Why confusing Clinical Investigation with PMCF fails reviews

Written by HATEM RABEH, MD, MSc Ing

Your Clinical Evaluation Expert And Partner

in
S

The confusion stems from a shared characteristic: both involve collecting clinical data on your device. But that is where the similarity ends.

Under MDR 2017/745, clinical investigations and PMCF studies sit on opposite sides of the device lifecycle. They serve different purposes, follow different procedures, and carry different regulatory consequences.

Mixing them up does not just delay approval. It exposes you to audit findings, ethics committee rejections, and competent authority objections.

The Regulatory Framework

Clinical investigations are defined in Article 2(45) of the MDR as systematic investigations involving human subjects to assess the safety or performance of a device.

They are what you conduct before CE marking, when you need clinical evidence to support conformity assessment. The procedure is described in Article 62 through Article 81 of the MDR.

PMCF studies, on the other hand, fall under Annex XIV Part B. They are post-market activities conducted to confirm the safety and performance of a device that is already CE marked and in use.

The distinction is temporal and functional. Clinical investigations generate evidence needed for market access. PMCF studies confirm and update that evidence after market access.

But here is where manufacturers trip.

They assume that because the device is CE marked, they can design any study as PMCF without restrictions. That is not how it works.

When PMCF Becomes a Clinical Investigation

MDCG 2020-7 on Post-Market Clinical Follow-up clarifies this point explicitly.

If your study involves an intervention that deviates from normal clinical practice, if it exposes patients to additional risks, or if it uses the device in a way not covered by the IFU or intended use, you are conducting a clinical investigation.

Even if the device is already on the market.

I have seen manufacturers design PMCF protocols that randomize patients to two treatment arms, include invasive follow-up procedures not part of routine care, or test performance parameters never claimed in the technical documentation. They submit these as PMCF studies.

Then the ethics committee rejects the protocol because it requires full clinical investigation review under national law.

The manufacturer loses months resubmitting under the correct pathway.

The trigger is not just the CE mark status. It is the nature of the intervention and the risk profile of the data collection itself.

Procedural Differences That Matter

Once you correctly classify your study, the procedures diverge completely.

For a clinical investigation, you must notify the competent authority of the Member State where the study will be conducted. In some countries, this is implicit approval if no objection is raised within a set period. In others, it requires explicit authorization.

You must submit the clinical investigation plan (CIP) to an ethics committee that meets the requirements of Article 62(4). The ethics committee has 60 days to issue an opinion, extendable in certain cases.

You must have a sponsor, defined and responsible under Article 62(2). The sponsor must have insurance or indemnity to cover liability under Article 69.

For PMCF studies conducted as part of normal clinical practice, there is no requirement to notify competent authorities in advance, though some Member States have national registries or notification schemes.

Ethics committee review depends on national law. In many countries, if the PMCF study does not deviate from normal clinical practice and does not add risk, it may be exempt from full ethics review or subject to simplified procedures.

You do not need a sponsor in the MDR sense. You do not need separate clinical investigation insurance.

The administrative burden is fundamentally different.

Misclassifying your study means restarting the clock under a different regulatory pathway. That is why Notified Bodies ask pointed questions about study design during CER reviews.

How Reviewers Spot the Confusion

When I review a PMCF plan or a clinical evaluation report, I look for signals that the manufacturer does not understand this distinction.

One signal is when the PMCF plan describes a study protocol with outcome measures that are not part of routine clinical use. For example, a cardiovascular device where the PMCF plan includes echocardiography follow-up at 6 months in all patients, but standard care for this device does not routinely include echo at 6 months.

That is an additional procedure. That adds risk. That requires ethics review as a clinical investigation.

Another signal is when the manufacturer includes a comparator arm in the PMCF study. Comparing your device to a competitor device or to standard therapy in a structured protocol is interventional research. It is not post-market surveillance of your device in normal use.

A third signal is when the PMCF plan involves patients who would not normally receive the device under the current intended use, but are included to explore a new indication. That is expanding the intended use. That requires a clinical investigation.

The Notified Body does not accept this. The reviewers will issue a finding that the study must be reclassified and resubmitted through the clinical investigation pathway.

This happens more often than manufacturers expect.

The Grey Zone: Registry Studies

There is a grey zone that creates confusion: registry-based PMCF studies.

Many specialties have clinical registries where patient data is routinely collected as part of quality assurance or national surveillance programs. If your device is used in patients enrolled in such a registry, can you use that data for PMCF without triggering clinical investigation requirements?

