A registry is cheaper per patient than a clinical investigation, and slower to a definitive answer. What decides the budget is not the number of patients. It is the number of countries, the number of variables you insist on collecting, and how long you need to follow each patient. Get those three right and the rest is arithmetic.
The people who build your evidence have sat on the other side of the table.
Charline PetitdemangeProject Delivery Lead, France and United KingdomStudy start-up and close-out
Dr Mark Da CostaChief Operating Officer and Head of CardiovascularFormer TÜV SÜD Team Leader
Sebastien Meier PiantanidaChief Data OfficerData management and EDC
A registry is cheaper per patient than a clinical investigation, and slower to a definitive answer. What decides the budget is not the number of patients. It is the number of countries, the number of variables you insist on collecting, and how long you need to follow each patient. Get those three right and the rest is arithmetic.
Leading manufacturers rely on Eclevar for registry-based PMCF and real-world evidence. Read all the client success stories.
A registry budget is read by an internal finance team that wants one number, and by a regulatory reviewer who wants to know the evidence will still be there at the end. Those two reflexes pull in opposite directions, and the budgets that break are the ones written for the first reader only.
We scope every program to the minimum defensible evidence a notified body will accept, then show what each design choice adds. That is a different exercise from quoting a cost per patient, and it is the one that survives both readings.
In an interventional study, cost follows visits. Every protocol visit carries a site payment, a monitoring cost and a data management cost, so the budget scales with the visit schedule. A registry under the EU MDR has no protocol visits. Care happens as it would have happened anyway, and the data are recorded around it.
That removes the single largest cost driver of a trial and replaces it with three others: setting the thing up across several countries, keeping sites entering data over years, and cleaning data that were not collected for research in the first place. The practical consequence is a budget that is front loaded and long tailed. Sponsors who budget a registry like a trial usually underfund the setup and the last two years, which are exactly the parts that decide whether the dataset is usable.
This page is about registry cost. The cost of an interventional study in Europe is a different question, covered on clinical trial cost in Europe and on what your trial will actually cost.
Which is why a cost per patient quoted without the design behind it is close to meaningless.
Protocol and data model, case report form build, electronic data capture configuration and validation, statistical analysis plan, and the documentation each country will ask for. Paid once, and it is where a registry is most often underfunded.
Ethics submissions, data protection filings and the institutional route in each country. Cost scales with the number of countries, not with the number of patients.
Contract negotiation, training and system access, per site. Adding a site is a fixed cost regardless of how many patients it ends up contributing, which is why a long site list is expensive before it is useful.
Usually modest compared with a trial, because there is no protocol visit to compensate. It is paid for years, which is what makes it add up.
Data management, query resolution, quality checks, site contact and project management. The recurring line, running for the full duration. It is the one that gets cut first and hurts most, as set out on registry data quality and monitoring.
Statistical analysis, the report itself, and whatever feeds the PMCF evaluation report or the clinical evaluation.
From our experience scoping these programs. The first three decide more than everything below them combined.
A registry that follows patients for several years costs several times one that follows them once, for the same number of patients. Duration multiplies the running line rather than adding to it.
Each country adds a regulatory route, a language, a contract template and a set of local rules. Two countries and six countries are not the same project.
Every additional field is a field someone has to enter, query and clean, for every patient, for the whole duration. Data models grow during design and nobody ever removes anything.
Reusing an existing registry or existing records removes collection cost and adds access negotiation and data quality work in its place.
An added imaging timepoint or a questionnaire on a fixed schedule changes the regime under Article 74 of Regulation (EU) 2017/745, and with it the cost.
A fixed cost per site at activation, then a running contact cost. More sites is not automatically faster, and it is always more expensive. Selection method on registry feasibility and site selection.
A validated instrument on a schedule adds licensing, translation, collection and chasing. An independent adjudication committee or a core lab is a real line, and the one sponsors forget until a notified body asks for it.
A registry is normally monitored far more lightly than a trial, but the level has to be decided and justified rather than left implicit. Joining the registry to another data source extends follow-up cheaply and adds legal, technical and custodian cost up front. Some of it can be reduced through decentralized methods.
Ask anyone who has run one of these programs which line went past the estimate, and the answer is almost always the same.
Consistently higher than what was budgeted at the outset. Across several sites, a gap of around ten thousand euros on this line alone is enough to be felt against the total budget. It is the first line we pressure test in a scenario.
Routinely underestimated. We have had to go back to a sponsor mid-study for money that was simply not accounted for in the original budget: travel costs, costs tied to a study extension, and similar items that only surface once the study is running.
It is often cheaper, in the end, to pay sites somewhat more to keep them engaged than to economize on indemnification and end up extending the study by six months. A six-month extension costs far more than the indemnification gap it was meant to save.
Evidence question, endpoints, data model, case report form, EDC build and validation, statistical analysis plan. Nothing downstream can start before the endpoints are fixed, which is why hesitation here is expensive later.
National routes, run in parallel across countries. The phase a sponsor controls least, and the one most often assumed away in a plan.
In parallel with the phase above where institutions allow it, sequential where they do not. Contracting is administrative rather than scientific, and it is regularly the longest single item.
Bounded by the real procedure volume at the sites, not by protocol acceptance. It cannot be accelerated by pushing harder, only by adding sites or extending the window.
The long tail. This is where a registry earns what it was built for, and where funding attention usually drifts away.
Fix the endpoints once and do not reopen them. Start the contracting route before the data model is finished. Everything else is second order.
Site activation is the biggest source of delay at launch, and one thing decides how badly it hits you: whether you have the right contact from day one.
