Integration in the literature

Integrated care has become a field of study in its own right for a dedicated academic community. What characterises these works is a shared interest for the elaboration of a common conceptual framework. Types, levels, breadth and degrees of integration, as well as the various mechanisms and strategies associated with it, have been identified, questioned and defined (Axelsson and Axelsson Reference Axelsson and Axelsson2006; Leutz Reference Leutz1999; Kodner Reference Kodner2009; Kodner and Spreeuwenberg Reference Kodner and Spreeuwenberg2002), leading to observation of the heterogeneity of the realities these can cover (Kodner Reference Kodner2006, Reference Kodner2009; Goodwin et al. Reference Goodwin, Dixon, Anderson and Wodchis2014; Somme et al. Reference Somme, Trouvé, Passadori, Corvez, Jeandel, Bloch, Ruault, Dupont and de Stampa2014a). Despite its various meanings, a common feature of definitions of integrated care is that ‘it seeks to improve the quality of care for individual patients, service users and carers by ensuring that services are well coordinated around their needs’ (Goodwin et al. Reference Goodwin, Dixon, Anderson and Wodchis2014: 2). From this perspective ‘it has been promoted as a means to improve access, quality and continuity of services in a more efficient way, especially for people with complex needs’ (Valentijn et al. Reference Valentijn, Schepman, Opheij and Bruijnzeels2013).

Various approaches to the study of integrated care have developed. Some focus on the analysis of specific integrative models or initiatives in an in-depth empirical case studies’ perspective: Kodner (Reference Kodner2006, Reference Kodner2009) explores health and social care partnerships in North America, Callahan (Reference Callahan2010) and Challis and Davis (Reference Challis and Davis1986) focus on pioneering experiences in the United Kingdom (UK) and United States of America (USA), and Williams and Sullivan (Reference Williams and Sullivan2009) on case studies from Wales. Others question overall national integration policies in specific national configurations – Sweden in Anderson and Karlberg (Reference Andersson and Karlberg2000) and Hedman, Johansson and Rosenqvist (Reference Hedman, Johansson and Rosenqvist2007); England in Humphries (Reference Humphries2015), etc. In these researches the objective is to identify structural reforms, social and economic drivers, legislative context and institutional framework. However, research projects can also combine both dimensions (Goodwin et al. Reference Goodwin, Dixon, Anderson and Wodchis2014; Williams and Sullivan Reference Williams and Sullivan2009), giving more or less importance to the national or local perspective.

In parallel, a comparative research agenda has developed in order to analyse integrated health and social care policies, measures, services or initiatives in different local and/or national contexts. Some studies pay most attention to the impact of national policies and pre-existing institutional, organisational and professional dynamics on the development of integrated health and social care initiatives in the elderly care sector (Leichsenring Reference Leichsenring2004; Mur-Veeman, van Raak and Paulus Reference Mur-Veeman, van Raak and Paulus2008; Trouvé et al. Reference Trouvé, Couturier, Etheridge, Saint-Jean and Somme2010). Similarly, although the work of Hardy et al. (Reference Hardy, Mur-Veemanu, Steenbergen and Wistow1999) focuses on the meso organisational level in integrated care inter-agency services in England and the Netherlands, the analysis also questions the integration process at a macro national perspective. These comparative research projects can focus on similarities and differences between national or local contexts, and even identify impeding and facilitating factors to integration, whatever the national context. In all these researches, accent is put on the importance of time and the necessity of considering integration as a process. As underlined by Goodwin et al. (Reference Goodwin, Dixon, Anderson and Wodchis2014: 1), ‘integrated care is a process that must be led, managed and nurtured over time’. By adopting the same diachronic and long-term perspective, Mur-Veeman, van Raak and Paulus (Reference Mur-Veeman, van Raak and Paulus2008: 181) consider integrated care development and delivery to be a laborious process.

