PMCF
PMCF literature reviews: cadence, scope, and a repeatable workflow
Direct answer: A PMCF literature review is a recurring, delta-focused update — not a rebuild of the original CER search. Freeze the search strategy in your PMCF plan, re-run it on a fixed cadence (at least annually for class III and implantable devices, aligned with PSUR cycles for the rest), and report what changed since the last run: new evidence on your device, equivalent devices, the clinical condition, and emerging risks. The hard part is keeping the re-run identical and the diff honest, which is exactly what software does better than a yearly manual scramble.
Where PMCF literature reviews sit in the MDR
Post-market clinical follow-up is defined in Annex XIV Part B of Regulation (EU) 2017/745 as a continuous process that updates the clinical evaluation. The PMCF plan must specify its methods, and the regulation explicitly lists screening of scientific literature and other sources of clinical data among the general methods available to manufacturers. The results are documented in a PMCF evaluation report, which feeds back into the clinical evaluation report and the technical documentation.
Two MDCG documents give the operational templates: MDCG 2020-7 (PMCF plan template) and MDCG 2020-8 (PMCF evaluation report template). MDCG 2020-7 expects each PMCF activity — including literature screening — to be described with its aim, rationale, procedure, and timeline. A vague line such as "literature will be monitored" does not meet that bar; the plan should state which sources, which queries, which screening criteria, and which cadence.
Cadence: what the regulation actually fixes
The MDR does not prescribe a single literature-review frequency, but it fixes anchor points you can derive a cadence from:
| Device class | Regulatory anchor | Practical literature cadence |
|---|---|---|
| Class III and implantable | PMCF evaluation report updated at least annually (Article 61(11)); PSUR at least annually (Article 86) | Annual full re-run minimum; many manufacturers screen quarterly or every 6 months for high-risk or novel devices |
| Class IIb | PSUR at least annually (Article 86) | Annual re-run aligned with the PSUR data lock point |
| Class IIa | PSUR updated when necessary and at least every two years (Article 86) | Re-run at least every two years; annually if the state of the art moves quickly |
| Class I | Post-market surveillance report updated when necessary (Article 85) | Risk-based; document the chosen interval and its justification in the PMS/PMCF plan |
Whatever interval you choose, the decisive point for an auditor is that the cadence is declared in the plan and then actually met, with dated evidence of each run. A brilliant review executed eighteen months after a promised twelve-month interval is a finding.
Scope: four strands, one strategy
A PMCF literature review is broader than "papers that mention our device." A defensible scope covers four strands:
- The device itself — clinical studies, case reports, complaints and field-safety notices appearing in literature.
- Equivalent and similar devices — safety signals on the device family are early warnings for yours.
- The clinical condition and state of the art — new therapies, revised clinical guidelines, or shifted benchmark outcomes can change your benefit-risk conclusion even if nothing was published about your device.
- Emerging risks and off-label signals — new contraindications, use errors, and misuse patterns.
These strands should reuse the search architecture of the original CER review — the one you built to MEDDEV 2.7/1 rev 4 standards (see our practical MEDDEV 2.7/1 rev 4 workflow) — with date limits set to "since the last run."
The weak point of manual processes: the re-run is never identical
The scientific value of a surveillance search depends on consistency. If the 2025 review used different queries, databases, or screening criteria than the 2026 review, the year-over-year comparison is meaningless: you cannot tell whether "no new safety signals" reflects reality or a narrower net. Manual processes drift for mundane reasons — staff turnover, a database subscription lapsing, a reviewer "improving" the query without versioning it.
The fix is to treat the search strategy like configuration: frozen, versioned, re-executable. Any change to the strategy is a documented protocol amendment, not a silent edit. This is the operational meaning of "systematic" in the MDR context, which we unpack in MDR Article 61: what "systematic" actually requires.
The delta report: what reviewers actually want to read
The PMCF evaluation report (MDCG 2020-8) asks for results and conclusions, including their impact on the clinical evaluation and risk management. For the literature activity, the most useful structure is a delta report:
- Records identified since the last run, per source, with the run dates of both searches.
- Screening outcome: included, excluded with reasons — presented in a PRISMA-style flow so the numbers reconcile.
- New evidence summarised per strand (device / equivalents / state of the art / emerging risks).
- The verdict that matters: does anything here change the benefit-risk determination, the claims, the IFU, or the risk file? "No change, and here is the evidence we looked" is a perfectly good conclusion — when the looking is documented.
Automating the re-run with AutoSearch
This workload profile — same strategy, fixed cadence, diff the results — is what AutoSearch is built for. A run executes the saved strategy across 12 source families (PubMed, ClinicalTrials.gov, Crossref, OpenAlex, and others) with date windows, logs every screening decision, verifies every citation's DOI against Crossref before it can enter the output, and generates the PRISMA flow diagram from the actual counts. Re-running next year means executing the same stored configuration with a new date window — the consistency problem disappears by construction, and the audit trail of each run is preserved. The methodology page describes the full disclosure model.
Because PMCF is recurring, the economics matter too: a subscription or credit plan sized to your review cadence is a fraction of the cost of commissioning each annual review externally — see pricing.
FAQ
Is a PMCF literature review the same as the CER literature review?
No. The CER review establishes the baseline evidence; the PMCF review is a periodic update against that baseline. It reuses the strategy but reports the delta, and its output flows into the PMCF evaluation report and the next CER revision (Article 61(11)).
Can literature screening be my only PMCF activity?
Usually not on its own. Annex XIV Part B lists literature screening among the general methods; depending on the device and the gaps in clinical evidence, notified bodies often expect specific activities (registries, surveys, PMCF studies) alongside it. The PMCF plan must justify the chosen mix.
What if a re-run finds nothing new?
Document it. A run with zero new included records is a valid PMCF result, provided the search log demonstrates the net was as wide as the plan promised.
Disclaimer. This article is general information, not regulatory advice. Responsibility for PMCF planning, execution, and conformity with Regulation (EU) 2017/745 remains with the manufacturer and its person responsible for regulatory compliance (PRRC, MDR Article 15). Verify requirements against the current regulation, MDCG guidance, and your notified body's expectations.
Set the cadence once, keep it forever
Explore the EU MDR/IVDR use case for the full CER and PMCF workflow, or launch a clinical evaluation literature run now and save the strategy as your re-usable PMCF baseline.