14-08-2020

IQC

WHAT ARE WE GOING TO LEARN RELEARN

UNLEARN TODAY ?
 What is Quality Control & Quality Assurance ?
 A bit of basic terms & statistics

 Why should we use Internal Quality materials?

Control How to choose QC materials ?
 How to establish lab mean & range?

 How often to run ?

 How to interpret data & identify problems ?

 What are Multi Rule QC & Westgard Rules

 How to take corrective action?

 What are CUSum Charts ? What is EWMA ?

 How patient results can be used as IQC ?

1
 14-08-2020

DEFINITION OF QUALITY

 “ PRODUCT OR SERVICES MEETING / FULFILLING

DESIRED & EXPECTED REQUIREMENTS OF THE
CUSTOMER ”

 CUSTOMER – PATIENT & REFERRING CLINICIAN

“ FIT FOR PURPOSE ”

WHAT IS QUALITY ?

 IT IS NOT VISIBLE
 IT CANNOT BE MEASURED
 IT HAS NO UNIT OF MEASUREMENT
 IT CAN ONLY BE COMPARED AGAINST A
STANDARD
 IT CANNOT BE ACHIEVED IN ONE STEP
 IT CAN ONLY BE ACHIEVED OVER A PERIOD
AND IN STEPS
THEN HOW DO WE ASSES QUALITY IN
MEDICAL LAB?

2
 14-08-2020

QUALITY IN MEDICAL LABORATORY

TESTING

 Right test performed by the

 Right method on the

 Right specimen after

 Right preparation and the

 Right result issued

 Based on Right reference data

 at the Right time

 & at the Right price

Quality Control

Is not

Quality Assurance

3
 14-08-2020

TOTAL QUALITY MANAGEMENT SYSTEM

MODEL

Quality Improvement

Quality Assurance

Quality Control

Quality Plan

WHAT IS QUALITY CONTROL IN MEDICAL LAB ?

Represents all procedures that monitor performance

to ensure that individual patient sample measurements

are within clinically acceptable limits and can be

released reliably (WHO 1981)

4
 14-08-2020

QUALITY CONTROL IS NOT …

just the ‘control material’ but rather it is a system that

helps lab provide the correct result each time a
measurement is made

The QC material is only part of analytical phase - Others

being

Pre Analytical Phase Quality Assurance

Post Analytical Phase Quality Assurance

Trained operators

Documents & Records

and a Procedural Manual

WHAT IS QUALITY ASSURANCE ?

is the sum of all activities done to ensure generation of
reliable & accurate patient results or data (WHO 1981)
Includes
 Pre-analytical, analytical & post-analytical processes
(including internal & external quality control)
 Equipment – maintenance, calibration, validation

 Reagents, kits, chemicals – validated, standard

 Environment monitoring

 Personnel –qualification, knowledge, training, competency

 Maintaining proper documents and records

 Compliance to legal, national guidelines and statutory

requirements

5
 14-08-2020

WHAT IS TOTAL QUALITY MANAGEMENT ?

 Total - made up of the whole

 Quality - degree of excellence a product or service

fulfils the needs & requirements of its users
 Management – a group of people working in a coordinated
manner towards quality

 Act or Manner of planning, directing, controlling &

improving an organisation towards the goal of Quality

 Monitoring & Improving

 Internal Audits & Quality Indicators

WHAT DOES QUALITY CONTROL

ACHIEVE…
Properly planned & executed Quality Control
procedures will
Detect errors
Their Source & Magnitude
& avoid false rejections
Alert laboratory personnel of the possible
deterioration of quality

Quality Control CANNOT PREVENT ERRORS;

CAN ONLY DETECT ERRORS

6
 14-08-2020

QUALITY CONTROL IN TOTAL TESTING

PROCESS (TTP)

1. Pre analytical phase

2. Analytical phase
3. Post Analytical phase

Lab Errors occur more in

1. Pre analytical ( 45-69 %) &
2. Post analytical phase ( 15-20%)

QUALITY CONTROL IN
PRE-ANALYTICAL PHASE
 Preanalytical phase - included all processes from the
time a laboratory request is made by a physician until
the sample is ready for testing.

 Why is it important ?
 Errors occurring at this stage often become apparent
only later in the analytical and post-analytical phases.

 Complex process - multiple people involved, most of the

steps are human, multiple tasks, not so experienced
staff, time pressure

 45 to 69 % of lab errors occur in pre analytical phase

7
 14-08-2020

QUALITY CONTROL IN
PRE-ANALYTICAL PHASE
 1. Test Request generation – Patient identification errors &
misidentification, wrong test request, wrong labelling.
 2. Patient preparation as appropriate must be done -time of
day, food intake & its time, type of food, drug intake
 2. Collection of specimen in right container – type of
additive, preservative, anti coagulant, sterile container,
 3. Appropriate blood-Additive or AC ratio
 4. Proper mixing– No. of times mixing to be done – 5
inversions
 5. Transport at appropriate temperature & time interval
 6. Proper processing of specimen – centrifugation speed &
time
 7. Appropriate storage conditions till analysis

QUALITY CONTROL IN
POST-ANALYTICAL PHASE
 1. Correlation of results with previous results & with
clinical condition of patient if available
 2. Transmission of correct results & patient details on
to final test report
 3. Minimizing transcription errors – decimals &
calculated values
 4. Timely Notification of CRITICAL results & URGENT
to the referring clinician
 5. Maintaining & Monitoring TAT – Turn Around Time
for tests
 6. Minimizing amendment & revision of test reports

8
 14-08-2020

COMPONENETS OF QUALITY
CONTROL IN ANALYTICAL PHASE

 Internal Quality Control

 External Quality Control

INTERNAL QUALITY CONTROL (IQC)

IQC REFERS TO THE SET OF

PROCEDURES UNDERTAKEN BY THE
LABORATORY STAFF FOR THE

CONTINUOUS AND IMMEDIATE

MONITORING OF LABORATORY WORK
IN ORDER TO DECIDE WHETHER THE
RESULTS ARE RELIABLE ENOUGH TO BE
RELEASED. PPT
C,
IT IS A DAILY PROCEDURE Welli
ngto
n,
NZ

9
 14-08-2020

INTERNAL QUALITY CONTROL (IQC)

• The main objective of IQC is to ensure
day-to-day consistency so that results can be
reliably released to make medical decisions
(WHO 1981)
• To detect errors immediately that occur due
to test system failure, adverse environmental
conditions & operator performance
• Their source & magnitude
• Quality control is not just analyzing Quality
control materials but identifying outliers &
taking suitable CA PPT
C,
Welli
ngto
n,
NZ

TERMINOLOGIES IN IQC & EQC

Precision: Closeness of agreement between independent test

results of a specimen
 Imprecision : How far apart are numerical values from each
other for a specimen
 Measure of imprecision – Standard Deviation in units &
Coefficient of Variation in %
 Repeatability: Closeness of agreement between independent
results of successive measurements under same conditions
 (called “Within-run precision”) - It is “short-term precision”

