Safety in the operating theatre - Part 2: H u m a n

error and organisational failure

J. Reason

Over the past decade, anaesthetists and h u m a n factors specialists have worked together
to find ways of minimising the h u m a n contribution to anaesthetic mishaps. As in the
functionally similar fields of aviation, process control and military operations, it is found
that errors are not confined to those at the 'sharp end'. In c o m m o n with other complex
and well-defended technologies, anaesthetic accidents usually result from the often
unforeseeable combination of h u m a n and organisational failures in the presence of some
weakness or gap in the system's m a n y barriers and safeguards. Psychological factors
such as inattention, distraction and forgetfulness are the last and often the least
manageable aspects of the accident sequence. W h e r e a s individual unsafe acts are h a r d to
predict and control, the organisational and contextual factors that give rise to them are
present before the occurrence of an incident or accident. As such, they are prime
candidates for treatment. E r r o r s at the sharp end are symptomatic of both h u m a n
fallibility and underlying organisational failings. Fallibility is here to stay. Organisational
and local problems, in contrast, are both diagnosable and manageable.

Introduction • Shifting, ill-defined, or competing goals.

• The need to maintain an up-to-date 'mental model' of
What do anaesthetists have in common with flight
what is often a rapidly changing situation.
crews, air traffic controllers, nuclear power plant opera-
• Reliance upon indirect or inferred indications.
tors, fire chiefs and battle tank commanders? David
• Ill-structured problems.
Gaba, 1an anaesthetist and a pioneer in the human factors
• Actions having immediate and multiple consequences.
of anaesthesia, claims that the practice of anaesthetics
• Moments of intense time stress interleaved with long
has more basic similarities to these activities than to other
periods of routine activity.
branches of medicine, excepting perhaps the related
fields of intensive and emergency care. These shared • Sophisticated technologies with many redundancies.
characteristics include the following: • Complex and sometimes confusing human-machine
interfaces.
• Uncertain, dynamic environments. • High stakes.
• Multiple sources of concurrent information (i.e., many • Multiple players with differing priorities.
data streams). • A working environment highly influenced by group
norms and organisational culture.
J a m e s Reason, PhD, Department of Psychology, University of One important difference between anaesthetics and
Manchester, Manchester, M13 9PL, UK these other activities, however, is the consequence of
Current Anaesthesia and Critical Care (1995)6, 121-126
© PearsonProfessionalLtd 1995 121
122 CURRENTANAESTHESIAAND CRITICALCARE

