When Your Legacy Device Has Almost No Clinical Data

Written by HATEM RABEH, MD, MSc Ing

Your Clinical Evaluation Expert And Partner

in
S

This is not rare. Many legacy devices were placed on the market under national routes or under MDD with limited scrutiny. They circulated for years. They were safe. They worked. But their clinical evidence base was never built systematically. Now, under MDR, they face a clinical evaluation requirement that assumes a body of evidence exists.

But what if it does not?

This is the third part of our series on legacy device transition. We are looking specifically at how to build an evidence generation strategy when historical data is sparse or absent. The challenge is not just technical. It is strategic. You need to decide what evidence to generate, in what order, and how to present the plan to your Notified Body so they see a coherent path forward, not a reactive scramble.

The MDR Expectation Is Clear

Under MDR Article 61 and Annex XIV, the manufacturer must demonstrate compliance with General Safety and Performance Requirements through clinical evidence. This evidence must be sufficient in terms of quality, quantity, and recency. Sufficient does not mean perfect. It means adequate to support the claims and to address the known risks.

MDCG 2020-6 provides the framework for what constitutes sufficient clinical evidence. It defines the evaluation pathways and the role of equivalence, clinical investigations, and post-market data. But the guidance assumes you have something to start from. It does not offer shortcuts when you have very little.

If your legacy device has sparse data, you are not excused from the requirement. You must generate evidence. The question is how to structure that generation in a way that is proportionate, scientifically sound, and acceptable to the Notified Body.

Assess What You Actually Have

Before you plan new evidence generation, you must inventory what already exists. This sounds obvious, but it is often done incompletely.

Start with your device history file. Look for any clinical data generated during development. This might include bench testing with clinical endpoints, design verification studies, usability studies, or early feasibility data. Even if these were not published, they may be useful if properly documented.

Then review your Post-Market Surveillance records. Complaints, vigilance reports, field safety notices, and service records all contribute to understanding real-world performance. If your device has been on the market for ten years with low complaint rates and no serious incidents, that is data. It needs to be systematically analyzed and presented, not just mentioned.

Next, search the literature. Even if there are no studies on your exact device, there may be literature on the technology class, the clinical application, or comparable devices. MDCG 2020-6 allows the use of equivalent device data under strict conditions. If equivalence is possible, that literature becomes part of your evidence base.

But often, you will find that none of this is enough. The literature is too general. The PMS data is thin. The historical files are incomplete. That is when you need to plan new evidence generation.

Decide What Evidence to Generate

The decision is not arbitrary. It must be based on the clinical evaluation itself. Your CER should identify the gaps. It should state clearly what questions remain unanswered and what evidence would close those gaps.

There are three main paths for generating new evidence: clinical investigations, PMCF studies, and real-world data collection.

Clinical Investigations

A clinical investigation is prospective, controlled, and designed to generate data on safety and performance under defined conditions. It is the gold standard. But it is also expensive, time-consuming, and regulated under MDR Article 62 and Annex XV.

You do not always need a clinical investigation. If your device is low-risk, well-established in its technology, and supported by adequate literature or equivalence data, a clinical investigation may not be proportionate. MDCG 2020-6 acknowledges this. The requirement is for sufficient clinical evidence, not necessarily for a clinical trial.

But if your device is novel, if it makes claims that are not substantiated by existing literature, or if it poses significant risks, a clinical investigation may be unavoidable. In that case, the investigation must be designed to answer the specific questions identified in the CER.

The study design, endpoints, sample size, and follow-up duration must all be justified. The protocol should be reviewed by a competent authority or ethics committee before you begin. This takes time. Plan for at least 12 to 18 months from protocol design to first patient enrollment, and longer for completion.

Post-Market Clinical Follow-Up

PMCF is required under MDR Article 61(11) and Annex XIV Part B. It is not optional. Every device must have a PMCF plan, and that plan must be implemented.