The answer depends on whether the registry itself involves interventions beyond normal care and whether your participation in the registry changes clinical practice.

If the registry is purely observational, collects data that clinicians would document anyway, and does not impose additional tests or visits, then extracting data from that registry for PMCF is acceptable under normal post-market surveillance.

But if the registry protocol requires additional imaging, lab tests, or follow-up intervals that exceed normal care, then participation in the registry is itself an interventional study. Your use of that data does not make it less interventional.

Some manufacturers assume that because the registry is run by a professional society or a national body, it automatically qualifies as normal clinical practice. That is not the logic that applies.

The logic is: does participation add risk or burden to the patient beyond what standard care would involve?

If yes, it is interventional. If interventional, any prospective enrollment tied to your device requires clinical investigation procedures.

I have seen PMCF plans rejected because the manufacturer planned to enroll patients into a registry that required annual CT scans not part of routine follow-up for that device class.

The Notified Body correctly identified this as a clinical investigation in disguise.

Impact on the Clinical Evaluation Report

The classification of your study affects how the data is evaluated in the CER.

Data from a clinical investigation conducted pre-market is appraisable clinical data under MEDDEV 2.7/1 Rev 4 methodology (still referenced under MDR for clinical evaluation principles). It is designed to answer specific safety and performance questions.

Data from PMCF studies is confirmatory. It validates that the performance and safety observed in pre-market studies hold true in real-world use.

But if you have labeled something as PMCF when it is actually a clinical investigation, the data may not be accepted for the purpose you intended.

For example, if you need clinical data to support a new intended use, and you conduct a study under a PMCF framework without proper ethics and competent authority procedures, that data may be deemed non-compliant during review.

The Notified Body will not accept non-compliant data as part of the clinical evidence base. You will have deficiencies in your CER related to insufficient clinical evidence.

This is not theoretical. I have reviewed CERs where the manufacturer conducted what they called PMCF studies to extend indications, but because they did not follow clinical investigation procedures, the data could not be used to support the claim. The file was incomplete.

Practical Decision Framework

Before you write your PMCF plan or clinical investigation plan, ask these questions:

Is the device already CE marked for the intended use in question?
If no, you are conducting a clinical investigation. If yes, proceed to the next question.

Does the study involve any intervention beyond normal clinical practice?
This includes additional visits, tests, imaging, or procedures that would not occur if the patient were not in the study. If yes, it is a clinical investigation. If no, proceed.

Does the study randomize patients or include a control arm with a different intervention?
If yes, it is a clinical investigation. If no, proceed.

Does the study test the device in a patient population or indication not covered by the current IFU or intended use?
If yes, it is a clinical investigation. If no, it is likely a PMCF study.

This framework is not exhaustive, but it catches most cases where manufacturers misclassify their studies.

When in doubt, consult the ethics committee early. Present the protocol outline before full submission. Many ethics committees will provide guidance on whether the study requires full review under clinical investigation rules or qualifies for simplified or exempt review under observational post-market surveillance.

That early conversation saves months.

Why This Matters More Under MDR

Under the Medical Device Directive (MDD), the distinction was less enforced in practice. Many manufacturers conducted hybrid studies that were not clearly classified, and Notified Bodies accepted data without strict procedural scrutiny.

MDR changed that. Articles 62 to 81 are detailed and prescriptive. Annex XIV Part B explicitly defines PMCF requirements. MDCG guidance documents clarify expectations.

Notified Bodies now have less discretion to overlook procedural non-compliance. Competent authorities audit clinical investigation notifications more rigorously.

If your study does not fit neatly into the correct regulatory pathway, you will face delays, findings, and potential rejection of your clinical data.

The stakes are higher because the clinical evaluation requirements are stricter. You cannot afford to lose valid clinical data because you misclassified the study.

That is why this distinction is not academic. It is operational risk.

Final Thought

The difference between a clinical investigation and a PMCF study is not semantic. It determines your regulatory pathway, your timeline, your ethical obligations, and whether your data can support your claims.

If you are unsure, classify conservatively. An ethics committee can tell you if a study qualifies for simplified review. But if you skip the process assuming it is PMCF, and you are wrong, the consequences are not fixable with a protocol amendment.

You restart from the beginning, under the correct procedure. And your CER timeline extends by months.

The device is the same. The patients are the same. But the procedure is not.

That is the distinction that matters.

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