UK RWE can support EU market access, but only when clinical, economic and payer evidence is translated for local HTA decisions.
The UK remains one of the strongest environments in the world for generating real-world evidence. The NHS is large, clinically credible, data-rich, and operationally constrained. It is not a laboratory version of healthcare. It is healthcare under pressure.
That is why UK evidence matters.
If an AI system, diagnostic, medical device, digital therapeutic, biotechnology product or pharmaceutical intervention can demonstrate value inside the NHS, that evidence should not be seen as merely local. It may be highly relevant to France, Germany, the Netherlands, Spain, Italy, the Nordics and wider European markets.
But there is a trap.
UK real-world evidence is not a European reimbursement passport.
It can travel scientifically. It can support European HTA and payer conversations. It can reduce uncertainty. It can strengthen clinical credibility, publications, registry strategy, investor confidence and early reference-centre adoption. But it rarely transfers unchanged.
It must be translated.
That distinction is now becoming one of the most important questions in market access: not simply can we generate evidence?, but can we generate evidence that survives another country’s decision problem?
The UK is an evidence engine, not a reimbursement passport
The NHS is valuable because it tests technologies in real clinical pathways. It shows what happens when a product meets referral pressure, staff shortages, IT friction, procurement constraints, coding ambiguity and everyday clinical behaviour.
That is especially important for technologies whose value is not captured by a classic pharmaceutical trial design.
For an AI imaging system, NHS evidence may show whether the technology improves detection, reduces reporting time, avoids unnecessary follow-up, or supports triage.
For a diagnostic, it may show whether better detection actually changes treatment decisions.
For a medical device, it may show procedure time, theatre burden, length of stay, complications, revision avoidance and patient-reported outcomes.
For a digital therapeutic, it may show engagement, adherence, symptom change, escalation avoidance and service impact.
For biotech and pharma, it may support disease burden, treatment sequencing, subgroup outcomes, external comparators, natural history and post-launch effectiveness.
NICE’s Real-World Evidence Framework is important here because it recognises the growing role of real-world data in evidence generation for health technology assessment. NICE says the framework is intended to improve the quality of real-world evidence used to inform its guidance.
https://www.nice.org.uk/corporate/ecd9
(NICE)
But NICE also makes the deeper point indirectly: real-world evidence is not valuable because it is “real world”. It is valuable when it is designed well enough to answer a decision-maker’s question.
That is the first lesson for companies.
Evidence is not automatically decision-grade. It has to be built that way.
Generalisability is not transportability
The central scientific issue is transportability.
Generalisability asks whether findings apply beyond the study sample to the wider population from which that sample was drawn. Transportability asks a harder question: can the findings be applied to a different population, setting or health system?
That is the UK-to-Europe question.
A UK study may be generalisable to NHS patients. But can it be transported to France, Germany, the Netherlands or Spain?
Only if the differences are understood.
Are the patients similar?
Is disease severity similar?
Is the comparator the same?
Is the pathway the same?
Are clinicians using the technology in the same way?
Are the outcomes measured and valued similarly?
Is the budget impact sitting with the same payer, hospital or department?
The causal-inference literature gives this question a proper scientific language. Degtiar and Rose’s review of generalisability and transportability describes how causal findings may be moved from a study population to a target population, and why external validity bias matters when the target population differs from the study population.
https://www.annualreviews.org/content/journals/10.1146/annurev-statistics-042522-103837
(Annual Reviews)
This matters commercially because every European payer is, in effect, asking a transportability question:
Why should evidence generated in the UK apply to our patients, our clinicians, our pathway, our costs and our reimbursement system?
HTA transferability is a recognised problem
This is not just an Odelle observation. It is a recognised academic and HTA problem.
Heupink and colleagues published a scoping review on the transferability of health technology assessments. They reviewed tools, methods and approaches used to transfer HTA information between settings, noting that transferring and adapting existing HTAs may reduce duplication but requires structured consideration of local context.
https://pubmed.ncbi.nlm.nih.gov/36321421/
https://www.cambridge.org/core/journals/international-journal-of-technology-assessment-in-health-care/article/considerations-for-transferability-of-health-technology-assessments-a-scoping-review-of-tools-methods-and-practices/D92FBBB3730E8BC42A12963E11D5D9AB
(PubMed)
A later systematic review by Ahmadnezhad and colleagues also examined tools and approaches for evaluating the transferability of HTA information.
https://www.ijhpm.com/article_4642_85beb369aabead9a8dfaa1a43bafa795.pdf
(IJHPM)
The message is clear: transferability is not a soft commercial concern. It is a methodological issue.
