Integration and continuity of primary care: polyclinics and alternatives - a patient-centred analysis of how organisation constrains care co-ordination

Care integration

A consequence of the successes of NHS acute care, improved living conditions and lifestyles is an ageing population, an increasing prevalence of multimorbidity and higher numbers of frail older people. Dementia is emblematic of this situation. It has become an increasingly pertinent health issue, to the extent that the Prime Minister has established a special working group to investigate and propose responses. However, it could equally be argued that what makes dementia a policy ‘problem’, or at least exacerbates it, are the insufficiencies, including fragmentation, of services caring for people with dementia.

The policy of integrating primary care so as to reduce the demand for hospital beds first appeared as a concomitant of transferring patients out of large mental hospitals into community care. Twenty years later, NHS acute hospitals also faced increasing pressures, if not to reduce, then to contain the growth of, demand for acute hospital services. Waiting times for hospital admissions had been a salient health issue since the 1950s and were becoming more so. Unplanned hospital admissions of frail elderly people were a growing part of a hospital’s workload,15,16 as were delayed discharges from hospital. 17 Patients less satisfied with access to their general practice appeared to be more likely to use emergency departments (EDs). 17 NHS policy came to assume that more integrated care might address all of these problems, reducing the numbers of unplanned admissions and reducing the length of hospital stays generally. 17

Fiscal stringency

After the 2008 banking crash, reducing hospital bed use was bound to be a pillar of fiscal control in the NHS, where 67.5% of costs arise from hospital care. 18 Cuts to local government expenditure fell heavily on social services, but that constrained hospitals’ ability to reduce lengths of stay. In response, it was proposed that £1.9B of NHS budgets be transferred to joint local authority and NHS control; however, this ‘Better Care Fund’ has (in mid-2015) not yet been created.

Provider diversification

Most general practices are professional partnerships, jointly owned by some or all of the GPs working in them and employing nurses and other staff. Since 1997 the policy of provider diversification has extended into NHS primary care, leading to the introduction of:19

• new general practice contracts [Personal Medical Services, ‘New’ General Medical Services, Alternative Provider Medical Services (APMS)]

• nurse- and allied health professional (AHP)-led general practices

• extending GP co-operatives to provide certain diurnal services

• commercial general medical services

• case management, both corporately provided and via the Community Matron programme

• social enterprises, whether independently founded or former NHS or local authority services

• PCT-owned and directly managed general practices

• GPs with special interests (GPSIs)

• walk-in health centres.

Increasing technical, and therefore occupational, specialisation within primary care also increased the number and variety of primary care providers. From 2009, PCTs were forbidden to provide services. Some followed the Welsh example of creating vertically integrated NHS trusts that provided both acute health services and CHS. 9 Both technical specialisation20 and providers’ increasing organisational diversity exacerbate the difficulty of co-ordinating care provision across multiple providers, fragment organisational responsibilities and increase the scope for service duplication. 21

General practitioners as commissioner-co-ordinators

Since the GP fundholding scheme (1991), NHS policy has increasingly made GPs responsible for care co-ordination, a dual trajectory of giving GPs budgets for other primary care services and most hospital services and of making GPs more openly accountable to NHS management for how they use these powers. GP fundholding implemented both of these elements at individual GP and at practice level. The subsequent evolution of PCTs made GPs collectively responsible, with others, for commissioning most secondary services and, later, CHS. After 2010, Clinical Commissioning Groups (CCGs), of which all general practices were members, were to:

[. . .] bring together responsibility for clinical decisions and for the financial consequences of these decisions [. . .] reinforce the crucial role that GPs already play in committing NHS resources [. . .] increase efficiency, by enabling GPs to strip out activities that do not have appreciable benefits. 22

The ‘activities’ that GPs would review explicitly included hospital services for which CHS or other forms of primary care might be substituted. In doing all this, CCGs were also to increase the integration of care and give patients a bigger role in (and more responsibility for) selecting and co-ordinating their own care.

Colocation

A physical rather than an organisational remedy was to colocate services. After Dawson’s report, the next main innovation (1935) was the Pioneer Health Centre at Peckham. This colocated preventative activities (physical exercise, swimming pool, games, workshops), a day nursery, cafeteria, laboratory, general practice and (after 1946) an infants’ school together in a purpose-built building. It was closed in 1950, partly because it offered more services than ordinary general practices and, therefore, required patient copayments.

From the 1950s the Ministry of Health supported the construction of health centres in which to colocate general practices with CHS, dental services and hearing services. Although many were built, they remain far from universal. Despite the Cumberlege Report’s37 advocacy of area-based community nursing teams, the colocation, or at least attaching, of CHS staff to general practices became widespread over subsequent decades. Less often, social workers were also colocated to general practices, and in a few areas social care teams came under joint NHS and local government management. 38

Stronger co-ordination

Our Health, Our Care, Our Say envisaged substantial expansion of community-based services, including ‘expanding the provision of community nursing services to support people at home’,39 a key element in the Darzi Report. 40 The Care Closer to Home initiative (2008) contained no one blueprint for such care but envisaged that it would be system-wide and delivered by enhancing the local workforce, working in partnerships with patients and communities, and investing in better community facilities, stronger CHS and GP-led health centres or ‘polyclinics’.

Darzi41 argued that colocating GPs with CHS and social services [or, failing that, using information technology (IT) to link services] would promote service co-ordination:

Integrating care is also a key driver of personalisation because [. . .] there are likely to be fewer appointments on a typical pathway, greater familiarity between patient and staff, better information for the patient, and a more ‘seamless’ experience for the patient. 41

The Next Stage review (2007) promised the creation of 150 ‘GP-led health centres’, open to patients not registered with GPs (which also addressed the finding that, on the basis of Allocation of Resources to English Areas-weighted populations, deprived populations were seriously underdoctored). These centres would have new buildings, free-standing or near an ED, providing diagnostic services, social advisers, consultants, physiotherapists and one or more general practices, serving a population of 50,000. 42 To some observers this sounded ‘like a GP surgery bolted on to an independent sector treatment centre’. 43 The proposal to offer APMS contracts to providers running these centres led many people, including the British Medical Association (BMA), to perceive, and hence oppose, the plan as extending corporate provision of NHS-funded primary care. 42,43

Although the government was quick to announce proposals to develop ‘polyclinics’ across the country,43 the concept remained vague, with no specific blueprint beyond Darzi’s initial vision. Developments in practice were more limited and ill defined. 42 Various developments that did not encompass all the key ‘Darzi’ characteristics were labelled ‘polyclinics’, or, in London, ‘polysystems’. Some were more like walk-in centres. Few managed to achieve the kind of integrated care ‘one-stop shop’ that early articulations of the polyclinic model envisaged. 44–46 They proved expensive and many attracted only few patients, giving poor value for money. 47 In London, however, the number of walk-in patients far exceeded expectations, although without reducing pressure elsewhere in the system. 48

London polysystems

NHS London took the Darzi centre idea a step further. A Framework for Action49 set out plans for between five and 10 ‘polyclinics’ to be established. The favoured option was locality-based ‘polysystems’, either physically colocated in a hub building or a ‘virtual polyclinic’, a care network of existing service providers and locations, with the emphasis on developing primary and community services. To general practice, pharmacy and some diagnostic services traditionally carried out in hospitals, A Framework for London added dentistry, physiotherapy, family planning and mental health services, non-NHS services such as social care, voluntary sector and other care providers, and newly created services to promote health and well-being. Darzi argued that London was suited to such concentrations of services because of its dense public transport network. However, the BMA criticised the imposition of the ‘super surgery’,50 challenged the assumption that polysystems would reduce ED attendances and questioned the lack of detail on costs and the implications for hospital survival. 51

A large-scale evaluation of London polysystems48 was commissioned, but remains unpublished. Its interviewees equated service with system integration, and both with multiple colocated services. The interviewees considered linking up services with primary care to provide improved patient pathways, an important factor for enhancing the quality of services. Commissioners aimed at developing a ‘one-stop shop’ offering a comprehensive range of services for the management of long-term conditions and making it easier for patients to obtain a GP appointment. Enabling GPs to obtain diagnostic and specialist services within the community would reduce ED attendances. Fewer than 10 polysystems actually appeared in London, many built on existing developments48 and more limited than originally envisaged;44 for example, developing single community-based services [for chronic obstructive pulmonary disease (COPD) or back pain]; a colocated GP and walk-in centre with some outpatient consultations; and a new general practice in a hospital ED.

It proved difficult to change GP referral patterns, leading to the underutilisation of community-based services, many of which were small-scale anyway. Community service developments demonstrated improvements in access, particularly geographical access, across a number of sites, but moving hospital services into the community resulted in the duplication and disconnection of services. There were also examples of colocated services related to the same care pathway still working in silos. Only limited numbers of patients used multiple services on the same day. Some GPs were concerned that colocating services had the potential to confuse patients for different services and to create inequities in treatment. Interviews with other staff also suggested that institutional and organisational barriers between different service providers prevented patients from using colocated services on the same day. Furthermore, patients were disappointed that the ‘polyclinics’ did not offer a comprehensive range of out-of-hospital services. Commissioners lacked guidelines and levers for creating integrated hubs and ‘one-stop shops’. Despite using various mechanisms to encourage integration between hubs (e.g. governance arrangements and policies, regular meetings with all providers), the commissioners interviewed for the evaluation considered it unlikely that hubs would be able to offer a truly integrated service in the near future. Most sites identified a variety of logistical, IT and clinical administrative barriers to fuller integration.

However, there were some successes. In one borough a new service was developed with consultant input and clearer patient pathways. It worked in partnership with a private organisation and other selected general practices across the borough to improve quality of care through mapping performance and facilitating co-ordination. Colocating services in the hubs did enable small-scale synergies to be identified. This facilitated more flexible delivery of care in hubs that had sufficient capacity and diagnostic services, encouraging collaboration and partnerships between different services and professionals. Examples included providers sharing clinical and support staff; nurses sharing expertise; informal clinical liaison; streamlining diagnostic and other cross-referrals; holding joint clinics; and better use of space. Colocation of services on one site provided opportunities for staff in one centre to direct patients between the available services, although barriers to capitalising on the opportunity included organisational ‘silos’ (physical separation, separate funding), organisational restructuring (which made providers inward-looking and hence less collaborative), and different IT systems. Although patients liked the option of accessing many services on the same site, few actually used more than one during each visit.

Integrated care pilots: care networks

Reflecting the tripartite NHS structure, general practice-centred and CHS-centred care networks have both developed, including the polysystems described above. The Integrated Care Pilot Scheme (2008) supported the formation of 16 care networks, coalitions of health-care providers who constructed care pathways across organisational boundaries,52 in ascending size from case management schemes to interorganisational care teams53 to care hubs, federations of general practices and NHS–local authority ‘partnerships’. Case management was often embedded in these networks, and indeed was the focus of six of them. 54 Since 2010, NHS policy has encouraged the development of federations of general medical practices that share certain other primary care services such as GPSIs, diagnostics, CHS and mental health services; and of hubs mediating referrals from ‘federated’, usually smaller, general practices to (other) NHS, social care or third-sector providers. For example, in one locality in Plymouth, social workers and long-term condition managers work as one team with a cluster of practices. The ‘Virtual Ward’ model is more vertically oriented, with (e.g.) GPs, CHS, social services and hospital consultants providing more acute care at patients’ homes than patients would usually receive there.

High-level partnerships between NHS planning (later, commissioning) bodies and local government have existed since the 1970s, originally for transferring patients from long-stay mental hospitals into community care. Later, ‘section 75’ (§75) agreements allowed the pooling of NHS and ‘health-related’ local authority budgets under common management. Nowadays, Health and Well-Being Boards (H&WBBs) undertake intersectoral health promotion and foster interorganisational care co-ordination,13,23 particularly for frail older people. Interorganisational co-ordination often appears strongest at the top managerial level and more ad hoc at an operational level. 55 However, there are exceptions. Care trusts come closer to organisational integration. In Torbay, managerial structures and budgets for CHS and social care were pooled below top management level, but general practices remained separate and were linked to other providers by care networks. The Think Local Act Personal scheme is piloting integrated health and social care personal budgets.

Patients as care co-ordinators

Most patients expect to have a role in co-ordinating their own care90 and, indeed, must when health professionals fail to co-ordinate it for them. 5 (See Appendix 1, which elaborates some of these points.) Wagner’s ‘pyramid’117 of care implies that patients in any case provide much of the care for themselves. Those who do are less likely to consult a GP. 118,119 An important way in which patients co-ordinate their own care is by mobilising their informal social support networks, differentiating between friends, relatives and others from whom they obtain different kinds and amounts of support. 120,121 Formal services can also help patients self-manage their care (e.g. by helping people with psychiatric problems manage their financial affairs122). Personal budgets supplement the patient’s resources, strengthening the patient’s role as care co-ordinator. An evaluation of the personal budget schemes for social care found that, on balance, clients thought that having a personal budget had increased their control over what services they received, and to that extent their quality of life. These responses were most prevalent among clients with a physical and/or mental disability. 123

However, there are practical limits to patients’ self-co-ordination of their own care. One is how far patients can or will travel to receive care,124 especially older people who have limited mobility and rely on public transport. Caregiver burden appears to be a predictor of the institutionalisation of frail older people, particularly those with dementia. 125 For more dependent older patients, care transitions (often, into a nursing home) may be initiated by their informal carer, especially when the carer’s skills are ill matched to the patient’s needs. A US study found that male informal carers (especially husbands and sons) were more likely than female carers to request such a transition, as were informal carers of the opposite sex to that of the patient. 126 The patients most needing co-ordinated care are often those whose multimorbidity1 and multiple medication make it hardest for them to co-ordinate their own care5 and for whom hospital admission is less discretionary. 127 The severity of the condition and the level of distress predict help-seeking for mental health problems. 128 Self-management programmes also have, in some contexts, high dropout and refusal rates (10% and 51%, respectively, in a Spanish study). 129 Cross-sectional, longitudinal and relational continuities of care are likely to break down when users are highly mobile, even within a small locality, for instance for patients with serious mental illnesses. 130,131 When these practical limits are reached, patients often seek help (an extensive body of research describes under what circumstances they do or do not do so). The task of co-ordinating their care then passes largely to their health-care providers.

Processes and structures for care co-ordination

In most health systems, the process of providing complex care consists of the following sequence:

1. A person co-ordinates their own care.

2. The person seeks help from formal services.

3. The patient attends a first responder, typically a primary care doctor (but alternatively the ambulance service, hospital ED, etc.).

4. Where needed, the patient is transferred to further providers (CHS, psychiatric care, hospital, etc., as relevant).

5. These ‘second-responder’ providers then provide and co-ordinate care.

This is a simplification. Further transfers to yet more providers may follow stage 5, including feedback loops returning the patient to an earlier provider. Further iterations of steps 3, 4 and 5 may occur (e.g. the ‘revolving door’ between outpatient, inpatient and community care of mental health problems). A patient or health worker may initiate, even duplicate, several of the above processes at once. The sequence may be truncated for some reason (e.g. the patient withdraws consent).

The ways in which health-care organisations conduct steps 2 to 5 produce the patient’s experience of continuities or discontinuities of care. After step 3, each transition is an opportunity for the care process to continue – or to be deflected or disrupted, endangering its continuity and tending to increase risk, adverse events and costs. 33 The increasing technical, and hence occupational, specialisation of health-care providers, and policies that promote more diverse ownership of providers, make it increasingly likely that patient episodes will be distributed across multiple professions and/or organisations, with correspondingly more transfers between them. 5 Studies of these processes (especially experimental studies) have tended to concentrate on chronic disease (cardiovascular disease, diabetes, asthma, COPD, acquired immunodeficiency syndrome/human immunodeficiency virus); mental health, including substance abuse; and care of the elderly132 – increasingly prevalent conditions92 whose treatment characteristically involves transferring both patients between multiple providers, including social care.

The more transitions an episode of complex care involves, the greater the need for a deliberate attempt to articulate the different carers’ interventions into a coherent care process133 and make the patient’s transitions between these providers as easy, prompt and non-disruptive as possible. 80 That is, the need is greater for care co-ordination5 as a ‘continuity mechanism’,80 for ‘continuity is how patients experience integration of services and coordination’. 90 Clinical care co-ordination requires the following steps:

1. A care co-ordinator (an individual or team) combines different providers’ separate inputs, with a clear division of labour between them, into a coherent care process13 that meets the patient’s care needs in the round (to produce cross-sectional continuity of care). A care plan documents these points.

2. The care co-ordinator maintains an ongoing relationship with the patient24 (to produce relational continuity).

3. The patient’s care plan, which includes social care,74,90,134 is shared between the different services involved (to produce cross-sectional continuity at each encounter between patient and provider, and longitudinal continuity over time).

4. The case co-ordinator reviews the patient’s circumstances and, if these change, alters the resources or services offered (to produce flexible continuity).

5. The patient has access (often facilitated by the care co-ordinator) to the range of services that they need, with no obstacles to access, missing services or interruptions of services during the period of the care plan (to produce continuity of access).

6. Transfer of information occurs between workers and between organisations and/or shared access to electronic clinical records (to produce informational continuity).

Observing the above features would be evidence of continuity and co-ordination of care.

However, the structures through which health systems undertake care co-ordination are nested. Working ‘outwards’ or ‘upwards’ from the patient level, the production of continuity of care typically involves four organisational levels:135,136

1. Patients co-ordinate their own care.

2. Co-ordination within one provider organisation (general practice, community mental health trust, etc.), which may involve a named care co-ordinator.

3. Interorganisational co-ordination through a care network of separate provider organisations;52 or ‘virtual’ integration (as in most NHS ‘Integrated’ care schemes). Horizontal co-ordination is that of multiple primary care providers, vertical co-ordination that of primary with secondary care. 136 (We do not equate ‘vertical’ with ‘hierarchical’ and ‘horizontal’ with ‘network’. 122)

4. The local health system sustains a set of overall governance structures within and through which provider organisations and care networks collaborate, controlled by higher-level bodies. 137

Evidence is accumulating that care co-ordination depends on the interaction of structures at all these levels. 92,138,139 ‘Coordination structures’ have ‘consequences for personnel’ and these have ‘consequences for clients’. 33 For instance, Wagner’s well-known model117 views chronic disease management as requiring six ‘pillars’: community resources and policies, health-care organisation, self-management support, delivery system design, decision support and clinical information systems. Programmes that contain elements of the Wagner Chronic Care model have been shown to be likely to improve clinical outcomes and care processes for both single diseases and multimorbidity. 92 A systematic review140 of studies concerning interventions based on the Wagner model found that 32 of the 39 studies reported improvement in at least one process or outcome measure for patients with diabetes. For diabetes, asthma and congestive heart failure care, 18 of the 27 relevant studies reported reduced health service use or reduced costs. Despite methodological problems with some of the studies (many evaluated time-limited special projects rather than typical health-care practice), the evidence suggested that implementing interventions based on the model appeared to support improved health-care outcomes. Evidence of the economic impacts of ‘integrated’ care is sparse and, overall, inconclusive. 141 More studies focus on general practice ‘characteristics’ (number of doctors, team climate, etc.) than on how such practices are organised. 142 In this research we focus on, and contrast, the intra- and inter-organisational levels of care co-ordination. 143

Care plans

Care plans are the standard way to assemble and define flexible, individually tailored combinations of services for each patient,24 possibly including elements of self-management,64 patient education and planned communication with clients and families (e.g. through joint consultations and reminders). 63 As a means of co-ordinating care148 they exploit the formalisation and standardisation of work processes found in ‘ideal-type’ bureaucracies, which one would expect to promote longitudinal continuity of care and, when care plans are reviewed, flexible continuity.

Nowadays care plans often involve the application of care guidelines,136 which facilitate informal relational co-ordination in hospital settings149,150 and so, perhaps, elsewhere. Thus, Swedish polyclinics use specific models of care (e.g. for preventing hospital admissions, disease management, case management, other chronic care and illness prevention models3), clinical indicators and standards, and agreements about what referrals are appropriate. Stepped care protocols, including ‘enhanced’ (accelerated) referral, can be used to define appropriate referral criteria for specific conditions. A systematic review151 found that such protocols produce clinical benefit in mental health services. However, there are technical limits to formalising and standardising complex care. The course of many non-cancerous diseases,144 for instance mental–physical multimorbidity,152 involves clinical complexity that is hard to predict and difficult for protocolised, routinised or pre-planned health-care processes to accommodate. 12,147 Many of the studies underlying evidence-based guidelines were designed expressly to exclude patients with comorbidities. A perception of receiving ‘production-line’ treatment may make patients less compliant. 153

Co-ordination mechanisms within a single organisation

‘Organisational integration’ means that a single organisation undertakes these activities, rather than separate organisations. 136 Patients’ care is co-ordinated by a unified management structure that applies a single set of objectives when planning, financing and delivering (producing) care. 20 Within integrated organisations, three mechanisms for co-ordinating patient care are widespread:

1. the consultation model

2. line management

3. internal interdisciplinary teams.

