PMCF Plan Under EU MDR Annex XIV Part B: Structure, Study Triggers, and the PSUR Link

A Post-Market Clinical Follow-up (PMCF) plan is the document a device manufacturer uses to specify how it will proactively and systematically collect and evaluate clinical data on a CE-marked device after it reaches the market. Under EU MDR (Regulation (EU) 2017/745), PMCF is not optional paperwork attached to a device file — it is required by Annex XIV, Part B, and its findings are meant to keep closing the loop between what a manufacturer claimed pre-market and what actually happens once the device is in routine use.

This guide walks through what Annex XIV Part B actually requires in a PMCF plan, when a literature-review-only approach is defensible versus when a dedicated PMCF study is unavoidable, and how PMCF output is meant to flow into the Periodic Safety Update Report (PSUR) and the wider post-market surveillance (PMS) system.

What a PMCF Plan Is, and Where It Sits in Your Technical Documentation

Annex XIV Part B frames PMCF as a continuous process, not a one-off study. The plan itself must specify the methods and procedures the manufacturer will use to proactively collect and evaluate clinical data from the use of the device in or on humans, within its intended purpose. The stated aims, per Annex XIV Part B, are to:

• Confirm the safety and performance of the device, including clinical benefit where applicable, throughout its expected lifetime;
• Identify previously unknown side effects and monitor known side effects and contraindications;
• Identify and analyse emergent risks on the basis of factual evidence;
• Ensure the continued acceptability of the benefit-risk ratio (Annex I, Sections 1 and 9); and
• Identify possible systematic misuse or off-label use, to verify the intended purpose remains correctly stated.

Structurally, the PMCF plan is not a standalone artefact. It is required to be part of the manufacturer’s post-market surveillance (PMS) plan, and the results of PMCF activity are written up in a PMCF evaluation report that becomes part of both the clinical evaluation report (CER) and the technical documentation. A notified body’s assessment of the clinical evaluation explicitly covers the adequacy of the PMCF plan and its execution — including, where a manufacturer has decided PMCF activity isn’t warranted, the justification for that decision. See CASRAI’s guides on the ISO 13485 quality management system that houses this documentation and the ISO 14971 risk management file that the PMCF plan is required to reference directly.

The Seven Sections MDCG 2020-7 Expects in a PMCF Plan

The Medical Device Coordination Group’s guidance document MDCG 2020-7, “Post-market clinical follow-up (PMCF) Plan Template” (April 2020), gives manufacturers and notified bodies a harmonised structure. It isn’t legally binding in itself, but it operationalises what Annex XIV Part B asks for, and notified bodies routinely check submissions against it:

• Section A — Manufacturer contact details: legal manufacturer, SRN, the person responsible for regulatory compliance, and the authorised representative if applicable.
• Section B — Device description and specification: intended purpose, intended users and patient population, indications and contraindications, Basic UDI-DI, classification and classification rule, and expected lifetime.
• Section C — Activities related to PMCF (general and specific methods and procedures): the substantive part of the plan — see the next section below.
• Section D — Reference to the relevant parts of the technical documentation: which open items from the CER and the risk management file this PMCF plan is meant to further analyse and monitor (or an explicit statement that there are none).
• Section E — Evaluation of clinical data relating to equivalent or similar devices: where PMCF conclusions lean on data from a comparator device rather than the device itself.
• Section F — Applicable common specifications, harmonised standards, or guidance documents.
• Section G — Estimated date of the PMCF evaluation report, on a schedule defined at least yearly (more frequently for higher-risk devices).

General and Specific PMCF Methods — What Actually Counts as Evidence

Section C is where a PMCF plan is won or lost during notified body review. MDCG 2020-7 lists the recognised categories of PMCF activity:

• Screening of scientific literature and other sources of clinical data — a systematic, protocol-driven review, not an ad hoc search;
• PMCF studies — new clinical investigations within the intended use, extended follow-up of patients from pre-market investigations, or retrospective studies;
• Registries — a manufacturer’s own device registry, or use of a suitable national public registry, with a pre-specified quantity and quality of data to collect;
• Real-world evidence (RWE) analyses — drawn from real-world data (RWD) that is of sufficient quality and comes from reliable sources; and
• Surveys — of healthcare professionals, patients, or users; and review of case reports for signs of misuse or off-label use.

