When specimens are analysed in the clinical laboratory, it is crucial that the results that are reported back to the clinician are accurate. If not, at best, results would not be comparable from the same laboratory on different days or between different laboratories at all. At worst, disease could remain undiagnosed or over-diagnosed with the consequence that patients are put on inappropriate management pathways. CLI caught up with Finlay MacKenzie, Consultant Clinical Scientist and Director of Birmingham Quality, the largest of the UK NEQAS Chemistry Centres, to find out more about how accuracy in clinical laboratory test results is ensured.
On a day-to-day or batch-by-batch basis, how is quality checked?
Essentially a clinical laboratory needs to reassure itself that the results it produces are fit to report back to their clinicians and on a day-to-day basis the first step in this process is internal quality control (IQC). In clinical chemistry, IQC is an ongoing process. The original theory behind IQC was developed for batch analysis where batches of specimens were analysed in blocks. Nowadays everything is done in a random access fashion, which is essentially multiple continuous batches, but IQC is still required and can be run at different frequencies throughout the day (depending on risk) and again there are a variety of approaches that laboratories can utilize and a whole plethora are used.
Most IQC protocols are variants of the so-called Westgard Rules (www.westgard.com) which have been used by clinical laboratories for decades. IQC material is often heavily manipulated and noncommutable since its function is primarily looking as consistency, not accuracy, as such. It tends to be manufactured in very large volumes to accommodate worldwide need.
There are some newer IQC software packages that are being run that allow laboratories to send in the results of their IQC material to either the IQC provider and/or the analyser/system manufacturer in real time to get some idea of peer agreement.
Is there any further oversight of quality assurance?
Yes, there is external quality assessment (EQA). At one level, IQC could be seen as consistency in the laboratory producing the same results time after time, and EQA could be considered as looking at the relative bias of results and if there are commutable reference materials, the EQA provider could give a view on absolute bias as well as just relative bias.
Laboratories that offer a clinical service are, by and large, accredited to an international standard ISO15189:2012 and as part of that accreditation process laboratories have to undertake some form of inter-laboratory comparison, and this is usually EQA from an ISO17043:2010 accredited provider. So EQA would be an adjunct to any IQC processes that the laboratory is undertaking and is a pivotal part of a laboratory’s quality management, generally. The term proficiency testing (PT) and EQA are deemed in some circles to be one and the same thing, and certainly the international standards that govern EQA often refer to PT. However, professionals in the field who provide EQA services as their job and career certainly distinguish between the two. They would assert that PT is really only the basics of what can be achieved by ‘value-added’ EQA.
The mechanics of PT and the basics of EQA is to distribute an identical material to all participating laboratories who then Role of external quality assessment in clinical laboratory quality assurance When specimens are analysed in the clinical laboratory, it is crucial that the results that are reported back to the clinician are accurate. If not, at best, results would not be comparable from the same laboratory on different days or between different laboratories at all. At worst, disease could remain undiagnosed or over-diagnosed with the consequence that patients are put on inappropriate management pathways. CLI caught up with Finlay MacKenzie, Consultant Clinical Scientist and Director of Birmingham Quality, the largest of the UK NEQAS Chemistry Centres, to find out more about how accuracy in clinical laboratory test results is ensured. April / May 2021 7 | analyse it as if it were a clinical specimen. They then return their results to the EQA provider who undertakes statistical analysis to look at the overall spread of data and how each participating laboratory has performed in terms of closeness to the expected target. There are expected limits of performance that need to be met.
EQA could be considered as part of an audit cycle. A laboratory is essentially working in isolation, it doesn’t know if it is getting the same results as other people using its methodology. Over and above what PT achieves, EQA looks at all methods and all laboratories nationally and even internationally.
So what does EQA achieve beyond PT?
