A specialist medical device CRO has to hold six things together in one accountable structure: clinical operations run to the applicable edition of ISO 14155, an EU MDR evidence strategy that connects the investigation to the clinical evaluation, therapeutic depth in your device area, accountable clinical execution across European countries, clinical data management and biostatistics, and the medical writing that produces the Clinical Investigation Report, the CER and the PMCF documentation. The selection question is not whether a CRO can run a study. It is whether it designs the study around the evidence you will eventually have to defend.
Most selection guides frame this as a competence contest, and it is not one. Large pharmaceutical CROs run complex international trials to a standard most device sponsors would be glad to have. The difference is structural: a drug program and a device program produce different evidence, on different timelines, for different readers.
Drug development typically progresses through phased clinical programs around a comparatively stable medicinal product. Medical-device development is often more iterative, with design evolution, operator effects and procedure-related variables that can materially affect the evidence strategy. A device is also evaluated against a clinical evaluation that has to be maintained for as long as it is on the market. That is not a harder problem or an easier one. It is a differently shaped one, and the shape decides what an experienced team does by reflex.
| Evaluation axis | What a device program demands |
|---|---|
| Governing framework | EU MDR 2017/745 and ISO 14155, not ICH-GCP by default |
| Development pattern | Iterative. Design changes during the program have to be handled as design changes, not as protocol noise |
| Procedure variability | Implantation and use technique vary between operators and sites, and the analysis has to anticipate it |
| Learning curve | Operator experience is a study variable, not a nuisance to be ignored |
| Device accountability | Serial-level traceability, device deficiency reporting, returned-device handling |
| Endpoint logic | Endpoints trace back to the intended purpose and the claims, and forward into the clinical evaluation |
| Post-market obligation | PMCF is planned before the pre-market study closes, not started after a certificate is issued |
| Reader of the evidence | A Notified Body assessor reading a technical file, not only a journal reviewer |
A pharmaceutical CRO with a genuine device operation may satisfy every row. The table is a list of things to test, not a conclusion about any provider.
Read that table as a list of things to test a CRO on, rather than as a verdict on any category of provider. A pharmaceutical CRO with a genuine device operation may handle every row. A specialist medical device CRO that has only ever run post-market surveys may not. The label is not the evidence. The answers are.
A major source of downstream rework is a disconnect between an early design decision and the evidence the clinical evaluation ultimately needs.
Figure 1 · The medical device evidence chain
Node 10 returns to node 04: the clinical evaluation is maintained across the market life of the device rather than closed at certification.
Every link is a place where a program can be sound in isolation and still fail as a whole. A perfectly executed study measuring the wrong thing is still a perfectly executed study. The practical test of a CRO is whether it reasons across that chain in the first conversation, or only from the point where it starts billing.
Each of these is designed to be hard to answer in marketing language. Hold any CRO to them, including this one.
"Life sciences experience" is not an answer. Ask what proportion of current work is device work, which device classes, which therapeutic areas, the split between pre-market and post-market studies, and whether the team has run programs on implantable devices. A specialist should be able to answer with concrete examples, defined scope and, where confidentiality permits, named programs or references. Ask specifically about Class IIb and Class III work if that is your device, because the operational demands do not transfer down from Class IIa experience.
The claim is easy; the evidence is not. ISO 14155 is not a management-system certification standard analogous to ISO 13485, so there is no certificate to ask for. Assess implementation instead, across six things: SOPs and templates, the monitoring approach, device accountability, risk management, safety escalation, and data-quality governance. Ask what changed in each when the edition changed, and what the gap analysis found for studies already running at that point. Eclevar sets out its own answers on the ISO 14155:2026 clinical-investigation CRO services page, and you should hold it to them as you would any other bidder.
Ask them to walk from your Clinical Investigation Plan to your endpoints, from your endpoints to your clinical evaluation, and from your clinical evaluation to your PMCF plan and your claims, out loud, using your device. A CRO that runs sites will describe visits and monitoring. A CRO that produces regulatory evidence will describe the argument the file has to make and what has to be measured to support it.