All these authors have raised the question of what is the best approach to achieve integration, but the issue of whether integrated care is more likely to happen through a top-down (national level) or bottom-up (local level) mechanism has not been settled yet in the literature. Exploring the interplay between ‘integrated care policy’ and ‘integrated care development’ in different European Union countries, Mur-Veeman, van Raak and Paulus (Reference Mur-Veeman, van Raak and Paulus2008) confirm the conclusions of Leichsenring's (Reference Leichsenring2004) comparative work, which pointed out the impact of proactive national policies on integrated care development. Their analysis of the situation in Finland, Sweden, the Netherlands and the UK shows that legislation and specific national programmes promoting health and social care integration facilitate the development of local integration initiatives, whereas in countries having no such policies, such as Spain or Austria, integrated care development is weak. Yet the authors also stress that ‘promoting governmental policy […] is not sufficient to arrive at successful integrated care delivery’ given the existence of “local actors” constellations’ (Mur-Veeman, van Raak and Paulus Reference Mur-Veeman, van Raak and Paulus2008: 180–1), which play a major role in shaping the different approaches and pathways towards integration.

This bottom-up perspective is highlighted in many comparative research projects. National policies remain central explanatory factors in making integration happen, but the accent is on the local dimensions of integration processes and the importance of factors such as professional interests and practices and the ‘spaces’ in which they are shaped (Williams and Sullivan Reference Williams and Sullivan2009: 9). The comparative work of Goodwin et al. also questions the top-down approach, emphasising the fact that ‘most of the case study programs were the result of a bottom-up initiative to improve the delivery of care; none resulted from a national policy on integration’ (Reference Goodwin, Dixon, Anderson and Wodchis2014: 17). National policies may not match the experiences or priorities of local actors, and can even be a considerable source of tension (Goodwin et al. Reference Goodwin, Dixon, Anderson and Wodchis2014; Williams and Sullivan Reference Williams and Sullivan2009). The issue of whether only a bottom-up approach leads to the development of integrated care initiatives has been raised, ‘organizational integration [… being considered …] as a consequence rather than as a cause, and may not occur at all’ (Goodwin et al. Reference Goodwin, Dixon, Anderson and Wodchis2014: 17).

A multi-level approach based on the analysis of integration as a policy process

This paper does not explore the conceptual dimension of integration. Following Kodner and Kyriacou, ‘we consider integrated care to be a discrete set of techniques and organizational models to create connectivity, alignment and collaboration within and between the cure and care sectors at the funding, administrative and/or provider levels’ (Reference Kodner and Kyriacou2000: 1). Based on part of the work of Valentijn et al. (Reference Valentijn, Schepman, Opheij and Bruijnzeels2013), which distinguishes between the macro-level of system integration, the meso level of organisational and professional integration, and the micro-level of clinical integration, the analysis will refer to four main dimensions of integration: institutional (vertical and horizontal collaborations between the policy institutions involved), organisational (collaborations between services and/or structures involved in elderly care) and professional (relating to interprofessional partnerships, within or between organisations). The clinical level will be included, in relation to professional integration based on shared competences, roles and responsibilities, to deliver a comprehensive continuum of care to older people.

Analysing integration as a process, the article adopts a qualitative diachronic perspective, from the 1990s to the present. This extended timespan is necessary to consider the whole range of major and minor policy measures – which include laws, schemes, programmes and mechanisms – related to integration within the elderly care sector, as their impact can only be analysed by taking into consideration their time and logical sequence (Pavolini and Ranci Reference Pavolini, Ranci, Constanzo and Pavolini2013: 18).

At the macro-level, our objective is not to choose comparable measures a priori, nor to select a single type of integration, nor to focus on one specific dimension of the elderly care sector (hospital, home-based care, residential care, Alzheimer's disease issues, etc.), but rather to ‘unroll’ elderly care policy and spot all the measures related to one aspect or another of the integration of the health and social sectors, in order to offer a global policy analysis of what integration is in the two countries; how it is shaped and what policy orientation(s) is(are) revealed by the measures identified. In this perspective we will not refer to pre-existing definitions making the distinction between co-ordination and integration. The analysis will identify policy measures which concern older people and which relate explicitly or not to integration of health and care sectors over the last three decades, going back earlier if necessary. Conversely, at the micro-level, in the same diachronic perspective (but over a shorter period) the article focuses on the development of specific integrated care initiatives in the two countries and investigates how local actors and local dynamics produce integration.