 Reproducibility: closeness of agreement between independent

results of measurements under changed conditions ( time,
operator, environment, reagent lots, calibrations)
 (called “Between-run precision” / “Intermediate precision” )

 Total Imprecision – sum of both = √S12 + S22

10
 14-08-2020

TERMINOLOGIES IN IQC & EQAS

Run : An interval (a period of time or a series of measurements)

within which the accuracy & precision of the measuring system is
expected to be stable - Usually an 8-hour to 24 – hour interval

 Within in run precision (imprecision) : precision calculated from

data collected from a single run

 Between run precision (imprecision) : precision calculated from

data collected from different runs ‘ intermediate precision

 Total Imprecision : Sum of within run & between run

TERMINOLOGIES IN IQC & EQAS

 Accuracy: Closeness of agreement between the result of a

SINGLE measurement and a assigned ‘true’ value of the analyte

 Accuracy therefore is influenced by both bias & imprecision &

 An estimate total ‘error’ or ‘variation’ in measurement procedure

 Uncertainty: Accuracy is a qualitative term & is expressed as

‘UNCERTAINITY’
 inversely related to the “uncertainty” of measurement

 Inaccuracy : The difference between the assigned ‘true’ value

and a single measured value is called inaccuracy a qualitative term

11
 14-08-2020

TERMINOLOGIES IN IQC & EQAS

Trueness of measurements is defined as closeness of agreement

between the average value obtained from a large series of results
of measurements and the ‘assingned’ true value.

 Trueness can be evaluated by comparison of measurements by

the new test and by some preselected reference measurement
procedure, both on the same sample or individuals

 Bias : Measure of trueness - and can be expressed in units of

analyte being measured or as a % of true value

TERMINOLOGIES IN IQC & EQAS

Statistical Terms
 Standard Deviation : Average of spread or dispersion of
values of repeated measurements about or around the
mean
 How far the repeat values are spread around the mean

 Coefficient of variation : SD expressed as a % –

measure of imprecision of a measurement

 Trend: sustained increase or decrease of a quality

control value over a period of four or more days with
the latest value at or above 2SD limits

 Shift A sudden and eventually a stable change in control

values and patient values -

12
 14-08-2020

TERMINOLOGIES IN QC
(FROM TEITZ TEXTBOOK OF CLINICAL CHEMISTRY &
MOLECULAR DIAGNOSTICS 6TH EDITION)

SYSTEMATIC VS. RANDOM ERRORS

Systematic Error Random Error

•Avoidable error due to • Unavoidable errors that
controllable variables in a are always present in any
measurement. measurement.
•Set pattern of error /
• No set pattern can be
mistake.
positive /negative
•Always in one direction - • Occurrence can’t be
either positive or
predicted
negative - never both
•Its occurrence can be
explained & corrected • Impossible to eliminate
but can be minimized
•Accuracy problem
• Precision problem –
• Expressed as Bias
• Expressed as SD or CV

13
 14-08-2020

Total Allowable Error (TEa)

Total Allowable Error (TEa) =

Systematic Error (inaccuracy / bias) + Random Error (Imprecision / SD/CV)

TEa = Lab Inaccuracy (in units) + 2 x SD (in units)

TE a lab % = Lab bias (%) + 2 x CV Laboratory

CV / SD can be maximum 1/3 for (between run) or 1/4 (within run) of

TEa

Guidelines for acceptable performance for Analytical Quality

Requirements
1. CLIA proficiency testing criteria for acceptable analytical performance

2. ‘Rili’BÄK’ - German Federal Medical Council Guidelines

3. Royal College of Pathologists of Australasia Analytical Quality

Requirements (RCPA)
4. ACBI CMC criteria for Indian labs – Desirable Coefficient of Variation

NORMAL (GAUSSIAN) DISTRIBUTION

–WHAT IS IT ?
 Assumed for all quality control
statistics

 All values are symmetrically

distributed around the mean

 Characteristic “bell-shaped”
curve
 68.2 % values – 1SD
 95.4% values – 2SD
 99.8% values – 3SD

14
 14-08-2020

STANDARD DEVIATION

 SD is used by manufacturer’s of control materials & labs

to set limits for the values of IQC materials

 3 SD limits are usually given by the manufacturer of IQC

materials – sometimes no SD iss given but as a %

 Reason why action to be taken if it falls outside these

3 SD limits – outside 99.8 % of values

15
14-08-2020

16
 14-08-2020

STANDARD DEVIATION
 When a QC measurement falls within 2 SD range, there is 95.5%
confidence that the measurement is correct

In other words

 1. if distribution is normal, 95.5 % of results would be within

mean +/- 2SD

 2. The probability that 95.5 of results have occurred by choice

and not by chance

 3. Only 4.5 % of results are outside this mean +/- 2 SD values

(errors)

HOW TO CALCULATE RANGE FOR QC

MATERIAL

Mean 200.6
SD 8.21
Range limits in our example
+ 1 SD- 200.6 + 8.21= 208.81
-1 SD- 200.6 - 8.21= 192.39
1 SD limits are 192.39 to 200.81
+ 2 SD-200.6 + (2x 8.21) = 217.02
+ 2 SD-200.6 + (2x 8.21) = 217.02 - 2 SD-200.6 - (2x 8.21) = 184.15
- 2 SD-200.6 – (2x 8.21) = 184.18
2 SD limits are 184.18 to 217.02

+ 3 SD-200.6 + (3x 8.21) = 225.24

- 3 SD-200.6 - (3x 8.21) = 175.93
3SD limits are 175.93 to 225.24

17
 14-08-2020

COEFFICIENT OF VARIATION

 The Coefficient of Variation (CV) is the standard

Deviation (SD) expressed as a percentage of the mean
( aka as Relative Standard deviation-RSD)
 CV % = SD x 100
Mean

CV = 8.21 x 100 = 4.09 %

200.6
 An imprecision of + /- 4.09 %
 MU = CV x 1.96 (approximately 2 times SD / CV)

 If you are reporting a value of 200 there is an error /

variation of +/- 4.09 % & MU is 16.42
 It can be either 216.42 183. 58

PRECISION : CV% VS. SD – WHICH IS BETTER ?

