Building research partnerships with local government: Reviewing five years of a Public Health Intervention Responsive Studies Team (PHIRST)

Background

The relationship between evidence, policy, and practice is nuanced, dynamic, political, and contested.¹ In England, local government is well placed to influence the determinants of public health and reduce inequalities following the transfer of responsibility from the NHS to local government in 2013 (though this differs in the devolved nations of the UK). However opportunities for intervention are often missed.² In addition, financial constraints on local government increases the need to identify interventions that are cost-effective in improving population health. Facilitating closer interaction between those working in public health policy, practice, and academic research is likely to increase the use of evidence in policy decisions to improve outcomes.

Previous research has identified numerous examples of the ways in which evidence may be used by local government, including but not limited to understanding the local context and local health need (including predicting future demands and planning public health services); evaluating the effectiveness of local services; informing policy and practice decisions; determining cost-effectiveness; and raising awareness of public health issues.³ Identified barriers to the use of research in policy-making (not limited to local government) include the absence of relevant research, lack of access to it, poor research dissemination, and lack of time to use research evidence.⁴ Research has also identified barriers to local authorities engaging in research activity, including limited time, analytical and financial resource; lack of staff for producing or managing research, and limitations in the capacity to align research with longer-term strategic needs or policy development.^5,6 The funding models in the public sector frequently means that research activity is framed around short-term project work, which can also make for difficulty in retaining the knowledge gained from such research.⁵ Partnering with public health academics and researchers may resolve some of these issues, but commissioning academic research is expensive, and there are often difficulties in identifying the relevant researchers to work with.⁷

To address this, Public Health Intervention Responsive Studies Teams (PHIRST) were established by the National Institute for Health and Care Research (NIHR) in 2020 to provide rapid evaluations of local public health interventions at no cost to local government. PHIRST is a national, competitive scheme open to local government public health teams across the UK (or equivalent organisations in the devolved administrations commissioning non-NHS, public health interventions). Criteria for prioritising interventions for support from a PHIRST team include that they: are likely to have important impacts on population health and health inequalities; have the potential to be widely generalisable; have an identified need; are readily evaluable; have funding identified and preferably secured; and for which evaluative research will provide timely evidence to inform practice or policy in local government.⁸ Four academic PHIRST teams were funded from 2020 and by 2025 ten PHIRST teams had been established, with over 80 studies underway or completed.

Several barriers to research partnerships between local government and academia have been identified in previous studies. Academic research has vigorous and time-consuming research governance demands around data sharing and ethics, and pressure to publish in peer-reviewed journals can make some research inaccessible.^3,7 The political nature of local government also means that public health priorities may be subject to rapid and continual change, requiring researchers to have relevant research evidence ready to respond to policy and political changes.^3,9,10 However, the lengthy time taken to produce academic evidence often means that it is perceived as outdated.^3,11 Local government staff want evidence that is timely and relevant to current practice and not ‘removed’ from what is happening on the ground.^2,5,12

Researchers’ existing commitments to long-term funding programmes may reduce their capacity to undertake rapid, localised studies,¹³ and building meaningful collaborations with local government practice and policy partners can come at a cost if the outputs of value to local government are not peer-reviewed journal articles needed for academic career progression (although this may be changing given the increased demand for research impact).¹² Different types of evidence may be valued, with practitioners favouring localised and in-house research, while academics value rigorous and transferable evidence with (inter) national relevance.¹² There is no guarantee that research commissioned to inform local government policy and practice will result in impact. The Local Government Association describes policy-making as evidence informed rather than evidence based.³ Factors influencing the use of evidence in local government include organisational, structural and staffing changes, financial pressures, availability of data and data sharing, and the national policy context and regulation.³ However, academics frequently cite a willingness to undertake solution-oriented research to address real-world problems, enabled by funding that supports collaboration between academia and practice.¹²

There are also drivers for local government public health teams to engage with academic research, including the perception that external evaluation may give greater credibility to their work¹² and be more likely to influence policymakers. There is also an increasing emphasis on adopting an evidence-based approach to justify financial decisions^5,7 and ‘proof-of-concept’ testing before embedding systemic change.^5,14 Personal contact and collaboration between policymakers and researchers, good research dissemination, and research outputs that include clear recommendations have all been found to support the use of research evidence in policy making.^4,11