The weak question is:
Can we reuse our UK evidence in Europe?
The better question is:
Which parts of our UK evidence are transferable, which parts require adaptation, and which parts must be rebuilt locally?
What travels well
Some evidence travels well.
Clinical performance often travels. If an AI system improves detection, if a diagnostic has strong sensitivity and specificity, if a medical device reduces complications, or if a medicine improves patient outcomes, that signal may be relevant internationally.
Safety evidence can travel, especially when adverse events, contraindications, failure modes and usability risks are well documented.
Patient-reported outcomes can travel when validated instruments are used and the target patient group is comparable.
Implementation evidence can travel if the study explains the conditions of use: training, staffing, clinical workflow, IT integration, interoperability, governance and adoption barriers.
This is particularly important for medical devices, diagnostics, AI and digital medical technologies. The OECD’s 2025 report on digital medical devices highlights the need for appropriate assessment methods, real-world data, transparency, adaptive HTA and international collaboration.
https://www.oecd.org/en/publications/towards-identifying-good-practices-in-the-assessment-of-digital-medical-devices_b485ee1f-en.html
https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/04/towards-identifying-good-practices-in-the-assessment-of-digital-medical-devices_e35198b0/b485ee1f-en.pdf
(NICE)
That is exactly the point: evidence travels better when it describes not only the result, but the context that produced the result.
A UK study that says “the technology reduced referrals” is useful.
A UK study that says which referrals, in which patients, from which clinicians, under which pathway rules, using which workflow, with which downstream consequences is much more useful.
What does not travel well
The economic case rarely travels intact.
UK costs are not French costs. NHS tariffs are not German DRGs. UK staff costs are not Dutch hospital costs. UK outpatient avoidance is not automatically valuable in the same way to a French hospital, German sickness fund or Dutch insurer.
This is where companies often go wrong.
They take a UK budget impact model, convert pounds into euros, and call it European evidence.
That is not localisation. That is currency conversion.
A European payer needs to know whether the product changes their own pathway, their own costs, their own coding, their own budget impact and their own incentives.
Economic transferability is a recognised issue in HTA. Weise and colleagues reviewed methodological guidance on assessing transferability in systematic reviews of health economic evaluations and found variation in how HTA organisations recommend assessing transferability.
https://pubmed.ncbi.nlm.nih.gov/35184733/
https://link.springer.com/article/10.1186/s12874-022-01536-6
(PubMed)
The implication is simple:
Clinical effect may travel.
Resource use may partially travel.
Costs, tariffs, coding and payment logic must be rebuilt locally.
EU HTA changes the landscape, but not the local payer problem
The EU HTA Regulation is important, but it is often misunderstood.
Joint Clinical Assessments are intended to support Member States’ national HTA processes by providing a scientific analysis of clinical evidence on the relative effects of health technologies.
https://health.ec.europa.eu/health-technology-assessment/implementation-regulation-health-technology-assessment/joint-clinical-assessments_en
(Public Health)
That is not the same as EU-wide reimbursement.
The Joint Clinical Assessment is concerned with relative clinical effectiveness and safety. It does not decide national pricing, reimbursement, affordability, budget impact or implementation. The Office of Health Economics summarises the point clearly: economic evaluation, pricing and reimbursement decisions remain the responsibility of individual Member States.
https://www.ohe.org/insights/one-europe-one-assessment-unpacking-the-european-joint-clinical-assessment/
(ohe.org)
Recent analysis of the EU HTA Regulation makes the same point: pricing, reimbursement and budget management remain under national jurisdiction.
https://pmc.ncbi.nlm.nih.gov/articles/PMC13078104/
(PMC)
So the new European reality is this:
There may be more shared clinical assessment.
There will still be national reimbursement decisions.
