ºÚÁϳԹÏÍø

Ìý

Introduction

Ìý

Commercial clinical trials are central to the UK’s life sciences sector, critical for patient access to innovative treatments, and represent substantial economic activity and scientific advancement. In 2022, industry clinical trials contributed £7.4 billion to the UK economy, generated £1.2 billion revenue for the NHS, and supported a total of 65,000 jobs across the UK.¹ Clinical trials offer patients the potential to access innovative treatments that may not be available through standard care, which can be life changing for those with limited options. Therefore, the UK’s success in commercial clinical trials delivery directly impacts patient outcomes, economic prosperity and the UK’s position as a global leader in life sciences.

In 2022 industry clinical trials contributed:

Ìý

Ìý

Ìý

Ìý

This report examines challenges facing recruitment into UK commercial trials, the status of existing data-enabled clinical trial services, and issues with harnessing NHS data to support commercial trial delivery at scale. Based on extensive consultation with pharmaceutical companies, data service providers, research delivery networks and NHS trusts, the report makes the case for a new UK-wide data-enabled clinical trials function aligned to industry, delivery system and patient needs. A function designed to meet the rigorous requirements of commercial trials would, by its nature, serve the needs of non-commercial research, ensuring benefits across the entire UK clinical research ecosystem.

Ìý

The trial recruitment challenge

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Current recruitment methods are inefficient

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

NHS data – an underused resource

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Fragmented approach using NHS data to support trial recruitment

Ìý

Ìý

Creating a competitive UK-wide offering

Ìý

Clinical trials account for about 50 per cent of a pharmaceutical company’s research and development spend.¹⁹ International competition to host commercial trials is growing as the widespread benefits trials bring to the economy and health outcomes are increasingly recognised.²⁰ Consequently, many countries are adopting a national or unified approach to incentivise industry investment.²¹

The potential for NHS data to accelerate trial delivery has been realised by multiple organisations that offer data-enabled services to different parts of the trial ecosystem. These services have grown organically or are privately funded with different governance arrangements, meaning it is not possible or practical to combine or retrofit their operations. Despite the enormous potential for NHS data to transform commercial trial delivery, to date there has been no UK-wide strategic approach to discerning how NHS data could enhance commercial trial recruitment at scale, directly benefiting the NHS trial delivery infrastructure where sponsors place their trials.

Delivering trials that include participants representative of the patient population that will ultimately benefit from the medicine is key to developing robust evidence on the safety and efficacy of new medicines.¹⁸ Identifying representative populations through current recruitment methods creates substantial challenges. Using privacy-preserving eligibility searches of NHS records would not only accelerate the recruitment process but would support recruitment of a more representative cohort.

The £600 million investment in establishing an HDRS⁶ provides, for the first time, a realistic possibility of creating a compelling UK offering to expedite recruitment into phase II-IV commercial trials. Fundamental to achieving this ambition is understanding from a pharmaceutical industry perspective, where the greatest gains could be made and how UK national data assets could be seamlessly connected with the existing trials infrastructure to improve trial delivery.

The resulting recommendation from the ºÚÁϳԹÏÍø and member pharmaceutical companies is to establish a dedicated UK-wide data-enabled clinical trial function that protects patient confidentiality and uses comprehensive NHS data to provide robust feasibility assessments and targeted participant recruitment to expedite trial delivery at scale across the UK.

Ìý

Ìý

Ìý

Industry requirements

Ìý

The pharmaceutical industry operates within global decision-making frameworks in which UK teams must compete for trial allocations against other markets. Understanding what commercial sponsors need to justify placing trials in the UK, and how a UK-wide data-enabled clinical trial service could deliver these requirements, is fundamental to realising the UK’s competitive potential.

Drawing on direct engagement with the pharmaceutical industry and NHS trial infrastructure, this section defines the operating principles for UK-wide data-enabled clinical trials services that would support the key requirements for pharmaceutical industry trial delivery. Findings are based on a desk-based review of clinical trials performance data and academic literature on eligibility criteria; semi-structured interviews with 20 stakeholders spanning UK-wide data providers (for example, CPRD, NHS DigiTrials, NorthWest eHealth, SAIL Databank, Research Data Scotland), research delivery networks and NHS trusts; and two Task and Finish Group workshops with ºÚÁϳԹÏÍø members (see the Appendix for further detail).

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Operating principles

Ìý

Three themes consistently emerged as core operating principles for a data-enabled clinical trials function (see Figure 3).

• Predictability. Sponsors return to sites and markets they trust. Reliable delivery justifies ongoing investment in the UK and will help UK teams to secure larger trial allocations in the future. The service must consistently prove it can support sites across the UK to deliver what has been promised.
  
