,
Au " 15
str b 9,
ali a 16
an l 3-
In a 16
sti n 5,
tut c 16
e e 8,
of d 17
He d 8
alt i ba
h e sel
an t in
d " e
We o m
lfa f ea
re, i su
73 n re
Au t m
str e en
alia r ts,
n v 17
Pat e 2
ien n be
t t ha
Saf i vio
ur s on
al t so
ch p ur
an r ce
gc, a s,
9, c 20
23 t bi
cli i as
nic c ,
al e 41
be , .1
ha 4 01
vio . bl
ur, 7 a
25
. m
,
1 e.
12
7 41
0,
. -
14
1 42
3,
9 ,1
15
. 93
8,
2 bo
20
4 ar
4
. d
m
1 of
oti
6 na
vat
6 m
io
. yi
n
for 1 uu
, 7 u
13 6 te
6 c e
rel r o
ap i H
se, t Mf
17 e E
4 r B
fip i M,
m a bo
t— . ar
uf 2 d
alf 0 Br
t - ist
2 ol
be 3 Ro
nc , ya
h 4 l
ma 4 Inf
rki i ir
ng. n m
9. f ar
19 o y,
.2 r 85
8. m B
30. a un
19 t da
be i be
rg i l,
Ba f 85
se o Ca
Ho r m
spi n pb
tal i el
, a ho
85 w
M n
C e Ho
d sp
C i ita
a c l,
l a 85
i l Ca
f na
o M da
r a .
n n 39
i a ,7
a g 4
e Ca
I m nb
n e er
s n ra
u t Ho
r sp
a A ita
n n l,
c a 85
e l ca
y se
F s m
e i an
a s ag
s , e
i m
b 5 en
i 2 t,
l , 14
i 5 5
t 3 ca
y b se
C m
S a an
t m ag
u d er,
d e 13
y n 9
. ca
H se
5 o mi
2 s x,
p 22
C i ch
a t a
l a m
pio 1 21
ns 6 -
. 1 12
15 d 6.
6- e 17
15 t 3-
7, e 17
16 r 5
0. m ch
16 i ec
3 n kli
ch a st.
an n 18
ge. t 2c
12 s -
0 . 18
ap 1 4c
pr 6 fra
oa 3 m
ch - ew
es 1 or
to. 6 k.
14 5 15
1. e 5-
14 v 18
2l i 7
bu d pl
dg e an
et, n .
16 c 16
0, e 2.
17 f 16
6 o 8-
co r 16
sts , 9.
an 1 17
d 5 1
be 6 str
ne i at
fit m egi
s, p es.
15 l 12
9- e 7,
16 m 14
0. e 1-
17 n 14
4 t 5,
cy a 16
cle t 8-
of i 17
ch o 2.
an n 22
ge. . 7
12 2 th
6. 5 eo
16 , rie
0- 1 s.
12 y ch
6- . an
14 1 ge
1 7 ,
sir 1 in
als . no
o 1 va
th 7 tio
eo 5 n
rie t m
s r he
of i alt
ch a h
an ll ca
ge i re,
im n ch
po g ec
na , kli
nt 1 st
poi 6 s
nt 1 ch
s. v ief
15 i ex
0. s ec
18 i uti
5 o ve,
an n 14
d .1
in f 60
div o .2
id r 04
ua . cli
ls, 1 ni
12 5 ca
9- 8 l
13 - ac
6 1 tiv
an 5 ity
d 9 ,
sta s m
ff. e on
10 t ito
9 a ri
sta l ng
ge s .5
s. o cli
12 b ni
8- e ca
12 h l
9. a au
13 v dit
0r i .
su o 5.
sta u 7-
in r 8.
ab a 10
iht l ,1
3.
14
.
21
-
23
ba
sel
ine
pe
rfo
rm
an
ce
tes
tin
g.
24
ch
cc
kli
st.
act
io
ns
,
32
def
ini
ng
,
17
i t m
a g
i M ,
f a
c m 2
r e 5
— n -
« d 2
a 6
a b .
n o 2
d o 9
s
h , d
t u
t 2 c
f 8 a
a l
r i
s c e
h s v
i u e
p k ,
, s
n 2
2 t d
6 p m
a c
o s al
c c
f a h
a n a
s d m
» pi
« i o
t w n
2 p s,
2 h 1
5
t 1 6
^ 3 -
o . 1
r 2 5
a 4 7.
o - 2
t 2 0
p 5 3
o , -
M 2 2
v 8 0
B 4
. c O
2 c m
2 p d
4 o M
i e
v t m
. e
2 s p
0 oi
d a n
k n ts
m d ,
a 1
k a 4
g c o
o t r
ve i g
r v m
n i is
a t a
n i o
ce e o
, s n
3- , fa
4, ct
3f 1 o
, 4 r
2 s,
2 h 1
3 e 0
a a -
cc k 1
o h 1
u o
nt s w
a e v
bi r k
lit v n
y i 1
1 c -
1- e 1
1 7
2 r r
c o e
o t p
m e o
p , rt
o i
n 1 n
e 2 g,
nt - 1
s, 1 1,
5- 3 1
1 8
1 i 9
c m -
oi p 1
e o 9
p r 0
ri t d
n a u
ci n sl
pl t g
ui 8 o
d , n
el 8 a
in 0 n
es - t
se 8 p