In practice these mechanisms often coexist, not always comfortably.

The consultation model of care co-ordination133 involves both a ‘captant’ profession that focuses on the patient’s care as a whole, in the round, and a complementary set of professions that focus on more specific tasks. 147 Typically, the captant health worker (classically a GP) transfers or refers patients to other professionals with more ‘appropriate’ expertise23,145 who, having made their contributions, return the patient to him or her. This referral pattern can be visualised as a star-like structure with the captant professional at its centre. The captant professional acts as care co-ordinator, unless he or she delegates that role, too, to another health worker. Within a hierarchy the most straightforward way to implement the consultation model is to make the captant professional line manager of the other professionals. One would expect the consultation model to produce high relational continuity of care with the care co-ordinator and low relational continuity with the non-captant professionals. Even within large NHS general practices, it is often possible to maintain a single GP care co-ordinator for most people with long-term conditions, but for individuals who are home bound or struggle to reach the practice and have multiple input from external teams (district nursing, mental health, social care), the GP is less likely to be able to fulfil this role.

Line management in an ‘ideal-type’ bureaucracy would allocate each patient’s care to one health worker or a group of them, and formalise and standardise procedures for hand-offs between care workers,154 assisting collaboration12 and care co-ordination. Besides using formal care plans (see p.13), an ideal-type bureaucracy would co-ordinate care by standardising:

1. Information transfer, producing informational continuity5,82,135,155,156 by implementing compatible, even shared, information systems (data definitions, IT protocols) across departments and the whole organisation.

2. Procedures for requesting and accepting (authorising) patient transfers between health workers, with uniform criteria defining which transfers are appropriate, so as to establish longitudinal continuity of care. (Evidence-based care pathways are one way to achieve this.)

3. Other working practices relevant to transfers, for example common hours of work. 89

Within a professional partnership of GPs, the partners stand at the apex of a hierarchy that line manages the non-partners’ (employees’) work. In these circumstances there is little practical difference between the way the consultation model works within a partnership general practice and conventional line management. However, as practice size increases, multiple GPs tend to become involved in one patient’s care.

A sufficient scale of service provision to implement the care plan is also required. For instance, to engage mental health service users and deliver improved mental health outcomes and social inclusion88 requires that, say, assertive outreach team staff have sufficient time to build and maintain working relationships with their clients. 157 Corporations and public bureaucracies154 can exploit the formalisation and standardisation of work processes to routinise, automate and deskill work, using cheaper, more flexible staffing arrangements, and to that extent increase labour productivity. Examples of this tendency include the introduction of call centres,158–160 telephone-based care management160 and the development of case-load norms for community matrons. 36 A larger organisation (e.g. the bigger US physician practice partnership vs. single-handed practices) is also more likely to have economies of scale permitting the use of, say, expensive diagnostic equipment. 161 However, the larger the provider organisation, the more difficult relational continuity might become to maintain. 75

The greater the range of professions and services contained within an organisation, the greater the flexible and access continuities of care one might expect. In theory, formalisation and standardisation of care plans (see above), information systems (see below) and a clear, explicit division of labour should produce efficient collaboration between specialisms.

In practice, however, health-care providers typically accommodate specialised care by developing an organisational structure162 in which uniprofessional hierarchies (‘semi-detached’ professional ‘silos’163) coexist in parallel, formally linked to each other and to ‘general’ management only at senior level via ‘hybrid’ clinical managers. 164 Each profession tends to cultivate the habit of other professions referring to them ‘good’ patients ‘appropriate’ to their established ways of working. 147 Certain professions see their role as limited to a determinate part of the care pathway; others see it as overseeing the whole patient journey. 147 In such organisations interprofessional demarcations and rivalries,56,165,166 reflecting ‘turf wars’ over status and resource control,33,167 often hinder the creation of common working practices (equated with control from outside the profession). Specialisation may fragment a general practice’s ‘procedural’ organisational memory of how to operate and co-ordinate care interventions. 148 Studies of psychiatric services in France and Sweden found that team members’ perception of the psychiatrist’s higher status impeded collaboration (they focused on the psychiatrist’s views rather than the patient’s needs122,147). Doctors especially are reported168 to resist what they perceive as overmanagement, at times equated with any non-medical management, which may actually reduce doctors’ ‘citizenship’ and ‘behavioural commitment’. However, managerial and interprofessional links created for practical operational purposes, and focusing on the health outcomes of collaboration, have the opposite effects. 151,169 In the USA, health maintenance organisations (HMOs) spread only by diluting the strong management and vertical integration of the likes of Kaiser Permanente. 170 Under the admitting-rights model or where primary medical care is organised on specialty lines (e.g. Germany, France147), the self-image of the doctor, even in primary care, is more as a specialist expert than as a care co-ordinator. A multiple case study171 reported GPs as being reluctant to relinquish what they saw as their central role in the clinician–patient relationship and wondering if interprofessional working would reduce their autonomy.

Internal interdisciplinary teams can co-ordinate care across professional ‘silos’ even when the care co-ordinator has no line-management authority over the other professionals135 (a ‘project team’ arrangement, common outside health care172,173). Published studies seldom say whether their findings apply to teams within an organisation, to teams working across organisations or to both. That often has to be inferred from the study setting. Paradoxically, studies of hospital teams may be more relevant to integrated primary care organisations than are studies of teams in fragmented primary care networks.

A recent systematic review23 describes interdisciplinary teams as a prerequisite for the effective, integrated care of older people. A care team can undertake assessment and care planning, implement disease-specific guidelines and support self-management,64 a collective care co-ordinator role. 147 The team may have a nominated co-ordinator who negotiates and agrees care plans with the other team members,174 for instance at case conferences or the equivalent. The impacts of team care co-ordination appear greatest for patients previously not linked with services, lacking knowledge of their condition and motivation to change their behaviour and lifestyle risk factors, but having some capacity for self-management. 64

Co-operation between team members depends on what place each occupational group has within the care pathway. 147 Many studies suggest that aligning ‘cultures’175 and team climate is important in influencing how far, and how effectively, the above activities occur. Mechanisms for bridging between professions with different occupational cultures include informal relationships176–178 and personal affinities among team members. 179 Good-quality information sent to other professionals was reported in a French study147 to improve relationships between other professionals and the referring doctor. Especially when the team is being set up,174 it is necessary for its members to commit themselves to shared assumptions about their roles, ways of interaction and common tasks122 (which does not require complete convergence of their mental models). Such agreement is more likely in long-established, highly interdependent teams with a small, stable membership. 175 The more varied the services or occupations involved, the harder it is to co-ordinate a team. 121 Scores for satisfaction, continuity of care and access to care are higher in general practices where staff report a better team climate. 180 One would expect co-ordination by a team to produce weak relational continuity but strong cross-sectional and flexible community.

Primary care teams’ reported imperfections include lack of ownership of the management of some conditions (e.g. dementia, mental health problems, multiple comorbidities). 181 Where the patient’s carers remain the same, relational continuity – in effect, the patient himself or herself transmitting information informally – can to a limited extent substitute for weak informational continuity in IT systems. 80 Team co-ordination may improve patient well-being, but it does not necessarily produce cost savings. 64

Vertical integration

Vertically integrated ‘systems’ combine primary with hospital care, for instance when hospitals acquire physician groups,182 financed by some form of shared budget (e.g. a capitation system). Transaction cost theory predicts that vertical integration reduces uncertainty and ‘opportunism’ (dishonest transactions), makes information about services transparent, and so reduces transaction (i.e. managerial) costs and sustains a common ‘mission’. 20 An extensive study24 found that the care co-ordination programmes that were more successful in reducing hospital admissions for older people with chronic conditions involved multidisciplinary assessment of a broad range of patient needs (medical, functional, psychosocial); following the patients up over time (at home, not just at the doctor’s premises); health education; supporting patients in making lifestyle changes; and co-ordinating care across different providers.

Having controlled for a ‘California effect’ (Medicare-financed hospital admissions in that state were half the NHS rate) and population age differences, Ham et al. 71 found that for 11 major care groups, Kaiser Permanente (which integrates health insurance, hospitals and ambulatory care doctors) had admission rates and average length of stay of around one-quarter of NHS levels (in 2003). 70 The researchers attributed these effects to primary care doctors having an interest in keeping Kaiser Permanente solvent; specialists working across hospital and primary care; a lack of incentives to maximise admissions; use of care pathways and specialist hospital discharge staff; and primary care doctors’ fast access to diagnostic services. Ham et al. also argued that the existence of competitors to Kaiser Permanente motivated these effects. Recent overviews of research on integrated health systems in the USA together suggest seven characteristics of successful vertically integrated health-care organisations:20,33

1. hierarchical governance or long-term relationships between organisations that approximate to that (e.g. micro-commissioning55)

2. ‘systemness’,20 that is continuity of care and provision by teams, enabling the substitution of less for more costly models of care. This is what Øvretveit et al. 33 call ‘clinical integration’, that is, how well services are co-ordinated between people and sites over time

3. capitation payment, creating incentives for preventative care

4. competing with other providers on quality rather than cost

5. good management and information systems, with ‘functional integration’ (i.e. support functions such as the management of finance, human resources, IT and planning closely co-ordinated across units)86

6. a sufficiently large organisation and population served to ensure long-term stability of the organisation and enable the development of care pathways for common diseases

7. doctors who actively participate in management and are economically linked to their organisation.

Horizontal integration also requires these conditions.

In theory, vertically integrated organisations have an incentive and the means to replace hospital with primary care, a development usually assumed to be cheaper and to improve access, relational, flexible and informational continuities of care, and, therefore, patient satisfaction. Regarding the cost of tests [but not magnetic resonance imaging (MRI) scans] and minor procedures, there is some US evidence to support this assumption. 161 Substitution also avoids disruptive transitions between care settings. In England, GPSI dermatology services were more costly than hospital outpatient treatment, but had similar outcomes and better access. 183 The NHS ‘Closer to Home’ project was a hospital outreach scheme that supplemented rather than replaced existing services. Patient-reported waiting times, quality of care (although some providers expressed concerns about worse quality), overall satisfaction and access improved compared with existing services; care co-ordination and interpersonal quality of care did not. The new services only treated less complex conditions, although they did this at a lower cost than the standard NHS hospital tariffs. 184 Technical developments (e.g. non-interventional cardiology, development of scanning and diagnostic equipment that could be used in non-hospital settings) aid in these substitutions. 161

Emergent networks

Emerging from habitual referral or information flows,12,52 a care network of interdependent providers151,185 co-ordinates health, and often social,186 care for a defined care group. Some call this ‘virtual integration’. At their simplest, care networks may work by nominating ‘link workers’ as care co-ordinators for individual patients and/or care groups. 187,188 Through repeated use147,188 or prior decision, such arrangements become normalised,189,190 even to the point of developing formalised systems for planning and co-ordinating patients’ care,12 requiring ‘a sound system of referrals’. 186 A Massachusetts study191 found that larger, care network-affiliated practices are more likely to have the ‘structural capabilities’ for patient assistance and reminders, a culture of quality, enhanced access to services, and electronic health records (EHRs). A systematic review63 found that the care co-ordination projects with the greatest impact on health outcomes were (in descending order) those addressing relationships between service providers, clinical co-ordination arrangements, and infrastructural support for co-ordination. Those with the greatest impact on patient satisfaction addressed relationships and communication between service providers, and gave ‘support’ to clinicians and patients. However, fewer than 20% of the studies in that review reported financial gains. Continuity of access in care networks depends on the network involving all the providers that patients need. 12 The greatest obstacle to care co-ordination is when key providers do not participate in the relevant care networks, which then lack links to other providers. 192 Successful programmes [e.g. SIPA, Integrated Care in Italy, Project for Advocacy, Counselling and Education (PACE), SA HealthPlus] have tended to include geriatricians, primary care doctors or both. 23

The consultation model may operate only within a single organisation, but classically it also extends across organisations. In NHS-like systems, GPs are usually the default care co-ordinator;193 this is an important part, the Royal College of General Practitioners (RCGP) says,194 of what makes GPs ‘generalists’ and the traditional mode of care co-ordination in general practice. The complementary non-captant roles may either be ongoing (e.g. nursing support) or transient (e.g. imaging services). English GPs sometimes delegate to nurses the management of more minor and stable medical conditions. 195,196

The consultation model has two limitations. First, it depends heavily on the skills of the captant professional, usually a primary care doctor. In some US projects a patient coach197 has been employed to check that the patient understands and agrees with the proposed course of treatment, and to help in co-ordinating care among the (other) providers involved, including informal carers. 5 Although patients value relational continuity, over 13% of patients in a study in London and Leicester stated that they had not experienced it in their general practice. 85 Survey data in 2002 suggested that 38% of a sample of UK patients said that their doctor had not stated the aims of treatment; 67% said that the doctor had not asked about the patient’s ideas on treatment and care [although 65% also rated their doctor(s) as ‘excellent’ for ‘Listening carefully to health concerns’ and 57% rated their doctor(s) as excellent in diagnosis]; and 75% reported long waiting times for care (which included hospital care). 198 In past decades, GP referral notes to consultants often had meagre clinical content. 20 One study suggested that GP interest in care co-ordination tended to trail off once patients had accessed other, specialist providers. 181 US studies suggest that the increasing pressure on primary care doctors’ time and the complexity of their tasks are practical obstacles to care co-ordination, which requires time between patient appointments;5 but that longer (not shorter) contact times for patients with chronic conditions are associated with better quality indicators, greater patient satisfaction and better patient education. 180,199 Patients often fail to understand the doctor’s advice, although checking that they do (which takes time) is associated with improved outcomes for diabetes care. 200 Primary care doctors tend naturally to have a clinical culture, and to be less receptive to interventions or strategies for areas that they cannot directly influence (housing, social care, etc.). 135

One response to these difficulties comes in case management schemes, an updated variant of the consultation model in which other clinicians can replace the primary care doctor as the first point of patient contact with formal care and as care co-ordinator. The case manager is often a primary care nurse64,150,201,202 (e.g. NP, ‘community matron’ or equivalent36), although AHPs, social workers203,204 or, when a patient is about to be discharged, a hospital-based ‘navigator’ can also fulfil the role. 205 In theory the case manager regularly reviews the patient’s condition (including compliance with treatment); plans, assembles and co-ordinates care inputs from a range of services (a gatekeeper role206); communicates with other professionals (e.g. orally, by sharing records, team meetings);207 and is the patient’s first port of call in a crisis. ‘Collaborative care’ for people with depression and anxiety typically has as the co-ordinator a trained health worker who, where necessary, involves a psychiatrist, other doctors or other clinicians to make a variety of interventions (screening, education, psychotherapy, etc.). 207 Care co-ordination might also be undertaken jointly by individuals from different organisations. 208 (The cited studies, however, relate to individuals with depression rather than to frail elderly people with multimorbidity.)

In the USA, case management schemes reduced unplanned hospital admissions31 and institutional placements of older people. 209 English case management schemes did not reduce unplanned hospital admissions210 because the UK schemes were applied to patients living in their own home, who are known to make more use of secondary care (ED, outpatient department and inpatients) than patients in residential care211 (the setting for the Medicare schemes). More effective case finding (also reported in Australia212) might have compensated for lower referral rates among the patients who already participated. However, patients, their families and NHS staff liked the English case management programmes. 204 The same applied to case management patients in Spain. A Spanish home-care service based on nurse-led case management increased access to health-care services and resources, patients’ functional ability and levels of satisfaction, while reducing caregiver burden. 213 A systematic review of randomised controlled trials for collaborative care207 found that collaborative care produced short-term (up to 24 months), but not longer-lasting, improvements in depression outcomes compared with ‘normal care’ (typically GP-based, using psychotropic medication) and also improved patient satisfaction, mental quality of life and non-reliance on medications, although it had little impact on physical quality of life. Regular supervision by a case manager was associated with improved outcomes in mental health care, although this effect depended on including large US studies in the systematic review and meta-analysis. 214

Separate providers naturally tend to work from different sites. Of itself, service colocation creates no new co-ordinating mechanisms and is feasible only when the providers’ size and proximity permit. 122 Nevertheless, one might expect colocation to increase continuity of access and create opportunities for informal information exchange (informational continuity) between professionals and organisations. 63 Colocation facilitates,151 but by itself is not sufficient to produce, collaboration across professional boundaries,215 including informal links. 12,80 Colocation can also be used to promote vertical co-ordination, for instance consultant consultations in the community and/or general practice (‘outreach’). 13 Conversely, many countries retain an admitting-rights model of hospital medical care, by which the primary care doctor (so to speak) follows the patient into hospital, with concomitantly increased specialisation, possibly loss of the generalist role and, in the USA, close control by the hospital. 20 However, joint clinics do not necessarily promote relational continuity of care unless the division of labour between professions is ‘carefully attended to’. 80 Indeed, colocation of general practices was in one instance reported as provoking rivalry and ‘bad behaviour’ (e.g. removing rivals’ notices, misdirecting patients). 216 As an alternative to colocation, evidence of the efficacy of telephone-based cognitive–behavioural therapy for depression, agoraphobia and obsessive–compulsive disorder is now reported by randomised controlled trials. 217

Internal MDTs are readily expanded to become interorganisational primary care teams, with the characteristics described above. Cultural continuity plays a major part in sustaining interorganisational teams. Strong norms about the urgency of providing care at the onset of a stroke promoted the development of stronger, more formal co-ordination between ambulance and hospital services in Sweden. The absence of such norms appears to have contributed to the weak clinical co-ordination of post-discharge care found in both the English and the Swedish health systems. 86

Managed networks

Emergent care networks may not necessarily work well or even form at all. In those cases, ‘umbrella organisational structures’ are required to create and co-ordinate them,23,218 to establish a network identity, policies, structures, care pathways, leadership and resources211 and to be an arena for meetings between the participants. Such an umbrella can ‘solidify the structure for collaboration and support individual professionals working in collaboration to provide care and service for individual patients or clients’. 122 Care networks become managed (to prevent confusion, we avoid the term ‘managed care network’). A network co-ordinating body can develop the sharing of:

1. Care plans, which in principle have much the same character, uses and content, with standardised and formalised pro formas, as within integrated providers. In a care network, care planning requires an explicit division of labour,151 defining each provider’s accountabilities,219 creating referral paths132 and harmonising the different providers’ working practices. These arrangements have to be negotiated and agreed, possibly even formally, between providers,12 as no higher manager exists to impose them. The co-ordination mechanisms so established typically include a single, co-ordinated entry point into an agreed care pathway for the whole care group, shared patient registers, a shared assessment and care plan for each patient, and joint consultations and case conferences.

2. Decision tools,132 for example standardised assessment, triage and care planning instruments,23 care protocols, periodic reassessments leading to early detection of health problems and consequent interventions,24 and care plans shared across all the providers in the network. Activities such as geriatric screening and multidisciplinary assessment promote communication between providers220 (informational continuity). A systematic review151 found that the combination of collaborative care with guideline-directed treatment (a treatment algorithm, stepped care or treatment escalation protocol) and enhanced communication systems improved clinical outcomes in mental health and (on the balance of evidence) reduced costs.

3. Funding streams,23,132 to align incentives among the network participants;175 a ‘demanding’ form of co-operation that inevitably involves formal budgeting, possibly legal arrangements,12 ‘even common ownership of resources’. 23

4. Support activities175 (e.g. integrated data systems23), to produce informational continuity. Standardised electronic records have considerable potential to ease transitions between organisations221 and to improve communications between them,151 including at clinical level (e.g. case conferencing63), although the reality of health IT systems often falls short of this and IT systems are no substitute for oral communication. 221

The network co-ordinating bodies may be either separate from (e.g. NHS referral hubs) or incorporated into (e.g. CCGs) the local or regional bodies responsible for overall health-system governance. Such a body can also co-ordinate the providers’ work61,219 at strategic and/or managerial and/or service delivery levels to ensure access for all the (eligible) population and efficient operations, and to ‘maintain overall accountability for service, quality and cost outcomes’. 23 These activities require ‘collaborative competence’ in handling the problems that typically occur in collaboration. 122 Such competences include communication with other agencies (necessary for establishing clear roles and responsibilities), understanding other agencies’ goals,12 arbitrating over any disagreements and harmonising working practices across the member organisations. 188 Sustaining care networks requires ‘outcomes-based advocacy’ (celebration of network achievements), ‘vision-focus balance’ (between goals and practical activities), systems orientation (‘big-picture’ understanding of the complex, systemic nature of community care) and community linkages. 222 Over time, trust develops between providers, and concomitantly territoriality and competition among them reduce,12 which aids in the transfer of tacit and informal knowledge.

