Background Interorganisational collaboration is currently being promoted to improve the performance of NHS providers. However, up to now, there has, to the best of our knowledge, been no systematic attempt to assess the effect of different approaches to collaboration or to understand the mechanisms through which interorganisational collaborations can work in particular contexts. Objectives Our objectives were to (1) explore the main strands of the literature about interorganisational collaboration and to identify the main theoretical and conceptual frameworks, (2) assess the empirical evidence with regard to how different interorganisational collaborations may (or may not) lead to improved performance and outcomes, (3) understand and learn from NHS evidence users and other stakeholders about how and where interorganisational collaborations can best be used to support turnaround processes, (4) develop a typology of interorganisational collaboration that considers different types and scales of collaboration appropriate to NHS provider contexts and (5) generate evidence-informed practical guidance for NHS providers, policy-makers and others with responsibility for implementing and assessing interorganisational collaboration arrangements. Design A realist synthesis was carried out to develop, test and refine theories about how interorganisational collaborations work, for whom and in what circumstances. Data sources Data sources were gathered from peer-reviewed and grey literature, realist interviews with 34 stakeholders and a focus group with patient and public representatives. Review methods Initial theories and ideas were gathered from scoping reviews that were gleaned and refined through a realist review of the literature. A range of stakeholder interviews and a focus group sought to further refine understandings of what works, for whom and in what circumstances with regard to high-performing interorganisational collaborations. Results A realist review and synthesis identified key mechanisms, such as trust, faith, confidence and risk tolerance, within the functioning of effective interorganisational collaborations. A stakeholder analysis refined this understanding and, in addition, developed a new programme theory of collaborative performance, with mechanisms related to cultural efficacy, organisational efficiency and technological effectiveness. A series of translatable tools, including a diagnostic survey and a collaboration maturity index, were also developed. Limitations The breadth of interorganisational collaboration arrangements included made it difficult to make specific recommendations for individual interorganisational collaboration types. The stakeholder analysis focused exclusively on England, UK, where the COVID-19 pandemic posed challenges for fieldwork. Conclusions Implementing successful interorganisational collaborations is a difficult, complex task that requires significant time, resource and energy to achieve the collaborative functioning that generates performance improvements. A delicate balance of building trust, instilling faith and maintaining confidence is required for high-performing interorganisational collaborations to flourish. Future work Future research should further refine our theory by incorporating other workforce and user perspectives. Research into digital platforms for interorganisational collaborations and outcome measurement are advocated, along with place-based and cross-sectoral partnerships, as well as regulatory models for overseeing interorganisational collaborations. Study registration The study is registered as PROSPERO CRD42019149009. Funding This project was funded by the National Institute for Health and Care Research (NIHR) Health and Social Care Delivery Research programme and will be published in full in Health and Social Care Delivery Research; Vol. 11, No. 6. See the NIHR Journals Library website for further project information. Plain language summary What was the question? How can the collaboration between health-care providers be improved? What did we do? There are continuing concerns about poor patient care across the NHS. One of the ways in which governments have tried to solve these issues is by getting services to work together, rather than separately, to solve any problems. The aim of our review is to learn about how, why and when different approaches to working together – which we call interorganisational collaboration – can be used to improve the performance of NHS providers. We reviewed published evidence and carried out interviews with NHS staff. We also carried out interviews and a focus group with patient and public representatives. What did we find? Our review finds that interorganisational collaborations can work well when a series of elements are in place, which includes the need to build trust between everyone involved. Having a belief in the collaboration is also needed to help inspire others to get involved. To try and reduce possible problems, setting priorities and having clear methods to show how improvements can be achieved are important, as well as having an agreed contract in place to ensure that any conflicts are resolved. If done well, collaboration can improve resource allocation, coordination, communication and shared learning about best practice. What does this mean? Our review provides valuable evidence of how different approaches to interorganisational collaboration can be used by NHS providers to work together to improve services in different situations. Our review provides different options for organisations to reflect on how well they are collaborating, which includes the involvement of key stakeholders, such as patients, the public and communities. Scientific summary Background Improving the collaboration and integration of services has become a mantra for health-care systems. Interorganisational collaborations (IOCs), such as alliances, groups, associations, networks and mergers, have been closely linked to policy contexts where governments have promoted collaboration as a solution for meeting the innovation, coordination, efficiency and quality challenges currently being faced. A variety of factors have been attributed to achieving success within such initiatives, including the importance of organisational capacity, having a shared vision, building trust and collaborative leadership. However, realising the advantages of collaboration is far from