e 3 oi
g n
ui d ts
d e ,
el f 8
in i 9,
es n 2
. i 2
H t 4
in i i
ifi o n
l n c
in , o
nj m
fn 6 pl
i 3 et
a r e
n e a
d p n
o d
a r bi
d t a
v i s
e n e
r g d,
s , 6
e 6,
6 6
e 3 9
v - li
e 7 m
n 1 it
t , a
s ti
, 7 o
3 n
6 , s,
4 8 6
9 5
d , -
a 1 6
t 9 6
a 2 o
b , n
a 2 li
s 2 n
e 4 e,
s 6
, i 6
m -
6 p 6
7 9 s
s - s
y 3 e
st 0 e
t i
m d n
, s te
7 u r
0- c v
7 a e
1 l n
di ti
ni i o
c n n
al f s
in o ri
di r i
c m n
at a r
or t al
s i jo
c o u
h n r
ar n
a s al
ct y s
e s s
n t e
s e t
o m m
cs . e
, di
2 2 c
9 0 al
d 6 jo
ef u
in c r
iti l n
o i al
n, n s
2 i di
9 c n
p a ic
ro l al
gr p
a i a
m n t
s, t h
3 e w
0- r a
3 v y
1 e s,
ty n 1
p t 9.
es i 7
. o 6
2 n -
Enhancing prtitnt
cm
7 1 al
7. 8 p
7 1 r
6f - a
. 1 ct
8 8 ic
9, 2 e
1 g
3 c u
9. h id
2 n el
0 i i
8 c n
fr a e
a l s.
m 1
e p 9.
w e 7
or r 6.
k f 1
fo o 7
r r 8
a m b
ct a di
io n n
n. c ic
1 e al
7 n
9- s s
1 e k
8 e m
4 a
li p n
m e a
it r g
at f e
io o m
n r e
s. m n
1 a t
8 n p
0- c r
1 e o
8 , g
1 r r
st j a
ro i m
k n ,
e i 2
c ^ 0
a i 6,
se d 2
st i 0
u n 7
d i -
y, c 2
0 t r
8. . a
2 p
1 2 r.
0 1 2
e 2 8.
d 1
u i 5
c m 5.
at p 2
io o 0
n. r 5
2 t -
0 a 2
6- n 0
2 t 6.
0 2
7 p 0
e o 9.
n i 2
c n 1
u t 0
o s r
vt . ef
s e
u 2 r
m 1 e
m 6 n
n . c
y, 2 e
2 2 p
3 4 a
2- n
2 i el
3 n .
4 f 2
i o 1
m r 9
pl m -
e a 2
m t 2
e i 0
n o a
n n n
n af
g, s l
2 h h
0 e o
3- a s
2 . oi
1 ia
6 2 l.
c 0 2
h 3 1
ec 7
kl m -
is a 2
 2 *
2 t .
i e
m a 1
p m 9
o s
r , c
t l
a 2 i
n 8 n
t i
c c
p l i
o i a
i n n
n i s
t c
s a a
, l n
d
2 t
2 r a
1 i d
a v
s l e
e s r
e , s
e
a 6
l , e
s 1 v
o 4 e
7 n
r t
i • s
s c ,
k l
i 9
m n 5
a i ,
n c 9
a a 7
g l
e a
m v p
e a p
n l o
t u i
e n
c t
l c i
i o n
n m g
i p ,
c a
a s 1
l s 2
 8 ty,
f - 88
e 1 sk
e 3 ill
d 9 s.
b , 9.
a 2 16
c 1 5-
k 0 16
, r 6
e an
2 s d
9 i te
m s a
oti t m
vat a m
in n e
g. c m
15 e be
7- t rs
15 o hi
8, c p,
17 h 16
7- a 3
17 n wi
9, g th
17 e dr
9c , aw
pe 8 in
rfo 1 g
rm . pri
an 1 vil
ce 2 eg
me 0 es,
as , 12
uri 1 "cl
ng 2 os
, 2 e
30 , ca
m 1 lls
on 4 "
ito 3 se
rin , e
g. 1 "n
12 6 ea
,2 1 r
27 r mi
qu e ss
ali s es
ty p "
m o Co
an n ch
ag s ra
em i ne
en b Co
t, il lla
13 i bo
rat 1 se
io 0 ev
n, 8 en
20 c ts.
co o 50
de m -
r m 51
tra u ,8
ini n 0.
ng i 21
, c 1.
68 a 21
co t 2
gn i br
iti o ea
ve n kd
th . ow
eo 9 n,
rie , 40
s, 2 .5
13 6 0.
0 , 85
col 4 th
la 0 eo
bo , rie
rat 1 s.
ive 2 13
s, 4 6
26 6 co
- a m
27 b or
co o bi
m u dit
me t ies
rci c ,
al h 6,
re a 28
qu n ,7
ire g 6,
me e 17
nt . 8
s, 1 co
re 7 m
po 3 pl
ni , ai
ng 1 nt
, 7 s
49 4 by
co , pa
m 2 tie
mi 1 nt
tte 1 s
e, a an
1, d d
11 v rel
,2 e ati
5, r ve
s, m
28 p 9
,7 l .
1. i 2
77 a 8
- n
78 c c
, e o
96 . m
- p
97 7 u
l , t
i 2 e
a 0 r
i , s
s 2 ,
o 2
n . 1
2 2
o 4 ,
f . 2
f 1 6
i 5
c 7 c
e o
r c n
. o f
m i