Principal investigators are always busy, and their natural network is other physicians. Those conversations matter, they are where a site's genuine interest gets tested, but they do not move the administrative start-up forward.
What actually determines speed is how quickly you reach the study coordinator. Everything administrative runs through that person, and every week spent looking for them is a week added to the critical path.
Even with the right contact, the file usually still goes through a hospital committee. It is common to wait several weeks just to get that meeting scheduled. Contract negotiation follows, then finding a date everyone can attend for the site initiation visit. Set-up almost always takes longer than the launch meeting assumed.
Reach the administrative and coordination contacts as early as the process allows, which means asking for them during feasibility rather than at activation. And schedule the site initiation visit as far in advance as the institution permits. The rest of the activation sequence is on initiation of clinical studies and investigations.
You pay for access, data preparation and analysis. You do not pay for site activation or years of data entry. You give up control of the data model and inherit whatever completeness the registry already has. Cheap, fast, and it answers only the questions the existing variables allow. Often the shortest route to real-world evidence.
You pay for everything, and you get exactly the variables your endpoints need, with device traceability designed in. Expensive, slow to start, and the only route when the variable you need was never recorded anywhere.
A middle path. The existing registry provides the population and the benchmark, a defined add-on provides what it does not carry. Watch the Article 74 line while designing the add-on, because that is where a cheap module quietly becomes a clinical investigation. Route conditions on PMCF registry CRO.
The question that settles this is whether your endpoints exist in someone else's dataset. If they do, the cost argument is decided for you. If they do not, no saving on the other lines will compensate for a registry that cannot reach its endpoint.
Not the first year with an intention to extend. Registries that are refunded annually lose sites in the gaps, and those sites are rarely recovered.
Setup, per country, per site, per patient, per year. It makes the effect of adding a country visible before someone adds one.
The number of variables is a budget decision disguised as a scientific one, and it is almost never treated as such.
Regulatory and contracting. It is the phase where the sponsor has least influence and where the schedule slips first.
That single paragraph prevents the most expensive conversation there is: the one that happens after the analysis.
Fixed and variable lines separated, so the effect of each design choice is visible before it is made.
Built on the actual regulatory and contracting route in each country you are considering, not on a generic Gantt chart.
Existing registry, sponsor owned, or a linked design, costed side by side against the same evidence question.
Setup, EDC, data management, site contact, monitoring at the level the design justifies, statistics and reporting.
What the chosen budget will and will not be able to conclude. We would rather have that argument at the start.
Budget and schedule sit between feasibility and delivery, inside the wider medical device CRO offer.
Whitepapers, client voices and publications produced by our teams and our partners (BSI, TÜV SÜD, RegenLab).
There is no honest single answer, and any figure quoted without the design behind it is a sales number. The total is driven by follow-up duration, number of countries, number of variables and whether the data already exist. Two registries with the same number of patients can differ by an order of magnitude on those four dimensions alone. What we can give quickly is a costed scenario for your specific question.
Per patient, almost always, because there are no protocol visits to fund. Per answer, not necessarily. If the registry cannot reach your endpoint, the cheaper option produced nothing. Compare the two on cost per usable answer, not cost per patient.
Duration, because it multiplies the recurring lines rather than adding to them. The second is the number of countries, which multiplies the regulatory and contracting work while adding no patients on its own. In day-to-day practice, the two lines that most often exceed the estimate are site indemnification and pass-through costs.
Yes, and it is often sensible, on one condition: the data model and the endpoint definitions are designed for the full ambition from the start. Expanding a registry is routine. Retrofitting variables into one already running is expensive and leaves you with two incompatible periods of data.
Plan for interim analyses as well as the final one. A registry that reports only at the end delivers nothing usable for years, which makes it hard to defend internally and hard to feed into a clinical evaluation that has to be updated in the meantime.
It removes the collection cost and replaces it with access negotiation, data preparation and analysis. It is genuinely cheaper. It is not free, and the timeline to access is the part most often underestimated. Talk to our registry team about which route fits your question.
There is no honest single answer, and any figure quoted without the design behind it is a sales number. The total is driven by follow-up duration, number of countries, number of variables and whether the data already exist. Two registries with the same number of patients can differ by an order of magnitude on those four dimensions alone. What we can give quickly is a costed scenario for your specific question.
Per patient, almost always, because there are no protocol visits to fund. Per answer, not necessarily. If the registry cannot reach your endpoint, the cheaper option produced nothing. Compare the two on cost per usable answer, not cost per patient.
Duration, because it multiplies the recurring lines rather than adding to them. The second is the number of countries, which multiplies the regulatory and contracting work while adding no patients on its own. In day-to-day practice, the two lines that most often exceed the estimate are site indemnification and pass-through costs.
Yes, and it is often sensible, on one condition: the data model and the endpoint definitions are designed for the full ambition from the start. Expanding a registry is routine. Retrofitting variables into one already running is expensive and leaves you with two incompatible periods of data.
Plan for interim analyses as well as the final one. A registry that reports only at the end delivers nothing usable for years, which makes it hard to defend internally and hard to feed into a clinical evaluation that has to be updated in the meantime.
It removes the collection cost and replaces it with access negotiation, data preparation and analysis. It is genuinely cheaper. It is not free, and the timeline to access is the part most often underestimated. Talk to our registry team about which route fits your question.
Whitepapers and publications produced by our teams with our notified body partners.
Written with Notified Body BSI: a practical reading of the clinical evidence expectations under EU MDR 2017/745, the same evidence your file has to support.
The services and topics connected to this page.
Send us the device, the claim and the deadline. You get a written answer within 24 hours.
Your documents are reviewed confidentially. An NDA can be signed before we receive any technical or clinical information.