Disease management programmes are a special case of the managed network. Interventions to improve care co-ordination may be more effective when focused on particular risk factors in comorbid conditions or on functional difficulties in multimorbidity. 223 Disease management schemes focus on patients with a single main diagnosis64 and often include supported self-care3 or the use of nurses as care co-ordinators. Often they develop partly because of the inadequacy of GP-based care co-ordination. 5

US evidence suggests that disease management programmes improve chronic disease outcomes, but not necessarily care co-ordination. 224 A meta-analysis of disease management combined with case management (see below) found that improved disease control was associated with educating the providers, reminders and feedback, with education of patients and with incentives. 225 A systematic review of disease management for depression found evidence of improvements in depressive symptoms and greater adherence to treatment, but more admissions to hospital and increased costs. 226 An Australian disease management programme based on generic assessment by nurses, but retaining GPs as care co-ordinators providing mentoring for eight (somewhat overlapping) care groups, produced small improvements in Short-form 36-item health survey scores. 64 In England, a disease management programme for patients with coronary heart disease and heart failure, administered through nurse clinics, produced the small gain of 0.03 quality-adjusted life-year (QALY) per year at an incremental cost of £13,158 per QALY compared with a control group,227 although this ratio easily met the National Institute for Health and Care Excellence (NICE) criterion of cost-effectiveness.

Disease management programmes thus appear well adapted to providing flexible, informational and access continuities of care, at least for the focal disease, and perhaps also relational continuity, but are ill suited to providing holistic continuity. 228

‘Vertical’ networks to manage hospital referral and discharge are often assumed to reduce unnecessary hospital admissions, making case management attractive to payers wanting to avoid the costs of unplanned hospital admissions. 31 In the USA, managed care involves utilisation controls and management practices (e.g. clinical guidelines, disease management programmes, reports on referrals, tests and admissions) that make primary care preventative services available to patients, but also restrict which other treatments or providers patients may access. Some of these schemes were implemented through a predetermined, selectively contracted care network of providers and negotiated lower prices from doctors wishing to join the network. 229 Some US attempts to introduced managed care through HMOs foundered on the problems of information sharing, transaction costs and fragmented provider interests. 20 Managed care schemes have also been tried in Germany, but with limited success in recruiting patients. 60 Such schemes did reduce preventable admissions of over-65-year-olds compared with non-managed care, especially for the more ill patients,127 thus helping to control health-care costs. 230

When patients leave hospital, primary care doctors can reduce the risk of unplanned readmission by contributing to a structured hospital discharge plan, reconciling hospital-prescribed with primary care-prescribed medications, and making an early assessment of the patient’s follow-up needs and resources. The hospital’s contribution is to initiate this plan, sending electronic discharge notifications and structured discharge summaries to the primary care doctor. 231 Preen et al. 232 described such plans being copied to the GP for review and then to the patient and other service providers. The GP made an appointment for the patient to attend within 7 days of discharge. Patients whose discharge was managed in this way rated discharge care arrangements as more ‘achievable’, were more satisfied with their input into discharge procedures and had a higher mental quality of life 7 days after discharge than patients who were discharged in the usual way. Medicare is introducing bundled payments to physicians to co-ordinate patients’ care on discharge from hospital and to provide a practice appointment 7–14 days after discharge. 233

Barriers to care co-ordination by network

Care co-ordination by network involves additional problems and tasks to those found within an integrated organisation:

1. To patients’ transitions between professional silos and between primary and secondary care, a care network adds transitions across organisational boundaries. Potential conflicts of interest and in cultures between organisations are added to those between professions. Differences in provider ownership compound the scope for conflict (see Appendix 2). Conflicting expectations about the role of liaison or link workers, and about appropriate referral criteria, increased professionals’ role ambiguity and impeded the work of mental health-care networks in England and Canada. 151 Explanations of under-co-ordination in care networks adduce ‘the costs, time and resistance to reduction in autonomy perceived by individual providers and services [. . .] differences in culture and ways of working’. 33 Systematic reviews that compare normal care with as many studies as they can find of a chosen model of ‘comprehensive’ or ‘integrated’ care, taken en bloc,92 often disregard these organisational differences; one reason, perhaps, why they often find the overall pattern of results ‘inconsistent’.

2. Networks wishing to strengthen informational continuity have to surmount the original separation of IT systems in each member organisation. 33 Heavier inputs of administrative time are required for running discrepant information systems for bookings, record keeping, audit and monitoring, although good software can mitigate these difficulties. 217 In England, NHS and social care departments use different IT systems. Social care departments also use different IT systems among themselves, with no common coding system. 211

Care networks’ co-ordination structures are added to those that exist within each member organisation, as are transfers across probably less harmonised interfaces. Often ‘invisible’, the greatest of the resulting transaction costs are those that arise from the failure of care co-ordination to occur at all. Transaction cost theory predicts that where services have uncertain outcomes, frequent non-standard interactions and asset specificity, markets (and by implication networks of market partners) evolve into hierarchies, which in these circumstances are more economically efficient. 234 These conditions are typical of care groups with long-term chronic multimorbid health problems. If that is the case, care networks may tend to evolve into integrated organisations, for instance in the way that a predominant provider in a care pathway diversifies into upstream and downstream activities. 185 Gradually, care networks may develop from information exchanges to running joint projects, changing ‘systemic rules’ and altering the health system as a whole. 235

Sampling

Although we intended to trace back from patients’ experiences of the continuities of care to the co-ordination structures that produced them, our aim was to contrast different types of co-ordination structure. We therefore selected a theoretically based, purposive maximum-variety qualitative sample of study sites, chosen as follows to instantiate the main types of co-ordination structures identified in our RQs:

1. Hierarchy, public ownership. Ideally, this would be a single, integrated organisation providing primary medical care, CHS, mental health services and social care. However, no such organisation existed in the NHS at the time of our fieldwork.

   1. The nearest NHS equivalents were trusts or foundation trusts that combined CHS provision with hospital services and/or mental health services and/or social care.

   2. In Sweden, however, such organisations did exist: the vårdcentral described in Chapter 10.

2. Two types of care network (‘federated’) structure pursuing the objectives originally proposed for English ‘polyclinics’ by combining:

   1. General practices and CHS.

   2. CHS and social care providers (sometimes with GP participation).

3. Hierarchical, privately owned commercial or joint-venture primary care providers. We selected a corporate private practice whose owning company also provided other primary and CHS in the same locality.

   1. A hybrid partnership, hierarchical and care network structure: concretely, the standard tripartite NHS model, as a reference point.

   2. A privately owned medical practice networked with public CHS. In the event, the study site mentioned at (3) also satisfied this criterion, since that general practice was networked with NHS-owned CHS and other providers. For further comparison we also examined (in less detail) a Swedish equivalent (see Chapter 10).

When our fieldwork began, no routinely published national data reported general practice ownership or the care co-ordination structures in each PCT (as local health economies then were). We therefore identified possible sites using the professional press (especially Pulse, GP Daily News, Medeconomics and the Health Services Journal), NHS Networks websites and mailing lists, Department of Health and think-tank websites, and national media (especially the BBC and The Guardian websites), supplemented with the researchers’ own knowledge and networks. In consequence, we refined the sampling strategy stated in the study protocol to distinguish two models of federated (networked) primary care co-ordination: networks of professional partnership general practices (e.g. CCGs, general practice ‘federations’)43 and networks of hierarchically organised public providers (confusingly, also called ‘federations’ in some places). During this period of greatly increased workloads for general practices, commissioners (and, therefore, NHS research governance administration) were also being restructured and restaffed. In all we approached 10 sites, and at the expense of much time and effort gained access to five, which together represented the main co-ordination structures that we needed to study.

These methods yielded the following sample of study sites in England. Three sites were whole counties, one was part of a county, and the fifth was part of a larger conurbation. We have pseudonymised the sites. Fuller details are in Appendix 3, but, briefly:

1. Eastmere instantiated the standard tripartite NHS co-ordination structure of separate primary medical care (professional partnership general practices), NHS trusts (CHS; acute hospital; mental health) and social services. The CCG divided its territory into 10 GP localities. The county council divided the same territory into six localities. Towards the end of the study period the county council was awarded integration pioneer status for its plans to join up health and social care services, and one study general practice was given a RCGP quality award. This study site was, therefore, a well-developed example of the standard model.

2. Whinshire was selected for its large, elaborate managed network for co-ordinating care across provider organisations. CHS and mental health services were integrated into a single NHS trust which, until 2012, also provided social services on behalf of the county council (under a §75 agreement). Thereafter, the site reverted to colocated, separately managed services, which had separate data holding but retained a shared referral model. It hosted a personal health budget pilot scheme. Its care network originated from a collaboration between the local practice-based commissioning group and the PCT, and developed into a community interest company. By the time of our fieldwork this network involved nearly all of the general practices in the county, all of the other health-care providers, and social services.

3. Sedgeton was selected for its virtual ‘polyclinic’ housing networked colocated services – that is, organisationally separate community health, acute hospital outpatient and mental health services, all in one building, with mental health inpatient facilities next door. Links were developing with social care. The general practices referring to all of these services were part of a ‘federation’ including these and other providers. Our case study practices were two of those that referred most patients to the colocated services.

4. Tarrow was a vertically and horizontally integrated health and social care provider (hierarchical governance) with an acute hospital, CHS and adult social care (under a §75 agreement), all managed by one chief executive. Mental health services were provided by a separate trust. When the §75 agreement ended, adult social care services reverted to separate county council management. Two virtual ward schemes were being piloted and, later, integrated health and social care personal budgets.

5. Fivecastles contained a corporate general practice belonging to a company that provided other primary and CHS in the county (hierarchical governance). The county also had a combined hospital and CHS trust. The general practice, integrated community and acute trust and mental health trust were, thus, three separate hierarchies, externally networked to each other.

Every site had a NHS treatment centre, a NHS ambulance trust and commercial and third-sector service providers, vertically integrated mental health services and, of course, other professional partnership general practices.

In each site integrated organisations and care network structures coexisted, and the organisations were components of care networks. Following the care process logic of our analytic framework (see Chapter 2), Table 3 summarises which structures coexisted. In each row, the bodies listed in the ‘Care network’ column co-ordinated the providers listed on either side of them. Where a combined organisation provided normally separate services during part or all of the study period, Table 3 brackets them together.

In each site we assembled a sample of key informants based on our patient interviews and discussions with a practice manager or GP, who in turn suggested further informants relevant to the study. This snowballing method yielded the sample of 80 informants summarised in Table 4.

Data collection

Data were collected from informant interviews (using the schedule given in Appendix 4) and managerial documents. Unless the interviewee requested otherwise, all interviews were digitally recorded and professionally transcribed, and interviewees were offered the chance to correct their transcript. Our findings chapters (see Chapters 5–9) identify these informants by the relevant case study prefix (E–W), followed by the letter P (for ‘practitioner’) and a numerical identifier. Our collection of managerial documents was guided by our interviewees, as they were best placed to say which documents were locally seminal. We also collected professional press reportage.

Initial analysis

We collated data about each organisation’s and each network’s care co-ordination mechanisms against the relevant headings in the analytic framework (Table 5), adjusting these categories when necessary.

Collating data against a prior framework revealed where any mechanisms described in the analytic framework were absent. It also provided a structured means of data triangulation so that we could supplement our data with ad-hoc e-mail and telephone enquiries when the data had gaps or ambiguities, or if we had reason to believe that circumstances had changed at a particular site. From any remaining data we inducted patterns, supplementing the analytic framework as necessary. Following the care process logic outlined above, we differentiated the care co-ordination structures and mechanisms found in the following:

1. first-responder organisations at each site

2. care networks and their managing bodies

3. main second-responder provider organisations, insofar as they contributed to the care of our focal care group.

Because of how the NHS is structured, this approach approximately corresponded to and analytically separated three contrasting governance structures for care co-ordination: professional partnerships (most general practices), networks (referral hubs, virtual wards) and hierarchies (NHS trusts, social services).

Sampling

As we wanted to trace the connections (and disconnections) between patients’ experiences of the continuities of care and local organisational and network structures, we sampled patients in the same study sites as the organisational case studies. We also needed to sample patients who had recently experienced complex care and hence who satisfied the following inclusion criteria:

1. being 65 years of age or older

2. having complex health-care needs, defined as having at least two of the following conditions: arthritis; heart failure; COPD; stroke; mental health problem (depression, dementia, psychosis, schizophrenia or bipolar disorder); and sensory deprivation (blind or deaf)

3. receiving care for at least 1 year before the study from at least two provider organisations

4. living in own home or with family.

Exclusion criteria were:

1. having a moderate or severe learning disability

2. receiving terminal care

3. being in the criminal justice system

4. being unable to give informed consent

5. being too frail to give an interview

6. living in residential care.

Our nine chosen GP practices each drew a sample of patients conforming to these criteria and sent out invitations, study information and reminders on our behalf. Patients who were interested in participating were then able to contact the research team using the reply slip and prepaid envelope supplied or, on occasions, by e-mail, telephone or their GP surgery. Because we intended to characterise patients’ experiences of the continuities and discontinues of care qualitatively, we interviewed small numbers of patients in depth until we approached data saturation (n = 66; Table 6). Interviews typically lasted about 45 minutes and, given the patients’ age and interest in the topic, often became somewhat discursive.

The wide inclusion criteria meant that patient characteristics varied across the study sites (see Chapter 5). The decision to take the patient as the starting point for the research meant, however, that they had not necessarily used the care networks that practitioners and managers subsequently described. For example, although the Sedgeton patients had all used services at the ‘polyclinic’, none of the Tarrow patients had used the virtual wards, as many lived beyond the relevant catchment area. Similarly, whereas the Eastmere interviewees were the most limited group in terms of activities of daily living and received the most social care support, Fivecastles was, in contrast, an affluent area with patients who were:

[. . .] very good at going and helping themselves [. . .] well supported by helpful, intelligent and well-resourced families.

GP F03

None of the Fivecastles patients had current social services or care agency support. Several had used private health care or local private hospitals. Patients in Eastmere, Tarrow and Fivecastles lived in ‘predominantly’ or ‘significantly’ rural areas, but the Whinshire patients lived largely, and Sedgeton patients lived entirely, in urban areas.

Data collection

Patient interviews were conducted using a semistructured interview schedule (see Appendix 5). With patients’ consent, their interviews were digitally recorded and transcribed professionally. When patients wished, their carers (for one, his 93-year-old sister; for two others, their spouses; and for another, his son) were present and contributed. We extracted data from the interview transcripts into a standard pro forma (Table 7), a deductive coding frame.

Free-text answers were thematically coded by the researcher who had conducted the field interviews [initially using NVivo (QSR International, Warrington, UK)] and then passed for validation to the principal investigator. One case study site was initially also coded by a third member of the research team to test for consistency of approach. Patient-level data were then collated into a site-wide form of Table 7, with additional categories suggested by the inductive analysis, so that patterns across patient experiences, and differences, could be seen. We also made an emergent (inductive) coding of the responses that the initial deductive coding framework could not accommodate, which included many of the more unstructured, ‘free’ responses. These additional codes (e.g. patient willingness to co-ordinate own care) and the coding of any ambiguous responses were agreed between the principal investigator and the researcher(s) most acquainted with the study site. We thus conducted both a framework and an inductive (‘ground-up’) analysis of the interview data.

Information on health care and referral pathways across the past year was perhaps the most difficult for patients to recall:

Who called the ambulance for you?

I don’t know. When I woke up the ambulance was here.

Did they take you to [hospital name]?

[Hospital name].

So how long were you in [hospital name] for?

I can’t say – I can’t remember.

E13

As validation and to obtain further data we therefore undertook a content analysis of the medical record for each patient for the preceding year, extracting data about three basic activities from the patient’s general practice record into a standard pro forma (Table 8):

1. encounters between general practice and patient (e.g. consultations, treatment and review in the surgery, home visits, telephone calls)

2. actions in which the practice engaged with external providers on the patient’s behalf (e.g. referrals, requests for advice, case meetings), usually without the patient being present

3. patient encounters with other providers (transfers), typically documented as reports, reviews, treatment, information, advice and discharge notices.

In extracting data we focused on pro tem transfers of responsibility for care between the patient, clinicians and organisations. We did not extract data about requests for repeat prescriptions, did-not-attend notices or practice requests to make appointments for, for instance, chronic condition reviews; we extracted data only about the resulting encounter. These data describe the activity captured by general practice medical records (and hence information readily available to GPs) but will tend to under-report community and social care, and will omit undocumented activities. In one practice, for example, CHS staff had read/write access to the GPs’ records, and so CHS activity was more fully recorded there. Practices had different ways of recording information depending, for instance, on which electronic system they used, practice protocols and individual idiosyncrasies. (One GP might, for example, write very full notes including a description of a patient’s holiday; another might write sparse notes relying heavily on abbreviations.) As with practitioner interviews, extracts from patient interviews are indicated by a case study prefix (E–W) and numerical identifier. Medical records are indicated in the same way, but with the additional letters MR (‘medical record’).

Initial qualitative analysis

Our analytic framework (see Chapter 2) listed observable features that would be evidence of continuity and co-ordination of care: the presence of a care co-ordinator and of a care plan; an ongoing relationship between care co-ordinator and patient; reviews; access to services; and information transfer. Combining the interview and medical records data for each patient, we checked for the qualitative presence or absence of each of these features in the patient’s experiences. Applying the same headings and collating within each study site, we wrote for each study site a case study of care co-ordination there, as collectively reported by its patients. We noted any differences between patients’ accounts, any aspects of care co-ordination not anticipated in our original framework, and the broad (qualitative) patterns of care co-ordination reported in the patients’ medical records. By comparing these case studies, we could then qualitatively compare study sites.

Initial statistical analysis

Quantifiable data from patient interviews and the matching medical records were tabulated and coded onto a spreadsheet, including coding for additional (e.g. organisational) variables constructed that might explain any variations between study sites or sampled practices. Where data sets were large enough, we ran simple tests of correlation between activities and sites [using IBM SPSS Statistics version 21 (Armonk, NY, USA)], declaring p-value ≤ 0.05 as the significance level. [The findings chapters (see Chapters 5–9) report which data sets were large enough to be so analysed. In those chapters we use the terms ‘significant difference’ or ‘no significant difference’ where statistical tests were made. Without that qualification, terms such as ‘difference’ refer to qualitative differences.] Simple measures of central tendency and cross-tabulations (‘crosstabs’) were used. Depending on data characteristics, the tests applied were Pearson’s chi-squared test, Fisher’s exact test and analysis of variance. This analysis contributed to answering RQs 1 and 4.

Design

Given our RQ, the ideal comparator against which to compare care co-ordination by care networks would have been a single, integrated organisation providing primary medical care, CHS, mental health services and social care. The NHS had no such providers at the time of our fieldwork, but they did exist in Sweden. A comparison of care co-ordination mechanisms available in these two settings would contribute to answering our RQs 1, 2 and 4. For comparability, we used the same analytic methods as for the English case studies and reused secondary data from surveys of Swedish polyclinic directors about their management of four chronic diseases (coronary heart failure, depression, diabetes and asthma). We did not analyse patient records but did interview a small number of patients (not reported here).

Sampling

Using the researchers’ own knowledge and networks to identify study sides, we selected the following PHCCs (polyclinics) whose organisational structures had no NHS equivalent:

1. three publicly owned, hierarchically structured polyclinics (vårdcentral) in Stockholm (Lisebergs, Kista, Djursholms Iäkarmottagning)

2. TioHundra AB, a vertically and horizontally integrated, publicly owned provider of hospital, primary care and social care (Norrtälje)

3. a not-for-profit trust providing primary medical care and other primary care services (Stockholms Sjukhem)

4. a corporation (private for-profit hierarchy) providing primary medical services (Solna Capio).

All were financed by Stockholms läns landsting (SLL: Stockholm county council). SLL directly line managed the three PHCCs under heading (1) above and commissioned the others through contracts. In size and function the PHCCs were similar to the larger NHS general practices.

Data collection

Data were collected (2011–13) at county and municipal (i.e. commissioner) and at provider level by mixed methods from the following sources:

1. Interviews and discussions with key informants. These included three representatives of SLL (polyclinic line managers and commissioners), the polyclinic head and two nurses at each provider, and at Norrtälje a half-day meeting with the polyclinic head for mental health services and three clinicians to discuss how care was co-ordinated across services. We interviewed six patients at the directly managed polyclinics, three at Norrtälje, four at Capio and two at Stockholms Sjukhem.

