When clinical investigation becomes unavoidable—and how to plan it

Written by HATEM RABEH, MD, MSc Ing

Your Clinical Evaluation Expert And Partner

in
S

The regulatory landscape has shifted. Under MDR Article 61(4) and (5), clinical investigation is no longer a last resort. For certain device categories—implantables, Class III devices, devices with novel features or mechanisms—clinical investigation becomes the expected path when existing data cannot sufficiently demonstrate safety and performance.

The question is not whether you need it. The question is whether you are planning it correctly.

Why equivalence collapses and investigation becomes necessary

Most manufacturers I work with arrive at clinical investigation after exhausting the equivalence pathway. They tried to demonstrate equivalence under MDR Annex XIV, Section 3. They failed.

The reasons are usually the same: technical differences that affect clinical outcomes, insufficient data on the equivalent device, or changes in the regulatory interpretation of what constitutes acceptable equivalence. The Notified Body requests additional clinical data. The manufacturer realizes that generating that data through literature or registries will take years—or is simply impossible.

At this point, clinical investigation becomes unavoidable.

But here is where the real challenge begins. Starting a clinical investigation without clarity on the regulatory endpoint is like building a bridge without knowing where it should land. The investigation may generate data, but if that data does not address the specific clinical questions required by MDR, it will not support conformity assessment.

What MDR actually requires from clinical investigation

MDR Article 62 specifies that clinical investigations must be conducted in accordance with Annex XV. The investigation must be designed to verify safety and performance. It must address the clinical claims made by the manufacturer. And it must generate data sufficient to demonstrate compliance with the general safety and performance requirements (GSPRs) in Annex I.

This sounds straightforward. In practice, it is not.

I have reviewed investigation protocols where the primary endpoint was selected based on clinical convention, not regulatory necessity. The study would have answered an interesting clinical question. But it would not have satisfied the Notified Body’s clinical data requirements.

The protocol must be designed backward from the clinical evaluation report. You need to know what gaps exist in your clinical evidence. You need to know which GSPRs cannot be addressed without new clinical data. You need to know what performance claims require verification through controlled investigation.

Only then can you design a study that will actually close those gaps.

Planning the investigation: regulatory alignment first

The starting point is not the clinical protocol. It is the clinical evaluation plan and the draft clinical evaluation report.

Before designing the investigation, complete a gap analysis. Identify which GSPRs lack sufficient clinical data. Identify which risks in your risk management file are not adequately mitigated by existing clinical evidence. Identify which performance claims in your IFU are not supported by clinical data of adequate quality and relevance.

This gap analysis defines the regulatory objective of the investigation. Every endpoint, every inclusion criterion, every safety assessment must map back to these gaps.

I recommend drafting the clinical evaluation report sections that will incorporate the investigation results before finalizing the protocol. Write the sentences you will use to claim conformity. Identify the data points you need to support those sentences. Then design the protocol to generate exactly those data points.

This approach ensures alignment between clinical objectives and regulatory requirements.

Choosing the right study design

MDR does not mandate a specific study design. Annex XV allows for various approaches: randomized controlled trials, single-arm studies, observational studies, or registry-based investigations—depending on the clinical question and the available comparators.

But the choice must be justified. If a randomized controlled trial is feasible and ethically acceptable, a single-arm study will face scrutiny. If a well-established comparator exists, the Notified Body will expect a controlled comparison.

The justification for the study design should be documented in the clinical evaluation plan and reflected in the investigation protocol. This justification becomes critical during Notified Body review.

Defining endpoints that satisfy regulators, not just clinicians

The primary endpoint must address a GSPR or a performance claim. It must be measurable, relevant, and sufficient to demonstrate conformity.

I often see endpoints chosen because they are standard in the clinical field. But clinical convention is not the same as regulatory sufficiency. A clinically interesting endpoint may not address the specific safety or performance question that the Notified Body identified as a gap.

The secondary endpoints should address residual risks and support the benefit-risk determination. They should align with the risk management file and the clinical evaluation plan.

