The level of clinical evidence describes how convincingly the totality of clinical data demonstrates the safety, performance, and an acceptable benefit-risk ratio of a medical device. It arises not solely from the number of studies, but from the quality, quantity, relevance, completeness, and statistical robustness of the available evidence.
Assessment of Evidence in the Product Context
The required evidence depends on the specific product and its intended purpose. Relevant factors include novelty, risk profile, invasiveness, duration of contact, clinical benefit, and whether robust data already exist. A large amount of data may be insufficient if it concerns a different patient group or does not cover the critical risks.
Clinical evaluation brings together literature, post-market experience, clinical investigations, and, if applicable, equivalent data. Crucial is the traceable connection between the collected data and the manufacturer’s claims. For a product with novel technology, the transferability of older data may be more limited than for a well-established product with a stable design.
Distinction from Evidence Level
The level of clinical evidence is not to be equated with the evidence level of evidence-based medicine. An evidence level categorizes study types in a hierarchy, such as randomized studies and observational data. It thus answers only part of the question of how strong the evidence for a particular medical device actually is.
For clinical evidence, it is also important whether the study covers the specific product, its indication, and the relevant endpoints. Consistency between sources, data recency, and unresolved safety issues also influence the overall level. A methodologically high-ranking publication can therefore only make a limited contribution for a different technology.
Gaps and Conclusions
An evaluation plan should first translate clinical claims into verifiable questions. This allows identifying which data are available, which data only indirectly support the claims, and what uncertainty remains. This gap analysis determines whether a clinical investigation is required or whether targeted post-market data can supplement the evidence.
The conclusion must not be stronger than the data basis. For example, in the case of rare but severe risks, a limited study size cannot rule out remaining uncertainty. The level of clinical evidence therefore requires an argumentation that combines positive results, methodological limitations, and the proportionality of further data collection.
The evidence is not a fixed threshold achieved by a certain number of subjects or publications. Its adequacy is measured against the risks and claimed performance. In the event of a product change, the question may therefore boil down to whether previous data are still transferable or whether the change creates a new clinical question.
Data from clinical post-market surveillance are also not merely collected but assessed for their suitability as evidence. Registry data, complaints, and feedback from use can provide important safety information, but may not capture a controlled comparative question. The clinical evaluation must justify each source’s contribution to the respective statement.
The evaluation is also dynamic. New findings from use, changed clinical guidelines, or a safety signal can alter the previous conclusion. The level of clinical evidence is therefore not determined once in the life cycle, but is updated through clinical evaluation and post-market surveillance. Especially for products whose benefit depends on a specific user group or treatment chain, this development must be assessed in the context of actual use.
Another aspect is the precision of the claims. The more specific a manufacturer’s claimed benefit, the more targeted the data must be to reflect that benefit. A general safety description, for example, makes no statement about faster diagnosis or a better therapeutic decision. The claimed outcome, not solely the risk class, helps determine the clinical evidence required for the evaluation.
Relevance for clinical trials
In a clinical investigation, the desired level of evidence determines the choice of population, endpoints, comparison, and follow-up duration. A protocol that only collects technical functional data cannot substantiate a claimed patient-relevant effect. The clinical evaluation should establish before the start of the study which specific evidence gap the investigation is intended to close, so that the evaluation and subsequent benefit-risk justification are consistent.
Full-service CROs like Mediconomics support manufacturers with evidence gap analysis, selection of clinically meaningful endpoints, planning of statistically sound evaluations, and the transfer of study results into the clinical evaluation report.
Frequently Asked Questions (FAQ)
Can a large number of publications compensate for a low level of evidence?
Only if the publications are truly relevant and methodologically robust for the product, the indication, and the open safety or performance questions.
Is a randomized study always sufficient?
No. A randomized study must also reflect the correct technology, appropriate endpoints, and the affected patient group.
When can an evidence gap necessitate a clinical investigation?
When existing data do not convincingly answer a significant safety, performance, or benefit question for the specific product.
Regulatory References
- Regulation (EU) 2017/745 (MDR) — requires a clinical evaluation based on sufficient clinical evidence.
- MDCG 2020-6, Guidance on sufficient clinical evidence — explains the assessment of sufficient evidence.
- MDCG 2020-5, Clinical Evaluation — Equivalence — addresses the transferability of comparative data.
- MDCG 2020-1, Clinical Evaluation of MDSW — demonstrates the application of evaluation principles to software.