Ask whether medical, regulatory, biostatistical and operational input all happen before the protocol is finalized, or whether the statistician sees the design after it is fixed. The most expensive errors in device programs are design errors that were operationally cheap to prevent and structurally expensive to correct. Ask who holds the pen on the synopsis and whether that person stays with the program.
Therapeutic depth is not a preference. It changes which sites can credibly run the study, which endpoints and follow-up windows are accepted in that field, what the comparator situation looks like, and what a monitor needs to understand to spot a problem. A cardiovascular structural heart program, an orthopedic implant program, a neuromodulation program and a wound care program make different demands on all four. Ask for programs in your area, with the named clinician who led the design.
Europe is not a single operational pathway, and treating it as one can create avoidable start-up delays. Each country has its own competent authority arrangements, its own ethics process and sequencing, its own site contracting norms, and its own language requirements for the file and for participant-facing documents. In several countries the applicable route also depends on the device class and on whether the device carries a CE mark for the purpose under investigation. Ask which countries the CRO has actually submitted in, who prepares the submissions, whether the clinical research associates are employed or subcontracted, and where those people physically sit.
Ask, item by item: who employs the CRAs, who owns monitoring, who owns project management, who owns data management, who owns biostatistics, who owns medical writing, and who remains accountable when two of those disagree. Subcontracting is not disqualifying. Undisclosed subcontracting, or a structure where no single party can be held to the integrity of the data, is a different matter. Some functions, core laboratory and imaging in particular, are routinely and correctly contracted to specialists. What matters is that you know which, and that governance sits in one place.
A device program does not end at CE marking. Ask how the pre-market design anticipates the post-market clinical follow-up that follows it, whether the same team can carry the evidence into a registry or a real-world data collection, and how the clinical evaluation report gets updated from what the studies produce. A registry or another PMCF method is not automatically a clinical investigation under ISO 14155, and a CRO that blurs those categories will blur them in your plan too.
Ask about the EDC and its validation documentation, the data management plan and the data validation plan, the monitoring model and how risk-based decisions are justified and recorded, the statistical analysis plan and who writes it, coding conventions where they apply, audit trails, and how the database is prepared for lock. Ask to see the validation file rather than the feature list. Clinical data management is where quiet problems accumulate for months before they surface.
Ask for programs in your area with named devices and described scope, the clinician who led them, and references you can verify. Ask for evidence of experience supporting documentation through Notified Body review cycles, including appropriately anonymized or sponsor-approved examples where available. Logos are not proof. Independent recognition carries weight only when it is externally judged and described at the scope it was actually awarded. A complete absence of any published client program is itself informative.
It changes the applicable regulatory route, the ethics pathway and its sequencing, the participant population you can reach, the investigators and sites available to you, the contracting workload, the language of the file, and in some markets the evidence later needed for reimbursement.