2. Grey material including official regulations and guidance.

3. Ad-hoc enquiries from individual experts.

4. Published research found by hand-searching journals.

Interviews were recorded and transcribed. Material from the Swedish study sites was obtained partly in Swedish and partly in English. The former material was translated or summarised for us by native Swedish speakers.

The surveys (2005 and 2011) replicated a US survey to discover the use of recommended care co-ordination practices and of IT for co-ordination in primary care. 252–254 The response rate was 37%, with a bias towards the larger centres. Whether or not there was a higher proportion of public than private PHCCs replying is unknown, as there are no reliable national data about PHCC ownership.

Initial analysis

To enable comparison we made the same initial analysis of the Swedish as of the English organisational case studies. Chapter 10 presents these findings.

Nested framework analyses

Given our ‘inside-out’ study design, the framework analyses were nested in the following order:

1. We started from the findings (see Chapter 5) about patients’ experiences of the observable features of continuity and co-ordination of care (see Chapter 2, Processes and structures for care co-ordination).

2. Next, we compared the pattern of patient experience at each site with the initial analysis of care co-ordination mechanisms in the first-responder organisations at that site. Chapter 6 presents these findings. Where they matched, the organisational case study provided a prima facie explanation of patients’ experiences. For example, patients often reported difficulties accessing ‘their’ preferred GP, whereas the general practices reported overloaded appointment systems and high GP workloads. Where patients reported discontinuities of care for which our case studies contained no apparent organisational explanation, we sought explanations in non-organisational causes255 such as patients’ own resources and decisions. However, our health worker informants more often reported care co-ordination mechanisms, or the absence of such, not mentioned in the patient experience data (e.g. case reviews which the patient was unaware of). We consider in Chapter 11 what that signifies.

3. We repeated step 2 for the case studies of the ‘second-responder’ provider organisations at each site. Again, we interpreted organisational case study findings that matched patients’ experiences as being prima facie explanations of those experiences. Chapter 7 presents the findings.

4. Next, we read off from our informants’ accounts how care networks (and other factors) had influenced the providers’ organisation and management. At this stage of the analysis the most informative data were usually our informants’ narratives (a local NHS history) of how care networks had developed in each study site, the rationale for those developments and the practical constraints on them. Chapter 8 presents these findings.

5. Finally, we traced the ways in which the governance of local health economies had shaped the working of the care networks. Here, too, informants’ narratives of recent NHS history in the study site were often the most informative data. These findings are in Chapter 9.

At each stage draft analyses were circulated among the researchers acquainted with a particular study site, to check for accuracy, completeness and face validity. We systematically compared findings across study sites to reveal similarities and differences.

Systematic comparison of organisational case studies

We systematically compared the initial analyses of organisational and care network structures in order to find common patterns in care co-ordination mechanisms across sites, and differences between sites that might reflect differences in organisational and network structures, including any differences between care networks and organisationally integrated providers. We therefore compared:

1. sites where acute and community services were integrated into one organisation (Fivecastles and Tarrow) with sites where they were separate organisations (Eastmere, Sedgeton and Whinshire)

2. sites where community and mental health services were integrated into one organisation (Whinshire) with sites where they were separate organisations (Fivecastles, Eastmere, Sedgeton and Tarrow)

3. sites where community services and social care had been integrated into one organisation (Tarrow) with sites where they were separate organisations (Fivecastles, Eastmere, Sedgeton and Whinshire).

In this way we addressed RQs 1 and 3. By comparing these findings with our original analytic framework and hypotheses, we also began to address RQ 4. We furthermore systematically compared the corporate general practice (Fivecastles) against the professional partnership general practices in the other four sites, thereby addressing RQ 2.

Cross-country comparison

We extended the systematic comparison of English organisational case studies by adding in our initial analysis of the Swedish polyclinics. The methods and process of analysis were essentially as described above.

Inducting patterns of managerial discretion

Given our RQs, we filtered our organisational data for any that suggested ways in which managerial decisions had produced or frustrated continuities of care. For this purpose, ‘bad’ managerial decisions (reducing continuities of care) are as informative as good. We collated these findings across sites, thereby answering RQs 3 and 5.

Different experiences of general practices and other first responders

Two main differences across sites appeared to be in regard to patients’ experiences of general practices and to other first responders.

Although patients everywhere reported difficulty in seeing the same GP, that appeared to be particularly difficult in the corporate general practice. As far as patients could ascertain, it had no full-time GPs, with the part-time core staff being supplemented by trainees and locums.

Now we have three main doctors and a lot of doctors who come in one a day or this day or that day [. . .] they are both very considerate and very efficient and effective. But trying to get an appointment with them is extremely difficult.

Patient F20

Patients explained:

[GP name] himself is only there on Mondays and he’s got a partner who manages the general practice as a business, they’ve got more than one surgery and he’s only there 1 day a week.

Patient F15

They run their practices very much more as a business than they used to [. . .] they certainly don’t want to overstaff the practice with more doctors than necessary, so obviously it becomes a little bit more difficult to get an appointment.

Patient F24

The medical records also indicated significant differences between professional partnerships in the proportion of patients seeing a ‘designated’ GP (see p. 43): 8% in Whinshire versus 67% or more in Sedgeton, Tarrow and Eastmere.

Concomitantly, although GPs dealt with patients mostly through appointments at the surgery or by telephone, there were significant differences in willingness to undertake home visits. Overall, 5% of encounters recorded in the medical records were GP home visits. This ranged from 1.6% in the corporate general practice to 8.3% in one of the more rural and dispersed study sites, where GPs still did make home visits to their patients. However, three professional partnership general practices recorded no home visits at all.

Patients also noticed that the corporate general practice often sent them pharmacological recommendations to change their medication. Mostly, patients felt that this surgery was thorough, up to date and efficient, with good information systems and GP records, and that it was ‘very good at checking things out. And I assume, and I think I’m right in assuming, that the hospital’s treatment gets advised back to the surgery and they’re kept’ (patient F21).

The medical records did not necessarily corroborate this impression. As noted, the quantity of recorded information flows on record was below average in Fivecastles, which was one of the three areas with an above-average incidence of poor information transfer.

Another difference between sites was that patients in Fivecastles and Sedgeton hardly commented on the idea of bypassing general practice for care reviews or queries, mentioning only standard recalls for urology and cardiology, for example, whereas in Eastmere and Whinshire patients maintained direct contact with the hospital in the event of exacerbations: ‘Well [the heart specialist] told me to ring there, his office, if I get serious problems’ (patient W40).

Different experiences of interorganisational co-ordination

In Fivecastles, patients’ experiences of transfers to other providers appeared to differ in two ways from the other study general practices. Fivecastles patients engaged mostly with the general practice or hospital and less so with a larger primary care network. CHS involvement was significantly lower (recorded for only 33% of sample patients) than in other sites (100% in Sedgeton). Furthermore, the Fivecastles practice seemed very willing to refer patients for NHS-funded treatment by private providers, including private diagnostic and treatment centres, but also to other private providers. Our study patients did not mention any of the other facilities in the county owned by the same company.

Patients at different sites reported contrasting experiences of their general practices’ willingness to refer them as private-payer patients to private health-care providers. The Fivecastles general practice readily did so. Several patients had GP referrals for private treatments at the DGH, local hospitals and London private hospitals, especially for cardiology and conditions such as a brain cyst where rapid treatment seemed important. Even one patient in social housing ‘went private’ after NHS providers were slow to diagnose a twisted gut. The Sedgeton general practices did not facilitate such referrals:

[I]f I have not heard in a few weeks’ time I shall make my own arrangements . . . I’ve never found a doctor who’d refer me privately, even though you say, ‘I’m happy to be referred privately’.

Patient S45

Nor were the study practices in Whinshire, Eastmere and Tarrow very active in referring patients for privately funded treatment. Patients in the Fivecastles sample were three times more likely than the study average to be referred to a non-NHS provider or to be treated privately. Patients in Whinshire (and Sedgeton) reported waiting several times longer than the NHS national target for secondary care.

Different experiences of ‘second-responder’ care co-ordination

Patients in Fivecastles described a certain freedom of movement between CHS and hospital care. After discharge to CHS, they were able to return to the DGH for a limited time period without needing further referrals, although one talked of the difficulty even then of getting more than one thing done at a time (e.g. blood tests while in hospital, the lack of which would have necessitated a return to primary care if he had been less assertive). In the same way, patients could access the day hospital, which they praised for its holistic approach:

[T]hey also look at your diet, they look at all aspects of your present living. You know, I mean there’s a whole team of nurses there and there’s a resident doctor.

Patient F26

This ‘one-stop shop’ gave access to physiotherapy and occupational therapy as a minimum, together with access to speech and language therapists, dieticians, nurses and geriatricians.

Whinshire patients were the only group describing unreliable hospital-to-hospital information transfer, and this was the only study site to refer patients consistently to two DGHs:

I got a letter from the haematologist at [town], she wanted me to see a neurologist and I’ve got the letter that she sent to my doctor saying she’s had no information from the [other town] neurologist.

Patient W42

A haematologist wrote:

[I] understand [patient] was seen by a neurologist at [other DGH] in order to investigate further funny turns: Unfortunately, I have received no correspondence whatsoever from [other DGH] and am at loss to understand exactly what is going on. [Patient] tells me [neurologist] does not think TIAs and has been put on a reveal cardiac monitor. I would be very grateful if [other DGH] consultants could copy me in on their clinic letters.

WMR42

In summary, the main similarities in patient experiences across our study sites were:

1. GPs were the main de facto care co-ordinator.

2. Few patients had designated care co-ordinators.

3. The one-item-per-consultation rule was seen as impeding holistic care.

4. Overall care plans existed for only a minority of patients (even in this frail care group).

5. Relational continuity was patchy.

6. Telephone contact with the GP was partly replacing face-to-face contact.

7. Nearly half of the patients had no regular reviews.

8. Patients reported having little time with the GP.

Differences were:

1. In Fivecastles and Whinshire we found:

   1. lower perceived relational continuity with the GPs

   2. fewer home visits, especially compared with Eastmere

   3. lower levels of engagement with CHS, especially compared with Sedgeton.

2. Higher use of private providers in Fivecastles.

3. Care co-ordination conducted predominantly by the general practice in Eastmere and Tarrow, with more mixed responsibility in Sedgeton, Fivecastles and Whinshire.

4. Lower levels of recorded information in Fivecastles and Tarrow, but apparently good use of IT systems in Fivecastles.

Care co-ordinator

The patients we interviewed usually felt that they knew where to seek additional health care and that it was from a GP, preferably their ‘usual’ GP. Few of them, however, were aware of having a designated care co-ordinator, an impression that their medical records confirmed. Nevertheless, GPs in our study practice believed both that being a care co-ordinator was the proper, normal role for a GP – ‘You are always a case manager for the patient you are seeing at any one time’ (GP WP01) – and that GPs were best qualified for it:

We need to take a more doctor-centred approach to chronic disease management. The nurses are good but if e.g. [they] follow diabetic protocol – such patients will inevitably also be hypertensive and have chronic artery disease (CAD). A GP appointment may not be so cuddly or client friendly but it does deal with the whole smorgasbord.

GP WP01

Frequently GPs’ co-ordinator role was tacit rather than openly discussed. GPs would be care co-ordinator by default when patients did not fit neatly into any other category with special care co-ordination arrangements.

For particular classes of patients, having a named GP with specific care co-ordination responsibilities was becoming more common. For example, one of the Eastmere practices had a named GP for each cancer patient. The corporate general practice had a flagged list of vulnerable patients for whom the receptionists were told to allow ready access to the GPs. Some general practices (Eastmere, Tarrow and Whinshire and briefly, but financially unsuccessfully, Fivecastles) provided personalised care by in-reach into nursing homes, with each nursing home having a responsible GP (Eastmere) and/or with one GP responsible for all nursing homes with which their practice dealt (Whinshire and Tarrow). When patients entered a nursing home, their ‘usual’ GP might or, more often, might not be the one in charge of their care. In one Tarrow practice, the GP(s) who had previously provided services to the nursing home continued to provide those aspects of care not covered by the General Medical Services (GMS) contract. Practices did not always have the option to provide GP cover for community hospitals (as in Eastmere). Although GPs valued their care co-ordinator role, they were content to delegate some of it to, for example, community matrons, advanced nurse practitioners (ANPs) or other external care co-ordinators (see Chapter 7) for some patients at the top of the Wagner pyramid. However, these were only a minority of their older frail patients.

Against this, the changes to the division of general practice labour described below were reducing relational continuity of care for the other patients.

Care plans

One reason why care plans were not very visible to patients (see Chapter 5) was that few had them. In the general practices we studied, this was because GPs were cautious or guarded about the relevance and utility of care plans, and because of the time required to produce and update them (and ‘Health Passports’). There was also insufficient time to encourage care planning more generally at practice level. Indeed, one GP suggested that written care plans were not a GP function (rather, they were a focus for nursing care, and not required if a good reactive system was in place). As with the care co-ordination role, nursing and care home residents were something of an exception. In Tarrow and in part of Eastmere, GPs’ coverage of whole nursing homes aimed to create integrated care plans and give standard weekly ‘ward round’-like medical care. In Eastmere, these personalised care plans were being piloted for every dementia patient on the Quality and Outcomes Framework (QOF) register in two general practices and:

every [. . .] nursing home and residential home bed [. . .] mainly around end-of-life care and anticipatory care plans as well as medication reviews.

GP EP19

There, and in Tarrow, these patients would get a care plan within 1 day of admission to a nursing home; their care plans would always be reviewed on hospital admission and discharge, and the nursing home GP became recognised as a GPSI-like expert in that activity.

Of our patient sample, 56% were aware of having their care reviewed regularly (if infrequently, for some). For those for whom this happened, our general practice informants told us that care reviews were made by recall, a home visit, a telephone call to the patient or a practice meeting (without the patient present). On occasion, selected patients were also the subject of a case meeting or MDT meeting. GPs’ review and recall processes (e.g. medication reviews) were driven largely by QOF targets:

I think QOF has worked very well, you know, in terms of people actually getting their checks. And an awareness of targets and, you know, management of diabetes.

Community team lead SP06

Or, more guardedly:

There are some useful things about QOF – it does flag things to be done – but some are utterly meaningless.

GP WP01

Among the meaningless things, this GP counted the depression questionnaire. Practice nurses did much of the routine monitoring, especially for QOF (annual reviews for COPD and diabetes). These were the care reviews that some patients found ‘superficial’ and did not count as overall reviews of their care. However, many of the study general practices went beyond what QOF required. In Tarrow, for example, one of our study general practices regularly monitored patients with arthritis (a condition not included in QOF until 2013). Regular GP visits to nursing homes were an opportunity for care review, and were becoming more common for housebound patients, too (e.g. in Tarrow). Care reviews were also sometimes made opportunistically when GPs responded to changes in a patient’s circumstances (e.g. in Eastmere). Reviews were also made at the instigation of a multiagency (care network) team.

The corporate general practice had protocols for how often patients’ care should be reviewed, depending on the patient’s condition and treatment, and it used the whitelist described above. A senior GP regularly led audits of condition prevalence and follow-up. This was explained as a necessary response to early discharge and the delegation of follow-up to primary care:

Because obviously patients do not tend to be kept on for outpatient appointments quite like they were, they tend to be discharged back to primary care now. So there is that concern about are we recalling these patients and keeping an eye on them. And we did find actually that there were definitely a few patients that were not being followed up as well as they should have been. So those sorts of processes can be quite useful.

GP P03

Partly this was driven by QOF and the CCG’s emphasis on audit, but some were ‘driven by a significant event that gets flagged up and we look at it and we think actually, you know, this is an opportunity to learn’ (GP FP03).

Our patient interviewees, however, often described the lack of follow-up or review by their general practice after they were discharged from hospital. Only in one of our five study areas did more than half of eligible respondents agree that they had been checked up on once back at home.

The uneven coverage of care reviews at the individual level thus appeared to reflect:

• single-condition reviews being given as mandated by QOF

• general practices using their own discretion in deciding when to do more care reviewing than this

• our inference that, as GPs thought care planning was laborious and of limited utility (see above), the same thinking applied to formal care reviews (a corollary of care planning).

Continuity of access: missing and extra services

Our patient interviewees cogently described their experiences of the one-item-per-consultation rule, irregular continuity of contact with their ‘own’ GP, little time with the GP and the role of telephone contact and triage. One immediate explanation was heavy and increasing GP workloads. Care co-ordination of older people with complex needs was only one – and not necessarily the biggest – cause. GPs also faced demands for quick appointment times. Workload pressures on GPs were manifested in:

1. The one-problem-per-consultation rule, which reduced cross-sectional continuity, although as noted GPs did not apply it rigidly. GPs at one study practice in Eastmere actively checked about patients’ earlier problems when they presented with new ones and maintained high relational continuity with these patients. In one of our Whinshire study practices, urgent appointments were for one issue only but ‘by negotiation’ several might be dealt with at other times. This was an evident source of tension for GPs as well as patients. One patient’s medical notes recorded that the one-problem-per-appointment rule was explained, but the patient had ‘unrealistic expectations about what can be safely managed in 10 minutes . . . [and] about being symptom free as she ages’ (SMR43).

2. Disinclination to provide ‘extended hours’ services. Patients who needed such services, therefore, had to use the separate out-of-hours service, ambulance or ED, reducing longitudinal continuity of care.

3. Practice lists of patients were generally replacing personal lists, so that responsibility for each patient was shared across the whole practice instead of each patient having a named GP. However, one of our Sedgeton practices did still operate personal lists and in others patients had a named ‘usual’ doctor. In one of our Tarrow study practices, medical records showed that the patient’s usual doctor made the home visit even if another GP had taken the telephone call that triggered it. Like their patients, our study GPs noticed the reduced relational continuity of care.

[W]e don’t have personalised lists which is a real shame . . . We don’t have time to do the things that make general practice good any more. We don’t have time for [. . .] domiciliary visits any more. . . that popping in doesn’t exist any more.

GP EP19

Over the years you develop a long-term relationship and they consider you ‘their doctor’. But I can’t deliver a personal service – the pressure on appointments is too much.

GP WP01

1. Restricted capacity to add services to the general practice (Whinshire) or to contribute to the meetings and service development activity of care network co-ordinating bodies (see Chapter 8). One of our study GPs had reduced his hours at his own general practice in order to assume a CCG role, something that the practice did not appreciate.

2. Use of medical labour-saving techniques such as telephone and e-mail consultations, and employing non-medical clinicians to do such work as QOF-based care reviews (and often more work than this).

3. In Eastmere and Fivecastles, both patients and health workers described receptionists as barriers to patients’, and even ANPs’, access to the GPs.

The study general practices’ additions of new services and specialisations reflected the GPs’ personal background and interests. Thus, both of the Whinshire study general practices offered minor surgery, to that extent substituting for hospital treatment. One practice had also developed further services for procedures that otherwise would have required a referral to hospital or treatment centre (e.g. spirometry, electrocardiograms and 24-hour blood-pressure monitoring). Another general practice in Whinshire had recruited a GP with a cardiology specialism, but not as a GPSI because of the cost. The presence of specialised GPs encouraged the recruitment of patients and resulted in a better service for them, but also tended to create more clinical and care coordination work in treating patients with complex care needs, because if a specialised GP was present the other GPs tended to use them. Similarly, the GPs felt that a NP worked in effect like a GP, creating their own workload and pulling patients into the practice.

Like others of their kind,256 the professional partnership practices employed NPs, practice nurses, health-care assistants and phlebotomists. The small corporate general practice had the lowest number of support staff; they had a part-time phlebomist and the practice nurse, but they also had a pharmacist who conducted medication reviews, an unusual occurrence in a non-dispensing practice and, effectively, a delegation of responsibility for signing off scripts. Because the pharmacist ‘knows a lot of the patients from being behind the counter at the chemist’ (GP FP03), this arrangement was also ‘able to give [the GPs] a different perspective on what’s going on with [the patients], you know, whether they’re remembering to take their medication, who’s picking up their medication’ (GP FP03). The post was funded by savings from the prescription budget. GPs might also have a clinical interest in working with older people with complex needs. For instance:

What’s required, I think, move common ailments and health advice and the, sort of, end of my job which I do which I don’t necessarily have to have had the degree of training that I’ve had to do it, taken away from me, giving me more time to spend with the frail.

GP SP10

In some instances where new services were being introduced (Eastmere), we found that, although GPs were contracted to provide the CCG with outcome indicator data for evaluating a new project, they were uninterested in doing so even though they would be paid for it:

The requirement to provide outcome data had been written into contracts by the CCG but GPs are saying, ‘No, can’t do it, don’t want to do it even for the money’. And I can only see that getting worse.

GP EP19

Heavy workloads had neutralised financial incentives at the margin. GPs’ personal interests might also orient their general practices towards expanding their income, increasing the range of practice activities accordingly. One of our study GPs, for example, had set up a private company in parallel with his general practice.