Referring to the definitions given above, the public policy focus adopted here explores how national and local interventions in favour of elderly care include and conceive the issue of integration of health and social care sectors. At both levels, this policy process approach will enable us to investigate the institutional, organisational, professional and clinical types of integration developed, as well as some of the mechanisms associated with it: control, funding, joint management, agreements, case management, care planning (Kodner and Spreeuwenberg Reference Kodner and Spreeuwenberg2002) – and the tensions they address or induce. The distinction between three ideal types of integration (hierarchy, market and network) proposed by Valentijn et al. (Reference Valentijn, Schepman, Opheij and Bruijnzeels2013) in reference to Thompson et al. (Reference Thompson, Frances, Levacic and Mitchell1991) will lead us to discuss these various co-ordinating mechanisms. For the authors, the flexibility offered in market governance situations is greater than that in hierarchies, yet the commitment between the organisations concerned is minimal (Valentijn et al. Reference Valentijn, Schepman, Opheij and Bruijnzeels2013: 5). Network modes of co-ordination, considered characteristic of health and social care integration, are presented as alternatives implying ‘more or less voluntary collaborations between organizations’ and depending ‘on relationships, mutual interests and reputation’ (Valentijn et al. Reference Valentijn, Schepman, Opheij and Bruijnzeels2013: 5).

Analysing both the national and local dimensions of integrated care processes, this multi-level approach leads us to investigate if and how integration is produced, by addressing the interplay between local integrated care initiatives and national policy measures in the elderly care sector. A range of questions can be raised: what impact do national policy measures have on the local development of integrated initiatives? Conversely, how does the development of local initiatives contribute to the definition of national integrated care policies in the sector of elderly care? Are local initiatives the implementation of national integrated care policy – or do they emerge at local level through the partnerships and collaborations organised by local actors?

The development of integrated elderly care in France: a chaotic trajectory and an overlaying of measures

Considered as a ‘welfare mix’ system combining public intervention and family care (Le Bihan and Martin Reference Le Bihan and Martin2012), the French elderly care policy has put forward the necessity to develop co-ordination at various levels. As a number of recent works recall (Argoud Reference Argoud, Hummel, Mallon and Caradec2014; Bloch and Hénaut Reference Bloch and Hénaut2014; Somme and de Stampa Reference Somme and de Stampa2011; Trouvé et al. Reference Trouvé, Couturier, Etheridge, Saint-Jean and Somme2010), it has been both a long-standing preoccupation of the state (Amyot Reference Amyot2006; Henrard and Vedel Reference Henrard and Vedel2003) and a permanent response to the highly fragmented organisation and funding of health, social and what are referred to as ‘medico-social’Footnote ¹ interventions (Bloch Reference Bloch2013; Henrard Reference Henrard2002) in the elderly care field.

The institutional settings of health and social care sectors and their evolution are characteristic of this fragmentation. Several levels of governance are involved in the organisation of the two sectors: the state, the region, the département and the municipality. The source of funding is also different: social insurance for the health-care sector and (national and local) taxation for social care. In France, the government defines national health and social policies through legislation, and different territorial levels are involved in the management and funding of the two sectors. Regional and local administrations execute national health policies under the strong supervision of the government, whereas the decentralised French local authorities – départements – are responsible for social policy. Within the elderly care sector, the départments have a statutory obligation to define the local policy orientations on their territory, to finance and implement the national allowance for autonomy (APA), and to regulate care services within their territory. In addition, municipalities can develop specific measures in favour of older people, on a voluntary basis.

This first line of divisions between the different territorial levels in the governance of the elderly sector has been reshaped by two laws. The 2004 law on solidarity and loss of autonomy which introduced the CNSA (the national solidarity fund for autonomy), a new national institutional actor in charge of older and disabled people; whereas a 2009 health law – ‘the Hospital, Patients, Health Territories’ Act – created a new regional institution representative of central government – the ARS (regional health agencies). Comprising all pre-existing regional and local health administrations, these new agencies extended traditionally health sector-only intervention to the social care sector. The introduction of a single institutional actor, representative of central government, dealing with both dimensions of care was a major objective of the reform. Yet, since the ‘départements’ maintained their leading role in the social policy field, the need for co-ordination between the two institutional levels dealing with different aspects of health and social care became a major challenge.

