Creating and facilitating change for Person‐Centred Coordinated Care (P3C): The development of the Organisational Change Tool (P3C‐OCT)

Abstract Background Person Centred Coordinated Care (P3C) is a UK priority for patients, carers, professionals, commissioners and policy makers. Services are developing a range of approaches to deliver this care with a lack of tools to guide implementation. Methodology A scoping review and critical examination of current policy, key literature and NHS guidelines, together with stakeholder involvement led to the identification of domains, subdomains and component activities (processes and behaviours) required to deliver P3C. These were validated through codesign with stakeholders via a series of workshops and cognitive interviews. Results Six core domains of P3C were identified as follows: (i) my goals, (ii) care planning, (iii) transitions, (iv) decision making (v), information and communication and (vi) organizational support activities. These were populated by 29 core subdomains (question items). A number of response codes (components) to each question provide examples of the processes and activities that can be actioned to achieve each core subdomain of P3C. Conclusion The P3C‐OCT provides a coherent approach to monitoring progress and supporting practice development towards P3C. It can be used to generate a shared understanding of the core domains of P3C at a service delivery level, and support reorganization of care for those with complex needs. The tool can reliably detect change over time, as demonstrated in a sample of 40 UK general practices. It is currently being used in four UK evaluations of new models of care and being further developed as a training tool for the delivery of P3C.


| INTRODUCTION
The current UK fiscal climate is demanding greater efficiency and cost-saving across public sector organizations. The NHS in particular is facing unrivalled challenges to do more with less and deliver better quality and more efficient care whilst reducing deficits. 1,2 It is in this context that a move away from disease-based models towards a more effective, integrated, and person-centred approach is perceived as a way to reorganize service delivery. This is particularly relevant for people with long-term conditions (LTCs), multiple LTCs and people with multimorbidity; the number of which is forecast to rise from 1.9 million in 2008 to 2.9 million in 2018. 3 Person Centred Care (PCC) is an approach to patients that embodies an individual's right to self-determination and highlights their role as an equal partner in the care exchange. 4 Recent work has identified care coordination rather than organizational integration as one of the essential components for the delivery of PCC. [5][6][7][8] This means that changes in the ways professionals work are required, 9 as it cannot be effectively delivered in a system that is confused, fragmented and lacking in continuity. 7,10 In a European context, Ekman et al have provided a guide on how to approach PCC through the development of three routines based around (i) eliciting the individuals narrative, (ii) the cocreation of a plan of care and (iii) documenting this plan within a care plan. 4 Lloyd has expanded this into four routines to fully encompass the needs of those with complex needs; (iv) an agreement to act in conjunction with the person and other professionals to coordinate care. 11 These two key concepts brought together reflect a possible way in which to achieve better outcomes of care for individuals: Person Centred and Coordinated Care (P3C). We define P3C as: Care and support that is guided by and organized effectively around the needs and preferences of individuals. [11] In detail, the following table provides a breakdown of the three elements of the current definition of Person Centred Coordinated Care (Table 1): P3C highlights the patient as an "expert," with access to both individual and environmental resources, and around all of which care should be coordinated. Anchored in the National Voices "I" Statements, 12 P3C places an emphasis on the individual and reflects what is important to them in relation to their care and support needs. This approach holds the promise of improved outcomes and experiences through the setting and attainment of personal goals based on the values and preferences of the individual (elicited through shared decision making). 13,14 The logic therefore follows that this approach produces care and support that is tailored to the individual and is more efficient at reducing waste and duplication.
Implementation of new models that seek to provide more integrated care has been hampered by conceptual confusion and a lack of practical guidance. As a result, this care is rarely delivered or implemented in a consistent manner. 4 The UK House of Care model 15 was developed with the aim of designing a partnership delivery model for Person Centred Care (PCC), encompassing coordinated services and available to all people with long-term conditions. It was established to move away from a single disease-focused reactive system towards a more pre-emptive, holistic view of the person that assigned an active role for patients. Its goal is to drive a whole system approach, based on the understanding that critical elements are required to deliver care in this way. Whilst the model provides a summary of areas where changes are required, few sites within the UK have achieved the implementation of the complete paradigm. 15 This may be in part due to the abstract constructs that abound the policy literature of this area, and which are difficult to implement without the specific detail of processes of change. For example, Coulter et al 16 state that the most robust barrier to the delivery of PCC is cultural change. This is a very real obstacle, but "culture change" within this model is not articulated in a way that identifies the behaviours and processes necessary to change cultures of practice.
A shift towards P3C also brings with it a requirement to measure and guide the development of services whilst considering organizational context (eg rural/suburban/urban) and how this influences the design and configuration of services. 6,17 However, at present there are no comprehensive tools that can achieve this within health and social care settings. 6,17 Our scoping exercise to identify ways in which P3C can be achieved failed to identify guidance that was sufficiently detailed to support implementation. We found evidence of only one co-created quality improvement organizational tool that encompassed an element of PCC. 18 This tool is intended for use in Australian general practice, and given the UK push for integrated health and social care, there remains the need for a tool which can be implemented across a range of services.

