Investigator-Initiated Studies in MedTech: The Good, the Bad, and the Ugly

Investigator-initiated studies occupy a peculiar place in the medtech ecosystem. Industry loves them because someone else does the work. Clinicians love them because someone else pays for it. Regulators tolerate them because the alternative, no independent evidence at all, is worse. And compliance departments lie awake at night because of them, for reasons that will become apparent below.

The premise is simple. A physician has a clinical question, designs a study, and asks a manufacturer for support: money, free devices, sometimes statistical help. The manufacturer says yes or no. If yes, the physician runs the study. That is the whole model. The complications start, as usual, with the small print.

One Sentence That Needs Repeating

In an investigator-initiated study, the company is not the sponsor.

This sentence needs repeating because almost nobody believes it the first time. The investigator, or more precisely the investigator's institution, takes on the sponsor role, and with it the full set of sponsor obligations: the CIP, the regulatory submissions, safety reporting under MDR Article 80, insurance, monitoring, the clinical investigation report, the archive. ISO 14155 has a name for this dual creature, the sponsor-investigator, and the standard is explicit that combining both roles does not halve the obligations. It doubles the workload for one party.

The manufacturer, meanwhile, is a supporter. It writes a grant agreement, ships devices, perhaps provides a biostatistician for an afternoon, and otherwise stays out of the way. On paper this division of labour is crisp. In practice, the moment something goes wrong, everyone involved rediscovers a strong intuitive belief that the company with the deep pockets and the regulatory department is surely responsible for something. Untangling that belief contractually, before the first subject is enrolled, is most of the job.

How It Should Work: A Framework in Four Stages

A functioning IIS program is not a mailbox where proposals arrive and cheques leave. It is a process, and the process looks roughly like this.

1. Proposal

The investigator submits a structured proposal through a defined channel: a portal, a standard form, a dedicated mailbox. Not through the sales representative who happens to visit on Tuesdays. The proposal contains a scientific rationale, a draft design with endpoints, a realistic recruitment estimate, a budget, and a statement of what is being requested (funding, product, in-kind support, or a combination). If the proposal cannot articulate a hypothesis, it is not a study proposal. It is a wish.

2. Review

An internal review committee evaluates the proposal against pre-defined, documented criteria: scientific merit, feasibility, alignment with legitimate evidence needs, budget at fair market value, and the qualifications of the sponsor-investigator to actually carry the sponsor role. The committee is staffed by medical, clinical, and regulatory functions, with compliance in the room. Sales and marketing should not be on the committee, should not score proposals, and should not get a veto or a fast lane. This firewall is not bureaucratic decoration; it is the load-bearing wall of the entire model, and the MedTech Europe Code of Ethical Business Practice is unambiguous that research support must never be a reward for, or an inducement to, purchasing decisions.

3. Approval and Contracting

Approved proposals get a written agreement that does the unglamorous work: it names the institution as sponsor, enumerates the sponsor obligations so nobody can later claim surprise, defines the support (amounts, devices, milestones), ties payments to milestones rather than to enrolment enthusiasm, secures the company's access to safety information it needs for its own post-market surveillance, and settles publication rights. The investigator publishes, positive or negative; the company gets a courtesy review window for confidential information and intellectual property, not editorial control. If a draft agreement gives the company approval rights over the conclusions, stop and start again.

4. Conduct and Oversight

The study runs under the sponsor-investigator's responsibility, but supported does not mean abandoned. The company tracks milestones, receives safety information per the agreement, and retains the contractual right to pause or end support if the study drifts from GCP or from the agreed protocol. The investigator delivers a final report and, ideally, a publication. Both sides archive everything, because in this field the question is never whether someone will ask, only when.

The Good

There is a reason every serious manufacturer runs an IIS programme despite the headaches.

Independent evidence is more credible than company evidence, and everybody knows it. A peer-reviewed publication from an academic group carries a weight in clinical evaluations, in guideline discussions, and frankly in the operating theatre, that a manufacturer-sponsored study with identical results does not. Under the MDR, where post-market clinical follow-up is a standing obligation rather than a polite suggestion, well-conducted IIS data is a legitimate and cost-efficient contribution to the evidence base, provided its limitations are honestly characterized in the clinical evaluation.

IIS also answer questions companies would never fund directly. Niche indications, long-term outcomes, head-to-head comparisons that no marketing department would dare to commission, real-world use in populations excluded from pivotal trials. The clinical community asks better questions than product managers do, mostly because it has to live with the answers.