Internal care co-ordination

Over one-fifth of our sampled patients felt unable to assess whether or not the people looking after them worked well as a team, but most of those responding (77%) thought that co-ordination (‘teamwork’) was evident. Many cited their general practice as evidence. Another general practice response to increasing workload was to delegate more of it to practice nurses, who ran many review clinics (e.g. diabetes, asthma/COPD control), did some care planning, provided routine health checks, managed common ailments and gave health advice. At one of our Sedgeton study practices, the practice nurse acted as case manager for MDT patients, reviewing their care at a home visit and putting them in touch with appropriate services, although we were told that this arrangement did not work well for patients with mental health problems. Salaried employment of one GP by others (practice partners) or by a corporation is also becoming more common. 19 This study included instances of both arrangements.

The GPs held practice-level ‘team’ meetings to deal inter alia with care co-ordination matters, especially expected hospital admissions or discharges for patients with complex needs, as well as to handle referrals. One of the Eastmere general practices held fortnightly hospital discharge meetings to discuss elderly frail patients and ensure that they were contacted and followed up. We were told that this is becoming more common as patients are being discharged with more tasks back to primary care. Both study general practices in Whinshire held formal weekly meetings (which also covered clinical teaching) as well as ad-hoc informal discussions, often daily:

[W]e have particularly good relationships between the partners and always discuss patients. We have what we call the ‘flight-deck’ with six computers and a big table where we congregate on an informal basis and bounce decisions.

GP WP01

Similarly, in the corporate general practice, the GPs, in-house pharmacist, administrator and practice manager also met weekly to manage referrals and to support locums and trainees:

[W]e can talk through whether a referral is legitimate or whether it’s something that we can sort out in house [. . .] I think it probably works well because we’re a small practice so, you know, there aren’t too many people at the meeting and we all do get on very well as a team.

GP FP03

Some, but not all, of the (internal) team meetings in the other study practices also involved a designated receptionist, district nurse and ANP with the designated GP (i.e. practice employees besides the GP partners, their employers) to discuss patients with multiple complex needs.

The study practices also hosted meetings for, and their GPs participated in, interorganisational primary care teams (see Chapter 8).

General practitioners in the study general practices made efforts to share information (in addition to patient records) with each other. This had been formalised especially in our Fivecastles study practice as their strategy for compensating for a fragmented GP workforce. Part-time salaried working is becoming more common among English GPs, and so this information-sharing culture seems to reflect a more general imperative. Practice IT systems were always substantially computerised. Staff in both Eastmere general practices, for example, talked of good communication and records. However, to varying extents, general practice information systems were also administratively closed to, and technically not interoperable with, other providers’ or care networks’ health records systems. The data available to the Eastmere CCG were reported as being ‘extremely poor’ (EP19). Partly this closure might have been because GPs anticipated, or had already experienced, that having too much information about a patient ‘clogs up the record and leaves us [GPs] to cut the wheat from the chaff’ (GP WP02).

This GP found that, as a consequence, it was becoming harder to spot the important action points quickly, which might raise risk management issues. Nevertheless, general practice information systems were gradually becoming more accessible to external health-care providers (see Chapter 5).

Despite all the above expedients, general practice resources were simply too small in scale and narrow in range to provide all the care that patients with complex health problems needed at home or in a community hospital. One GP saw organisational integration of a wider range of services within his general practice as an ideal ‘Oh, my dream . . . I would love us to hire our own district nurse, health visitor, mental health nurse’ (GP EP19).

Referral to additional providers, and the ensuing care co-ordination, still took place by default through the consultation model. GPs had a monopoly on some referrals and authorisations to providers outside their general practice, in particular to consultants. Post-discharge hand-back of the patient from consultant to GP was usually immediate, but in one of the Tarrow virtual wards (see Chapter 8) the consultant retained responsibility for patients.

A corporate general practice

Some of our patient interviews had noticed, and speculated about, the biggest organisational difference among our study general practices: that one of them was owned by a corporation that owned and ran other general practices and other health services. In that study site, corporate ownership appeared to have three main consequences for co-ordinating the care of people with complex care needs:

1. Because the doctors also worked elsewhere, each made only small inputs to this practice, resulting, as their patients noticed (see Chapter 7) and the GPs acknowledged, in low personal continuity of care. Its well-developed IT system (on which patients remarked) was in part a compensatory mechanism to produce at least some longitudinal continuity of care. There were also, as patients had again noticed, few home visits.

2. Two doctors were accountable to the third, who had wider clinical management responsibilities across the company.

3. The company provided other services in the locality. An ophthalmology clinic owned by the same corporation was on the floor above the study general practice, but run completely separately (although the practice’s doctors could refer to it). Contrariwise, patients at a nearby community hospital had received services from the corporation’s doctors rather than their own GPs, many of whom worked in the (independent) general practice across the road. Furthermore, the corporation reportedly sent different doctors each day of the week. The company had also taken over some local care homes and used its doctors to provide and co-ordinate care there, although that project had been discontinued partly for financial reasons. Our study practice had once hosted private physiotherapy and chiropractor services, but was too small to host a NHS physiotherapist or to exploit the easy links to additional services that could, in theory, have resulted from being part of a wider organisation. It thus appeared little different from other nearby general practices in terms of its patients’ access to other services run by the company – an unexpected finding.

Professional partnership general practices also have made certain other innovations that the corporate general practice pursued. 256 It recruited an in-house pharmacist, which released doctors from the reviews associated with signing off requests for prescriptions to do other work, and readily referred patients to private diagnostic services. It was not obvious to us whether readiness to refer to private providers of diagnostic services reflected its corporate ownership, small size or normal referral patterns in that CCG. In addition, the corporate study practice had a whitelist of patients who were allowed easy access to the GP:

[I]f a patient rings up [and] that name means something to [the reception team] and they will make sure that they give them an urgent appointment [because . . .] what you want to be sure you’re not missing are the people who do not like bothering the doctor but, you know, are quietly sitting at home and not doing very well.

GP FP03

In such a small general practice, ownership seemed to have less influence than organisational size on care co-ordination; in most aspects of care co-ordination except those indicated above, the corporate general practice was similar to the professional partnerships.

Integrated information systems

In Whinshire, some general practices used shared medical records to which health and social care staff of other organisations, besides GPs and general practice staff, had read–write access. The Sedgeton general practices had an urgent clinical care ‘dashboard’ that, as part of the local anticipatory planning process introduced by the integrated primary care teams (IPCTs) (see Appendix 5), could be used to highlight patients whose needs or service use were increasing. They planned to give the out-of-hours and the ambulance services real-time access to it, and to include care plans on it, but at the time of our fieldwork these additions had not yet been implemented.

Other first responders

It was noticeable how little the patients we interviewed mentioned out-of-hours services, in whose use there was a marked decline nationally during the period of our fieldwork. 17 In Eastmere, out-of-hours medical cover was simply scarce. In contrast, EDs faced increasing demand. The care networks described below responded in Sedgeton and Whinshire by introducing ‘front-door’ teams to triage incoming patients with complex conditions, so as to divert (i.e. refer on) those whose condition allowed it to other providers. The care network co-ordinating bodies (‘hubs’) in Eastmere and Sedgeton also accepted self-referrals, but as their main role was care network co-ordination (routing or diverting referrals originating from other provider organisations), we describe them in Chapter 8.

Organisation-level co-ordination structures invisible to the patient

Some co-ordination mechanisms were, for patients, behind the scenes. Patients might notice the effects but not the co-ordination mechanism itself. Thus, some case conferences (or equivalents) and care reviews involved only general practice staff and not the patients, who were not always aware that these were taking place. Patients knew that non-doctors at the practice provided, and even co-ordinated, some of their care, but had neither occasion nor reason to know what line-management arrangements achieved this; nor, similarly, what information systems transmitted (or lost) information. With our patient sample, it so happened that few had used the additional (e.g. minor surgery) or colocated (e.g. mental health support) services that some of the study general practices provided.

Non-organisational influences on care co-ordination

Some aspects of patient experience of care co-ordination had obviously non-organisational explanations: patients liking to have their ‘own’ GP; whether patients took the initiative in co-ordinating their own care, passively complied with or resisted health workers’ initiatives; and what personal capacities and resources, including their state of health, patients brought to their care. GPs’ own backgrounds and interests also influenced how care was co-ordinated and what resources were available for it.

Higher-level constraints on care co-ordination

At health-system governance level, the contractual and QOF framework in which GPs worked, CCG imperatives and national policy and guidance were all reflected in GP workloads and priorities. More immediate external constraints on general practices’ co-ordination of patients’ care were as follows:

1. availability of budgets or payments for additional care co-ordination work

2. the presence and scope of multiagency teams to provide access to services that general practices did not provide

3. the availability of external providers willing to provide additional services on general practice premises

4. the openness or closure of other providers’ information systems to general practices, and vice versa

5. the co-ordination of post-discharge care by hospitals alongside, not instead of, general practices, especially in Eastmere and Whinshire.

General practices depended on care networks (see Chapter 8) to address these issues.

To summarise, general practices in our study sites had extended services and introduced new working practices for, among other reasons, co-ordinating complex care. However, the greatest constraint – seldom mentioned in research or policy despite being so obvious – was the small scale and narrow scope of general practice services. Patients requiring complex or round-the-clock care were bound to be referred to external providers.

Care co-ordinator

Only one of our patient interviewees considered that they had a nurse as their care co-ordinator. All of the study site CHS identified case managers for patients with complex, long-term conditions, but there was considerable variation across sites and between services as to which patients had one, the criteria for having one and arrangements for assigning that role. To illustrate the complexity, in Eastmere, patients with two or more unplanned admissions in the past year and taking four or more medications could be assigned a CHS care co-ordinator. In Whinshire, post-acute patients were covered, with each ANP covering (as in Eastmere) a number of designated general practices. In Sedgeton, patients referred to a short-term CHS service (intermediate care, intravenous community therapy, community rapid response) were variously allocated a clinician to assess them and then transfer them to the necessary professionals. Outside the virtual wards, case management in Tarrow was so limited as to make the question of the criteria for having it ‘a bit academic’ (TP02). These eligibility criteria suggest that patients were sometimes right to think that they did not have a nurse (or similar) care co-ordinator. One virtual ward scheme uncovered high levels of unmet need, with some complex housebound patients seeing only a district nurse:

[W]e’re identifying patients that have never seen anyone. And actually I find it shocking a little bit that I’ve met really poorly people with end stage lung diseases, diabetes that, I suppose, then the district criteria’s they’re housebound and that they have a nursing need.

Nurse TP10

Usually the co-ordinator was an ANP or equivalent (e.g. community matron). In Sedgeton, each discipline had a professional responsible for each client, so ‘a therapist who is allocated a case will stay with that client right the way through their journey whilst they’re with us’ (clinical services manager SP07).

In Fivecastles, the care co-ordinators were usually nurses, but there, as elsewhere, a therapist, specialist nurse or other professional might undertake the role:

I’m in charge of that patient, it’s my responsibility to make sure that we reach the goals we’ve set . . . because the intervention is in their home, you’re treating that patient as a whole human being, so you’re looking at not just physiotherapy needs, you’re looking at what that patient needs to keep them safe and independent within their home.

Physiotherapist F08

Health workers’ propensity to take on the care co-ordinator role was influenced by the length of the intervention and the degree to which it was task focused. For hospital outreach care, the consultant was the de facto care co-ordinator.

Community mental health services tended to have a single point of access with onward referral to teams or localities and, because of resource constraints, an often pragmatic approach to case allocation. Those assigned a case [usually community psychiatric nurses (CPNs)] tended to act as assessor and care co-ordinator, and to make onward referrals. In Tarrow, the community mental health team usually acted as care co-ordinator, but the older persons’ community mental health team meeting would allocate its new referrals one professional, as would the intermediate care team if ‘I don’t think anything’s moving along’ (TP04). Two care networks (see Chapter 8) also assumed some responsibility.

Care plans

Only a minority of patients believed that they had care plans, and these were plans made by separate providers rather than an overall care plan. Our health worker informants confirmed that each profession or service tended to make its own unidisciplinary treatment plan rather than ‘read off’ its contribution from an overall holistic care plan. Copies of the unidisciplinary plans for each patient would sometimes be bundled together at the patient’s home. Various documents were termed ‘care plans’, ranging from a checklist agreed with clients (Age UK) to personal care plans, clinical management plans, contingency plans, advanced care plans and district nurses’ patient records. For example, Whinshire district nurses made:

1. anticipatory (generic) care plans

2. more specific (e.g. wound management) care plans

3. ‘goal-centred’ or ‘management’ plans for complex patients.

The Sedgeton specialist nursing teams’ plans set goals, but without strict time limits. Achieving the outstanding goals (if they remained realistic) was what mattered. Such plans typically included patient and GP inputs. In Eastmere, ‘personalised’ care plans assisted longitudinal continuity of care by indicating, for example, what oxygen level was normal for a patient, which aided timely discharge by showing when the patient’s condition had restabilised at its normal level. The Eastmere community mental health team attached considerable weight to ‘care’ (contingency) plans as they helped in managing mental health episodes. Prompts such as ‘have you made yourself a cup of coffee?’ or ‘have you had a cigarette?’ (EP13) promoted longitudinal continuity by enabling unfamiliar carers (e.g. out of hours) to assume the role of more familiar ones.

As case management was not universal, neither were overall care plans, especially for patients on intensive programmes of care to prevent hospital admission or expedite discharge. On a generous definition, most services had some kind of care plan, the key exceptions being short-term services (e.g. housing, OTs). Despite the colocation of so many services there, Sedgeton patients were unlikely to have one overall care plan. The CHS assembled a set of service-specific plans or goals on paper in a ‘yellow folder’ at the patient’s home, but ‘in a crisis it takes a good carer or relative to recognise the need to bring any of the individual care plans into hospital’ (geriatrician SP15). Patient involvement in the care planning process was also often limited, particularly if mental health workers were involved, because we are ‘talking about very vulnerable individuals by this stage’ (team facilitator TP04).

National guidance recommended the Care Programme Approach for planning complex mental health care, but our informants explicitly mentioned it only in Sedgeton and Tarrow. Even among the patients with long-standing complex needs on whom we focused, it was noticeable how few (14%) received any social care, let alone a plan for it extending beyond a 6-week reablement programme.

Care review

About half of the patients in our sample reported having care reviews. Care reviews, where they did exist, were more common for the simpler, condition-specific interventions. Apart from patients in a proactive care scheme, it was the complexity of care that was not reviewed. Review frequency, and which patients had them, again varied across study sites.

Whinshire CHS patients could always request a care review but, we were told, rarely did. For patients who had them, their ‘goal centred district nurse care plans’ specified timed care reviews. Post-acute care patients were reviewed whenever an ANP visited them, with other patients reviewed when they came to a clinic. In Fivecastles, patient reviews were being formalised into ‘visit standards’ ‘so if the patient’s seen daily in a 7-day period, at least one visit is undertaken by a qualified practitioner’ (locality manager FP04). Review frequency reflected individual need:

I mean sometimes I’ll review somebody monthly, sometimes it’s weekly, basically our role is teaching a patient to manage themselves.

Physiotherapist FP08

However, what this physiotherapist described was not necessarily universal. The review process was increasingly being linked to timely discharge (as with Eastmere’s community matrons and all district nurse teams) while, at the same time, staff shortages impeded care reviews. Responsibility for review was also being traded, at times, between different parts of the health system (sometimes without financial compensation). District nurses, for example, were often performing medication reviews, diabetic reviews and other long-term reviews for housebound patients. In Fivecastles this was formally commissioned, but in Eastmere it was not funded. Sedgeton CHS staff asked GPs to make the 6-month checks that NICE recommended for patients with heart failure. Our informant was unsure if those checks actually happened.

In contrast, review was normal for reablement patients (commonly on 6-week care packages), who would typically have each visit written up, be monitored weekly and formally reviewed just after the mid-way point to establish whether or not they needed ongoing care.

Review was also integral to mental health care. The Eastmere community mental health service reported that they aimed for 3- or 6-monthly care plan reviews depending on status, but would review weekly during crises, and daily if patients’ medication or behaviour warranted it, updating ‘their risk assessment/care plan, every other day, every week, because it needs to be up to date’ (CPN EP13). They also reviewed dementia patients 6-monthly rather than yearly (as in NICE guidelines).

Service profile: shortage and specialisation

Patient interviewees reported instances of hurried home visits, often at unpredictable times, insufficient contact time with health workers, often-changing health workers and (for some) prompt cessation of CHS support. All of our study sites reported having insufficient CHS staff, especially specialised nurses, community matrons (or the equivalent) and OTs. Staff shortages reduced the scope for transferring patients even within a provider, and hence reduced flexible and longitudinal continuity of care. ‘Blocking’ not only of beds but of skilled workers (e.g. district nurses) occurred when a reduced staff complement stopped patients being transferred onwards. Delays in assessment, care and treatment were reported in Eastmere, particularly for physiotherapy. Nevertheless, some clinicians in Eastmere regarded large numbers of patient referrals to them as a sign of success, showing demand for their services. In Tarrow:

I think there’s a delay across most services to be honest in that everybody’s [community, secondary, social care] under the same pressures.

Nurse TP05

There, CHS were mostly limited to working hours of 5 days per week and lacked sufficiently qualified community matrons for the virtual wards:

I am the only person at the moment, for the matrons, who’s got health assessment and who can prescribe.

Nurse TP10

The CHS heart failure service at Sedgeton was two-thirds under strength, relying heavily on the health-care assistants as ‘sort of care co-ordinators’. Such shortages directly affected longitudinal continuity of care because ‘there’s a memory for the person who deals with that particular area of care that obviously improves outcomes for the patient’ (speech therapist SP01).

Occupational therapists and carers were in short supply in Fivecastles CHS. Whinshire CHS staff said that they had good specialist services (e.g. speech and language), but other services were stripped of staff (rehabilitation services had only 2 hours of podiatry input per week; OTs were scarce). There were insufficient services (e.g. night sitting) for patients entering residential care. Staff shortages leading to expedients such as ad-hoc staff redeployment may explain why patients reported discontinuity of the staff who made home visits, as may high numbers of part-time staff.

Staff shortages were more severe in mental health and social care. In Eastmere, this bottleneck stopped CHS transferring patients onwards. Redundancies and downgrading had made it difficult for the Whinshire mental health team for older adults even to deal with patient crises. Mental health services there faced a shortage of rehabilitation beds, there were no psychotherapy services in the county, and the well-being service was of high quality but had too little capacity, which became a problem when making a care plan for someone with acute psychological problems or anxieties. In Sedgeton, the shortages had a partly financial cause, but it was also difficult in any case to recruit specialised staff, and indeed community nurses generally, with the right experience. The CHS often backfilled vacancies, but that only depleted another team. In Fivecastles there were also reported shortages of welfare benefits advisers and befrienders (volunteers).

Workload pressures had two opposite effects on the specialisation of primary care labour. On the one hand, new specialisations appeared, such as primary care dementia practitioners (Eastmere). To work more flexibly (e.g. in acute care at home schemes) staff required higher skills and hence they cost more to employ, which had to be weighed against any savings resulting from higher productivity:

Up skill for flexibility – especially for rural areas, with some of the roles that we have, I’d like to tweak so that people could do more, especially our support workers, they can’t take blood and things like that.

Nurse TP10

Against this, financial pressures made CHS managers reluctant to set ‘precedents’ for additional or different services, and encouraged task delegation (e.g. in Eastmere, delegating 6-monthly reviews of dementia patients’ care to a health-care assistant). Such delegation could have unforeseen advantages in that more regular contact generated trust and helped reveal patients’ needs ‘once you get to speak to a carer there’s often other things going on’ (CPN EP14). Similarly in Whinshire CHS, the caseload was stratified from low- to high-complexity patients, with district nurse work increasingly concentrated on complex patients. In some district nurse teams, the health-care assistants scheduled the nurses’ caseloads. Sedgeton CHS employed ‘roving GPs’ who cared above all for (mostly elderly) patients recently discharged from hospital, where exacerbations would otherwise have resulted in readmission.

Co-ordination mechanisms

Line managers tended to allocate district nurses’ work in task-oriented ways that maximised output (‘efficiency’) rather than cross-sectional continuity. Partly they resorted to ad-hoc rostering in response to staffing pressures, but to varying extents at all study sites CHS had been centralised on a locality basis. Economies of scale were one motive, and the desire for ‘seamless’ CHS (enabling, say, district nurses to work more closely with community matrons) was another.