  
• Accuracy. Feasibility assessments must provide an accurate estimate of eligible populations that will translate into participants who enrol in a trial. Current screen failure rates can approach 90 per cent in some therapeutic areas due to inaccurate feasibility assessments and inability to target the right patients to invite for screening. According to ºÚÁϳԹÏÍø members, reducing screen failures to 50 per cent by more accurate feasibility assessments and participant identification would nearly halve screening costs and site burden. In addition, it would reduce patient time wasted in attending screening, only to be told they are ineligible for a trial, based on medical information already captured in their records.
  
  
• Speed. Global teams need feasibility responses within two weeks, or they allocate trials elsewhere. Rapid participant recruitment will increase the likelihood of UK sites successfully delivering their trial quotas within faster timeframes. Providing each site with a pre-screened list of pseudonymised potential participants during the regulatory approvals process would speed up subsequent identification of eligible participants at sites, enabling recruitment to begin as soon as regulatory approval has been received.

Ìý

Ìý

Ìý

Ìý

Ìý

Minimum datasets required

Ìý

According to ºÚÁϳԹÏÍø member companies consulted, information contained within structured GP data linked to Secondary Uses Service (SUS)ⁱ data in England or equivalent national secondary care datasets in the devolved nations, encompass approximately 60 per cent of selection criteria in most industry trial protocols, excluding oncology and rare diseases. GP records capture 90 per cent of patient contacts with the healthcare system,^14,15 and contain longitudinal data on chronic diseases and acute conditions, prescribing histories, laboratory test results, diagnoses and referrals. This information supports identification of patients who may be suitable for trials in many chronic conditions. However, for most trials, comprehensive GP data alone is insufficient and must be supplemented with national secondary care data. Together these datasets provide an overview of patients’ interactions with GP and hospital settings, enabling more reliable targeting of participants who may be eligible for a wide range of commercial trials.

The categories of key information captured within GP and SUS datasets is illustrated in Figure 4 below.

Ìý

Ìý

Figure 4: Key variables within the minimum viable dataset.

Ìý

Ìý

i The SUS serves as England’s single comprehensive repository for healthcare data, enabling reporting and analysis that supports NHS service delivery. Most national level datasets in England, such as the Hospital Episodes Statistics, are extracted from SUS data.

ii Information captured within national secondary care data in each devolved nation will differ from the SUS.

Ìý

Ìý

Ìý

Ìý

In addition to extensive coverage of patients’ medical histories and interactions with the health service, these two data sources exist in structured, queryable formats. For these datasets to be of value for facilitating recruitment into industry trials, the data need to be updated with sufficient frequency to align with a range of protocol selection criteria. As such, GP data will need to be updated weekly, as a minimum, while the current monthly SUS data updates, with validation typically lagging by approximately six months, should suffice to support data-enabled trials services.

At a minimum, comprehensive UK-wide GP data coverage, linked to national secondary care data and mortality data is required to provide meaningful feasibility assessments and to locate participants for commercial pharmaceutical trials based around delivery sites across the UK. Adopting this minimal essential dataset approach will lead to more targeted, efficient recruitment and significantly reduce costly and time-consuming screen failure rates.

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Ìý

Essential services

Ìý

In discussions ºÚÁϳԹÏÍø member companies identified the following three services that should be provided to support UK-wide data-enabled clinical trials, based on access to timely UK-wide GP and secondary care data, that would accelerate commercial trial delivery and increase global competitiveness of the UK.

Ìý

Ìý

Service 1

Feasibility assessments with self-service option

Currently it is not possible to obtain accurate feasibility assessments of the number of patients across the UK who may be eligible for a trial, due to the inability to access comprehensive population-wide clinical data. Consequently, UK feasibility assessments for trials are either generated locally at a selection of sites from current caseload lists and subsequently collated, derived nationally using tools based on proxy indicators of real patient numbers or extrapolated from access to incomplete patient databases. Many services in the UK and internationally still rely on form-based requests to interrogate patient records, even for straightforward questions about the number of patients fitting within the inclusion and exclusion criteria of a trial.²²

By centrally matching comprehensive GP data linked to national secondary care data to key selection criteria in the trial protocol, it will be possible to rapidly obtain more accurate UKwide feasibility assessments. This will enable UK affiliates to bid for a UK trial allocation and be confident that these estimates will subsequently translate into patients enrolled. It is not uncommon for companies needing to submit feasibility assessments within a two-week timeframe, so accuracy combined with speed is vital. Equally important, by using this approach companies can more reliably assess whether there are sufficient patients in the UK to take part in a trial and not waste effort bidding for trials that will not deliver. More accurate feasibility assessments that reliably predict the number of participants who could enrol in a trial will also increase efficiently in the system by enabling NHS staff at sites and industry to focus their resources on trials that have the greatest likelihood of successful delivery. 