PHIRST Insight, a collaboration between the Universities of Bristol and Cardiff, was established in 2020 and completed 10 studies (see Table 1 for details) with local government (or equivalent) in its first five years. After being allocated potential local government interventions for evaluation by the NIHR, PHIRST teams have six weeks to work with local government partners and other stakeholders to determine if the intervention is evaluable, and must develop a research protocol in collaboration with local partners within 12 weeks of allocation. Studies are not allocated based on geographical proximity to PHIRST academic teams, with the result that researchers are often unfamiliar with the local context, geography, socio-demographics and the public health priorities for the local government team involved. They are also building new relationships with public health commissioners, practitioners, and community members, creating a challenging context within which to develop a research study.

While the ten local interventions allocated to PHIRST Insight are diverse in terms of their intended health outcomes, target population, delivery method, and locality, Insight undertook these studies with three common principles: co-production of the research with local government partners, the embedding of public involvement across all aspects of the studies, and a focus on maximising the impact on local government policy and practice.

In 2024, the Insight team began a review of their work to date to determine whether these three principles had been implemented effectively. A full review of public involvement will be published in a separate paper. This article reports on local government partners’ (including third-sector agencies commissioned to deliver interventions) experience of co-producing studies with Insight, and the perceived impact of PHIRST Insight studies. The study also makes recommendations for improving research partnerships with local government and supporting impact within the context of the PHIRST scheme.

Methods

This qualitative study consisted of semi-structured interviews with people involved in each of the 10 Insight studies. Participants were recruited from two groups; local government partners (including some from third-sector organisations commissioned to deliver interventions); and members of the PHIRST academic team. The data were collected between May 2024 and March 2025. The study was approved by the Research Ethics Committee of the University of Bristol Faculty of Health Sciences (ref: 17879). The COREQ criteria for reporting qualitative research were observed in the reporting of this study.¹⁵

Local government participants were identified and contacted by the Insight researcher who led each of the 10 studies and asked if they would be interested in taking part. If they agreed, their contact details were shared with a member of the Insight team undertaking fieldwork for this study but who had not been involved in the original Insight study with them. Participants were sent a detailed information sheet (PIS) containing information about the study, its aims, confidentiality, use of data, and expected outputs. All participants signed an online consent form for the study prior to interview.

The lead academic researcher(s) from each of the 10 studies were also invited to participate, and a PIS shared and consent form was completed in the same way. Again, interviews were undertaken by a researcher who had not been involved in the original Insight study.

A topic guide for local government respondents was developed and piloted which included questions on applying to the PHIRST scheme, experience of working with Insight, co-production of the study, public involvement, and the impact of the study. A similar guide was developed for academic participants, which omitted questions about applying for PHIRST support and included questions about working with local government partners within the context of PHIRST. Data collection was undertaken by TJ, HR, GW, CF, HL, JH and JM, all public health academics trained in qualitative research. All had previously worked on PHIRST Insight studies with the exception of GW. Data collection was conducted mostly on MS Teams, although a small number of interviews with academic researchers were conducted face-to-face. Interviews lasted between 40 and 90 minutes.

Interviews were recorded, transcribed verbatim, and analysed using the Framework method,¹⁶ a thematic analysis approach. The lead researcher (TJ) developed a draft analytical framework that included the key themes identified in the data and relevant to the research questions. This analytical framework was reviewed by other members of the research team (GW, who had no involvement in PHIRST prior to commencing this work and HR, who had undertaken a small amount of work for one of the 10 Insight studies) and revised until agreement was reached. A systematic approach to data management was adopted, with the lead researcher coding the transcripts into the analytical framework using NVivo software.¹⁷ The analysis continued using Framework matrices as a detailed and accessible overview of the data populating each theme from every interview.

Results

In total, 25 participants took part in the study. Thirteen were from local government or third sector partners, and 12 were academic researchers. All 10 Insight studies had at least one local partner and researcher involved interviewed for this study; some researchers had been involved in multiple studies (see Table 2).

Findings are reported under the following themes; a) reasons local authority teams applied for PHIRST support, b) developing research partnerships, c) co-production, public and community engagement, involvement and participation, and d) outputs and impact. Local authority respondents (including those from third sector organisations) are identified as ‘LA’ when quotations are used, academic researchers as ‘R’. The numbers allocated to participant quotes bear no relation to the order in which studies are presented in Tables 1 and 2.