That means a UK evidence package may help with the clinical story, but France, Germany, the Netherlands, Italy, Spain and the Nordics will still ask local questions.
Who pays?
Which code?
Which budget?
Which comparator?
Which patient subgroup?
Which hospital or payer benefits?
Is the benefit clinically meaningful in this system?
Does the technology create cost, save cost, or shift cost?
The companies that succeed will not merely generate evidence. They will generate evidence that can be interpreted at both EU clinical-assessment level and national reimbursement level.
RWE is welcome, but not automatically trusted
Europe is not anti-real-world evidence. NICE, EMA, HTA bodies, payers and regulators are all becoming more sophisticated about real-world data.
EMA has developed a real-world evidence capability through DARWIN EU, intended to provide timely and reliable evidence on the use, safety and effectiveness of human medicines from real-world healthcare databases across the EU.
https://www.ema.europa.eu/en/about-us/how-we-work/data-regulation-big-data-other-sources/real-world-evidence
(European Medicines Agency (EMA))
EMA’s third annual report on RWE to support EU regulatory decision-making describes the progress of regulator-led studies between February 2024 and February 2025.
https://www.ema.europa.eu/en/documents/report/real-world-evidence-framework-support-eu-regulatory-decision-making-3rd-report-experience-gained-regulator-led-studies-february-2024-february-2025_en.pdf
(European Medicines Agency (EMA))
But RWE is not accepted simply because it is observational, large or collected in routine care.
Murphy and colleagues argue that RWE can support HTA and payer decision-making when it is relevant, transparent, methodologically appropriate and trusted by stakeholders.
https://pmc.ncbi.nlm.nih.gov/articles/PMC12018852/
(PMC)
Sarri and colleagues describe the current RWE guidance environment as a “maze”, showing that many regulatory and HTA organisations have published guidance, but the landscape remains complex and variable.
https://pubmed.ncbi.nlm.nih.gov/39132748/
https://becarispublishing.com/doi/10.57264/cer-2024-0061
(PubMed)
This is important for industry.
A registry is not automatically evidence.
A pilot is not automatically evidence.
A hospital case series is not automatically evidence.
A dashboard is not automatically evidence.
A patient app dataset is not automatically evidence.
Evidence becomes useful when it reduces uncertainty for a decision-maker.
The Odelle Evidence Translation Test
Before taking UK RWE into Europe, companies should test it across six domains.
1. Clinical portability
Does the clinical claim survive movement from the NHS to the target European system?
For example, if a device reduces complications in an NHS study, is the baseline complication rate similar in Germany or France? If an AI system improves detection, does it work across different scanners, data formats, populations and clinical thresholds? If a digital therapeutic improves symptoms, were validated measures used and are those measures accepted locally?
The evidence should define not only that the technology works, but for whom, compared with what, and under which conditions.
2. Population portability
Are the UK patients similar to the patients in the target country?
This matters for almost every technology, but especially for AI, diagnostics, oncology, rare disease, digital health and precision medicine.
Transportability can fail if the receiving country has different patient demographics, disease severity, comorbidities, ethnicity, referral timing, diagnostic access, treatment sequencing or baseline risk.
For AI, this is not only an HTA issue. It is a performance and safety issue. An algorithm evaluated in one population may not perform the same way in another if the data distribution changes.
This is why transportability and external validity methods matter. The Degtiar and Rose review is useful here because it explains why internal validity is not enough if the study population differs from the population in which decisions will be made.
https://arxiv.org/abs/2102.11904
(arXiv)
3. Comparator portability
Was the UK comparator the same as the European comparator?
This is one of the most common evidence gaps.
A technology may look valuable against NHS standard care but less valuable against a different pathway in Germany, France or the Netherlands. Conversely, it may look modest in the NHS but highly valuable in a country where the current pathway is slower, more expensive or more procedure-heavy.
The comparator is not a technical detail. It is the heart of the reimbursement argument.
For pharmaceuticals and biotech, this becomes especially important in the context of EU Joint Clinical Assessment, because companies may face multiple PICOs across Member States. The clinical evidence package may need to speak to more than one comparator and more than one national clinical practice pattern.
