A MedTech startup needs one evidence stack that simultaneously satisfies the notified body (MDR Article 61 and Annex XIV), post-market surveillance obligations (MDR Articles 83 to 86 and Annex III), HTA bodies, and payers. Building three separate evidence plans is the most expensive mistake a founder can make. Building one integrated plan is the difference between shipping and burning out.

By Tibor Zechmeister and Felix Lenhard.

TL;DR

  • MDR clinical evidence (Article 61, Annex XIV), post-market surveillance evidence (Articles 83 to 86, Annex III), HTA evidence, and payer evidence are four audiences with overlapping but distinct requirements.
  • The overlap is large. A well-designed clinical investigation can generate data that feeds all four audiences if you plan for it before locking the protocol.
  • The gaps typically sit in comparator data, patient-reported outcomes, and health-economic modelling — things the MDR does not require but HTA bodies and payers demand.
  • Subtract to Ship logic: design one study, not three. Cut evidence activities that serve no specific audience. Invest the savings in the study that does.
  • A real evidence strategy maps every piece of data to at least one audience, every audience to at least one data source, and every data source to a budget line.
  • The Reality Check below is the exercise to run before your next clinical investigation plan is finalised.

Why an evidence strategy is the most valuable document a MedTech startup owns

We sat in a meeting with a founder who had just burned through €1.2M on a clinical study. The CE mark was secured. The notified body was happy. Then the German reimbursement consultant looked at the dataset and said, quietly, that the trial had used a surrogate endpoint IQWiG would not accept, had no comparator arm, and had not collected any patient-reported outcomes. "You'll need to run a second study."

That founder did not have an evidence generation problem. The founder had an evidence strategy problem. Data was generated — plenty of it — but not against a plan that accounted for all the audiences the business needed to satisfy.

An evidence strategy is a one-page answer to four questions. What does the notified body need? What do post-market surveillance obligations need? What do HTA bodies need? What do payers and procurement need? And where do those needs overlap enough that one trial, one registry, or one dataset can serve more than one of them?

Get this right and a 12-person startup can ship with one clinical study. Get it wrong and the same startup ends up running two or three in sequence, each delayed by the previous.

What MDR actually says about evidence

MDR Article 61 requires clinical evaluation for all medical devices, structured around a clinical evaluation plan and leading to a clinical evaluation report that demonstrates conformity with the General Safety and Performance Requirements in Annex I. Annex XIV Part A sets out the procedural steps of clinical evaluation, including the identification of available clinical data, appraisal, analysis, and conclusion.

For most Class IIa and higher devices — and certainly for implantables and Class III — Article 61 expects clinical investigation data unless equivalence to another device can be robustly demonstrated (Annex XIV Part A point 3) or specific exemptions apply (Article 61(4) to (6), further clarified by MDCG 2023-7). Clinical investigations themselves fall under MDR Chapter VI (Articles 62 to 82), Annex XV, and Good Clinical Practice under EN ISO 14155:2020+A11:2024.

Evidence does not stop at the CE mark. MDR Articles 83 to 86 require a post-market surveillance system proportionate to risk class, with a PMS plan (Annex III), periodic safety update reports (PSURs) for Class IIa and above, and post-market clinical follow-up (PMCF) under Annex XIV Part B for most devices. MDCG 2025-10 provides the current guidance on PMS. PMS is not a reporting formality. It is a continuous evidence generation obligation.

Risk management (EN ISO 14971:2019+A11:2021) sits across all of this, feeding into Annex I GSPR 1 to 9 and requiring that clinical and post-market data flow back into risk assessment.

HTA evidence requirements are set outside the MDR, by Regulation (EU) 2021/2282 at EU level and by national HTA bodies. Payer requirements are set nationally, by statutory insurers, hospital procurement committees, and regional bodies. Neither group is bound by MDR definitions of "sufficient clinical evidence." They set their own bar.

The key insight is this: MDR defines the regulatory floor. HTA and payer requirements usually sit higher. If you design for the floor you will fail the ceiling. If you design for the ceiling you automatically clear the floor.

A worked example

A startup is developing a Class IIb closed-loop insulin delivery system. The team initially plans a single-arm safety and performance study, 120 patients, 6 months, glycaemic variability as the primary endpoint.

We build the evidence map.

Notified body needs (Article 61, Annex XIV): safety, performance, clinical benefit, benefit-risk. Single-arm data with glycaemic variability can work if the state of the art is well characterised. Probably sufficient for CE mark.

PMS needs (Articles 83 to 86, Annex III): ongoing safety signals, complaint data, PMCF under Annex XIV Part B. The single-arm trial provides a baseline; PMCF will extend it.

HTA needs: comparator data versus current standard of care (multiple daily injections or standard pumps), patient-relevant endpoints (HbA1c, severe hypoglycaemia events, quality of life, time in range), 12-month follow-up, health-economic modelling.

Payer needs: budget impact model, reimbursement dossier, real-world adherence data.

The original plan answers the first audience. It partially answers the second. It fails the third and fourth.

The redesigned plan: randomised controlled trial, 240 patients, 12-month follow-up, primary endpoint HbA1c (patient-relevant), secondary endpoints time in range and severe hypoglycaemia, pre-specified health-economic sub-study, validated quality-of-life instrument, structured data collection to feed PMCF from day one of market launch. Cost goes up by roughly 60%. Timeline slips by four months.

But the startup runs one trial. The CE mark, the PSUR, the PMCF start, the HTA dossier, and the payer submission all draw from the same dataset. Total cost is lower than Plan A plus a second study. Timeline to reimbursement is 18 months faster.

