PSUR quality is becoming the weak point in post-market files

Periodic safety update reports are supposed to be an analytical document, not a data dump. Reviewers across several member states have noted a recurring problem: manufacturers list complaint counts, sales volumes, and literature findings, but stop short of stating whether the benefit-risk profile has changed and what action follows. In the weakest examples the report repeats the same conclusion for several reporting periods without explaining the underlying data.

The MDR requires the report to include an assessment of the benefit-risk determination, the volume of sales, and an estimate of the population exposed. Where the report is thin, it is usually because complaint coding is inconsistent, which makes trend analysis unreliable and pushes authors towards narrative description instead of quantitative reasoning. Correcting the coding scheme is unglamorous work, but it is what makes the analysis possible at all.

Class IIa devices with a lower risk profile are less frequently reviewed, which can hide slow deterioration. Several notified bodies now ask for a short trend commentary even when the formal interval has not elapsed, particularly where a component supplier change has occurred or where a software update altered the user interaction. Manufacturers that pre-empt the request tend to avoid a formal question round and the associated delay.

The remediation is organisational rather than documentary. Manufacturers that link complaint categories to the risk file, and that use the same terminology across CAPA records and the safety report, spend far less time reconciling numbers. Those with separate coding schemes tend to produce reports that are internally inconsistent, and the inconsistency is exactly what a reviewer notices first when comparing two consecutive submissions for the same device.