For every activity, MDCG 2020-7 expects the manufacturer to document: where the need for the activity originated (notified body request, CER, PMS, risk management report, or a prior PMCF report); whether it is a general or specific method; which of the aims above it serves; a rationale for why the chosen method, sample size, timescale and endpoints are appropriate, including a statistical justification tied to the residual risk; and a defined timeline, reviewed at least quarterly or yearly. The guidance is explicit that a retrospective survey justified only as “this should demonstrate the expected quality of evidence we require,” with no statistical rationale behind it, is not acceptable.

When a Literature-Review-Only PMCF Is Defensible

Screening scientific literature is a legitimate, named PMCF method in its own right — it is not automatically a lesser substitute for a study. A literature-review-driven PMCF plan tends to hold up under notified body scrutiny when several conditions line up together: the state of the art for the device’s intended use is well established and stable rather than actively shifting; the device relies on an equivalence or similarity claim to a device (or devices) with a substantial, still-relevant clinical data trail, documented per Section E of the plan; the residual risks identified in the risk management file and CER are already well characterised and not expected to change materially with more real-world exposure; and the device is not implantable and not Class III with a pre-market data gap (see the mandatory-study trigger below). In that combination, a systematic literature search protocol — with defined search terms, databases, inclusion/exclusion criteria and an appraisal methodology, the same rigour Annex XIV Part A requires of a pre-market clinical evaluation literature search — can credibly confirm continued safety and performance without a dedicated study.

What a literature-only approach cannot do is manufacture confidence that doesn’t exist in the published record. If the equivalent-device literature is thin, dated, or drawn from a device whose design has since diverged, or if a CER identified a genuine evidence gap rather than a merely theoretical one, “we will monitor the literature” is not, on its own, an adequate PMCF plan — and notified bodies are specifically tasked with assessing that adequacy, not just checking that a Section C table exists.

When a Dedicated PMCF Study Is Required

MDR Annex XIV Part B sets one bright-line trigger directly: for implantable devices and Class III devices where clinical investigations were not performed pursuant to Article 61(4), the PMCF plan must include post-market clinical studies to confirm the safety and performance of the device. Article 61(4) is the provision under which some Class III and implantable devices can, in narrow circumstances, avoid a pre-market clinical investigation (for example, on a well-documented equivalence basis); Annex XIV Part B closes that gap by moving the clinical-data obligation into the post-market period instead of removing it.

Beyond that explicit trigger, a PMCF study becomes the practical necessity — even without a literal legal mandate — whenever: the device is genuinely novel, so there is no representative equivalent-device literature to lean on; a notified body or the manufacturer’s own CER has flagged a specific residual-risk or evidence-gap question that existing literature cannot answer; an equivalence claim underpinning the pre-market file is borderline or contested, so PMCF needs to independently establish clinical performance rather than assume it; or post-market signals (complaints, vigilance data, prior PMCF cycles) suggest something the literature record doesn’t cover. Clinical investigations run as part of PMCF are still expected to meet the same good clinical practice standard as pre-market ones — see ISO 14155, the GCP standard for clinical investigation of medical devices in human subjects.

Justifying PMCF as “Not Applicable”

A manufacturer can conclude that PMCF activity, beyond routine literature monitoring, is not warranted for a given device — but MDR treats this as a documented, reviewable decision, not a default. The notified body’s clinical evaluation assessment explicitly covers “the justification in relation to non-performance of PMCF,” meaning the absence of a PMCF study has to be argued on the same evidentiary footing as its presence: stable state of the art, a mature and directly comparable equivalent-device data trail, low residual risk, and no Article 61(4) trigger. A thin or boilerplate non-applicability justification is one of the more common points notified bodies push back on during technical documentation review.

