The aspiration and principles of care planning developed in 2003 were translated into 5 steps – Preparation, Conversation, Recording, Actions and Review – to enable a robust and reproducible delivery model of CSP (Figure 2). The CSP model is a cyclical process replacing current approaches to planned care. It is expected to occur regularly, usually but not necessarily annually. Space for an uncluttered CSP conversation is achieved by separating out the tasks of disease surveillance and care delivery (e.g. teaching inhaler technique) to be provided as part of the preparation step or in subsequent ad hoc task-oriented appointments. The CSP process is designed to make space for the conversation and enable the person to have the same personal information as the professional with time to prepare before the conversation (Figure 3). The structure of the conversation was derived from a number of consultation models and reflects common elements rather than favouring one approach . The conversation starts by acknowledging any preparation work and the concerns the person has identified, adds relevant clinical issues if not covered and moves to discuss, debate and prioritise, identify personal goals, develop action plans and review arrangements.
The steps of the model were adapted successfully from diabetes for other single and multiple conditions, prevention and the biopsychosocial complexity associated with ageing and frailty. This offers a generic approach for teams that brings service users and providers together in co-production  ‘across a lifetime’  however many issues and conditions may emerge. This approach is also outlined following the TIDieR  template (Additional file 2).
The 5 steps of the CSP model
People in a chosen group (LTC/s, age etc.) are identified from the practice register, usually in their birth month. A small number (current care, malignancy etc.) are excluded in a clinically driven triage process. The majority are invited to take part, with explanatory material on the first occasion.
Preparation: The preparation step has two components for most people. The first involves an ‘information gathering appointment’ or sometimes a home visit with a healthcare assistant (HCA) with completion of tests, examinations or assessments needed for condition surveillance. The value of preparing for a new sort of consultation is emphasised with the person. Secondly results are sent to them with simple explanations alongside reflective agenda setting prompts one to two weeks before the CSP conversation. Where no direct measurements are involved, the prompts are included with appointment details, and simple information. The practitioner will also prepare by reviewing results, records, and information from colleagues identifying the few issues of real importance for that person from the wealth of clinical information previously collected on condition specific templates.
Conversation: The structured conversation with a trained practitioner, usually a nurse or GP, focusses on what is important to each person, bringing together the technical expertise of the professional with the lived experience of the individual, in a solution focused and forward-looking discussion. It starts with concerns the individual has identified, adds relevant clinical issues if not covered, and moves to discuss, debate and prioritise, identify personal goals, develop action plans and review arrangements. Contingency planning, referral for specific clinical or medication review, and links to activities within a supportive community (‘social prescribing’ ) may also emerge. Follow up is agreed individually and may not be at routine intervals.
Recording: The ‘care plan’ summarises the discussion and is made available for both the person and the system. It includes the issues of importance to the person, ideally in their own words, and can act as an aide memoire for self-management and contingency arrangements (e.g. actions during asthma exacerbations). It may include health professional activities to coordinate ongoing care for those with complex support needs.
Actions: The CSP conversation shifts the focus of planned activities away from a medical model towards a social model of ongoing care . Self-management may involve individual actions alone as well as non-traditional formal and informal support from groups and peers. For those with complex needs link workers  can build on the issues identified in CSP and support individuals to access an increasing range of community activities. For those who also need specialist, or traditional MDT and social care involvement, CSP initiates an important coordinating function.
Review: The review process is determined by the actions agreed during the conversation and may involve self-monitoring of a behaviour change or clinical indicator such as blood pressure, a laboratory check, a more formal repeat visit to support motivation or to ensure plans are on track.
Implementing CSP: initial lessons
Ensuring people and practitioners had the same personal information, such as test results, with time and encouragement to reflect prior to the consultation was an early aim of CSP in diabetes. Articulating this as a critical ‘preparation’ step took us longer. Early positive feedback from patients and professionals  on a range of benefits was regularly confirmed including by those living with multiple conditions . This included signalling and valuing a new role for the person with time to reflect, leading some to make changes prior to the conversation or raise issues they had not previously thought relevant (such as pain in diabetes). Separating out tests and tasks, done in advance, released time within the consultation for these new topics to be discussed and reduced the need to focus on the computer. Staff concerns that sharing results might increase patient anxiety did not materialise.