straightforward, with notable barriers including the influence of historical events, competitive behaviour, the regulatory environment and a lack of organisational resources. Despite the burgeoning evidence base and increased policy emphasis on collaborative working, notable gaps in knowledge persist. As a result, our understanding of the mechanisms and processes for spreading and sustaining evidence about how IOC relationships work in practice is limited. Many questions remain about how interorganisational arrangements work, for whom and in what circumstances. Given the complexities of collaborative arrangements, contributions identify how ‘theories of change’ approaches provide a way to assess how collaboration synergies are shaped by contexts, behaviours and structures. Realist approaches to the study of IOCs are advocated; however, applications within health-care settings have, hitherto, remained an underdeveloped area. Objectives The research has the following inter-related objectives: •To explore the main strands of the literature about IOC and to identify the main theoretical and conceptual frameworks that can be used to shed light on the conditions and antecedents for effective partnering across sectors and stakeholders •To assess the empirical evidence with regard to how different interorganisational practices may (or may not) lead to improved performance and outcomes •To understand and learn from NHS evidence users and other stakeholders about how and where IOC can best be used as a mechanism to support turnaround processes •To develop a typology of IOC that considers different types and scales of collaborative ventures that are appropriate for particular NHS provider contexts •To generate evidence-informed practical guidance for NHS providers, policy-makers and others with responsibility for implementing and assessing IOC arrangements in the NHS. Methods A realist methodology is employed to provide useful intelligence regarding how, why and in what circumstances different approaches to IOC can improve the performance of NHS provider organisations. Data sources Given the large, multifaceted and complex nature of IOCs, an ‘initial rough theory’ was developed by combining a review of grey and narrative literature, along with systematic reviews of evidence, to capture key definitions, typologies, ingredients and outcomes. Subsequent systematic searches were conducted to gather evidence about how IOC works and the contextual factors shaping a range of entities, such as alliances, buddying, mergers, acquisitions and hospital groups. Searches were run between 7 October 2020 and 4 March 2021 on databases, including the Healthcare Management Information Consortium, MEDLINE, Social Policy and Practice and PsycINFO, and on Google Scholar (Google Inc., Mountain View, CA, USA). Reference-scanning and citation-tracking were also employed. We conducted a realist evaluation to further test our refined programme theory by exploring the experiences of a range of stakeholders, comprising the leaders or architects of IOCs, regulators, policy-makers, professional bodies, front-line staff and patient representatives. Inclusion criteria The realist review used the following inclusion criteria for the title and abstract stage: ‘the paper clearly relates to collaborations between one or more public sector organisations on either a structural or individual level’ and ‘the paper is a case study, evaluation, opinion, or review’. Full-text screening also included ‘propositions about the success or failure of collaboration in the public sector, mechanisms underlying how collaboration works or include information about “entry points” (i.e. drivers of collaboration)’. For the refinement stage, we included papers that (1) were case studies or evaluations (defined as reporting results of arrangements using descriptive methods), (2) report on an IOC between health care-providing organisations and (3) were in English (because of resource limitations of the study). A purposive sampling strategy identified participants through contacts via our Study Advisory Group and from direct contact with potential individuals and organisations identified through scoping work. Participants were chosen based on their likelihood of being able to provide rich information about various aspects of IOCs from being engaged in formulating, influencing, implementing or experiencing such arrangements. The final sample comprised 37 interviews and one focus group with 8 patient and public representatives. Data extraction Selected studies were subject to rigour and relevance checks in accordance with realist synthesis methodology. The screening for rigour was ongoing and primarily involved including context–mechanism–outcome configurations (CMOCs) only when supported by (1) clear data in included studies and (2) multiple sources. For theoretical sources of evidence, only theories that had seen significant use in the literature since publication were used in the building of our middle-range theory (MRT) and CMOCs. No studies or extracts were excluded on the basis of trustworthiness. Analysis of the realist evaluation interview data was performed in NVivo 12 software (QSR International, Warrington, UK) by one coder (JAA), with the coding logic independently verified by a second coder (RM). Coding was performed retroductively, combining inductive and deductive logic. Data synthesis Theory gleaning synthesised document evidence according to whether or not the evidence shed light on entry points into collaboration, contextual factors, mechanisms or other elements relating to collaborations that helped elucidate the underlying ideas and assumptions regarding how collaboration was intended to work. Theory refinement aimed to test the identified CMOCs against case studies, and improve our programme theory. The realist interviews and focus group provided further refinement to the CMOCs relating to collaborative functioning, as well as to glean novel CMOCs relating to collaborative performance. Interview data were retroductively analysed in NVivo 12. Results The realist synthesis incorporated reviews, MRTs, case studies and organisational evaluation literature. A total of 86 papers produced a refined realist theory that surfaced the inter-related roles of trust and risk tolerance, faith, task complexity, interpersonal communication, cultural integration, perception of progress, etc., and how these roles causally