PMCF is designed to confirm safety and performance throughout the lifecycle of the device. It can take many forms: registry participation, surveys, retrospective chart reviews, or prospective observational studies. The format depends on the device, the risks, and the data gaps.

If your legacy device has sparse historical data, PMCF becomes one of your primary tools for evidence generation. You can design a PMCF study that collects structured data on clinical outcomes, user experience, and adverse events in real-world use.

The advantage of PMCF over a clinical investigation is flexibility. You do not need ethical approval in the same way. You do not need the same level of control. You can collect data faster and at lower cost. But the trade-off is that the evidence may be less robust. You must design the study carefully to ensure the data will be credible.

PMCF studies should have clear objectives, defined endpoints, and a sample size justification. They should specify how data will be collected, analyzed, and reported. And they should feed directly back into your clinical evaluation. That is the loop: PMCF generates data, the CER integrates that data, and the CER informs the next PMCF cycle.

Real-World Data Collection

Real-world data is not the same as PMCF, though they overlap. Real-world data refers to information collected outside of a controlled study: sales data, complaint trends, user feedback, registry data, or claims databases.

This data is useful but often difficult to structure. It is retrospective, uncontrolled, and variable in quality. But if you are starting with almost no clinical data, real-world evidence can fill some gaps, especially for understanding long-term safety.

The key is to formalize the collection. Do not wait for complaints to arrive passively. Design a process to actively gather data from distributors, users, and service teams. Structure that data so it can be analyzed statistically. Present it in your CER as part of the safety profile.

But be realistic about its limitations. Real-world data is good for detecting signals and trends. It is less good for proving efficacy or comparing outcomes. Know what questions it can answer and what it cannot.

Build a Phased Evidence Generation Plan

You cannot generate all the evidence at once. You need a plan that unfolds over time, with each phase building on the previous one.

Start by defining what is absolutely necessary for initial certification. What is the minimum evidence required to demonstrate that the device is safe and performs as intended? That is your Phase 1. This might include a targeted literature review, initial PMS analysis, and a small PMCF pilot study.

Once you have initial certification, you move to Phase 2: deeper evidence generation. This could include a full PMCF study, a registry analysis, or a clinical investigation if needed. This phase should be outlined in your PMCF plan and referenced in your CER.

Phase 3 is ongoing surveillance and periodic update. Your CER must be updated at least annually or when new information arises. Each update should integrate new data from PMCF, PMS, and literature.

The Notified Body needs to see this roadmap upfront. They need to understand that you are not claiming the evidence is complete, but that you have a structured plan to close the gaps. Present the plan clearly in the CER. Reference it in the PMCF plan. Show the timelines, the deliverables, and the decision points.

Justify Every Decision

When your evidence base is thin, every decision you make will be scrutinized. Why did you choose PMCF over a clinical investigation? Why is the sample size 50 and not 100? Why is the follow-up period six months and not twelve?

You must justify each choice. The justification should be based on risk, clinical relevance, and feasibility. If the device is low-risk and well-understood, a smaller PMCF study may be sufficient. If the device is higher-risk or makes novel claims, you need more robust evidence.

Reference the guidance. MDCG 2020-6 provides the framework for proportionality. Use it. Show that your approach is consistent with the guidance and appropriate for your device.

And be honest about limitations. If your PMCF study has a small sample size because of budget constraints, say so. But also explain how you will compensate: longer follow-up, more detailed data collection, or staged expansion of the study.

Transparency builds credibility. The Notified Body is not expecting perfection. They are expecting a rational plan, honestly presented, with clear milestones and realistic timelines.

Anticipate Notified Body Questions

When you submit a clinical evaluation with a sparse evidence base and a generation plan, the Notified Body will ask questions. Anticipate them.

They will ask why the evidence is sparse. You need a clear answer: the device is legacy, it was placed on the market under MDD with less stringent requirements, and the manufacturer is now systematically building the evidence base under MDR.