Introducing the preparation step can, however, be a significant organisational challenge affecting the whole practice team. It may involve skill mix changes and take weeks to set up. In some practices this concentration of effort led staff to perceive CSP as a system rather than a cultural change and the functional link between a prepared patient and the new style of conversation was sometimes lost. Staff described the activity mechanistically as ‘sending test results’ rather than the enabling function of ‘preparation’. On occasion it was abandoned in the interests of convenience, undermining the integrity of the CSP process. These observations led us to change the language we used in descriptions of the model, training and fidelity metrics. For instance, we relabelled the information sharing leaflets as ‘preparing for your CSP conversation’. As we introduced CSP for people living with increasingly complex issues such as frailty the preparation step also became more complex. HCPs traditionally use specific tools to assess function in these situations. We redesigned these for self-completion by the person in the spirit of ‘working with’ rather than ‘doing to’ and supported practices to identify where best to include these in care pathways.
Over time we learnt that the length of the conversation (20 to 40 minutes) and the practitioner chosen (usually practice nurse or GP) is influenced by the complexity of issues, prior knowledge of the person, what emerges from information gathering, and the skills and experience across the team. Sites began to build in continuity and patient preferences. We observed that a focus on goal setting sometimes led practitioners to confuse CSP with motivational interviewing  unless the role and purpose was explored explicitly. The emphasis on ‘lifestyle change’ in clinical guidelines  and perceptions of professional responsibility could create tension for practitioners over the ownership of goals and frustration when traditional ‘clinical goals’ were not met. On occasion staff blamed ‘lack of motivation’, age or social issues when their own priorities were not acted upon. Reframing the purpose of CSP from ‘helping the person to manage their condition’ towards ‘helping the person to manage their life with their condition’  - which might include specific condition management tasks - helped practitioners to value a wider range of components of the conversation. These included building a therapeutic relationship, an empathic approach , and recognising the person’s problem-solving abilities; as well as striving to increase the person’s specific knowledge, skills and confidence . This was particularly relevant to those with multimorbidity, but it was a core issue for everyone.
Implementing CSP: working with organisations and teams
We devised the Year of Care House (‘The House’) (Figure 4) as a checklist of essential enablers of CSP, and a metaphor emphasising that the walls, roof and floor must be addressed together. YOC pilot sites assigned all the issues they identified to four groups which became the walls, roof and foundations of the Year of Care House. This emphasises that effective CSP consultations rely on these elements working together in the local healthcare system: an engaged, empowered person working with health care professionals (HCPs) committed to a partnership approach (the walls), supported by appropriate/robust organisational systems (the roof) and underpinned by responsive whole system commissioning. In Scotland the foundations of The House were adapted to reflect the different organisational arrangements within the Scottish NHS. The House acts as a check list for what needs to be in place; is a metaphor for the interdependence of each part (if one is weak or missing the structure is not fit for purpose); and provides a flexible framework to enable communities to get started and design the sort of House that suits their population. When used in the training programme, a blank outline of The House is provided, and participants are asked to consider what needs to be in place to deliver and support the CSP process and then reflect on current local services. The concept of the Year of Care House proved useful in helping local groups focus their support programmes and was taken up by others as a way to organise person centred programmes of work . The components shown in Figure 4 are those that have been consistently identified by teams across multiple diverse health communities .
Many of the elements that make up The House were common across sites. These included the need for IT support to get the letters and codes embedded in local systems, senior clinical leadership to secure changes in practice organisation and oversight groups which were visibly led and welcomed feedback. Other components often benefited from local design, and this led to an increasing repository of adaptations to be shared across the CSP community. Exploring the use of language proved critical. For example, staff may assume that ‘care plans’ and ‘treatment plans’ have the same purpose, and that templates are lists of traditional ‘clinical’ issues to be ticked off during the conversation. The term ‘care and support planning’ is often not recognised even by ‘patients’ who have participated in CSP and value it. Local terms (e.g. ‘yellow letter’, ‘MOT’) need to be agreed and used across the team. Partnerships with voluntary organisations and patients influenced the language we used, replacing ‘consultation’ with ‘conversation’. Where confusion between ‘end of life planning’ and ‘care planning’ was identified, local documentation was changed to ‘health planning’.
A focus on the left wall (‘engaged informed individuals’) led sites to explore local demographic, cultural and health literacy issues. Some vulnerable communities  needed additional resources to engage but once involved they participated fully . Examples included practice and faith based introductory sessions, links with education programmes, preparatory material in a variety of formats, including colour coding, home visits and greater use of the telephone. Although many workforce components of the right wall (‘health care professionals committed to partnership working’) were common, these needed tailoring to practice skills and demographics. Expanding the role and number of administrative and non-qualified staff made more time for nurses during the conversation. Working at the top of their grade, HCAs could also reduce non-attenders by engaging patients at the outset of the process, leading to their greater recognition within the team, job satisfaction and career development .
