The clinical investigation plan (CIP) is the master document that specifies, before a single subject is enrolled, how a clinical investigation of a medical device will be conducted. Under MDR Article 62(4), every clinical investigation must be conducted on the basis of a clinical investigation plan. The required content of the CIP is set out in Annex XV Chapter II of Regulation (EU) 2017/745, and the operational requirements are reinforced by the harmonised standard EN ISO 14155:2020+A11:2024. Writing a CIP is not a formatting exercise. It is the act of committing, in writing and in advance, to the scientific question, the design, the endpoints, the statistics, the risk controls, and the conduct rules of the study — in a form the ethics committee, the competent authority, and the Notified Body will all hold you to.

By Tibor Zechmeister and Felix Lenhard. Last updated 10 April 2026.

TL;DR

  • The clinical investigation plan (CIP) is required by MDR Article 62(4) for every clinical investigation conducted to demonstrate conformity.
  • The required content of the CIP is specified in Annex XV Chapter II of Regulation (EU) 2017/745 and is enforced section by section by the ethics committee and the competent authority.
  • EN ISO 14155:2020+A11:2024 is the harmonised Good Clinical Practice standard that operationalises the CIP contents and is the practical template most sponsors follow.
  • The CIP is not a living marketing document. Changes after authorisation trigger substantial amendment procedures under Article 75.
  • A well-written CIP answers one scientific question clearly. A poorly written CIP tries to answer five and fails all of them.
  • The most common CIP failures at submission are vague endpoints, missing risk-benefit justification, under-specified statistical analysis, and a monitoring plan that does not match the sponsor's actual capacity.

Why the CIP is the document that decides the study

Every clinical investigation under MDR stands or falls on its CIP. The ethics committee reviews it. The competent authority authorises the study against it. The investigator runs the study by it. The monitors audit compliance to it. The final clinical investigation report is written against it. And the Notified Body, years later, reads it back against the clinical evaluation to decide whether the evidence in the file was generated by a scientifically sound study.

If the CIP is vague, everything downstream is vague. If the CIP commits to an endpoint that the device cannot meet, the study fails publicly and permanently. If the CIP is written to "keep options open," the ethics committee will send it back. If the CIP is copy-pasted from a template that does not match the actual device and actual intended purpose, the authorisation package will unravel on first scrutiny.

The CIP is also the discipline mechanism. Writing it forces the sponsor to decide — in one document, before any patient is touched — what the study is actually for, what it will measure, how it will measure it, who is accountable, and what will be done with the data. Founders who skip the discipline of writing a tight CIP end up running loose studies. Loose studies produce data the Notified Body cannot use.

This post walks through the required contents of the CIP as specified by MDR Annex XV Chapter II, section by section, in the order they must appear. Use it as a checklist for your own draft.

The legal anchor — MDR Article 62(4)

MDR Article 62(4) sets out the requirement. A clinical investigation conducted to demonstrate conformity of devices with the applicable requirements shall be conducted in accordance with a clinical investigation plan. The plan must be set up before the investigation begins. Any substantial modification during the investigation goes through the amendment procedure of Article 75.

Article 62(4) is short. Annex XV Chapter II is where the detailed content requirements live. Chapter II lists the specific sections that the CIP must contain, and every one of them is non-negotiable. A CIP that is missing a Chapter II section is an incomplete CIP, and an incomplete CIP cannot be authorised.

EN ISO 14155:2020+A11:2024 — Clinical investigation of medical devices for human subjects — Good clinical practice — is the harmonised standard. It provides the practical template for a CIP that satisfies both Annex XV and the Good Clinical Practice expectations of every European competent authority. Use the standard as the operational floor. Use Annex XV Chapter II as the legal ceiling. The two agree on almost everything; where they diverge, Annex XV wins.

Walking Annex XV Chapter II — the required sections of a CIP

What follows is a section-by-section walk through the CIP content required by Annex XV Chapter II. Under each heading is a plain-language explanation of what the section must contain, why it matters, and what Notified Bodies and ethics committees expect to see. This is the structure every CIP for a conformity-assessment clinical investigation under MDR must follow.

1. General identification of the investigation, the sponsor, and the investigators

The CIP opens with identification. The title of the investigation, a unique identifying number, the sponsor's name and address, the identification of the principal investigator at each site and the coordinating investigator where relevant, and the identification of the monitor. This section is short but unforgiving — every named role carries legal responsibility under Article 72 and under EN ISO 14155:2020+A11:2024, and every name has to be a real person with documented qualification.

