Your clinical claim sounds safe until someone asks for evidence

Written by HATEM RABEH, MD, MSc Ing

Your Clinical Evaluation Expert And Partner

in
S

The problem is not that manufacturers lack data. Most do. The problem is they do not understand what the MDR considers a clinical claim, how specific it must be, and what type of evidence is required to support it.

This matters because clinical claims define what you promise to the market. They shape your intended use, your instructions, your labeling, and your risk-benefit analysis. If your evidence does not match your claims, your clinical evaluation fails at the structural level.

Let me walk you through how this works in practice.

What the MDR Means by Clinical Claim

A clinical claim is any statement about the clinical performance, clinical benefits, or safety profile of your device. It includes what the device does, how well it does it, and what outcomes it produces.

Article 2(43) of the MDR defines clinical performance. Annex XIV Part A describes what must be demonstrated. But most manufacturers miss the operational meaning: a claim must be specific, measurable, and linked to patient-relevant outcomes.

Saying “our device is safe” is not a clinical claim you can defend. It is a conclusion without structure.

Saying “our device results in surgical site infection rates below 2% in colorectal procedures” is a claim. It has a measurable outcome, a defined population, and a threshold you can compare.

This is where most deficiencies begin. Manufacturers underestimate what counts as a claim. They treat descriptive language as neutral. But from a regulatory perspective, any statement that influences clinical decision-making is a claim that requires evidence.

The Evidence Hierarchy That Notified Bodies Apply

The MDR does not treat all evidence equally. Annex XIV Part A, Section 1 establishes a hierarchy. Clinical data from your own device is preferred. Equivalence data is acceptable only under strict conditions. Literature alone is rarely sufficient.

But hierarchy is not enough. You also need congruence.

Congruence means the evidence you cite must directly address the claim you make. If your claim is about infection rates, your evidence must measure infection rates. If your claim is about patient-reported outcomes, your evidence must include validated PRO tools.

I see CERs where manufacturers claim reduced recovery time, then cite studies measuring pain scores. Or they claim superiority over standard of care, then provide only bench testing. The evidence may be high quality. But it does not match the claim.

When this happens, Notified Bodies issue deficiencies. Not because the data is weak, but because the logic is broken.

This deficiency reflects a reasoning error. The manufacturer assumed related outcomes could substitute for the claimed outcome. But unless you demonstrate the relationship, that substitution is not valid.

How to Structure Your Claims Around Available Evidence

The safest approach is reverse engineering. Start with the evidence you have. Identify what outcomes it measures, in which populations, under what conditions. Then write claims that match.

Most manufacturers do the opposite. They write the claims they want, then search for evidence to support them. This creates tension between marketing ambition and regulatory reality.

If your evidence shows a 15% reduction in procedure time compared to predicate, your claim should state that. Not “significantly faster.” Not “improved efficiency.” The specific magnitude your data supports.

If your evidence comes from a single-center study with 50 patients, your claim should reflect that limitation. You cannot generalize to all populations unless your evidence base supports it.

This does not mean you must publish weak claims. It means your claims must be defensible with the evidence you currently hold. You can always strengthen claims later with additional data. But you cannot defend claims that exceed your evidence.

What Counts as Sufficient Evidence

MDCG 2020-13 provides guidance on sufficient clinical evidence. But sufficiency is context-dependent. It depends on your device class, your novelty, your risk profile, and your claims.

For a well-established Class IIa device with incremental design changes and modest claims, a robust literature review combined with bench data may suffice.

For a Class III device with new mechanism of action and claims of superiority, you will need prospective clinical data from your device, in your intended population, measuring your claimed outcomes.

But even within those boundaries, the question remains: how much is enough?

The answer is not a number. It is a judgment. The Notified Body assesses whether the body of evidence, taken together, reduces residual uncertainty to an acceptable level.

This means you cannot simply count studies. You must evaluate the quality of each study, the relevance to your device and population, the consistency of findings, and the magnitude of observed effects.

I have seen CERs with 200 references that failed because none addressed the core claim. And I have seen CERs with 15 references that succeeded because each one directly answered a key clinical question.

The Role of Comparators in Supporting Claims

Most claims are inherently comparative. When you say your device is safe, you mean safer than doing nothing or safer than alternatives. When you say it is effective, you mean it achieves outcomes that matter.

The MDR requires you to demonstrate performance relative to the state of the art. This is not optional. Annex XIV Part A, Section 1 and Annex I Chapter I require comparison.

But comparison introduces complexity. You must define your comparator, justify why it represents state of the art, and demonstrate that your comparison is valid.

If you claim superiority over standard of care, you must show data comparing your device to that standard under similar conditions. If you claim equivalence to a predicate, you must demonstrate equivalence according to MDCG 2020-5.

