IVDR Clinical Performance Studies | When They Are Needed and How to Plan One

Do you need an IVDR clinical performance study for your diagnostic test? Many manufacturers ask this question too late. The answer decides your budget, your timeline, and your path to CE marking.

The short answer is this: it depends on your claims, your existing data, and your device type. This guide helps you decide. You will learn when a study is needed, which approval route applies, and how to plan one without costly surprises.

We keep the language simple and the steps practical. Use it as a checklist before you talk to your notified body or a regulatory consultant.

What Is an IVDR Clinical Performance Study?

An IVDR clinical performance study tests whether your in vitro diagnostic (IVD) gives results that match a real clinical condition. It uses human samples and follows a written plan.

The Regulation (EU) 2017/746 defines a performance study as a study that establishes or confirms the analytical or clinical performance of a device. The Dutch CCMO suggests a simple test: ask whether your study will establish or confirm analytical or clinical performance. If yes, the IVDR applies.

Here is a point many teams miss. A study does not escape the IVDR because a university runs it, because nobody funds it, or because you do not plan to file for CE marking yet. As one MDCG 2025-5 summary explains, the study counts as a performance study if its design aims to establish or confirm performance.

How a Clinical Performance Study Fits Into Performance Evaluation

Your study is one source of data inside a bigger process. That process is called performance evaluation. It rests on three pillars:

• Scientific validity: the link between the analyte and the condition
• Analytical performance: how well the test detects or measures the analyte
• Clinical performance: how well results match the patient’s clinical state

Together, these pillars form your clinical evidence. You plan them in a performance evaluation plan and summarize them in a performance evaluation report.

Importantly, not every performance evaluation needs a performance study. A study is one possible data source. Literature, routine diagnostic data, and earlier studies can also support your claims.

That point leads to the key question: when do you actually need to run an IVDR clinical performance study?

When Is an IVDR Clinical Performance Study Needed?

You need an IVDR clinical performance study when existing data cannot prove your clinical performance claims. The IVDR says clinical performance studies shall be performed unless you justify relying on other data sources.

That wording matters. The default is to run a study. Skipping it requires a written, reasoned justification.

Ask yourself these three questions:

1. Does published data cover my exact intended purpose, population, and sample type?
2. Is the evidence recent enough to match the current state of the art?
3. Can I defend every claim on my label with data I can show?

If you answer “no” to any of them, plan a study. Your device risk class also matters. Higher-risk tests, such as companion diagnostics or blood screening assays, face tougher review.

Cases Where You Usually Need a Study

A study is likely when:

• Your analyte is new, or your marker has little literature behind it
• You claim a new intended purpose or a new patient group
• You change the sample type, such as moving from serum to capillary blood
• A significant device change affects clinical results
• Your performance evaluation plan shows a gap you cannot close with literature

For novel markers, the IVDR is direct. If other approaches are unavailable, a clinical performance study comparing the new device with current clinical practice is required.

Cases Where Other Data May Be Enough

You may not need new work when strong literature covers a well-known marker. Similarly, routine diagnostic data or an earlier study may support your claims, if it matches your intended purpose.

However, you must justify every shortcut. Notified bodies often find their biggest IVDR gaps in clinical performance data. Weak justification invites questions.

Application, Notification, or Neither: Which Route Applies?

After you decide to run a study, you must pick the right regulatory route. This choice shapes your timeline more than almost any other decision.

The European Commission published MDCG 2025-5 on 18 June 2025. It holds 54 questions and answers on IVDR performance studies. Appendix 1 gives a flow diagram that helps you decide between application, notification, or no submission.

Three articles matter most. Article 57 applies to all performance studies, Article 58 adds rules for studies with extra risk, and Article 70 covers studies on CE-marked devices.

One warning: the guidance does not harmonize national authority requirements. Always check the rules in each Member State where you will run the study.

Studies That Need an Application

You need to file an application with the competent authority in these cases, according to the MDCG 2025-5 overview:

• Sampling is surgically invasive and happens only for the study
• Test results may influence patient management or treatment (an interventional study)
• The study adds invasive procedures or other risks for participants
• A companion diagnostic study uses samples that are not left over

Studies That Need Only Notification

A companion diagnostic study that uses only left-over samples follows the notification route. Other studies may need no IVDR submission to authorities at all. Even then, ethics, data protection, and other legal rules still apply.