human failure. The release of radioactive material from a Of more significance from a human factors viewpoint
nuclear power plant, as at Chemobyl, or the destruction is the percentage of surgical cases in which, despite prior
of large commercial aircraft, as at Tenerife, creates planning, some unanticipated problem will arise. Cooper
an immediate public demand for investigation and and his co-workers estimated that 18% of cases will
remediation. The accidental death of a single patient involve an unexpected problem requiring intervention by
during surgery or shortly after usually attracts little atten- the anaesthetist, and 3-5% of cases will involve a serious
tion beyond the hospital concerned and the immediate unplanned event calling for substantial anaesthetic inter-
family. It is no coincidence, therefore, that until recently vention. 6 This rate of problem occurrence is much higher
human factors research has focused largely upon the high- than would be expected by pilots, for example, and indi-
consequence fields of aviation, nuclear power generation cates that one of the key features of an anaesthetist's
and military operations. These were the people that had skill is to know when and how to intervene to thwart an
the resources and the political muscle to fund the work. accident sequence. Of the three phases of surgical anaes-
Over the past decade, there has been a growing aware- thesia - induction, maintenance and emergence - 45%
ness on the part of both anaesthetists and psychologists of incidents occur during the maintenance phase. 7 This
that what has been learned from these high-profile, non- suggests that patient monitoring problems, along with
medical accident investigations and their associated very high workload in the event of an emergency, can
research applies remarkably well to the study and pre- make excessively high demands upon the limited
vention of anaesthetic mishaps. This paper takes these attentional resources of the anaesthetist.
domain similarities as its starting point, and considers
how knowledge of the human contribution to system
Classifying human failures
breakdown, acquired in these other fields, can be use-
fully applied to anaesthetics. There is no one error taxonomy. Different error classifi-
cations serve different needs. In many domains of appli-
cation, two kinds of categorisation are used together: a
The nature and likelihood of anaesthetic classification by consequences, and a classification by
mishaps psychological origins. In the case of anaesthetics, a
Surveys give somewhat differing values for the involve- consequential classification would identify which aspect
ment of human error in anaesthetic incidents and acci- of the anaesthetist's performance was less than adequate
dents, but there is growing agreement that the figure is (e.g. not checking equipment, wrong intubation, missing
b e t w e e n 7 0 - 8 0 % . 2,3 Data from the Australian Incident critical signs, inappropriate dosage, misinterpreting
Monitoring Study, 3 based upon 2000 anaesthetic inci- rapidly changing physiological parameters, failing to
dents, identified the following as the 12 most commonly recognise complications associated with congestive heart
occurring contributing factors. disease, carrying out an ill-advised intervention, etc.). A
psychological classification, on the other hand, would
• Misjudgement (16%) focus upon the mental antecedents of the error. It is this
• Failure to check equipment (13%) latter type of classification that will be considered here.
• Fault of technique (13%) Three distinctions are important.
• Other human factors problems (13%)
• Other equipment problem (13%)
• Inattention (12%) 1. Slips and lapses versus mistakes
• Haste (12%) There are many ways of defining error) For our present
• Inexperience (11%) purposes, we can say that an error is the failure of
• Communication problem (9%) planned actions to achieve their desired goal. There are
• Inadequate pre-operative assessment (7%) basically two ways in which this failure can occur:
• Monitor problem (6%)
• Inadequate pre-operative preparation (4%). • The plan is adequate, but the associated actions do not
go as intended. These are failures of execution and
The prominence of equipment-related problems in this are commonly termed slips and lapses. Slips relate to
list is in keeping with earlier findings that 48% of anaes- observable actions and are associated with attentional
thetists use new equipment without reading the manual, failures. Lapses are more internal events and relate to
and 60% do not follow the manufacturer's check proce- failures of memory.
dure. 4 Mayor and Eaton reported that 30--41% of anaes- • The actions may go entirely as planned, but the plan is
thetists perform no checks at all. 5 inadequate to achieve its intended outcome. These are
As Gaba points out, estimates of the frequency of failures of intention, termed mistakes.
adverse outcomes related to anaesthetic care are very
difficult to obtain. 1The available data suggest that deaths All errors involve some kind of deviation. In the case
due at least in part to anaesthetic factors are of the order of slips, lapses, and fumbles, actions deviate from the
of 1 in 2000 cases. However, deaths due solely to anaes- current intention. Here, the failure occurs at the level of
thesia lie somewhere between 1 in 100 000 and 200 000 execution. In the case of mistakes, the actions may go
entirely as planned, but the plan itself deviates from
cases.
 SAFETYIN THE OPERATINGTHEATRE- PART 2 123