Where manufacturers struggle is in selecting appropriate comparators. They choose outdated devices, or devices used in different populations, or devices with different intended uses. Then they claim equivalence or superiority based on weak comparisons.

Notified Bodies see through this. If your comparator is not truly representative of current practice, your comparative claim is not valid.

This deficiency reflects a misunderstanding of what state of the art means. It is not the device you happened to have data for. It is the best available option that patients currently receive.

How to Document the Evidence-to-Claim Pathway

The most effective CERs include an explicit table mapping each claim to its supporting evidence. This table should list the claim, the evidence source, the outcome measured, the population studied, and the level of certainty.

This is not decoration. It is a reasoning tool. It forces you to verify that every claim has support, and that the support is appropriate.

When a Notified Body reviews your CER, they perform this mapping mentally. If you provide it explicitly, you demonstrate that you have already done the work. You reduce their uncertainty and their review time.

But the table is only useful if it is honest. Do not list evidence that loosely relates to your claim and hope no one checks. Reviewers check. And when they find mismatch, they question everything else.

Better to acknowledge gaps. If you lack direct evidence for a specific claim, say so. Explain why you consider the claim reasonable based on related evidence, bench testing, and risk analysis. Then commit to gathering direct evidence in your PMCF.

This approach shows regulatory maturity. It demonstrates you understand the evidence requirements, you are transparent about limitations, and you have a plan to address them.

What Happens When Claims Exceed Evidence

When your claims exceed your evidence, you face two options: reduce your claims or generate additional evidence.

Most manufacturers resist reducing claims. They believe it weakens their market position. But maintaining unsupported claims weakens your regulatory position, which ultimately affects market access anyway.

If you cannot defend a claim, remove it. Focus on claims you can prove. Then use PMCF to gather evidence for additional claims over time.

This is not failure. It is strategy. You enter the market with defensible claims, establish your device in clinical practice, and expand your claims as evidence accumulates.

The alternative is prolonged review, repeated deficiencies, and potential rejection. That delay costs more than cautious claims.

This principle applies equally to new devices and to legacy devices updating their technical documentation. Under MDR, you must re-justify every claim. If you cannot, you must revise or remove them.

The Connection Between Claims and Risk-Benefit Analysis

Your clinical claims feed directly into your risk-benefit determination. The benefits you claim must outweigh the residual risks you identify.

If your claims are vague or unsupported, your benefit side of the analysis is weak. Even if your risks are well-characterized and mitigated, the balance becomes questionable.

Notified Bodies assess whether the benefits you promise are real, meaningful, and sufficient to justify exposure to risk. If your evidence for those benefits is thin, the balance tips.

This is why clinical claims are not a documentation exercise. They are the foundation of your entire clinical evaluation. Everything connects back to what you claim the device does and whether you can prove it.

When I audit a CER, I start with the claims. If they are clear and defensible, the rest of the evaluation usually follows logically. If they are unclear or unsupported, the entire CER is at risk.

How This Affects Your PMCF Strategy

Your PMCF plan must address gaps in your clinical evidence, including gaps related to specific claims. If you make a claim based on limited pre-market data, your PMCF should confirm that claim in real-world use.

MDCG 2020-7 requires that PMCF be device-specific and address residual uncertainties. Claims are a primary source of residual uncertainty.

If you claim reduced complication rates, your PMCF must monitor complication rates. If you claim suitability for a specific population, your PMCF must include that population.

This is not optional follow-up. It is required confirmatory evidence. And if your PMCF contradicts your claims, you must update your technical documentation and potentially revise your labeling.

I have seen manufacturers surprised when PMCF data does not support their pre-market claims. But this should not be surprising. Pre-market studies are often limited in size and setting. Real-world use reveals variability.

The solution is not to ignore contradictory PMCF data. The solution is to make conservative pre-market claims that you can confidently confirm, then expand claims as post-market evidence accumulates.

Final Considerations for Regulatory Teams

Clinical claims are not marketing statements that regulatory reviews afterward. They are regulatory commitments that must be established first, then communicated.

Your regulatory and clinical affairs teams should define and validate claims before they reach marketing. Every claim should be mapped to evidence and assessed for sufficiency before it appears in labeling or promotional materials.

This requires coordination. It requires that clinical evaluation drives commercial messaging, not the other way around.

When this alignment exists, regulatory submissions are faster, deficiencies are fewer, and the device enters the market with claims that reflect reality. When alignment is missing, you face repeated review cycles and potential compliance issues post-market.

The question is simple: can you defend every claim you make? If the answer is yes, you are ready. If the answer is uncertain, your CER needs more work.

The time to fix this is before submission, not during review.

Peace,
Hatem
Clinical Evaluation Expert for Medical Devices
Follow me for more insights and practical advice.