2. Identification and description of the device

The device that is the subject of the investigation must be described in enough detail that the reader understands what is being tested. Intended purpose, classification under Annex VIII, principle of operation, materials in contact with the patient, the parts that are investigational as opposed to parts that are already CE-marked and used as intended, and the identification of any accessories used in the study. If the device is software, the version, the environment it runs on, and the functions under investigation are explicit.

3. Justification for the design of the investigation

This section is where the study design is defended. Why a randomised controlled trial, or why a single-arm study, or why a cross-over, or why an observational design. Why the control or comparator chosen. Why the duration of follow-up. Why the study population. The justification has to hold up against the scientific question the study exists to answer. Notified Bodies read this section hardest when the design looks unusual or lean — if the design is minimal, the justification must be strong.

4. Risks and benefits of the device and the clinical investigation

The risk-benefit analysis. The known and foreseeable risks of the device under investigation, the known and foreseeable risks of the procedures involved in the investigation itself, the expected benefits to subjects or to future patients, and the reasoned conclusion that the benefits justify the risks. This section connects directly to the risk management file under EN ISO 14971:2019+A11:2021 and to the pre-clinical evidence package. A risk-benefit section that is not traceable to the risk file will not be accepted.

5. Objectives and hypotheses of the investigation

The primary objective of the investigation, written precisely. Any secondary objectives. The hypotheses that the investigation is designed to test, written in a form that allows the statistical analysis to either confirm or reject them. This is the section where "we want to explore" language must be replaced with "we will test whether." If the objective cannot be written as a testable claim, the study is not ready for authorisation.

6. Design of the investigation

The full design in structured form. Type of study (interventional, observational, pivotal, first-in-human, confirmatory, post-market). Number of arms. Allocation procedure. Blinding if any. Randomisation if any. Number of subjects and the justification for that number. Number of investigation sites. Duration per subject. Total duration of the investigation. This section is the skeleton the whole study hangs on, and the ethics committee will compare every downstream detail back to it.

7. Statistical considerations

Sample size calculation with the assumptions behind it. Primary endpoint and its statistical analysis. Secondary endpoints and their analyses. Handling of missing data. Interim analyses if any, and the stopping rules that attach to them. The level of significance and the power of the study. If the study uses Bayesian methods or adaptive design, the statistical framework must be specified in advance and not improvised during conduct. A CIP with a hand-waved statistics section is a CIP that will come back from review with the same comment every authority writes: "specify."

8. Data management

How data will be captured (paper case report forms, electronic data capture, device logs, imaging). How data will be stored, secured, and backed up. Who has access. How data will be monitored and reconciled. How queries are handled. How the database is locked. How data is archived at the end of the study and for how long — the retention requirements of EN ISO 14155:2020+A11:2024 and national law both apply and both are long.

9. Information on subjects

How subjects will be selected. Inclusion and exclusion criteria, written precisely enough that a screening coordinator can apply them at a site without interpretation. How informed consent will be obtained in line with MDR Article 63 and the requirements for specific vulnerable populations where applicable (Articles 64 to 68). What information subjects receive, in what language, and in what form. How withdrawal is handled.

10. Information on treatment and procedures

The full description of what happens to each subject. Screening procedures. Baseline assessments. Device use procedures — who does what, when, with what training. Follow-up visits, their timing, and what is done at each one. Rescue treatments if applicable. Concomitant therapies and medications allowed or prohibited. This section is written so that a new investigator at a new site could run the study correctly on the strength of this section alone.

11. Monitoring plan

How the sponsor will verify that the study is being conducted in accordance with the CIP and the applicable regulations. Frequency of monitoring visits. What is monitored (source data verification, consent documentation, device accountability, adverse event reporting, deviations). Who monitors — by name and qualification. For a startup, the monitoring plan must match real capacity. A plan that calls for visits a small sponsor cannot actually staff is a plan that will be broken, and broken monitoring plans generate findings that poison the study.

12. Safety evaluation and adverse event reporting

How adverse events and device deficiencies are identified, documented, assessed for causality, and reported. The definitions of adverse event, serious adverse event, adverse device effect, and serious adverse device effect, as used in EN ISO 14155:2020+A11:2024 and MDR Article 80. The timelines for reporting to the sponsor, the competent authority, and the ethics committee under Article 80 — short, absolute, and the same for small sponsors as for large ones. The escalation path and the stopping rules.

