Professional Documents
Culture Documents
Laboratory Quality Assurance
Laboratory Quality Assurance
1
Positive predictive value (PPV) and negative predictive value (NPV)
of a serologic test, based on an assay assumed to be 95% sensitive and 90% specific &
PREDICTIVE VALUE assuming disease prevalence of 0.01%, 0.1%, 1%, 5%, 10%, and 50%
2
ACTIVITIES WITHIN EACH PHASE OF
QUALITY CONTROL THE TOTAL TESTING PROCESS
(MMWR 2005, Vol.54/RR13)
3
QUALITY ASSESSMENT
LABORATORY QUALITY CONTROL
• Quality Assessment (also known as proficiency
• Quality Control - QC refers to the measures that testing)
must be included during each assay run to verify that the - is a means to determine the quality of the results
test is working properly
generated by the laboratory. Quality assessment
• Quality Assurance - QA is defined as the overall is a challenge to the effectiveness of the QA and
program that ensures that the final results reported by the QC programs.
laboratory are correct
• Quality Assessment may be external or internal.
4
SELECTING CONTROL MATERIALS SELECTING CONTROL MATERIALS
CALIBRATORS CONTROLS
INTERNAL CONTROLS
TYPES OF CONTROLS
INTERNAL, PROCEDURAL, OR BUILT-IN
CONTROLS
• INTERNAL CONTROLS • Evaluate whether certain aspects of the test
• EXTERNAL CONTROLS system are working properly.
• They are designed to verify that the test system is
working as expected, that sufficient specimen
was added and, for unitized test devices, whether
is migrated through the test strip properly.
• Certain test systems might have electronic
internal controls to monitor electronic functions.
5
INTERNAL CONTROL EXTERNAL CONTROLS
• Mimic patients specimens and monitor the
Reference material: testing process, from specimen application
to result interpretation, to assure proper
– POSITIVE CONTROL test performance.
– NEGATIVE CONTROL
– CUT-OFF CONTROL • They might be provided as liquid or other
materials similar to patient specimens and
might be included with the test system or
purchased separately.
• Controls should be tested concurrent with patient • Record of control results should be periodically reviewed
specimens by the personnel who routinely to detect shifts or changes in performance over time.
perform patient testing.
6
TEST RESULT INTERPRETATION ERRORS IN MEASUREMENT
• When the test is complete, interpret the results • TRUE VALUE - this is an ideal concept which cannot be
according to instructions in the product insert, as achieved.
• Input data required - such as standards used, calibration • OPERATOR ERROR / HUMAN FACTORS
values, and values of physical constants.
• Use of incorrect specimen type
• Inherent characteristics of the quantity being measured
• Incorrect application of the specimen on the device
- e.g. CFT and HAI titre.
• Incorrect placement of device (e.g., non-level surface)
• Instruments used - accuracy, repeatability. • Incorrect placement of reagents including strips, or other components
that contain reagents
• Observer fallibility - reading errors, blunders, equipment
• Use of incorrect reagents, for example, reagents that are not specific for
selection, analysis and computation errors.
the particular device or lot, or generic reagents
• Environment - any external influences affecting the • Incorrect order of reagent application
measurement. • Use of incorrect amount of reagent
• Theory assumed - validity of mathematical methods and • Incorrect timing analysis (e.g., specimen application, running the test, or
approximations. reading results)
• Incorrect reading test results
7
•SPECIMEN INTEGRITY AND HANDLING •REAGENT INTEGRITY
(REAGENT VIABILITY)
• Error in specimen collection
• Use of inappropriate anticoagulant
• Use of improperly stored reagents
• Clotted specimen
• Error in specimen processing and handling • Use of outdated reagents
• Incorrect specimen transport and/or • Use of improperly mixed reagents
storage • Use of contaminated reagents
• Presence of interfering substances
• Presence of bubbles in the specimen
8
RANDOM ERROR
•ENVIRONMENTAL FACTORS
• An error which varies in an unpredictable manner, in
• Impact of key environmental factors (heat, magnitude and sign, when a large number of
measurements of the same quantity are made under
humidity, sunlight, surface angel, device effectively identical conditions.
movement, etc.) on reagents, specimens, • Random errors create a characteristic spread of results
and test results for any test method and cannot be accounted for by
• Impact of key environmental factors applying corrections.
(including electrical or electromagnetic • Random errors are difficult to eliminate but repetition
reduces the influences of random errors.
interference) on instruments, if
– Examples of random errors include errors in pipetting and
appropriate. changes in incubation period.
• Random errors can be minimized by training, supervision
and adherence to standard operating procedures.