How PMCF Feeds the PSUR and the Rest of the PMS System

PMCF doesn’t terminate in its own report. Two downstream links matter:

Into the CER and technical documentation. The PMCF evaluation report is explicitly part of the clinical evaluation report, not a separate appendix filed alongside it. New PMCF findings — a previously unknown side effect, a shift in the benefit-risk balance, evidence that off-label use is occurring — are what typically trigger a CER update outside the routine review cycle.

Into the PSUR (or PMS report). The PMCF plan is part of the manufacturer’s broader PMS plan, and PMCF findings are one of the substantive inputs into the periodic reporting MDR requires under Chapter VII: a PMS report for Class I devices, updated when necessary and made available to the competent authority on request, and a Periodic Safety Update Report (PSUR) for Class IIa, IIb and III devices, summarising the main findings of the PMCF, the conclusions of the benefit-risk determination, and the volume of sales together with an estimate of the population using the device and, where practicable, its frequency of use. Class III and implantable devices update the PSUR at least annually; Class IIa and IIb devices update it when necessary and at least every two years. For Class IIb and III devices, the PSUR is submitted to the notified body and made available via EUDAMED; for Class IIa, it is made available to the notified body on request.

The same PMCF-to-vigilance link applies laterally: where PMCF or PMS activity surfaces a safety issue serious enough to need a field correction, that action is reported and tracked separately as a field safety corrective action (FSCA), not folded silently into the next PSUR cycle.

Notified Body Review: What Gets Checked

A notified body’s review of PMCF is not limited to confirming a document exists. Per Annex XIV Part B and the MDCG 2020-7 template, reviewers are expected to check that: each Section C activity has a stated origin, aim, method, and a rationale that includes a real statistical or comparator justification rather than an assertion; Section D accurately reflects open items from the current CER and risk management file version, not a stale one; Section E’s equivalent/similar device claims are consistent with how the same devices are treated elsewhere in the technical documentation, per the equivalence criteria set out for a device’s clinical evaluation more broadly; and, where PMCF is deemed non-applicable, the justification is substantive. Device software carries its own parallel lifecycle obligations under IEC 62304 that a PMCF plan for a software-containing device should stay consistent with, particularly where post-market monitoring surfaces a defect that is a software issue rather than a clinical one.

Frequently Asked Questions

Is a PMCF plan required for every CE-marked device?

Yes — Annex XIV Part B applies across risk classes, though the intensity of the plan scales with risk. What varies by device is whether PMCF activity beyond literature screening, and specifically a dedicated PMCF study, is required or can be justified as unnecessary.

What’s the difference between the PMCF plan and the PMCF evaluation report?

The plan specifies what will be done (methods, rationale, timelines); the evaluation report documents what was found once the planned activities have run, and that report is what becomes part of the CER and technical documentation.

How often does a PMCF plan need to be revisited?

MDCG 2020-7 expects PMCF activity timelines to be defined at least yearly, more frequently (quarterly) where risk warrants it, and the plan itself is revised as new PMCF, PMS, or CER findings emerge.

Can a manufacturer rely on a UDI-based registry instead of a bespoke PMCF study?

A registry can be a legitimate PMCF method under Section C, provided it is pre-specified for quality and quantity of data and tied to the device (or a genuinely comparable device) under evaluation; it does not substitute for a mandatory PMCF study where the Article 61(4) implantable/Class III trigger applies. Registry data quality depends heavily on consistent UDI capture at the point of use.

Does PMCF apply the same way under IVDR?

IVDR has an analogous but separately numbered post-market performance follow-up (PMPF) requirement rather than PMCF by name; the structural logic — a plan, an evaluation report feeding the performance evaluation report, and periodic safety reporting — parallels the MDR PMCF/PSUR relationship described here, but the article numbers and terminology differ and should be checked against IVDR text directly rather than assumed to be identical.