Handling OOS results in stability testing

OOS and OOT are not interchangeable

FDA defines OOS as a result outside an established specification or acceptance criterion. OOT is a separate quality-system concept: an unusual result or trend may still be within specification. Firms should define justified OOT criteria and escalate to the OOS process whenever the result is also outside specification.

Start with the laboratory assessment

Retain the original sample preparations for investigation, notify supervision, and review calculations, raw data, the approved method, system suitability, audit trails, standards, reagents, sample preparation, instrument status, sample identity, and chamber history. Phase I is an evidence review, not a search for a convenient reason to discard data.

Invalidate only with clear evidence

An initial OOS may be invalidated only when the investigation establishes a specific laboratory root cause with clear evidence. A passing repeat result, an undocumented assumption, or a hypothesis is not enough. Retain the original data, correct the cause, assess broader impact, and document CAPA.

Phase II follows the evidence

If Phase I does not establish laboratory error, the quality unit should expand the investigation to production, sampling, materials, packaging, stability trends, related batches, distributed product, and scientifically justified additional testing. An investigation may be inconclusive; a root cause must not be invented.

Control retesting and resampling

Retesting and resampling must answer a defined scientific question under a written plan with a stopping rule. A passing retest does not erase the initial result. Do not average an initial OOS with later results to hide variability; averaging is appropriate only when the approved method defines replicate measurements that form one reportable result.

Apply stability conditions correctly

ICH Q1A(R2) lists general-case long-term, intermediate, and accelerated conditions, but these are not universal defaults. Refrigerated, frozen, impermeable-container, and semi-permeable-container products require different considerations. Commercial monitoring follows the approved protocol and registered conditions.

Assess Field Alert Reports immediately

For distributed NDA or ANDA drug product, 21 CFR 314.81(b)(1) requires specified information within three working days of receipt. The clock does not wait for Phase II. Unless the distributed-batch OOS is invalidated within that period, FDA guidance recommends an initial FAR and a later follow-up.

Primary references

• FDA, Investigating Out-of-Specification Test Results for Pharmaceutical Production, May 2022
• 21 CFR 211.192 and 21 CFR 314.81(b)(1)
• ICH Q1A(R2), Stability Testing
• ICH Q2(R2), Validation of Analytical Procedures