interact to drive collaborative behaviour. The results demonstrate that in mandated or highly integrative collaborations the locus may be shifted from trust towards contractual obligation and a sense of confidence that the partner will act collaboratively. These chains of CMOCs were situated within a ‘causal web’ to depict how distant contextual items and their mechanisms work to affect the outcomes underpinning organisational performance. Stakeholder interviews supported the CMOCs identified within the review. In doing so, stakeholders further articulated how building and sustaining trust was connected to the leadership skills and behaviours of authenticity, empathy, visibility and generosity. A commitment to place-based approaches also featured, along with the importance of stakeholder engagement, data analysis and project management. The findings also show how a delicate balance is required for building faith, where energising leadership is tempered by the stark capacity issues facing current NHS contexts. The importance of priority-setting and data analytics features in building faith; however, increasing task complexity can reduce faith, particularly when working across boundaries. Interviews also stress the need for confidence and memoranda of understanding in particular types of IOC. The results present the first comprehensive realist evaluation of how well-functioning IOCs can drive performance improvements. Drawing on the domains of collaborative performance, the interviews and focus group identify how cultural efficacy mechanisms prove to be particularly important in driving improved communication, better coordination, shared improvement strategies and reputation management. Organisational efficiency mechanisms highlight the causal links between collaboration and improving financial and workforce resource allocation, as well as better coordination to increase responsiveness and reduce duplication. Technological effectiveness sheds light on the benefits collaboration can bring for research and development and working across clinical pathways. Conclusions Through analysis of theoretical, empirical and stakeholder evidence, the research presents a synthesis of MRTs and CMOCs to better understand how, why and in what circumstances IOCs are effective for NHS providers. Our study finds that the core mechanisms of collaborative functioning comprise the development of trust, faith and confidence. The extent to which task success or failure is achieved is mediated by supporting mechanisms related to capacity, legitimacy, complexity, conflict and risk tolerance. Performance improvement from collaboration can be achieved when mechanisms underlying organisational efficiency (e.g. reduced duplication of effort), cultural efficacy (e.g. enhanced coordination in local health system) and technological effectiveness (e.g. sharing clinical expertise) are activated. The findings conclude that performance improvements occur in a context of collaborative functioning, which, in turn, drive improvements in long-term outcomes, including care quality, safety, efficiency and experience. The findings provide a range of practical steps that organisations can take in the development of IOCs, including the development of diagnostic surveys for assessing collaboration to help organisations assess their readiness for collaboration, as well as for diagnosing collaborations that are already progressing. A pilot of a survey with a mental health provider collaborative shows promising signs for its utility in providing a valuable means for stimulating discussion with regard to the perceived readiness for collaboration. A maturity index for collaboration was also presented to assess levels of collaboration and encourage critical discussion and reflection. A range of theoretical, empirical and policy implications arise from this research. Specific issues that warrant further consideration and investigation are as follows: •Where much of the analysis of IOCs has captured the experiences, processes and outcomes from the perspective of those leading programmes and initiatives, further research is needed to gather workforce perspectives regarding how new processes are understood and operationalised, and how IOCs shape patient and user interactions. •Research is required to better grasp how IOCs can engage and improve population health by further involving patients and communities through drawing on principles of co-design and co-production. •The COVID-19 pandemic has been a driver for activity using digital platforms for communication, yet further research is needed to better understand and nurture ‘interpersonal communication’ across digital platforms and to better understand the role of digital technology in facilitating collaboration. •Further research is needed to investigate the applicability and adaptability of a number of the elements raised by this project, such as the roles of faith, trust and other mechanisms within the formation and maintenance of place-based partnerships. Learning from other national contexts could facilitate such efforts, with further comparative studies of IOCs from across the United Kingdom and beyond. •A review of regulatory models and perspectives for overseeing collaborative ventures is required, learning from other sectors and health-care contexts, where appropriate. •Building on our realist theory of collaborative performance, further research is needed to disentangle the motivators and drivers from the ‘outcomes’ associated with IOCs. Such analysis can support the current policy landscape, placing greater emphasis on measuring the outcomes and social value generated from collaborative working. •Further research is required to articulate the cross-sectoral relationships within the current IOC policy agenda. The place and positioning of social care and third sector requires further development. Furthermore, greater attention to the role of public/private partnerships, and the private sector more specifically, within collaboration and integration agendas is needed. Study registration The study is registered as PROSPERO CRD42019149009. Funding This project was funded by the National Institute for Health and Care Research (NIHR) Health and Social Care Delivery Research programme and will be published in full in Health and Social Care Delivery Research; Vol. 11, No. 6. See the NIHR Journals Library website for further project information.