| Country | Decision variables |
|---|---|
| France | The applicable route depends on the investigation category, the device class, its invasiveness, its CE-marking status and the purpose of the study, and investigations intended to demonstrate conformity do not follow the same route as post-market studies on CE-marked devices used within their intended purpose. Investigations are subject to scientific and ethical review, with ANSM authorization applying according to the case and CPP ethics review alongside it. Determine the applicable route before the dossier is built. Where French reimbursement is commercially relevant, it is worth knowing during country selection that the evidence a reimbursement assessment will later look for is not identical to the evidence a conformity assessment looks for. Eclevar sets out how it runs programs there on the medical device CRO in France page |
| Germany | A favourable opinion from the competent ethics committee, and either approval from or notification to the competent higher federal authority, are required depending on the case. The procedure is sequential: the application goes to the ethics committee first, and the favourable opinion must accompany the application to the federal authority. The competent ethics committee is the one responsible for the investigator under federal state law. Applications are submitted through the national system while the corresponding EUDAMED functionality is unavailable |
| United Kingdom | Great Britain operates outside EU MDR, under the UK Medical Devices Regulations 2002. Where an application is required, the MHRA must be notified at least 60 days before the intended start, and a research ethics committee opinion from a committee appointed by the Research Ethics Service is required, sought before or in parallel with the notification. Not every study requires an application: studies of validly UKCA or CE marked devices used within their marked indication are exempt in Great Britain, which covers many PMCF studies. Northern Ireland follows EU rules. UK-generated clinical data may be relevant to an EU clinical evaluation where scientifically and regulatorily applicable, but the UK regulatory route is separate and should not be described as an EU MDR clinical-investigation route |
| Spain | Clinical investigations run under Royal Decree 192/2023. The national competent authority authorizes those investigations that require competent authority authorization. The ethics opinion is issued by a CEIm accredited by the relevant autonomous community, and for multicenter studies a single CEIm opinion applies nationally and is binding. The agreement of the management of each participating site is also required, which is a start-up workstream in its own right. Investigations of CE-marked devices used outside their intended purpose fall under the same provisions as conformity investigations |
| Italy | Applications for investigations of devices not CE marked for the intended purpose go to the national ministry under the national framework adapting the MDR. The application is accompanied by a favourable opinion from a territorial or national ethics committee, which is valid nationally and binding on every site in the application. For higher-risk devices, meaning Class III or invasive Class IIa and IIb, a ministerial authorization decision is required, normally issued within 45 days of validation, extendable by 20 days where expert consultation is needed. A sponsor may file with a conditional opinion or with the opinion request pending |
| Sweden | Clinical investigations are submitted to the Swedish Medical Products Agency and undergo ethical review by the Swedish Ethical Review Authority through the coordinated national procedure. MPA authorization or notification depends on the device status and investigation route. Where human biological samples fall under the Swedish Biobank Act, the applicable biobank workstream should be initiated in parallel. |
| Other European markets | Other European markets should be assessed individually against the device, participant population, investigator availability, applicable submission route and operational requirements |
Decision considerations by country, not a ranking and not a coverage map. Requirements vary by device and therapeutic area, and there is no single uniform European route.
The instinct to add countries to accelerate recruitment is often wrong at the margin. Each additional country adds a submission, a contract set, a translation workload and a monitoring footprint, and the enrollment it contributes may arrive after the enrollment you already had. Add countries where the participants, the operators or the evidence requirement genuinely are.
This is a real choice with a real answer that depends on the program, and any provider who tells you their category always wins is selling.
The program is large and genuinely global, the geographic footprint extends well beyond Europe, procurement requires a global master services agreement or an established vendor relationship, the therapeutic infrastructure needed spans several areas at once, or the sponsor has an internal clinical organization capable of directing the work and needs execution capacity rather than design partnership.
The device is Class IIb or Class III, particularly an implantable, when the program is early feasibility or first-in-human and the design is still moving, when therapeutic expertise materially changes the protocol, when EU MDR evidence integration is the central problem rather than trial logistics, when post-market and registry work has to connect back to the regulatory strategy, when senior people need to stay reachable, and when the sponsor's team is small enough that handoffs between departments become their problem to manage.
Most sponsors already know which description fits them. The failure mode is not choosing the wrong category. It is choosing a provider whose actual structure does not match the category it presents.
No honest guide quotes a price for a European device program, because the range between a single-country post-market data collection and a multi-country Class III pre-market investigation is wider than any average could usefully describe.
What actually drives the number: the number of countries and the number of sites; the participant count and the recruitment period; the monitoring model and visit frequency; device complexity and whether procedures require proctoring; endpoint complexity, including any imaging, core laboratory or independent adjudication; electronic clinical outcome assessment and patient-reported outcomes; the EDC build and the data management workload; medical monitoring; statistical work; safety reporting infrastructure; trial master file management; the number and type of regulatory and ethics submissions; and the duration of post-market follow-up.
The lowest bid is frequently the one that omitted the work. The comparison that matters is not price against price, but assumption against assumption.