Even under single management, working practices were often limited to a single occupational ‘silo’, not standardised or co-ordinated organisation wide. Each CHS team in Sedgeton had different ways of selecting a patient’s key worker (after assessment or triage; whoever was available; by locality). All of this variation occurred within one manager’s span of control. Information management was often equally disparate. In Eastmere ‘we have cardiac, we have respiratory, all using different paperwork, diabetes, different paperwork, Parkinson’s’ (team facilitator EP07). Paper-based care plans in the home then became the only way of communicating between staff. The need to care for patients around the clock also increased discontinuity of care (e.g. in Fivecastles patients went to dressings clinics at the weekend instead of a district nurse visiting them).

In all study sites, providers used ‘horizontal’ MDTs to manage complex care. Nevertheless, the ‘silo’ mentality reported among professionals at all sites was an obstacle, commonly taking the forms of not trusting other professionals’ assessments and of non-communication. The acute care at home scheme in Eastmere lost early recruits because ‘they were A&E [accident and emergency] nurses and had it tattooed on them’ (team facilitator EP07). Another area had a history of non-cooperation between equipment and learning disability OTs. In Sedgeton:

The word ‘team’ has begun to mean people sitting in the same office, rather than virtual multidisciplinary teams which I think are actually the foundation of what most people do. Because you can’t all sit in one office [laughs] all of the time.

Speech therapist SP01

Then the word ‘team’ was little more than a euphemism for a hierarchical, uniprofessional department whose members tended not to see themselves as part of a larger care pathway. Consequently, one ‘team’ (e.g. heart failure) might have no direct access to services provided by another (e.g. physiotherapy, occupational therapy):

So unless a patient has COPD then the OT in the respiratory team is bound not to take those patients on and equally if the patient doesn’t have cancer, the therapists within the palliative care team can’t take them on.

Nurse SP04

Physiotherapists could refer to podiatrists but not vice versa, yet other ‘teams’ in the trust could readily transfer patients to one another.

Vertical integration

Second-responder providers substituted primary for hospital care through a combination of:

1. diverting potential admissions from the ED ‘front door’

2. expediting hospital discharge at the margin

3. hosting and/or staffing care network co-ordinating bodies (see Chapter 8).

Mental health services were vertically integrated, with consultants working in both community and inpatient mental health services. When a CPN thought that a patient needed consultant attention, the referral still had to be routed back via the GP, conforming to the consultation model of care co-ordination. However, once referred to a consultant, patients could readily be transferred between services within the trust, for instance when hospital-based services were transferred to community clinics, even to support groups.

Community health services with acute services

Our patient interviewees in Fivecastles described their transfers between CHS, acute hospital and the various day hospital services. The same geriatricians worked across the wards and day hospital. Patients could immediately ‘step up’ or ‘step down’ between community, intermediate and acute care without external referral. The day hospital gave ‘a sort of one-stop shop’ for GP referrals (70% of its caseload) and patients liked it. They could access a range of equipment in a safe environment, promoting confidence and socialisation, and follow a flexible treatment programme throughout the day rather than treatment at allotted times.

However, conventional line management did not reach very far into the day hospital. Its geriatricians were based in the two DGHs and:

[N]o one sort of takes charge of the day hospital [. . .] It’s very much team working and I think that’s one reason why it works, because there isn’t a manager to oversee everything, so everyone pulls their weight and works [. . .] We meet every morning to discuss patients and will set up a meeting if there are problems but [it’s] difficult to schedule team meetings with part-time working, rotas, patients in, etc.

Therapist FP14

Perhaps unwittingly, this regime resembled the Netherlands’ Buurtzorg system257 of giving community health workers high levels of autonomy and responsibility. Transfer-of-care co-ordinators worked in the ED assessment unit to divert patients and avoid admissions:

Somebody else had fallen, OK we’ll see to her laceration, we’ll now send her to an occupational therapist to see where do they need their grab rail, do they need their couch raising, you know things like this, put them in the Falls Prevention Programme.

Locality manager FP04

Even within the trust, lack of communication could sometimes inhibit continuity of care. CHS were not always informed when their patients returned to hospital, and the care of those patients who missed outpatient department appointments (even when this occurred as a result of sickness with which CHS were already involved) was referred back to the GP.

Because it was the same organisation, community and day hospital patients were able to return to the DGH for limited periods without needing another GP referral. As noted, CHS transfer-of-care co-ordinators provided an ED ‘front-door’ referral diversion service. Specialist nurses had become hospital based but ‘they’re still very accessible to us’ (locality manager F04) and would visit patients with the community nurses when required. The local hospice-at-home was officially an external provider, but:

We sort of classify them with us [combined acute and CHS trust] because I think there is some reciprocal funding and [. . .] they’re partly charity funded as well.

Locality manager F04

By contrast, the main community hospital for our Eastmere study general practices was not just managed by a separate trust from the CHS, but by a trust located in, and mainly serving, another CCG.

Community health services with mental health services

Community health and mental health services in Whinshire were organisationally integrated as one trust. This trust was rolling out a primary mental health model under which mental health services were seen as a normal part of primary care. Consequently, mental and other community health care was integrated through interdisciplinary teams. The community elderly care team included geriatricians, psycho-geriatricians, CPNs, OTs and nurses. The district nurse team lead’s role was also to better integrate physical with mental health care, especially for patients who were in and out of hospital with long-term conditions. In the absence of common information systems, this included working with a mental health counterpart to identify patients who were on both caseloads or could benefit from joint input. However, an additional referral from the GP was still required before the patient could access additional services within the same trust that the GP had already referred the patient to for other reasons. Representatives of the older adults’ mental health team were meant to attend weekly meetings at the mental health hospital, although pressure of work curtailed this to discharge planning meetings only. Until recently, social care staff were seconded into CHS teams, but financial austerities had scuppered this arrangement.

Nevertheless, the parallel, occupationally based ‘silos’ for CHS, mental health services and therapies provided by social services were resilient. The well-being team and psychological therapy teams ‘just work very differently [from us in CHS], even though we’re all in the same organisation’ (nurse WP03). The older adult mental health team still did not have non-mental health nurse members and had lost its social worker members, which made it harder to maintain longitudinal continuity between health and social care. Mental health services were vertically integrated. Day hospital mental health services had largely been closed in favour of cross-sectional groups, whose expansion was limited mainly by the trust’s resources.

Whinshire was unusual in having quite integrated medical records for frail elderly patients, with read–write access for other health and social care staff besides GPs. Nevertheless, data input remained laborious. One informant claimed that it could take up to half of the working day:

[I]f you’ve done it and haven’t recorded, you know, you haven’t done it, whether it’s in the patient’s own notes it has to be on [proprietary system name] (the spiros and the phone calls).

Nurse WP03

Mental health services still had a separate information system whose records were not directly accessible to other trust staff, but the trust was considering a combined data warehouse.

Community health services with acute care and social services

When social workers at another site (Tarrow) began working as part of the mental health trust team, the benefits of organisationally integrating health and social care were immediate: ‘[D]elayed transfers of care were eradicated within 6 weeks’, moving the trust from being ‘about the worst in the Strategic Health Authority to the best’ and this where there was ‘a high performing acute sector and [. . .] an underinvested in community sector’ (manager TP01). Expenditure had to be shifted so that there were no overall savings, ‘So in a place like [county], high performing and quite small, the benefits are clinical and quality, clinical quality, not economic’ (manager TP01). The subsequent experience of organisational disintegration revealed, with hindsight, how much easier organisational integration had made co-ordinating and maintaining longitudinal continuity of care:

[W]e learnt so much about each other, adult social care and health, because we were together for several years [. . .] We weren’t given the 5 years to really make it embed into practice [. . .] it’s such a great shame.

Nurse manager TP06

Towards the end of the study period, the integrated organisation in Tarrow was partly dismembered. Social services reverted to independence, but acute and CHS services remained organisationally integrated in one NHS trust. Our account of the organisational integration of health and social care, therefore, reports the integrated period, illuminated by contrasts with the subsequent separation.

This trust ran a pilot ‘virtual ward’, organised within the trust. (See Chapter 8, which describes a second, separate ‘virtual ward’ that GPs ran.) It was consultant run and staffed mainly by nurses and community matrons, ‘assertive, highly skilled nurses who are well versed at risk assessment’ (nurse manager TP06), who worked largely in the acute hospital; they identified, in the ED, the clinical assessment unit and ward rounds, patients who could be cared for at home. On discharge:

[T]he medical responsibility lies with the consultants in the hospital, not the GPs, so that’s been a real change for the consultants and these matrons have spent a great deal of time building the confidence of the consultants.

Nurse manager TP06

Patients, therefore, did not necessarily have to be medically stable to return home. A community sister took responsibility for the patient’s ongoing care (including a 24/7 response capacity). Both virtual wards had integrated electronic patient records with read–write access for all the relevant health professionals in the trust, increasing informational continuity.

The virtual ward scheme described above initially focused on expediting discharges and so reducing hospital average length of stay, rather than entirely substituting primary for inpatient care. The team of consultants, community matrons (who could prescribe), NPs and other nursing and support workers provided care in discharged patients’ homes. However, the virtual ward now also receives direct referrals from the ED, the clinical assessment unit and outpatient clinics, including podiatry and orthopaedics.

In this site the integration of CHS, social care and acute hospital care had little impact on primary care services themselves, but, as reported above, it did make it easier to access some of them.

Interorganisational information exchange

Referrals require information exchange between organisations. The exchange of health records between providers under different ownership was, in every study site, incomplete, even lacking, owing to:

1. Information governance rules preventing data exchange (e.g. for risk stratification) and sharing care plans, especially (but not only) between social services and CHS. In Eastmere we were told that the lack of a common care plan was because of legal requirements connected with individualised care budgets.

2. Variegated information systems, some ill designed for information sharing, even if that was wanted. Thus, nursing records and care plans might be held electronically (e.g. in Eastmere), but even when access had been negotiated data still had to be scanned or retyped into general practice systems. Because hospital, social service, mental health and third-sector systems rarely communicated:

We don’t have one record for one patient. One patient could have five different services calling in. All those records are archived somewhere different and I may not know what the cardiac nurse has said, you know. I may not know what the consultant has said, in [hospital], I may not know lots of things.

Nurse EP08

1. Non-automated systems requiring dual data entry. For example, Eastmere and Whinshire CHS staff made notes on paper at the client’s home and then re-entered them onto computers back at the office. (Sedgeton staff had no electronic systems and were, therefore, spared this task.) CHS staff often had to go to the general practice to read patients’ records there. There were separate (partly duplicated) financial, clinical and management systems.

2. Unreliable connectivity, especially in rural areas:

Some days it will work; other days it won’t.

Manager EP18

1. Clinical coding inconsistencies: different general practices coded similar activities differently. Hospital coding of mental health problems was erratic.

Colocation

To varying extents, all the study sites had replaced colocation at general practices with basing CHS staff at centralised locality offices. In Eastmere, Fivecastles and Tarrow, CHS teams that included district nurses, OTs, therapists, assistant practitioners and support workers were each linked to clusters of general practices. In contrast, organisations were colocating individual staff members. In Eastmere, for example, primary care dementia practitioners employed by the hospital trust were colocated in GPs’ surgeries, providing outreach to community hospitals and nursing homes. Elsewhere, CHS, mental health services and voluntary organisations were providing in-reach counselling, podiatry or care navigation services to GP surgeries, social service offices and district nursing teams. In one instance, a voluntary organisation provided a follow-up check for patients who had been discharged from hospital because:

people were coming to our attention 6 weeks after they were discharged from hospital with actually quite a lot of need that hadn’t been referred through to the hospital based services.

Manager WP08

Given time for trust and good working relationships to form, the benefits of colocation were described to us as being quick, easy access to recent information about patients; easy access to advice; joint problem-solving; shared, consistent decisions about patients; mitigating the problem of inaccessible patient records; and easier informal (‘permissive’) referrals. For co-located CHS and social care staff (Eastmere) ‘you can walk up and down, you can discuss easier face to face’ (manager EP06). The Fivecastles general practice noticed the value of colocation once it ceased:

[T]he district nurse would come in 11.30, 12.00 each morning when we finished our surgery so that we could liaise, you know, face to face a proper relationship [. . . Now] you have to fax some central place [. . .] So you send off this form into what feels like the ether and you haven’t got a clue really whether somebody’s going to get back.

GP FP03

The experience was similar in relation to colocating the memory service and intermediate care team:

[I]t makes for easy referrals doesn’t it? And you’re on hand and you can deal with things quite quickly.

Team facilitator TP04

Pharmacists were in some areas also colocated with general practices and undertook patient reviews.

Colocation could short-circuit the effects of separate record systems and circuitous (‘consultation-model’) referral routes, but it could not remove the underlying organisational blockages. Neither was colocation sufficient to stimulate communication between staff, even within the same organisation, if they were simply personally uncommunicative. At Sedgeton, some hospital outpatients, CHS and mental health services were colocated in a large, purpose-built building, which (some informants told us) made it easier for them to obtain information and advice and talk informally across disciplines rather than referring patients into ‘the ether’. However, others told us that all of this colocation did not seem to have made CHS teams especially communicative with each other, nor were they when different services operated out of, for instance, the same community clinics. Colocated staff would not necessarily communicate with each other unless they saw a practical reason. Although described as a ‘polyclinic’, the building was more a venue for large-scale hospital outreach than a polyclinic in the classic sense. It was treatment and clinic focused rather than focused on cross-sectional continuity of care.

Interorganisational care process

Accordingly, the care networks assembled care pathways for care groups, or even individual patients, across separate providers. These pathways were often time limited (e.g. to 6 weeks for most post-discharge patients). Networks established a single entry point for referrals, with swift, formalised assessment and triage for most patients and then referral to a ‘second-responder’ provider. Two networks (Sedgeton, Whinshire) used ‘front-door’ teams at hospitals to ‘pull out’ incoming patients whose admission could be avoided. (The CHS rather than an interorganisational care network did this in Fivecastles.)

Only certain patients had cross-organisational care plans (see Chapter 5), but it was usually those whose care was co-ordinated by a network. Whinshire, for example, had developed an integrated EHR to which some services, including hospital wards, had read–write access, although not all did. High-risk patients had personalised care plans that included advice about what to do should they suffer an exacerbation. In Tarrow, the virtual ward teams used such plans to review patients’ condition daily. A MDT reviewed ‘key’ patients in Whinshire to decide when the next review was due and who would update the patient’s care plan. One locality team also shared its social care plans with voluntary services. In Fivecastles, some general practices (although not our corporate example) had an electronic care record for patients approaching the end of their life and shared it with all urgent care providers. However, for most patients and health and social care providers this did not apply, causing ‘quite a lot of duplication, potentially going in for support of service users and not very, potentially, not very well co-ordinated in places as well’ (manager FP09).

Network co-ordinating body

Care network co-ordinating bodies went under various names: ‘referral hub’, ‘assessment service’, ‘emergency intervention service’ ‘proactive care’, ‘virtual ward’. In our study sites, either CHS hosted the co-ordinating body or the providers with the largest or most problematic interfaces jointly did so (Eastmere and Whinshire). In some sites the network co-ordinating body had its own budget or (in Eastmere) its member organisations supported it only in kind. Elsewhere there were pooled budgets for jointly managed services.

Care co-ordinators

Care networks had also assumed care co-ordination responsibilities to various degrees, sometimes co-ordinating their patients’ care themselves, but sometimes assigning it to an existing service. The networks usually co-ordinated care through a combination of case manager and interorganisational MDT.

Case management by non-doctors was available at all sites, typically for a limited time, to avoid hospital admission or expedite discharge. Within the limits of the information available, case management was often based, at least in part, on formalised risk assessment (but not always: GPs frequently felt that the available tools did not identify any patients of whom they were not already aware, and relied on their own clinical knowledge). Sites varied in terms of which patients were eligible for case management, although in no sites were all patients eligible. In Eastmere, the ‘grey area’ of patients who seemed still to lack a care co-ordinator caused staff a lot of stress. However, some patients had no designated care co-ordinator because (we were told) their needs might be too complex, high level and unstable; responsibility for care had to remain flexible.

Care co-ordination by case management was, in our study sites, a hybrid arrangement, transitional between the consultation model and care co-ordination by teams. It was consistent with the consultation model principle in that the GP might delegate tasks – here, care co-ordination – to others, and in that the new case manager replaced the GP at the hub of a ‘star-pattern’ configuration of referral links. However, in doing so, it also partly transferred the captant role. (In some places, although not our study sites, nurses have entirely replaced doctors in that role.) New or exacerbated health problems might still be referred back to the GP, but more selectively now:

[B]efore, they [patients] would have gone back to their GP for any health queries, the absolute expectation now is [. . .] they must first, in the initial instance, come to ICT [intermediate care team]. Because obviously we wouldn’t want to be escalating it to a GP when it should be dealt with in-house, and the majority of those would have never have needed GP view at all.

Manager WP16

This arrangement seemed most developed in Eastmere, and least so in the more silo-like Sedgeton CHS.

The care co-ordinator role might shift between case manager and multiagency team:

I would say yes, we’re the case managers, but if you feel like there’s an exacerbation, need more intensive support then that care will transfer to ICT [intermediate care team] and then, so there’s not a duplicate, and then they would maybe refer that patient back to me.

Nurse WP03

A patient in Sedgeton might first be assessed by, say, a neurology team, and then transferred to another team to assemble an overall package of care. Except for urgent referrals, which were allocated on a duty rota system, referrals to the Tarrow older persons’ community mental health team were triaged at a weekly team meeting, which included a consultant, and were allocated depending on area and type of need. The person who first assessed the patient would, if appropriate, become their care co-ordinator.

Although patients might have a named care co-ordinator, much of the actual co-ordination work was undertaken by interorganisational, multidisciplinary primary care teams. Referrals between CHS and social care were often decided by case conferences and reviews. The nomenclature (‘multiagency group’, etc.), membership, scale and scope of these teams varied. Some were organised by locality; others were CCG wide or on some other scale, or with different teams concurrently serving populations of different sizes (e.g. when small specialised teams could not be divided among localities). Often several teams were available with different specialist skills (e.g. mental health, stroke, frail elderly). Typically – but not always – teams met at general practices with at least one GP attending, although in one of our Tarrow study practices such meetings were (unusually) attended by virtual ward staff and several of the practice GPs:

[Y]ou’ll get whoever’s available on the day, so you may end up with three, you may end up with six, eight.

Nurse TP10

However, there were still communication gaps between the (other) district nurses and the community ward team, and so weekly meetings between the district nurse and the virtual ward’s community matron were also instituted:

[H]opefully it’s only about half an hour. And we go through anyone who’s new to the virtual ward, and anyone who’s been discharged and needs follow-up from district nursing.

Nurse TP10

Indeed, multiagency care networks undertook most of the MDT activities that occurred at the study general practices. Our study general practice in Fivecastles also continued to hold monthly ‘Gold Standard Meetings’ with palliative care nurses and district nurse case managers for case reviews (in Sedgeton these became part of the IPCT meetings).

Network resources

The resources available to each care network depended on the service configurations within the member-organisations (see Chapter 8), which organisations participated and how actively. Care networks in this study variously included general practices, CHS, mental health, acute trusts and social services. Some used community hospital beds for step-up/-down care, with consultant or GP medical input. Hospital doctors participated in some of the Sedgeton and Whinshire teams. They also varied in scale. For example, the ambulance service and a practice-based commissioning consortium covering about 90% of the county’s population participated in Whinshire, whereas the Tarrow virtual ward served only part of the county. Care networks often reached into the ED ‘front door’, sometimes into the hospital wards or meetings with consultants.

Eastmere GPs referred patients to the care network, but thereafter made little further input to it (similar to GP behaviour after referral to a specialist181). GPs were, at times, conservative about referring to new services, fearing that they would lose control over quality of care (Fivecastles and Eastmere). We were frequently told that some GPs and nurses found care networks easier to accept than others.

Social services participated actively in care networks in Eastmere, Whinshire and Fivecastles, but more tenuously in Tarrow (despite previous collaboration) and Sedgeton. Social care providers also participated as, for example, in Whinshire, where an arm’s-length trading company had assumed the county council’s reablement and care functions. Social services in Fivecastles had a programme co-ordinating services for providing social support for, for instance, bereaved or socially isolated people. Funded by the NHS, voluntary-sector organisations managed practice-based community workers in collaboration with a central intelligence hub. Whinshire had a range of preventative programmes driven by the voluntary sector, but across the study sites voluntary-sector involvement was often reported as being underdeveloped or undervalued, and initiative specific.

Mental health services participated in Tarrow, fragmentarily in Eastmere and Fivecastles, and not yet in Sedgeton. (A single trust provided both mental health services and CHS in Whinshire.) The well-being service in Sedgeton was a collaboration that included a GP-led social enterprise. Several informants involved in other teams in Sedgeton (and Fivecastles) noted, however, the neglect of mental health in their meetings, despite the significance of psychiatric problems in ‘frequent flyers’. All of the networks were also attempting to involve third-sector providers more in service delivery, although sometimes from a low base. Patient participation was ‘planned’ in Whinshire but not reported elsewhere.