Discussion

This review of the first five years of working with local governments through the PHIRST scheme has demonstrated that, despite the challenges of the scheme, it has been possible to develop strong and successful research partnerships and co-produce studies that have had an impact on policy, practice, and local government research capacity.

Across the ten Insight studies, the most common reason local government sought external research support was to evaluate the effectiveness of interventions. Many of these would not have been evaluated without PHIRST support, with local government staff citing barriers to evaluation also seen in other studies, including lack of time, staff, and financial resource.^5,6 The PHIRST scheme has helped address these barriers, as well as removing the potentially burdensome tendering process for those local government teams who may have considered commissioning a research study from an external partner. This review also supports earlier studies in the finding that local government staff perceive evaluations conducted by external partners as providing greater rigour and credibility, and hence more likely to have influence and impact than those conducted ‘in house’.¹² Furthermore, this study finds evidence that local governments seek partnerships with academia as a means of building their profile as research-engaged and evidence-based organisations.

The identified disadvantages of working with an external research team include that they may be unfamiliar with the local context, politics, public health challenges and staff involved in commissioning and delivering the intervention. The allocation of PHIRST projects is important here, as some projects are allocated some distance from where the PHIRST team is based. This often means that it is not possible to rely on pre-existing networks. As a result local government staff members have had to allocate considerable time and resource at the start of PHIRST studies to bring the academic team ‘up to speed’. Where feasible, the Insight team travelled to meet local teams in person, which was valued by local staff. Nevertheless, the time and resource required to explain intervention rationales, identify and share potential sources of data for studies, develop logic models, support public involvement and agree the scope of each study might be reduced if PHIRST studies were allocated with greater regard to existing relationships between PHIRST academic teams and local governments. Such an approach would also reduce travel time and costs of in person meetings and fieldwork and make the work more sustainable from an environmental perspective. This would also support better and more inclusive PCIEP, aiding the recruitment of public partners who are more representative of the population of interest. A further constraint identified across these first ten studies is that intervention timescales are frequently misaligned with local governments’ applications for support from PHIRST, often impeding the collection of baseline data and/or the collection of long-term outcome measures that would facilitate impact and cost-benefit evaluations. Finally, the research governance constraints faced by academic teams, in particular gaining ethical approval and agreeing data sharing arrangements, is frustrating for local governments seeking timely evidence to support local commissioning and practice.³

This study identified several factors that have facilitated successful research partnerships between the Insight team and their local partners in such circumstances. These include Insight researchers facilitating regular, transparent and responsive communication; taking responsibility for project management; proactively researching local factors likely to influence the interventions under evaluation; good interpersonal skills; and accommodating local staff changes.

PHIRST Insight studies have been co-produced, with researchers working in collaboration with those involved in commissioning, delivering, and receiving the intervention, including members of the public. Oliver et al.’s review of the literature sets out four main arguments for coproducing research.¹⁸ There is a normative argument that co-production has intrinsic value, supporting the production of research in the public interest, and that sharing power and expertise is ethical. Certainly, there is evidence from this study that local partners perceive that Insight studies have been ‘done with’ rather than ‘done to’ them, supporting their engagement. Co-production can improve the quality of research by helping researchers and policymakers develop holistic understandings of the context, issues, and interventions designed as potential solutions. This has been particularly valuable to Insight researchers who have sought to engage with and learn from local partners more familiar with the context and intervention, and has supported the development of relevant research questions and better research design (in particular, by co-producing data collection tools with public partners). Co-production may also help to increase the impact of research by building research capacity among non-academics and building engagement with research users. There is evidence in this study of local government staff improving their monitoring and evaluation skills, and increasing their motivation to engage with research in the future. The fourth argument for co-production is the intention for research to be inclusive and empowering, increasing a sense of ownership and thus making research findings more likely to be utilised.¹⁸ While there is some evidence of this from our study amongst local government staff, it is less clear that Insight studies have been successfully inclusive of the target populations of each of the ten interventions. While public partners supported co-production in all studies, there is some scepticism amongst local government staff that the recruitment methods used resulted in the engagement of truly representative public members. Again, this is likely the result of the challenges of working remotely from the locations where interventions are being evaluated and the pace at which research protocols need to be developed within the PHIRST scheme. In addition there was little evidence of public involvement in the design or commissioning of any of the ten interventions evaluated by Insight, meaning that researchers were building public involvement from scratch.

