Enhancing understanding of interventions to increase relational continuity in general practice: a realist review protocol

Step 1: Locating existing theories

The first step involves identifying existing theories and frameworks that explain how RCC is achieved and sustained in general practice. This foundational work supports the development of an initial programme theory (IPT), which will guide the realist review. To inform this, we engaged our ES group — including researchers, GPs, and policymakers — and our PPIE group. These consultations provided valuable insights into current practice, perceived challenges, and assumptions about what enables or hinders RCC.

Informed by this engagement, we reviewed key literature identified by the ES group, including the scoping review by Fox et al⁷ and analytical articles by Pereira Gray et al¹³ and Sidaway-Lee et al.¹⁴ These sources helped surface existing theoretical models and frameworks for RCC interventions (see Figure 1). Using this evidence and feedback from both the PPIE and ES groups, we developed and refined the IPT (Annexe 1), which outlines candidate mechanisms — such as relational, informational, and longitudinal continuity — and contextual factors, including practice structure, team composition, and patient characteristics.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Initial programme theory: a synthesis of key initiatives. EHR = electronic health records. MDTs = multidisciplinary teams. RCC = relational continuity of care

The IPT also considers intended and unintended consequences, recognising that interventions may support RCC in some contexts but disrupt it in others. It will be iteratively tested and refined as the review progresses, supporting a nuanced understanding of how RCC functions across diverse settings.

Step 2: Searching for evidence

We will search for literature to inform the development and refinement of the IPT created in step 1. Systematic searches will be conducted in MEDLINE, Embase, and CINAHL to identify peer-reviewed articles on RCC interventions in general practice. Search strategies will combine free-text terms and subject headings (for example, MeSH) such as ’relational continuity’, ’continuity of care’, ’general practice’, ’primary care’, ’patient–provider relationships’, and ’healthcare interventions’. No study design restrictions will be applied, allowing inclusion of qualitative, quantitative, and mixed-methods studies.

In addition to database searches, we will review the grey literature using Overton and organisations such as the NHS, Department of Health and Social Care, Royal College of General Practitioners (RCGP), The Health Foundation, and The King’s Fund, alongside relevant policy reports and conference proceedings. Citation tracking of key papers will be undertaken, and the PPIE and ES groups may suggest additional sources. This comprehensive strategy will ensure the capture of relevant interventions and outcomes. Results will be imported into EndNote21 for deduplication, then transferred to Rayyan software¹⁵ for title and abstract screening.

Inclusion criteria

• Articles will be included if they provide information on the description, contexts, mechanisms, or outcomes of RCC interventions, specifically in UK general practice and in settings similar to UK primary care, according to the OECD and The Health Foundation report.¹¹

• Articles published in English from inception of individual databases to 2024 will be included.

Exclusion criteria

• Articles focusing on non-primary care settings.

• International studies pertaining to countries with systems that are significantly different from UK primary care.

• Studies that do not address RCC.

Step 3: Article selection

Unlike narrative reviews and meta-analyses, realist reviews prioritise understanding generative causation through retroductive theorising over methodological quality.¹⁶ Following the literature search, a three-step screening process will be applied. First, reviewer 1 will screen titles for relevance. Remaining articles will then be assessed by title and abstract to ensure they address relational continuity in primary care. Rayyan software will support this screening. Where abstracts are unclear, full texts will be reviewed.

Reviewers 1 and 2 will screen all abstracts. To ensure inter-rater reliability, reviewer 3 will independently assess a random 10% sample. Reviewers 1 and 2 will then evaluate all relevant full texts for inclusion, based on their discussion of RCC, related interventions, or outcomes in general practice. The final set of included articles will be uploaded to NVivo (version 11) for storage and coding.

Disagreements or uncertainties about inclusion will be resolved through team discussion. Final inclusion will be guided by each document’s relevance, richness, and rigour in relation to the developing programme theory.¹⁷

Step 4: Data extraction

Once the final set of articles is selected, full texts will be uploaded to NVivo (version 11) for coding. Data will be extracted inductively (emerging from the data), deductively (guided by the initial programme theory), and retroductively (to identify underlying causal mechanisms). The aim is to capture the contexts in which RCC interventions are implemented, the mechanisms through which they operate, and the outcomes they produce.¹⁸

The realist review is an iterative process, with searching, appraisal, and synthesis occurring concurrently and contributing to ongoing programme theory development. As the theory evolves, further data will be collected to test emerging concepts and uncover causal mechanisms, which may lead to refinement of previously extracted data.

Extracted data will include details of intervention components, provider and patient experiences, system-level impacts, and contextual factors affecting intervention effectiveness. Coding will also be iterative, adapting as evidence accumulates. As with article selection, a subset of data extraction decisions will be reviewed by two team members to ensure rigour and consistency.

Step 5: Synthesising evidence and drawing conclusions

We will synthesise the extracted data using a realist logic of analysis to develop CMO configurations. These configurations will explain how and why RCC interventions succeed or fail in different contexts. For example, we will compare evidence from settings where RCC interventions have improved patient satisfaction and health outcomes with those where interventions have been less successful, aiming to understand the factors that contribute to these differing outcomes.

The synthesis will use cross-case comparison to explore the influence of various contextual factors, such as practice organisation, patient demographics, and the composition of multidisciplinary teams. The outcome of this analysis will be a refined programme theory, explaining how RCC interventions function in general practice and suggesting ways they can be tailored to different settings to maximise their effectiveness.¹⁹

Finally, the review will conclude with recommendations for practice and policy, offering actionable insights on how to design, implement, and adapt RCC interventions to improve relational continuity in general practice. The refined programme theory will guide these recommendations, and it will be shared with healthcare practitioners, policymakers, and researchers to enhance the sustainability and effectiveness of RCC interventions in primary care.

Stakeholder and patient and public involvement and engagement (PPIE)

The review will engage GPs, patient representatives, policymakers, and administrators identified through expert panel recommendations. Stakeholders will help refine the programme theory and ensure the review addresses practical and policy-relevant concerns. Patients with and without experience of RCC will contribute, ensuring diverse perspectives. An equality of opportunity approach will guide the selection of public contributors to maximise involvement. PPIE group members will shape the programme theory, contribute to findings, and help develop accessible outputs such as documents and infographics. They will also play a key role in creating a dissemination strategy and ensuring solutions are both feasible and acceptable in real-world contexts.