The care networks obtained varying degrees of compliance with their preferred working practices. To varying extents, unclear and inconsistent appropriateness criteria, and different administrative procedures for referrals, were described in all our study networks. In Eastmere, some staff had the problem of ‘not knowing properly where we can, what we can access, when we can access it and how we do that’ (manager EP20).

Seeing the care network’s referral rules as a cause of delay, staff would work around them. Staff in the new acute care at home service, for example, encouraged direct referrals to themselves so that they could conduct a full assessment and refer on as necessary, rather than do so via the referral hub. Member organisations’ roles were sometimes confusing. For example, both CHS and the county council in Fivecastles ran something called a ‘falls service’. None of our informants was sure what the difference was or when to refer to each. Without clear referral or triage criteria, parts of CHS could (and, in Eastmere, did) become catch-all destinations for all patients except those who obviously needed specialised (e.g. mental health) care, thereby acquiring a large caseload.

Informal care networks continued to develop alongside the new official ones. The manager of the acute care at home team in Eastmere said of his staff, ‘everybody refers, should refer to everybody else in the community and that’s what makes the community’ (manager EP06).

Local knowledge facilitated what our informants called ‘permissive referrals’. The carers’ service in Eastmere, for example, informally referred carers to adult social care, to the community mental health team and to voluntary agencies because they had established good personal relations with those providers. Against this, voluntary-sector workers in Tarrow encountered both health and social worker resistance to their participation at first, although relationships gradually improved.

As noted, the study networks differed in which services were integrated into a single organisation and hence in which interorganisational boundaries remained:

• Eastmere, Sedgeton and Whinshire had an interorganisational boundary between CHS and acute services. Fivecastles and Tarrow did not.

• Fivecastles, Eastmere, Sedgeton and Tarrow had an interorganisational boundary between CHS and mental health services. Whinshire did not.

• Fivecastles, Eastmere, Sedgeton and Whinshire had an interorganisational boundary between CHS and social care. In Tarrow, this boundary, which had been absent for several years, was reinstated during the study period.

However, these different configurations of interorganisational boundaries made little difference to what care co-ordination mechanisms existed and how they were used across the remaining interfaces.

Creating care networks: obtaining provider participation

In all study sites, the PCT, or later CCG, had either established new care networks (Fivecastles and Tarrow) or adopted and formalised one initiated by another organisation (Eastmere, Sedgeton and Whinshire). The same applied to case management schemes. Table 3 outlines the resulting care networks. PCTs, and later CCGs, also played a part in organisationally integrating formerly separate providers.

Senior manager advocacy was a precondition for these activities. Whinshire PCT – where the most developed care network appeared – had had a Director of Integration who was a driving force and could

[S]ee the interests of all the partners and seek to identify solutions that yet didn’t run against the interests of all the partners

Voluntary organisation director WP07

and had some like-minded colleagues, as had the chief executive at Tarrow. The difficulty came in sustaining this support during the conversion of PCTs into CCGs, and the concomitant staff ‘churn’, and during the similar ‘churn’ at the end of time-limited special projects to promote integration. These individuals had moved on before relationships between the organisations that had collaborated became institutionalised. After the pilot collaboration with the county council ended in Whinshire, it became:

[I]mpossible to keep track of the different boards and subgroups [. . .] the same people are sitting in four different meetings [. . .] no one knows which meeting actually has the delegated authority to decide anything.

Social care manager WP17

Managers also used techniques such as gaining Pioneer Site status and facilitating access to knowledge and examples of good practice to legitimate interorganisational working. Performance targets had mixed effects on interorganisational collaboration:

It’s a system which is very much operating in flitting between ‘how are we going to solve it’ through to ‘who are we going to blame for it?’

Social care manager WP17

For instance, the care co-ordination hub in Eastmere was required to hold a case conference for each new patient within a week of referral, irrespective of whether the necessary participants (e.g. CPNs) did (or even could) participate. This tension between organisational targets and those of the care networks was also reported in Whinshire and Fivecastles, but not in Sedgeton or Tarrow.

Clinician advocates were equally important for introducing and promoting projects and maintaining close working relationships at senior medical management levels across organisations. A GP lead explained how he had ‘got the consultant on board and he very much has the same vision’ for frailty care teams ‘both within the hospital setting and the community’ (EP19). Our study practices either maintained frequent contact with CCG leads or had them as partners (and were, in that respect, atypical).

Providers that could not be integrated had, nevertheless, to be recruited to the care networks. Eastmere CCG’s experience with its mental health provider shows that this was not a foregone conclusion. All of our study site CCGs encouraged voluntary bodies that provide additional services and participate in care co-ordination work.

The CCGs also mandated, and in some instances designed, the cross-organisational care pathways that the care networks operated. Whinshire CCG constructed two reablement/recovery pathways for older people with mental health problems (e.g. dementia) to ‘step down’ into bedded care at either a NHS subacute service or a county council residential home with NHS in-reach mental health staff, followed in either case by a step down to domiciliary care. Sedgeton CCG’s frailty pathway was a common strategy for the acute trust, CHS and general practices. The CCG wanted GPs ‘to be much more responsible for a co-ordinated way for, particularly, people with complex needs’ (manager SP09).

Sedgeton CCG funded two ‘interface geriatricians’ to co-ordinate the acute hospital and community services. Eastmere CCG introduced a scheme for designating GP leads for residential and nursing homes.

Constructing a care pathway involved defining which patients were eligible for case management, corresponding referral criteria and procedures, professions’ and providers’ roles, and disseminating these to all of the providers and health workers involved. The example of Eastmere shows that achieving all of this was not a foregone conclusion, either. In Whinshire, the absence of shared key performance indicators across health and social care, and the different contractual arrangements, made joint triage sometimes problematic, and even a cause of disagreements.

Commissioning played little direct part in the care co-ordination mechanisms that we observed. At the time of this study CCGs did not commission most general practices, but all of the study site CCGs used discretionary directly enhanced services to pay general practices for additional work on frail elderly, especially dementia, patients; for instance, risk assessment of 5% of GPs’ patient lists in Fivecastles. Eastmere and Tarrow CCGs devolved the budget for this nursing home in-reach to those general practices that undertook it.

Paying hospitals and mental health services by tariff was not always helpful to interorganisational care co-ordination. For hospitals these tariff payments rewarded activity rather than its prevention, and were an incentive to retain caseload and services. 55 In Tarrow, we were told, tariff payments were a disincentive to accept probably high-cost patients within each tariff group. For our focal care group, ‘high cost’ did not necessarily reflect technology-intense treatment; only, perhaps, that transport would be costly because of where the patient lived. Similarly, for the more activity-based payments to CHS providers, a home visit by a mixed CHS team (e.g. nurse and therapist) still only counted as one visit when it came to paying the provider. Providers were at times unresponsive to contractual ‘fines’ for late discharges. In Fivecastles and Eastmere ‘a lot of them just took the hit and took a fine’ (GP EP19).

One activity that was not reported to us was the use of CCGs’ commissioning powers to equalise capacity across providers for our focal care group, so as to even out bottlenecks in the care pathways (e.g. to match hospital and CHS throughput). Providers seldom pooled budgets. Previous chapters have reported the difficulties of information sharing, which applied to managerial as well as clinical purposes. We found few systems, other than ‘planned’ patient representation in care networks, for managers (and then, not commissioners) to capture patient stories and experiences. That said, it should be remembered that the new commissioning system was still young. Joint Strategic Needs Assessments did not yet seem to have had much practical impact on the care networks.

Paywalls for social care, and the consequent assessment of patients’ needs and finances, were, as noted, widely reported bottlenecks and obstacles to maintaining flexible access and cross-sectional continuities of care. Personal budgets were blamed in Eastmere for social services producing their own care plans separately from NHS care plans. Commissioners wanted to find workarounds to ensure that day centres were not accidental victims of the personal budget system (as day centres were not easy to finance from personal budgets).

Information systems

Policy documents state CCGs’ governance role as including the promotion of IT systems to support current policy goals, including sharing managerial information across providers,22 which might reasonably be taken to include standardised, shared EHRs. Whinshire had fairly integrated medical records for frail elderly patients, with read–write access for GPs, community nurses, teams and therapists, other health and social care staff, and indeed for patients and carers. If patients consented, the information collected about them became available to hospital wards. That experience provides proof of concept that such systems are feasible in NHS settings. A conspicuous absence even in these systems, however, was of mechanisms for capturing evidence of patient experience for use in designing care co-ordination.

Managerial discretion at provider level

For care co-ordination to occur, it was necessary not only for providers to join the relevant care networks, but also, as a Service Delivery and Organisation (SDO) study on networks258 suggested, to implement internally the networks’ rules and working practices for care co-ordination. Some of our patient interviewees noticed (see Chapter 5) that it was one thing for one organisation to send information or requests to another, but another thing again for that information or request to reach the relevant frontline workers or the patient, or to be acted on.

For managers within care providers, care co-ordination posed various prioritising decisions and trade-offs:

1. Whether or not to target interventions at those most in need. In the rest of the county (although, curiously, not the part where our study general practices were located), Eastmere CHS concentrated its NPs on high-admitting nursing and residential homes only.

2. Whether or not to mandate attendance at interagency team meetings; in Eastmere and Fivecastles attendance was otherwise erratic.

3. Whether or not to standardise admission criteria. Different parts of the same organisation could react differently, even inconsistently (e.g. Whinshire county council therapy services had different criteria for accepting CHS and hospital referrals), in co-ordinating care with external providers.

4. Whether to trade off relational continuity of care against ‘efficient’ scheduling of staff caseloads by allowing staff time to travel to maintain continuity with patients who move home, or to maintain contact, as least for a while, when patients cross organisational boundaries.

5. How to trade off the use of less skilled and cheaper staff against the use of more flexible, costly, specialised and productive staff, especially in rural areas or for more complex care such as acute care at home, as discussed in comparisons of the relative costs of general medical and advanced nursing practitioners as case managers. 201

6. How to trade off larger provider scale and scope for cross-sectional and flexible continuity against a possible cost in relational continuity (although, as the corporate general practice we studied illustrates, even having a small-scale provider does not guarantee relational continuity).

7. How to trade off the presumed (although there is little published evidence) economies of case in centralising, for instance, community nursing services at local level against the informational, flexible and relational continuity advantages of colocating them at general practices.

Care co-ordination, both internally and externally, also at times confronted the obstacle of professional cultures focused on who ‘owned’ which part of which services, and they were distrustful of assessments and judgements made by members of other professions. Against this, we were also told that practitioners appreciated being appreciated, not least by those in other professions:

What’s nice as well is you’re valued, is like sometimes the GP will ring you up and they ask for you, you know, and it’s one of your patients and the GP will ring up and say, ‘[name], I’ve tried this, tried that, what do you think?’

CPN EP13

Insofar as managers can manage organisational culture and climate (a moot point), such comments suggest that provider managers faced decisions about managing organisational cultures in favour of interprofessional and interorganisational collaboration. For instance, our informants valued the often small meetings that brought bits of services together and enabled them to get to know, through personal contact, who was who in the local health system and what was happening in everyday service provision.

Integration and disintegration

In Whinshire, the joint management of health and social care services had run its allotted time as a pilot project, but left some enduring legacies (e.g. the integrated information system, multiagency teams) that continued to evolve, indeed to expand to other care groups, after the project formally finished. There was, however, dissatisfaction within social care in parts of Whinshire.

Joint management of primary, acute care and county council social care had been reversed in Tarrow, with adult social care returning to county council management. We were told that the reasons were:

1. Financial: central demands for 30% cost cuts over 5 years, compounded by lack of robust financial models at the outset.

2. A protectionist coalition: the trust was not culturally integrated, but more a coalition with some protectionist members. A social care view was that money was spent on social care not to achieve social care goals, but ‘short-term responses to the tier four emergency issues that [trust name] were having’ (social care manager TP09) prompting social care to leave. (The same tension was reported in Whinshire.) Some other leading members were preoccupied with preserving the DGH so that the trust never ‘actually broke away from the currency of the health service being dominated by the acute sector’ (NHS trust manager TP01).

3. Ownership and vision: there was turnover among senior managers and non-executive board members compounded, in a site that always had difficulty recruiting (and this was not the only one in our study), by recruiting staff who instead of the integrated trust vision had more traditional, even divisive, attitudes: ‘what I saw was a very traditional and very unhelpful social worker versus GP clash’ (NHS trust manager TP01) and a board that, some suggested, had little experience of or support for the public sector.

4. Role and adequacy of CHS and primary care: the trust was not an implementation of, say, the Kaiser Permanente model, but based on ‘a reactive, a traditional reactive primary care, not a proactive integrated system identifying patients who were at risk of admissions, intervening early, stabilising, allowing patients to continue to be independent in their own homes’ (NHS trust manager TP01). ‘Traditional’ appeared to mean some GPs thinking ‘If in doubt, I’m going to use the hospital’ (NHS trust manager TP01) and (as others also thought) ‘the desire to preserve professional independence and economic independence’ (NHS trust manager TP01). A fundamental misalignment of incentives, structures, institutions and interests limited GP sign-up beyond the leaders (indeed, limited GP engagement with the CCG more generally): ‘Almost irrespective of the financial situation, there would have come a point at which questions were raised about primary care and the interface between the new model of care we were running in [trust] and how that fits in with general practice’ (NHS trust manager TP01).

5. Timing: before the banking crisis struck there was too little time to properly ‘develop the relationships [. . . and] develop the mechanisms [. . .] in terms of getting a more integrated approach it is going to take us a while’ (social care manager TP09).

Certain NHS-wide governance structures also both enabled and constrained care co-ordination. The relevant higher-level governance structures were:

1. the consultation model, which is heavily institutionalised within the NHS

2. information governance regulations and law

3. paywall barriers to accessing services

4. the fact that after the 6-week time limit patients had to be financially assessed before receiving ongoing social care

5. institutionalised ‘cultural’ differences between professions

6. not least, the tripartite structure with its radical separation of primary medical care from CHS, and of CHS from social care.

Care co-ordinator

Having chosen a PHCC, patients are assigned a primary care doctor, whom they can change. In practice, they may have to wait to see their chosen doctor, so many simply see the first doctor who is available. At Kista, patients also had an assigned lead nurse.

The consultation model of care co-ordination predominates. Typically, the doctor will transfer the patient to the polyclinic-employed or -based nurse. Referrals to a physiotherapist, a speech therapist, an OT or diagnostic service must be made by a doctor if that consultation is to be publicly funded. Polyclinic-based nurses cannot refer patients directly to a specialist or therapist, but must either ask the doctor to do so or ask the patient to ask the doctor. Most PHCCs say that one of their nurses ‘co-ordinates care’, especially for complex cases, but the amount of co-ordination done by the nurse varies between centres and none operates a formal care manager model. As noted, nurses cannot refer directly to specialists or other professions. Our survey254 suggested that case management schemes were not widely used. Only one-quarter of PHCCs responded using case management for diabetes, and much lower proportions for other chronic diseases. Where case management is used, it is generally underdeveloped, underspecified and subordinated to the consultation model of care co-ordination. No one else provides such a service instead, in which case this low use is significant for generating high costs and reducing quality of care, especially for heart failure patients.

There is a more systematic care co-ordinator role in some mental health services. A psychiatrist or nurse ‘case co-ordinator’ will invite others to a needs assessment and care-planning meeting. In one area in Stockholm (Sodertälje), a psychiatric care co-ordinator and a municipality case manager will work together, each co-ordinating other personnel from their respective health or municipality services. When these co-ordination responsibilities are defined, the specification is of the very general job; authority and accountability are not sharply defined. Again, neither can refer directly to other specialists.

Care plans

All patients have a doctor and nurse care plan, but there is variation between patients and between professionals in how much the patient is involved and decides on the plan contents or services to be used. The doctor and nurse usually consult patients about whether they want or need different services, and patients can ask for a service professional to be changed. Beyond this, patients have little choice as to when the services are provided or other details (e.g. when home care or meals are delivered).

Services and specialisation

At Lisebergs Vårdcentral, five primary care doctors, two trainees in general medical practice, six district nurses and two speech therapists provided services. A neurologist routinely provided outreach sessions. Djursholms Läkarmottagning had six doctors and four district nurses but no specialist outreach clinics. No private practitioner services or social service staff were based there. Kista had nine family doctors, five ‘residents’ (consultant equivalent), two trainee doctors, an outreach geriatrician clinic, seven district nurses, six unqualified staff and extensive physiotherapy services. A privately owned diagnostic laboratory was colocated there.

Vårdcentral nurses generally provide both ‘basic’ and ‘advanced’ (e.g. palliative) home nursing services and services at the centre itself. They visit patients at home within the defined area, and outside the area if not too distant, provided that the nurse deems the patient to be a ‘home care patient’; that is, somebody who needs nursing services more than once every second week. If they need nursing services less frequently, Liseberg Vårdcentral will not provide them, but this criterion was decided by Liseberg Vårdcentral itself and not by the county council, and no other vårdcentral adheres to it. In general, the nurses at each polyclinic can set their own local criteria for prioritising and scheduling home visits. Officially, vårdcentral nursing services are provided equally to all registered patients, whether local or distant, but in practice nurses appeared not to visit the more distant patients outside their area. Details vary, but typically a nurse will work a week at the vårdcentral clinic itself and then make home visits for the next week. It was not clear to us whether or not nurses also made home visits to patients who were inside their area but registered with other polyclinics. Vårdcentrals also undertake population-oriented public health activities such as maternity and well-baby clinics, vaccinations and care in the community.

Therapists (physiotherapists, OTs, speech therapists, clinical psychologists) were based at certain polyclinics, but were managed by therapy managers rather than the polyclinic manager and provided services to a wide area, although some did also provide regular sessions at the PHCC where they were based. Some (but not all) PHCCs shared a building with a laboratory and specialist outpatient departments. There is no separate out-of-hours service like in the NHS. ED visiting rates are high.

Co-ordination mechanisms within a single organisation

Vårdcentral PHCCs have a hierarchical organisational structure in which the doctors, nurses and (in some) therapists employed are managed in one unit and subdivided into MDTs. All personnel, doctors included, are employed by the county council, which also owns and manages the building. Each vårdcentral has for its manager a clinical director (usually, but not always, a doctor), who has limited discretion over expenditure within defined budget headings. This clinical director reports to a co-ordinating director, who in turn reports to the director of the Stockholm county division of primary health care and mental health.

The vårdcentral director ‘manages’ the other doctors in it, as also occurs in the private PHCCs. The key issue, however, is what ‘manage’ and ‘line manage’ mean, particularly regarding clinicians’ scope for clinical autonomy, which is pre-defined (constrained) by various policies, procedures and clinical guidelines. The vårdcentral director has ‘management autonomy’; that is, authority and scope to decide management issues. In practice, management autonomy is fairly limited. The doctors have medical (clinical practice) autonomy in the sense of latitude to make treatment and clinical decisions by themselves without checking with anyone else, but only within the aforementioned policies and procedures. Managers’ decisions can be reviewed and over-ridden by a higher authority, but doctors’ decisions taken in exercise of their clinical autonomy rarely are. Notwithstanding the salaried status of (most) Swedish primary care doctors and the independent status of (most) English GPs, the clinical autonomy of Swedish polyclinic doctors remains less constrained than in the UK.

As noted above, MDTs are usually formed among the staff working (and mostly, line managed) within each polyclinic. However, there is no one standard model, and generally there are informal working arrangements between personnel, apart from the formal referral arrangements noted above and the employment role descriptions that specify the work of employed staff.

All authorised employed health personnel (licensed professionals) at the polyclinic have access to some common patient data stored by the vårdcentral, but not to all data held in the patients’ EHR. Polyclinic doctors have access to a patient’s EHR and electronic medical record (EMR, a subset of the EHR) showing the patient’s history, tests, diagnosis, treatments and referrals, and to the polyclinic’s nursing record containing the patient’s nursing needs assessment and care plan. Parts of the patient’s record (e.g. the care plan) can be shared with social services and private providers, but only with the patient’s consent. Most polyclinic doctors do not have access to the patient’s hospital EHR or EMR, but do have access to tests, diagnostics (e.g. radiology) and discharge reports.