some adequate path towards its intended goal. Here, the intended. Deliberate violations differ from errors in a
failure lies at a higher level: with the mental processes number of important ways.
involved in planning, formulating intentions, judging
• Whereas errors arise primarily from informational
and problem solving.
problems (forgetting, inattention, incomplete knowl-
Slips and lapses occur during the largely automatic
edge, etc.), violations are more generally associated with
performance of some routine task, usually in familiar
motivational problems (low morale, poor supervisory
surroundings. They are almost invariably associated with
examples, perceived lack of concern, the failure to re-
some form of attentional capture, either distraction from
ward compliance and sanction non-compliance, etc.).
the immediate surroundings or preoccupation. They are
• Errors can be explained by what goes on in the mind
also provoked by change, either in the current plan of
of an individual, but violations occur in a regulated
action or in the immediate surroundings. 9
social context.
Mistakes can begin to occur once a problem has been
• Errors can be reduced by improving the quality and
detected. A problem is anything that requires a change or
delivery of the necessary information within the
alteration of the plan. Mistakes can be further subdivided
workplace. Violations generally require motivational
into two categories: rule-based mistakes and knowledge-
and organisational remedies.
based mistakes.
Rule-based mistakes. These occur in relation to familiar
3. Active versus latent failures
or trained-for problems. A large part of the anaesthetist' s
expertise is made up of rules of thumb or heuristics of the The distinction between active and latent failures owes a
kind: if X (local signs of a problem exist) then it is great deal to Mr Justice Sheen's observations regarding
probably Y (a particular condition to be managed), or if the capsize of the Herald of Free Enterprise. In his
X (local signs) then do Y (a particular intervention). inquiry report, he wrote: 11
Human beings are furious pattern matchers. We are
At first sight the faults which led to this disaster were
extremely good at making rapid and largely automatic
t h e . . , errors of omission on the part of the Master, the
assessments of complex situations based upon matching
Chief Officer and the assistant bosun . . . But a full
features of the world to patterns stored in long-term
investigation into the circumstances of the disaster
memory. But this process can go wrong in two ways. We
leads inexorably to the conclusion that the underlying
can misapply a good rule (i.e. one that is frequently ap-
or cardinal faults lay higher up in the Company . . .
plicable) because we fail to notice the contraindications.
From top to bottom the body corporate was infected
Or we can apply a bad rule that has remained uncorrected
with the disease of sloppiness.
in our stored repertoire of problem solutions.
Here, the distinction between active and latent failures
Knowledge-based mistakes. These occur when the prac-
is made very clear. The active failures - the immediate
titioner encounters a novel situation that lies outside the
causes of the capsize - were various errors on the part of
range of his or her stock of pre-packaged problem-
the ship's officers and crew. But, as the inquiry revealed,
solving routines. Under these conditions, practitioners
the Herald was a 'sick' ship even before it sailed from
are forced to resort to slow, effortful, on-line reasoning.
Zeebrugge on 6 March 1987.
This process is extremely error-prone for several rea-
Active failures are unsafe acts (errors and violations)
sons. First, our capacity for conscious thought is highly
committed by those at the 'sharp end' of the system
resource-limited; we can only attend to and manipulate
(e.g. anaesthetists, surgeons, nurses). They are the people
one or two discrete items at any one time. Second, we
whose actions can have immediate adverse consequences.
have to rely upon a mental model of the current situation
Latent failures are created as the result of decisions,
that is nearly always incomplete and, in parts, incorrect.
taken at the higher echelons of the organisation. Their
Third, we have a marked tendency in these circum-
damaging consequences may lie dormant for a long time,
stances to 'fixate' upon a particular hunch or hypothesis
only becoming evident when they combine with active
and then select features of the world to support it, while
failures and local triggering factors to breach the sys-
neglecting contradictory evidence. This has been called
tem's many defences.
'confirmation bias' or 'cognitive lock-up' and has been
Thus, the distinction between active and latent failures
frequently observed in nuclear power plant operators and
rests upon two considerations: first, the length of time
others during attempts to recover from an emergency. 1°
before the failures have a bad outcome and second,
where in the organisation the failures occur. Generally,
2. Errors versus violations active failures are committed by those in direct contact
with the patient, and latent failures occur within the
Violations are deviations from safe operating practices, organisational and management spheres.
procedures, standards or rules. Such deviations can either
be deliberate or erroneous (e.g. speeding without being
aware of either the speed or the restriction). However, we
Some problems with the interpretation of
are mostly interested in deliberate violations, where the
human error data
actions (though not the possible bad consequences) were The statement that around 80% of anaesthetic incidents
124 CURRENT ANAESTHESIA AND CRITICAL CARE