Every endpoint must have a clear regulatory purpose.

Engaging the Notified Body early

MDR Article 61(5) and MDCG 2020-6 recommend early consultation with the Notified Body regarding the necessity and scope of clinical investigation. This is not optional if you want to avoid protocol rejections.

The Notified Body needs to confirm that the proposed investigation will satisfy their clinical data requirements. They need to see that the endpoints address the gaps they have identified. They need to understand how the results will support the clinical evaluation.

I recommend submitting a pre-investigation consultation package that includes:

• The gap analysis from the clinical evaluation plan
• A summary of the GSPRs and performance claims that require new clinical data
• The proposed study design with justification
• Draft primary and secondary endpoints with regulatory rationale
• An outline of how the results will be integrated into the CER

This package allows the Notified Body to provide feedback before you finalize the protocol and initiate the study. Adjusting the protocol after enrollment is costly and often infeasible. Adjusting it during planning is straightforward.

Building the investigation into the PMCF framework

Under MDR, clinical investigation is not separate from post-market clinical follow-up. They are both components of the continuous clinical evaluation process required by Article 61(1) and Annex XIV, Part B.

The investigation generates initial clinical data. PMCF confirms that the safety and performance observed during the investigation are maintained in real-world use.

This means the PMCF plan must be designed in parallel with the investigation protocol. The PMCF endpoints should align with the investigation endpoints. The follow-up duration should be sufficient to capture long-term outcomes that the investigation could not assess.

I have seen investigations that succeeded in demonstrating short-term safety and performance, but failed to establish a coherent PMCF framework for long-term surveillance. The Notified Body approved the CER based on the investigation data, but flagged the PMCF plan as insufficient. The result was delayed certification.

Plan the investigation and the PMCF as integrated components of the same clinical evidence generation strategy.

Understanding what success looks like

Success is not publication. Success is not statistical significance. Success is generating data that closes the gaps identified in your clinical evaluation and satisfies the Notified Body’s requirements for conformity assessment.

This requires clarity on the acceptance criteria before the study begins. What results will be considered sufficient to demonstrate safety? What performance thresholds must be met to support the intended use claims? What adverse event rate will be acceptable given the benefit-risk profile?

These criteria should be documented in the clinical evaluation plan and reflected in the investigation protocol. They define the regulatory endpoint.

I recommend drafting the sections of the CER that will present the investigation results before you finalize the protocol. Write the conformity claims you intend to make. Identify the data points required to support those claims. Ensure the protocol is designed to generate those data points with adequate statistical power and clinical relevance.

This approach ensures that the investigation will succeed regulatorily, not just clinically.

Final considerations: timeline and resource planning

Clinical investigation under MDR is not fast. Between protocol development, ethics committee approval, competent authority notification, site initiation, enrollment, follow-up, data analysis, and CER integration, you are looking at a minimum of 18 to 24 months—often longer for implantables or long-term outcomes.

The timeline must be realistic. It must account for regulatory review cycles, site activation delays, enrollment challenges, and data maturation. It must also account for the possibility of protocol amendments if early results reveal safety concerns or enrollment challenges.

Resource planning is equally critical. Clinical investigation is expensive. Budget for site costs, monitoring, data management, statistical analysis, and regulatory submissions. Budget for potential delays and protocol amendments.

But most importantly, budget for the clinical and regulatory expertise required to design the investigation correctly from the start. An investigation that fails to satisfy regulatory requirements is more expensive than one that was properly planned.

When clinical investigation becomes unavoidable, the path forward is not about generating data. It is about generating the right data—data that closes specific gaps, satisfies specific requirements, and supports specific conformity claims.

This requires alignment between clinical objectives and regulatory requirements. It requires early engagement with the Notified Body. It requires integration with the PMCF framework. And it requires clarity on what success looks like before the first patient is enrolled.

Most manufacturers approach clinical investigation as a clinical project. It is not. It is a regulatory project with clinical execution. Treat it accordingly, and you increase the probability that the investigation will succeed where it matters: in supporting conformity assessment under MDR.

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