Hospitals do not buy because your device is CE marked. They buy because a clinical champion wants it, a value analysis committee has cleared it, procurement has scored it against MEAT criteria, and finance has confirmed the budget line. CE marking is the ticket to enter the room. Everything that happens inside the room is a separate game, and most founders underestimate it by eighteen months.

By Tibor Zechmeister and Felix Lenhard.

TL;DR

  • Under MDR Article 5, a device may only be placed on the EU market after CE marking, but CE marking creates no obligation on any hospital to purchase it.
  • Hospital procurement typically moves through clinical need identification, value analysis, tender or framework agreement, and contract award against MEAT criteria (Most Economically Advantageous Tender).
  • Value analysis committees scrutinise clinical evidence, total cost of ownership, training burden, and integration with existing workflows, not the marketing deck.
  • Your intended purpose statement under Article 2(12) ends up quoted in tender documents word-for-word, so procurement-aware wording matters before you lock the CE file.
  • A startup that plans for procurement during regulatory design ships faster into hospitals than one that treats procurement as a downstream sales problem.

Why this matters

A founder Felix coached last year got CE marked in fourteen months on a Class IIa imaging add-on. Clean technical file, competent notified body, first-try approval. They expected to be selling into three German university hospitals within a quarter. Fifteen months later, two of those hospitals were still in the value analysis stage and the third had pushed them into a public tender where they were competing on price against a legacy incumbent whose device was clinically worse but ten years embedded in the workflow.

This is the normal outcome. CE marking and hospital purchasing are two entirely different processes run by different people with different incentives on different timelines. Founders who treat the CE mark as the finish line and the purchase order as a formality are the ones who burn their Series A runway in the waiting room.

The good news: most of the procurement pain is predictable, and a lot of it can be pre-empted by decisions you make while the device is still in design.

What MDR actually says

MDR Article 5(1) is blunt: "A device may be placed on the market or put into service only if it complies with this Regulation when duly supplied and properly installed, maintained and used in accordance with its intended purpose." The Regulation governs whether you are allowed to sell. It does not govern whether anyone must buy.

The intended purpose itself is defined in Article 2(12): "intended purpose means the use for which a device is intended according to the data supplied by the manufacturer on the label, in the instructions for use or in promotional or sales materials or statements and as specified by the manufacturer in the clinical evaluation". That sentence ends up in your tender responses, in your value dossier, and in the questions a hospital clinical engineer asks before the purchase order is cut. If your intended purpose is narrow, you cannot sell to adjacent indications. If it is sloppy, procurement will mark you down. If it contradicts your marketing, you have created a promotional material conformance problem under Article 7.

Everything else about how a hospital buys sits outside the MDR. It lives in EU public procurement law (Directive 2014/24/EU for public contracting authorities), national reimbursement rules, hospital bylaws, and the internal politics of value analysis committees. None of that reduces the relevance of your MDR file, because every serious procurement process will ask for it, read it, and cross-check the claims.

A worked example

Imagine a Class IIa wearable monitoring device for post-surgical ward patients. CE marked in Q1 2026. First hospital target: a 900-bed university hospital in Bavaria.

Month 1: A cardiologist the founders met at a conference champions the device internally. She submits a New Product Request to the hospital's value analysis committee.

Month 2-4: The value analysis committee asks for the CE certificate, the declaration of conformity, the summary of the clinical evaluation, biocompatibility evidence for skin contact, EMC test reports, a cybersecurity dossier, a cleaning and reprocessing protocol, training plan, total cost of ownership over three years including consumables, and an integration spec for the hospital's patient data management system. The founders have six of those ten artefacts. The other four take eight weeks to produce.

Month 5: Value analysis committee approves clinical fit but flags that the contract value exceeds the threshold for direct award, which triggers an EU tender under Directive 2014/24/EU because the hospital is a public body.

Month 6-11: Tender is drafted, published with a minimum six-week response window, evaluated against MEAT criteria. MEAT (Most Economically Advantageous Tender) weights price against quality criteria chosen by the hospital: clinical performance (25%), total cost of ownership (25%), service level (15%), training and implementation (15%), cybersecurity and data protection (10%), environmental footprint (10%). A competitor enters the tender with a legacy device. The startup wins on clinical performance and loses on service level because they have no local service engineer within four hours of the site.

Month 12: Contract award, framework agreement signed, first rollout unit delivered. Month 14: Second unit. Month 18: The budget line is up for renegotiation because the hospital's finance office reclassified the line item.

That is eighteen months from CE mark to meaningful revenue, with no single villain in the story. Every step was procurement doing its job correctly.

The Subtract to Ship playbook

Here is how you pull months out of that timeline without cutting corners on Article 5 compliance.