2 p d
0 l e
5 i n
c c
c a e
o t
m i i
p o n
l n t
e s e
x . r
i v
t 5 a
y 2 l
, s
t 5 .
h 3
e b 2
o . 5
r
y 1 c
. 1 o
1 n
1 r f
3 a i
9 t d
e e
c s n
o . t
i . c
a 7 o
l 8 n
i , s
t i
y 1 d
9 e
d 3 r
a . e
t 2 d
a 1
. 9 j
u
2 c d
4 o g
n e
p s m
a e e
t n n
i s t
e u ,
n s
t . 2
, 0
1
2 3 c
5 7 o
- n
2 c s
6 o t
r n r
e s a
p e i
o n n
n t t
i s
n s .
g e
e 1
a 4
n i 8
d n con
f sul
e o tati
v r ve
i m cou
d e ncil
e d an
n d
c c co
e o m
. n mit
s tee
5 e s.
. n 78.
6 t 23
7 2
c 1 cat
o 6 ion
n 2 ,
s 8-
u c 9
m l co
e i nti
r n nu
i ity
a c of
d a ca
v l re,
i 9
s g co
o o nti
r v nu
y e ou
r s
p n qu
a a ali
n n ty
e c im
l e pr
, ov
p em
1 r en
3 o t
g se
c r e
o a qu
n m ali
s , ty
u im
m 1 pr
e 3 ov
r . em
1 en
i 4 t
n c co
v o ntr
o n act
l ti ing
v n th
e u eor
m i y,
e n 14
n g 0
t m co
. e ntr
d oll
5 i ed
. c tri
9 a als
- l se
1 e e
0 d ra
. u nd
om i egi
ise o ng,
d , 5,
co 1 8,
ntr 5 12
oll 9 -
ed - 13
tri 1 ,
als 6 14
Cor 0 .2
onia c 7
l o co
Co s m
mm t mi
uni s tte
que, , e,
79 7 29
Co . cul
ro 2 pa
nia 2 bili
l . ty,
Lia 2 19
iso 4 3
n . cul
Ser 2 tur
vic 5 e
e , fle
in 1 xib
Vic 9 le,
tor 6 19
ia, , 3
79 2 he
cor 0 alt
oni 5 h
al , ser
rep 2 vic
ort 0 e.
s, 9 20
78 c .6
- r 9.
79. e 12
23 d 4b
2 e ,
cor n 14
rec ti 0,
tiv a 19
e ll 6,
act i 19
ion n 8-
, g 19
1,4 a 9
cos n inf
t- d or
be p me
nef ri d.
its v 19
rat il 2-
19 2 ,
4 1 23
jus - sm
t. 2 all
19 2 sa
3,2 , m
11 2 ple
lea 3 col
rni - lec
ng, 2 tio
19 4 n.
4 , 25
org 1 -
ani 7 26
sat 3 da
ion a ta
al. n col
19 a lec
1- l tio
19 y n.
2.1 s 27
92 i -
b. s 28
21 , .1
1 m 70
rep e .1
oni a 72
ng, s .1
19 u 92
3 r .
saf e 22
ety m 4
. e ele
19 n ctr
1- t. oni
20f 2 c.
. 6 23
22 , ,2
5 6 4
var 6 sy
iab e ste
ilit n ms
y. t .
19 r 13
8 y da
. ta
D 2 ba
3 se
dat - s,
a, 2 21
15 4 .2
8.1 f 8.
73 u 60
acc m ac
ur i ce
ac n ss.
y. g 14
Au 1 ro
str 5 up
ali 7 s.
a. . 18
80 2 .3
cli 2 0,
nic 0 76
al d ,1
inc e 09
ide c dia
nts i gn
. s ost
68. i ic
80 o co
- n de
83 s s.
cli u 52
nic p dir
al p ect
tri o or
als n of
,8 t me
cri o dic
tic o al
al l ser
rev s vic
iew . es.
s, 1 69
7,1 9 .1
2 D 05
Un e ,
ite n 13
d m 5.
Ki a 16
ng r 0
do k dis
m. , ch
80 3 ar
- 9 ge
82 d su
Un i m
ite a ma
d g rie
Sta n s.
tes o 59
. s .
82 i 74
- s .1
83 - 48
de r .
ath e 20
s, l 7.
pat a 22
ien t 0
t. e ele
54. d ctr
62. g oni
c.
59
d e
i , d
s o
c 4 c
l 0 t
o o
s d r
u i s
r s f
e e ail
, m ur
p e
o o to
p w fol
e e lo
n r w
m gu
s e id
e n eli
e t ne
, s,
o 12
p 8 2,
e 5 17
n 8b
D re
d i sis
i v ta
s i nc
c s e
l i to
o o ch
s n an
u ge,
r o 81
e f se
e
d G al
i e so
s n cli
c e ni
o r ci
n a an
t l s
i do
n P cu
u r m
i a en
t c tat
y t io
i n,
o c 96
f e ,2
, 06
c ,2
a 7 07
r 4 dr
ug i r
ale n e
rts i s
se c .
e a
ph l 5
ar 0
m r ,
ac i 1
eu s 7
tic k 7
al -
ale m 1