In addition to a managed service offering, carried out by staff using pseudonymised GP and secondary care data, the feasibility service should include a self-service option, enabling industry researchers to also directly run queries on these pseudonymised datasets. The use of pseudonymised data and privacy techniques ensures that patient identities are concealed from all researchers using the feasibility service.

Ìý

Ìý

Service 2

Geomapping and site selection

The UK could gain further competitive advantage by understanding where relevant potential participants are physically located across the UK and therefore which trial delivery sites may hold the greatest recruitment potential. Trial site selection could consequently be based not only on site’s capacity and performance information, as is currently the case, but also on geomapping sites to locations with the highest relevant patient populations. Using a more evidence-based approach for site selection would drive efficiencies by placing trials in areas with greater certainty of delivery. This would reduce costly site set-ups and wasted NHS staff time on sites that fail to recruit or only recruit very small participant numbers. Geomapping would be particularly advantageous for rare disease trials where locating sufficient eligible patients within site catchment areas is an even greater challenge.

Ìý

Ìý

Service 3

Recruitment pre-screening

Although there is increasing recognition of the value of using patient records in pre-screening potentially eligible participants, currently trial sites in secondary care must rely on access to local available hospital datasets, usually in the absence of GP records. The current practice of each site carrying out bespoke searches on incomplete datasets that lack most of patients’ medical histories is inefficient and labour intensive. Significant time and cost savings can be made by centralised pre-screening of linked GP and national secondary care datasets, such as SUS in England. For each trial, the data could be modelled to the industry protocol once, using methods that safeguard patient confidentiality, and the resulting pre-screened pseudonymised lists of potentially suitable participants could be used by sites for more detailed local clinical review. Supplying individual sites with a pre-screened list would not only reduce the burden of each site having to do searches for suitable individuals from scratch for every study, but it would expedite trial startup if the centralised lists were generated during the regulatory approval process. This would free up resources at sites, enabling NHS staff to focus on selecting the most appropriate participants and inviting them for screening immediately after the site receives regulatory approval for the study to commence.

Ìý

Ìý

Ìý

Additional service offerings

Ìý

Additional services could be offered to supplement the essential three functions of the data-enabled clinical trials service.

Offering 1

Protocol design consultation

A protocol design consultation service could offer sponsors evidence-based guidance on which protocol inclusion and exclusion criteria could be reliably modelled using pseudonymised, structured NHS data, which would require manual abstraction, and which are not available in clinical records. For some companies, there are limited opportunities to adjust pre-determined protocols that have been optimised for US healthcare delivery models. However, where there is an opportunity to shape protocol development to be more aligned with UK systems, this service offering, in conjunction with existing services that inform UK protocol design, could contribute to more efficient downstream feasibility assessments and rapid patient recruitment.

Ìý

Ìý

Offering 2

Feedback loop to support continuous service improvements

The ability to continually refine and improve services through a feedback loop will underpin successful evolution, ensuring services are fit for purpose to meet the needs of trial participants, delivery infrastructure and industry sponsors. Analysis of the utility of pre-screened, pseudonymised potential participants lists for facilitating identification of suitable participants at sites to invite for screening, and subsequent conversion rates of screened individuals into enrolled participants, should be carried out on an ongoing basis. Monitoring the translation rates of feasibility assessments into real-world participants screened and enrolled at sites will be crucial for validating and continually improving centralised modelling of pseudonymised national datasets against protocol inclusion and exclusion criteria.

Ìý

UK-wide, centralised data-enabled function that integrates with the UK clinical trial delivery infrastructure

Ìý

A new data-enabled clinical trial function must enhance the capability of the UK’s existing clinical trials delivery infrastructure and not create a standalone data service operating in isolation from established clinical trial workflows and delivery systems. It should seamlessly interface with research delivery networks, UK Clinical Research Delivery Centres, site infrastructure and trial coordination mechanisms to increase efficiencies and boost recruitment at secondary and primary care trial delivery sites across the UK. The function should provide a unified and coordinated UK-wide offering, carrying out centralised data searches on pseudonymised NHS data, modelled against industry protocol criteria, to support subsequent clinical validation, patient identification and recruitment by clinical trial delivery sites. This integrated approach contrasts with current fragmentation, where industry navigates multiple disconnected services, approaching CPRD SPRINT, NHS DigiTrials, regional services, therapeutic area registries and trials networks separately. While each offers effective services within its remit, coordination across different services is limited. Sponsors lose time navigating different interfaces, governance requirements and timelines, resulting in trial sites receiving data from multiple sources without sufficient context.

