Benefit-Risk Analysis That Actually Convinces Reviewers

Written by HATEM RABEH, MD, MSc Ing

Your Clinical Evaluation Expert And Partner

in
S

I have reviewed hundreds of benefit-risk sections. Most follow the same pattern: list the benefits, list the risks, state that benefits outweigh risks. Done. Except it is not done. Because listing is not analyzing, and stating is not demonstrating.

The benefit-risk determination is where your entire clinical evaluation comes together. Every piece of evidence you gathered, every study you appraised, every claim you made—they all converge here. If this section fails, everything before it becomes suspect.

What Benefit-Risk Analysis Actually Requires

MDR Article 61 requires that clinical evidence demonstrates conformity with relevant general safety and performance requirements. The benefit-risk determination is not a formality. It is the regulatory answer to a fundamental question: is this device acceptable for use in the intended population?

MDCG 2020-6 reinforces that the benefit-risk analysis must be proportionate to the device’s risk class and the nature of the clinical claims. This is not about filling a section. It is about building an argument.

Quantifying Benefits: Beyond Marketing Claims

Benefits must be clinical, measurable, and supported by evidence. Marketing claims do not count. If you claim improved accuracy, show the data. If you claim faster recovery, cite the studies. Every benefit statement should trace back to clinical evidence in your evaluation.

The challenge is that many manufacturers confuse device features with clinical benefits. A feature is what the device does. A benefit is what that means for the patient. Faster processing time is a feature. Reduced procedure duration leading to lower anesthesia exposure is a benefit.

Quantification matters. Saying your device is effective is meaningless. Stating that clinical studies demonstrate 94% sensitivity with 95% confidence interval of 91-97% gives reviewers something to evaluate.

Characterizing Risks: The Complete Picture

Risks must come from your risk management file. The clinical evaluation does not replace ISO 14971—it feeds into it and draws from it. Your benefit-risk section should reference the residual risks identified after all mitigations are applied.

But here is where many manufacturers fail: they only consider device-specific risks. A complete benefit-risk analysis includes procedural risks, risks from the clinical context, and risks compared to alternative treatments. If your device replaces surgery, the surgical risks it eliminates are part of the benefit calculation.

Probability and severity matter. A rare but catastrophic risk is different from a common but minor one. Your analysis should reflect this. Using risk matrices from your risk management file helps, but the clinical evaluation must interpret what those risks mean for patients.

The Comparison That Matters

Benefit-risk is always comparative, even when you do not explicitly compare. Reviewers ask: compared to what? The answer shapes your entire analysis.

For equivalent devices, you compare to the equivalent device’s established profile. For novel devices, you compare to the state of the art—current treatment alternatives, including no treatment. For devices with no alternatives, you must establish absolute acceptability thresholds.

MDCG 2020-6 emphasizes that the acceptability of risks depends on the severity of the condition being treated and the availability of alternatives. A higher risk profile may be acceptable for a life-threatening condition with no other options. The same profile would be unacceptable for a cosmetic indication.

Building the Argument

The structure of your benefit-risk section should follow a logical progression:

First, summarize the clinical benefits with quantification and evidence references. Second, summarize the residual risks with probability and severity assessments. Third, compare to alternatives and establish the decision context. Fourth, provide the determination with explicit reasoning.

The determination is not just a statement. It is a conclusion that follows from the preceding analysis. A reviewer should be able to trace your logic. If they cannot, you have not demonstrated—you have merely asserted.

Some manufacturers use scoring systems or quantitative frameworks. These can help if applied consistently, but they are not required. What matters is that your reasoning is explicit and defensible.

What Reviewers Look For

Notified Body reviewers approach benefit-risk sections with specific questions: Are the benefits clinical and quantified? Are the risks complete and characterized? Is the comparison appropriate? Does the conclusion follow from the analysis?

They also look for consistency. Your benefit-risk section must align with your intended purpose, your clinical claims, your risk management file, and your PMCF plan. If your risk management identifies significant uncertainties, your benefit-risk should acknowledge them. If your PMCF is designed to address specific questions, your benefit-risk should explain why those questions remain.

Inconsistency is a red flag. If your benefit-risk says residual risks are well-characterized, but your PMCF plan includes studies to better understand those risks, reviewers will notice.

Moving Forward

Building a convincing benefit-risk analysis requires more than data assembly. It requires argumentation. You are not just presenting information—you are making a case for why your device should be used in the intended population.

Start with the end in mind. What would convince a skeptical expert that your device is acceptable? Build your analysis to answer that question. If you cannot articulate the argument, neither can your document.

In the next post, we will look at how to structure the overall CER document—how to organize your evidence and argumentation so that reviewers can follow your reasoning from beginning to end.

Benefit-risk analysis is one chapter in the broader CER structure. For the complete methodology, see our comprehensive CER writing guide under MDR.