rts a 7
n 8

E a
g p
ec e a
on m s
o e s
mi n i
c t v
th , e
eo ,
rie 2
s. 0 9
14 6 p
0 - r
ed 2 i
uc 0 n
ati 7 t
on , e
. d
8- 2
9, 1 m
12 8 a
5b - t
, 2 e
16 1 r
6 9 i
ch m a
an e l
ge e s
th t ,
eo i
rie n 1
s. g 4
13 s 5
0
cli a s
nic n a
al d f
au e
dit l t
, e y
26 c
c t a
l u n
d l
d st
q e ud
u p ies
a a ,
l r 52
i t ,1
t m 01
y e ,1
, n 09
t eq
1 s ui
7 , p
7 m
3 en
e 7 i,
l , 40
e 1 ,6
c 4 4,
t 8 86
r s eq
o c ui
n r p
i e m
c e en
n t
d i al
a n ert
t g s,
a c 49
r ,8
c i 4-
o t 85
l e er
l r ro
e i r
c a ca
t , us
i 5 e
o 5 sc
n , al
, 5 e,
5 10
2 b 2b
3 e er
, p ro
2 i rs,
4 d 63
e .6
e m 9,
m i 75
e o ac
r l tiv
g o e,
e g 40
n i -
c c 41
y a cl
as - r
sifi 1 e
cat 5 d
io 8 u
n, , c
10 2 i
3 3 n
cli 3 g
nic .
ia p
n r 4
pe i 2
rce o ,
pti r 1
on i 1
of, t 1
15 i ,
7- s 1
15 i 6
8 n }
co g
nfi r
de a e
nti c s
ali t p
ty, i o
5 o n
def n s
ini , e
tio
n, 1 t
40 0 o
ed 9 ,
uc -
ati 1 4
on 1 1
,8 4 -
p 4
r p 2
e r ,
v o 4
e b 6
n a ,
t b 2
i i 1
o l 1
n i
, t t
y h
8 , e
7 r
- 9 a
8 8 p
8 , e
, 1 u
1 9 t
5 5 i
7 c
, r di
v ci
7 e ne
5 n ,
t 5,
s i 6-
e o 7,
e n 14
;
a e 76
l f ,
s f 11
o e 2.
c 13
a t 4
d i an
v v d
e e ch
r n an
s e ge.
e s 14
s 1.
e , 14
v 6.
e 1 17
n 1 8
t 0 de
s - fin
1 iti
e 1 on
v 1 ,6
i , li
d mi
e 1 tat
n 1 io
c 0 ns
e b ,
. , 6-
7
2 1 re
0 5 se
, 6 ar
e ch
7 v ,
2 i m
, d ed
1 c ic
4 n al
6 c jo
, e ur
1 - na
5 b ls,
9 a 18
s ex
i e ec
n d uti
t m ve.
e e 10
,2 n ,
11 s 53
.2 e
20 n F_________________
ex s
ec u fac
uti s ilit
ve o ato
ch p r.
a i 13
m n 9.1
pio i 41.
n. o 16
20 n 3-
4, . 16
21 6 5,1
0 , 78,
ex 2 19
ec 0 9
uti e fail
ve x ure
su p mo
m e des
m r ,
ar t 87
y. p fail
23 a ure
2- n to
23 e dia
4 l gn
ex s ose
pe , or
cta 1 pre
nc 4 ve
y 7 nt,
th - 41
eo 1 fal
ry. 4 se-
13 8 po
2- e siti
13 x ve
3 t rat
ex e e,
pe r 60
rt n fati
cli a gu
nic l e,
al c 40
ju a fea
dg u sib
em s ilit
en e y,
t, c 7,2
6 o 4*
ex d 15
pe e 9-
rt 16
co 0
fee- fo an
for- r ge,
ser i 16
vic m 0.1
e, p 76
14 le
0 m G
fee e
db n ga
ack ti ps
. n in
67, g qu
68. c alit
69. h y,
14 a 15
4,1 n 6,1
76, g 63,
17 e, 22
7+ 1 5
cli 5 str
nic 5 ate
ian - gie
s. 1 s
25. 8 for
29. 7 clo
17 i sin
3- m g,
17 p 16
4,2 o 7-
27 rt 16
gen a 8,1
era n 85
l t ge
pra p ner
ctit oi al
ion n pra
ers ts ctit
. , ion
74, 1 ers
89, 8 .
22 5 14
4 F 8
foc r cli
us a nic
gro n al
up c ma
s. e, na
13. 3 ge
20. 9 me
77- f nt
78. u pro
16 n gra
4,1 d m,
98 i 2)7
fra n ,
me g 21
wo c 8,2
rk h 19
f 1 t
e 4 als
e , o
d 1 cli
b 6 nic
a 2 al
c , gov
k 1 ern
. 6 an
6 ce
7 - gov
4 1 ern
- 6 ing
7 7 bo
5 p dy,
. r 10
8 o gra
9 g ph
, r s.
e 25,
2 s 26
2 s, gui
4 1 del
7 ine
G 3 s,
o - 12
a 1 2,1
l 7 48
- 4 gui
s , lt,
e 1 96
t 7
t 5 H
i -
n 1 ha
g 7 nd
6 ove
T g r,
h o pat
e v ien
o e t.
r r 40
y n Ha
, a rva
n rd
1 c Me
3 e, dic
1 2 al
, Pra
g 1 cti
o 4 ce
a , St
l 1 ud
s 2 y.
, 4 38-
b 39,
1 s 50.
, e 56.
58. 1 m,
10 - 37-
9 2 38
err , red
or 2 esi
cla / gni
ssif h ng,
ica e 17.
tio al 42.
n, t 22
10 h 4
3,1 c t
03 a he
b r ori
scr e es
een s dd
ing y rni
crit st ge.
eri e 12
a m 71
for , 29
adv 2 he
ers 2 alt
e 3 h
eve a tnf
nts s
an
, s
n*
51. e
k
51 s
m«
b, si
er
55, n
vK
20 g,
xs
7 1
.
Ha 5
16
wt 8
1,
hor c
20
ne h
6.
effe e
20
ct. c
7
17 k
He
2 s
haz i alt

