Whether the system described on paper is the system the organization actually runs. An investigator follows a small number of records end to end — a complaint, a deviation, a batch, a change — and checks that each handoff happened when the procedure said it would, by someone qualified to make the decision.
The system is judged by its records, not its procedures
A procedure states an intention. A record shows what happened. When the two disagree, the record wins — and the disagreement itself becomes the finding, because it demonstrates the quality system is not in control of its own commitments.
That is why an experienced reviewer rarely reads the manual first. They pull recent records from the systems most likely to expose weakness — investigations, CAPA, change control, complaints — and trace each one back to the procedure that governed it.
What gets examined first
- Deviations and investigations opened in the last twelve months, and how many closed on time
- CAPA records where the corrective action is retraining, and what the effectiveness check measured
- Change controls that touch validated processes, equipment or suppliers
- Complaints and, for devices, the decision record behind each reportability determination
- Data review practices in the laboratory, including audit trail review
The questions behind the questions
Most follow-up questions are variations on three: who decided, what evidence did they have, and how would you know if the decision were wrong. A quality system that can answer those three quickly, from the record itself, holds up.
A system that requires someone to reconstruct the reasoning from memory does not — regardless of how thorough the procedures look.
Reviewed by the named consultant. Last updated July 14, 2026. General information only — not regulatory or legal advice for a specific situation.