Despite their initial objective of simplifying the institutional context through vertical integration, these recent evolutions have further contributed to the complex and changing institutional setting. In practice, elderly care depends on a wide range of regulations and funding authorities whose overlapping competences create a number of tensions: at the national and local decision-making level, both between and within the successive health and social affairs ministries and local councils concerned; at the implementation level, between the various administrations (national agencies, local administrative services) concerned. These tensions are further accentuated by fragmentation of responsibilities for the service provider organisations – hospitals, social care services, nursing care services, residential care homes, collective homes – as well as for the different professionals involved, who often work separately. As a consequence, a vast range of entry points and pathways into the health and social care systems co-exist (Bloch and Hénaut Reference Bloch and Hénaut2014) and users are confronted with a wide variety of often-overlaying schemes, depending on the different (local and/or national) administrations.

Since the 1980s, a number of dedicated measures have addressed the concrete need for co-ordination as a response to this highly fragmented institutional, organisational and professional context (Trouvé et al. Reference Trouvé, Couturier, Etheridge, Saint-Jean and Somme2010). At that time, various structures – such as residential homes, home care services, day care solutions, temporary accommodation – had developed in parallel within the health or the social care sectors often ‘in a haphazard manner’ (Amyot Reference Amyot2006: 17). A range of co-ordinating mechanisms aimed at creating links between existing organisations and professionals were introduced in the 1980s: ‘gerontological action co-ordinators’ for the social sector and ‘gerontological health networks’ for the health sector. During the 1990s, multi-disciplinary care teams were introduced in the département as a part of a national reform creating a specific allowance for older people, in a bid to offer them financial support in setting up an adapted care arrangement (Le Bihan and Martin Reference Le Bihan, Martin, Ungerson and Yeandle2007). Created within the ‘départements’, these multi-disciplinary care teams connected health and social professionals, representing the first form of case management procedure for older people in France (Frinault Reference Frinault2009). In the same period, ‘local information and co-ordination centres’ (CLIC: Centres Locaux d'Information et de Coordination) developed, to help care users find their way through the maze of mechanisms by introducing a one-stop shop for social care services (Frossard et al. Reference Frossard, Genin, Guisset, Villez, Leichsenring and Alaszewski2004). Within the health sector too, different co-ordination schemes seeking to develop multi-disciplinarity while facilitating links between the various health professionals multiplied. In 2010, with the MAIA methodFootnote ² – initially part of the 2008–2012 Alzheimer plan, later extended to the whole elderly care sector – the co-ordination objective was reaffirmed. Moreover, the new term established was integration – with this semantic distinction marking a break, in the official discourse, with earlier attempts at co-ordination that had been found wanting (Somme et al. Reference Somme, Trouvé, Passadori, Corvez, Jeandel, Bloch, Ruault, Dupont and de Stampa2014a). Since then, two recent laws – the 2015 Act on adapting society to an ageing population and the 2016 health law – have introduced, independently of one another, new measures which aimed at incentivising joint decision-making between regional and local institutional actors, new forms of territorial co-ordination for health and social care organisations, and further clinical integration – in which the concept of health pathways became central.

This specific trajectory of national co-ordination policy measures continuously calling into question the effectiveness of pre-existing schemes, yet without removing them, is a major characteristic of the French context (Amyot Reference Amyot2006). It determines the chaotic trajectory of French integration policies in which the successive schemes contribute to institutional, organisational and clinical complexity.

The development of integrated elderly care in Sweden: co-ordination efforts falling into a broader movement towards marketisation

In Sweden, characterised by its universalist dimension and strong decentralisation (Anttonen Reference Anttonen and Simonen1990), responsibilities between health and social care are shared out between three levels of governance: central government, county councils which are responsible for health and nursing services, and municipalities which manage social sector care services. In the late 1970s, with the emergence and gradual consolidation of home care in the elderly care sector, a debate on the necessity of rationalising responsibilities at municipal and regional levels was initiated. However, it was only in 1992 that the significant organisational Ädel Reform targeted in-depth transformation of the elderly care sector (Trydegård Reference Trydegård2003). As analysed by Hedman, Johansson, and Rosenqvist, improving quality outcomes for older people with complex needs was a major objective of the reform, alongside the need ‘to find more efficient ways to organise old-age care under one authority … [and] to increase the integration of health and social care’ (Reference Hedman, Johansson and Rosenqvist2007: 2). Transferring all home and institutional nursing care services to the municipal level, the essence of the Ädel Reform was to concentrate all elderly care outside hospitals within the municipalities. Initiated by central government, the reform is based on financial incentives to create collaboration between county councils and municipalities, as well as between health professionals and care workers, on a voluntary basis (Berglund et al. Reference Berglund, Blomberg, Dunér and Kjellgren2015).