The HDRS has a mandate to reduce fragmentation and increase availability of, and access to, NHS datasets for research. Given that three of the four stated priority datasets that the HDRS will initially make available – GP data, national secondary care data and mortality data – are the proposed minimal datasets to support data-enabled clinical trials services, it makes logical sense for a data-enabled clinical trials function to operate within the HDRS. This UK-wide data-enabled capability housed within the HDRS would, for the first time, provide a comprehensive overview of UK recruitment potential and offer a single engagement point for sponsors, trial coordinating infrastructure and individual sites seeking support from data-enabled approaches.

The proposed architecture for an end to end service design maintains appropriate information governance boundaries between centralised pseudonymised feasibility assessments, site selection and pre-screening in the HDRS and subsequent patient identification following clinical review at NHS sites (see Figure 5). Pseudonymised datasets would be centrally matched against industry protocol inclusion and exclusion criteria within the HDRS to determine trial feasibility, and to geomap where the highest number of potential participants may be located to help inform site selection. Following more in-depth data modelling and clinical interpretation against complex protocol eligibility criteria, a pre-screened potential participant lists would be generated and passed on to selected delivery sites within the NHS. The identity of individual patients would only be known to NHS staff.Ìý

Trial sites would receive pre-screened, pseudonymised participant lists from the dataenabled trials function in the HDRS, accompanied by clear contextual information indicating which eligibility criteria have been modelled in data searches and which require further clinical review by local NHS site teams. The provision of prescreened lists would negate the need for each delivery site to search available information from scratch, which would drive system efficiencies by saving valuable NHS resources and time. Local clinical teams at sites would then validate prescreened lists and identify suitable individuals to invite for screening when the regulatory approvals have been granted. Participants would be identified and contacted by appropriate people with a legitimate basis for access to identifiable health data. This process of matching patient eligibility criteria to their medical history prior to invitation reduces screen failures and frustration for patients, who are contacted only to subsequently find out they are not eligible based on existing information in their health records.Ìý

The centralised data-enabled clinical trial services provided by the HDRS should be appropriately staffed with researchers, such as epidemiologists and data scientists, as well as individuals with clinical knowledge who understand NHS clinical practice, coding conventions and data-quality issues, including when coding errors impact data interpretation.^9,10,23,24

Ìý

Ìý

Ìý

Ìý

Ìý

The efficiency funnel: from protocol to recruitment

Ìý

Current methods for identifying participants to recruit to a trial often creates overwhelming volumes of false positives at the screening stage. Not only is this costly, but it is time consuming for all involved. The goal should be to increase efficiency by more effectively identifying suitable patients prior to inviting them to take part in a trial. The operating model for the data-enabled clinical trials function should therefore follow a funnel efficiency approach, modelling NHS data to protocol inclusion and exclusion criteria where possible. Systematic application of information within the medical record will progressively narrow the population pool to a more targeted group of individuals with a higher likelihood of being eligible to participate in a trial, before inviting individuals for recruitment screening.

Ìý

Ìý

Figure 6: Efficiency funnel from feasibility to recruitment.

Ìý

Ìý

Ìý

Ìý

Ìý

Developing trial protocol archetypes

Ìý

While there is clear potential for improving recruitment efficiency and inclusivity by mapping trial eligibility to NHS data sources, the reliability of meaningfully translating the array of NHS data codes into protocol inclusion and exclusion criteria is not well understood. Existing data-enabled services have gained a wealth of experience interpreting pharmaceutical industry protocols. However, this knowledge largely remains siloed within these organisations.

To retain system knowledge and avoid unnecessary duplication of effort, it is recommended that a mapping exercise is carried out characterising applicability of NHS data to common phase II-IV selection criteria in the trial protocol. The aim would be to generate reusable protocol archetypes, organised by therapeutic area, documenting the criteria that can be reliably interpreted from coded data in the minimum recommended datasets, which could be extracted from specialist hospital data sources, and which would only be obtained at the screening stage. The mapping should capture diagnostic coding accuracy, data completeness and generate metadata on data quality, data velocity and known limitations of NHS data by therapeutic area, in particular in relation to common protocol temporal requirements. 

This reusable evidence base would transform current methods of bespoke data modelling, which currently generate novel algorithms requiring clinical interpretation for every trial protocol, into a system that matches against validated archetypes, significantly reducing the workload for each new study. For instance, for cardiovascular trials that consistently require recent lipid panels, the archetype would capture which laboratory systems record this, with what completeness, at what update frequency.

