• September 24, 2026

The Silent Validation Leaks Costing Your Lab Time, Accreditation and Trust

Most laboratories are good scientists and poor historians. The bench work is careful and the results are sound, but the record of how those results were reached often is not.

The result is a familiar pattern. A new test performs perfectly in the laboratory, and then an assessor asks a single question the validation file cannot answer. None of this shows up while the test is running well. It shows up at the worst possible moment, in front of an assessor or a client audit.

We see the same seven gaps again and again when we train pathology and medical laboratories in test verification and validation. Here they are, in case one sounds familiar.

1. Verify versus validate confusion

Verification asks whether a test is being performed correctly. Validation asks whether the correct test is being performed at all. Those are different questions with different evidence requirements, and the line between them is easy to misjudge under time pressure. Get it wrong one way and a laboratory burns weeks of bench time on a full validation that only needed verification. Get it wrong the other way and a genuinely novel test is waved through with a quick check, and the gap only surfaces at assessment.

2. Skipping the planning phase

Planning is where a validation earns its credibility. Skip it, and the team goes straight to bench work: running samples, comparing results, hoping it holds together. Acceptance criteria get defined after the data is already in, when it is far too easy to set them to match whatever result was already produced.

3. Incomplete performance characteristics

Not every characteristic, accuracy, precision, sensitivity, specificity, limit of detection, matrix effect, applies to every test, but the selection has to be a deliberate, defensible decision. Left to instinct, teams default to the characteristics they find easiest to measure rather than the ones the test actually needs, and the gap only shows up when someone asks why a particular characteristic was left out.

4. Measurement uncertainty as an afterthought

Measurement uncertainty is a required, ongoing part of validation under ISO 15189, not a one off certificate to produce on request. Done properly it needs the right approach for the data available, the right coverage factor for the sample size, and a clear record of why it matters for results near a clinical decision limit. Done the week before an assessment, with a coverage factor pulled from memory, it rarely holds up to scrutiny.

5. The documentation gap

This is the one we see most often. The validation itself was sound. The report is a folder of raw data with no summary, no clear structure, and no way for an assessor, or a colleague six months later, to follow what was tested and why it passed. Many laboratories validate a test well and document it poorly, and to an assessor those look identical.

6. Rushed implementation

Deciding a test is ready involves real judgement about performance, turnaround time and cost, but that judgement still has to be checked against a concrete implementation checklist. When a test goes live before the SOP is finalised, staff are trained and signed off as competent, or internal and external quality control are actually running, problems move from the validation file into live patient results.

7. Risk thinking bolted on

A risk assessment that gets attached to the validation file at the end, as a document to satisfy the audit, tends to restate the obvious and miss the risks specific to that test. Risk based thinking is meant to run through validation planning from the start, shaping which samples get tested and what the acceptance criteria actually need to cover, not sit beside the file as a separate form.

Closing the gaps

None of these seven gaps are about scientific competence. They are process gaps, and very few laboratory professionals are taught this specific discipline before they are asked to lead it.

Our one day course, Verification and Validation for Pathology, works through all seven, from the difference between verification and validation through to planning, performance characteristics, measurement uncertainty, report writing, implementation and risk based thinking. For laboratories that want to go deeper on measurement uncertainty specifically, Introduction to Measurement Uncertainty is a good companion course.

View upcoming dates and enquire at neas.com.au, or contact our team directly.