Oliver et al. also identify one of the challenges of co-production as developing study teams that are genuinely devoid of power imbalance and decision-making that is not dominated by particular dominant voices or interests.¹⁸ While we found little evidence in this study that any of the respondents, either within the research team or from local government partners (including public partners) felt disempowered, we did find evidence that some local partners did not want to co-develop research design for fear of damaging the integrity and independence of studies so valued by local government staff. This meant that the research design was led in large part by the research team rather than being genuinely co-produced. Furthermore some local partners were new to logic model development and required significant support from the research team which likely led to some power imbalance. A further risk of co-producing research is disagreement between the research team and local partners over the research findings, how they are framed and/or disseminated, with associated risks for research credibility.¹⁸ Some of the studies revealed unexpected findings or indicated that interventions were ineffective, and researchers had to be mindful of the reputational risk to local government partners while preserving research integrity. The development of supportive and trusted relationships between the research team and local partners has helped mitigate this challenge, although it remains a risk for future PHIRST studies.

We find evidence in this study that Insight studies have successfully provided evidence that has informed public health policy and practice in local government and more widely. This has been facilitated by early discussions with local partners about the impact they hope to achieve, who or what they hope to influence, and the format of research outputs most likely to achieve these aims. The co-production of DIICE plans is valued by local partners, and outputs have been produced in a range of accessible formats including short research and policy briefings, research reports, and infographics. Many of these were co-authored with local government staff and public partners. The production of peer-reviewed papers, while an important output for academic researchers, is also valued by most local partners as a means of building their reputation as an evidence-based organisation, although there is some scepticism about the utility of these outputs in communicating research findings effectively and accessibly. Reported impacts from Insight studies have included using evidence to inform commissioning decisions (including in local governments not involved in the original PHIRST study), influencing wider programmes of work (e.g., local travel plans) and influencing national policy. Barriers to impact remain, which include the resource and financial constraints on local government and rapidly changing public health priorities. Respondents in this study also identified that research outputs are more likely to have an impact if they make clear and feasible recommendations to local government, something that PHIRST teams should be mindful of in the future.

We make the following recommendations. We believe that the central NIHR PHIRST team should support local government (and equivalent organisations in the devolved nations) to make timely applications to the PHIRST scheme to ensure that the period of evaluation aligns with intervention timescales. We also advocate that where possible allocating studies with better regard to the geographies of PHIRST teams would reduce the burden for research teams and local partners in scoping and developing research protocols and aid in reducing the carbon footprint for research. As an academic team Insight should continue to take responsibility for project management of PHIRST studies but be mindful of the impact of this on co-production and shared decision-making. Local government partners often need support to understand and develop logic models and Insight has developed more detailed guidance on this that will be shared in future studies. Researchers should also encourage local staff to fully engage with the co-development of research design and methodology in ways that do not risk the integrity or credibility of the study. Researchers should, where possible, maximise efforts to reduce the time taken to achieve ethical approval and agree data sharing arrangements for studies. We also identified here a need to improve the recruitment of local public partners to ensure a more diverse, inclusive and representative sample. Finally, Insight should further explore with local partners the production of more creative, audio-visual outputs that can be hosted on social media platforms.

This study is limited by the involvement of most of the authors in the PHIRST Insight studies, with the exception of GW. We have tried to mitigate this by interviews with participants being conducted by researchers who were not involved in each of the ten individual studies, and the oversight by GW of the analysis and interpretation. The study was strengthened by the recruitment of at least one researcher and a local government partner for each of the ten studies.

This review of the first five years of work as a PHIRST team, co-producing studies of a diverse range of local interventions, finds evidence that local government teams perceive a range of benefits from PHIRST support and that Insight studies have met the aim of the PHIRST scheme in providing timely evidence to inform local government practices or policies. Adopting the findings and recommendations made in this review is likely to support the maintenance of this success, improve the experience of local government partners supported by the PHIRST scheme, and maximise the impact of studies as the Insight team commences a further five years of work.