The introduction of Integrated Care Systems (ICSs) in England marked a shift toward partnership working and shared responsibility, by bringing together National Health Service (NHS) organisations, Local Authorities, Voluntary Sector Organisations and others to deliver health and social care across specified geographical areas [1, 2]. The main purpose of establishing ICSs was to improve population health and mitigate inequalities. Partnership working across the health and social care sector supports this by allowing a mutual understanding of local needs and the use of collective resources to meet those needs [3]. As a result, government policy requires ICSs to use data from across the health and care system to inform their decision-making [4, 5]. This aligns with the concept of a ‘learning health system’, which describes how data from across the health and care system can be used to generate new knowledge, to directly improve service delivery [6].

Effective use of health service data is recognised as critical to responding to the current challenges facing the NHS [7]. However, as the wider determinants of health extend beyond clinical settings, there has also been a drive to link health records with non-health sources (such as education data) to achieve a holistic view of individual and population health [8]. Integrating these broader concepts of health into health planning requires a ’systems thinking’ approach, where local decision-makers move away from viewing public services in isolation and instead, consider health as part of a complex, interconnected web of services, where changes in sectors like education or social care can also influence health outcomes and health inequalities [9]. As such, several population level datasets have been established to facilitate this (e.g. Whole Systems Integrated Dataset, Lambeth Datanet, Connected Bradford, the Discover Dataset) (8, 10–12] and there have been significant investments in linking data across services to support children and families [13, 14]. This includes the Born and Bred in (BaBi) Network; a collection of local birth cohorts across England [15]. In contrast to traditional population-based research that relies on national ’opt-out’ models to use and link routine records, BaBi cohorts utilise a consent-based approach, where pregnant women are invited to join a local BaBi cohort by a trained midwife during a routine antenatal appointment. This enables recruitment to be ongoing and permits the linkage of their and their child’s routine health, education and social care records for research purposes. In this paper, we use the term ‘routine data’ to refer to data that are collected by public services as part of their service delivery or clinical practice and research that uses data from more than one source as ‘linked data research’. Each BaBi cohort aims to harness the power of routine data about local families for research to support local decision-making [15]. Local evidence can provide important insights for local decision-makers, including whether local services are working, how they are working and who for [15].

While this opt-in model means the BaBi cohorts are not strictly population-based, it represents a deliberate response to England’s complex and highly contested history regarding health data sharing. Previous research with members of the public in the United Kingdom (UK) has highlighted that trust and transparent use of data for research is crucial for public acceptance [5, 16, 17] and the BaBi consent model prioritises building this trust. BaBi participants also consent to being contacted about future research opportunities, which therefore enables the creation of a sustainable research platform that promotes inclusivity in research and transparent use of service-user data [15].

The BaBi Network was founded on the success of the Born in Bradford research cohort, which successfully linked primary research data to routinely collected data to inform local policy and practice [18, 19]. While the original Born in Bradford cohort has established a proven track record of influencing local service delivery, the BaBi cohorts were created with the aim of scaling this potential. They provide a cohort for every generation, through continuous, ongoing recruitment, which can capture a real-time picture of families experiencing policy changes to support policy evaluations and changes to service provision. The BaBi Network began with five pilot sites (Bradford, Doncaster, East London, Leeds and Wakefield) and following the success of these cohorts, has expanded to include 13 active sites at the time this paper was written, representing approximately 10% of the pregnant population in England. An up-to-date list of local areas with BaBi cohorts can be found at: https://www.babinetwork.co.uk/babi-sites.

Despite increased investment in linking routine data and making this accessible for research, there is limited evidence regarding how local decision-makers perceive and utilise such research. Existing studies have explored stakeholder views on population-level data sharing without explicit consent, under the lawful basis of ‘task in the public interest’ [20, 21], however, the utility and current use of locally owned, consented, longitudinal routine data remains under-researched.