involve some type of human failure is potentially human mind is prone to suppose the existence of more
misleading. Taken at its face value, it suggests that order and regularity in the world than it finds', iv One
anaesthetists have a major human error problem, though of the many ways of simplifying complex events is to
similar values are found in most other domains as well. 12 assume a symmetry of magnitude between causes and
A natural reaction to these high error numbers, both in consequences. When confronted with horrific man-made
anaesthesia and elsewhere, is to direct the majority of catastrophes, like Bhopal and Chernobyl, or the accidental
remedial measures towards the 'sharp end' (i.e. those in death of a young, healthy patient during minor surgery, it
direct contact with the patient): to 'blame and train', to seems natural to look for some equally monstrous act of
write additional procedures and to engineer people more irresponsibility or incompetence as the primary cause.
and more out of the loop with further automation and What we usually find, however, is the chance and largely
'intelligent' displays. But these person-specific measures unforeseeable concatenation of many different causal
have little impact on well-trained, experienced and factors, none of them sufficient or even especially re-
highly motivated professionals. Indeed, they can sow the markable by themselves, but each necessary to bring
seeds of future m i s h a p s . 13 about the outcome.
There are a number of reasons why people at the As we shall see below, errors are not so much causes
'sharp end' get blamed for bad outcomes. 13,14 as consequences. 13,21 The contributing errors, just as
First, they are obvious targets; their actions were usu- much as their bad outcomes, require an explanation.
ally those closest in time and space to the bad outcome. Errors are the product of a chain of causes, in which the
Second, human beings are prone to the illusion of free individual psychological factors (momentary inattention,
will. People, especially in Western cultures, place great forgetting, haste, etc.) are the last and often the least
value in the belief that they are the makers of their own manageable link.
fates. Naturally, they also attribute this autonomy to
other people. They too are seen as being able to choose
between right and wrong, between correct and erroneous Modelling the aetiology of organisational
actions. But no one chooses to make an error, nor are accidents
all the circumstances influencing performance under a
The thesis to be presented in the remainder of this paper
person's control.
is that anaesthetic accidents, in common with accidents
Third, it is extremely difficult to trace the causes of
in other low-risk, high-hazard systems, are usually orga-
accidents back to their organisational roots. The natural
nisational accidents, i.e. multiple-cause events whose
tendency of most accident investigators is to stop the
origins can be traced to decisions taken some time before
search as soon as they have identified some less than
the accident. The Australian Incident Monitoring Study TM
adequate performance on the part of those on the spot.
found that system-based or organisational factors were
They are not to be blamed for this, since the tools for
implicated in 90% of the incidents (or 97% if human
carrying out such in-depth analyses are only now being
factors are included).
fashioned. Moreover, the legal aspects of such inquiries
The technological advances of the last 20 years, par-
are usually best satisfied by the identification of 'respon-
ticularly in regard to engineered safety features, have
sible' individuals.
made many hazardous systems largely proof against
Fourth, reviewers of past events are subject to hind-
single failures, either human or technical. In order to
sight bias. ~5 Knowledge of the outcome causes us to
breach all defences, it now requires the unlikely combi-
simplify the problems facing the practitioner, who was
nation of several contributing factors.
armed only with foresight. Mistakes are apparently easy
The aetiology of an organisational accident is shown
to spot in retrospect, but extremely difficult to detect at
in the Figure. A case study illustrating the ways in which
the time. In simulated nuclear power plant emergencies,
these various organisational and human factors combine
mistakes were rarely spotted by their perpetrators. It
to create an anaesthetic accident is described in detail
usually takes someone else, with a fresh view of the
elsewhere. 19The direction of causality is from left to right.
situation, to detect a deviation from some adequate path.
In a recent study, two groups of anaesthetists were • The accident sequence begins with the negative conse-
asked to judge the quality of care described in a set of quences of organisational processes (i.e. decisions
written cases. 16 The description of the events was the concerned with planning, scheduling, forecasting, de-
same for both groups, but the outcomes seen by one signing, specifying, communicating, regulating, main-
group were bad, while the outcomes for the other group taining, etc).
were neutral. The judges consistently rated the perform- • The latent failures so created are transmitted along
ance in the bad outcome cases as substandard, whereas various organisational and departmental pathways to
the identical care provided in the neutral outcome cases the workplace (e.g. the operating theatre or intensive
was assessed as being adequate. As Cook and Woods 13 care unit) where they create the local conditions (e.g.
point out, 'the judgement of whether or not a human undermanning, fatigue, technical problems, high work-
error occurred is critically dependent on knowledge of load, poor communication, conflicting goals, inexperi-
the outcome, something that is impossible before the fact'. ence, low morale, teamwork deficiencies, etc.) that
In 1620, Sir Francis Bacon observed that: ' . . . the promote the commission of errors and violations.
 S A F E T Y IN T H E O P E R A T I N G T H E A T R E - P A R T 2 125

Organisation Workplace Person Defences Bad

outcome

Fig.--Stages in the development of an organisational accident.