Send us your protocol, synopsis or evidence gap. We will review the program and identify the operational and regulatory workstreams it requires.
Score each criterion from 1 to 5, multiply by the weight, and compare candidates on the total rather than on the impression left by the last meeting.
| Selection criterion | Weight | The question that tests it | Red flag |
|---|---|---|---|
| Device specialization | 15% | What share of your current work is device work, and in which classes? | Device experience described only as part of a broader life sciences portfolio |
| Therapeutic expertise | 15% | Which programs in my area, and which clinician led the design? | Therapeutic depth claimed at company level, never at program level |
| EU MDR evidence integration | 15% | Walk me from my intended purpose to my endpoints to my CER | Fluency about study conduct, silence about the clinical evaluation |
| ISO 14155 implementation | 10% | What evidence of implementation can you show across SOPs, monitoring, device accountability, risk management, safety escalation and data governance? | Reference to the standard as a logo or a marketing line |
| European operations | 10% | Which countries have you submitted in, and who prepared those submissions? | Europe described as a single operational pathway |
| Site and recruitment capability | 10% | How was the recruitment forecast built, and on what data? | A forecast with no stated basis, or a guarantee |
| Data and biometrics | 10% | Show me the EDC validation documentation and the data management plan | Feature demonstration offered instead of validation evidence |
| Senior team access | 5% | Who works on this after signature, and for what proportion of their time? | Senior people present at the pitch and absent from the org chart |
| Budget transparency | 5% | Unit rates, assumptions and pass-through, separated | A single fixed number with no visible build |
| Relevant evidence and case studies | 5% | Named programs I can verify, and evidence of Notified Body review experience | Logo wall, no described program |
Total weight 100%.
A scoring instrument, not a ranking. Weights should be adjusted to the device: therapeutic expertise matters more for an implantable than for a post-market survey.
Eclevar is a medical device CRO. Not a pharmaceutical CRO with a device division, and not a regulatory consultancy that also books monitors.
Figure 5 · One accountable program governance
The structural choice behind the model is that the people who reason about regulatory evidence and the people who run the study sit inside the same program governance. Clinical strategy, protocol and endpoint design, feasibility and site selection, submissions, monitoring, data management, biostatistics and medical writing are planned together and reported together, so that a decision about endpoints is taken with knowledge of what it will cost to collect, and a decision about monitoring is taken with knowledge of what the file will have to defend.
The team includes people who previously assessed clinical evidence inside a Notified Body, alongside practicing clinicians in cardiovascular and in orthopedics and spine. That combination is the reason the evidence conversation and the operational conversation can happen at the same table rather than sequentially.
To date, Eclevar has delivered more than 30 device evidence programs covering more than 2,000 participants, with regulatory and ethics submission experience across six European jurisdictions. Therapeutic concentration is in cardiology and structural heart, orthopedics and spine, vascular devices, neuromodulation and advanced wound care.
In February 2026, Milo, Eclevar's clinical data platform, received the Platinum Award in the xShare open call organized by EUCROF and co-funded under Horizon Europe. The award recognized the submitted clinical research data use case. It is not a ranking of contract research organizations and should not be read as one.
A randomized post-market clinical follow-up program covering 160 participants across 14 sites in five countries, addressing both diabetic foot ulcer and venous leg ulcer indications, with data captured on the Milo platform. Eclevar designed and is managing the program, including the standardized wound assessment across countries, sites and assessors that makes the data comparable.
Evidence intended to support EU MDR clinical evaluation and PMCF activities. No results or regulatory outcomes are presented here.
A randomized crossover clinical investigation of an intermittent catheter, 72 participants planned across sites in three countries, registered as NCT05814211. Eclevar structured the program architecture and supported delivery, including the multi-country submission work and the home-use data collection design that a crossover comparison requires.
Description limited to Eclevar's contracted scope. No results or regulatory outcomes are presented here.
Further programs, including orthopedic and cardiovascular work, are described on the client success stories page.
