How would you interpret a fall in deviations when reporting volume also drops?

Why

Deviation counts measure two things at once: how often things go wrong and how willing people are to say so. If both events and reporting fall, the count cannot tell which one drove the change. Start with the denominator, since fewer batches made means fewer deviations.

Then look at the mix. Real improvement usually lowers serious deviations while minor reports and near misses stay steady or rise, because a healthy culture keeps reporting small things. A drop concentrated in minor reports suggests people stopped reporting.

Compare with signals people cannot suppress: QA review findings, audit observations, equipment alarms, complaints, OOS results and yield losses. If those are flat while deviations fall, be suspicious. Check what changed around the drop, such as a new target that rewards low deviation counts, a reclassification of events as "incidents" outside the system, or a new reporting tool that is harder to use.

The MHRA points out that performance indicators and management behaviour shape what gets recorded, and that a good system actively encourages reporting of errors. ICH Q10 asks management review to assess indicators like deviation and CAPA processes, which is where a misleading metric should be challenged. The analysis cannot prove events happened that nobody recorded, so it should end with a clear statement of what is known and a targeted check, such as a focused audit.