Continuity has been a key feature of the integration process within the elderly care field since the beginning of the 1990s. The Ädel Reform has been implemented progressively, with 56 per cent of municipalities concerned in 2007 (Berglund et al. Reference Berglund, Blomberg, Dunér and Kjellgren2015). Besides, in the 2000s other policy measures have been taken to facilitate seamless care which had an impact on the elderly care sector: a specific law on joint political boards in health and social care in 2003 and specific incentives to promote collaboration between health and social care providers and professionals. These resulted in different forms of inter-organisational local systems, often with a focus on older people (Ahgren and Axelsson Reference Ahgren and Axelsson2011). Since the beginning of the 2010s, the issue of integrated care has been high on the political agenda, with a strong national pressure for transferring home health care to all municipalities (Meagher and Szebehely Reference Meagher, Szebehely, Ranci and Pavolini2013). In parallel, a governmental initiative called ‘Co-ordinated health and social care for the most severely ill people’ (2011–2014) was launched, targeting the co-ordination of health and social care for the ‘most severely ill elderly’.

National elderly care policies developed in Sweden are also linked to a second policy objective: its rationalisation and cost-effectiveness improvement. Indeed, undertaken in a context of strong financial pressure, the integration objective of the Ädel Reform also falls within this perspective. It seeks to better articulate hospital and social care services in order to shorten hospital stays for older people and improve the performance of hospital services (Hieda Reference Hieda2012), as does the above-mentioned governmental initiative, which targets the reduction of hospital (re)admissions through better co-ordination between health care, home care and residential care. This rationalisation issue leads us to the issue of marketisation of care. As explained by Szebehely and Trydegård (Reference Szebehely and Trydegård2012), the Swedish elderly care sector has been greatly influenced by the global wave of New Public Management reforms since the second half of the 1980s. Important steps in this direction include the local government act and the act on public procurement (LOU) of 1992 – which established the principle of separation of purchaser and provider and introduced competition within the care market on the supply side. A further step in this ‘choice revolution’ (Blomqvist Reference Blomqvist2004) was the 2009 act on free choice (LOV) which made it easier for municipalities to introduce a customer-choice system and competition on quality on the demand side. In December 2012, 133 of the 290 Swedish municipalities had implemented LOV, and another 42 municipalities had decided to do so (Erlandsson et al. Reference Erlandsson, Storm, Stranz, Szebehely, Trydegård, Meagher and Szebehely2013; Socialstyrelsen 2013). However, because it is implemented on a voluntary basis, further financial incentives have been introduced in 2011 to push municipalities towards this model.

Integration policies are linked to marketisation policies in a manifold way. Further the rationalisation objective, by opening the care sector to private (for-profit) providers, marketisation policies have introduced a diversity of care providers belonging to different organisations having different rules and cultures. Differentiation of responsibilities is therefore accentuated, and new forms of organisational and professional fragmentation have emerged. As exposed by many researchers (Ahgren Reference Ahgren2010; Ham Reference Ham2012; Leichsenring et al. Reference Leichsenring, Rodrigues, Winkelmann and Falk2015), the development of the two policies – one promoting competition and consumer choice, and the other the integration of health and social care – can appear contradictory. Indeed, associated with vertical concentration, integration can reduce the user's options. Conversely, a consumer-choice perspective fosters the diversity of competing care providers, while integrated care promotes co-operation between a few providers. Besides, as argued by Kremer (Reference Kremer2006), the major role given to professionals in the integration process can lead to user disempowerment and contradict the consumer-choice objective. The dichotomy between marketisation and integration policies can also be illustrated by the development within the health-care sector of ‘chains of care’ in the 1990s and ‘local health systems’ in the 2000s. Presented as alternative quality models to the market-oriented policies implemented since the 1990s (Ahgren and Axelsson Reference Ahgren and Axelsson2011), they rely on a strong focus on patient needs and the use of evidence-based care pathways.