SYSTEMATIC ERROR
RANDOM ERRORS •An error which, in the course of a number of measurements
of the same value of a given quantity, remains constant
when measurements are made under the same conditions, or
varies according to a definite law when conditions change.
x
9
SYSTEMATIC ERRORS MONITORING QC DATA
• Use Shewart chart or Levey-Jennings chart
10
Levey-Jennings Chart Levey-Jennings Chart -
Calculate the Mean and Standard Deviation;
Record the Mean and +/- 1,2 and 3 SD Control Limits Record and Evaluate the Control Values
115 115
+3SD
+3SD
110 110
+2SD
+2SD
105 105
+1SD
+1SD
100 Mean 100
Mean
95 95
-1SD -1SD
90 90
-2SD
-2SD
85 -3SD 85
-3SD
80 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 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Day Day
• Imprecision = large amount of scatter about the – +/- 1 SD 68.2% of the 68.2%
mean. Usually caused by errors in technique time
– +/- 2 SD 95.5% of the 95.5%
• Inaccuracy = may see as a trend or a shift, time 99.7%
usually caused by change in the testing process
– +/- 3 SD 99.7% of the -3s- 2s -1s Mean +1s +2s +3s
time
• Random error = no pattern. Usually poor
technique, malfunctioning equipment
11
STANDARD DEVIATION AND INTERNAL QUALITY CONTROL
PROBABILITY PROGRAM FOR SEROLOGICAL TESTING
• Only 4.5% of the time will a value fall outside of 3. the use of statistical methods for interpretation
that range due to chance; more likely it will be (Westgard rules)
due to error
12
SHEWHART CHART 3. WESTGARD RULES
or Levey-Jennings Chart
• The formulation of Westgard rules were based on
100 statistical methods.
90 • Commonly used to analyse data in Shewhart control
80
Antibody Units
70 +3 sd charts.
+2 sd
60
50
+1 sd
Target value
• Westgard rules are used to define specific performance
40 -1 sd limits for a particular assay and can be used to detect
-2 sd
30
-3 sd
both random and systematic errors.
20
10 • There are six commonly used Westgard rules of which
0
– three are warning rules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Assay Run
– and the other three mandatory rules.
VZV IgG ELISA: Target Value = 49 U/ml
WARNING RULES
THE VIOLATION OF RULES
1. Warning 12SD:
– It is violated if the IQC value exceeds the mean by 2SD
• The violation of warning rules should
trigger a review of test procedures, 2. Warning 22SD:
reagent performance and equipment – It detects systematic errors and is violated when two
consecutive IQC values exceed the target value on the
calibration. same side of the mean by 2SD
• The violation of mandatory rules should – Patients results cannot be reported; reject the run
results in the rejection of the results
3. Warning 41SD:
obtained with patient’s serum samples in – It is violated if four consecutive IQC values exceed the
that assay. same limit (mean 1SD) and this may indicate the need
to perform instrument maintenance or reagent
calibration.
13
MANDATORY RULES
Westgard Rules: Mandatory 13SD
1. Mandatory 13SD:
• It is violated when the IQC value exceeds the mean by 3SD.
• Run must be rejected.
100
2. Mandatory R4SD: 90
80
•
Antibody Units
It is only applied when the IQC is tested in duplicate (one exceeds +3 sd
70
the mean by –2SD, and the other by +2SD. +2 sd
60
• This rule is violated when the difference in SD between the +1 sd
50 Target value
duplicates exceeds 4SD. -1 sd
40
• Random error has occurred, test run must be rejected 30
-2 sd
-3 sd
20
3. Mandatory 10x: 10
• This rule is violated when the 10 consecutive IQC values are on the 0
same side of the mean or target value. This detects systematic
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
errors.
• May happen when a new test batch or introduced or changes in the Assay Run
calibration of equipment VZV IgG ELISA: Target Value = 49 U/ml
+3 sd
70
60
+2 sd
+1 sd
applies when only a warning rule is violated.
50 Target value
-1 sd
40
-2 sd
2. Reject the whole test run - this applies only when
30
-3 sd a mandatory rule is violated.
20
10
0 3. Enlarge the greyzone and thus re-test range for
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 that particular assay run - this option can be
Assay Run considered in the event of a violation of either a
VZV IgG ELISA: Target Value = 49 U/ml warning or mandatory rule.
14
WHEN A RULE IS VIOLATED INTERNAL ASSESSMENT
• Warning rule = use other rules to inspect • Objective internal assessment offers
the control points flexible, low-cost options for evaluating
quality such as self-conducted inspections,
• Rejection rule = “out of control” supervisory review of documented
– Stop testing problems that occur in the different phases
– Identify and correct problem of the testing process, review of QC
– Repeat testing on patient samples and controls
– Do not report patient results until problem is solved documentation, and testing and reporting
and controls indicate proper performance procedures.
Summary:
How to implement a QC program?
15