Left-Over Samples: A Small Detail With Big Impact

The phrase “left-over samples” can cut months from your timeline. It can also trap you if you misread it.

According to MDCG 2025-5, left-over samples are remnants from routine clinical practice that are discarded because no clinical need remains. They can also come from tissue banks and commercial sample vendors.

Here is the trap. Samples you collect prospectively for your study do not count as left over just because the collection would also happen in normal care. Collection for the study makes them study samples.

The distinction also sets your start date. For a prospective study, “start” means the first act of recruitment. For a left-over-sample study, it means the first study-specific analysis.

Because of this, define your sample source in the plan early. Do it before you contact any authority.

How to Plan an IVDR Clinical Performance Study: 7 Steps

A clear plan protects your schedule. Use these seven steps as your roadmap.

Steps 1 to 3: Lay the Foundation

1. Define your intended purpose. Manufacturers carry the responsibility for setting the intended purpose. State what the test detects, in whom, and who uses it.
2. Write the performance evaluation plan. List each claim, the evidence you hold, and the gaps. This plan follows IVDR Annex XIII, Part A.
3. Close the earlier pillars. Confirm scientific validity and finish your analytical performance work first. A test that fails in the lab will fail in the clinic.

Steps 4 to 7: Build and Run the Study

4. Pick the study type and route. Use MDCG 2025-5 to decide between application, notification, or neither.
5. Design the study. Choose the comparator, sample source, sites, sample size, and endpoints.
6. Write the study plan and get approvals. Submit to ethics committees and authorities where required.
7. Run, analyze, and report. Follow good study practice, then feed results into your performance evaluation report.

The Johner Institute’s 8-step process follows a similar flow if you want a second view.

Study Design Choices That Shape Your Results

Design choices decide whether your data convinces a reviewer. Make them with your claims in mind.

Comparator. Choose a reference method that reflects current clinical practice. A weak comparator weakens every result.

Sample selection. Your samples must represent the intended population. Include the age groups, disease stages, and interfering conditions your label covers. Biased selection is a top reason for rejected data.

Sample size. Link the number of samples to your claims. Statistics should show that you can meet your acceptance criteria, not just hit a round number.

Endpoints. Match each endpoint to a performance claim. Typical measures include sensitivity, specificity, and predictive values.

Sites. Multiple sites show your test works in different labs and workflows. More sites also add complexity, so balance both.

Data protection. Studies, including those that use left-over samples, must follow applicable data protection law. Plan consent and data handling before the first sample arrives.

One document can cover several devices. The MDCG guidance says one study plan may cover multiple IVDs if each device, objective, and analysis is unambiguously documented.

Writing the Clinical Performance Study Plan

Every IVDR clinical performance study needs a clinical performance study plan. Annex XIII says the plan defines the rationale, objectives, design, analysis, methodology, monitoring, conduct and record-keeping of the study.

A strong plan covers:

• The objective and the claims it tests
• The target population and selection criteria
• Sample handling, storage, and traceability
• The comparator and the statistical methods
• Acceptance criteria set before the study begins
• Risk management and ethics points
• Monitoring, deviation handling, and record-keeping

Write it so a stranger could run the study from your plan alone. Vague wording leads to vague results.

When the study ends, you write a report. That report must carry the signature of a medical practitioner or other authorised person. It includes the plan, results, and conclusions.

ISO 20916 and Good Study Practice

ISO 20916:2019 is the main standard for running an IVDR clinical performance study. Its full title covers clinical performance studies that use specimens from human subjects, with a focus on good study practice.

The Johner Institute notes that ISO 20916 overlaps with the IVDR on the content of the study plan. The standard calls the plan a Clinical Performance Study Protocol.

Treat both documents as a pair. The IVDR sets the legal minimum. ISO 20916 shows how to run the work well. Following both makes your study easier to defend during review.

Also read MDCG 2022-2 on general clinical evidence principles and MDCG 2022-19 on study application documents. Together, they fill many gaps the regulation leaves open.

Changes, Reporting, and Closing Out Your Study

Your work does not stop once the first sample arrives. Plan for changes and reporting from day one.

Substantial modifications. If you change the study in a way that affects participants or data quality, you may need approval first. MDCG 2025-5 includes an appendix that lists performance study changes that may count as substantial.

Safety events. Studies that fall under the application route must report serious adverse events as the IVDR requires. Set up a clear reporting process before you begin.

