What this article helps you consider
- Recognise two different signals
- Preserve the evidence
- Test hypotheses rather than search for passing numbers
Recognise two different signals
An out-of-specification result falls outside established acceptance criteria. An out-of-trend result differs from an expected pattern even if it remains within specification. OOT criteria should be established through product and method understanding rather than used as a retrospective label for any inconvenient observation.
A hypothetical assay result of 97% may meet a specification yet warrant review if a stable process has consistently produced results near 100% and the shift exceeds expected variability. The decision depends on the measurement capability and historical context. OOT does not automatically mean batch failure, but it should not be ignored because a limit has not been crossed.
Preserve the evidence
Notify the appropriate supervisor and quality function promptly, retain relevant preparations when feasible and preserve instrument data and audit trails. Review whether the system and method performed as intended. Check calculations, standards, reagents, sample preparation, instrument condition and observations recorded during analysis.
FDA’s OOS guidance describes an initial laboratory assessment followed by a broader investigation when a clearly causative laboratory error has not been established. Contract laboratories should communicate their findings and supporting records to the responsible manufacturing quality unit. R17
Test hypotheses rather than search for passing numbers
A suspected cause should explain the magnitude and direction of the result and be supported by evidence. If incomplete extraction is proposed, design an experiment capable of examining extraction behaviour. If dilution error is suspected, review preparation records and use retained material where suitable. A possibility mentioned in a checklist is not an assigned cause.
Retesting should follow an approved, scientifically justified plan. Repeated analysis until a passing value appears introduces selection bias. A passing retest does not automatically invalidate the original result, and averaging a failure with passing results can obscure meaningful variation. Treat resampling as a separate decision requiring a rationale about sample representativeness.
Extend the investigation beyond the laboratory
When a laboratory cause is not demonstrated, examine manufacturing, materials, storage, transport and sampling. Review related batches, products and time periods according to the plausible scope. A rejected batch may still reveal a systemic issue that affects material already distributed or other work in progress.
For a hypothetical dissolution failure, evidence of correct apparatus performance directs attention toward the product but does not identify the root cause. Compare granulation, drying, compression, coating and packaging records with the release pattern. Use additional testing to answer defined questions rather than generate a large volume of unrelated data.
Make disposition transparent
The responsible quality unit should evaluate all relevant evidence, including the original result, investigation findings, additional testing and unresolved uncertainty. Clearly distinguish a documented invalid laboratory result from a confirmed product failure and from an inconclusive investigation. Each situation requires a reasoned disposition under the applicable quality system and regulatory obligations.
The illustration shows these outcomes as separate branches. It does not turn an inconclusive result into a presumptive pass. The investigation record should allow a reviewer to reconstruct why evidence supported the final decision.
Follow the evidence to a quality decision
Retain relevant preparations, records, instrument data and audit trails.
Conceptual illustration. Select a topic to explore its role. The visual does not establish formulation performance or prescribe universal regulatory requirements.
Verify corrective action effectiveness
A corrective action should address the demonstrated cause. Training alone is a weak response when the procedure is ambiguous or equipment design creates repeated errors. Consider method clarification, process controls, maintenance or supplier action as appropriate. Define an effectiveness check with a relevant observation period and measurable outcome.
Stabicon can support structured laboratory investigations and trend review while providing clients with the data needed for their quality decisions. The laboratory’s technical conclusion and the manufacturer’s batch-disposition responsibilities should be clearly defined in the quality agreement.
References & further reading
Read the source in its stated scope. Requirements, amendments and compendial editions depend on the product and target market. Draft guidance is identified separately.
- R06ICH Q10 Pharmaceutical Quality System
Final guideline, 2008. Pharmaceutical quality system and lifecycle management.
- R17FDA Investigating Out of Specification Test Results for Pharmaceutical Production
Final guidance, May 2022. Laboratory investigations and evaluation of results.
- R18ICH Q9 R1 Quality Risk Management
Final guideline adopted January 2023. Scientific and risk-based quality decisions.
From insight to a defined programme
Bring your next technical question.
Share your formulation, testing objective or development challenge. Stabicon can review the context and define a suitable scope of work.
Discuss this topic with Stabicon