13. Suspension or early termination

The conditions under which the investigation will be suspended or terminated early, whether for safety, futility, or logistical reasons. Who decides. How subjects in the study at the time of suspension are handled. How the competent authority and the ethics committee are notified. Article 77 governs end-of-investigation reporting including after early termination.

14. Amendments to the CIP

The procedure for changes to the CIP during the investigation. Substantial amendments under MDR Article 75 require notification to and, in many cases, authorisation by the competent authority and the ethics committee before they take effect. Non-substantial amendments are logged and tracked but do not require authorisation. The CIP itself must describe how the sponsor distinguishes between the two, who approves amendments internally, and how versions are controlled.

15. Publication policy

The sponsor's policy on publishing the results of the investigation, whether positive, negative, or inconclusive. MDR commits sponsors to transparency expectations around clinical investigation results, and the CIP states the publication plan in writing. A CIP that implies results will only be published if favourable is a red flag for reviewers.

16. References and supporting documents

Scientific literature cited. The investigator's brochure. The risk management file extract relevant to the investigation. The pre-clinical evidence supporting the safety of first human use. Any prior clinical data on the device or on equivalent devices. The reference list is not decoration — reviewers read it.

Common CIP mistakes we see at submission

A short, non-exhaustive list of the failures that come back most often when a CIP is reviewed by ethics committees, competent authorities, or Notified Bodies.

Vague primary endpoint. "Improvement in patient outcomes" is not a primary endpoint. A primary endpoint names the specific measurement, the specific instrument, the specific timepoint, and the specific analysis. If the endpoint cannot be written in one sentence that a statistician can program, the CIP is not ready.

Missing or hand-waved sample size justification. Reviewers want to see the assumptions: expected effect size, variance, power, significance level, drop-out rate. "Based on clinical judgment" is not a justification.

Risk-benefit section disconnected from the risk file. The risks listed in the CIP must be the same risks controlled in the risk management file under EN ISO 14971:2019+A11:2021. Divergence between the two documents is a finding every time.

Monitoring plan the sponsor cannot actually execute. If the CIP commits to monthly on-site monitoring visits and the sponsor has one part-time monitor, the plan is fiction. Better to write a plan that matches reality and defend it, than to write an aspirational plan the study will violate.

Inclusion and exclusion criteria that do not match the intended purpose. If the intended purpose targets one population and the study enrols another, the data generated will not support the intended purpose. This is one of the most expensive mistakes possible — it surfaces at Notified Body review, years and hundreds of thousands of euros after it could have been fixed.

Copy-pasted text from a template CIP. Placeholder company names, wrong device descriptions, references to adverse event terminology that does not match MDR. Template CIPs are a starting point, not a destination. Every sentence must be true of this device, this sponsor, this study.

Statistical analysis plan deferred to "a separate document to be developed." The CIP needs the statistical analysis in enough detail to support authorisation. A full separate SAP can exist, but the CIP itself must specify the core analysis, not defer it.

The Subtract to Ship angle on the CIP

Our Subtract to Ship framework for MDR applies to the CIP the same way it applies to every other regulatory document. The test is not "is the CIP long." The test is "does every section of this CIP earn its place against Annex XV Chapter II and EN ISO 14155:2020+A11:2024."

A subtracted CIP is not a short CIP. It is a CIP with no padding, no duplication, no aspirational sections, and no hedging language. One scientific question. One primary endpoint. One statistical analysis plan. One monitoring plan that matches real capacity. One risk-benefit argument traceable to the risk file. One population that matches the intended purpose.

The contrast we see in practice: sponsors who write the CIP as a ninety-page artifact meant to impress reviewers, versus sponsors who write it as a thirty- to forty-page working document designed to run the study. The working document almost always wins authorisation faster and generates cleaner data, because the investigators can actually read it and follow it. Length is not rigour. Traceability to Annex XV Chapter II is rigour.

For the broader frame on when a clinical investigation is required in the first place, see What Is a Clinical Investigation Under MDR?. For the decision of whether to run one at all, see Do You Need a Clinical Investigation? The MDR Decision Tree. For the step-by-step design work that feeds the CIP, see How to Design a Clinical Investigation as a Startup.

Reality Check — Where do you stand?