• Many of these unsafe acts are likely to be committed, so on. The model shown in the Figure relates to a given
but only very few of them will penetrate the defences hospital, but it must be appreciated that the reality is
to produce damaging consequences for a patient. The considerably more complex, with influences from other
fact that engineered safety features, standards, controls, organisations impinging on the sequence at many different
procedures and the like can be deficient due to latent points.
as well as active failures is shown by the arrow con-
necting organisational processes directly to defences.
Remedial implications of the model
The model presents the people at the sharp end (the
anaesthetists, surgeons and nurses) as the inheritors Unlike theories in the natural sciences, whose value is
rather than as the instigators of an accident sequence. assessed by the amount of experimental interest they
This may seem as if the 'blame' for accidents has been provoke, theories in the safety sciences are judged by the
shifted from the sharp end to the system managers. But much harsher criterion of practical utility. In what ways
this is not the case for the following reasons. can the organisational accident model lead to safer
anaesthetic practice?
• The attribution of blame, though often emotionally
The key to effective safety management in any haz-
satisfying, hardly ever translates into effective counter-
ardous enterprise is to target what are invariably
measures. Blame implies delinquency. Delinquency is
limited remedial resources at the most tractable problems:
normally dealt with by exhortations and sanctions. But
in short, to manage the manageable. In most organi-
these are wholly inappropriate if the individuals
sations, a disproportionate amount of these resources is
concerned did not choose to err in the first place.
directed at individual practitioners in an effort to prevent
• High-level decisions are shaped by economic, political
the recurrence of past errors, through sanctions, exhorta-
and financial constraints. Like designs, decisions are
tions, stricter procedures, tighter selection, additional
always a compromise. It is thus taken as axiomatic
training, improved certification, and the like. But these
that all strategic decisions will carry some negative
measures are only appropriate if the people who commit
safety consequences for some part of the system. This
the active failures are especially error-prone, inexperi-
is not to say that all such decisions are flawed, though
enced, undermotivated and ill-trained. This is rarely the
some of them will be. But even those decisions judged
case, either in anaesthesia or in the fields of aviation,
at the time as being good ones will carry a potential
process control and military command. A common fea-
downside for someone, somewhere in the system.
ture of all of these domains is that the best people can
Resources, for example, are rarely allocated evenly.
sometimes make the worst mistakes.
There are nearly always losers. In judging uncertain
Central to the accident model presented earlier is the
futures, it is inevitable that some of the shots will
notion that the psychological antecedents of unsafe acts
be called wrongly. We cannot prevent the creation
(i.e. what goes on in the head of the practitioner) are -
of latent failures, we can only make their adverse
consequences visible before they combine with local beyond a certain point - extremely difficult to control.
triggers to breach the system's defences. Distraction, momentary inattention, forgetting, losing the
picture, preoccupation and fixation are entirely natural
These organisational root causes are further compli- human reactions to the kind of working environment
cated by the fact that the medical system as a whole in- described at the beginning of this paper. What is remark-
volves many interdependent organisations: legislators, able is not that dangerous errors happen, but that they
manufacturers, maintainers, administrators, medical de- happen so rarely.
fence and standards organisations, professional bodies, Whereas active failures at the sharp end are unpredict-
civil servants, medical institutions, primary carers, and able in their precise details and therefore hard to manage,
126 CURRENT ANAESTHESIA AND CRITICAL CARE

latent failures existing within the work context and the remedial efforts switches from dimension to dimension
institution at large are, by definition, present before the as previously salient factors improve and new ones come
occurrence of any incident or mishap. For this reason, into prominence. Like life, effective safety management
and because (in the terms of the model) they are the pre- is 'one damn thing after another'.
cursors of unsafe acts, they represent the most suitable
cases for treatment. Unsafe acts are like mosquitoes. Acknowledgements
They can be swatted or sprayed, but they still keep
As the many references to their work will testify, I owe a special debt
coming. The only effective remedy is to drain the of gratitude to Dr David Woods, Dr Richard Cook, Dr David Gaba,
swamps in which they breed. In the case of anaesthetic Professor Jan Davies and Professor Bill Runciman. They led the way.
practice, the nature and location of these swamps is both This paper merely followed.
well known and universal. They include:

