Introducing a diagnostic test is not simply a matter of installing an analyser and obtaining a plausible result. Before reporting patient samples, a laboratory needs evidence that the examination procedure performs adequately for its intended clinical use, under local conditions, with the people, equipment and specimen types involved.
Two terms are central to that evidence: verification and validation. They are sometimes used interchangeably, but they answer different questions. Getting the distinction right helps laboratories make better decisions, avoid unnecessary work and recognise situations where a larger study is needed.
At a glance: Method verification confirms that a laboratory can achieve established performance claims under its own conditions. Method validation establishes that a procedure is fit for its intended clinical use, particularly when it is laboratory-developed or substantially modified. Both require predetermined acceptance criteria, documented evidence and appropriate authorisation before patient reporting.
Verification asks whether established performance can be achieved locally
Suppose a laboratory introduces a commercially available quantitative assay, using the manufacturer's specified instrument, reagents and specimen type. The manufacturer has established performance characteristics under defined conditions. The laboratory still needs to demonstrate that relevant claims can be met in its own setting.
That is the central idea of verification. A local study might consider precision, bias, analytical measuring interval and other characteristics selected according to the test and its clinical risk. Acceptance criteria should be set before the study, not after seeing the results.
The CLSI EP15 guideline provides one recognised protocol for user verification of precision and estimation of bias. It does not, by itself, cover every aspect of a test's fitness for clinical use. A satisfactory precision result cannot compensate for a sample-handling problem, inappropriate reference interval or unsafe reporting rule.
Validation asks whether a procedure is fit for its intended use
Validation is usually the more extensive question. It is particularly relevant when a laboratory develops its own examination procedure, substantially modifies a commercial method or proposes to use a test outside its stated intended purpose.
Consider a molecular assay intended for one specimen type but proposed for another. Even if the instrument runs successfully, the change may alter extraction efficiency, interference, sensitivity and interpretation. The previous evidence may no longer be sufficient. A risk-based validation plan should establish what must be demonstrated for the new use.
The 2026 Validation Guidance for Medical Laboratories, developed by UK professional and regulatory bodies, addresses manufacturer claims, modified or off-label use, laboratory-developed tests and research-use-only components. It emphasises proportionate evidence and clear responsibility rather than a single template applied to every method.
Start with the clinical purpose, not the study spreadsheet
A useful implementation plan begins with five questions. What clinical decision will the result support? Who will receive it? Which specimen types and patient groups are included? What can go wrong if a result is wrong or delayed? And which performance characteristics matter most for that risk?
For example, an assay used near a clinical decision limit may need particular attention to bias and imprecision around that limit. A low-concentration infectious disease assay may need careful investigation of detection limits, contamination risks and false negatives. A semi-quantitative or qualitative examination requires different acceptance criteria from a routine numerical chemistry test.
The plan should identify sample numbers, concentrations, comparison procedures, data analysis, rejection rules and responsibilities. It should also record what happens if acceptance criteria are not met. Repeating an experiment until it passes is not a substitute for investigating unexpected performance.
Evidence must include the surrounding workflow
A technically sound assay can still produce unsafe results if specimens are incorrectly identified, calibration is poorly managed or reporting logic is wrong. Implementation therefore extends beyond the analytical bench.
A complete introduction considers sample collection and transport, reagent acceptance, staff training, internal quality control, external quality assessment where available, analyser and middleware interfaces, reference intervals, interpretive comments, critical-result escalation and continuity arrangements. The laboratory should document who reviews the evidence and who authorises routine use.
ISO 15189:2022 provides the overall requirements for medical laboratory quality and competence. It is not a substitute for method-specific technical guidance, and laboratories must interpret its requirements in the context of their actual scope and applicable national rules.
Make the evidence useful after go-live
Verification or validation should not become a folder that is never opened again. The baseline evidence helps the laboratory recognise later shifts in performance. Changes to reagents, equipment, software, specimen pathways or intended use may trigger a documented impact assessment and additional verification or validation.
The best outcome is not the thickest report. It is a defensible decision: the laboratory understands the procedure's limitations, has demonstrated performance appropriate to clinical need and can identify when that conclusion must be reconsidered.
Where Optymum SS fits: Optymum SS can support assay evaluation planning, technical evidence reviews, verification and validation documentation, and implementation readiness. Responsibility for clinical approval, accreditation and patient testing remains with the relevant laboratory and authorised professionals.
Related on this site: Research, Development & Scientific Advisory, Testing & Diagnostics.
Key takeaways
- Choose verification or validation based on intended use and existing evidence.
- Set acceptance criteria before testing and document deviations.
- Include specimen pathways, interfaces, staff and result reporting in implementation.
Frequently asked questions
What is the difference between verification and validation in a medical laboratory?
Verification checks established method claims under local laboratory conditions. Validation establishes fitness for intended use when suitable performance evidence is not already available or the use differs materially from the validated claim.
Does ISO 15189 require every new test to undergo full validation?
No. The evidence required depends on the method, intended use, available manufacturer evidence, modifications and laboratory risk assessment. Local verification may be appropriate for an unchanged established method.
Who authorises a diagnostic method for routine use?
The medical laboratory must follow its documented governance arrangements and applicable regulatory requirements. External advisers may support the evaluation but do not replace the laboratory’s clinical and quality responsibilities.
