POCT EQA management: running external quality assessment across sites

What external quality assessment is

External quality assessment (EQA), known in the United States as proficiency testing (PT), is a scheme in which a provider sends the same samples to many participating laboratories and point-of-care testing (POCT) sites. The values are not disclosed. Each participant runs the samples as if they were patient samples, submits the results by a deadline, and receives a report comparing their results with a target value, usually the consensus of participants using the same device or method, and with the spread of the peer group.

EQA answers a question that internal quality control (QC) cannot. Internal QC uses material of known value, run on a schedule, to show that the device is stable from day to day; the Levey-Jennings chart and Westgard rules are its tools. EQA assesses performance against the scheme's assigned target and against the peer group using the same method, and it can show whether the device at one site agrees with the same device at another; how far that speaks to trueness depends on how the target was assigned and how suitable the material is for the device. A device can pass internal QC for a year while drifting away from the truth, because the QC targets drift with it. EQA is how that is caught.

Providers of EQA can themselves be accredited, to ISO/IEC 17043, the standard for proficiency testing providers. Participation in an accredited scheme is what most accreditation bodies expect where one exists.

Why EQA is required

ISO 15189:2022, the standard the United Kingdom Accreditation Service (UKAS) assesses medical laboratories against, deals with EQA in clause 7.3.7.3, part of clause 7.3.7 on ensuring the validity of examination results. The laboratory must participate in EQA appropriate to the examinations it performs, the EQA samples must be handled in the same way as patient samples by the personnel who routinely perform the examinations, results must be reviewed, and where performance criteria are not met the laboratory must investigate and take corrective action. Annex A applies the same requirement to every POCT device under the laboratory's responsibility. Where no suitable EQA scheme exists, the clause allows an alternative approach, such as exchanging samples with another laboratory or using certified reference materials, provided it is documented. Our page on ISO 15189:2022 and POCT summarises the surrounding clauses.

UKAS expects participation for every examination on an accreditation schedule, POCT included, and publishes its policy on participation in proficiency testing as TPS 47. At assessment, the assessor will ask for the list of schemes per test and site, sample the returns, and read the investigation of any poor return. A poor return with no investigation is a finding; a poor return with a documented, closed investigation is the system working.

The Medicines and Healthcare products Regulatory Agency (MHRA) guidance Management and use of IVD point of care test devices recommends EQA participation for POCT alongside internal QC, and expects the POCT committee to review performance.

For international readers: in the United States, proficiency testing is a regulatory requirement under the Clinical Laboratory Improvement Amendments (CLIA) at 42 CFR Part 493, Subpart H, for non-waived testing of regulated analytes, with scored events through the year and sanctions for repeated failure. CLIA also prohibits discussing PT results with other laboratories before submission and referring PT samples elsewhere for testing. Waived tests are exempt from the PT requirement, though the College of American Pathologists checklist still expects some form of accuracy assessment. The ISO 15189 requirement is less prescriptive about scoring but identical in intent.

UK EQA providers and what to look for

Three providers cover most POCT in the UK. UK NEQAS is a consortium of scheme centres organised by discipline. WEQAS is a provider based in Wales with a range of clinical chemistry and POCT schemes. RIQAS is the scheme run by Randox. Between them, schemes exist for the common POCT analytes: blood glucose, blood gas and electrolytes, HbA1c, INR, C-reactive protein, haemoglobin, urinalysis and others. Check the current catalogue, because scheme coverage changes. This article makes no claim about the relative performance of providers; they are named only as examples of where to look.

When choosing a scheme for a POCT device, ask:

• Does the scheme have a peer group for your device model, so that your result is compared with like devices and not with laboratory analysers?
• Is the sample matrix suitable for the device? Some POCT devices behave differently with stabilised or lyophilised material than with fresh blood, and the report should say how that is handled.
• How many distributions a year, and how many samples per distribution? More frequent distributions catch drift sooner but cost more operator time at every site.
• Can the scheme register many sites and devices under one organisation, with results submitted per device and a consolidated report for the coordinator?
• Is the provider accredited to ISO/IEC 17043?
• What does the report show: target, peer group mean and spread, your bias, a score, and a trend over distributions?