Privacy laws forbid patient data being communicated between municipality and health-care IT systems, although this position is currently changing. In the meantime, there is no common case record. Neither may patient data be shared between the polyclinics and, for instance, private nursing homes or private home-care services, although in these cases it is also because the record systems technically cannot exchange data. At present, patient information is also not fully shared between the polyclinic, the hospital and other authorised health-care providers. (There is no shared EHR.) In Stockholm, and most other regions, one health IT system that does enable co-ordination and collaboration between patients and their care providers is Mina vårdkontakter 264, an encrypted service for patients to access and help co-ordinate their own care. This enables patients to contact services electronically; add notes about scheduled doctor visits, immunisations and more in a calendar; and get text message or e-mail reminders on the day before a visit. Patients can access their personal health record and, in some cases, contribute to it. Nevertheless, at present the system is too complicated for many patients to use.

Vertical co-ordination

Formally, patients can also choose their hospital. The patient choice policy is being extended to elective surgery, allergy testing and other specialist services. Patients usually choose on the basis of distance and travel times (which may limit their choice), primary health-care doctors’ recommendations and those of friends. As in England, patients have no direct access to specialists, who are mostly employed by the public hospitals. All PHCCs are gatekeepers to secondary care. In most polyclinics, formal systems for all the above are less standardised and developed than in the NHS, apart from agreements (clinical guidelines) about when it is appropriate for primary care doctors to refer patients to specialists. Most hospitals send discharge notifications electronically to the polyclinic and the referring doctor, and notify municipality social services. If the patient is ‘medically fit for discharge’ but cannot be discharged because the municipality has not made arrangements, the municipality is charged for each day the patient remains in hospital. However, this ‘fine’ is sometimes lower than the cost to the municipality of arranging nursing home or home care. Nårsjukvard health centres are larger, and are also publicly funded and managed. They are run by local hospitals typically offering planned day surgery, other day treatments and limited inpatient care besides the primary care described above.

Östra Vårdcentralen

This service is owned and run by a private for-profit company, a partnership of eight GPs. The PHCC clinical director (one of the doctors) is accountable for centre services and budgets to a board of directors, that is, the other seven doctors. The doctors’ company employs the nurses, who provide home-care services to registered patients and to a home for disabled people. The doctors rent the PHCC building from SLL. In it are colocated other council-run services: a laboratory, a hospital, psychiatric services, a hospital outpatient gynaecological clinic (mottagning) and a geriatric clinic. A radiography service, orthopaedic service and private pharmacy are also colocated there, but there are no social services, hospital specialist clinics or private practitioners.

The doctors buy managerial support services from a larger partnership, Praktikertjänst AB, of which ‘Our 2000 shareholders are dentists, dental technicians, dental hygienists, physicians, physical therapists/physiotherapists, psychoanalyst/therapists, psychologists, nurses, midwives, speech therapists, occupational therapists, social workers, chiropractors, nutritionists and other health care practitioners who are themselves responsible [i.e. professionally registered] staff and working in the company.’264 To these shareholders, Praktikertjänst distributes profits. It employs another 6780 people (in 2014) and provides private dentistry (circa 30% market share) and primary medical care (circa 20% market share) besides publicly funded primary health care. It centralises managerial support services in pursuit of economies of scale, but provides primary care through numerous ‘small-scale’ clinics (which tend, however, to be larger than most English general practices).

Stockholms Sjukhem

This PHCC is owned and run by the not-for-profit private foundation Stockholms Sjukhem. A doctor is clinical head of the unit, accountable to the foundation’s director. Based in the grounds of a hospital, a PHCC with two doctors and one district nurse serves a list of registered patients. No social services, hospital specialist clinics or private practitioner services are colocated there. Stockholms Sjukhem has a contract with the county council to provide the primary care doctors, and employs the doctors and the nurse. The wider foundation within which the centre is incorporated also provides district nursing services (‘Advanced Care in the Home’) and a rehabilitation centre, nursing home and memory clinic (which also provides home visits by OTs). Rehabilitation services are accessible without referral for primary care patients (but a hospital doctor’s referral is required for hospital dischargees). Palliative care is accessed by referral; other nursing home care is not.

Solna Vårdcentral

This PHCC is one of nine in Stockholm run by Capio AB, a corporation leasing the building from a public construction company. It contracts to SLL to provide primary care (specified by the ‘rule book’ – ‘what we should do, and how this is checked’) between 08.00 and 22.00 on weekdays, with evening cover at the nearby SLL Solna Vårdcentral. The centre’s income depends on the numbers of patients registered and patient visits, but the per-visit payment decreases once the average number of visits per patient per year exceeds 1.9. Capio’s district nurses provide home nursing to home-care patients. Assistant nurses provide home-care services (hemtjanst). Social care assessors (biståndsbedömaren) arrange separate home services and personal care services from the county or municipality.

To co-ordinate care for some patients, the PHCC district nurses needed to co-operate with the social care assessors. Both parties considered themselves care co-ordinators, but there are no formal systems to ensure co-ordination, which does not always occur but is more likely to do so if the following apply:

1. The different services cover the same patient population or nearly so (often similar geographical areas). Then, the Capio nurses are likely to be repeatedly in contact with others caring for the same patients and to care for other patients that they have in common.

2. Continuity of personnel.

3. Good personal relations between staff.

4. Staff are not overworked and hence do not tend to pass tasks on to others when the option arises.

5. There are stable systems (e.g. telephone numbers do not change) for faxing, telephoning and exchanging information. In acquiring another firm, Capio also acquired its electronic ‘deviation’ (adverse event reporting) system, which was initially not interoperable with Capio’s own.

Most interviewees thought that a well-defined care co-ordinator role would improve care for patients who needed multiple services and would avoid conflict and duplication or missed services, but that it would require funding.

Organisational characteristics

In Stockholm, where a higher proportion of PHCCs are privately owned and/or managed, nearly 20% more patients were registered for each centre (10,400) than for the Sweden sample (8500). The mean number of doctors was 30% more and nurses nearly 50% more per polyclinic.

Fewer Stockholm heads of polyclinics than those in the national sample reported experiencing a favourable financial impact from investments to improve quality of care for four chronic diseases (coronary heart failure, depression, diabetes and asthma). Thirty per cent reported a positive financial impact for investments in raising the quality of diabetes care compared with 34% for the national sample, and 4% for heart failure improvements compared with 13% nationally. Equally, however, Stockholm centres were less likely to report negative financial impacts (2% Stockholm vs. 4% Sweden). In both Stockholm and elsewhere, 60% and 64%, respectively, of PHCC heads reported that their centre had undertaken no financial investment in quality improvement for heart failure. Significant differences between centres in the Stockholm sample and those nationally were greater in respect of information systems:

1. 9% more Stockholm PHCC heads said that they made reports to another organisation about their adoption or use of IT (36% vs. 27% for all of Sweden)

2. 23% fewer in Stockholm said that they had a patient discharge summary available (70% vs. 93%)

3. 27% more in Stockholm said that they had laboratory results available electronically (94% vs. 67%)

4. 10% more had warnings of potential drug interactions (80% vs. 70%)

5. 8% more had reminders at the point of care (46% vs. 38%)

6. 4% fewer had some e-mail contact with patients (47% vs. 51%).

In Stockholm, where a higher proportion of polyclinics were privately owned than elsewhere:

1. Fewer sent patients reminders for preventative or follow-up care. Fourteen per cent fewer did this for diabetes than in Sweden as a whole (25% of PHCCs vs. 39%), 12% fewer for coronary heart failure (27% vs. 39%), 9% fewer for depression (22% vs. 31%) and 20% fewer for asthma (25% vs. 45%).

2. Fewer non-medical services were used, especially for coronary heart failure, which was 19% less than the national average (14% vs. 33%). For diabetes it was 5% less, for depression it was 11% less and for asthma it was 14% less.

3. Less use was made of case management for coronary heart failure: only 4% of PHCCs in Stockholm reported doing so, compared with 8% nationally. Otherwise there were few significant differences between Stockholm and the national pattern. Nationally, only 26% of sites used this proven practice for diabetic patients, 17% used it for asthma and none used it for depression.

4. More doctors received guideline-based reminders about which services patients should receive. According to patients’ EMRs, such reminders were used in 7–11% of Stockholm PHCCs, compared with 5% across Sweden generally. Again, however, even the top figures were low.

5. More PHCCs had lists of patients with the four diseases (coronary heart failure, depression, diabetes and asthma) than in the national sample. The largest significant difference was that 12% more Stockholm centres had lists of patients with coronary heart failure (68% vs. 56% for all of Sweden).

6. More PHCC heads reported giving feedback to their physicians about their quality of care: 14% more used this for quality of care indicators for coronary heart failure patients (38% vs. 24% for all of Sweden) and 11% more used it for diabetic patient quality indicators (60% vs. 49%).

The above patterns appear consistent with the hypothesis that where there is a higher proportion of privately owned and/or managed polyclinics, informational continuity within the polyclinic is higher than elsewhere, but external informational continuity and case management (longitudinal and flexible continuity) are lower.

Research question 1: integration versus network

Research question 2: ownership

The corporate general practice in England had low relational continuity of care despite being a small practice (which ordinarily may have greater relational continuity75,269), because its doctors were part-timers with other activities elsewhere. Although this practice was more willing than others to refer its patients to self- (or insurance-) paid private care, its patients had no easier internal access to other services run by the same firm; referrals were required just as if the doctors worked for another organisation. In contrast, the corporately run Swedish PHCC did give its patients access to the CHS that the firm also owned, which were organisationally integrated with its primary medical care service. Its English counterpart appeared to make less use of external CHS and, therefore, of any case management or other services that they supplied. It had been financially unable to provide personalised care in-reach to nursing homes. Our English and Swedish data were both consistent with the hypothesis that privately owned and/or managed primary care providers tend to have greater informational continuity of care internally, but less externally (to patients and hospitals).

The professional partnerships in England were distinctive in giving their partner-doctors flexibility to develop their own interests or specialisations and latitude to refuse projects, and in their generally smaller scale and scope than the Swedish PHCCs (but not smaller than the corporate English general practice).

Both of the Swedish general practices run by social enterprises (one charity, one health worker owned) gave ready access to other services run by different branches of the same organisation. In Sweden, the integrated, publicly owned PHCCs also gave patients direct access to other services run by the organisation (with external referral), and typically provided a wider range of services than the English general practices, however owned. Except for the clinic director, their doctors had no equivalent to the practice management workload of English partner-GPs.

In summary, ownership differences in this sample of providers affected the range of services to which patients had direct access, primary care doctors’ managerial responsibilities (relevant to care co-ordination because of its impact on GP workload) and the scope for medical innovation.

Research question 3: managerial discretion and performance

We found little difference between integrated organisations and care networks in terms of managerial discretion and performance. A larger difference was between NHS and Swedish primary care organisations. These differences reflected varying assumptions about the demarcations between managerial and medical jurisdiction, and the transparency of medical practice to managerial scrutiny. In these respects, the Swedish PHCCs resembled NHS practice some years earlier.

Senior manager advocacy was a precondition for activities both to establish care networks and to amalgamate the managerial structures of previously separate services (Whinshire and Tarrow). Clinician advocates were equally important for introducing and promoting projects and maintaining close working relationships at senior medical management levels across organisations. 270

As more concrete managerial practices that promoted care co-ordination or substitution of primary for secondary care, patients and health workers mentioned:

1. Patients seeing the same GP or other clinician as far as possible, for relational and cross-sectional continuity, for example by deploying a patient’s ‘normal’ health worker for home visits even if another worker takes the initial request.

2. When (1) is not feasible, team members cross-briefing each other about the patient (e.g. by ‘live handover’).

3. Ensuring that patients know what to do and whose help to seek in a crisis or exacerbation.

4. Overall care plans, from which individual providers or unidisciplinary plans follow.

5. Ensuring that hospitals can effectually refer patients directly to CHS.

6. Informing patients that their care has been reviewed (if the patient was not present).

7. Whitelists of vulnerable patients who are prioritised for access to professionals (for relational and cross-sectional continuity) and for care plans.

8. Regular audit and follow-up reviews of recently discharged and other vulnerable patients.

9. Opening health records and care plans, as far as possible, to other providers caring for the same patients.

10. Interoperable intermediate care team (e.g. conforming to open standards for data sharing).

11. Hosting or outposting services, or seconding staff, to achieve colocation.

12. Allowing time and stability for working relationships to develop across organisations.

13. Mandating staff participation in interagency and interdisciplinary care teams when necessary.

14. Education, training and culture management to breach interprofessional ‘silos’ and promote interorganisational working.

We accept that in times of austerity and high GP workloads some of the above may seem counsels of perfection. To that list, therefore, we add reducing GPs’ non-clinical workload by delegation to other clinicians or managers, which the feminisation of, and the looming recruitment and retirement ‘crises’ in, the GP workforce will also necessitate.

Research question 4: predicted differences between integration and networks

Wider implications of this research for organisational theory and other disciplines will be addressed in future peer-reviewed publications arising from this project. Here, we focus on the main hypothesis arising from our overview (see Chapter 2) of earlier research: an integrated organisation containing a wide range of services (above all, primary medical care) is more likely to favour the development of care co-ordination, and, therefore, continuities of care, than co-ordination by care network.

Some detailed findings that tend, on balance, to support this hypothesis are outlined above. Care co-ordination in the NHS relies heavily on care networks because the tripartite NHS structure (see Chapter 1) is a radically fractured form of consultation model. Because of the small scale and scope of general practice services, patients who require round-the-clock or multiple services have immediately to be referred from their general practice to other providers. Care network co-ordinating bodies exist partly as a workaround for this problem. It might, therefore, appear that a single-provider model of primary care would improve care co-ordination, compared with care network structures. ‘Single provider’ connotes one organisation (not a network) providing primary care medical, nursing and other clinical services (therapies, mental health, pharmacy, diagnostics, etc.). Its internal structures would include interdisciplinary teams based on care groups or care plans, and hence relying heavily on ‘matrix’, ‘organic’ or ‘project’ methods of management, with highly standardised and (so far as is consistent with patient confidentiality) open information and communication systems. It would depend much less on multiagency networks than do current general practices. That raises the question of who might own and operate such a provider. This study describes the following possibilities:

1. A publicly owned polyclinic on, for instance, the Swedish PHCC model. The corporate general practices and Swedish PHCCs gave further proof of concept (if it were still needed) of the feasibility of mainly salaried general practice in NHS-like settings.

2. A corporate primary care provider.

3. A polyclinic operated by a co-operative, clinician-owned or other ‘third-sector’ organisation.

4. A professional partnership of larger scale and, above all, scope than is now usual in the NHS.

Any of the above could be managed, and where applicable owned, by doctors, nurses, other clinicians or a mixture. Further options exist. 271

Our empirical findings also suggest that structure is not everything when it comes to improving care co-ordination. 208 Personality (patients’ and clinicians’), patients’ resources, rurality, organisational size and managerial performance and discretion also have impacts.

Research question 5: implications for managerial practice

We present these implications by domain of practice.

Case management

For frail elderly patients with complex chronic health problems:

• Who coordinates their care? [Case Manager? GP? Someone else (who?)?]

• Are there designated case managers? What is their role?

• Or is there an informal system of case management?

• For such a patient, how often does the case manager change?

• Why does the case manager change?

• Do patients have written care plans? (Could we see an anonymised example or a blank pro-forma?) How are these integrated with the records?

• When do case-managed patients have their needs reviewed?

  • Periodically?

    • With formal risk management?

  • In response to health events?

[PROMPTS]

1. exacerbation

2. progress of disease

3. new health problem

4. major life event (e.g. loss of spouse or other carer)?

   • At patient’s request?

   • At clinician’s discretion?

   • On other occasions? (Which?)

External referral

• What are the main transfers or referrals to external organisations or other sources of care?

  • Are any of these colocated with you (i.e. based in the same clinic or workplace but employed by different organisations?)

    • IF YES: what are the practical benefits/disbenefits of colocation?

• What factors help in ensuring that patients referred to these organisations receive prompt service?

• Do you face any incentives – or disincentives – to refer to particular services? What are these dis/incentives?

• What are the main causes of any delays or restrictions to the referred patient(s) actually receiving the service?

• Are any services only available externally, although it would be possible to provide them ‘in-house’?

• Are there any services to which you would like to be able to refer frail elderly patients with complex chronic health problems, but cannot?

  • IF SO: why can you not make these referrals?

[PROMPTS] Because these services are:

1. Not available locally (at all, or enough)?

2. Available, but not budgeted/contracted for (‘rationed’)?

3. Administrative preconditions (authorisations, documentation, etc.)?

4. Evidence-based medicine-based limitations?

• Are there any other problems with using external services?

[PROMPTS]

1. Transfer of records and data between workers and organisations?

2. Access to information about what care the patient received?

3. Non-organisational factors (e.g. practitioners’ local knowledge, personal skills)?

• Specifically, are there any problems with [service(s) reported as problematic by documents or patient informants]?

Other aspects

Finally, is there any other important aspect of the integration and continuity of the primary care services in [organisation name] which we have not yet asked about?

Your health

We are interested in your recent experience of healthcare locally, particularly how the different people and organisations responsible for your care work together.

I’d like to start by asking you a few questions about your health in the last 12 months.

1. Can you tell me what health problems and/or disabilities you have at the moment, or in the last 12 months?

[Prompt Card 1] Enter on Table

1. Have there been any major changes in your health/personal circumstances (e.g. loss of spouse or other carer) in this time?

Thinking about everyday activities:

1. Have you had any difficulties in the last 12 months doing any of these because of your health?

[Prompt Card 2]

1. Do you have any support from social services/care agency etc. because of your health/to help you cope at home?

2. Over the past 12 months how would you say your health has been. . .?

Excellent/Very Good/Okay/Not so good/Poor

Your health care

1. Can you talk me through how your health has been over the last 12 months?

Use Table to record health episodes. Chart how the episode started (e.g. patient contacted GP), who they saw, where they saw them, who they were referred onto and how this overlapped with other health problems and interventions. (Cards 3–5 contain codes.)

As a result of discussion note:

Whether this resulted in any changes to their care or treatment

Enter on Table

1. Have there been any treatments, after-care/follow-ups that you thought you should have received/would like to receive but haven’t? YES/NO

IF ‘YES’: enter unmet need details on Table

1. How important is it for the health service to find out how you are after any change in treatment or hospital care?

Very important/Important/Not very important/Not important at all.

IF care over the last year includes time as hospital inpatient/outpatient.

1. How far do you agree with the following statement?

After treatment in hospital, a health professional has always checked all is well once I am back home

Strongly Agree/Agree/Disagree/Strongly Disagree.

1. Are you satisfied with the care you received in the last 12 months? YES/NO.

Ask what aspects of care have produced satisfaction/dissatisfaction.

Record on Table

1. Overall, do you find it easy to see someone about your health care? YES/NO.

Discuss why:

[Free text].

Organisation of health care: care manager

1. Do you think it is important that you have someone who is responsible for co-ordinating your care?

2. Who do you think has the best overview of your care at the moment?

   1. PROMPT: do you have a designated care manager/is more than one person responsible/what about the quality of this relationship?

   2. Has this person changed in the last six months? YES/NO

   3. IF YES: why was that?

   4. [Free text].

3. How important do you think it is that your medical team appears interested in your health, explains treatment clearly and is open to discussion?

Very important/Important/Not very important/Not important at all

1. Can I ask you how far you agree with the following statements?

   1. My doctor/care manager always has time for me and treats me as an individual.

Strongly Agree/Agree/Disagree/Strongly Disagree.

1. It is very important to me that I see my regular doctor/care manager.

Strongly Agree/Agree/Disagree/Strongly Disagree.

Prompt card 1: list of health issues

Prompt card 2: daily living

1. Bathing or showering.

2. Dressing (including putting on shoes and socks).

3. Cooking and eating.

4. Managing medications.

5. Making telephone calls.

6. Driving.

7. Walking outside your home.

Prompt card 3: how is treatment initiated?

1. Patient.

2. Family member, friend or neighbour.

3. GP.

4. Out-of-hours medical service.

5. Ambulance.

6. Hospital: emergency admission.

7. Hospital: elective admission.

8. Social worker.

9. Other (who?).

Prompt card 4: which professionals did you see?

1. GP.

2. Practice nurse.

3. Community nurse.

4. Community mental health worker.

5. Physiotherapist.

6. Occupational therapist.

7. Speech therapist.

8. Podiatrist.

9. Dentist.

10. Optician.

11. Ambulance crew.

12. Chemist.

13. ‘Alternative’ therapist or practitioner.

14. Social worker.

15. Home help/carer.

16. Voluntary sector (e.g. support group).

17. Other (e.g. nutritionist, counsellor?).

Prompt card 5: where did you receive health care?

1. At home.

2. At GP (or equivalent) clinic.

3. At another primary care clinic.

4. A&E.

5. Outpatient at hospital.

6. Inpatient at hospital.

7. Community hospital, nursing home or similar (intermediate care).

8. Residential care.

9. Other place(s) (where?).