Transcript Quality Assurance in Haematology

Quality Assurance in
Haematology
Introduction
 QA is the sum of all those activities in which the
laboratory is engaged to ensure that information
generated by laboratory is correct.
 QA is not restricted to the development and retention
of quality control charts but rather includes all aspects
of laboratory activities that affects the results
produced, from the choice of methods, to the
education of personnel, to the handling of specimens
and reporting results.
 The real purpose of QA activities is to determine how
correct or incorrect the results emanating from the lab
are, and to allow those managing the lab to determine
whether or not the lab is fulfilling its functions
satisfactorily.
Introduction – cont’
 3 major activities of QA :
1 ) Preventive – those activities that are done prior
to the examination of the specimen or sample and
that are intended to establish systems conducive
to accuracy in analytical testing ( eg : preventive
maintenance and calibration of instruments,
testing of media, orientation and training of
personnel )
Introduction – cont’
2 ) Assessment – those activities that are done
during testing to determine whether the test
systems are performing correctly ( eg : the use of
standard and controls, maintenance of control
charts )
3 ) Corrective – those activities that are done, when
error is detected, to correct the system ( eg :
equipment troubleshooting, recalibration of
instruments )
QA in Haematology Laboratory
 QA in haematology lab is intended to ensure the
reliability of the lab tests.
 The objective is to achieve precision and
accuracy
 4 components of QA programme :
1 ) Internal Quality Control ( IQC )
2 ) External Quality Control ( EQC )
3 ) Standardization
4 ) Proficiency surveillance
 Accuracy
- the closeness of the estimated value to the true
mean
- can be checked by the use of reference
materials which have been assayed by
independent methods of known precision
 Precision
- reproducibility of a results, whether accurate or
inaccurate within a define frame time ( eg:
within the same day, from week to week etc )
- can be controlled by replicate tests, check tests
on previously measured specimens and
statistical evaluation of results
Internal Quality Control ( IQC )
 based on monitoring the haematology tests
procedures that are performed in the lab
 includes measurements on specially prepared
materials, and repeated measurements on routine
specimens, statistical analysis and day by day
data obtained from the tests which have been
routinely carried out
 is intended to ensure that there is continual
evaluation of the reliability of the work of the lab
and that control is exercised over the release of
the results
 it is primarily a check of precision but not
necessarily accuracy
External Quality Control ( EQC )
 is the objective evaluation by an outside agency
of the performance by a number of laboratories
on material which is supplied specially for the
purpose
 is usually organized on a national or regional
basis
 analysis of performance is retrospective
 the objective is to achieve comparability, not
necessarily accuracy unless the specimens have
been assayed by reference lab, using methods of
known precision, alongside a reference
preparation of known value
Standardization
 Refers to both materials and methods
 A material standard or reference preparation is
used to calibrate analytic instruments and to
assign a quantitative value to calibrators
 A reference method is an exactly defined
technique which provides sufficiently accurate
and precise data for it to be used to assess the
validity of other methods
 International reference preparations are not freely
available for routine use but are intended to act as
standards for assigning values to commercial ( or
lab produced ) ‘secondary standard’ or calibrators
Proficiency surveillance
 Implies critical supervision of all aspects of
laboratory tests ( collection, labelling,
delivery, storage of specimens before the
tests are preformed and of reading and
reporting of results )
 Also includes maintenance and control of
equipment and apparatus
 It is necessary for correct interpretation of
test results, for the lab to establish normal
reference values that are valid for their test
methods and their local population
Control materials
 Specially prepared
 It may be anticoagulated WB, preserved pooled red cells,
plasma or serum
 It can be used to check for accuracy if the value has been
reliably determined ( eg : reference centre )
 Should have controls of high, normal and low values
 At least 1 control specimen should be used for every
batch
 If large specimens, use 1 control for every 20 specimens
 The same material is used for EQC or also as calibrator
 A reference standard is a substance ( eg: Hb std ) that has
been characterized by chemical or physical means and
has been specific values.