Where no scheme exists for a test, document the alternative under clause 7.3.7.3: split patient samples with the laboratory at a defined frequency, or run reference material with an assigned value, and review the results the same way.

The EQA cycle in practice

The cycle below is the same for one site or fifty. What scales is the number of hand-offs, and each hand-off is where a sample or a deadline gets lost.

1. Register. Enrol each test, at each site, for each device (by serial number where the scheme allows), and keep the registration list in the quality system. A device that is not registered will produce a result that has no peer group and cannot be scored.
2. Receive material. Material arrives at a central point or direct to sites. Log the receipt, check the storage conditions, and note the deadline for submission on the day it arrives.
3. Distribute to sites. Send samples with instructions, a named recipient at each site, the deadline, and a way to confirm they arrived. Record the date sent and the date confirmed.
4. Run within the window. The operators who normally run patient tests run the EQA samples, on the registered device, following the routine procedure, with no extra care and no repeats unless the procedure would allow a repeat for a patient. Record who ran each sample, on which device, with which reagent lot, on what date.
5. Submit. Enter results per device, in the scheme's units, before the deadline.A second person checks the entry against the device record: transcription slips are a frequent and avoidable cause of a poor return.
6. Review the report. When the report arrives, the coordinator reviews every device's performance, not only the flagged ones, and looks for trends across distributions: a bias that is growing but still within limits is a warning.
7. Investigate poor performance. Open an investigation for every result outside the scheme's acceptable limits, and for any pattern the coordinator judges significant. The next section describes the investigation.
8. Record and close. Document the cause, the corrective action, the person responsible, and the check that the action worked at the next distribution. Consider whether patient results were affected and what was done about it.
9. Report to the committee. Summarise performance by site and device, open investigations, and any device or site whose performance is persistently poor.

What goes wrong in multi-site POCT

EQA in a single laboratory is a routine. EQA across a district of GP practices, a chain of clinics or a group of community diagnostic centres (CDCs) fails in predictable ways, and the failures are almost never analytical.

Samples not run. The material arrives at a site where nobody owns it. It sits in a fridge, or on a desk, past the deadline. The scheme records a non-return, which counts against the site in the same way as a wrong result, and the coordinator finds out when the report arrives.

Results submitted late. The samples were run, but the results stayed on the device or on a form and were entered after the deadline, or not at all. Late submissions are excluded from scoring and are a common finding at assessment of multi-site services.

No record of who ran them. Either the coordinator ran all the samples on a site visit, which satisfies the deadline but not the requirement that routine operators perform the testing, or a site ran them and nobody recorded who. Both remove the value of EQA as evidence of operator competence.

Poor performance never closed. The report flagged a device, the coordinator noted it, and nothing else happened. At the next distribution the device fails again. The assessor reads both reports and asks what was done in between.

Quieter failures. Results discussed between sites before submission, which defeats the purpose and is prohibited under CLIA and contrary to the intent of ISO 15189. Samples run on a spare device that is not the one in routine use. The same enthusiastic operator running every distribution. A device replaced under warranty without updating the registration.

The common thread is ownership across distance, the same problem described in our guide to managing POCT across multiple sites: a cycle that depends on one coordinator remembering fifty deadlines will miss some of them.

Investigating and closing poor performance

A poor EQA return is a nonconformity under ISO 15189:2022 clause 8.7, and the investigation should be recorded as one, with a cause, an action and a check that the action worked. Work through the likely causes in order of frequency.

1. Transcription and submission. Compare the submitted value with the device record. Wrong units, a decimal in the wrong place, results entered against the wrong device or sample.
2. Sample handling. Was the material stored, reconstituted and brought to temperature as instructed? Was it run within its stability window? POCT devices are more sensitive to matrix effects than laboratory analysers, and the provider's report often comments on this.
3. Operator. Who ran it, and were they competent and in date on that device? Check their internal QC history for the same period. A poor return from one operator on a device that other operators run well is a competency question.
4. Device and lot. Internal QC on the day of testing, the reagent or strip lot in use, calibration status, any error flags, and whether the device has since been serviced. Compare with the same device model at other sites in the same distribution.
5. Method. If every device of that model at every site shows the same bias against the peer group, the problem is the method or the material, and the conversation is with the manufacturer or the scheme provider.

