Guide
EQA failure investigation, step by step
An EQA flag is a signal, not a verdict. Investigate in a fixed order: check the return itself for clerical error, check what IQC was doing at the time, reconstruct how the sample was handled, then look at method and lot effects across the peer group. Document each step, decide whether patient results could have been affected, fix the cause, and check the fix worked at the next distribution.
First, breathe
Single poor scores happen to good services. Assessors are rarely alarmed by a flag; they are alarmed by a flag with no investigation behind it. The difference between a finding and a footnote is your paperwork.
The investigation sequence
- Clerical check. Was the right result submitted for the right sample on the right device? Transposed digits and unit errors cause a meaningful share of all EQA failures. Five minutes here can close the whole investigation.
- IQC at the time. Pull the QC records bracketing the EQA sample. If IQC was clean, a large one off error becomes less likely and handling moves up the list. If IQC was drifting, you may already have your answer, and a look back at patient results is now on the table.
- Sample handling. EQA material is not a patient sample: reconstitution, storage, timing and matrix all bite. Reconstruct who ran it, when, and exactly how. Interview without blame; you want the truth, not a confession.
- Method and peer group. Read the report properly: how did your peer group do, is there a lot wide or method wide effect, is your bias consistent across distributions? A persistent bias against the peer group with clean IQC points at calibration, lot or a genuine method limitation.
- Device and environment. Maintenance records, error logs, temperature excursions, software changes around the distribution date.
The patient question
Every investigation must answer one question explicitly: could patient results have been affected, and if so over what period? If yes, define the look back window, review the results, and escalate clinically where needed. Write the reasoning down even when the answer is no.
Closing properly
A closed investigation has: the cause (or the honest statement that none was found), the corrective action, the patient impact decision, and the check that the next distribution scored acceptably. That last step is the one services forget and assessors love to ask about.
We take a realistic worked case end to end, with the full investigation paperwork completed step by step, in IQC and EQA Mastery for POCT.
Questions, answered
Do I need to investigate every EQA flag?
Yes, proportionately. A single borderline flag may close after the clerical and IQC checks; a persistent or extreme failure needs the full sequence. Every flag needs something written down.
When does an EQA failure require looking back at patient results?
When the investigation suggests the error was present during routine testing, for example drifting IQC or a systematic bias. Define the window, review, and record the clinical decision.
What if we find no cause?
Say so honestly, increase surveillance on that analyte, and confirm acceptable performance at the next distribution. An honest null finding with follow up is defensible; a blank space is not.