Ìý

Ìý

Ìý

Therapeutic areas most suited to modelling protocols against NHS datasets

Ìý

The utility of using NHS data to facilitate clinical trial recruitment will vary by therapeutic area depending on which datasets the most relevant clinical information is recorded in, whether or to what extent salient information is captured, and how accessible these data sources are to researchers. Not all therapeutic areas will be equally suited to a data-enabled approach that is predicated on central modelling of national NHS datasets against protocol criteria, supplemented by local data sources at delivery sites.

Trials in therapeutic areas that most lend themselves to this two-tiered approach align with conditions where a large proportion of the protocol criteria are routinely recorded in GP and national secondary care datasets. Many of the protocol inclusion and exclusion criteria for conditions commonly managed in primary care, such as cardiovascular, metabolic and respiratory conditions, can be mapped to diagnoses, prescribing, test results and interventions recorded in GP data, supplemented by national secondary care data. These two datasets can contribute to neurology and immunology protocol prescreening, however, some of the trial selection criteria for these conditions depend on access to specialist hospital datasets that currently cannot be accessed centrally. For these trials, there would need to be greater emphasis on site delivery teams consulting relevant local data sources to augment the initial pre-screens carried out on national data held within the HDRS. The situation for oncology trials is different. Key eligibility selection criteria for oncology trials are entirely dependent on information within datasets held in specialist oncology centres or individual trusts.

Understanding applicability of different datasets for supporting trials in various therapeutic indications, informed by development of trial protocol archetypes, will define the scope of where data-enabled clinical trials services can realistically support commercial trial recruitment. Annex Table 1 provides an overview of how modelling NHS data via the proposed centralised approach might support trial pre-screening in different therapeutic areas.

Ìý

Ìý

Ìý

Oncology trials require a different approach

Ìý

Oncology trial screen failure rates often reach 90 per cent because matching patients to the complex trial eligibility criteria requires access to data held within hospitals and specialist oncology centres in multiple different NHS data sources. These can include histopathology reports confirming tumour type and grade, radiology reports documenting tumour progression, reports recoding biomarkers present, and prescribing data containing treatment sequencing. This information is outside the scope of GP and national secondary care data sources and, consequently, a centralised modelling approach using these datasets has limited utility for oncology trial recruitment. Annex Table 2 provides further detail on oncology trial specific data requirements. The most recent ºÚÁϳԹÏÍø annual clinical trials report highlighted that oncology was the leading therapeutic area among UK pharmaceutical industry clinical trials, accounting for three in every 10 phase II and III trials initiated in 2024. Significant gains in UK competitiveness could be made through more effective identification of eligible patients for these trials. Due to the challenges of collating information from the range of specialist data sources required to support oncology protocol pre-screening, a dedicated strategic approach is needed to inform the specification for an oncology data-enabled clinical trials offering.

To understand the potential for data-enabled services to improve recruitment into commercial oncology trials, a network of four to five major tertiary oncology centres across the UK should initially be formed. The aim of the network would be to develop a standardised data framework for broader rollout, by defining which datasets are of most value and what standards should be used, and to practically test data-sharing approaches across sites.

Ìý

Ìý

concentrated trial-eligible populations: major tertiary centres that have a high through-put of cancer patients who could be eligible for a wide range of trials

Ìý

holding necessary data sources: centres holding relevant histopathology, radiology, genomics, and treatment datasets locally

Ìý

existing research capabilities: already research active with clinical academics leading on numerous industry oncology trials.

Ìý

Ìý

Ìý

Factors that will determine impact

Ìý

The primary driver for establishing a dedicated function to support data-enabled clinical trials at scale, is to increase UK competitiveness by improving efficiency of recruitment into commercial trials. Evaluating the impact of data-enabled services should therefore be based on measures of timely and accurate feasibility assessments and subsequent rapid, targeted participant recruitment by the trials delivery system. The following performance indicators are proposed:

• Feasibility assessments within two weeks, based on an overview of whole population national datasets. Provision of feasibility assessments within two-week timeframes will improve the UK chances of being selected in global allocations of trials. Accurate as well as timely assessments will be critical to instilling confidence among global pharmaceutical companies in choosing the UK as a location. To evidence whether the data-enabled feasibility service is improving quality of feasibility predictions, the accuracy of estimates generated by this service that subsequently translate into real patients enrolled in trials, compared with current baseline estimates using incomplete or proxy data, should be measured.

