Collaborative action for person-centred coordinated care (P3C): an approach to support the development of a comprehensive system-wide solution to fragmented care

Health systems are responding actively to policy imperatives for ‘integration’ and person-centred care, but these efforts are commonly piecemeal, untested, unevaluated and carried out in isolation, often in response to local leaders or crises. In the United Kingdom, whole system redesign has also begun to gain traction and, with this, the recognition that ‘culture’ change towards more integrated or coordinated care, which is also person centred, requires support and energy spanning different organisational levels, i.e. from individual practitioner to multi-professional team to whole provider-organisation, commissioner (payers) and beyond [5, 42]. Without concurrent and rigorous formative evaluation [43], the learning from these innovations is often short lived or ignored. Furthermore, with little understanding of how the context influences implementation, opportunities to create a wealth of insight are often lost. Consequently, ‘whole system’ implementation knowledge is well timed and requires the appropriate melding of traditional approaches to research (e.g. large randomised controlled trials, etc.) with those that can flex and explore the role of context (e.g. quality improvement, realist evaluation, action research). In spite of this, there are few comprehensive programmes that aim to (1) support the implementation of evidence-informed practice (where it exists) and (2) develop practice-based evidence about what works.

We have set about bringing together a ‘whole systems’ realist evaluation [44] incorporating service redesign, implementation, education, evaluation and research. We are using co-designed ‘Collaborative Action’ [45] in the context of the South West Peninsula Collaboration for Applied Health Research and Care (SW CLAHRC). Our approach shares much with that of Glasgow et al.’s [46] ‘Evidence Integration Triangle’, namely, prioritising actionable feedback to practice, a participatory implementation process, the shared understanding by participants of key components, and practical and ongoing measurement within a multilevel context. We hope that this region-wide initiative of collaborative groups (researchers, commissioners, providers and practitioners) will continue to bring a clarity of purpose and continued use of our common evaluation framework. Through this supportive process we are putting P3C into practice locally whilst also providing a receptive context to support more research, thus contributing to a coherent body of knowledge.

The Five Year Forward View developed by NHS England presents a radical vision of reformed and integrated public services working alongside communities, social networks and the voluntary sector to support people living with long-term conditions [39]. A number of national programmes with local initiatives (Integrated Personal Commissioning programme [47], Vanguards [48], Realising the Value [49]) aim to embed person- and community-centred approaches to deliver this vision. To support these local initiatives, we have developed a taxonomic framework for P3C [41, 50, 51], and tools to monitor and develop this at a practice level [41, 51]. This framework has been developed from our multidimensional definition of P3C, which is presented in Table 1.

Using ‘Collaborative Action’ [45] and our evaluation framework, we are supporting the implementation of P3C. This involves expert practitioners, managers and researchers working together to help ensure progress in service provision by investigating how, where and why models work. This strategy presupposes that valuable knowledge can be derived from the evaluation of innovations. It also entails accepting that, while we can inform practice by what we do know (using evidence where we have it), we (researchers, practitioners, managers, patients and their carers) also acknowledge what we do not know, and are ready to try things out and create opportunities for experiential learning (including formalising already existing tacit or undocumented knowledge). This ‘Collaborative Action’ is depicted in Fig. 1 (where the arrows represent a flow and exchange of different types of knowledge or activities).

This collaborative approach facilitates the transfer and synthesis of different types of knowledge (i.e. published evidence and that developed from local practice) between specific local and wider national/international settings. Generating knowledge this way has the potential to make research more relevant to local ways of delivering care, i.e. by generating more practice-based evidence [52]. These strands are then synthesised and fed back into an emerging understanding of what P3C is and how to support its implementation. This knowledge can also help develop an overarching causal framework (programme logic) to track how changes (e.g. team working, practitioner-patient interactions, etc.) occur and how they relate to improvements in outcomes (Fig. 2).

Generating useful knowledge and improving practice models is facilitated by the use of a similar evaluation framework across settings. Based, therefore, on the changes anticipated in the overarching programme logic (Fig. 2), we are using a combination of co-selected metrics across core domains of interest (e.g. practitioner and patient experience [51, 53], patient activation [54], patient well-being [55], morbidity, mortality, cost and organisational processes [41]). Local services benefit from the results of the evaluation as these are used to shape the ongoing development of the model. We have found that supporting the development of specific evaluation frameworks for projects based on their objectives and targets improves quality, competency and ownership. Supporting health economies to develop the core measurements (e.g. cost of care, admission rates, experience of care, patient reported outcomes) and embed ways to collect data is helpful to monitor and understand the whole system. This will help create the knowledge that we need about how local care providers produce – or fail to – continuity of care for people with multiple chronic health problems. Crucially for P3C, this approach helps identify the mechanisms that explain how causal relationships occur (through the interaction of people’s reasoning and the resources available to them) [56, 57].

Over the past 2 years, our collaborative effort has involved engagement and working with diverse local sites (Fig. 3) using a consistent multi-level and multi-perspective evaluation framework (Fig. 4) for the evaluation of specific P3C innovations and system change. This framework reflects the complex interventions and organisational level changes that occur during P3C service reconfigurations. More importantly, however, it captures and uses the voice of patients and professionals to shape redesign efforts. For this purpose, we also collect multi-perspective data at both individual and group level (Fig. 4).

This evaluation framework assesses if service change and developments are achieving pre-specified outcomes towards P3C. Using a three tiered and multi-perspective longitudinal mixed methods approach, the framework encompasses qualitative data, questionnaire data and service use and data (Fig. 4). Service use data capture allows for the assessment of patterns of service contact (GP attendance, hospital admissions, etc.) following exposure to an intervention with the intention to derive cost calculations. Staff experience data provides an insight into their perspective, whilst individual patients’ perspectives are sought to assess if care has improved or worsened [51], and if there are any perceived improvements in health, wellbeing, loneliness and self-management. We have included loneliness and self-management as measurement dimensions as these were hypothesised as important to the target population of older people at risk of isolation with long-term conditions. This framework is being used in the English counties of Somerset and Devon (Fig. 3) and was co-designed with local system commissioners and providers in each of the three Clinical Commissioning Groups. This is being used both to monitor the system as a whole over time, and to evaluate particular initiatives and innovations. Generalisable knowledge is derived firstly through reviewing literature, using realist synthesis where feasible, and by comparison of data across the evaluation sites focusing on practitioner-patient interaction and organisational processes. We are also examining the process of implementation, both through the lens of existing knowledge about changing practice and quality improvement in the NHS, and also identifying specific organisational constraints and facilitators for developing system-wide P3C.