Post-market clinical follow-up under EU MDR: from obligation to strategic asset
PMCF has outgrown its reputation as a formality. How manufacturers are turning real-world safety and performance data into intended-purpose extensions, competitive positioning, and dossiers regulators want to read.
Insight · · 3 min
Under the Medical Device Regulation, Regulation (EU) 2017/745, applicable since 26 May 2021, a device's evidence obligation does not end at CE marking. It begins there in a second form. Post-market clinical follow-up is a continuous process, not an occasional study, and the clinical evaluation is expected to be updated throughout the device's lifetime as exposure accumulates.
The machinery is concrete rather than aspirational, and worth naming precisely. Annex XIV Part B requires a PMCF plan and a PMCF evaluation report, proactively collecting clinical data from the device in normal use to confirm safety and performance and to identify previously unknown side effects. Annex XIV Part A requires the clinical evaluation report to be updated with that data. Article 86 requires a periodic safety update report, annually for Class IIb and III devices, with the PSUR for Class III and implantable devices submitted through EUDAMED and assessed by the notified body. MDCG 2020-7 provides the PMCF plan template. Class III and implantable devices also face the summary of safety and clinical performance under Article 32. Transitional deadlines were extended by Regulation (EU) 2023/607, which relieved timing pressure and changed none of the substance.
These are serious obligations, and regulators treat them as such. What is worth questioning is not the requirement but the operating model most manufacturers use to meet it: a periodic effort against a deadline, a literature review, a small PMCF study commissioned to close a gap the notified body identified, and a CER updated by someone who was not involved in collecting any of it. The data is assembled once, used once, and left behind. When the next cycle arrives, much of the work is repeated, because nothing was built to persist.
The alternative is architectural rather than procedural. A device instrumented for continuous, structured real-world capture, device telemetry where the product supports it, connected peripherals, ePRO from users, registry participation and, where governance permits, EHR-derived outcomes, produces a live evidence stream rather than periodic snapshots. The same stream services PMCF, feeds the CER update, populates the PSUR, and is available when a notified body asks an unexpected question mid-cycle.
Once the stream exists, it earns in three directions manufacturers routinely fund separately.
Intended-purpose extensions. Real-world performance data documents how a device behaves in populations, settings and indications adjacent to the ones on the current certificate. Where that data is structured, traceable and prospectively planned, it becomes the substance of an argument to extend the intended purpose, rather than a reason to commission a new study from scratch. This is the clearest case of an obligation paying for something the commercial roadmap wanted anyway.
Reimbursement and health technology assessment. Payers and HTA bodies increasingly ask for performance in routine use rather than in the pivotal study population, and Regulation (EU) 2021/2282 on joint clinical assessments extends to certain high-risk devices from 2028. A manufacturer already holding structured real-world outcome data is answering, with evidence they were required to collect, a question that would otherwise trigger a separate programme.
Competitive positioning and product development. Failure modes, usage patterns, adherence and workflow friction appear in continuous data long before they appear in complaints. That is design input while the next iteration is still specifiable, and it is also the material for a comparative story that a purchaser can verify. A successor device does not start from zero when the predecessor's field performance is documented and curated.
The precondition is that the data be structured and traceable from the outset, which is a decision made at product and system design, not at the point a report is due. Retrofitting provenance onto data collected without it is expensive and frequently impossible. Manufacturers who plan for continuous capture pay once, at design time, for something otherwise paid for repeatedly that never accumulates.
For MedTech, unlike pharma, the evidence obligation genuinely runs both pre-market and post-market, and the two are meant to be continuous with one another. The regulation assumes a manufacturer who knows how their device performs in the field. The realistic question is not whether continuous evidence arrives, it is already required, but whether it is captured as an asset or reassembled each cycle as a cost.
Sources. Regulation (EU) 2017/745 (MDR), applicable 26 May 2021, Annex XIV Parts A and B, Articles 32 and 86. Regulation (EU) 2023/607 (transitional provisions). MDCG 2020-7, PMCF plan template. Regulation (EU) 2021/2282 on health technology assessment.
Written by the Evidilya scientific team. For interviews, references or a full publication list, use the contact page.
Talk to our team.
Bring us the evidence question you're wrestling with, we'll tell you the shortest defensible path.