From deviation to CAPA: actions that remove the cause and effectiveness checks that prove it

The deviation system exists to produce CAPA that works. Everything before the CAPA, the investigation, the root cause, the impact assessment, is preparation. If the CAPA does not remove the cause, the whole record is a well-documented failure to learn.

Correction, corrective action, preventive action

• Correction: fixes the immediate problem. Rejecting the batch, relabelling the bin, recalibrating the balance. Necessary, but it prevents nothing.
• Corrective action: removes the root cause so this event does not recur. Redesigning the label, adding an independent line clearance check, changing the calibration interval.
• Preventive action: removes the same cause elsewhere before it produces an event. Reviewing all label pairs for differentiation, not just the two involved.

Why retraining is usually a correction, not a corrective action

Retraining an individual addresses that individual's knowledge on that day. It does nothing about the procedure that was unclear, the equipment that made the error easy or the check that failed to catch it. Retraining is appropriate when the root cause is genuinely a knowledge gap, and the record shows how that was established. It is not appropriate as a reflex.

Designing an effectiveness check

An effectiveness check verifies that the CAPA worked, not that it was completed. 'Training completed on 14 March' is a completion record. 'No recurrence of label mix-ups on line 3 in the six months following implementation, verified against the deviation log on 15 September' is an effectiveness check.

1. Define what success looks like before implementation: a measurable outcome (recurrence rate, audit result, metric) over a defined period.
2. Set the check date far enough out for the event to have had the opportunity to recur. For a monthly process, six months; for a daily one, perhaps six weeks.
3. Name the evidence that will be reviewed and who will review it.
4. If the check fails, reopen the investigation. A failed effectiveness check means the root cause was wrong or incomplete.

Trending closes the loop

Individual effectiveness checks tell you about individual CAPAs. Trending by failure mode, area, product and cause category tells you whether the system is learning. The management review under ICH Q10 should see deviation trends, not just counts, and the trend should show the effect of CAPA over time.

Module 6 of the course ends with a CAPA and effectiveness check workshop using the three investigations from Module 4.