Like PT and in contrast to IQC, EQA is performed retrospectively. We send a small number of samples, typically three samples a month, to a laboratory. They would then analyse these specimens as if they were clinical specimens and return the results to the EQA scheme organizers. After doing statistical analysis the scheme organizer would be able to tell the laboratory if their results were close to the true value, whether they were similar to other people using that methodology and could give a sense of trend data as to how that laboratory had been performing over the previous number of months. The reason that EQA is required is that despite the fact that IQC is performed regularly and hopefully rigorously, there are all sorts of variations between the biological reagents that are used – there are lot-to-lot differences in both the reagents, standards and calibrators used. The maintenance and set-up of the instrument itself can also lead to variation.
Everything that I have said so far is true for both PT and EQA. However, EQA delivery would do all of that and more, looking at the pre-analytical and post-analytical as well as analytical phases. It would be distributing more frequent and more probing specimens. EQA scheme design is robust enough to accommodate and react to observations and changes, planned and unplanned, in current methodologies in real time. The material used is minimally manipulated so is as commutable (behaves the same) to a real patient specimen as possible. EQA is much more educational in delivery and would have things like scientific commentaries and adds to the evidence base.
The role of EQA is to push for harmonization of results (not just report the differences), with the idea that any result produced anywhere would be not only the same as another result, but would be the same correct result – so getting result agreement and also agreement on the right answer.
A well-designed EQA programme should be able to identify poorly performing laboratories. A higher aim, from my point of view, is that EQA should give reassurance to the vast majority of laboratories but can still, nevertheless, be able to pick up on the small number of laboratories who are not doing so well. The scheme design which includes challenging laboratories with replicates, adequate concentration range coverage, recoveries and sensitivity should be able to do both. It is important to distinguish between problems that are unique to a particular laboratory and those which are common across all users of that methodology. EQA providers work in a triangular relationship with diagnostic manufacturers and participants. In the UK there are other bodies, such as the National Quality Assurance Advisory Panels, which generally can assist with encouraging individual laboratories to better engage with performance issues. Likewise, if it’s a manufacturer issue and if there are really serious concerns about an in vitro diagnostic device these concerns would be reported to the MHRA (the UK government Medicines and Healthcare products Regulatory Agency).
What sort of extra information is provided with EQA schemes?
UK NEQAS (UK National External Quality Assessment Service) has the widest coverage of EQA across the field of laboratory medicine. UK NEQAS is divided into centres that run EQA programmes that specialize in different areas of clinical laboratory testing. I am the Director of Birmingham Quality, the biggest of several UK NEQAS centres in clinical chemistry, which is operated as an NHS department | 8 Expert Opinion from one of the country’s major teaching hospitals. These different UK NEQAS centres across the UK share a common code of practice and is a charitable consortium of EQA providers to provide harmonization, though all the service provision still resides at the hospital/ university level. We aim to improve patient care through monitoring the quality of tests and their reporting, in an independent manner and on a not-for-profit basis. We are here to help ensure optimal quality in testing for the benefit of patients. We believe that the result of tests should be comparable, safe and clinically useful to the patient no matter where or when they are performed. We believe that the customer of our services is ultimately the patient, not the laboratory.
To give you a better idea of the work that we do, I’ll briefly mention some of our programmes.
Serum Indices (HIL) programme
For example, one of the newer EQA programmes that we run concerns serum indices and we look at the impact of hemolysis, icterus and lipemia (HIL indices). We evaluate the quality of a serum specimen, and determine whether or not a laboratory can get the right result even if the sample is hemolysed, icteric or lipemic, both in terms of the spectral interference and, in the case of hemolysis, the impact of cell content leakage into the serum. So we would be probing to see if the assay will give the right answer under those situations. This pre-analytical area is one where laboratories are in control of their local processes and there are a number of situations that we have highlighted in case studies in clinically important analytes such as troponin, creatinine and potassium, which can have life and death consequences.
Steroid Hormones programme
In the Steroid Hormones programme in the endocrinology field where immunoassay is still the predominant methodology, we might be looking to see if there are other interfering compounds, for example the commonly prescribed steroid prednisolone, that could give rise to a compromised result in cortisol assays as it is structurally similar. We have looked at a range of structurally similar homologues and metabolites to assess the accuracy of such assays in challenging situations. In general, having endogenous levels and not trying to have every analyte in a single tube really is the best approach, scientifically.