  1. Can you state the primary objective of your planned investigation in one sentence that names the specific measurement, the specific instrument, and the specific timepoint?
  2. Is your primary endpoint written in a form a statistician can program without asking clarifying questions?
  3. Have you run the sample size calculation yourself, and can you defend every assumption in it?
  4. Does the risks section of your draft CIP match, line for line, the risks in your risk management file under EN ISO 14971:2019+A11:2021?
  5. Does your inclusion and exclusion criteria list match the population in your intended purpose, without gaps and without overreach?
  6. Is your monitoring plan something your sponsor team can actually execute on the schedule you wrote?
  7. Do you have a named principal investigator at every planned site, with documented qualification on file?
  8. Have you walked your current draft CIP against Annex XV Chapter II section by section and confirmed every section is present and non-trivial?
  9. Is your adverse event reporting procedure written with the Article 80 timelines, or with a vague "we will report promptly"?
  10. Have you decided, and written in the CIP, the rules under which the study will be suspended or terminated early?

Frequently Asked Questions

What is a clinical investigation plan under MDR? A clinical investigation plan (CIP) is the document required by MDR Article 62(4) that specifies the rationale, objectives, design, methodology, monitoring, statistical considerations, organisation, and conduct of a clinical investigation of a medical device. Its required content is set out in Annex XV Chapter II of Regulation (EU) 2017/745, and its operational form follows the harmonised Good Clinical Practice standard EN ISO 14155:2020+A11:2024.

Is the CIP the same as a study protocol? In practical use, yes. The CIP is the MDR-specific term for the document that Good Clinical Practice traditions call the study protocol. The content required by Annex XV Chapter II aligns closely with the protocol content expected by EN ISO 14155:2020+A11:2024, and most sponsors produce a single document that serves as both.

Who writes the CIP? The sponsor is responsible for the CIP under MDR Article 72. In practice the drafting is done by the sponsor's clinical team, often with input from biostatisticians, the principal investigator, and regulatory affairs. For small sponsors, external clinical research organisations or experienced regulatory partners frequently co-draft. Authorship is a team exercise, but accountability under MDR sits with the sponsor.

When does the CIP need to be finalised? Before the application to the competent authority under MDR Article 70 and before submission to the ethics committee. Both reviewers assess the CIP as the authoritative plan of the investigation. No investigation may begin until both authorisations are in place, and both are granted against the CIP as submitted.

Can the CIP be changed after authorisation? Yes, but changes that are substantial require notification and, in many cases, authorisation under MDR Article 75 before they take effect. Non-substantial changes are logged through the sponsor's version control. The CIP itself must contain the procedure the sponsor will follow to distinguish substantial from non-substantial amendments and to manage version control.

What is the difference between the CIP and the investigator's brochure? The CIP specifies how the investigation will be run. The investigator's brochure compiles the clinical and non-clinical information about the device that is relevant to the study, so that investigators understand the device they are using. Both documents are required by Annex XV — Chapter II specifies the CIP contents; other parts of Annex XV specify the investigator's brochure and the documentation of the device under investigation.

How long should a CIP be? There is no page count in Annex XV Chapter II. A well-written CIP for a focused startup investigation is typically in the range of thirty to fifty pages, with a separate statistical analysis plan and appendices. CIPs that run far beyond this are almost always padded. CIPs that run far short of this are almost always missing Annex XV content and will come back from review.

Sources

  1. Regulation (EU) 2017/745 of the European Parliament and of the Council of 5 April 2017 on medical devices, Article 62 (general requirements regarding clinical investigations), Article 62(4) (requirement for a clinical investigation plan), Article 63 (informed consent), Article 70 (application for clinical investigations), Article 72 (sponsor obligations), Article 75 (substantial modifications), Article 77 (end, temporary halt, or early termination reporting), Article 80 (adverse event recording and reporting), Annex XV Chapter II (clinical investigation plan content). Official Journal L 117, 5.5.2017.
  2. EN ISO 14155:2020+A11:2024 — Clinical investigation of medical devices for human subjects — Good clinical practice.
  3. EN ISO 14971:2019+A11:2021 — Medical devices — Application of risk management to medical devices.

This post sits in the Clinical Evaluation & Clinical Investigations cluster of the Subtract to Ship: MDR blog. Authored by Felix Lenhard and Tibor Zechmeister. If you are drafting a CIP under MDR Article 62(4) and Annex XV Chapter II and want a second pair of expert eyes before the ethics committee sees it, Zechmeister Strategic Solutions reviews CIPs for exactly that moment — after the design is set, before the submission is locked.