Person Centred Care
The cocreation of care between the patients, their family and informal carers, and health professionals. This definition is becoming widely used by many international organizations including the WHO, and has been translated into a proven approach and used at the Gothenburg University Centre for Person Centred Care (GPCC). Person-centred care strives to see an individual as bio-psycho-social whole, as a person and not an illness or a collection of conditions Resources Psycho-social and environmental resources are non-clinical and have a community focus. This is commonly being referred to as "Community-centred approaches" that complement other types of interventions that focus more on individual care and behaviour change or on developing sustainable environments. These approaches acknowledge the importance of social capital for health and well-being to flourish Coordinated Care Care coordination is the deliberate organisation of patient care activities between two or more participants (including the patient) involved in a patient's care to facilitate the appropriate delivery of healthcare services. Organizing care involves the marshalling of personnel and other resources needed to carry out all required patient care activities and is often managed by the exchange of information among participants responsible for different aspects of care. From a person or family perspective, care coordination is any activity that helps ensure that the individual's needs and preferences for health services and information sharing across people, functions and sites are met over time T A B L E 1 Three elements of the current definition of P3C

| Aim
Our aim was to develop a practical tool to support organizations and practitioners to provide personalized and coordinated care for people with multimorbidity. This tool is based in on the principles of promoting person-centred relationships with service users and between practitioners, and highlights how organisations can support its achievement. 19,20  21 Its advancement was iterative and progressed during three phases (see Figure 1).

| Phase Two: Validation and endorsement of components, and testing for relevance and readability
Phase two used codesign and ran concurrently with phases (1) and (3) to validate and endorse the clustering of domains, subdomains and components, and pilot test for readability and content validity.
Literature and policy were repeatedly examined to ensure latest findings continued to be incorporated, and questions were interrogated and adapted (if necessary) to ensure domains and subdomains encom-

| Phase Three: Design of questions and response codes
Phase three ran concurrently with phase 2 and comprised the development of questions and response codes iteratively in response to stakeholder guidance. Configurations of possible components (actions/ behaviours etc.) were explored to ascertain how these acted as potential mechanisms to achieve each subdomain. This process continued until saturation was complete (ie no new configurations could be identified).

| Ethics
Stakeholder involvement and the gathering of pilot data were approved by Plymouth University Faculty Research Ethics Committee.

Patient representatives
Co-design was achieved through two stakeholder workshops in conjunction with the PenCLAHRC Public Involvement Group (PenPIG).
Potential participants were sent an invitatory email and study information sheet to help inform their decision to take part. Consent forms were completed and their right to withdraw explained. All participants were diagnosed with multiple LTCs.
Workshop with Health and Social Care professionals and patients Feedback on question items was received as part of a wider workshop on outcome measures for P3C. Participants were known to the research group and were invited to take part either by email, telephone or in person.

Cognitive interviewing
Participants were recruited either through the academic team (n = 1) or from evaluation work where the tool was being piloted (n = 2). Participants were sent an invitation email and study information sheet to help inform their decision to take part. Consent forms were completed and their right to withdraw explained.

| Development of the P3C-OCT
The  Table 2).

| Scoring the P3C-OCT
The (ii) Clarification of domains/subdomains and components through the identification of the level within which they operate (patient/practitioner/ organisational). (iii) Identification of domains/ subdomains and components as a key action or behaviour, an interaction between people, or an element of organisational support in order to provide a practical understanding of how they achieve the aims of P3C.
Phase 3 Analysis/examination of prototype causal mechanisms and context features resulting in P3C-OCT question and response code design.
Phase 2 Establishing face and content validity through a cocreated analytical and discursive process. According to this scoring mechanism, if all activities are being performed (+10 points for objective) but are "not working," the subjective score will be −10. This results in an aggregate objective + subjective score of 0, so that evidence of an activity (and it not working) has the same score as not implementing an activity at all. See question example in Table 3 below: Once scores have been derived for each question, they can be aggregated to derive a total score for the P3C-OCT. The total score is normalized to 20 so that the overall score is out of a maximum of 20. Furthermore, scores can be derived for only objective components (eg a summary of activity towards P3C) or subjective components (eg how well things are working). Scores can be also derived according to domains/subdomains of P3C, by aggregating questions that correspond to these domains.
All questions follow the above schema, with a number of exceptions where the question format requires an idiosyncratic scoring. This is achieved in the most parsimonious manner. For instance, question 6 has two objective components (the second part is about using personal budgets), and these two components are aggregated (as if they were a single question) so that the maximum objective score remains equal to 10. Nonetheless, the "equally weighted" scoring mechanism has been retained so that all questions still retain a maximum of 10 points for objective and 10 points for subjective.
Feedback is delivered in the form of an interactive "dashboard" of results and a set of instructions to assist practices in its navigation. The  The tool is designed to be completed in the form of a paper/ electronic/or online document. In its current form, it is best suited to completion by organisational managers and clinical/service leads.

