A clinical investigation plan has to justify its design

A clinical investigation plan reads, in a lot of files, like a research protocol with a regulatory cover page — objectives, procedures, statistics, signatures. ISO 14155:2020's Annex A asks for something more specific than that. A CIP under the standard doesn't just describe what the investigation will do; in its early substantive sections it has to justify why this design, on this device, at this point in its development, is the right one — and the justification has to trace back to a document the plan doesn't itself generate: the device's own clinical evaluation.

Annex A runs eighteen sections, and the justification comes early

ISO 14155:2020's Annex A is normative — eighteen numbered sections a CIP has to address, not a template a sponsor can trim to taste. General administrative information opens the annex, but the document's own sequence puts the scientific and rationale justification right behind it: Annex A.3 requires the plan's justification for undertaking the investigation to rest on the conclusions of the device's clinical evaluation, informed by an evaluation of relevant preclinical testing and any prior clinical investigations of the device. A CIP that leads with objectives and endpoints before establishing why the investigation is needed at all has the annex's own sequence backward.

The investigational device section isn't the technical file, borrowed

Annex A.4 asks for information about the investigational device itself — its description, its intended purpose, the manufacturer, and enough technical and functional detail for an investigator and ethics committee to understand what they're evaluating. It's tempting to satisfy this section by attaching the device's technical documentation wholesale, but a CIP's device section has a narrower job than a technical file's: it has to give a reviewer what they need to assess this specific investigation, not everything a notified body will eventually review under MDR Annex II. A CIP padded with technical-file content isn't more thorough for it; it's harder for an ethics committee to find the study-specific risk the section is actually there to surface.

The risk analysis section leans on the device's risk file; it doesn't rebuild it

Annex A also requires a risk analysis specific to the investigation — the risks the study itself introduces, on top of whatever risk the device already carries in ordinary use. That's a narrower document than the full risk management file this blog has already covered, built to ISO 14971's own per-hazard structure. A CIP that tries to reproduce that entire risk management file inside the plan is doing work Annex A never asked for; what it needs is the subset of that risk picture the investigation changes — new procedures, an unapproved use, a monitoring burden the marketed device doesn't carry — with the rest incorporated by reference rather than copied in.

The justification has a real audience, and it isn't only the competent authority

MDCG 2024-3, the Commission's 2024 guidance on CIP content under the MDR, reinforces the same emphasis Annex A already carries: the plan has to give both the ethics committee and the competent authority reviewing the application enough of the design's own rationale to assess it on its merits, not just enough procedural detail to run it. MDCG 2024-3 speaks to MDR investigations specifically, but the underlying discipline — justify the design before describing it — runs through ISO 14155 wherever the standard applies, including studies that never touch an EU file at all.

The plan cross-references evidence it doesn't restate

None of Annex A's eighteen sections asks the CIP to reproduce the device's own safety and performance data in full; that's the job of the investigator's brochure this blog covers in a companion note. What the CIP has to do instead is point to that evidence and use it in its own risk analysis — which is exactly where a plan that treats the IB as a separate, unrelated document tends to go wrong. A CIP whose risk-benefit reasoning doesn't visibly draw on the IB it's meant to sit beside is missing the connection Annex A assumes the two documents will make together.

Where this meets the file

A CIP outline that treats Annex A.3's justification as the section to write first, not the section to backfill once the protocol is drafted, is previewed in the launch catalog, alongside the investigator's brochure and the informed consent records the same investigation depends on. If your program sequences this differently, the shelf takes that correction directly.