ard n h

sco , tns

re. 4 ur
86 1 mc
/ d eA
hea ef a
lth i 19
car n 73
e i (O
ma n fc
na g h).
ge a 20
me s 6
nt, y he
rol st alt
e. e h
ser m h
vic i Se
e c rvi
cli a ce
nic l s
al o Ad
go m 19
ver o 88
na b (Vi
nc c c).
e , U.
rol 2 10
e. 0 7.
12 - 20
- 2 6
13 1 He
,1 h rz
4 e be
hi a rgt
^w l M
eli t otr
ab h ai
ilit on
y. s -
19 e Hy
5- r pc
19 v ne
7* i Th
m c eo
on e ry.
ito a 13
rin o 1
g. c ho
5 r sp
re a ita
de k l
sig a cli
n. t ni
20 i ca
5 o l
saf Q gr
ety ou
sp s ps
ect u .
ru r 15
m. v 7
19 e in
5- y pa
19 s tie
7 . nt
str 7 sc
uc 1 re
tu H en
re e in
an a g
d l cri
d t ter
ia. 0 nt
51 7 set
- r in
53 e an
,5 - jgr
2b a nif
qu d nf
ali m ho
ty i sp
im s ita
pr s ls,
ov i pu
em o bli
en n c.
t, s 80
12 , ho
2. 7 sp
12 4 ita
5 h ls,
ser o s
io s m
us p all
ad i .
vei t 28
se a .6
ev l 7
en i cli
t. n ni
52 q ca
ho u l
spi i ris
tal r k
ad i m
mi e an
ssi s ag
on , e
s, 8 m
18 5 en
.3 - t
9. 8 pi
50 6 op
.8 h an
6. o .
20 s 21
7- p 7-
20 i 22
8 t 2
di a hu
ag l m
no m an
sti a -
c n fa
co a ct
de g or
s. e pri
52 m nc
,2 e ipl
es. y
Ill . o
f
I 1
0 c
i 7 h
a - a
t 1 n
r 0 g
o 8 e
g .
e 2 s
n 0 e
i 6 e
c
I c
h m h
a p a
r e n
m r g
. i e
a ,
3 l
8 i
C m
I o p
C l l
D l e
- e m
I g e
0 e n
- t
A L a
M o t
, n i
d o
5 o n
2 n i
i , m
m p
i 9 r
m 9 o
i v
n i e
i m m
r p e
y l n
, e t
m
s e s
t n e
a t e
t a
u t q
t i u
o o a
r n l
i 9 po
t 2 rti
y c ng
h ,
i a 65
m r -
p a 66
r c on
o t lin
v e e.
e r 66
m i -
e s 67
n t in
t i de
c m
i s nit
n , y
c 6 se
i 4 e
d - m
e 6 ed
n 5 ic
t i al
m de
r p fe
e o nc
p r e
o t oc
r a gn
t n us
i t aii
n p an
g o s
i in
s n di
y t ca
s s tor
t , s
e 8 se
m 9 e
. . cli
2 ni
2 2 ca
8 4 l
. l in
6 o di
3 w ca
, tor
7 r s
0 a inf
- t ec
7 e tio
1 o n
. f co
r nt
1 e rol
m n
an s he
ag e alt
er. n h
28 t ca
inf , re
or 1 ch
m 0 ar
ati . ac
on 5 ter
so 3 ist
ur . ics
ce 7 .
s 9 12
an I 3b
d n
flo n
w. o
19 v
- a
20 t
, i
14 o
8- n
14
9, D
19 if
2 f
inf u
or s
m i
ati o
on n
sy
ste T
m. h
20 e
6 o
inf r
or y
m .
ati 1
on 3
tec 4
hn i
olo n
gy n
23 o
.2 v
6, a
20 t
6 i
inf o
or n
me
d i
co n
 Mincing patient c*r»

checklists, 32,147-148,182*-18V 206. United States, 83 manuals, 7,67,68,220 O

208,212-215 ikMWMs, 121-126,121/. measurement, patient care, 5 media
123k. 124b, 125b reports, adverse events, 49,71,83-84, 98 obstetric screening criteria, 52
Atom. 122,134-136 medical defence organisations, 79,232 online policy and procedure manuals.
pcKcpbons about. 135 claims, 51 medical director, 204 medical 7,67
online reponing. 66-67
vmms. 134,133.164 errors see errors medical journals,
open disclosure. 9.10,50.95-96
iapMw.M screening 20,21,84,232
documentation. 96 operational
criteria 51-53 evidence-based research, 18 medical
hazards, 195 opinion leaders.
Wtamen Heahh Cut Group, 52b management analysis severity scale.
134.137,144.156-157. 171
inquires. 49v 85-86 ftwnct. 63.72.79- 57b, 207
technical. 135 organisational
80,232 wmpmed cue, 139 ■Mens* care medical records, 70.148 completion of.
accident model. 39/ organisational
unit, 61.62.148,164 157-158 high-risk records. 51-52
factors. 5.86.138-140 and change.
from general ward, 54 recommended for review, 53, 54 review.
124b. 130(. 138.163-165 clinical
*wrnhon>ny unsafe act", 82-83 morn 51. 55-58, 73, 74.89.98.99. 102-103.
governance, 10-11,14 culture. 191-
training supervisor, 28 bttenononal 204, 224 form modifications, 58-59
192.192b, 194-195.194b
Cbsafkabon of Diseases set ICD-10-AM limitations, 59-60 small hospitals. 217-
Organisational Learning Theory. 139-
imnimHinal indicator. 21 interventions, 220 trigger tools, 60-62 screening
140 outcomes. 9.10. 20. 52 health
18 19. 110-111. 198-199 analyss of process. 49.105.206.207.218 criteria.
care system. 37-38 outcome
determinants. 163-165 balanced diet 51-55 medical reviewers, 55-
measures. 22-23,29,170 screening
approach. 114 harness, 127,143- 58.208.219,220 medical staff see staff criteria. 53b. 218 ownership of the
144.159.168,178 strategies for. 141- medication errors. 41,64,65 MEDLINE. issue. 157-158
145.167-168.226-227 success, 120- 19 mental illness, doctors. 9 meta-
121,135.144.225.226 mttrviews,
20.164,198 intranet, hospital, 171
analysis, 7,19.20 methodology, 27
military model. 42
P
Monash University Centre for Rural Pareto Principle. 18 patient
J Health, 217.218 monitoring clinical and adverse events, 39.87,95-
activity. 5 clinician performance. 97.99,225 and change. 140-141.163-
job suitability, 132 Jobs 9.12,227 compliance. 7 165 characteristics, 124b and
Characteristics Model, 132 journals see health service performance. 1 morbidity. clinicians, 137 complaints. 28.71.77-
medical journals junior ambcal suff. 19,22.30.195 mortality. 19.22-23.30, 78 identification data, 23 passive
23.28,59,69.146. 157.164.178 111. 195 recipients, 87-88 privacy. 25-26
30-day. 23 motivating, clinicians. 157- resistance, 41 sampling, 25 patient
K 158.177-179 motivation theories. 131- caie, 52,76,99,179 dimensions, 5,170
133 motivation to change and learn, gaps in quality. 155 improving,
King Edward Memorial Hospital. Penh.
85 130. 133.157 3.140-141 meetings, 50 patient-
multidisciplinary team. 7.21. centred, 13 standardisation, 76