They will ask whether the device can be certified with the current evidence. You need to demonstrate that the existing evidence, combined with the committed generation plan, is sufficient to support initial certification with ongoing obligations.

They will ask about the timelines. Be specific. Do not say “we will conduct a PMCF study.” Say “we will initiate a PMCF study in Q2 2025, enroll 50 patients over 12 months, and report results in Q3 2026.”

They will ask how the new evidence will be integrated. Explain the CER update process, the trigger points for review, and the procedure for escalating findings if safety concerns arise.

Prepare these answers in advance. Include them in the CER. Do not make the Notified Body extract them through questions. That slows the process and signals that you have not thought it through.

Accept That This Takes Time

Generating clinical evidence is not fast. A PMCF study can take two years from design to final report. A clinical investigation can take longer. If you are starting from a sparse evidence base, you are committing to a multi-year process.

Plan accordingly. Budget for it. Resource it. And communicate the timeline to your stakeholders. Do not promise your sales team that the device will be certified in six months if the evidence generation plan requires 18 months of data collection.

The Notified Body may grant initial certification with conditions. They may require interim reports, periodic updates, or staged evidence submission. This is normal when the evidence base is being built. Accept it as part of the process.

What matters is that the plan is credible, the timelines are realistic, and the commitments are met. Deliver on what you promise. If you say you will submit a PMCF report in Q3 2026, submit it in Q3 2026. Consistency builds trust.

Document Everything

When you are generating evidence to close gaps, documentation becomes even more important. Every study protocol, every data collection form, every analysis plan must be documented and retained in the technical file.

The CER should reference these documents explicitly. Do not just say “a PMCF study is planned.” Say “PMCF Study Protocol XYZ, dated January 2025, is included in the technical file as Document 4.3.2.”

When the study is complete, the report should be integrated into the CER. The integration should not be superficial. It should analyze the findings, compare them to the expected outcomes, and assess their impact on the benefit-risk profile.

This level of documentation shows that the evidence generation process is systematic, controlled, and auditable. It is what the Notified Body expects. It is what an inspector will look for.

Learn From What Is Accepted

I have seen manufacturers succeed with very limited initial evidence. The key was always the same: they presented a clear gap analysis, a rational generation plan, and a commitment to deliver on the plan.

One case involved a Class IIb surgical instrument with almost no published literature and minimal historical PMS data. The manufacturer designed a multi-center PMCF study, defined clear endpoints, and committed to a two-year data collection period. The Notified Body granted initial certification with the condition that the PMCF report would be submitted within 30 months. The manufacturer delivered the report on time. The certification was maintained.

Another case involved a diagnostic device with no clinical investigation. The manufacturer used equivalence data from a predicate device, supplemented it with a structured real-world data collection program, and initiated a registry study. The clinical evaluation was accepted because the plan was proportionate to the risk and clearly documented.

What these cases have in common is clarity. The Notified Body could see what was being done, why it was being done, and when it would be done. There were no surprises. There was no ambiguity.

That is what you are aiming for.

Closing Thoughts

When your legacy device has sparse clinical data, the path to compliance is longer and more structured. You cannot rely on shortcuts. You must build the evidence base systematically, phase by phase, with clear justification at every step.

The MDR does not excuse legacy devices. It requires them to meet the same clinical evidence standards as new devices, but it allows proportionality in how that evidence is generated. Use that flexibility. Design a plan that is appropriate for your device, realistic for your resources, and acceptable to your Notified Body.

And remember: the evidence generation process is not separate from the clinical evaluation. It is part of it. The CER should describe the current evidence and the plan to strengthen it. The PMCF plan should feed into the CER. The technical file should document every step.

This is how you turn a sparse evidence base into a compliant one. Not by avoiding the requirement, but by meeting it strategically.

In the next part of this series, we will look at how to handle equivalence claims for legacy devices when the predicate device is also transitioning.