This document discusses quality control in clinical laboratory testing. It emphasizes that quality control is essential to provide reliable diagnostic reports and cost-effective patient care. Quality control involves monitoring precision, accuracy, and sources of variation through internal quality control, external quality assessment, and statistical analysis of control values using tools like control charts and Westgard rules. The goal is to minimize laboratory errors and reliably distinguish pathological variations in patient samples.

1. CLINICAL LAB QUALITY
CONTROL
Dr.S. Sethupathy, M.D,Ph.D.,
Professor of Biochemistry,
Rajah Muthiah Medical College,
Annamalai University.

2. PURPOSE OF LAB TESTING
1. Single patient testing for diagnosis
2. Repeated testing in a patient for
monitoring
3. Testing a population –
Epidemiological work

3. QUALITY LAB SERVICE
Importance
• To produce reliable reports and timely reports
for diagnostic, prognostic and screening
purpose to achieve best patient care
• Cost-effective service to reduce burden on the
patient.

4. QUALITY CONTROL
• Good laboratory practice (GLP) requires Internal
Quality Control (IQC)
For Lab accreditation , QC is a must. Why?
• Quality must be designed from the front end and
not on the back end.

6. TOTAL QUALITY MANAGEMENT
• It is a continuous approach to improve quality and performance
towards meeting out the demands and satisfaction of the
customers.
• It involves :
• Management
• Work force
• Suppliers
• Feed back from customers
• Plan, do, review and act- sequence of QC

7. GOAL FOR QC
• To provide cost-effective , quality lab service
for the best patient care
How to achieve?
1. By increasing the error detection, reliability
and quality is assured.
2. By reducing the false rejection of runs, it will
be cost effective.

8. GOALS OF MEASUREMENT
• Accuracy and Precision
• Accuracy is the closeness of measured value with the
true value.
• The method adopted for clinical lab has certain level of
inaccuracy.
• To minimize inaccuracy ,Quality assurance program is
run .

9. PRECISION
• It is the closeness among replicative
measurements under specified conditions.
• The wide and random dispersion of values of
the replicative measurements indicate
imprecision of the assay.
• Imprecision leads to loss of reliability and
affects serial measurements. eg- bl. glucose.

11. VARIABLE
• Measurement is a varying one.
Variation can be due to
• a. Analytical variation – under control .
• b. Biological variation – Physiological
• c. Pathological variation - our concern

12. VARIABILITY OF LAB MEASUREMENTS
Due to :
• Sample collection
• Sample transport and storage
• Preparation of reagents, performance
• Maintenance of instruments
• Type of method, etc.
• Variation inherent to the biological sample.

13. To obtain reliable results
• All the sources of variation should be
minimized to differentiate pathological
variation.
• To achieve this, Minimize laboratory error.
• Perfectly estimate the components of
biological variation ( The inherent variation of
the sample)

14. CHECK POINTS OF CONTROL
• Pre-analytical
• Analytical
• Post-analytical

23. IQC AND EQA
• IQC is done on daily basis .
• It is useful for accepting or rejecting the runs based on
the variation limits.
• It is done for monitoring the quality of analysis.
• EQA is used to compare the quality between
laboratories of that area and to confirm the IQC
results.

25. HOW TO ACHIEVE QUALITY RESULTS?
• Adopting approved methods
• Adopting proper calibration protocol
• Adhering to maintenance schedule
• Training the technicians
• Internal quality control
• Adopting quality assurance program
• Proper documentation and regular audits for improvement

34. INTERNAL QUALITY CONTROL
• Internal Quality Control (IQC). It includes all
SQC methods which are performed every day by
the laboratory personnel with the laboratory’s
materials and equipment.
• It checks primarily the precision (repeatability
or reproducibility) of the method.

35. EXTERNAL QUALITY ASSESSMENT (EQA)
• External Quality Control (EQC). It includes all SQC
methods which are performed periodically (i.e. every
month, every two months, twice a year) by the
laboratory personnel with the contribution of an external
center .
• It checks primarily the accuracy of the laboratory’s
analytical methods.
• Certain EQC schemes check both the accuracy and
the precision.

36. HOW MUCH VARIATION IS ALLOWED IN
MEASUREMENTS?
• Variation is unavoidable.
• It should be controlled for clinical utility.
• It is needed to reject or accept analytical runs for
reliable reports.
• Fixing up the control limits is the first step of IQC.