The co-existence of these two major policy trends – integration and marketisation – is a main characteristic of the elderly care policy in Sweden. Nevertheless, analysis shows that they have different positions, marketisation constituting an overwhelming framework shaping other orientations within the sector (Meagher and Szebehely Reference Meagher, Szebehely, Ranci and Pavolini2013) – including those related to co-ordination of care.

The French MAIA scheme

The local MAIA integrated scheme is part of the third national plan on Alzheimer's disease (2008–2012) in favour of the integration of health and social care services for older people. Tested in 17 localities (2009–2011), it was later expanded to cover the whole territory. Inspired by the Canadian Prisma programme (Somme et al. Reference Somme, Trouvé, Perisset, Corvol, Ankri, Saint-Jean and de Stampa2014b), it includes three integration mechanisms: a consultation process between institutional (decision-making and financing) and professional actors; an ‘integrated entry point’; and intensive case management for complex cases. The MAIA scheme seeks to achieve integration at all four of the interdependent levels explored in this article: institutional, organisational, professional and clinical. Its local implementation is a two-stage process: a call for projects financed by the regional health agencies and a project proposal initiated by local actors from the health and/or social sectors. For each project selected, a dedicated professional – the MAIA ‘pilot’ – is recruited to foster the integration dynamics across the concerned territory, alongside between two and five case managers, who are in charge of up to 40 complex cases each. These managers perform clinical integration with the other professionals of the health and social care sectors on the territory.

Local integration processes over time: the importance of pre-existing dynamics and local actors’ investment in elderly care

The local MAIA scheme investigated here was created in 2012. It involves one MAIA pilot and three case managers. It was initiated by four local actors from the social care sector: the four local information and co-ordination centres (CLICs): ‘We were aware that we were working in the same directions. And when we saw the national definition of the MAIAs, we thought that we were legitimate to initiate the scheme. We were already dealing with the co-ordination part, and the MAIA scheme was a way to go further’ (CLIC Chief Director). The creation of the new MAIA scheme is thus closely linked to pre-existing local dynamics.

The local MAIA scheme selected for our study was created in a département having a strong political investment in the elderly care sector. Since the 1980s, it has implemented successive national social care measures – mainly local autonomy committees, multi-disciplinary teams, and the aforementioned local information and co-ordination centres – in a bid to maintain as much connection as possible between their respective actions within the social care sector: ‘We had anticipated on this, much before the creation of the regional health agency. We had – and still have – a well-defined and a pro-active co-ordination policy in elderly care’ (High official, département). In the health sector, national legislation led to such network-based co-ordination initiatives as gerontological and hospital geriatric networks, which have been implemented in parallel. Over time, this parallel development of schemes (independently of one another) reinforced segmentation between the health and social sectors. At the institutional level, each actor managed and financed its own co-ordination policies, whereas at the organisational level, health and social care schemes and their associated professionals co-existed across the same territory.

Creation of the scheme: reception and constraints

As analysed at the national level, national efforts have been directed, since the 2010s, at linking the health and social care sectors. The creation of both the regional health agency and the MAIA model fell within this objective. Nevertheless, according to the various local actors interviewed, the creation of the MAIA scheme was initially perceived as yet another co-ordination initiative, coming on top of the previous ones. As such, its introduction corresponded, at the local level, to the overlaying process identified at the national level. At the institutional level, the consultation process between institutional and financing actors organised by the MAIA pilot implied specific representation, particularly from the département and regional health agency. As explained in the interviews, the multiplication across the territory of meetings requiring institutional representation as a direct consequence of the multiplication of schemes deprives the designated representatives of their concrete decision-making power: ‘We participate to so many different working groups […] and it is always the same people going. We have the feeling to work on the same things in many different places […] It is always slightly different. We spend a great deal of time and energy on this!’ (High official, département). As such, it is perceived as a real obstacle to the development of a structured decision-making ‘space’ that was targeted by the scheme. At the organisational level, the creation of an integrated entry point also implied representation from the various local health and social care providers. In addition, critics considered it to overlap with the one-stop-shop function assigned to the existing local information centres: ‘Because “entry point” is often understood as “one-stop shop”, which corresponds to the place where the user goes’ (High official, regional health agency).