The founder's reaction was the one we hear every time we do this exercise properly: "This is obvious in hindsight. Why doesn't anyone teach it upfront?"

The Subtract to Ship playbook

Step 1 — List your four audiences explicitly. Write them on a whiteboard: Notified body. PMS obligations. HTA bodies. Payers. Under each, write the decisions they will make and the evidence they need to make those decisions.

Step 2 — Write the evidence matrix. Rows are evidence items (clinical investigation, literature review, bench data, PMCF, registry, patient-reported outcomes, health-economic model). Columns are the four audiences. Mark each cell with Required, Useful, or Not relevant. The pattern that emerges tells you where to invest.

Step 3 — Identify the overlap zone. Evidence items that serve three or four audiences get priority investment. Evidence items that serve only one audience get scrutiny: is it really needed, or is it habit?

Step 4 — Kill single-purpose evidence when you can. This is the subtraction. Bench tests no auditor will look at. Literature reviews that are not structured to feed the CER. Surveys that no HTA body will accept. Cut them.

Step 5 — Upgrade the core study to serve multiple audiences. Usually this means adding a comparator arm, patient-relevant endpoints, and a health-economic analysis plan to what would otherwise have been a minimum-MDR trial. The marginal cost of adding these to an existing trial is far lower than running a second trial later.

Step 6 — Plan PMCF as live evidence generation. Article 83 to 86 plus Annex XIV Part B PMCF should not be designed as post-launch compliance — they should be designed as a continuous evidence pipeline that feeds the next HTA submission, the next PSUR, and the next reimbursement review.

Step 7 — Budget the full stack. Clinical investigation, PMS system, PMCF plan, HTA submission, payer dossier. Put numbers against each line before your next funding round.

Step 8 — Write a one-page evidence strategy document and keep it updated. This is a living document. It should be reviewed at every major clinical milestone and at every funding round. The most useful thing we do with founders is sit them down in front of this page.

Reality Check

  1. Can you list every audience your clinical evidence will need to satisfy over the next five years?
  2. Have you mapped each evidence item to at least one specific audience and each audience to at least one specific evidence item?
  3. Is there a comparator in your current or planned clinical investigation, or only a single arm?
  4. Do your primary endpoints include at least one patient-relevant outcome, or only surrogate markers?
  5. Is your PMCF plan designed as evidence generation or only as compliance reporting?
  6. Do you have a health-economic analysis plan — even a rough one — before the clinical investigation plan is locked?
  7. Has anyone on your team read a recent HTA appraisal in your therapeutic area end-to-end?
  8. Can your current evidence stack fit on one page in a form a notified body auditor, a PMS reviewer, an HTA assessor, and a reimbursement consultant could each recognise?
  9. Is the total evidence budget — clinical, PMS, PMCF, HTA, payer — a line item in your next funding round?

Frequently Asked Questions

What is the difference between clinical evaluation and clinical investigation? Clinical evaluation under MDR Article 61 and Annex XIV Part A is the ongoing process of collecting, appraising, and analysing clinical data from all sources. Clinical investigation is one specific source of data: a controlled study performed under MDR Chapter VI (Articles 62 to 82) and Annex XV, in accordance with EN ISO 14155:2020+A11:2024. Clinical evaluation is broader than clinical investigation.

Is an RCT always necessary? No. An RCT is not a regulatory requirement under MDR, and for some devices literature-based evaluation or single-arm data is sufficient. But if HTA submissions and payer dossiers are part of your business model, a comparator is usually needed somewhere in your evidence stack, and an RCT is the simplest way to get one.

What is PMCF and do I have to do it? Post-market clinical follow-up is described in MDR Annex XIV Part B. For most devices — certainly implantables, Class IIb, and Class III, and commonly Class IIa — PMCF is expected. Exemptions can be justified (see MDCG 2023-7) but must be documented. A PMCF plan is required even where active follow-up activities are minimal.

Do I need patient-reported outcomes for MDR? MDR does not explicitly require patient-reported outcomes for CE mark. HTA bodies and payers increasingly do. If your business model depends on reimbursement, PROs belong in the evidence stack.

How early should we finalise the evidence strategy? Before the clinical investigation plan is locked. Once the CIP is signed, changing endpoints or adding a comparator becomes expensive and slow.

Can AI or real-world data replace a traditional clinical investigation? Real-world data and registries play a growing role, particularly in PMCF and HTA dossiers. They rarely replace a pivotal clinical investigation for CE mark but can reduce the size of one and extend its value over time.

Who writes the evidence strategy — regulatory, clinical, or commercial? All three. An evidence strategy written in isolation by any one function tends to miss at least one audience. The best evidence strategies we see are written in a room with regulatory, clinical, and commercial leads present and signed off by the founder.

Sources

  1. Regulation (EU) 2017/745 on medical devices, consolidated text. Articles 61, 62-82, 83-86; Annex III; Annex XIV Parts A and B; Annex XV.
  2. MDCG 2020-5 (April 2020) — Clinical evaluation equivalence.
  3. MDCG 2023-7 (December 2023) — Clinical investigation exemptions under Article 61(4) to (6).
  4. MDCG 2025-10 (December 2025) — Post-market surveillance.
  5. EN ISO 14155:2020+A11:2024 — Clinical investigation of medical devices for human subjects — Good clinical practice.
  6. EN ISO 14971:2019+A11:2021 — Medical devices — Application of risk management to medical devices.
  7. Regulation (EU) 2021/2282 on health technology assessment.