While the challenges of using routine data for research are well documented [7, 21, 22], there is evidence to suggest these data can still be useful for informing policy decisions, particularly around child and maternal health. For example, project data have been linked with routine health and education records to evaluate the Flying Start childcare programme in Wales [23]. In England, routine health, education and social care data were linked to evaluate the Family Nurse Partnership, which provides support to young mothers from pregnancy until their child is aged between one and two years [24]. Given this potential, and because the overarching aim of BaBi cohorts is to link wide-ranging data to inform policy and practice, understanding decision-makers’ attitudes around using linked routine data for research is essential, if linked data research outputs are to be used effectively to address key challenges facing the health and care system. This aligns with the core principles of implementation science, which focuses on the methods and strategies used to systematically promote the uptake of research evidence into routine, real-world practice [25]. As highlighted in this field, establishing the effectiveness of an innovation does not guarantee it can be used in practice and from this perspective, simply creating and linking datasets does not mean that decision-makers will use these datasets as evidence. Therefore, it is important to understand the human and organisational factors that can influence use of these data. This provides the rationale for the current study, as understanding these perspectives is a necessary first step toward realising the network’s goal of informing local practice.

Hence, this study explored the perspectives of local clinical decision-makers and commissioners, through qualitative semi-structured interviews, regarding:

This paper is based on work originally presented in the lead author’s doctoral thesis focused on how research using linked routine data could support local decision-making [26].

To ensure comprehensive reporting of methods, this section is structured according to the Consolidated Criteria for Reporting Qualitative Research (COREQ) checklist [27]. The study protocol was published prior to data collection [28].

Interviews were conducted with fifteen local early years decision-makers. The sample comprised individuals from clinical (n = 8), commissioning (n = 5) and BaBi network (n = 2) roles, distributed across four areas: Bradford (n = 4), Doncaster(n = 4), Leeds(n = 4), and Wakefield (n = 3). Participants represented local government, primary care, midwifery, health visiting and local perinatal and early life health services. Table 1 describes the participants’ decision-making responsibilities [26].

Due to the implementation of the ICSs, decision-makers often covered multiple local areas. Many participants had cross-cutting roles within their organisations and had experience making decisions in several areas of early life health. The resulting depth and breadth of these data provided high information power, which facilitated a comprehensive address of the research aims with a more concentrated sample size.

Fourteen interviews were held online, and one interview was held over the phone due to connectivity issues.

Analysis of interview data with local early years decision-makers has provided an understanding of their views regarding the utility of linked data research for informing local decision-making and how researchers can effectively engage and support decision-makers to make use of these data as evidence [26].

Our findings demonstrate that local decision-makers perceived value in using local linked routine data for research and to inform decision-making, viewing it as a critical “starting point” for evidence-based practice. Participants highlighted the unique evidence contribution of linked data, that separate, unlinked datasets cannot. Rather than looking at individual services in isolation, linking data across health, education, and social care allows decision-makers to gain better insights into how the wider determinants of health impact families over time. Furthermore, because this evidence is generated locally, it provides local commissioners with a powerful tool to justify funding and protect services, in contrast to research evidence generated from populations that may not match their local demographic. By proving that specific problems exist within their own populations, local decision-makers can confidently target resources to reduce health inequalities.

It was expected that participants would attribute some value to linked datasets given the recent large investments in data linkage made by local government [13, 43]. They were also either involved in a local BaBi steering group or were contacts of those involved in a BaBi steering group, meaning that many participants had heard of linked data and its potential benefits. Furthermore, local BaBi teams were working with partners across the NHS and Local Authorities to promote BaBi. The experiences of decision-makers involved in a BaBi steering group broadly resonated with those not as closely involved in BaBi.

While the appetite for linked data was high, participants highlighted that its practical application is often undermined by poor data quality. This study identified two main reasons for this: 1) inadequate electronic health record systems used to record these data, and 2) limited understanding by health and care professionals regarding how these data are used beyond clinical care. These discussions reflect the current research evidence, which suggests that the quality of routine data is limiting its potential for research and decision-making [7, 20, 44, 45]. From an implementation science perspective, accurate data entry at the frontline is itself a critical behavioural intervention that requires targeted implementation strategies. To address these challenges, participants suggested that data literacy should be a core part of professional training, whereby health and care professionals develop an understanding of how routine data is used for research and service planning beyond the immediate clinical appointment. The eHealth Swiss initiative in Switzerland [44] offers a valuable example of how to integrate digital health topics into professional education and learning from this could be used to develop training for local policymakers and health care professionals on the importance of accurate routine data capture.

However, the benefits of improving data literacy among health professionals is limited by the current constraints of electronic record systems. Hence, to improve the quality of routine data for research, electronic record systems must be co-designed and developed to meet the needs of all end-users. This includes the needs of researchers, senior managers, national policymakers, patients or carers [7]. The NHS long-term plan has a strong emphasis on digital transformation, which provides the opportunity to improve data sources and data linkage across the health system [46].