4. Workflow portability
Does the workflow evidence map to the local care pathway?
A technology may save radiologist time in one country but not another. A diagnostic may accelerate treatment in a system where treatment capacity exists, but not in a system where the bottleneck is downstream access. A device may shorten length of stay in one system, but if another country already has very short stays, the value argument must shift to complications, readmissions or procedure avoidance.
Workflow evidence travels only when the original study describes the workflow in enough detail.
For AI, diagnostics and devices, the question is often not “does it work?” but “where does it sit?”
Before referral?
At triage?
Inside diagnostics?
During surgery?
At discharge?
In rehabilitation?
In chronic disease monitoring?
In primary care?
In specialist care?
A receiving country cannot translate UK evidence if the UK pathway is poorly described.
5. Economic portability
Can the resource-use findings be re-costed locally?
This is where the UK evidence package must be deliberately designed.
The company should collect resource use in natural units, not only in UK costs:
staff minutes,
number of visits,
admissions avoided,
bed-days,
imaging episodes,
theatre time,
consumables,
adverse events,
reoperations,
follow-up intensity,
patient travel burden,
caregiver burden.
Those units can then be costed in France, Germany, the Netherlands or Sweden.
If the UK study only reports “NHS savings”, the evidence is much less portable.
A French payer does not need an NHS saving. A German hospital does not need an NHS tariff. A Dutch insurer does not need an English budget-impact assumption.
They need local economics built from transferable resource-use evidence.
6. Reimbursement portability
Is there a plausible route to payment in the target country?
This question must be asked before the evidence programme starts.
For Germany, the company may need to consider DRG absorption, OPS coding, NUB, ZE, EBM, GOÄ, hybrid DRG, selective contracts or private reimbursement.
For France, the route may involve LPPR, CCAM, RIHN, PECAN, Article 51, forfait innovation, liste en sus or hospital procurement.
For the Netherlands, it may involve DBC logic, insurer contracting, appropriate care, conditional reimbursement or hospital procurement.
For Spain and Italy, the national evidence package may need to be made usable by regional decision-makers.
For the Nordics, registry linkage, procurement, regional implementation and population-level productivity may be decisive.
Evidence must meet the payment route. Otherwise, the evidence may be scientifically impressive but commercially stranded.
Example: UK AI imaging evidence into Europe
Imagine an AI imaging system evaluated in the NHS.
The UK study shows improved detection, reduced reporting time, fewer unnecessary follow-ups and safe use in clinical workflow.
That is a strong foundation.
But in Germany, the question may be whether the AI is funded through hospital procurement, whether it affects inpatient or outpatient billing, whether radiology workflow savings matter to the hospital, and whether there is a coding or selective-contract argument.
In France, the question may be whether the AI sits inside an existing clinical act, whether it could support a digital pathway, whether HAS would view the evidence as clinically meaningful, and whether procurement or innovation funding is the realistic first route.
In the Netherlands, the argument may need to focus on appropriate care, insurer priorities, reduced low-value imaging and substitution of unnecessary specialist activity.
Same evidence. Different decision problem.
Example: UK device evidence into Europe
Consider a medical device evaluated in NHS practice.
The study shows safety, procedure time, length of stay, complications, patient-reported outcomes and avoidance of a more invasive procedure.
That evidence may be valuable in every European country.
But the reimbursement question changes.
In Germany, the device may be absorbed into an existing DRG unless an additional payment route is created.
In France, it may need device listing or procedure recognition.
In the Netherlands, the argument may need to show substitution away from higher-cost care.
In the Nordics, the strongest case may be registry-based outcomes and regional productivity.
The device evidence travels scientifically. The payment logic does not.
Example: UK digital therapeutic evidence into Europe
Digital therapeutics show why evidence translation is becoming urgent.
European assessment routes for digital therapeutics remain heterogeneous. Arcà and colleagues compared HTA criteria for digital therapeutics across Germany, the UK and France and concluded that assessment and access frameworks remain fragmented and heterogeneous.
https://journals.sagepub.com/doi/10.1177/20552076241308704
https://pubmed.ncbi.nlm.nih.gov/39845519/
(Sage Journals)
That means UK evidence on engagement, symptom improvement, adherence or quality of life may be useful, but each country may ask for something different.