And yes, there is the relationship dimension. Supporting a clinician's research is a form of customer engagement, and pretending otherwise insults everyone's intelligence. The honest position is that engagement is a by-product, never the purpose, and that the review process exists precisely to keep the by-product from quietly becoming the purpose. Which brings us to the other sections.

The Bad

The structural weakness of the model is that it hands the most demanding role in clinical research, the sponsor role, to the party least equipped to perform it. A university surgical department is staffed for surgery, not for safety reporting timelines, monitoring plans, device accountability logs, and the statistical sections of Annex A. The standard does not care. ISO 14155 imposes the same sponsor obligations whether the sponsor is a multinational with a clinical operations department or a professor with a shared secretary and a PhD student who leaves in March.

The predictable results: under-powered designs, recruitment estimates from a parallel universe, missing data handled by not mentioning it, safety reporting that happens eventually, and final reports that arrive years late or, in a respectable fraction of cases, never. None of this is malice. It is resourcing. The 2020 edition of ISO 14155 raised the methodological bar, and the trajectory of the standard is clearly toward more statistical rigour, not less, which means the gap between what the standard requires and what an unsupported academic sponsor can deliver is widening, not closing.

A supporting company has limited tools here. It can require evidence of sponsor capability during review, fund monitoring or statistical support explicitly, and structure milestones so that non-performance has consequences. What it cannot do is quietly take over sponsor duties, because then the careful legal separation collapses and the company owns a study it never controlled in design.

The Ugly

The bad is incompetence. The ugly is intent.

The ugly version of an IIS is the seed study: a thinly designed protocol whose real endpoint is familiarizing twenty surgeons with a device, with funding routed through the research budget because the marketing budget would have raised questions. The tell-tale signs are reliable. The proposal arrives via the sales channel. The "research question" is indistinguishable from the product's launch messaging. The site selection maps suspiciously well onto the key account list. The budget contains line items that are honoraria wearing a lab coat.

This is not a victimless arrangement. It exposes the company to inducement findings under national anti-corruption law and the MedTech Europe Code, it pollutes the literature with studies designed never to find anything inconvenient, and it burns the credibility of the entire IIS category, including the legitimate studies. The related, slightly subtler pathology is selective support: funding only investigators whose prior publications point the right way, or letting support quietly lapse when interim signals disappoint. The funding decision is made before the data exist; the integrity test is whether the company's behaviour is identical when the results come in negative.

And negative results will come in. An honestly run IIS programme will, with statistical certainty, eventually fund a study that concludes the company's device is no better than the comparator, or worse. What happens next, whether the publication proceeds, whether the data enters the clinical evaluation, whether the investigator is ever funded again, is the single most informative fact about a company's IIS programme. Everything else is process documentation.

A Compact Summary

Dimension	The Good	The Bad	The Ugly
Evidence	Independent, credible, PMCF-relevant	Under-powered, late, unreported	Designed not to find anything
Sponsor role	Genuine investigator ownership	Obligations exceed academic resources	Company pulls strings without accountability
Funding	FMV, milestone-based, firewalled review	Optimistic budgets, weak controls	Marketing money in a research costume
Publication	Investigator publishes regardless of result	Results vanish into drawer	Selective support, selective silence

The Bottom Line

Investigator-initiated studies are probably not too bad, provided they are well managed, which is a sentence that does a lot of lifting. The model works when three conditions hold: the proposal review is genuinely firewalled from commercial influence, the sponsor-investigator is honestly assessed for the capacity to carry the sponsor role and supported where capacity is missing, and the company's behaviour is demonstrably indifferent to the direction of the results.

Where those conditions hold, IIS deliver something the industry cannot buy any other way: independent evidence with independent credibility. Where they do not, the programme is either a slow-motion compliance finding or a charitable donation to the global archive of unfinished studies. The difference is not luck. It is the process, written down, followed, and applied with equal enthusiasm to the proposals one likes and the proposals one's sales team likes.

Key References

1. ISO 14155:2020. Clinical investigation of medical devices for human subjects — Good clinical practice. International Organization for Standardization.

2. Regulation (EU) 2017/745 of the European Parliament and of the Council (EU MDR). Articles 62, 69, 80; Annex XV.

3. MDCG 2021-6. Questions and answers regarding clinical investigation under Regulation (EU) 2017/745.

4. MDCG 2020-13. Clinical evaluation assessment report template — PMCF context.

5. MedTech Europe Code of Ethical Business Practice.

6. Declaration of Helsinki. World Medical Association.

Dr. Marc Harms MH-Analytics