Medtech surveillance: an annual report is only as sound as its signals

A safety report may be up to date yet still be weak if complaints, incidents, distributor returns and clinical data never lead to a consistent decision. On 23 April 2026, Swissmedic informed Swiss manufacturers and authorised representatives of a new targeted action covering post-market surveillance documentation for class IIa, IIb and III devices. The authority will sample devices and may request plans, analyses and reports. Receipt of the general letter does not mean that an organisation or device has been selected, and it does not imply a deficiency. This insight follows field evidence through to the periodic safety update report. It separates collection, trend analysis, clinical evaluation, vigilance, corrective action and approval . A practical first engagement is a mock review of one device or family, producing a gap register and closure plan. Outreach must remain measured: the offer tests the evidence chain without claiming to predict Swissmedic's sample or using patient safety to create fear. General analysis updated on 7 August 2026. It does not replace a Swissmedic decision, the manufacturer's assessment, or device-specific regulatory or clinical advice.

The 2026 targeted action examines post-market surveillance documentation for sampled class IIa, IIb and III devices. Swissmedic requests the records from the relevant Swiss manufacturer or Swiss authorised representative. The general letter requires no immediate response; an organisation whose device is selected receives a separate request.

Swissmedic restates the duties to operate a surveillance system, plan the relevant activities and document the analyses periodically in safety reports. The authority says that a 2024 action found that many manufacturers had difficulty achieving full compliance. That finding supports prudent preparation, but it gives neither a probability of inspection nor a verdict on any individual business.

A PSUR can be formally complete yet remain inconsistent when its sources use different populations, periods or categories. Commercial complaints, vigilance records, returns and literature must be reconciled. An apparent increase may reflect better collection, while apparent stability may conceal a missing channel. The reasoning and its limits must remain visible.

Sales data is a valid denominator only when its coverage matches the signal being assessed. A change in coding can create a false break in a trend. The manufacturer must also know who approves the conclusion and how an unresolved signal is tracked through to closure.

The first practical engagement is a mock review of one device or family: PMS plan, sources, responsibilities, trend analysis, PSUR, clinical evaluation and links to vigilance. The output ranks gaps and assigns actions. It does not certify the whole organisation or guarantee that Swissmedic will accept a particular file.

The partner starts with a scope that matches its expertise and secure access to data. It records versions, reporting periods and owners. The parties agree the fee, timing and sample size before work begins. If a medical or statistical judgement falls outside the partner's remit, the question is escalated to the specialist appointed by the manufacturer.

PMS gathers and analyses post-market experience; vigilance handles events and actions governed by its rules; clinical evaluation incorporates relevant safety and performance evidence. Their outputs inform one another without becoming copies. Each item has one source, while every process explains how it affects the decision within that process.

A dependency matrix prevents the same incident from being counted differently in three reports. It records the source system, extraction date, population and version. Where conclusions diverge, the difference is resolved explicitly rather than hidden behind generic wording.

The Swiss authorised representative must also know which records it holds and which remain with the foreign manufacturer. A formal responsibility is fragile if transmission times are uncontrolled. The review should therefore test a real scenario: a document request, version check, any required translation and final approval. This does not transfer the manufacturer's obligations. It verifies that the chain can produce the agreed evidence under realistic conditions.

A review should reopen when a signal, trend, serious incident, corrective action, design change, new indication or clinical finding changes the assessment. The periodic timetable remains necessary, but it must not delay an urgent decision. Every trigger needs a threshold, an owner and a response time.

How to read the diagram. The final report retains the decision trail; it never replaces the immediate response to an urgent signal.

The audience comprises Swiss manufacturers and Swiss authorised representatives for class IIa, IIb and III devices, together with their specialist partners. The 2026 letter went to all relevant registered organisations, but only some devices will be sampled. A campaign should target an observable documentation responsibility, never a supposed selection or weakness.

Useful commercial signals include a new device, a change of authorised representative, international expansion, recruitment to a PMS team or an approaching PSUR cycle. Technical content addresses active research questions. Professional associations and networks can provide context and trust. Direct outreach offers a mock review without referring to a patient, incident or confidential file.

Past the midpoint of this insight, the Medtech market page connects surveillance with the MDR framework, the Swiss Medical Devices Ordinance and in vitro diagnostics. The purchase decision is different: market-access work supports placing a product on the market and representation, while this review tests evidence generated after real-world use.

Value is measured through reviews ordered, gaps closed on time, margin and renewed cycles. Correcting a document does not automatically mean that an incident was prevented. Regulatory and clinical outcomes remain confidential. A public example may show the process, but it must never extrapolate an inspection probability or a reduction in harm.

The working metrics should therefore stay close to the service delivered: scope accepted, records reviewed, actions assigned, evidence of closure and any subsequent paid cycle. The manufacturer owns the underlying safety judgements. The partner reports its work without turning confidential risk information into a marketing claim.

A review can recur with each PSUR cycle, new signal, product change or transfer of responsibility. The partner may test a sample and follow actions through to closure. Every engagement needs a defined population and conclusion. General monitoring without data, thresholds or decisions is not enough to justify recurring revenue.

Swissmedic establishes the action dated 23 April 2026, the device classes concerned and the sampling approach. The Medical Devices Ordinance sets PMS and PSUR obligations, while official guidance covers surveillance and vigilance. These sources do not identify selected devices, approve a particular report or set a price for consultancy work.

Official references and verification dates remain in the private source file. This public article names the authorities without linking externally. Before every campaign, the partner confirms the applicable scope and requirements again.

The complimentary eligibility assessment reviews your expertise, independence, proposed initial review, acquisition challenge and secure capacity. It does not validate a PSUR. It establishes whether getfishnet and your team could build a tailored strategy around an identifiable market, an official action and a recurring need.