At the professional and clinical level, intensive case management has been difficult to set up. In dealing with older people in ‘complex situations’, the new case managers encountered professional resistance from social workers of the multi-disciplinary teams providing personalised follow-up of those older people eligible for the département’s personal autonomy allowance. The reluctance to refer older people to new case managers is related to both the social workers’ professional interest in complex situations and an unwillingness to share their follow-ups with the incoming professionals, who call into question their legitimacy: ‘When the case managers explain their work, it looks so much as what we do. I find it really hard. I also feel bitterness: they have a case load of only 40 situations. What does this mean for our work?! It is completely discrediting it! Our work must be really pathetic if we manage 458 follow-ups a year!’ (Social worker).

Integration activities and local recompositions

The qualitative analysis of interviews performed at the three levels studied leads us to identify different evolutions in the initial positions and practices of the local institutional and professional actors, since the creation of the MAIA scheme.

At the institutional level, the evolution concerns relations between the département (initially opposed to the creation of the scheme) and the regional health agency. The argument against the overlaying effect of the MAIA scheme's creation was not the only criticism formulated. Since the MAIA scheme is financed and supervised by the regional health agency, its implementation was perceived as a clear competitor for the département’s traditional role within the field of elderly care. In addition, in terms of funding, it feared that financing of the scheme would be transferred to the local level, as had previously been the case for the local information centres. Nevertheless, the regular meetings between the two institutions helped clarify the positions of both actors. First, the role of the département in defining the agenda and the content for the meetings was reinforced, which helped balance the political power between the two institutional actors. Second, with the confirmation of the national rollout of the scheme, the département repositioned itself as a project initiator for future MAIA schemes to be created to cover the rest of its territory. The conditions for this investment were directly negotiated with the regional health agency, which agreed to extend the scheme's target population to disabled people, as was already the case for other schemes supported by the département. The initial constraint represented by the national rollout objective for the scheme was thus translated into an opportunity for the local authority to reinforce its leading position in the long-term care sector.

At the organisational level, the regular meetings between the different health and social care providers organised by the MAIA pilot built up the actors’ mutual knowledge and understanding of the scheme's objectives and operational modes. These last have been criticised for their lack of clarity and conceptual ambiguity regarding concrete implementation of the scheme: ‘Everyone was saying “What is this!”’ (High official, département); ‘How is it supposed to work in people's real lives? It is a conceptual tool which hasn't been confronted with the reality’ (Social worker). Yet, the professionals succeeded in developing a number of shared tools that clarified its different components:

Similarly, the formal meetings enabled representatives of the département to participate in drawing up criteria for the inclusion of older people in intensive case management, and thus defend the professional interests of the social workers performing follow-up for complex cases within the département’s multi-disciplinary teams, who feared that this task would be exclusively assigned to the new MAIA case managers. As a social worker told us: ‘Strong initial resistance created a hostile climate with the new case managers, but was ultimately useful, since it made it possible to defend pre-existing practice and local specificities’.

The Norrtälje model

The Norrtälje model was created in 2006 as a new health and social care model for the municipality's general population, for a period of five years. It was then extended until 2013 and again until 2015, finally to be established on a permanent basis in 2016.

Local integration processes over time: the importance of pre-existing dynamics and local actors’ investment

One of the main reasons for the creation of the model was the threat of closure of the local hospital in 2002, within a context of strong financial constraints. Several local key actors – which can also be referred to as ‘boundary spanners’ who manage ‘within inter-organizational theatres’ (Williams Reference Williams2002: 103) – played a major role in exploring the available options for saving the hospital, designing a project that capitalised on good local collaboration culture, on the basis of existing legislation on joint political boards that had been adopted in 2003:

The model jointly agreed upon by the county council and the local authority, and implemented in January 2006, had two main components. At the institutional level, a political board comprising representatives from the municipality and the county council jointly decided on health and social care policies financed through pooled local and county council taxes. The existence of such a single decision-making and financing body explains the centrality of political and financial issues to institutional integration in Norrtälje. At the organisational level, a single public company in charge of providing health and social care for the whole population was created, gathering services formerly provided separately (primary care, hospital, home care, residential care), now comprising the different units of the company. Lastly, clinical integration is closely related to organisational integration, since all the professionals involved in elderly care are employed by the same public provider, and deliver all the services in the care chain.