CV IS BETTER

 The CV is more accurate comparator than SD because SD

typically increases as the conc. of analyte increases

 The CV is useful for comparisons of precision at different

concentrations
 Can be easily misled if you are comparing precision for two
different methods (e.g. by instrument, method, reagent, etc.)
by using standard deviation
 E.g SD for cholesterol is 8.1mg/dL at a mean cocn. of 212
mg/dL and for TGL is 4.8 mg/dL at a mean cocn of 104
mg/dL. Based on just SD, one might conclude that the TGL is
is more precise than the cholesterol
 However, CV % for CHO is 4.01 & for TGL is 4.61 which shows
that cholesterol estimation is more precise than TGL
estimation

18
 14-08-2020

PRECISION : CV% VS. SD -

CV IS BETTER
 Cholesterol
LEVEL l LEVEL 2

MEAN 87.7 223.4

SD 3.9 9.8

Which is better?
Based on just SD, one might conclude that the
Level 1 is more precise than the level 2

PRECISION : CV% VS. SD -

CV IS BETTER

 Cholesterol
LEVEL l LEVEL 2

MEAN 87.7 223.4

SD 3.9 9.8
CV 4.45 4.39
However, CV % for level 2 is 4.39 & for level 1 is
4.45 which shows that Level 2 is more precise than
level 1
Precision is better at higher concentration than at
lower concentration

19
 14-08-2020

SOURCES OF
INTERNAL QUALITY CONTROL MATERIALS

1. DIAGNOSTIC PATIENT SAMPLES – ‘ in house control’ material

from inside the lab - values not established
o Ideal source of ‘in-house’ control material - Master Health
Checkup samples, donor blood samples, pre-anesthetic & pre-
operative patient samples
o Body fluids - CSF, Pl..Fluid, Perit. Fluid, Low Electrolytes,
High Creatinine & Bilirubin

2. COMMERCIAL QC SAMPLES - Assayed & values established by

the manufacturer – lyophilized, liquid

3. EQAS SAMPLES - Assayed & Values established by consensus by

EQAS provider – small pool still useful for calibration verification

20
 14-08-2020

SOURCE OF IQC MATERIALS -

MANUFACTURER’S VS. THIRD PARTY

• Manufacturer controls: of equipment or reagent or

calibrators
• Third party controls - Not specifically designed for a
specific instrument or reagent or calibrator
• Manufactured independently of the equipment,
calibrators or reagents
• Preferable
Why ?
• Independent assessment of the entire test system
• Better ability to detect errors and shifts with new
reagent lots & calibrators esp. if same lot is available
over a long period
• 1 year supply ideal . optimal 6 months.

SOURCE OF COMMERCIAL CONTROLS

THIRD 4 MONTHS
FIRST 4 MONTHS SECOND 4 MONTHS

Inst.
Manufacturer LOT 2
LOT 1 LOT 3
Reagent
Inst.
Manufacturer LOT 1 LOT 2 LOT 3
Control
Inst. LOT 2
Calibrator LOT 1 LOT 3

Third Party
Control (SAME) LOT 1

21
 14-08-2020

HOW TO CHOOSE A INTERNAL QUALITY

CONTROL MATERIAL
 Values should cover entire medical decision points – ‘low, normal, high’

 At or near the Analytical Measuring Range of the measurement procedure

because defects may affect these concentrations before others

 Also Equipment need to be validated for the entire range and likely patient
values

 If unable to assay all three levels daily or with each shift, use a
combination

 Should resemble the patient sample or test specimen (matrix & interfering
substances apt. conc.) – ‘commutability’

 Available in large quantity - Ideally, should last for at least 1 year –

 After vial has been opened and material prepared, it should remain stable
for the period of use –
 Material should be low vial-to-vial variability.

HOW TO CHOOSE INTERNAL QUALITY

CONTROL MATERIALS ?

o If the results near the interval limits are acceptable it can be

assumed that their performance over the entire AMR will be
acceptable – for linear response assays

o For nonlinear analytical response assays it may be necessary to

use additional controls at intermediate concentrations

o Imprecision near LoQ may be high (20%) – Choose

concentrations where it is practically possible

o For procedures with extraction (DBS) or other pretreatment

steps (HbA1c), controls used must include these pretreatment
steps too.

22
 14-08-2020

FREQUENCY TO MEASURE QUALITY

CONTROL SAMPLES
 To prevent risk of harm to a patient from clinical decision not
being taken or wrong medical decisions made before a
significant error is detected at the next scheduled QC event
 Based on
 • 1. Analytical stability of the measurement procedure

 • 2. Events such as recalibration or major maintenance /

breakdown that may alter the current performance condition
of the measurement procedure

 3. When New Reagent Lot is put into use / installed

 4. As per manufacturer’s schedule

FREQUENCY TO MEASURE QUALITY

CONTROL SAMPLES
 Daily as a routine:

 According to NABL 112 “ Specific criteria for

accreditation medical laboratories document”

 Minimum 2 levels daily at peak hours

 And atleast 1 level every 8 hours for labs operating 24

hours

 Irrespective of the number of test requests received for

each test

 There should be IQC for every reporting analyte.

23
 14-08-2020

RECONSTITUTION OF IQC MATERIAL

 “ Exactly as per manufacturer’s instructions”
 Same person(s) to reconstitute every time
 Bring to RT before reconstitution
 Use Reagent Grade 1 (CLRW) water for reconstitution
 Use graduated glass volumetric pipette ‘Type A’
 Not automated pipette esp if more than 1 mL of
water is required ;
 ‘Less diluent, More analyte;More Diluent Less Analyte’
 Allow to stand for 20-30 minutes - Do not, shake
immediately, vigorously, foam
 Gently swirl between palms of hand every 10 minutes
 Keep inverted for 5 minutes before aliquoting

ESTABLISHING
RANGE FOR IQC MATERIALS
 Use product insert ranges only as a guideline

 Manufacturer Ranges are based on reagent lots

available at the time of assignment of values.

 During the life of the control lot, manufacturers

may reformulate tests or begin using a new source
of raw materials for kit/reagent production.

 Manufacturer Ranges cannot account for variables

such as instrument , software updates or
performance differences over time, operator
variability, variable environmental conditions, water
quality used for reconstitution etc.

24
 14-08-2020

ESTABLSIHING
RANGE FOR IQC MATERIALS

• Lab mean & SD for IQC material should be established

• 1. When systematic bias is small as possible

• 2. Measurement procedure is calibrated correctly

• 3. Adequate number replicate measurements over a

sufficient time interval to include all the typical
sources of measurement variability so that a true
representative mean is obtained

ESTABLSIHING LAB
RANGE FOR IQC MATERIALS
 Provisional Mean – Collect minimum 20 data points from separate
analytical runs on different days (10 to 20 days)

 Purpose - covers all possible sources of variability & day to day

variability in the measurement procedure & are reasonably
represented in the mean value
 To account for calibration, change of reagent or reagent lot,
operator technique, temperature / humidity of testing location,
daily/weekly maintenance, etc.

 If 20 data points from 20 separate analytical runs are not

available, provisional values may have to be established from data
collected over fewer than 20 runs
 Using the manufacture's guidelines, remove the obvious outliers
and blunders.
 After the trimming there should be at least 20 data entries for
each analyte.

25
 14-08-2020

ESTABLSIHING LAB MEAN

RANGE FOR IQC MATERIALS

 Using the collected data, calculate the mean & SD

 More the data, better will be the target value.

 Use your discretion to decide on the decimals.