Glomerular Filtration Rate (GFR) Estimations programme
In the Glomerular Filtration Rate (GFR) Estimations programme the most common test for kidney function is creatinine measurement. There has been a shift to better, more specific assays over the last 15 years but historically the most common method principle (Jaffe) in use in the UK is interfered with by the presence of glucose. For some methods you get 1 μmol/L of apparent creatinine for every 1 mmol/L of glucose. So if you had a diabetic patient with a glucose of 30 mmol/L then the apparent creatinine would be 30 μmol/L higher than it actually was. That patient could then easily be put into the wrong treatment category because the clinician might understandably think that the kidney function was a lot worse than in reality. So that’s the sort of value-added experiment that we would do to show that this is the case. The manufacturers do generally know the limitations of their assays and this information can be found in the instructions for use but not everyone has read the small print or appreciates its significance.
Pre and Post Analytical Quality Monitoring Service
As well as the concentration ‘numbers’ that we assess, we collate and analyse all sorts of interpretations from, for example, negative, equivocal or positive as part of the regular schemes, all the way through to dedicated schemes that we have asking clinical questions and getting the answers marked by a peer-group of experts in the field – the Interpretive Comments EQA programme. Taking assessment to its logical limits, working with our colleagues across all the disciplines, we also address pre- and post-analytical issues in the UK NEQAS Pre and Post Analytical Quality Monitoring Service. This is a genuine pan-discipline UK NEQAS activity, looking at preand post-analytical error rates, which is more of a bench-marking activity where labs can see if their failure rate (for example, number of mislabelled samples, or hemolysed samples, or samples that are not reported within the right time period) is better or worse than their peers. Essentially, there is not a part of the testing service that we do not cover.
With the recent and ongoing COVID-19 pandemic, many new tests have been developed. What has UK NEQAS had to do in response to this?
I think that UK NEQAS responded very quickly and thoroughly to the national pandemic. The UK went into a period of national lockdown in March 2020 and by the end of May 2020, several UK NEQAS centres offered Covid-19 related services. The Public Health England Centre in Colindale, London had launched a SARS-CoV-2 molecular exercise and the UK NEQAS Immunology, Immunochemistry & Allergy Centre in Sheffield had launched a SARS-CoV-2/COVID-19 Antibodies service, which was quickly ISO17043:2010 accredited.
My centre assisted with some of the computing elements for these. The Edinburgh-based GenQA molecular genetics centre offered educational and on-line competency assessment which included the basics of sample collection and analysis. My centre kept a full service operating throughout the lockdown, which was a bit of a logistical challenge, but we felt that it was more important than ever to keep our services going. We knew that some departments had to move staff around and re-jig their own services and so it was important to give them, and the patients they serve, the reassurance that the quality of laboratory testing was not compromised.
How do you see the future development of EQA?
At the local UK Clinical Chemistry level, there are many changes to service provision across the country that we have to respond to. Whether this is producing reports for the newly formed Networks or Dashboards for everyone, we always adapt and April / May 2021 9 | change to remain relevant. We are particularly proud of working with the National Pathology Exchange (NPEx) to offer a fully minimal hands-on EQA service. We have an Order Comms approach that means that when our EQA samples arrive in the laboratory they are already booked-in. The results get released through the laboratory’s usual electronic systems without the laboratory’s manpower requirement usually associated with EQA specimen handling. We have been running this since 2017 and have a hundred analysers registered over 25 laboratories accepting over a quarter of a million results since launch and I believe we are the only centre offering this type of service.
EQA is becoming more international. Because the same diagnostic manufacturers are providing assays globally, it would be no surprise that the same issues affect more than one country. I was until recently the secretary for EQALM (the European Organisation for External Quality Assurance Providers in Laboratory Medicine), which is very international in its outlook. Although it’s European, it has got Australians, Americans (north and south) in it as well, so it is truly worldwide. There are many international harmonization initiatives going on related to assay systems as well and in which the UK, as a large and longstanding provider of EQA, is well represented.