L 76.100.139, 145.208
mundane adverse events see adverse
patient deaths, 54.72.207 patient
management plan, 40 patient
laboratory results. 60-61 leadership. events, mundane outcomes see outcomes patient
10.26.85.86.124b. 138 learning N- representation, 162 patient safely. 4-
culture. 194 Learning from Defects 5.50,87 alerts. 80.81,83 culture. 191-
tool. 102 learning, small group. 9 National Coordinating Council lor 201
learning theories. 120.139-140 Medication Error Reporting and levels. 194b information, 80-81
lectures. 50.177 legal action. 5.49.50- Prevention, 57,58b monitoring. 13,87-88 notices. 80-81
51.79.99 legislation. 5. 11.49. 72. National Institute of Clinical Studies, practice. 110-111 spectrum. 195-197
78.107 length of stay. 55 bterature 18 National Primary Care Research and surveys. 198-199 patient safety
search. 143.159.181 hnganon tee legal Development Centre, UK, 21 "near manager. 100 patient safety programs.
action misses", 28,63,64,65,82,192,195 177-179.179c patient satisfaction, 20,
negative consequences of change, 22, 77-78.89 patient surveys, 20,78
38,119, 168,209,223 negligence, 50 peer review, 27,85.135,220
M____________________
neonatal units. 61,81 Netherlands, 39 performance, 1,132 clinical, 21,27,
malpractice. United States. 79 New South Wales, 80 Quality Use of 29,30 monitoring, 5,9 standards, 4
management. 11.28. 1)8-1)9.192.212 Medicines in Australian Hospitals, 30 clinicians, 21,27,29, 30.174 dealing
board. 11.14,203.204-205,206.211 New Zealand, 39.80,233 non-clinical with poor performance. 14
structure, 25 suppon for change, 160 personnel. 23 non-compliance, 196 improvement. 172/ measures. 170
management causation scale, 56,56b non-randomised trials, 6 non-reponing monitoring. 1. 5.9. 12.176 variation
Manchester Patient Safety Framework, of incidents, 41,65 Nonhem Territory, graphing, 25 personal digital
199.199b 80 nursing staff. 55.146.205 director, assistants, 23 pharmaceutical alerts,
204 education, 65 failure in care 49.84-85 pharmaceutical companies.
mandatory requirements, 49,72,81,83
delivery. 50 137 pharmacy committcc. 28 plan-do-