Do not use decimals for analytes that have values in three digits
e.g., Na, Cholesterol etc. Similarly do not roundup the values for
Creatinine, Potassium, Magnesium etc., use 2 digits

 Make sure the calculated target value is not far away from
the manufacturer's target.
Lab mean has some uncertainty & may need to be redefined if
measurement conditions fluctuate (e.g new Reagent lot / Calibrator
lots, equipment maintenance, new calibration etc)

SETTING A SD FOR QUALITY CONTROL

MATERIAL

• SD of IQC material expresses measurement procedure

RE over time
• Since SD decides range of QC range values &
acceptability of an individual QC value and application of
QC rules , it must be as realistic to represent all these
variability over time
 Setting SD for New Lot of QC material - it is
appropriate to use mean value for the new lot of
manufacturer for a time along with the well-established
lab SD from the previous lot. (Or higher of previous 2
lots)
 Imprecision is a property of the measurement procedure
& equipment used & is unlikely to change with a different
lot of QC material

26
 14-08-2020

SETTING A SD FOR QUALITY CONTROL

MATERIAL
 SD for new QC material from a different manufacturer - it is
possible for the observed SD to be different than the historic value
because of matrix differences. - Manufacturer SD can be used
 SD for a new measurement procedure – no historical SD performance
is available
 1. Can use SD obtained during method verification
 2. A minimum of 20 observations on 20 different days recommended
for initial estimate of SD
 3. But reagent Lot changes should not be included in initial estimate
of SD – QC results are artifactually influenced by RGT lot changes
 4. SD of the existing procedure can be used provided SD of existing
method was appropriate for taking suitable clinical decisions – esp.
when initial estimate SD of new method is smaller than existing
method

L-J CHART WITH MANUFACTURER & LAB

CALCULATED MEAN & RANGE
Glucose (Jan 2011) with Manufacturer Mean & Ranges
339
(SD – 21 ; CV- 7.6%)
318
Mean - 276
297 Range 213-339
SD - 21
276 CV 7.6 %
255

234

213
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

Mean – 274.77
Range 266.58-280.23
Glucose (Jan 2011) with Lab defined Mean & Ranges
(SD – 2.73; CV- 0.99%) SD – 2.73
CV 0.99 %
280.23

277.5

274.77

272.04

269.31

266.58
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

27
 14-08-2020

L-J CHART WITH LAB CALCULATED MEAN & RANGE

HDL CHOLESTEROL Level: 1 Date/Month/Year:01 April 2015-30 April 2015

Established Lab Mean Day to Day Manufacturer Mean

Equip. Name: BIOSYSTEMS A25 Lot.# 14481

QC. Manufacturer.: Exp. Date Mean

BIORAD 31/05/2017 64.95 Mean 66.01 Mean 56.6
SD 4.09 SD 2.48 Low Range 43.4

Control CV(%) High Range

Days
Values
6.29 CV(%) 3.76 69.9
(+)3SD (+)2SD (+)1SD MEA
N (-)1SD (-)2SD (-)3SD Mu 7.37 SD 4.42
1-Apr 68.8 77.21 73.13 69.04 60.86 56.77 52.69
77.21
2-Apr 67.9 73.13 69.04 # 60.86 56.77 52.69
3-Apr 68.3 77.21 73.13 69.04 # 60.86 56.77 52.69
4-Apr 65.6 77.21 73.13 69.04 # 60.86 56.77 52.69
6-Apr 66.4 77.21 73.13 69.04 # 60.86 56.77 52.69
7-Apr 72.1 77.21 73.13 69.04 # 60.86 56.77 52.69
8-Apr 67.2 77.21 73.13 69.04 60.86 56.77 52.69
#
Levey Jennings Chart
9-Apr 66 77.21 73.13 69.04 # 60.86 56.77 52.69
10-Apr 63.8 77.21 73.13 69.04 # 60.86 56.77 52.69
11-Apr 67.4 77.21 73.13 69.04 # 60.86 56.77 52.69
13-Apr 63.9 77.21 73.13 69.04 # 60.86 56.77 52.69
15-Apr 67.9 77.21 73.13
73.14
69.04 60.86 56.77 52.69
#

16-Apr 62.8 77.21 73.13 69.04 # 60.86 56.77 52.69

17-Apr 69 77.21 73.13 69.04 # 60.86 56.77 52.69
18-Apr 63.5 77.21 73.13 69.05
69.04 # 60.86 56.77 52.69
20-Apr 64.4 77.21 73.13 69.04 # 60.86 56.77 52.69
21-Apr 64.7 77.21 73.13 69.04 # 60.86 56.77 52.69
22-Apr 66.1 77.21 73.13 64.96
69.04 # 60.86 56.77 52.69
23-Apr 63.1 77.21 73.13 69.04 # 60.86 56.77 52.69
24-Apr 65.4 77.21 73.13 69.04 # 60.86 56.77 52.69
25-Apr 69 77.21 73.13 69.04 60.86 56.77 52.69
60.87
#

27-Apr 62.8 77.21 73.13 69.04 # 60.86 56.77 52.69

28-Apr 63.3 77.21 73.13 69.04 # 60.86 56.77 52.69
29-Apr 63 77.21 73.13 69.04 60.86 56.77 52.69
56.78
#

30-Apr 67.8 77.21 73.13 69.04 # 60.86 56.77 52.69

77.21 73.13 69.04 # 60.86 56.77 52.69
77.21 73.13 69.04 60.86 56.77 52.69
52.69
#

Date

L-J CHART WITH MANUFACTURER MEAN & RANGE

HDL CHOLESTEROL Level: 1 Date/Month/Year:01 April 2015-30 April 2015

Established Lab Mean Day to Day Manufacturer Mean

Equip. Name: BIOSYSTEMS A25 Lot.# 14481

QC. Manufacturer.: Exp. Date Mean

BIORAD 31/05/2017 64.95 Mean 66.01 Mean 56.6
SD 4.09 SD 2.48 Low Range 43.4

Control CV(%) High Range

Days
Values
6.29 CV(%) 3.76 69.9
(+)3SD (+)2SD (+)1SD MEA
N (-)1SD (-)2SD (-)3SD Mu 7.37 SD 4.42
1-Apr 68.8 77.21 73.13 69.04 60.86 56.77 52.69
77.21
2-Apr 67.9 73.13 69.04 # 60.86 56.77 52.69
3-Apr 68.3 77.21 73.13 69.04 # 60.86 56.77 52.69
4-Apr 65.6 77.21 73.13 69.04 # 60.86 56.77 52.69
6-Apr 66.4 77.21 73.13 69.04 # 60.86 56.77 Levey Jennings Chart
52.69
7-Apr 72.1 69.86
77.21 73.13 69.04 # 60.86 56.77 52.69
8-Apr 67.2 77.21 73.13 69.04 # 60.86 56.77 52.69
9-Apr 66 77.21 73.13 69.04 # 60.86 56.77 52.69
10-Apr 63.8 65.44
77.21 73.13 69.04 # 60.86 56.77 52.69
11-Apr 67.4 77.21 73.13 69.04 # 60.86 56.77 52.69
13-Apr 63.9 77.21 73.13 69.04 # 60.86 56.77 52.69
15-Apr 67.9 61.02
77.21 73.13 69.04 # 60.86 56.77 52.69
16-Apr 62.8 77.21 73.13 69.04 # 60.86 56.77 52.69
17-Apr 69 77.21 73.13 69.04 # 60.86 56.77 52.69
18-Apr 63.5 56.6
77.21 73.13 69.04 # 60.86 56.77 52.69
20-Apr 64.4 77.21 73.13 69.04 # 60.86 56.77 52.69
21-Apr 64.7 77.21 73.13 69.04 # 60.86 56.77 52.69
22-Apr 66.1 77.21
52.18 73.13 69.04 # 60.86 56.77 52.69
23-Apr 63.1 77.21 73.13 69.04 # 60.86 56.77 52.69
24-Apr 65.4 77.21 73.13 69.04 # 60.86 56.77 52.69
25-Apr 69 77.21 73.13 69.04 60.86 56.77 52.69
47.76
#