37. CALCULATION OF CONTROL LIMITS
• Control limits : They consist of a center value (CL) and
an upper and low control limit (UCL &LCL).
• They are created by repetitive measurements of control
samples.
• In internal SQC two or more control samples are
assayed every day and at least once per day before the
patients’ samples.

38. The steps for their calculation are the following:
1. Collects 20 – 30 successive measurements from any control
level.
2. Standard deviation (s) and mean value (μ) are calculated. The
range μ±3s is considered as “trial limits”
3. If any of the measurements exceed the range μ±3s, the outlier
is rejected. The standard deviation and mean value are calculated
once more.
4. Repeat the previous procedure until no measurement exceeds
the range μ±3s.
5. The final μ and s are the mean value and the standard
deviation of the control limits.

39. • Control limits correspond to a normal distribution.
• “μ” is considered a true value of the daily control
values .
• “s” is equal to the inherent error.
• The mean value of the daily control values is
symbolized as x bar and its standard deviation as
“SD”.
• SD = the inherent error and any other RE.

40. ERRORS AND MISTAKES
• Errors: Non-conforming results with “statistical
meaning” due to non- human action.
• Mistakes: Non-conforming results with “no statistical
meaning” due to the human errors e.g. mixing up
samples, the special samples, the control samples.

41. • Random error- It is any positive or negative
unexpected deviation from the calculated
mean.
• Systematic error- It is a change in the mean
of control values.
• It may be gradual (Trend ) or abrupt (Shift ).
They are reproducible inaccuracies and are
often due to a persistent problem .

42. •TE (total error, total
analytical error -TAE)
The sum of random error
(imprecision) and systematic
error (bias).

43. TOTAL ANALYTIC ERROR

44. LEVEY- JENNINGS CHART
• It is used to plot control values for monitoring
• Done for successive day-to-day or run-to-run .
• A chart is prepared for each test.
• For level of control, the decision limits are first fixed.
• The limits are ± 1s, ± 2s, ± 3s from the mean.

45. LEVEY-JENNINGS CHART

46. RANDOM AND SYSTEMATIC ERRORS

47. RANDOM ERRORS

48. SYSTEMATIC ERRORS

49. QC STATISTICS
Mean
• To find out mean, the sum of the values in the data set
is divided by number of values in the data set.
Formula : ∑ Xn/n
• ∑ - sum
• Xn – Each value in the data set
• n – number of values in the data set
• Two groups having same mean can have different
spread or range.

50. VARIANCE
• Variance is the average squared deviation from the
mean:
• s2 = Σ ( xi - ẋ )2 / ( n - 1 )
• where s2 is the sample variance, ẋ is the sample mean,
xi is the ith element from the sample, and n is the
number of elements in the sample.
• The standard deviation is the square root of the
variance.

51. STANDARD DEVIATION (S)
• It is a measure of precision.
• It indicates how close the QC values are to
each other.
• Wide s indicates the imprecision and loss of
reliability.
• It is also useful for monitoring daily
performance.

52. STANDARD DEVIATION (S)
Formula
• s = √ ∑ (xi - ẋ) 2 / n-1
• s- standard deviation
• ẋ - mean
• ∑ (xi - ẋ) 2 - The sum of the squares of the
differences between individual QC values and the
mean
• n- number of the values in the data set

53. IF ANALYTICAL PROCESS IN CONTROL
• 68% of QC values are within ± 1s
• 95.5% of QC values are within ± 2s
• 99.7% of QC values are within ± 3s.
• 0.3 % of QC values are outside ± 3s and are
associated with significant error . So patient
results should not be reported.

54. STANDARD DEVIATION
• In case of serious loss of precision,
the following things are to be looked in to.
1. Change of reagent
2. Regarding maintenance schedule
3. Recent Calibration
4. Reagent , sample pipettes status
5. Any change of operator

55. COEFFICIENT OF VARIATION
• It is the ratio of standard deviation to the mean of the data
set expressed as percentage.
• CV = ( s / ẋ ) 100
• It is used
1. To compare precision of two methods.
2. To compare two different methods with
different s .
3.To compare instrument performance.