Germany may ask whether evidence meets DiGA expectations.
France may ask whether the product fits PECAN, LPPR digital reimbursement, Article 51 or hospital pathways.
Belgium may ask whether it fits mHealthBelgium logic.
The Netherlands may ask whether it reduces avoidable care and fits insurer priorities.
The evidence may be the same. The threshold for action is not.
Example: UK pharmaceutical or biotech evidence into Europe
For pharmaceuticals and biotech products, the UK-to-Europe question is different but equally important.
A UK real-world dataset may show treatment sequencing, persistence, discontinuation, adverse events, unmet need, external control performance, natural history or subgroup outcomes. This can be extremely useful, particularly in oncology, rare disease, advanced therapies and areas where RCT evidence is immature or single-arm.
But the same RWE may be interpreted differently in Europe.
A national HTA body may ask whether the UK standard of care reflects its own comparator. A European clinical assessment may ask whether the evidence addresses the relevant PICO. A national payer may accept the clinical signal but reject the economic extrapolation. A country with different treatment sequencing may see the UK data as informative but not decisive.
This is why RWE should be designed around the decision problem from the beginning, not retrofitted after the study has finished.
How to design UK evidence so it can be used in Europe
The answer is not to run a UK study and then hope.
The answer is to design the study for transferability from the beginning.
A company should do the following.
First, define the European target countries before the UK evidence programme begins. France, Germany, the Netherlands and Sweden may all want different economic and pathway evidence.
Second, define the target decision-maker. Is the evidence for HTA, a national payer, a hospital procurement committee, a regional health authority, an insurer, a clinical society, or an investor?
Third, collect resource use in natural units, not only in UK financial values.
Fourth, document the comparator in detail. “Standard care” is not enough.
Fifth, use validated patient-reported outcome measures where possible.
Sixth, capture workflow and implementation burden: staff time, training, IT integration, failed use, clinician adherence and patient acceptability.
Seventh, pre-specify subgroup analyses. European payers often care less about the average effect and more about the first reimbursable subgroup.
Eighth, plan local expert validation before submission. A small panel of French, German or Dutch clinicians, coders and payers can identify whether the UK evidence is believable locally.
Ninth, design the registry or post-market evidence plan so that European sites can join later.
Tenth, separate the evidence dossier into portable and local modules.
The evidence package should have three layers
The first layer is the scientific core.
This includes clinical performance, safety, outcomes, patient selection, subgroup effects and uncertainty.
The second layer is the implementation layer.
This includes workflow, usability, training, governance, integration, clinician behaviour and patient acceptability.
The third layer is the market translation layer.
This includes country-specific comparators, costs, tariffs, codes, budget impact, payer incentives and route to reimbursement.
The scientific core can often travel.
The implementation layer can travel if it is well described.
The market translation layer must be rebuilt locally.
That is the practical meaning of evidence transferability.
The Odelle position
The companies that win in Europe will not necessarily be the companies with the largest evidence package.
They will be the companies with the most interpretable evidence package.
A large UK dataset can still fail if it does not answer the French, German or Dutch question.
A modest UK study can become powerful if it captures the right clinical endpoints, the right resource-use units, the right workflow detail, the right patient subgroup and the right comparator.
The commercial skill is not simply producing evidence.
It is knowing what evidence will still be useful after it crosses a border.
Conclusion: evidence must survive another country’s question
UK real-world evidence can be a powerful foundation for European market access.
But the NHS is not Europe. NICE is not HAS. An NHS trust is not a German sickness fund. A UK tariff is not a French act or a German DRG. A UK pathway is not a Dutch insurer pathway.
The mistake is not using UK evidence.
The mistake is assuming that the evidence means the same thing everywhere.
The winners will be companies that design UK evidence so that it can be translated: clinically, economically, operationally and commercially.
They will ask, before the study begins:
Which claims will travel?
Which claims are fragile?
Which costs must be rebuilt?
Which comparator matters locally?
Which payer objection will this evidence answer?
What minimum local evidence will complete the case?
The next frontier in market access is not simply more real-world evidence.
It is more translatable evidence.
