Literature Review (Clinical) is a systematic, documented search and appraisal of published scientific literature used to gather clinical data on a medical device, its equivalent, and the state of the art. It forms a core input to the clinical evaluation report (CER) under EU MDR 2017/745, supporting the device’s safety and performance claims.
What is a Literature Review (Clinical)?
A clinical literature review is the structured process of finding, selecting, and critically appraising published evidence relevant to a device’s intended purpose. Manufacturers use it to demonstrate two things: the safety and performance of the subject device (or a justified equivalent), and the current state of the art for the condition the device treats or diagnoses. It sits within the wider clinical evaluation activity, alongside any clinical investigation data and post-market data.
The review is not a one-time literature dump. It follows a written protocol, runs reproducible database searches, applies pre-set inclusion and exclusion criteria, and records every appraisal decision so an auditor can repeat the work and reach the same conclusion.
Why Literature Review (Clinical) matters in medical device development
For most CE-marked devices, published literature is the main source of clinical evidence. A weak or non-reproducible review is one of the most common reasons a clinical evaluation report gets flagged during notified body review under EU MDR. The consequences are concrete: certification delays, requests for more data, or refusal of the technical file.
Beyond approval, the review feeds the benefit-risk determination and the risk management file. If the literature surfaces a known complication or a better-performing alternative, that finding has to flow into the risk analysis (ISO 14971) and, where relevant, the post-market surveillance plan. Under-scoping the search exposes a manufacturer to audit findings and, more seriously, to safety gaps that only appear once the device is on the market.
How the clinical literature review process works
The methodology most teams follow comes from MEDDEV 2.7/1 Rev 4, which remains the practical reference even though it predates the regulation. EU MDR Annex XIV Part A sets the legal requirement; the MEDDEV explains how to meet it. Several MDCG documents (for example, MDCG 2020-5 on equivalence and MDCG 2020-6 on sufficient clinical evidence for legacy devices) point back to its methods.
A defensible review usually runs through these steps:
Scope and plan
Define the device, its intended purpose, the clinical claims, and the review questions in the clinical evaluation plan (CEP).
Write the search protocol.
Specify databases, search strings, date ranges, and inclusion/exclusion criteria before running anything.
Run two search streams.
One covers the subject or equivalent device; the other covers the state of the art and benchmark devices for the same indication.
Screen and select
Apply the criteria, log every excluded record with a reason, and keep an audit trail (a PRISMA-style flow diagram is common).
Appraise the data
Score each included paper for methodological quality and scientific validity, following Section 9.3 of MEDDEV 2.7/1 Rev 4.
Analyze and document
Weight the evidence, address gaps, and feed the synthesis into the CER.
Clinical investigation data appraised alongside the literature is expected to meet good clinical practice under ISO 14155. The current edition is ISO 14155:2026, though EN ISO 14155:2020 remains the version harmonized to the EU MDR at the time of writing, so confirm which applies to your submission.
Common challenges and best practices
The most frequent failure is a search that cannot be reproduced: vague strings, no recorded dates, or undocumented exclusions. Notified bodies treat a non-repeatable search as no search. Write the protocol first and freeze it before searching.
A second problem is equivalence. Under EU MDR, the bar for claiming a device is equivalent to one already supported by literature, which is high, covering technical, biological, and clinical characteristics. Teams often lean on equivalence data they cannot fully access, then lose it during review. Decide early whether equivalence is realistic or whether you need your own clinical data.
Other practical points:
- Keep the review live rather than treating it as a pre-market task, since the CER must be updated through post-market surveillance and PMCF.
- Separate the device-specific search from the state-of-the-art search so each question is answered cleanly.
- Tie every finding back to the risk file and the general safety and performance requirements (GSPR) in MDR Annex I.
How SJML helps with Literature Review (Clinical)
SJML’s QARA team provides clinical evaluation support as part of its Compliance-as-a-Service offering. That includes building the clinical evaluation plan and report (CEP and CER), running structured literature searches and appraisal aligned with MEDDEV 2.7/1 Rev 4 methodology, and managing the clinical data behind a submission. The team also connects the review to post-market surveillance planning, vigilance, and risk management files (ISO 14971), so clinical evidence stays consistent across the technical documentation. Support scales from startups preparing a first CE submission to established OEMs maintaining large device portfolios.
Frequently asked questions
The literature review is one input; the clinical evaluation report (CER) is the output. The review gathers and appraises published evidence, while the CER synthesizes that evidence together with clinical investigation and post-market data into a documented benefit-risk conclusion for the device. A CER cannot be complete without a defensible literature review behind it.
EU MDR does not mandate specific databases, but a defensible search under MEDDEV 2.7/1 Rev 4 typically uses MEDLINE/PubMed plus Embase for European coverage, and often the Cochrane Library. PubMed alone is usually treated as a supplement rather than a standalone source, because full-text access and European journal coverage can be limited.
The clinical evaluation, including its literature review, must be updated across the device lifecycle as new data arrives from post-market surveillance and PMCF. For well-established, lower-risk devices, intervals of two to five years are common; higher-risk and implantable devices need more frequent updates. The risk profile and post-market signals set the cadence, not a fixed calendar.
Sometimes. For many lower-risk, well-established devices, appraised literature plus equivalence can provide sufficient clinical evidence. For Class III and implantable devices, EU MDR generally expects device-specific clinical investigation data, and the Article 61(10) literature-only route cannot be applied to them. The device class and intended purpose decide whether literature alone is acceptable.
Related terms
- Clinical Evaluation Report (CER)
- Post-Market Clinical Follow-up (PMCF)
- State of the Art (SOTA)
- Clinical Investigation
- Risk Management (ISO 14971)