Within the CSP conversation attending to the ‘person’ not the ‘condition’, meant practitioners needed to be ‘expert generalists’  handling a range of health, social and behavioural issues, during one discussion. We observed that this was easier for GPs than practice nurses, whose role and expertise in the UK has been incentivised towards single disease surveillance using a template driven approach (QOF) . GPs could also draw on core communication skills, but this is not part of basic training for practice nurses, to whom much of LTC management has been delegated. Master classes in conditions that nurses were unfamiliar with and further training in handling difficult conversations proved useful. The most successful practices used in-house support and mentorship to smooth the transition to new roles and ways of working.
Implementing CSP: enabling scale and spread
The core challenge driving the changes in our developing approach to spread was how to maintain fidelity to the CSP model during transfer to other communities, while building on local ‘practical wisdom’  to secure ownership and success. Our learning over the whole programme was that this tension could be used constructively if explicitly recognised, and our approach, resources, training and support were adapted to manage this positively.
In parallel with the pilot phase a few independent sites resourced local programmes to introduce CSP (one across a whole regional economy), using early published descriptions of the model  as part of local strategic plans. They had limited contact with the YOC team, training was not yet available, and they stressed ‘don’t tell us what to do’. Using didactic approaches, they found that CSP was difficult to introduce, with poor clinical engagement or culture change; none were sustained. As the core training programme and train the trainers became available, we developed productive relationships with local managers who requested ‘please tell us how to start’ and effective CSP began to be established. Apart from training days we had no direct contact with practice delivery teams. Two members from each practice were expected to attend, return to the practice following training, convince colleagues to embrace the approach and make personal and practical changes. Although it was recommended that attendees should be ‘people with power to make change’ this proved too big a challenge for some.
Local managers also struggled to use the ‘critical success factors’ identified in the pilot programme to support local practices after training. We turned to theory to reframe issues and this became a turning point both for local managers and our own team. Normalisation Process Theory (NPT)  questions assumptions about the delivery of complex interventions in a complex environment, using a sociological perspective to challenge the over-reliance on psychological theories of behaviour change in health settings  which had dominated our thinking. With support of the NPT innovators, we reframed the task orientated list of ‘critical success factors’ under the four theoretical concepts of NPT. Using NPT tools  in workshops enabled project managers to recognise barriers based on human relationships and team interactions, and to design local solutions. We promoted ‘coherence’ as an essential principle for action rather than the aspiration of ‘consistent vision across the organisation’. This engaged steering groups and practice teams with the importance of fidelity. A redesigned training exercise focussed participants on what was different about the new way of working enabling clinicians to discard the belief that ‘we do this already’ and stimulated discussion and greater consensus about ‘purpose’.
Despite these improvements this cascade model of support was slow, exacerbated by frequent changes in local staff, short term funding and a prevailing emphasis on ‘pace and scale’ to be achieved predominantly via training. However, we observed that where individual practices embraced CSP for the majority of their LTC patients and the whole practice was involved, they acted as exemplars, hosting visits from other communities and proved a powerful driver for adoption both locally and beyond . We changed the way we worked to capitalise on this, working directly with ‘early adopter’ practices identified at ‘taster sessions’, using our experienced facilitators to support them to map processes, discuss roles and challenge individual issues  either immediately before or after they attended core training. The positive experience this generated spread by word of mouth and helped to recruit subsequent cohorts.
Local capacity building needed to embrace this facilitation approach . This required an acknowledgement of the importance of context  and the personal skills and attributes involved in ‘holding the flame for CSP' while ‘supporting local tailoring’. Helping steering groups to distribute the roles of project manager, trainer and practice facilitator across a local long-term support programme became an early action at new sites. We developed a 3 day ‘facilitator programme’ and direct work with practices provided opportunities to offer a ‘mentorship’ model for local support staff. YOCP facilitators demonstrated how to use observed behaviours and language to identify cultural barriers at team level and how to share practical solutions to common problems collated from the wider community of practice. Debriefing immediately after practice sessions and support by phone as issues emerged helped develop local staff. This current approach to implementation is outlined following the TIDieR template  (Additional file 3).
Impact of CSP
The achievement of improved patient experience, practitioner job satisfaction, health behaviours and outcomes, team work, practice organisation and resource use reported in the pilot  have been repeated throughout the programme . CSP is now ‘business as usual’ in a majority of practices across several CCGs . The demonstration of greater savings for a practice as patient complexity increases  has helped to engage new teams. In diabetes, improved biomedical outcomes take time to emerge across populations but have been sustained . Large managed programmes have seen stepwise improvement in completion of the core processes of diabetes care  moving from within the bottom 10% to first place  and from 119th to 5th place  in national comparisons .