Article 88 asks for a threshold, not a PSUR

A vigilance file built around two categories — individual serious incidents reported under Article 87, and the periodic aggregate report covered under Article 86 — is missing a third duty MDR imposes on exactly the events those two categories don't reach. This blog has already drawn the line between a PSUR and a PMS report; Article 88, trend reporting, sits on a third axis again, and it's the one most trackers build last, if they build it at all, because unlike the other two it doesn't come with a number already attached.

What the other two articles already cover

Article 87 covers the individual, serious event: a single incident meeting MDR's serious-incident definition gets reported on its own timeline, tiered by severity — no later than 15 days after the manufacturer becomes aware of most serious incidents, no later than 10 days for one involving death or an unanticipated serious deterioration in health, and no later than 2 days for a serious public health threat. Article 86 covers the aggregate, periodic report — the PSUR, updated at least annually for Class IIb and III devices and at least every two years for Class IIa, rolling up the device's safety and performance picture across the reporting period. Between them, those two articles answer “did something serious just happen” and “how has the device performed over the period.” Neither one is built to answer a third question: is a pattern of minor events quietly becoming a safety signal.

What Article 88 actually asks for

Article 88 requires manufacturers to report, through the electronic system in Article 92, any statistically significant increase in the frequency or severity of incidents that are not serious incidents, or of expected undesirable side-effects, where that increase could have a significant impact on the benefit-risk analysis described in Annex I and could lead to unacceptable risk when weighed against the device's intended benefits. The events in scope are, by definition, the ones that individually wouldn't trigger an Article 87 report at all — a trend report exists precisely because a run of individually unremarkable events can add up to something an incident-by-incident view never surfaces.

The threshold is the manufacturer's own to set, in writing, in advance

Article 88 doesn't hand manufacturers a fixed percentage or count to trigger against. The significant increase has to be established against the foreseeable frequency or severity of that kind of incident for the specific device, or device category, over a specific observation period — and the article requires the manufacturer to specify, in the post-market surveillance plan required under Article 84, exactly how it will manage these incidents and exactly what methodology it will use to determine a statistically significant increase, including the observation period itself. That's a real, checkable requirement sitting inside a document most PMS plans treat as boilerplate: a plan that doesn't name a baseline frequency, an observation window, and a statistical method for this specific device hasn't actually met Article 88's own precondition for filing a trend report at all, whether or not a trend ever needs reporting.

Competent authorities can still push back on the manufacturer's own math

Setting the threshold isn't the last word. Article 88(2) allows competent authorities to run their own assessment of a submitted trend report and require the manufacturer to adopt appropriate corrective measures regardless of what the manufacturer's own methodology concluded. A trend-reporting methodology built to make triggering as unlikely as possible isn't a defensible reading of Article 88 — it's a document a competent authority can challenge on its own terms, against the device's actual incident history.

Where this meets the file

A vigilance tracker built around this split needs Article 87's individual-incident log, Article 86's PSUR cycle, and a third record most files skip entirely: the trend-reporting methodology itself, sitting inside the Article 84 PMS plan, with its baseline frequency, observation period, and statistical method stated before any event ever needs to be run against it. The same discipline applies downstream — a field safety notice reports the corrective action, not the incident that prompted it, a distinction that only holds up if the incident and trend records feeding it are kept separately in the first place. A PMS-plan worksheet built around Article 84's own required elements, trend-reporting methodology included, is previewed in the launch catalog. If your program defines its trend threshold differently, the shelf takes that correction directly.