To further support local decision-makers to utilise linked data research, participants advocated for engagement strategies that prioritise convenience and relationality. This is consistent with the wider literature suggesting that the “research-to-policy” gap is best bridged through building collaborative relationships and presenting research in easily accessible formats [29–32]. The preference for layered information, such as infographics and executive summaries, backed by full research papers, allows decision-makers to perform rapid evidence appraisals without being overwhelmed. For researchers, this implies that the dissemination phase of a project is just as critical as the analysis. The “right conversation at the right time” (I15) suggests a move toward consultative research models where researchers are available to provide brief, high-impact verbal syntheses of their findings to time-poor stakeholders. Further research is required to understand whether communicating the research outputs in the way participants have described is effective at influencing the uptake of the linked routine data research in policy or service provision.

Time and capacity constraints were frequently discussed as barriers to engagement with linked data research, a finding echoed in the wider ‘research use’ literature [32]. Our findings suggested that these constraints are driven by how decision-makers’ perceive their professional responsibilities. Specifically, while they valued linked data, many viewed research as being outside of their core remit. Consequently, research engagement may be deprioritised in favour of tasks seen as more central to their primary roles. This demonstrates that an understanding of the users’ individual and behavioural context is critical for designing effective linked data implementation strategies.

Our findings also imply that some decision-makers feel it is the researcher’s responsibility to promote their findings in a way that decision-makers can easily take action. This leaves a gap between research findings and practical solutions, and it is currently unclear who is responsible for filling that gap.

Existing evidence has suggested that policymakers often lack organisational support or training to use research effectively [33] and this could explain why some decision-makers’ do not view research as part of their responsibility. This implies that sustainable evidence uptake cannot rely solely on individual motivation and requires implementation strategies aimed at organisational and structural shifts. Specifically, to influence how linked data research is utilised, these organisational strategies must formally embed research engagement into decision-makers’ core job descriptions. Further exploration into if and why some decision-makers perceive research as outside of their core responsibilities and if this is the reason they feel unable to give up their time to engage with linked data research is essential for designing more effective collaboration strategies.

The evolving landscape of the UK health system provides a timely opportunity to institutionalise research engagement. The transition to ICSs encouraged cross-sectoral collaboration and research is a key element of their strategies [4]. Simultaneously, the establishment of National Institute for Health and Care Research (NIHR) Health Determinants Research Collaborations (HDRCs) specifically aimed to embed a culture of evidence-based policymaking within local authorities [47]. These structural shifts represent a move toward collective accountability for research and evidence. Moreover, evidence from a project scoping research activity in Bradford Metropolitan District Council suggests that they are willing to respond to invitations to collaborate in research, are willing to share data and there is evidence of some use of research in policy developed in some departments, meaning they have the potential to benefit from BaBi data and research [48]. Consequently, as research becomes an embedded consideration within local government and health and care services, the perceived lack of time to engage with research may diminish.

This study explored how local early years decision-makers perceive and utilise linked data research. As the BaBi Network represents a novel data linkage infrastructure for local authorities, understanding these perspectives is vital for ensuring that linked data research translates into meaningful local impact. Our analysis suggests that for routine data to truly influence policy and practice, a fundamental shift is required in how data systems are designed and how research responsibility is defined.

To achieve this, electronic health record systems should be co-designed to ensure they are technically capable of capturing high-quality data that meets the needs of both frontline practitioners and researchers. Furthermore, effective research engagement can be supported by structured communication channels, which facilitate a continuous, two-way exchange between researchers and decision-makers. However, these technical and communicative improvements may only succeed if the narrative around research responsibility is transformed. For engagement efforts to have a lasting impact, research use must be embedded into the core professional identity of decision-makers, rather than being viewed as an additional or peripheral task.

This research offers insights for early years data linkage studies on how to better support and engage with local decision-makers. As such, the BaBi Bradford team has already begun applying these findings to their ongoing collaboration with local services, working to refine data quality at the source.

The current rise in evidence-users within the UK and the increased investment in data linkage makes this research timely. By addressing the disconnect between researchers and policymakers through better system design and clearer professional accountability, we can move toward a landscape where routine data is effectively utilised to inform the best possible care for children and families.