MDCG 2020-7: PMCF plan template, explained

Who this applies to

You are affected if you manufacture a medical device under the MDR and need to address post-market clinical follow-up (PMCF). PMCF is the clinical part of your post-market surveillance system: a continuous process for updating the clinical evaluation with data from the real-world use of your CE-marked device.

The plan sits inside the PMS plan, and the results later go into a PMCF evaluation report (MDCG 2020-8). If you conclude that no PMCF activity is needed, that decision still needs a device-specific justification which your notified body can assess.

What PMCF is supposed to achieve

The template starts with five aims straight from MDR Annex XIV Part B:

1. Confirm safety, performance, and clinical benefit throughout the device's expected lifetime.
2. Find previously unknown side-effects and monitor known side-effects and contraindications.
3. Identify and analyse emerging risks using factual evidence.
4. Confirm that the benefit-risk ratio remains acceptable.
5. Detect systematic misuse or off-label use and check whether the intended purpose is still correct.

That is the useful test for every activity in your plan: which of these questions will it actually answer?

What the template contains

Sections A and B identify the manufacturer and device. The real work begins in Section C:

For every activity in Section C, state where the need came from (for example the CER, risk management, PMS, a previous PMCF report, or a notified-body request), whether it is a general or specific method, and its aim. Then describe the method, sample size, endpoints, comparator where relevant, sources of bias, known limitations, and a justified schedule.

General and specific PMCF methods

PMCF is not automatically a new clinical investigation. The right mix depends on the device, its risks, and the gaps left by your clinical evaluation:

Calling an annual customer survey "PMCF" does not make it useful. The template explicitly expects statistical reasoning where appropriate: why the sample size, timescale, endpoints, comparator, and study design can produce evidence strong enough for the residual risks. It even singles out retrospective surveys without a statistical rationale as unacceptable.

Data from equivalent or similar devices can help update the state of the art or identify safety outcomes. It should not quietly replace data from your own device when the objective is to confirm its continued safety and performance.

What this means for you, practically

1. Start with the gaps, not the template. Pull unresolved questions from the CER and risk management file first; then choose activities which can answer them.
2. Give every activity a question and an output. "Review literature annually" is vague. Define the search, the safety or performance outcome, the review date, and what happens to the result.
3. Justify the evidence strength. Record likely bias, missing data, incomplete follow-up, sample-size reasoning, and why the result will still be usable.
4. Use your own device data. Similar-device data is useful for state of the art and new hazards, but it is not a permanent substitute for learning from your product in real use.
5. Connect the document chain. Use stable activity numbers in the PMCF plan, carry them into the evaluation report, and state a quarterly or at least yearly reporting date. That makes it obvious whether each promised activity actually happened.