• Reduced proportion of non-recruiting sites by geomapping sites with the highest recruitment potential. Reducing the number of sites that fail to recruit any participants or recruit far fewer than their projected numbers will lead to cost savings for sponsors and will reduce time wasted by the NHS setting up sites that do not deliver. Comparing the proportion of nonrecruiting sites in current and recent studies to trials where sites are selected through evidence-based geomapping will measure the performance of the geomapping service.

• More efficient and targeted recruitment through centralised evidence-based pre-screening. Supplying delivery sites with a pre-screened lists of pseudonymised participants who may be suitable for a trial should enable NHS staff at these sites to concentrate their time on reviewing a short-list of patients. This twotiered approach drawing on existing information within NHS data should lead to fewer screen failures among those invited for screening, resulting in a higher proportion of successfully enrolled participants. Conversion rates of invited participants to those invited for screening to similar types of trials using current methods can be used to evaluate the data-enabled prescreening service performance. The approach of using centralised pre-screening searches should also facilitate faster trial set-ups, enabling sites to more rapidly identify suitable participants to invite for screening as soon as they receive regulatory approval. Time taken post-regulatory approval to recruit the first patient, as well as time to reach the target number of patients, compared with current timeframes could be measured to determine the data-enabled prescreening service’s contribution to expediting commercial trial delivery.

Ìý

Ìý

Ìý

Additional benefits

Ìý

Other benefits that a data-enabled clinical trials function should bring include time saved by NHS staff at each site carrying out prescreening from scratch, and increased inclusivity and representativeness in clinical trials due to the ability to review all potential suitable trial participants. These benefits may be more difficult to quantify, as relevant metrics are not routinely collected. The Health Research Authority, government and trial delivery partners are currently developing metrics to assess diversity and inclusion in trials. When implemented, these metrics may provide a suitable baseline to assess whether data-enabled approaches contribute to the shared goal of increasing representativeness and inclusion in trials.

Ìý

Conclusion

Ìý

Government has recognised the immense value that industry clinical trials bring to patients, the NHS and the economy and the need to make the UK a more appealing and reliable place to deliver industry global trials. As a consequence, ambitious government targets have been set to reduce commercial trial set-up times and significantly increase commercial trial recruitment. To achieve these goals, a series of measures are being implemented to streamline processes, reward good practice and increase performance transparency across the delivery system. Undoubtedly, these actions will lead to demonstrable improvements that should make the UK a more attractive destination for global industry.

However, for the UK to seriously compete with preferred European markets in Spain, France and Germany, globally with China and the US, and with emerging markets such as South Korea and Canada, a transformational change is required. Fortunately, the UK has an opportunity to enact this transformational change by strategically aligning the two major government commitments to establish an HDRS and to improve the environment for delivering commercial clinical trials. NHS records contain longitudinal medical information on the UK’s population. If successfully implemented, creation of a data-enabled clinical trials capability that harnesses this globally unique NHS asset, offers the UK a competitive advantage in accelerating recruitment into commercial trials via the UK’s established trial delivery infrastructure.

Public trust in responsible use of health data in research is paramount. The service design proposed for data-enabled clinical trials is protects patient confidentiality, with only staff at participating NHS organisations being able to access identifiable patient data. The proposed central data feasibility assessments and recruitment pre-screening services would be carried out on pseudonymised data within the HDRS, where researchers do not have access to identifiable patient information. Use of pseudonymised data to support clinical trials should be part of HDRS planned discussions with patients, the public and healthcare professionals, which are aimed at building trust in transparent use of patient data for research.

Although the rationale for this proposal is to create industry-standard data-enabled services that will improve speed and predictability of industry trial delivery, the services do not have to exclusively support commercial studies. Industry clinical trials must comply with the highest regulatory standards and deliver to globally competitive timeframes. In setting the bar high to establish an operational standard required to fulfil industry’s requirements, the resulting service capabilities should also be fit for purpose to support the needs of non-commercial trials, resulting in benefits for the whole UK trial ecosystem.

The seven recommendations outlined in this report to drive UK-wide data-enabled clinical trials capability, would unlock the potential of NHS data to significantly increase efficiency of recruitment into commercial clinical trials. Critical to success is centrally collected GP data, as is the need to seamlessly integrate the data service with the existing trial infrastructure. Implementing the full suite of proposed UK-wide data-enabled services is a major endeavour and will take time to be fully operational. In the meanwhile, a modular stepwise approach, building on existing data-enabled trials services or national data services, in particular those that have access to GP data given its criticality for matching to industry protocols, should be adopted to inform end-to-end service design and workflows.