Then decide on the patient impact. If the cause could have affected patient results between the last acceptable return and now, the clinical lead decides whether a look-back is needed. Record the decision either way. Close the investigation only when the corrective action has been checked, usually at the next distribution, and keep the whole chain, from the report to the closure, in one place; a nonconformance and corrective action record in the quality management system is the natural home. Persistent poor performance on one device, across distributions and after corrective action, is a reason for the POCT committee to suspend it.

The cycle as a table: step, owner, evidence

The table sets out who owns each step and what an assessor will ask to see. Adapt the owners to your structure; the point is that every step has one.

Where software helps

Most EQA failures in multi-site POCT are failures of tracking: which site has which samples, when they are due, who ran them, whether the investigation was closed. A spreadsheet and a calendar reminder can manage a handful of sites. Beyond that, the coordinator needs a system that holds the scheme register per device and site, tracks each distribution from receipt to submission, reminds the site and the coordinator before the deadline, records the operator, device and lot for each sample, stores the provider's report, and links a poor return to an investigation and corrective action.

The strongest version of this is where the EQA record is joined to the rest of the quality data: the operator's competency, the device's internal QC on the day, the lot in use and the previous distributions on the same device, so that the investigation starts from the evidence instead of from a phone call. Catenix provides EQA management for POCT along those lines, with deadlines, distribution tracking, investigations and committee reporting across sites. Whatever software you use, the test is the assessor's: for any device at any site, can you show the last three distributions, who ran them, what the report said and, if it was poor, how it was closed.

Questions people ask

What is EQA in point-of-care testing?

External quality assessment (EQA), called proficiency testing in the United States, is a scheme in which a provider sends samples of undisclosed value to many participants, who run them as patient samples and submit results. The provider compares each result with a target and with the peer group using the same device. For POCT it shows whether a device at one site agrees with the truth and with the same device elsewhere.

Is EQA mandatory for POCT in the UK?

For a service accredited to ISO 15189:2022, yes. Clause 7.3.7.3 requires participation in EQA appropriate to the examinations performed, and Annex A extends this to POCT under the laboratory's responsibility. UKAS assessors ask for the scheme list per test and site and sample the returns. Where no scheme exists, a documented alternative such as split samples with the laboratory is acceptable. MHRA guidance recommends EQA for all POCT.

What is the difference between internal QC and EQA?

Internal quality control uses material of known value, run on a schedule, to show that a device is stable from day to day; it is reviewed with Levey-Jennings charts and Westgard rules. EQA uses samples of undisclosed value sent by an external provider to show whether results are accurate and comparable with other sites and methods. A device can pass internal QC while drifting; EQA catches the drift.

Who provides EQA schemes for POCT in the UK?

The main providers are UK NEQAS, a consortium of scheme centres organised by discipline, WEQAS, based in Wales, and RIQAS, run by Randox. Between them they cover common POCT analytes such as blood glucose, blood gas, HbA1c, INR, CRP, haemoglobin and urinalysis. Choose by whether the scheme has a peer group for your device, the sample suitability, the frequency and whether the provider is accredited to ISO/IEC 17043.

What should you do after a poor EQA result?

Open a nonconformity and investigate in order of likelihood: transcription and submission errors, sample handling, the operator and their competency, the device and its internal QC and lot on the day, and finally the method if every device of that model shows the same bias. Decide whether patient results could have been affected, record the corrective action, and close only after the next distribution confirms the action worked.

What if there is no EQA scheme for my POCT test?

ISO 15189:2022 clause 7.3.7.3 allows an alternative approach where no suitable scheme exists, provided it is documented and reviewed in the same way. The usual options are splitting patient samples with the main laboratory or another site at a defined frequency and comparing results, or running certified reference material with an assigned value. Record the rationale, the frequency, the acceptance criteria and the review, and check periodically whether a scheme has become available.

Sources and further reading

• ISO 15189:2022 Medical laboratories: requirements for quality and competence (clause 7.3.7.3 and Annex A)
• ISO/IEC 17043:2023 Conformity assessment: general requirements for the competence of proficiency testing providers
• UKAS TPS 47, UKAS policy on participation in proficiency testing
• MHRA, Management and use of IVD point of care test devices
• CLIA regulations, 42 CFR Part 493, Subpart H: participation in proficiency testing