An out-of-specification (OOS) result is a test result that falls outside the specifications or acceptance criteria set in a registered specification, official compendium or company standard. It applies to raw materials, in-process samples, finished products and stability samples alike.
The FDA guidance on investigating OOS results describes a staged approach. A phase I laboratory investigation looks for an assignable laboratory cause — a calculation error, an instrument fault, a sample preparation mistake. If none is found, a phase II full-scale investigation extends into manufacturing and may include a defined, justified retest or resample plan.
Why it matters in GMP
Testing into compliance — repeating a test until a passing result appears and discarding the failure — is one of the most serious data integrity findings an inspector can make. 21 CFR 211.192 requires every unexplained discrepancy to be investigated, and the conclusion must be scientifically justified and documented, whether the batch is ultimately released or rejected.
- Keep the original result: it is never overwritten or deleted.
- Separate the analyst who produced the result from the people who review and approve the investigation.
- Link the investigation to any CAPA, and to the batch release decision.
How Swat iQua Suite handles it
Swat iQua Suite runs OOS investigations in up to three levels, each with its own review and approval. The server works out the current level, so users cannot skip one, and the approver must differ from the creator and that level’s reviewer. E-signatures on review and approval can be required per site, a Finalize step locks the record before controlled printing, and when CAPA is required one CAPA is raised and approved with the final level. Lab incidents, OOT and other lab quality events follow a similar levelled workflow.