 International reference standard are manufactures by or
on behalf an international organization (eg : WHO and
International Council for Standardization in
Haematology)
 A reference material is a substance ( eg : plasma ) or
device ( eg : reference beads ) that conform to national ar
international reference std. This material is used to ensure
accuracy of the test procedure
 A calibrator is a substance or device that conforms to a
reference std or specification. It is used to calibrate an
apparatus or to adjust a measurement to obtain accurate
results
Statistics of QC
 Mean ( x ) the total score of all the measurements
divided by the number of measurements
 Standard deviation ( SD ) – variation in measurement
obtained in lab tests
formula: SD =
Σ ( x – x )2
n–1
n : number of measurement
 Coefficient of variation ( CV ) – relation of SD to the
actual measurement
formula : CV = SD X 100%
x
Analysis of data
 Standard deviation of control specimens
- if value assigned to a specimen a number of
times, dispersion of results around the mean will
indicate the error of reproducibility
- 95% of results on the same specimen should be
within ± 2 SD and 99.7% within ± 3 SD
- by chance, 1 in 20 of measurement might
expected to fall outside 2 SD and only 1 in 333
outside 3 SD
- If measurement more widely dispersed, this
indicates an error in the test
 Control Charts
- originally described by Shewhart, 1st applied in clinical
chemistry by Levey and Jennings
- Samples of the control specimen are included in every
batch of patients’ specimens and the results checked on a
control chart
- To check precision, it is not necessary to know the exact
value of the control specimen
- Value has been determined reliably by a reference
method, the same material can be used to check accuracy
or to calibrate an instrument
- If possible, controls with high, low and normal values
should be used
- Advisable to use at least one control sample per batch
even if the batch is very small
- The results obtained with the control samples can be
plotted on a chart
 Cumulative Sum Method ( CUSUM )
- is the running total of the difference bet each
measurement and the established mean of the
control tests
- It provides another way to display the data
obtained in the precision test
- More sensitive than control chart – faulty
technique or instrument
- Useful for detecting a change in performance d/t
drift ( a consistent error in one direction ) or a
slight progressive drift away from the original
mean
- The size of change that the test is design to detect
can be varied for individual test, clinical
significance
 Duplicate Tests On Patient’s specimens
- provides another way of checking the precision
of routine work
- Test 10 consecutive specimens in duplicate under
careful conditions
- Calculate the differences between the pairs of
results and derive the SD
- Subsequent duplicate tests should not differ from
each other by more than 2 SD
- This methods will detect random errors but it is
not sensitive to gradual drift nor will it detect
incorrect calibration
- Impractical for routine blood counts in a busy lab
- A few consecutive specimens in a batch should
be tested from time to time as a rough check
 Use Of Normal Haematological Data For QC
- in healthy individuals the blood count remains virtually
constant day by day
- It is possible to use observations on healthy individuals
for quality control I routine lab
- Analyze 5-10 healthy subjects at intervals and calculating
mean and SD for MCV, MCH and MCHC
- On each occasion the mean should not vary by more than
2 SD and the SDs should remains constant
- A significant difference in mean indicates a constant error
eg : incorrect calibration
- A random error will result in an increase in SD although
the mean may be unaffected
 Use Of Patient Data For QC
- to start this programme it is 1st necessary to assay
samples from at least 300-500 patients in an
automated blood counter and to establish the
means of MCV, MCH and MCHC
- Analyse the results in successive batches of 20
specimens
- By plotting this results on graph, any drift from 3
indices can be readily recognized and used to
identify instruments faults
- This method appears to be as accurate and
precise as the of preserved blood controls and is
especially sensitive to changes in RBC
Internal QC
 Intended to monitor various aspects of test
performance which performed in the lab
 Measurement on specially prepared materials,
 repeated measurements on routine specimens as
well as statistical analysis, day by day, of data
obtained from the routine tests which have been
carried out in the lab
 Measures provide a way to achieve precision
External QC
 The major purpose – to achieve harmonization
concordance between labs
 The principle is that the same material is sent
from a national or regional centre to a large
number of laboratories
 All the labs send the results back to the centre
where they are analysed and interpreted by one of
several procedures
 From the results returned from the participants,
the median or mean and SD are calculated
 An individual lab can then compare its
performance in the survey with that of other labs
and with its own previous performance ( using
deviation index )
 A deviation index ( score ) =
actual results – weighted median / mean for test
weighted SD
Interpretation :
0.5 – excellent
0.5-1.0 – satisfactory
1.0- 2.0 – acceptable
> 2.0 – defect requiring attention
 Assigned values
- values are assigned to the test materials by
expert lab who also establish the SD of the
method under optimal condition
- To ensure that results are clinically
reliable, they should be within a certain
percentage of the assigned value
 Organizations –
A ) National External Quality Assurance
Programme ( NEQAP )
Pusat Darah Negara
B ) Royal Collage of Pathologists of
Australia ( RCPA )
C ) International EQA Scheme for
Haematology ( IEQAS )
References
 Good Laboratory and Clinical Practices –
Techniques For Quality Assurance
Professional, P A Carson and N J Dent,
1990
 Quality Assurance Practices for Health
Laboratories, Stanley L. Inhorn, 1978
 Royal Postgraduate Medical School
University of London, Dr. S. M. Lewis