Final report. Finish the report, then link its results to your performance evaluation report. Show how the data supports each claim.

After launch. Evidence collection continues. A post-market performance follow-up (PMPF) study under Article 70 can assess a CE-marked device further. Therefore, build a PMPF plan before you launch, not after.

Mistakes That Delay an IVDR Clinical Performance Study

Teams repeat the same errors. Avoid these:

• Choosing the wrong route. Filing an application you do not need wastes weeks. Skipping one you do need can stop the study.
• Mixing up left-over samples. Prospective collection is not left over.
• Setting acceptance criteria after seeing data. Reviewers spot this quickly.
• Using a weak comparator. It undermines every result.
• Ignoring state of the art. You must compare your results with current medical practice.
• Missing the link to claims. Each claim needs matching data.
• Forgetting national rules. MDCG 2025-5 does not replace local requirements.

QbD Group reports that clinical evidence gaps make up a significant share of deficient technical files in notified body reviews. A good plan prevents most of them.

How to Choose Support for Your IVDR Clinical Performance Study

Should you run the study alone or hire help? The right answer depends on your team’s experience, budget, and risk class.

Compare three common options:

• In-house team. You keep control, but you need IVDR, statistics, and clinical operations skills.
• Contract research organization (CRO). A CRO runs sites and data collection well, but it may not write your regulatory strategy.
• Regulatory consultant. A consultant links the study to your technical file, your notified body, and your market plan.

Many manufacturers combine a consultant with a CRO. Whatever you choose, ask each partner these questions:

• Do they know IVDR Annex XIII, MDCG 2025-5, and ISO 20916?
• Can they review your performance evaluation plan for gaps?
• Have they supported IVDs in your risk class?
• Can they help answer notified body questions?

NKB Regovanta supports manufacturers with EU MDR and IVDR regulatory services, global IVD registration, and technical documentation. You can also explore our IVD industry page or book a consultation to review your study plan.

Transition Deadlines and Your Study Timeline

Legacy IVDs moving from the old IVDD to the IVDR face tight deadlines. BSI explains that the extended timelines run up to the end of 2027, 2028 or 2029, depending on device class.

Extra time does not mean extra comfort. A study can take many months to plan, approve, run, and report. Notified body queues add more time.

Therefore, run your gap assessment now. Decide which devices need new studies, and start the long steps first, such as ethics approval and site setup. Our Global Regulatory Knowledge Hub has more guidance on IVDR transition planning and Class D verification.

Frequently Asked Questions About IVDR Clinical Performance Study

What is an IVDR clinical performance study?

It is a formal study that tests whether an IVD’s results match a clinical condition in the intended patient group. It uses human samples, follows a written plan, and ends with a signed report.

When do I need an IVDR clinical performance study?

You need one when existing data cannot support your clinical claims. The IVDR says studies shall be performed unless you justify using other data. Novel markers almost always require one.

Does every performance evaluation need a performance study?

No. A performance study is one possible data source. You can also use literature, routine diagnostic data, or earlier studies if you justify them and they match your intended purpose.

What is the difference between application and notification?

An application needs authority approval before the study starts, for example for interventional or invasive studies. A notification only informs the authority, for example for companion diagnostic studies using only left-over samples.

What counts as left-over samples?

They are remnants from routine clinical practice that are discarded because no clinical need remains. Samples collected prospectively for your study do not count, even if normal care would also collect them.

Is an academic or unfunded study exempt from the IVDR?

No. If a study aims to establish or confirm the analytical or clinical performance of an IVD, the IVDR treats it as a performance study, no matter who runs it or pays for it.

Which guidance helps me plan an IVDR clinical performance study?

Use IVDR Annex XIII, MDCG 2025-5, MDCG 2022-2, MDCG 2022-19, and ISO 20916:2019. Check national authority rules in each country where you run the study.

Who signs the clinical performance study report?

A medical practitioner or another authorised person responsible for the study signs it. The report includes the study plan, results, and conclusions.

Conclusion

An IVDR clinical performance study turns your claims into proof. Knowing when you need one, and which route applies, saves months of work.

To recap, start with a clear intended purpose. Check whether literature can support your claims. If it cannot, plan a study, pick the right route using MDCG 2025-5, and write a detailed study plan that follows IVDR Annex XIII and ISO 20916.

Begin early, and treat approvals and sample sources as part of the design. If you want expert review before you commit, contact NKB Regovanta to talk through your IVDR pathway.