Or put another way:
A UK study may prove that a technology works. Evidence translation proves why another country should pay for it.
Reference list
- NICE. NICE real-world evidence framework. 2022.
https://www.nice.org.uk/corporate/ecd9
https://www.nice.org.uk/corporate/ecd9/resources/nice-realworld-evidence-framework-pdf-1124020816837 - Heupink LF, Peacocke EF, Sæterdal I, Chola L, Frønsdal K. Considerations for transferability of health technology assessments: a scoping review of tools, methods, and practices. International Journal of Technology Assessment in Health Care. 2022.
https://pubmed.ncbi.nlm.nih.gov/36321421/
https://www.cambridge.org/core/journals/international-journal-of-technology-assessment-in-health-care/article/considerations-for-transferability-of-health-technology-assessments-a-scoping-review-of-tools-methods-and-practices/D92FBBB3730E8BC42A12963E11D5D9AB - Ahmadnezhad E et al. Systematic Review of Tools and Approaches for Evaluating Transferability of Health Technology Assessment Information. 2024.
https://www.ijhpm.com/article_4642_85beb369aabead9a8dfaa1a43bafa795.pdf - Degtiar I, Rose S. A Review of Generalizability and Transportability. Annual Review of Statistics and Its Application. 2023.
https://www.annualreviews.org/content/journals/10.1146/annurev-statistics-042522-103837
https://arxiv.org/abs/2102.11904 - European Commission. Joint Clinical Assessments under the EU HTA Regulation.
https://health.ec.europa.eu/health-technology-assessment/implementation-regulation-health-technology-assessment/joint-clinical-assessments_en - Office of Health Economics. One Europe, one assessment? Unpacking the European Joint Clinical Assessment. 2026.
https://www.ohe.org/insights/one-europe-one-assessment-unpacking-the-european-joint-clinical-assessment/ - Meregaglia M et al. Implementing the EU HTA regulation and joint clinical assessments. 2026.
https://pmc.ncbi.nlm.nih.gov/articles/PMC13078104/ - EMA. Real-world evidence.
https://www.ema.europa.eu/en/about-us/how-we-work/data-regulation-big-data-other-sources/real-world-evidence - EMA. Real-world evidence framework to support EU regulatory decision-making: third report on experience gained with regulator-led studies, February 2024 to February 2025. 2025.
https://www.ema.europa.eu/en/documents/report/real-world-evidence-framework-support-eu-regulatory-decision-making-3rd-report-experience-gained-regulator-led-studies-february-2024-february-2025_en.pdf - Murphy LA et al. Real-world evidence to support health technology assessment and payer decision-making: is it now or never? 2025.
https://pmc.ncbi.nlm.nih.gov/articles/PMC12018852/ - Sarri G, Hernandez LG. The maze of real-world evidence frameworks: from a desert to a jungle. Journal of Comparative Effectiveness Research. 2024.
https://pubmed.ncbi.nlm.nih.gov/39132748/
https://becarispublishing.com/doi/10.57264/cer-2024-0061 - Weise A, Büchter RB, Pieper D, Mathes T. Assessing transferability in systematic reviews of health economic evaluations: a review of methodological guidance. BMC Medical Research Methodology. 2022.
https://pubmed.ncbi.nlm.nih.gov/35184733/
https://link.springer.com/article/10.1186/s12874-022-01536-6 - OECD. Towards identifying good practices in the assessment of digital medical devices. 2025.
https://www.oecd.org/en/publications/towards-identifying-good-practices-in-the-assessment-of-digital-medical-devices_b485ee1f-en.html
https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/04/towards-identifying-good-practices-in-the-assessment-of-digital-medical-devices_e35198b0/b485ee1f-en.pdf - Arcà E et al. Comparison of health technology assessments for digital therapeutics in Germany, the United Kingdom and France. Digital Health. 2025.
https://journals.sagepub.com/doi/10.1177/20552076241308704
https://pubmed.ncbi.nlm.nih.gov/39845519/ - Tunaru F et al. Use of real-world data and real-world evidence in NICE oncology technology appraisals. 2025.
https://pmc.ncbi.nlm.nih.gov/articles/PMC12481278/