27-Apr 62.8 77.21 73.13 69.04 # 60.86 56.77 52.69

28-Apr 63.3 77.21 73.13 69.04 # 60.86 56.77 52.69
29-Apr 63 77.21 73.13 69.04 60.86 56.77 52.69
43.34
#

30-Apr 67.8 77.21 73.13 69.04 # 60.86 56.77 52.69

77.21 73.13 69.04 # 60.86 56.77 52.69
77.21 73.13 69.04 # 60.86 56.77 52.69

Date

28
 14-08-2020

L-J CHART WITH MANUFACTURER & LAB

CALCULATED MEAN & RANGE
Levey Jennings Chart
69.86

65.44

61.02

56.6

52.18

47.76

43.34

Date

Levey Jennings Chart

73.14

69.05

64.96

60.87

56.78

52.69

Date

PARALLEL TESTING TO ESTABLISH LAB

RANGE FOR IQC MATERIALS

 Parallel Testing : New lot of IQC material is

analyzed for each analyte of interest in parallel with
the lot of control material in current use - Better
practice

 Minimum 20 data points / days / runs

 Establish a provisional mean using data points –

provided current lot QC values are within 2SD range
/ not violating Multi Rule QC

29
 14-08-2020

GUIDELINES ON PLANNING QC PROCEDURE

 QC results help to validate whether the instrument

is operating within pre-defined specifications,
thereby inferring that patient test results can be
released reliably

 Set the Quality Requirement for the test –Total

Allowable Error

 Based on lab’s Precision (CV/SD) & Inaccuracy (Bias)

ASSESSING IQC PERFORMANCE

 Calculate your IQC values’ SD and CV at the end of
every month

 How much SD or CV is allowable ? 25%to 33% of

TEa for the analyte
 TEa = RE + SE = (50 % RE + 50% SE)

 RE is 2 times CV Hence SD / CV can be 25% of Tea

 Based on Interindividual & Intraindividual Biological

Variation –
 Optimal 50% of BV
 Desirable 25 % BV
 Minimal 75% BV

30
 14-08-2020

PLOTTING & MONITORING

QC RESULTS – “LEVEY-JENNINGS CHART”

 Use of Control charts – popular “Levey-Jennings

chart” (L-J chart)

 From ‘Control Charts’ Originally developed by Walter

A Shewart in 1920s while working for bell labs, a
telephone transmission company-

 Later in 1950s adopted to monitor performance in

medical labs by Stephen Levey & Edward.R Jennings

PLOTTING & MONITORING QC RESULTS

 A graphical method for displaying IQC results and to
decide whether a run is in-control or out-of-control

 Control values of each run are plotted on y-axis are

plotted versus time (day / run) on x-axis
 Connecting lines are drawn from point to point to identify
any trends, shifts, or random variations and make
decisions whether to accept or reject the run
 Useful to monitor & evaluate the imprecision (RE) / &
inaccuracy (SE) of repeated IQC measurements

 The scale should be selected in such a way that the

plotting of data is not crowded.

 The chart should be displayed in the laboratory.

31
 14-08-2020

LEVEY-JENNINGS CHART –
PLOT TIME ON X-AXIS & THE CONTROL VALUES ON Y-AXIS
C o n tro l Valu es (e.g . m g /d L )
115
110
105
100
95
90
85
80
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Time (e.g. day, date, run number)

LEVEY-JENNINGS CHART
STEP 1. CALCULATE THE MEAN AND STANDARD DEVIATION
STEP 2. PLOT THE MEAN AND +/- 1,2 AND 3 SD CONTROL LIMITS ON Y-AXIS

115 113
+ 3 SD
110 108
+ 2 SD
105 + 1 SD
103
100 MEAN
95 93 - 1 SD
90 88
- 2 SD
85 - 3 SD
83
80
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Time (e.g. day, date, run number)

MEAN -98 MG/DL SD- 5 MG/DL
1 SD; 93-103 MG/DL 2 SD; 88-108 MG/DL 3 SD; 83-113 MG/DL

32
 14-08-2020

LEVEY-JENNINGS CHART –
STEP 3 : PLOT CONTROL VALUES FOR EACH RUN
STEP 4 : LOOK FOR QC RULE VIOLATION
C o n tro l Valu es (e.g . m g /d L )
115
110
105
100
95
90
85
80
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Time (e.g. day, date, run number)

MONITORING & DETECTING QC PROBLEMS –

MINIMUM CRITERIA

• Set QC rules - ‘Multi QC rules’ or ‘Westgard Rules’

• Though there are many rules, six are very popular.

• 1 2s, 2 2s, 1 3s, R 4s, 4 1s, 10x

• The ‘1 2s’ is a ‘warning rule’ and the

• 2 2s, 1 3s, R 4s’ are ‘rejection rules’

• 4 1s, 10x – indicate a bias – warning rule

• The Westgard rules are more useful when two level controls are
used in single run.

33
 14-08-2020

PRE REQUISITES FOR ESTABLISHING AND

INTERPRETING QC RULES

 Measurement system must be correctly calibrated &

 Operating within acceptable performance specifications

 before the statistical parameters to establish QC

interpretive rules can be established

 The Mean & SD used for QC interpretive rules should

adequately represent all sources of variability that is
likely to occur over time

Monitoring & Analysing Internal QC

Multi Rule QC (WESTGARD RULES)

• Given by Dr. James O.Westgard, of the University of

Wisconsin in an article in 1981 on laboratory quality control
that set the basis for evaluating analytical run quality for
medical laboratories.

• Six basic rules in the Westgard scheme: 1-3s, 2-2s, R-4s, 1-

2s, 4-1s, and 10x.

• These rules are used individually or in combination (multi-

rule) to evaluate the quality of analytical runs.

• Detect Random error (imprecision or CV) & Systematic

errors (bias or inaccuracy, shifts & trends)

34
 14-08-2020

Monitoring & Analysing Internal QC

Multi Rule QC (WESTGARD RULES)

• Given by Dr. James O.Westgard, of the University of

Wisconsin in an article in 1981 on laboratory quality control
that set the basis for evaluating analytical run quality for
medical laboratories.

• Six basic rules in the Westgard scheme: 1-3s, 2-2s, R-4s, 1-

2s, 4-1s, and 10x.

• These rules are used individually or in combination (multi-

rule) to evaluate the quality of analytical runs.

• Detect Random error (imprecision or CV) & Systematic

errors (bias or inaccuracy, shifts & trends)

WHY USE MULTI RULE QC ?

 ‘False Rejections’ is avoided & ‘Error Detection’ is
maximized

 They provide better performance than the commonly

used 12s and 13s single-rule QC procedures.