Implementation of marketisation legislation and impact on the local model

Ever since its creation, the characteristics of this model have evolved in response to both national and local dynamics. As partially documented in the literature (see mainly Andersson Bäck and Calltorp Reference Andersson Bäck and Calltorp2013; Øvretveit, Hansson and Brommels Reference Øvretveit, Hansson and Brommels2010; Schön, Hagman and Wanell Reference Schön, Hagman and Wanell2012), these changes correspond to different periods in the development of the Norrtälje model, and are strongly related to national marketisation policies. The first major change occurs in 2006, just six months after its creation, with the split of the public company into two different entities: a financial organisation responsible for the joint purchasing, planning and monitoring of health and social care, and a public company providing health and social care services. This corresponds to implementation of the purchaser provider legislation of 1992, accelerated by the change in the local authority's political majority in 2006. The second change is linked to the arrival of private providers on the territory in 2008, anticipating the national legislation on choice in 2009 (LOV). In both cases, the changes are due to county council politicians’ political concern to keep – for reasons of equity – the financing and organisation of health services in Norrtälje as similar as possible to what is in place in the rest of the county.

These two turning points have modified the model's original features. At the institutional level, integrated purchasing of services was challenged by the arrival of new providers – the purchasing unit had to extend its contractual relations to the rest of the service providers present across the territory. The need to develop a more cross-boundary contractual arrangement with all providers was underlined in a report (Sjögren and Åhblom Reference Sjögren and Åhblom2012). Service organisation and established professional practices were also challenged. Health and social care professionals from the public company had to adapt their work habits and invent new routines with the new providers in order to limit service discontinuities:

Integration activities and local recompositions

Analysis from interviews with local institutional and professional actors shows that the implementation of national measures has led to concrete changes at all four interdependent levels of integration explored here.

At the institutional level, right after the introduction of the purchaser/provider split, the accent was on planning new financing by linking reimbursement mechanisms to three patient groups, by age (0–18, 18–64 and 65+) as a further step towards financial integration (Øvretveit, Hansson and Brommels Reference Øvretveit, Hansson and Brommels2010). Yet, deeper financial integration was blocked at the higher, county council level:

The absence of a satisfactory solution led to the resignation of the directors of both entities and to the appointment of new directors in 2013.

Similarly, at the organisational level, new local rules for service provision were negotiated between local authority and county council politicians prior to the introduction of the choice model in 2008. Their aim was to limit service fragmentation and preserve the integrated delivery of elderly home care. First, new care providers had to be able to supply three types of services – home health care, home social care and home rehabilitation – as a mandatory precondition to registering as authorised providers (Schön, Hagman and Wanell Reference Schön, Hagman and Wanell2012; Ström Falk Reference Ström Falk2015). ‘The most successful was the agreement with all the providers which leads them to deliver social care, home health care and rehabilitation for the same person […] It was also quite new’ (Formal high official, Purchasing unit). Using the same perspective, the territory covered was divided into four geographic areas with a limited number of home care providers in each (see Ström Falk Reference Ström Falk2015). Finally, in order to avoid discontinuity between hospital and home care following a hospital stay, a new customer procedure was introduced at hospital for those patients unwilling or unable (Meinow, Parker and Thorslund Reference Meinow, Parker and Thorslund2011) to choose a home care provider. The choice is then performed by professionals from the hospital – limiting the risk of breaking the continuum of care – according to geographic area and day of the week, to remain respectful of competition between providers.

Finally, at the micro-clinical level, the focus was on integration activities performed within the public company. Aimed at creating a smoother path for the patient, these are of various types: restructuring within the company's different units; creation of a new reception system in primary care (2013) or introduction of care planning meetings with home care professionals at the hospital (2014). Though some of these activities are related to macro institutional integration (Øvretveit, Hansson and Brommels Reference Øvretveit, Hansson and Brommels2010), the nomination of the new company head in 2013 has accelerated this evolution. A reorganisation of the company's governance was instigated, more closely associating top managers of the various units with decisions, and different initiatives were developed, to favour mutual knowledge between different units.

In this sense, the arrival of the new directors helped draw attention away from the controversial issue of financial integration, transforming the conflictual situation related to financial issues into an opportunity to pursue deeper professional and clinical integration.