23*
study-act cyclc. 27.138.172.172/
Planned Behaviour, Theory of. 131
qittkty manager. 205
Quality and S^eiy m Health Gas. 59
s
political and economic theones. 140 Quality Um of Medum to Amalmm safety alerts and notices, 80,81,83
preadmission clinic. 55 preventability Hocp*ak. 30 safety assessment code. 100 safety at
scalc. 58b preventive health strategies. Queensland. 52,80.102 work, 12
50.88 prioritising action adverse events. questionnaire, safety tftiudr. IM-M see ciso patient safety safety
103-105. 109-114. 113b. 225 committee, 28 safety framework,
external programs application, 112-113 R 199.199b safety programs see patient
risk management, 43 privacy, patient. safety programs scientific evidence.
25-26 privileges, withdrawing. 12 randomised controlled wtk, 6.22. 6.19.81.172.176, 179.226
privileging see credentialling and 111-112.127,134,174 scientific literature see literature
privileging probability, adverse events. rapid response teams. 112 search Scottish IntercoUegule
98,104 procedures, standardising. 146- re-ceitifiadon, 8.27 Guidelines Network. 20
147 process control charts, 25 process, Reason. James. 39.98 sentinel event, 71-74.89.98
indicators. 29 process measures, 22- recovery times. 76 Sentinel Event Alerts. 72.73.74
23,29,170 Process Redesign Theory, 139 reference panel. 219-220 %vemy scale. 57k. 207
process theories, 127-128 professional daft lengths. 13,40.86
references, dmictms, 8
development, 5,8-9,12, 14,137 - SIGN, see Scottish Intercollegiate
regulating. 1.14
professional experience, 8 professional Guidelines
Reinforcement Theoty 131 it . I
organisations. 29 professional relapse. 174.175.176 nenpont
qualifications. 8 * relatives and adverse events. 95.96-97. gmubux laboratories. 9
professional responsibility. 11 99.225 sue visas. 20
professionals motivating, 133,157-158 reminders. 147-148.208 reponing small group learning. 9
rewarding, 133 project name, 171 advene events. 177 audit results, 25- small-scale tests. 23.24-
project officer. 160 26.29 behaviour. 68 25.161.172.227
project team. 161-163,166-168.174,175 clinical governance program. 11. 189- SMART goals. 166-167
evaluate the project, 176 protocols. 190 sodal influence theories. 136-137
20.146 publications, 18.20,21,84,159 clinical incident. 63-71.193 incomplete social mesxaoas. 38.136-138
punitive response. 41.46.69,83,99,149. nd bistil, 66 limitations, 65-66 online, Social Learning Theory, 136
211-212 66-67 Social Network and Influence Theory,
under-repoiting, 41.65,69 commercial 136-137
Q requirements, 49 culture. 193 suuupotocal content, 125b. 163-165
South Australia. 80. 102 Off. 23. 157
mandatory, 49
qualifications, 8 ad advene events. 97.99. 211.225 and
time, 71
Quality in Australian Health Care change implementation, 160 and
voluntary; 49.67 research. 79.120.159
Study, 39, 56.109 clinical risk management, 206-207,
Residential Aged Cm Cervrid
error cause scale, 102b quality of care. 209,226
Communique 79
medical staff group report, Wimmera,
3-4,10.85-86 components, 20 gaps. resources. 7.13,20-21.25.26.110-111.
108 supervision, 40
156,163.208-209 measuring, 22 quality 125b. 210 allocation, 39-40.59-
see also clinicians; doctors; junior
improvement, 11,13-14,17-35. 60.205.223.226 analysis. 165-166 medical staff; nursing stafT staff-
111.138.145, 209-210 clinical area developed elsewhere. 227 monitoring. patient interface, 42 stakeholder 156-
selection, 18-19 collaboratives. 26-27 176 saving. 18 responsibility, defining. 157,161,164,171,176 standards,
committees. 5.28,206 cycle. 17/ 12 responsible autonomy; 10 results. 1,95,104 of care, 19.21 clinical
definition, 17,20-22 7 clinkal audit. 13.24-25 replicating. performance, 4 data collection, 24-25
evidence and theory for improving, 119 25 revenue. 209 nsk procedural, 146-147 professional, 29
hospitals, 122,125 assessment. 104-105.104t. 233 Standards Australia/New Zealand,
important points, 33, 223, 224 detecting. 40-93.225 104,233 storage of data, 24 strategic
siraicgics. 145 rienafywg ra of risk. 49.78.88. 156-17* plans, 1 stroke, 28,77/, 181-182
Wimmera model, 26f, 43-44, 45/ quality mtaong, 167-168 nsk —f iiW!W. 5. structural change, 160 structural
improvement programs. 191*201 10.11.12.14.28 advent events. 37-47 indicators, 29 substance abuse,
components. 19-20, 179 imponant important points, 46,223,223 proactive, doctors, 9 suicide, inpatient, 72
points. 200 motivating doctors, 157- 49-50.78-88.179 important points, supervision of clinicians, 12,13 intern
158,177-179. 179f 89,225 ■ process, 43.43f, 44 ■active, training, 28 surgery, 27-28 deaths, 54
quality management. 119. 138-139. 49.50-78 important points. 89,224 wrong-site, 72 surveillance committee.
145-149 Wimmera model. 43(. 44,45/. 89 see abs 58.69,74,105, 208,209
clinical risk management program risk statutory immunity, 107-108 surveys,
register. 105.1061.210 role models. 134 13,20,71.164.198 presurvey
root-cause analysis. 71.73.98,100-102 questionnaire. 20 system failure, 85
bmitanans. 101-102 Royal Australasian identifying gaps, 156 simplifying
College of Surgeons. 27 Royal Children* system, 146 system performance. 170
Hospital, Melbourne, 54 Royal College of system redesign, 17,179 systematic
Physicians. 28 determinant analysis, 122 systems
analysis. 98-100 systems engineering.
120

T
tables of results and trends, 26 uigei
Enhancing
group. 122.130(. patient
171 analysis. care
164
characteristics, 123b. 134,135.160
education, 166 identifying. 156-166
informing, 158-159 vision for change,
158-159,163-165 teams, 11-12, 162-
163 changes, 145
quality improvemenl, 26, 137
see.also project team telephone
reporting, 67 tests, small-scale, 23,24-
25,172,227 TGA see Therapeutic Goods
Administration theories of change in
health care and individuals, 129-136
attitude theories — Theory of Planned
Behaviour, 131 cognitive theories, 130
education theories, 130 motivation
theories, 131 expectancy theory, 132-
133 Goal-setting Theory, 131 Herzbergt
Motivation-Hygiene Theory, 131
Innovation Diffusion Theory, 134
Jobs Characteristics Model, 132
motivating professionals, 133
Reinforcement Theory, 131 Three
Needs Theory, 131 organisational
context, 138-140 complexity theory,
139 innovative organisations, 138
integrated care, 139 organisational
culture, 140 organisational learning,
139-140 Process Redesign Theory,
139 quality management. 138-139
and patients. 140-141 political and
economic context, 140 contracting
theory. 140 economic
(reimbursement) theory, 140 process
theories, 127-128 and social
interaction, 136-138 communication
theories, 136 leadership, 138
professional development, 137 Social
Learning Theory, 136 social network
and inlluence theories, 136-137
team effectiveness. 137 Stages of
Change Theory. 128-129.130t
Therapeutic Goods Administration, 85
Three Needs Theory, 131 lime. 14. 76
timetables, 169
total quality management, 119,138-
139, 145
training, 68, 166,178,218
clinicians, 8 data analysis,
26 treatment errors, 41
trends. 24,26 mailing
change, 144,161 trigger
tools, 60-62,60b
 u
ENHAN
and Paediatric Mortality and Morbidity,
23,
under-reporting, 41,65 unintended 78
Department of Health. 78 Department

CING
injury, 36 United Kingdom.
21,27.39.54 databases, 80-81 Imperial of Human Services. 78.95, 96-97.105
College London, 99 Inquiries into Clinical Risk Management Strategy,
health services, 85-86 Lothian Surgical 72-73
Incident Severity Rating Classification, 68
PATIENT
Audit in Scotland, 27 National Health
Service, 3 Belter Care. Better Value small hospitals program, 220-221 State Coroneri
Indicators, 30 Central Alerting System, Office, 79 Victorian Consultative Council on