 The false-alarm rate with a 12s rule, when N is 2,

probability is 9% of good runs will be falsely rejected;
 with N=3, it is even higher, about 14%;
 with N=4, it's almost 18%

 Advantage of multirule QC procedures - false

rejections can be kept low while at the same time
error detection can be kept high.

35
 14-08-2020

Monitoring & Analysing Internal QC

Multi control QC rules (WESTGARD RULES)
When more than 1 level of IQC are used
12SD or 1- 2s: ‘Warning’ Rule
A single or any one of 2 or 3 control results falls outside ±2SD
Either side - +/-
Applicable across runs & across QC materials
Do not reject run

“Warning rule” - Alerts technologist to possible problems ‘

Start looking for other errors – look at other control & previous
day

13SD or 1-3s: ‘Rejection’ Rule

• A single or any one of 2 or 3 control results falls outside 3SD

• Either side - +/-

• Applicable across runs & across QC materials

• The 1-3s rule identifies unacceptable random error or possibly the

beginning of a large systematic error.
• .

• If this is not violated check 2-2S

36
 14-08-2020

• 22SD or 2-2s: ‘Rejection’ Rule

• This rule detects systematic error -
• Applicable across runs & across materials.
• It is violated ‘within run’ when values of both levles of
control materials or 2 out of 3 control materials exceed
the 2SD limits in "same“ direction Both are mean + 2s or
mean - 2s
• It is violated ‘across runs’ when the previous value for the
same control material level exceeds the mean + 2s or
mean – 2s limit in the same direction.

Within run violation Across run violation

R4s: Rejection’

• This rule indicates only Random

error

• Applicable only across materials &

within the run and NOT for the
same material and between runs .

• It is violated within the run when

value of one of control materials
exceed the mean + 2s or mean – 2s
and the other control materials
exceed the mean + 2s or mean – 2s
but in the opposite direction

37
 14-08-2020

10x Rule : ‘Need not be Rejection’ Rule

• Requires control data from previous runs
• This rule detects systematic error or bias
• Accuracy problem
• Indicates a bias trend
• Applicable to both ‘across run’ & ‘across materials’.
• It is violated across control materials if the last 5 consecutive values
are on the same side of the mean within 1SD (+ / - direction)
regardless of control level
• The rule is violated within the control materials of the last 10 values
for the same control level are on the same side of the mean within
1SD (+ / - direction)

41SD or 4-1s: indicates a Bias Trend

– Sytematic Error - Need nor be a rejection
• Requires control data from previous runs
• Four consecutive IQC results for one level of control exceed mean
+/- 1SD are outside ±1SD (or)
• Both levels of control have 2 consecutive results that exceed mean
± 1SD are outside ±1 SD in the same direction (both + 1 SD / - 1 SD)

38
 14-08-2020

Look for a Trend

Identifies – Bias trend need not be a

Systematic error Warning Rule

FOR CONTROLS RUN IN MULTIPLES OF 2 (typically

chemistry controls)

QC rules are 1 3s ,2 2s R 4s 4 1s, 10x

FOR CONTROLS RUN IN MULTIPLES OF 3 (typically

haematology, coagulation, immunoassay, hormones, blood
gas controls)
1 3s ,2 of 32s R 4s 3 1s, 12x

39
 14-08-2020

Westgard Multirule QC Procedure

3 are Warning Rules 3 are Rejection Rules

40
 14-08-2020

 whenever you reject a run and correct a problem,

 you have to start over again & collect the necessary
number of control measurements to assess control
status of the corrected process.

 You can't use earlier measurements that were

collected prior to correcting the problem

RESOLVING QC PROBLEMS
• Take action in a sequential manner to identify a problem – not
in a random manner

• Reanalyse QC material – using a fresh aliquot

• Check if problem is resolved – If QC problem still persists

take next step

• Record the problem, action taken

• Technologists encouraged to perform corrective action steps

by themselves

41
 14-08-2020

STEPS IN RESOLVING QC PROBLEMS

• 1.Check whether QC is out of control for many analytes using
same wavelength

• 2. Check whether the QC is out- of- control for both levels

or any one level for same analyte

• 3. Check whether reagent has reached its onboard stability

period / nearing it

• 4. Check whether QC material has reached its open vial

stability period / nearing it

• 5. Check whether latest calibration was OK

STEPS IN RESOLVING QC PROBLEMS

• 1. Repeat Assay on control specimen using fresh aliquot of QC
pool

• 2. Repeat assays on control specimen using a newly

reconstituted set of control

• 3. Look for obvious problems – clots, blocks in probes,

carryover reagent levels, mechanical fault,

• 4. Recalibrate instrument for the analyte in question reassay

all controls

• 5. Install a new lot of reagent bottle ( one or all),recalibrate

& reassay all controls

• 6. Perform machine maintenance , recalibrate & reassay all

controls

42
 14-08-2020

CHECKLIST FOR TROUBLE SHOOTING

What is unique to the case ? What is common to
 New reagent the situation ?
 New control  More than one
 New calibrator parameter affected
 Recent part replacement in  Related parameters
the equipment also affected
 Maintenance recently  Multiple levels of
performed control material affected
 Problem occurring on
another instrument also


CHECKLIST FOR TROUBLE SHOOTING

Check Controls Check Reagent
 Controls mixed properly Correct reagent(s)
Reagent(s) loaded and
connected correctly
 Control properly presented
New reagent(s) Lot number
Reagent(s) properly
 Ccontrol failures become
prepared
apparent after recent
Reagent(s) expired
calibration
Reagent(s) exceed the
open/closed container
 Control failures after stability
service Reagents near expiration
date / near end of open vial
stability


43
 14-08-2020

CUSUM CHARTS & EWMA-

• CUmulative SUM Charts (CUSUM) &
• Exponentially Weighted Moving Averages (EWMA)

• Preferred to monitor for bias trends –

• More exact quantitative method to detect SE from RE

and bias trends than Multi Rule QC
 Provides a warning - that may not require immediate CA
but rather an alert to a potential problem
• Disadvantage – less sensitive for RE

• Better to use it along with L-J chart

 CUSUM – is the difference between a QC result and its

target mean in terms of SD (as an SDI, or z-score)
 The fraction of SD represented by the difference

 “Decision Limit technique’

CUSUM CHARTS -
 Set a mean
 Set ‘Threshold’ to ‘initiate’ Cusum calculations

 Only when a IQC result exceeds the threshold initiate

cusum
 Seta decision / control limit to end Cusum – ‘to reject
run’
 Easier to judge & does not require experience

 Consistent interpretation

 More objective

 Manual – Threshold is 1.0 s Decision limit 2.7s

 Computer aided - Threshold is 0.5 s Decision limit 5.1s

44
 14-08-2020

CUSUM CHARTS -

CUSUM CHARTS -

45
 14-08-2020

EXPONENTIALLY WEIGHTED MOVING AVERAGES (EWMA)

EWMA

o EWMA operates by taking the average of the most

recent QC result and previous result

o More “weight” is given to the more recent values in

determining the “average

o Increases the weightage exponentially to recent results

of QC

o Recent results contribute a greater proportion, and

older results contribute very little to the current EWMA
value

YOUDEN PLOT (TWIN PLOT)