• Teamwork and communication problems. 1:s,2°

References
• Problems with the design, construction, maintenance 1. Gaba D M. Human error in dynamic medical domains. In: Bogner
M S, ed. Human Errors in Medicine. Hillsdale, New Jersey:
and standardisation of equipment. 21
Erlbanm, 1994.
• Problems with drugs: labelling, purchase, stock con- 2. Wilson M. Errors and accidents in anaesthetics. In: Vincent C,
trol, delivery to and from storage, etc. z2 Ennis M, Audley R, eds. Medical Accidents. Oxford: Oxford
University Press, 1993: 61-79.
• Problems with the assessment and scheduling of 3. Williamson J A, Webb R K, Sellen A, Runciman W B, Van Der
patients. 18,w Walt J H. Human failure: An analysis of 2000 incident reports.
• Problems with the planning and coordination of anaes- Anaesthesia and Intensive Care 1993; 21: 678-683.
4. Weir P M, Wilson M E, Are you getting the message? A look at
thetists and their co-workers) ,19,2°
the communication between the Depamnent of Health,
manufacturers and anaesthetists. Anaesthesia 1991; 46: 845-848.
Measures to combat these problems are being imple- 5. Mayor A H, Eaton J M. Anaesthetic machine checking practices: a
mented in a number of institutions. Less well understood, survey. Anaesthesia 1992; 47: 866-868.
however, are the factors that determine an organisation's 6. Cooper J B, Cullen D J, Nemeskal R et al. Effects of information
feedback and pulse oximetry on the incidence of anaesthesia
general 'safely health'. complications. Anesthesiology 1987; 67: 686-694.
In medicine, there is no single, definitive measure of a 7. Chopra V, Bovill J G, Spierdijk J, Koornneef F. Reported
person's health. It is an emergent property inferred from significant observations during anaesthesia: A prospective analysis
over an 18-month period. Br J Anaesth 1992; 68: 13-17.
a selection of physiological signs and life-style indica- 8. Senders J W, Moray N P. Human Error: Cause, Prediction and
tors. The same is also true for complex hazardous Reduction. Hillsdale, NJ: Erlbaum, 1991.
systems. Assessing an organisation's current state of 9. Reason J. Human Error. New York: Cambridge University Press,
1990.
'safety health', as in medicine, involves the regular and 10. Woods D D. Some results on operator performance in emergency
judicious sampling of a small subset of a potentially events. Institute of Chemical Engineers Symposium Series 1984;
large number of indices. 90: 21-31.
11. Sheen Mr Justice. MV Herald of Free Enterprise. Report of Court
At the University of Manchester, we have tried a No. 8074 Formal Investigation. London: Department of Transport,
variety of diagnostic approaches in our applied research, 1987.
carried out mainly in the fields of oil exploration and 12. Hollnagel E. Human Reliability Analysis: Context and Control.
London: Academic Press, 1993.
production, railway operations and aircraft engineering. 13. Cook R I, Woods D D. Operating at the sharp end: The complexity
The individual labels for the assessed dimensions vary of human error. In: Bogner M S, ed. Human Errors in Medicine.
from industry to industry, but all of them have been Hillsdale, New Jersey: Erlbaum, 1994.
14. Reason J. The Mariner's Guide to Human Error. London: Shell
guided by two principles. First, we try to include those International Tankers, 1993.
organisational 'pathogens' that have featured most 15. Fischhoff B. For those condemned to study the past: Heuristics and
biases in hindsight. In: Kahneman D, Slovic P, Tversky A, eds.
conspicuously in well-documented accidents (hardware
Judgment Under Uncertainty: Heuristics and Biases. New York:
defects, incompatible goals, poor operating procedures, Cambridge University Press; 1982: 335-354.
undermanning, high workload, inadequate training, etc.). 16. Caplan R A, Posner K L, Cheney F W. Effect of outcome on
physician's judgments of appropriateness of care. JAMA 1991;
Second, we seek to encompass a representative sampling
265: 1957-1960.
of those core processes common to all technological 17. Bacon Sir F. The New Organon. Anderson F, ed. Indianapolis:
organisations (designing, building, or specifying equip- Bobbs-Merrill, 1960 (Originally published 1620).
18. Runciman W B, Webb R K, Lee R, Holland R. System failure: An
ment, operating, maintaining, managing, communicating, analysis of 2000 incident reports. Anaesthesia and Intensive Care
goal-setting, etc.). 1993; 21: 684--695.
In all three industries, the measurements are summa- 19. Eagle C J, Davies J M, Reason J T. Accident analysis of large scale
technological disasters applied to an anaesthetic complication.
rised as bar graph profiles. Their purpose is to identify Canadian Journal of Anaesthesia 1992; 39:118-122.
the two or three organisational factors most in need of 20. Helmreich R L, Schaefer H-G. Team performance in die operating
remediation and to track changes over time. Instead room. In: Bogner M S, ed. Human Errors in Medicine. Hillsdale,
New Jersey: Erlbanm, 1994.
of dwelling upon the last accident, and trying to find 21. Woods D D, Johannesen J J, Cook R I, Sarter N B. Behind Human
local 'fixes' for what was probably a unique occurrence, Error: Cognitive Systems, Computers, and Hindsight. CSERIAC
the attention of safety managers is now directed towards State-of-the-Art Report. Wright-Patterson Air Force Base, OH:
Crew Systems Ergonomics Information Analysis Center, 1994.
eliminating the worst of the current latent problems. 22. Currie M, Mackay P, Morgan C et al. The 'wrong drug' problem in
Maintaining adequate safety health is thus comparable anaesthesia: An analysis of 2000 incident reports. Anaesthesia and
to a long-term fitness programme in which the focus of Intensive Care 1993; 21: 596-601.