Write your intended purpose with procurement in mind. Before you lock it into the Article 2(12) clinical evaluation, show the draft wording to two clinical champions and one hospital procurement contact. Ask them whether the wording matches how they would describe the use case in a New Product Request. If it does not, rewrite it while the CE file is still open. Narrow intended purposes are cheaper to certify but expensive to sell against; broad intended purposes are the opposite. You are balancing a regulatory cost against a commercial one, and the balance point is specific to your beachhead.

Build the procurement dossier in parallel with the technical file. The ten artefacts the value analysis committee will ask for overlap massively with your Annex II technical documentation. If you know that in advance, you produce each artefact once in a form that serves both readers. Biocompatibility evidence, cybersecurity documentation, training materials, service plans, and the summary of clinical evaluation are all dual-purpose. Total cost of ownership, integration specifications, and environmental footprint are procurement-only and need to be built separately. Build them while the CE file is in notified body review, not after.

Identify public versus private hospitals early. EU public procurement thresholds trigger tender obligations above roughly EUR 215,000 for supply contracts to central public authorities and around EUR 443,000 for sub-central authorities (thresholds as of 2024, updated biennially). Below those thresholds, a hospital has direct award flexibility. Above them, a tender is mandatory. A private clinic or a public hospital buying under threshold is a ten-week sale. A public hospital buying over threshold is a nine-month process with a six-week minimum publication window. Sequence your beachhead accordingly.

Have a MEAT answer ready for every criterion. Do not wait for the tender document to arrive. Write your own standard MEAT response template now: clinical performance, total cost of ownership, service level, training, cybersecurity, environmental footprint. Populate it with evidence. When the tender lands, you are answering it, not researching it.

Solve the service problem before you sell. The single most common reason European MedTech startups lose hospital tenders on quality criteria is service level. A four-hour on-site response commitment is worth more in MEAT scoring than most founders realise. If you cannot offer it directly, partner with an established distributor who can. This is a commercial decision with a regulatory footprint: distributors trigger Article 14 obligations, and you need to document the chain of responsibility in your QMS under EN ISO 13485:2016+A11:2021.

Anchor on a clinical champion, but never on only one. A champion who leaves, retires, or changes departments kills the deal. Two champions in different specialties at the same hospital is the minimum durable configuration.

Reality Check

  1. Have you shown your draft intended purpose to a hospital procurement contact before locking the Article 2(12) wording in the clinical evaluation?
  2. Can you produce a total cost of ownership model for your device over three years, including consumables, service, training, and disposal?
  3. Do you know whether your first five target hospitals are public, private, or mixed status under EU procurement law?
  4. Do you know the procurement threshold in euros that triggers a mandatory tender in each of your target countries?
  5. Do you have a written MEAT response template with evidence populated for clinical performance, TCO, service, training, cybersecurity, and environmental footprint?
  6. Do you have a service-level commitment you can honour for every target site, or a distributor partner who can?
  7. Do you have at least two clinical champions per target hospital, or are you single-threaded on one relationship?
  8. Is your summary of clinical evaluation written in language a value analysis committee can read without a translator?

Frequently Asked Questions

Does a CE mark obligate any hospital to buy my device? No. MDR Article 5 governs whether you may place a device on the market. Purchasing decisions are made by hospitals under their own rules and, for public contracting authorities, under EU public procurement law. A CE mark is necessary but not sufficient.

What is MEAT and who decides the weightings? MEAT stands for Most Economically Advantageous Tender, the default award principle under EU public procurement Directive 2014/24/EU. The contracting authority (usually the hospital or a regional purchasing group) sets the weightings between price and quality criteria before the tender is published. You cannot negotiate the weightings, only respond to them.

What is a value analysis committee? A cross-functional hospital body, typically including clinicians, nursing leadership, procurement, infection control, clinical engineering, and sometimes finance, that evaluates whether a new product should be added to the hospital's formulary. It sits upstream of the tender process and is often the real decision point.

Can I skip tenders by targeting private hospitals first? Often yes for the first commercial traction, but private hospital networks in Europe increasingly run their own group purchasing processes that resemble public tenders. Treat the distinction as a scheduling tool, not an exemption.

How much does my MDR technical file help in procurement? A lot, if the summary of clinical evaluation and intended purpose are written to be readable by non-regulatory clinicians. If they are written for the notified body only, procurement will ask you to translate everything, which delays the deal by weeks.

Should I lower my intended purpose scope to make CE marking cheaper? Only if the narrower scope still sells. A narrow intended purpose that cannot be stretched later locks you out of adjacent indications without a new conformity assessment. That is a commercial decision dressed as a regulatory one.

Sources

  1. Regulation (EU) 2017/745 on medical devices, consolidated text. Article 5 (placing on the market), Article 2(12) (intended purpose), Article 7 (claims).
  2. Directive 2014/24/EU on public procurement, MEAT principle and threshold framework (thresholds updated biennially by Commission Delegated Regulation).
  3. EN ISO 13485:2016+A11:2021, clause 7.2 on customer-related processes.