How would you investigate a GMP deviation?

Why

A good deviation investigation begins before anyone writes a root cause. First ensure that affected product, equipment or material is controlled according to procedure. Preserve original readings, timestamps, instrument audit trails, environmental conditions and batch history.

Determine whether other lots or processes might share the same problem. Build a causal timeline, consider alternatives and test them against actual evidence. Assign appropriate product-quality and patient-risk implications, including whether escalation or reporting is required.

A scientifically justified conclusion distinguishes confirmed causes from unresolved hypotheses. Then choose containment, corrective or preventive measures proportional to the risk, complete necessary approvals and test effectiveness over a meaningful observation period. A passing retest does not automatically explain an original out-of-specification result.

Nor is retraining sufficient if the procedure or equipment was designed in a way that invites recurrence.