EQA and IQC Flashcards
What is EQA and IQC and what the importance of them in the clinical laboratory?
EQA is retrospective analysis of performance of an assay in comparison to other users (assesses whether you are getting the same result as someone in a different geographical location. Hence can determine if the patient have continuity in results if they move around hospitals). It is important for accuracy.
IQC is used in real time to monitor the performance of an assay, therefore can measure the precision of an assay (trying to maintain the reproducibility of analysis, to verify that the procedure is happening as best as possible)
EQA and IQC are clinical practices used to measure quality of testing performed in the clinical laboratory, which ensures: The best patient outcome The best patient experience Safe practice Confidence in results produced
What is the process of choosing IQC material?
IQC material is generally bought commercially but it depends on analyte as to whether this is possible.
It can contain one or multiple analytes.
Analytical results monitored by using a computer programme - if it is manual you have to keep up to date with this (eg COBAS IT for WEQAS) - this is for the audit trail required by UKAS.
When choosing IQC material it is important to consider the following:
The matrix:
Close matrix match, eg serum matched with serum (not aq)
Human preferable, but not always possible
Third parties:
At least one QC used should not be made by the assay manufacturer, better test as to whether assay is working properly. Same goes for working with kits.
The cost:
Most cost effective without compromising on quality
The levels available:
Minimum of 2 – normal (physiological) and high (pathological)
Near cut off value
The range of analytes available:
Able to QC multiple assays
The different QCs you select need to cover your whole assay range
Save money if your using one IQC with lots of analytes
The stability:
Use repeatedly, must be usable over a long period of time so that you can compare over this period
IQC materials don’t contain everything, as some analytes cant be present together e.g. high levels of bilirubin interfere with the measurement of other analytes as it makes the sample very icteric and interferes with colourimetric assays - must buy a separate.
Assayed IQC material is more expensive (range of targets for different methods), whereas unassayed IQC material, which has no target values, is much cheaper.
What is the process of analysing IQC material?
Common practice is to accept IQC values within + 2SD of the mean and report patient results
If IQC is >+ 2SD but + 3SD, reject the run
Practice varies between labs
Depends on the assay - eg with a screening test, you may lower the cut off, erring on the side of caution.
Also if if two QC levels pass and the third fails, but all your patient samples are close in concentration to the first two then you may authorise these
What is the process of EQA?
- Receive a set of samples (normally 3 or 4 depending on the scheme)- don’t know the values.
- Processed as normal samples, in order to assess the entire testing process.
- Results reported to the EQA scheme.
- Report generated.
You keep a record of all your results in a managed system - booked in as a patient sample so will be recorded as such. If you do not meet the performance criteria determined by the EQA scheme,
the scheme will warn you, with a letter of poor performance. The schemes have put in place poor performance criteria which expect a 100% return rate. Return rate fail, or non-response to formal SO contact are becoming an automatic panel referral.
What may a poor EQA result reflect?
EQA assesses inter-lab variation (is a method consistent
between labs?) and inter-method variation (do different methods
perform differently?)
What is the problem? Poor A, B or C score, is it a random error?
Is there a pattern? Ie is the bias pattern proportional or constant?
Things to consider for isolated incidents:
- was the result correctly transcribed?
- was the correct sample analysed?
- was there a sampling error?
- was it reported in the correct units
Things to consider for persistent issues:
- is there a method flatted issue?
- is the equipment correctly calibrated?
- are you in the correct method group?
To elucidate whether it was a random error leading to a poor A or B score, and why, check the IQC:
- Single random error, was the QC all OK?
- What does the precision of the assay look like?
- If this one sample has had a random error, could this be happening at other times?
- If you have a continuous bias, you are not performing acceptably- this should show up in your IQC
- If not, target and SDs of your IQC should be re-evaluated – the current settings are not picking up issues as they should
A poor C score reflects the consistency of bias:
- is the positive/ negative bias always to the same degree?
- is it concentration-dependent?
- has there been a stepwise shift (new calibrator, maintenance, lot change, etc)
- is there analytical imprecision?
What is the impact of poor IQC/EQA results on the patient?
IQC ensures patients analysed in the same lab will receive continuity of results and therefore clinical care.
Patients analysed in different labs should have equivalent results QA provides confidence that this is true.
Developments in EQA are bringing in MAPS (minimum acceptable performance standards) scores to determine more clinically the impact on the patient. This is a single scoring system with specific standards for each analyte. Derived from intra- and inter-individual biologic variation data, currently applied only to analytes where values have particular clinical significance, eg HbA1c and HDL cholesterol. HOWEVER, this evidence-based medicine can only be practiced if results are commutable between labs.
What is EQA material?
Obtained by signing up to an EQA scheme (every assay SHOULD be subject to one - maybe in another country!)
All assays should be subject to EQA - though each needs a certain number of participants
If not sample exchange scheme - not as good, but at least you are comparing - requirement for UKAS
How often does IQC need to be performed?
Depends on frequency of analysis
- Batch analysis requires IQC at least at the beginning and end of a run
- If large number of samples also benefit from IQC in the middle of a batch as well.
If you know your assay drifts after a certain amount you will customise
Continuous flow analysis requires IQC throughout the day
Depends on local policy
With new calibration/reagent lot
As you can only compare two IQCs that you know to be right
You can only guarantee the accuracy of results between acceptable IQC results
How often is IQC material used?
Depends on frequency of analysis
- Batch analysis requires IQC at the beginning and end of a run
- If large number of samples also benefit from IQC in the middle of a batch as well. Or if you know your assay drifts after a certain amount you will customise.
- Continuous flow analysis requires IQC throughout the day
- Frequency depends on local policy, but it must be with each new calibration/reagent lot.
As you can only compare two IQCs that you know to be right
You can only guarantee the accuracy of results between acceptable IQC results.
- eg if calibration 1 and 2 are acceptable, but 3 fails, you only know test some up until 2 are okay and you must repeat all there after.
Also depends on the analyte:
- Chemistry/Immunoassay
- Stability of calibration
- Reagent lot changes
Operator change (manual assays) can have a big effect on assay quality - should perform IQC between operator change over
What are the target values for IQC?
In order to work out what the target values are you need to run up IQC (determine intermediate precision - different days and different batches, idealistic to do 20 tests on different days to get a mean, but depends on frequency of use and cost of assay use)
Assayed IQC material is provided with target values
Range of values tends to be huge!
Assay should be able to perform in a much more reproducible form within controlled conditions of your laboratory
New IQC material requires mean to be determined
Standard deviation should be checked but should not change between IQC lots.
Target value range is usually defined as the mean +/- 2 standard deviations. Standard deviations will vary greatly between different assays, dependent on method and difficulty of precision - must be appropriate for the analyte.
What kind of errors can result in failure of an IQC?
Systematic:
Appear as a shift or trend representative of a proportional or constant bias that affects the accuracy. This can be caused by:
- Incorrectly assigned calibrator
- Calibration lot changes
- Reagent lot changes
- Light source deterioration
- Reagent/calibrator deterioration
There are three types of systematic errors:
1. Proportional (divergence at high concentrations)
2. Constant (same amount of error at all concentrations)
3. Mixed (a combination of proportional and constant)
Random:
Appear as random deviation from the lab mean. Affects precision/reproducibility. This is caused by:
- Inaccurate pipetting
- Analyser mis-sampling
- Sample contamination
Be aware, repeated random errors may indicate a problem!