 Plotting the values of

both normal & abnormal
controls in the same
graph

 Helps to differentiate
between SE & RE

 Yellow Square - 1SD

 Green Square - 2SD

 Blue Square - 3SD

46
 14-08-2020

YOUDEN PLOT (TWIN PLOT)

 Control Normal (1) is

plotted on ‘x’ axis and 3
Control Abnormal (2) 2
plotted on ‘y’ axis
1

 Yellow Square - 1SD -1

 Green Square - 2SD
-2
 Blue Square - 3SD
-3
-3 -2 -1 1 2 3

YOUDEN PLOT (TWIN PLOT) –

CONTROL NORMAL (1) IS PLOTTED ON X AXIS AND
CONTROL ABNORMAL (2) PLOTTED ON X AXIS

 Point H QC1 > - 2SD

QC 2 > + 2SD
 Random Error

3
 Point A –Both Levels 2
Normal & -Abnormal are
1
> +3SD
 Systematic Error

-1
 Point C –Both Levels
-2
Normal & Abnormal are
> - 2SD
 Systematic Error

47
 14-08-2020

ALTERNATE INTERNAL QUALITY CONTROL

 If control material is not available

 1. Retesting of any randomly chosen retained samples

normal & abnormal

 2. Retained EQA /PT sample/ reference material like

calibrator

 3. Replicate test of same sample by different

method or different equipment / same method but
another equipment

ALTERNATE INTERNAL QUALITY CONTROL

 Calculation of the anion gap.

 simple formula can be applied to electrolyte

results of patient samples at specific intervals
(time, runs, or tests) to monitor possible analytic
error. - Spot check performance of electrolyte
analyzers

 If several (eight samples) consecutive samples

have an increased or decreased anion gap during
any one day or run, then all of the patient results
in those data should be reviewed for possible
analytical error.

48
 14-08-2020

Westgard Procedure Flowchart

Some of the causes of Random Errors:

Sample Calibrations
-carryover - too frequent
-interference -variable age

Reagent
QC material -variable reagent mixing
-varying tests
- variable preparation -splashing
- variable age

Instrument
-variable cuvettes
-intermittent failure
- Sudden Voltage problems
- Air bubbles & Short sampling

49
 14-08-2020

Sources of Systematic Errors:

 Improper alignment of sample or reagent pipettes, dirty cuvettes
 Drift or Shift in incubator chamber temperature
 Inappropriate temperature / humidity levels in the testing area
 Change of reagent or calibrator lot
 Deterioration of reagent, calibrator while in use, storage or
shipment
 Incorrect handling of control product (e.g. Freezing when not
recommended, wrong reconstitution, ageing QC material)
 Inappropriate storage of control products in frost free freezers
 Failing light source
 Use of non-reagent grade water in the test system
 Recent calibration
 Change in test operator
 Specimen carry-over,- inadequate cleaning / washing of probes
 Obstruction of tubing

USE OF PATIENT SAMPLES IN QUALITY

CONTROL PROCEDURES - 1. COMMUTABILITY OF
& EQUIVALENT RESULTS
 Patient samples can be used to verify the consistency of
results for the same measurand between
 1. Multiple instruments using same method
 2. Multiple instruments using different methods
 3. Same instrument to verify calibration & calibration stability
 4. Same instrument using same method
 Why ? To Ensure that calibration of the different methods /
different instruments produces consistent results.
 Sometimes may be necessary to modify the calibration
settings so that equivalent results for patient samples can be
given –
 Also helps to fix common BRI to be used, among different
laboratory testing locations, & avoids clinical confusion while
interpreting laboratory results.

50
 14-08-2020

EMPIRICAL CRITERIA FOR PATIENT TEST RESULT AGREEMENT BETWEEN

REPEATED ASSAYS & FOR AGREEMENT AMONG RESULTS FOR A SINGLE
PATIENT SAMPLE MEASURED ON MULTIPLE INSTRUMENTS
Analyte Acceptable Acceptable
difference in difference in units
%
AST,ALT,ALP, LDH, CK, yGT, 10 % 10 U/L & 10 ug/dL
Lipase & Iron
Mg2+,K+ 0.3 mmol/L
Amylase 10 % 15 U/L
Bilirubin, 10 % 0.3 mg/dL
Protein, PO4 Albumin, Uric acid 0.4 mg/dL, 0.4 g/dL
Calcium 0.5 mg/dL
Creatinine 0.2 mg/dL
Na Cl & CO2 4 mmol/L
Cholesterol 5%
TGL 10%
Glucose 10% 6 mg/dL
BUN 10% 3 mg/dL

USE OF PATIENT SAMPLES IN QUALITY

CONTROL PROCEDURES –
2. FOR CALIBRATION STABILITY & CALIBRATION
MONITORING

 The median or mean value for a sufficiently large

number of patient sample results may be sufficiently
stable to be used as an indicator of method’s
calibration stability over time.

 By calculating the mean & SD for the distribution of

results, periodically for a specified time period (eg, 1
month), & comparing one time period Vs another to
determine whether any changes have occurred.

51
 14-08-2020

USE OF PATIENT SAMPLES IN QUALITY

CONTROL PROCEDURES –
2. CALIBRATION STABILITY & MONITORING

USE OF PATIENT SAMPLES IN QUALITY

CONTROL PROCEDURES –
2. CALIBRATION STABILITY & MONITORING
 Can identify changes in calibration stability or in overall
imprecision for a measurement procedure.

 Can also be compared for consistency among two or more

methods or instruments for the same analyte

 Target Mean ( Median) & SD can be established as a QC

sample value. And can be used to establish acceptance
rules similar to those used to interpret IQC results.

52
 14-08-2020

LIMITIATION IN USING PATIENT SAMPLES AS QC

 Can supplement traditional surrogate IQC samples
 To monitor a method’s calibration stability & to monitor
calibration uniformity among different methods
 Limitation –

 1. Results selected for mean / median should be

Physiologically homogeneous. - Nonhomogeneous sample of
patients results cannot be used hospital general medicine
inpatients Vs. outpatient clinic / dialysis pts
 2. Lack of consensus guidelines to determine the number
of sequential patient results (in terms of time duration or
number of results) to be included in a single group for
which the mean or median is to be calculated
 3.Lack of computer support from Instrument suppliers
and software support from LIS suppliers.

USE OF PATIENT SAMPLES IN QUALITY

CONTROL PROCEDURES – 3. DELTA CHECK
 by comparing a patient’s current test result against a
previous result for the same analyte
 most useful

 1. to detect mislabeled samples

 2. Samples altered by dilution with intravenous fluid.

 3. Might be useful to identify an interfering substance

 (eg, from a drug) in a patient’s sample.