No of question items in each domain
Although one or two people may take responsibility for its completion, they will need to gather information from several key professionals (eg GP's, nurses, community matrons) and members of other teams which come together to provide multidisciplinary care. For this reason, questions may be circulated across the relevant professionals, or the tool can be completed as a team.

| Validation
In response to stakeholder workshops and cognitive interviewing, a number of changes were made to the P3C-OCT (see Table 4 for examples). In particular, changes related to ambiguous wording (for example, "care providers" was changed to paid care providers for clarity) and clarification on specific terminology. For example, originally, the tool referred only to a care plan. Feedback suggested that this was too restrictive and ignored the construct of care planning in terms of discussions with patients and colleagues.  T A B L E 3 P3C-OCT example question improvement is significant at P = .034 on a paired t test (preliminary data; full publication of evaluation studies is undergoing preparation).

| Pilot testing
The tool is currently being trialled across a number of South West evaluations. It has also been used to design, analyse and interpret organisational processes towards achieving P3C, and was used as a framework for the construction of questions to elicit the impact of change on practices opting out of QOF. 23

| DISCUSSION
The delivery of person-centred coordinated care has been enmeshed in an environment of conceptual confusion and ambiguous language, resulting in a lack of tangible guidance on its implementation at an organizational level, and difficulties in real-world application. 24,25 Components and domains of P3C identified in the literature range from broad themes to specific actions across domains which become unwieldy when combined into an assessment framework. 26 Although aims of programmes appear similar, that is the reduction in fragmentation and the enhancement of continuity and coordination through the placing of the person at the centre of health-care delivery, 25 the processes through which to achieve these are less so. 27 The P3C-OCT reflects the importance of committing resources to the development of policies and processes and adds to the consensus that multiple components are involved in its successful implementation. 28 Crucially to the delivery of P3C, the tool supports organisations to better understand their own practice 29 and to identify whether "I" statement and House of Care principles are being delivered. 12 The P3C-OCT unpicks the conceptual confusion of how to do P3C at an organisational level and provides guidance in a single toolkit. It is the first comprehensive evidence-based tool that brings together a set of actions and behaviours to achieve the domains/ subdomains of P3C, and which can be implemented as a means to achieve routines evidenced as necessary to its accomplishment. 4

| Further validation of the domains and components of P3C
Currently, we have little evidence for the optimal configurations of behaviours, interactions and system support to produce P3C and further research using the OCT tool or similar approaches is required to address this. Further work also needs to consider fully the impact of contextual features (eg practice size, rural/urban) and their impact on the achievement and implementation of components.

| Understanding and development of how the tool is used in practice
The development of the P3C OCT will also benefit from widespread in-practice use. Given the wealth of data collected for each completion, the dashboard requires refinement to maximize its ability to be user-friendly. Work is currently being undertaken to deliver the dashboard on a web-based platform. This will allow users to log in and access their results and will enhance its interactivity. Feedback from professionals has also highlighted a wish for a portfolio of generalizable intelligence to guide improvements to practice; it is envisaged that this too could form part of the web-based platform.
To advance the development of the tool, consideration also needs to be given to its ability to adapt to the ever-changing landscape of health care. For example, pilot data suggest that practices are increasingly employing new types of health-care professionals such as Health & Wellbeing coordinators. In order for the tool to remain relevant to practice, collaborative work with stakeholders needs to continue to ensure changes are incorporated in a timely manner.

| CONCLUSIONS
Implementing P3C is a complex and multifaceted intervention that The tool is currently being tested and used as a monitoring and change instrument in four evaluations of P3C across a range of UK sites and models of care. Pilot testing will continue and feedback will be used to adapt and improve the tool. We theorise that ongoing interrogation of the interaction between domains/subdomains (question items) and components (response codes) from implementation data will allow the development of a more comprehensive theory of what works for whom and in what situations to best accomplish P3C.