What is post-market surveillance?
Post-market surveillance is the proactive, systematic collection and analysis of data on your device once it is on the market. Articles 83 to 86 of the MDR require every manufacturer to run a PMS system as part of its quality management system, proportionate to the device's risk class.
The aim is simple: spot problems early, confirm that benefits still outweigh risks, and keep your technical file true to how the device behaves in real use.
Why does it matter?
Pre-market evidence only goes so far. Trials are limited in size and duration, and real users, real settings and rare events show up later. PMS is how you catch what testing could not.
It also matters for compliance. A missing or weak PMS system is an audit finding, and the Notified Body expects PMS output to feed your clinical evaluation, risk file and labelling. For manufacturers, though, complaint patterns, service data and user feedback are some of the most useful product intelligence you will ever get.
The building blocks
The MDR sets out a connected set of documents and activities. It helps to see how they relate.
| Element | What it is | Applies to |
|---|---|---|
| PMS system | The overall process for gathering and using post-market data | All classes |
| PMS plan (Annex III) | Your written plan for what data you collect, how and how often | All classes |
| PMS report (Article 85) | A summary of PMS results and conclusions | Class I |
| PSUR (Article 86) | Periodic safety update report with benefit-risk conclusions | Class IIa, IIb and III |
| PMCF (Annex XIV, Part B) | Proactive clinical follow-up to confirm safety and performance | As justified by risk and evidence |
| Vigilance (Articles 87 to 92) | Reporting serious incidents and field safety corrective actions | All classes |
Writing a PMS plan that works
Annex III says what the plan must contain. In practice, a plan is useful when it answers a few direct questions:
- What data will you collect? Complaints, service and repair records, returns, literature, registries, user feedback and information on similar devices are typical sources.
- How will you collect it? Name the processes, owners and tools.
- How will you analyse it? Define how you detect trends and what counts as a signal.
- What triggers action? Set thresholds in advance, so a decision does not depend on who happens to notice.
- How does it feedback? Show the link to the risk file, clinical evaluation, labelling and design changes.
A plan that lists data sources but not decisions is only half a plan.
PMCF: closing the clinical evidence loop
Post-market clinical follow-up is the clinical part of PMS. Your PMCF plan states what clinical questions remain open after certification and how you will answer them, for example through registry data, follow-up studies or surveys of users. If you decide PMCF is not needed, you must justify it.
The results go into a PMCF evaluation report, which updates your clinical evaluation. MDCG 2020-7 and 2020-8 provide templates for the plan and the report.
Reporting cycles: PSURs and PMS reports
How often you report depends on class.
- Class I: a PMS report, updated when necessary.
- Class IIa: a PSUR, updated when necessary and at least every two years.
- Class IIb and III: a PSUR, updated at least annually.
For Class III and implantable devices, the PSUR goes to the Notified Body through EUDAMED for evaluation. MDCG 2022-21 gives guidance on content and structure. A good PSUR draws a clear conclusion about benefit-risk rather than just listing numbers.
Vigilance: when speed matters
PMS is steady and planned. Vigilance is the part where the clock runs. You must report serious incidents to the competent authorities, and the deadlines are short.
- Serious incidents: within 15 days of becoming aware
- Death or unanticipated serious deterioration in health: within 10 days
- Serious public health threat: within 2 days
You must also report field safety corrective actions, and statistically significant increases in non-serious incidents under the trend reporting rules. The clock starts when you become aware, not when your investigation ends, so complaint handling needs to recognise a reportable event quickly.
Turning data into decisions
Data are only useful if someone acts on them. Strong PMS systems tend to share a few habits:
- Complaints are logged, classified and trended, not just closed
- Reportability is assessed within days, with the reasoning recorded
- Findings route automatically into the risk file and CAPA process
- Clinical and regulatory staff review the data together, regularly
- Management sees the results, so resources follow the findings
Common mistakes
- Writing a PMS plan and never using it
- Collecting complaints but not analysing trends
- Waiting for investigation to finish before judging reportability
- Leaving PMCF unjustified when no activity is planned
- Missing PSUR deadlines after certification
- Not feeding findings back into the risk file and clinical evaluation
Is your PMS system ready for audit?
We help manufacturers write PMS and PMCF plans, prepare PSURs, and set up trending and vigilance processes that work in daily use. Talk to us about where your system stands today.