 Delta Check values can be established in 3 ways

 1. Empirically on experience

 2. Based on a % difference

 3. Calculating & establishing RCV using Cva & BVi & BVg

 CLSI EP33 guideline for using delta checks in medical lab

53
 14-08-2020

USE OF PATIENT SAMPLES IN QUALITY

CONTROL PROCEDURES – 3. DELTA CHECK

EMPIRICAL CRITERIA FOR PATIENT TEST RESULT AGREEMENT BETWEEN

REPEATED ASSAYS AND FOR AGREEMENT AMONG RESULTS FOR A SINGLE
PATIENT SAMPLE MEASURED ON MULTIPLE INSTRUMENTS
Analyte Acceptable Acceptable
difference in difference in units
%
AST,ALT,ALP, LDH, CK, yGT, 10 % 10 U/L & 10 ug/dL
Lipase & Iron
Mg2+,K+ 0.3 mmol/L
Amylase 10 % 15 U/L
Bilirubin, 10 % 0.3 mg/dL
Protein, PO4 Albumin, Uric acid 0.4 mg/dL, 0.4 g/dL
Calcium 0.5 mg/dL
Creatinine 0.2 mg/dL
Na Cl & CO2 4 mmol/L
Cholesterol 5%
TGL 10%
Glucose 10% 6 mg/dL
BUN 10% 3 mg/dL

54
 14-08-2020

THANK YOU !

VERIFYING QUALITY CONTROL EVALUATION PARAMETERS

AFTER A REAGENT LOT CHANGE

QC values for the high-concentration control shifted after

the change to a new lot of reagents, (positive bias) but
there was no change in results for the low control
Comparison of patient samples assayed using the new and old
reagent lots, shows patient results were the same when either
lot of reagents was used - Slope 1.000 & y intercept -3 mg/dL
Why this happens ?

55
 14-08-2020

VERIFYING QUALITY CONTROL EVALUATION PARAMETERS

AFTER A REAGENT LOT CHANGE

Change in QC values for the high-concentration material is

DUE TO A DIFFERENCE IN MATRIX-RELATED BIAS
BETWEEN THE QC MATERIAL AND EACH OF THE
REAGENT LOTS & therefore acceptable

No real problem inspite of shift of QC results – can continue to

run patient samples

VERIFYING QUALITY CONTROL EVALUATION PARAMETERS

AFTER A REAGENT LOT CHANGE

Therefore MUST USE PATIENT SAMPLES & NOT QC

SAMPLES TO VERIFY the consistency of results between old
& new lots of reagents during reagent lot change because of the
unpredictability of a matrix-related bias being present for QC
materials.
Only patient sample results, not the QC sample results,
provide the basis for verifying that the new reagent lot is
acceptable for use.

56
 14-08-2020

VERIFYING QUALITY CONTROL EVALUATION PARAMETERS

AFTER A REAGENT LOT CHANGE

If a problem is identified

1. Calibration of the new reagent lot must be investigated &

corrected

2. Recalibration done

3. New reagent lot may be defective and should not be used

VERIFYING QUALITY CONTROL EVALUATION PARAMETERS

AFTER A REAGENT LOT CHANGE

How many patient samples to use for verifying the performance of

a new reagent lot ? Depends on
1. The Measuring Interval
2. The Imprecision of a MP
3. The concentrations at which clinical decisions are made.
CLSI document EP2655 recommends a minimum of 3
patient samples and more patient samples

57
 14-08-2020

VERIFYING QUALITY CONTROL EVALUATION PARAMETERS

AFTER A REAGENT LOT CHANGE

What are acceptable differences between results ?

no well-established clinical acceptance criteria for agreement
between results;
Lab must establish acceptance criteria based on
1. the relatively small number of samples used
2. the analytical performance characteristics of the MP
3. The clinical requirements for interpreting results.

EMPIRICAL CRITERIA FOR PATIENT TEST RESULT AGREEMENT BETWEEN

REPEATED ASSAYS & FOR AGREEMENT AMONG RESULTS FOR A SINGLE
PATIENT SAMPLE MEASURED ON MULTIPLE INSTRUMENTS
Analyte Acceptable Acceptable
difference in difference in units
%
AST,ALT,ALP, LDH, CK, yGT, 10 % 10 U/L & 10 ug/dL
Lipase & Iron
Mg2+,K+ 0.3 mmol/L
Amylase 10 % 15 U/L
Bilirubin, 10 % 0.3 mg/dL
Protein, PO4 Albumin, Uric acid 0.4 mg/dL, 0.4 g/dL
Calcium 0.5 mg/dL
Creatinine 0.2 mg/dL
Na Cl & CO2 4 mmol/L
Cholesterol 5%
TGL 10%
Glucose 10% 6 mg/dL
BUN 10% 3 mg/dL

58
 14-08-2020

VERIFYING QUALITY CONTROL EVALUATION PARAMETERS

AFTER A REAGENT LOT CHANGE – QC TARGET VALUE TO
CHANGE OR REMAIN SAME ?

Next evaluate QC results for each QC material to determine if its

target value is correct for use with the new lot of reagent(s).
If the target value has changed, it must be adjusted to correct for
the change in matrix-related bias between old and new lots of
reagent(s).
This adjustment keeps the expected variability around the QC
target Value so that QC interpretive rules will remain valid

VERIFYING QUALITY CONTROL EVALUATION PARAMETERS

AFTER A REAGENT LOT CHANGE

If target QC value is not adjusted it will introduce an artifactual bias

in subsequent QC results resulting in wrong decisions while
interpreting QC rules, causing both an increased false-alert rate and
a decreased ability to detect some rejection or error rules

59
 14-08-2020

VERIFYING QUALITY CONTROL EVALUATION PARAMETERS

AFTER A REAGENT LOT CHANGE – WILL SD CHANGE?

The SD used to evaluate QC results WILL NOT

TYPICALLY CHANGE when a new lot of reagent(s)
is put into use. The SD represents expected variability
when the MP is stable and is performing according to
specifications. - In most cases, the variability of a MP
will be the same with any new lot of reagent(s).

WHEN SHOULD IT BE CARRIED OUT ?

Evaluation is usually not required when changing to a new bottle of reagent

or calibrator from the same lot as the constituents of each bottle within a lot
should be almost identical, with a negligible impact on patient results. If
there is significant vial-to-vial instability this can be checked with internal

quality control.

60
 14-08-2020

VERIFYING MEASUREMENT PROCEDURE PERFORMANCE AFTER

USE OF A NEW LOT OF CALIBRATOR

When a new lot of calibrator is used, with no change in reagents, there

is no change in matrix interaction between the QC material and the
reagents.
In this situation, QC results provide a reliable indication of calibration
status with the new lot of calibrator.
If the QC results indicate a bias after use of a new lot of calibrator,
the calibration has changed and needs to be corrected to ensure

VERIFYING MEASUREMENT PROCEDURE PERFORMANCE AFTER

USE OF A NEW LOT OF CALIBRATOR

If Reagents, Calibrators & QC materials are from the same source QC results may fail to
identify a shift in calibration
In such cases it is recommended to use third party QC materials that are independent of the
kit & calibrator lot and
Avoid changing lots of QC material at the same time as changing lots of reagent or
calibrators.
Measuring patient samples always provides a reliable approach to verify the consistency of
results after changes in lots of reagents or calibrators changes in other measurement
conditions

61