The requirements have been developed by the pharmaceutical industry and from discussions with numerous stakeholders, including therapeutic conditions that would be most suited to a data-enabled approach to recruitment and where additional scoping, such as for oncology trials, would need to be done. With participation in commercial clinical trials in the NHS at its lowest point in seven years and the HDRS currently formulating its business plans, the time to act on these recommendations is now.

Ìý

Ìý

Ìý

Ìý

Ìý

Establishing a service design group

Ìý

The critical next step is to convene a UK-wide service design group (SDG) bringing together government representatives from the four nations, the HDRS team, key parts of the clinical trials delivery system and the pharmaceutical industry to define how a the delivery of data-enabled commercial trials at scale could be achieved. This collaborative design process must start with the problem to be solved – rapid, targeted recruitment into commercial trials across the UK.

The SDG would have five core objectives:

1. Validate the proposed design with clinical research delivery systems to confirm that it meets real-world needs for facilitating rapid feasibility assessment, site selection and recruitment across the UK.
   
   
2. Set out the operational approach and workflows needed to connect HDRS data searches with trial coordination and delivery systems and recommend governance structures with clear responsibility for implementation.
   
   
3. Recommend a phased roadmap to guide prioritisation of set-up and implementation of the key data-enabled services, including access to comprehensive UK-wide GP data linked to national secondary care and mortality data to form the minimum service dataset.
   
   
4. Expand the Clinical Practice Research Datalink (CPRD) as a minimum viable product to demonstrate proof of principle, building on its existing industry-facing trial services and linking secondary care data with CPRD’s UK-wide GP data to provide quick wins and inform UK service design.
   
   
5. Define success metrics translating the key performance indicators into measurable targets.

Convening the SDG in the first half of 2026 would signal a strong commitment to translating the UK’s structural advantages into coordinated strategic actions aimed at significantly improving predictability and efficiency of commercial trial delivery. Successful delivery of data-enabled clinical trials would reverse declining recruitment, secure international investment and deliver on government commitments. Importantly, creating services that is aligned with industry requirements will finally realise the potential of NHS data to increase the UK’s competitive advantage and reinstate its position as a global leader in clinical trials delivery.

Ìý

Methodology

Ìý

The research for this report was conducted for the ºÚÁϳԹÏÍø by Newmarket Strategy and employed a mixed-methods approach combining deskbased literature review, stakeholder interviews, and participatory workshops to examine the NHS data landscape and challenges in feasibility and recruitment for clinical trials.

The literature review encompassed desk research on the NHS data infrastructure, systematic examination of academic papers focusing on inclusion and exclusion criteria in trial protocols, and analysis of participant recruitment challenges. This was supplemented by a review of grey literature on the UK’s clinical trial landscape and associated recruitment difficulties, providing context on the operational environment within which electronic health records might be deployed. Primary data collection comprised semi-structured interviews with 20 stakeholders from UK-wide data providers including SAIL Databank, Research Data Scotland, NHS DigiTrials, and CPRD. Additional perspectives were gathered from research delivery networks and NHS trusts to capture both the national infrastructure view and local operational realities.

Ìý

Ìý

Ìý

ºÚÁϳԹÏÍø member engagement

Ìý

Two Task and Finish Group workshops were conducted with ºÚÁϳԹÏÍø members to validate findings and co-develop recommendations. The first workshop focused on analysing inclusion and exclusion criteria from several trial protocols to identify which trial types would be most amenable to using electronic health-record data for participant identification and exploring data requirements for a data-enabled clinical trial recruitment service. The second examined ºÚÁϳԹÏÍø members’ existing experience with feasibility and recruitment services and defining service requirements for a new data-enabled clinical trial functionality, including how it might integrate with existing infrastructure.

Findings were presented through four sessions to ºÚÁϳԹÏÍø data and clinical trials member groups to gather feedback on findings followed by recommendations and structured follow-up discussions.

Ìý

Ìý

Ìý

Ìý

Ìý

Interviewees

Ìý

The views expressed in this report are those of the ºÚÁϳԹÏÍø. Interviews were conducted as part of the background research and the perspectives of individual interviewees may differ from the positions presented here:

Ìý

Ìý

Ìý

Ìý

Table 1: Overview of how using a centralised NHS data modelling approach might suit different therapeutic areas.

Ìý

Ìý

Ìý

Ìý

Table 2: Data required to support oncology trial eligibility criteria that are not held within national GP or secondary care datasets.