Chief Operating Officer and Head of Cardiovascular, Senior Consultant Surgeon
Mark leads cardiovascular clinical strategy, combining 25 years of Consultant Cardiac Surgery experience with first-hand senior leadership Notified Body experience.
Chief Medical Officer, Orthopedics and Spine
Practicing NHS orthopedic surgeon with Notified Body clinical review experience in Class III orthopedic and spine implants, and associate principal investigator on randomized trials in robotic arthroplasty and personalized osteotomy.
Head of Medical Writing
Leads clinical evaluation reports, clinical investigation reports and Notified Body deficiency responses.
Head of Data Management and Biostatistics
Leads EDC design, data management planning and statistical analysis across the program portfolio.
The engagement team is confirmed for each program according to the device, therapeutic area, countries, study design and contracted scope. Meet the full leadership team.
A medical device CRO is a contract research organization that designs, runs and reports the clinical evidence a device manufacturer needs. For European medical-device investigations, ISO 14155 is the device-specific good clinical practice standard used alongside the applicable MDR, national and ethical requirements. ICH-GCP is primarily structured around medicinal-product clinical trials. Its scope typically runs from clinical strategy and protocol design through regulatory and ethics submissions, site selection, monitoring, data management and statistics, to the clinical investigation report and the post-market evidence that follows.
Test device specialization, therapeutic depth, demonstrated ISO 14155 practice, the ability to connect the investigation to the clinical evaluation, real country-by-country execution, what is subcontracted, data and biometrics capability, senior team access, budget transparency and verifiable program evidence. Weight the criteria that matter most for your device, score candidates against them, and treat any criterion answered in marketing language as unanswered.
Not as of the review date shown below. ISO 14155:2026 is the current international edition, published in March 2026 and replacing ISO 14155:2020 at the ISO level. For presumption of conformity under the MDR, the version cited in the Official Journal is EN ISO 14155:2020/A11:2024, added by Commission Implementing Decision (EU) 2026/193. Sponsors should document which edition and requirements are applicable to the investigation, taking account of ISO 14155:2026 as the current international edition, the harmonized EN standard cited under the MDR, applicable national requirements, and expectations associated with the relevant regulatory and conformity-assessment pathway. Note also that ISO 14155 does not stand alone: European, national and ethical requirements apply alongside it.
Not yet. Four EUDAMED modules became mandatory on 28 May 2026: actor registration, UDI and device registration, notified bodies and certificates, and market surveillance. Neither of the two remaining modules is among them, and they are not at the same stage: the vigilance and post-market surveillance module remains in development, while the clinical investigations and performance studies module is currently under analysis. Clinical investigation applications therefore continue through national competent authority and ethics routes. EUDAMED registration and device data readiness are operational regulatory requirements to incorporate into EU market access planning.
There is no meaningful average, because the range from a single-country post-market data collection to a multi-country Class III pre-market investigation is too wide for one to be useful. Cost is driven by countries and sites, participant numbers, recruitment period, monitoring model, device and endpoint complexity, imaging or adjudication requirements, data management scope and the duration of post-market follow-up. Compare bids on itemized assumptions and unit rates rather than on totals.
There is no country that is right regardless of device. The right question is which countries contain the participants, the operators and the evidence requirements your program needs, and what each adds in submission work, ethics sequencing, contracting and translation. In several countries the applicable route also depends on the device class and on whether the device is CE marked for the purpose under investigation, so the route determination comes before the country shortlist is fixed. Great Britain sits outside EU MDR under a separate framework.
It should be able to, and the connection is where the value sits. Post-market evidence is more defensible when it was anticipated during the pre-market design rather than reconstructed afterward. Note that a registry or another PMCF method is not automatically a clinical investigation under ISO 14155, and the two should not be described interchangeably.
Regulatory position reviewed: 10 August 2026. National routes and harmonized standard citations change; this page is not legal or regulatory advice.
Looking for a specialist medical device CRO in Europe? Bring your device, your target markets and the answers other CROs gave you to the ten questions above. We will tell you where the gaps are.
