Handle an out-of-specification result as an OOS investigation, not as a general deviation.
What we are working through:
  • A laboratory-phase investigation that rules out analytical error before any manufacturing-phase investigation begins
  • Defined retest and re-sampling steps, with the authorisation for each recorded
  • An explicit, signed decision on whether the original result is invalidated, and on what grounds
  • Links to the batch or lot affected and to any deviation or CAPA that follows
Why this matters: an OOS has a specific expected shape, and a reviewer who finds one recorded as an ordinary deviation will notice that the lab-phase investigation and the invalidation decision are missing. This matters most for operators doing release testing against an importing country's requirements, where an OOS holds a shipment.