Out-of-Specification Results: How to Run a Laboratory Investigation That Holds Up

An out-of-specification result is not a retest waiting to happen. It is a potential quality signal, and how you handle the first hours decides whether the investigation will be credible.

What the guide says

Chapter 6 of the EU GMP Guide requires a procedure for the investigation of out-of-specification (OOS) and out-of-trend results, and states that out-of-specification results or significant atypical trends should be investigated. Where a confirmed OOS result or significant negative trend affects batches already on the market, the relevant competent authorities must be informed. The EU Guide is short on step-by-step method, so many companies also align their procedures with the FDA guidance on investigating OOS results, which is widely used as a reference for the staged approach.

Phase 1: the laboratory assessment

The first step is to establish, before any retesting, whether an obvious laboratory error occurred. This should be done promptly by the analyst and supervisor and documented. Points to check include:

  • calculations, dilutions and transcription of results;
  • instrument qualification, calibration and system suitability;
  • reagents, standards, solutions and their expiry dates;
  • whether the correct method version and sample were used;
  • audit trails and raw data for the run.

If an assignable laboratory cause is demonstrated, the result may be invalidated on documented grounds and the test repeated. If not, the original result stands and the investigation moves beyond the laboratory.

Phase 2: the wider investigation

This extends to manufacturing, sampling and materials. It should consider other batches and products that could share the cause, and it should involve Quality Assurance from the start. Any retesting or resampling must follow a pre-defined plan, with the number of tests and the way results will be interpreted set out in advance.

Common pitfalls

  • Testing into compliance: repeating tests until a passing result appears, then averaging or discarding the failing one.
  • Hypothetical causes: invalidating a result on the basis of a suspected but unproven error.
  • Delayed escalation: leaving the investigation inside the laboratory when the evidence points elsewhere.
  • Data integrity gaps: unrecorded trial injections or missing audit trail review.

Linking to disposition and reporting

The outcome feeds directly into batch disposition by the Qualified Person, the deviation and CAPA systems, and, where a marketed product is affected, notification obligations. A clear decision tree in the procedure avoids ad hoc judgement under time pressure.

A short self-check

  • Does your procedure state what must happen before any retest is permitted?
  • Is audit trail review part of every laboratory assessment?
  • Is the Qualified Person informed at the point an OOS is confirmed, not at batch release?

How PQRA helps

PQRA helps quality control laboratories and quality assurance teams write or revise OOS and OOT procedures, build decision trees and templates, and review recent investigations for the weaknesses described above. We can also support training of analysts and supervisors on the first-hour decisions that matter most.

Contact PQRA to discuss how we can support your quality system.

CATEGORIES:

Uncategorized

Tags:

No responses yet

Αφήστε μια απάντηση

Η ηλ. διεύθυνση σας δεν δημοσιεύεται. Τα υποχρεωτικά πεδία σημειώνονται με *