IVDR Performance Evaluation: The PMPF Plan Isn't a Copy of Your MDR PMCF Plan

Manufacturers running both a device portfolio and an in vitro diagnostics line reach for the same instinct twice: reuse the post-market plan. Under MDR, the PMCF plan tracks clinical performance and safety in use. Under IVDR, the equivalent document — the PMPF plan — has to support a different, three-part evidence standard, and a plan built on the MDR template leaves two of those three pillars uncovered.

Performance evaluation is not clinical evaluation

IVDR replaces the clinical-evaluation vocabulary MDR uses with performance evaluation, governed by Article 56 and Annex XIII, and the difference is not just terminology. A performance evaluation has to establish scientific validity — the association between the analyte and the clinical or physiological condition it is intended to detect — alongside analytical performance and clinical performance. An MDR device's clinical evaluation has no real equivalent to scientific validity; it does not need to independently establish that a biomarker means what the literature says it means. An IVD does, because the device's value depends entirely on that association holding.

Where an MDR PMCF template falls short

A PMCF plan built for a device is, structurally, a clinical-data collection plan: registries, real-world outcome studies, user surveys, complaint trending. Every one of those can still belong in an IVDR PMPF plan — clinical performance is one of the three pillars, and clinical-outcomes data still matters there. What a device PMCF template has no section for is scientific validity, because a device's clinical evaluation was never required to demonstrate that a biological signal means what the manufacturer claims. Porting the MDR template into an IVDR PEP without adding that section is not a formatting shortcut; it is silently dropping a third of the evidence standard.

• Scientific validity. Published literature on the analyte-condition association, clinical guidelines that reference the analyte, and, where the evidence base is thin, targeted literature reviews built into the PMPF program rather than left as a one-time literature search.
• Analytical performance. External quality assessment (EQA) scheme results, proficiency testing data, and real-world analytical performance monitoring — sources a device PMCF plan has no reason to include.
• Clinical performance. Real-world outcome data, clinician and laboratory user feedback, and post-market clinical performance studies where the initial evidence base leaves a gap — the pillar that does resemble MDR PMCF.
• Cross-pillar signals. Complaints, vigilance data, and scientific validity report updates that flag when any one pillar's conclusions no longer hold.

A device's clinical evaluation asks whether the device performs safely. An IVD's performance evaluation has to ask a question underneath that one first: does the signal it measures mean what we say it means? Why scientific validity has no MDR equivalent

Writing the PMPF plan into the PEP, not around it

PMPF is not a standalone post-market document that happens to sit near the performance evaluation plan — it is specified inside Annex XIII Part A as the mechanism for closing whatever gaps the performance evaluation identifies, and Part B is where the manufacturer proactively collects and evaluates that data throughout the device's lifetime to confirm safety, performance, and scientific validity and to keep the benefit-risk determination current. A plan that lists PMPF activities without tracing each one back to a named gap in the performance evaluation — the same discipline MDR requires of PMCF objectives — will read as generic to a notified body, IVDR or not. The difference under IVDR is which gaps: expect at least one PMPF workstream addressing scientific validity, not only clinical performance, as part of the wider EU MDR & IVDR compliance program it sits inside.

None of this requires abandoning what a PMCF program already does well — the clinical-performance pillar genuinely does resemble it. It requires not stopping there. Manufacturers who run both device and IVD portfolios save real effort by sharing PMPF infrastructure across product lines; they lose that advantage the moment the shared template quietly drops scientific validity because the device side never needed it. If your PEP hasn't named its scientific-validity gaps as explicitly as its clinical-performance ones, that is the fastest place to find where the plan is still wearing an MDR PMCF badge.