CARE
85 Medicine and Healthcare Products Anaesthetics Mortality and Morbidity, 78,232
Regulatory Agency. 84-85 National Victorian Maternity Services Performance
Institute for Health and Clinical Indicators, 30
Excellence. 20 Victorian Minister for Health, 78, 107 Victorian
Scotland Clinical Indicators. 30 Surgical Consultative Council. 78. 232 If you would like to purchase copies
National Health Service Trust, 65 voluntary reporting, 27,67, 100
of this publication, please contact
National Patient Safety Agency, 80
National Primary Care Research and
w MJA Books at AMPCo 277 Clarence
Development Centre, 21 National Street, Sydney NSW 2000
wirnng timet. 19
Reporting and Learning System, 80-81
websaes. 59.77.99.100.102,210 Telephone: 02 9562 6666 •
National Sentinel Audit of Stroke, 28
VMaUfnJtufuab.80
Royal College of Physicians' Clinical sales@ampco.com.au
Audfe of Suipcal Monday. 27 Western
Effectiveness and Evaluation Unit.*28 or visit our secure online medical bookshop
Victoria Divaton of Gcneoi Practice.
Scottish Audit of Surgical Mortality, 27
217,218.220 whisdeblowtis. 85 whok- at http://shop.mja.com.au
Scottish Intercollegiate Guidelines
of-pnoice mhOu. 27-29 Wunmera Hokh
Network. 20 United States.
Care Gtoup. vu. 23.52.
52.135.157.173 5 Million Lives
78,84,175.204-207 advene event
Campaign. 112-113 Agency for Healthcare
analysis. 57.89.102-103. 211.229-
Research and Quality, 4,20.21.40,58.210
231
Quality Indicators, 30 California Insurance form, 56 changc strategies, 145-149
Feasibility Study, 52 California Medical Management clinical incident reporting. 66-70.70f If You Need to Get Published, We Can
Analysis, 52 dinial indicators, 31 clinical pathways. Help!
Department of Health and Human Sen-ices, 20 76f. Tlj. 179-194 discharge summary, Print and Online Publishing Opportunities
Harvard Medical Practice Study, 38-39,50. 51, 51b, 59.74 general practitioner tmfhark fonn, are available with AMPCo.
55. 56. 58, 103, 103b, 109, 207 Institute for 74.75f Patient Care Commute*.
Healthcare Improvement. 27, 61. 113 105.107-100. 108b
(Australasian Medical Publishing Company Pty Ltd)

Institute of Medicine. 191 Joint Commission on quality improvement and risk AMPCo offers a wide range of
Accreditation of Hcalthcarc Organizations, 30, 71- management modd. 26f. 43(. 44.45f. specialised services to suit
73,86, 95,98,113,191 46.89 clients wanting to produce
NASA see United Sutes. National
risk assessment matrix, 104c medical education and support
Aeronautical Space Administration
screening criteria emergency
National Aeronautical Space material. We provide total
department. 55k. 63f inpatients. 52b
Administration. 83 National Center for publishing solutions — our team
medical record. 54.58-59.62f. 217-
Palienl Safety, 80 National Guidelines 220 small hospitals program, 217-221 of specialists can help with
Clearinghouse. US. 20 National website. 77 medical copywriting, peer review,
Hospital Quality Measures, 30, 71- Wimmera Clinical Ride Matagemem editing, design, publishing and
73.86.95.98.113.191 Patient Safety Project. 232 work environment. distribution.
Network. 210 Pennsylvania, Patient 39.49.195 working conditions.
Safety Authority, 83 Safety Assessment All topics related to health and medicine
12.14.96
Code score, 82,86 sentinel reporting, latent (actors. 86-87 workloads. will be considered. Print and Web • Books •
71-72,73 13.40.86.88.159 Warld Health Journals • Newsletters • Supplements
VA-NASA Patient Safety Reporting
System. 83 Organization. 3 Reprints • Focus Documents • Monographs
Enhancing patient care
Veterans Health Administration Please contact MJA Editorial Department:
Health Care Failure Mode and Effect
Telephone: 02 9562 6666 • Email:
Analysis. 86-87
medjaust@ampco.com.au
incident reporting system. 82-83,86,
98.100 University College London, 99
unsafe acts, 82-83.193

V
variance analysis, 76-77,89
Victoria, 11.52,54.79.80.101 102.
174-175
Consultative Council on Obstetric
ENHANCING
PATIENT C A R E
We all want the best possible standard of care for our patients.
But high-quality, safe health care can be difficult to achieve. Above all, it requires
cultural change, which in turn can only occur through the dedicated, consistent
efforts of clinical safety champions.
Alan Wolff and Sally Taylor are two such champions. Based on their well recognised
work in the Wimmera Health Care Group in Victoria, Australia, they have created a
commonsense guide to quality improvement and risk management. Their experience
at a regional hospital has led to a pragmatic framework that guides other health
services through the relevant evidence and theory, down to the finest details.
Enhancing Patient Care will be of use to anyone who wants to set up or improve a
quality improvement and risk management program, regardless of size and budget.

About the authors

Alan Wolff is the Director of Medical Services at Wimmera Base Hospital. He has 27 years' experience
in medical administration, and has a research interest in quality improvement and risk management
methods and their use in hospitals. He was involved in the development of Wimmera Base Hospital's
adverse event screening program, which began in June 1989 and was , Av' one of the first
commenced in an Australian hospital.

Sally Taylor has worked in clinical risk management

for 10 years and is currently the Clinical Risk Manager of the Wimmera Health Care Group. She has
experience ' with quality and risk programs in both regional and small rural hospitals. She has a clinical
background and qualifications in midwifery and inter care nursing.
Mjtm