56. Reagent hCG (low level)
CV
hCG (Normal)
CV
hCG( high)
CV
Reagent 1 6.0% 4.5% 12%
Reagent 2 5.7% 5.0% 11%
Reagent 3 15.2% 5.7% 10%

57. DECISION ABOUT OF THE CHOICE OF
THE KIT
• Since the clinical decision about pregnancy
is based on low level, only kit 1 or 2 is
useful and so the kit 3 is not selected.
• Imprecision and inaccuracy are most
important at the clinical decision levels.

58. WHAT IS AN ACCEPTABLE CV?
It is decided based on
• Precision information given in the product insert or
instrument manual
• Inter-laboratory comparison (QC) programs
• Proficiency surveys
• Evaluations of methods, instruments published in journals
• CLIA proficiency limits (USA)

59. COEFFICIENT OF VARIATION RATIO (CVR)
• It is useful for comparing the precision of the lab test with peer
lab group using the same reagent, same instrument.
• CVR = Within lab CV / peer group CV
• If ratio is less than 1, the precision is better than peer group lab.
• Ratio greater than 1.5 needs investigation for the cause of
imprecision.
• Ratio greater than 2 needs troubleshooting and corrective
action.

60. CONTROL CHART
• The control chart is a graph used to study how a process
changes over time. Data are plotted in time order.
• A control chart always has a central line for the average, an
upper line for the upper control limit and a lower line for the
lower control limit.
• Lines are determined from historical data. By comparing current
data to these lines, you can draw conclusions about whether the
process variation is consistent (in control) or is unpredictable
(out of control, affected by special causes of variation).

62. WESTGARD RULES
• Dr.James Westgard’s six rules are used
individually or in combination to evaluate the
quality of analytical runs.
• The rules are expressed as NL where N stands
for number of control observations and
subscript L stands for statistical limits for
evaluating control observations.

63. WESTGARD RULES
• Rule 12s - single control observation lies outside the ± 2s
limits. It is a warning sign and the run is not rejected.
• In the absence of error, 4.5% of quality results lie between ± 2s
and ± 3s limits.
• The relationship between current, previous analytical runs
should be examined .
• If no source of error is found, it is an acceptable random error
and the patient results can be reported.

65. APPLICATIONS OF RULE 22S
Within run : It affects the current analytical run. If both
level 1 and level 2 controls are greater than 2s limits
on the same side of the mean, this violates within run
for systematic error.
Across runs : If level one is -1s and level 2 is +2.5s,
then previous run is examined and if level 2 is more
than + 2s, then it violates for systematic error and the
current run is rejected.

67. • Rule 13s - Any QC value outside
3s limits violates this rule for
random error and the run is
rejected. It may be the beginning
of a large systematic error.

69. Rule R4s : It identifies random error within
run only .
• If there is more than 4s difference between the
controls in the same run, this rule is violated for
random error. Eg: If level 1 is +2.8s and level 2 is
– 1.4 s , the difference 4.2s . It violates the run for
random error and the run is rejected.
• For three level controls, two of three shows more
than 4s difference, the run violates for random
error and it may be the beginning of systematic
error. So it should be rectified.

71. • Violation of the following rules does not require rejection of
analytical runs. These indicate smaller systematic error or
analytical bias which may not be clinically relevant. These can be
eliminated by calibration and or instrument maintenance.
• Rule 31s : Three consecutive results greater than 1s on the
same side of the mean.
• Rule 41s : Four consecutive results greater than 1s on the
same side of the mean.
• Violation of these rules - applications
• If level 1 control is used, it is systematic bias over a narrow
concentration. If three level controls are used in combination , it
violates for systematic error over a broader concentration.

73. • Rule 10 ẋ : 7 ẋ, 8 ẋ, 9 ẋ, 10 ẋ, 12 ẋ
• These rules are violated when 7or 8 or 9 or 10 or
12 control results are on the same side of the
mean regardless of the s.
• Applications
• If level 1 control is used, it is systematic bias over
a narrow concentration. If three level controls are
used in combination , it indicates systematic bias
over a broader concentration.

75. QC LOG
It should contain :
• Name of the test
• Name of the instrument
• Units
• Date
• Initials of the person performed
• Results of each level of control assayed
• Actions taken in case of out of control QC values
• Name of the method
• Assay temperature in case of enzymes
• Supervisor’s notes

76. Thank you