Ìý

Ìý

Ìý

Ìý

Ìý

References

Ìý

1. The ºÚÁϳԹÏÍø, ‘The value of industry clinical trials to the UK - extended report’, December 2024, available at www.abpi.org.uk/publications/the-value-of-industry-clinical-trials-tothe- uk-extended-report
2. The ºÚÁϳԹÏÍø ‘UK industry clinical trials: translating actions into impact’, December 2025, available at www.abpi.org.uk/publications/uk-industry-clinical-trials-translating-actions-intoimpact
   
   
3. The ºÚÁϳԹÏÍø, ‘The value of industry clinical trials to the UK’, September 2024, available at www.abpi.org.uk/publications/the-value-of-industry-clinical-trials-to-the-uk
   
   
4. Department of Health and Social Care, Prime Minister’s Office, Department for Science, Innovation and Technology, Office for Life Sciences, Office for Investment, ‘Prime Minister turbocharges medical research’, April 2025, available at
   
   
5. Department for Business and Trade, Department of Health and Social Care, Department for Science, Innovation and Technology, Office for Life Sciences, ‘Life Sciences Sector Plan’, July 2025, available at
   
   
6. Diwakar V, Vickers L, ‘Health Data Research Service – unlocking the potential of health and care data to transform lives’, NHS England, August 2025, available at
   
   
7. Atkins V, May P, Morgan E, Matheson M, ‘Full government response to the Lord O’Shaughnessy Review into commercial clinical trials’, December 2023, available at
   
   
8. Ross J, Tu S, Carini S, Sim I, ‘Analysis of eligibility criteria complexity in clinical trials’, Summit on Translational Bioinformatics, 2010, pp46–50
   
   
9. Dong H, Falis M, Whiteley W, Alex B, Matterson J et al., ‘Automated clinical coding: what, why, and where we are?’ NPJ Digital Medicine, 2022, 5, p159
   
   
10. Tsopra R, Peckham D, Beirne P, Rodger K, Callister M et al., ‘The impact of three discharge coding methods on the accuracy of diagnostic coding and hospital reimbursement for inpatient medical care’, International Journal of Medical Informatics, 2018, 115, pp35–42
    
    
11. Goldacre B, Morley J, ‘Better, broader, safer: using health data for research and analysis’, Department of Health and Social Care, April 2022, available at
    
    
12. Department of Health and Social Care, ‘Data saves lives: reshaping health and social care with data’, June 2022, available at
    
    
13. Brooks S, ‘The NHS data challenge: how to unlock insights from a sea of data’, Health Service Journal, June 2024, available at
    
    
14. McCarthy M, ‘Reimagining the NHS must focus on restoring general practice’, BMJ, 2024, 387, q2295
    
    
15. Baker M, Ware J, Morgan K, ‘Time to put patients first by investing in general practice’, British Journal of General Practice, 2014, 64, pp268-9
    
    
16. Lensen S, Macnair A, Love SB, Yorke-Edwards V, Noor NM et al., ‘Access to routinely collected health data for clinical trials – review of successful data requests to UK registries’, Trials, 2020, 21, p398
    
    
17. Sydes MR, Barbachano Y, Bowman L, Denwood T, Farmer A et al., ‘Realising the full potential of data-enabled trials in the UK: a call for action’, 2021, BMJ Open, 11, e043906
    
    
18. Zhang J, Morley J, Gallifant J, Oddy C, Teo JT et al., ‘Mapping and evaluating national data flows: transparency, privacy, and guiding infrastructural transformation’, 2023, Lancet Digital Health, 5, e737-48
    
    
19. Pharmaceutical Research and Manufacturers of America, ‘PhRMA 2025 Annual Membership Survey’, July 2025, available at
    
    
20. The European Federation of Pharmaceutical Industries and Associations, ‘The economic impact of industry clinical trials across Europe’, February 2026, available at
    
    
21. The ºÚÁϳԹÏÍø, ‘Creating the conditions for investment and growth: pharmaceutical industry investment competitiveness framework’, September 2025, available at www.abpi.org.uk/publications/creating-the-conditions-for-investment-and-growth
    
    
22. Callahan A, Polony V, Posada JD, Banda JM, Gombar S et al., ‘ACE: the Advanced Cohort Engine for searching longitudinal patient records’, Journal of the American Medical Informatics Association, 2021, 28, pp1468-79
    
    
23. Mahbubani K, Georgiades F, Goh EL, Chidambaram S, Sivakumaran P et al., ‘Clinician-directed improvement in the accuracy of hospital clinical coding’, Future Healthcare Journal, 2018, 5, pp47–51
    
    
24. Pankhurst T, Evison F, Atia J, Gallier S, Coleman J et al., ‘Introduction of systematized nomenclature of medicine clinical
    terms coding into an electronic health record and evaluation of its impact: qualitative and quantitative study’, JMIR Medical Informatics, 2021, 9, e29532