Sweden is regarded as a universal welfare state in terms of its systems of benefits and services linked to the lifetime needs of its citizens. The formal goals of the Swedish welfare system comprise universalism and extensive coverage [16]. Since 1992, the municipalities have the overall responsibility for social care and services for older people and people with disabilities, including special housing [17]. However, the division of responsibility for health care is complex and varies depending on local agreements [18]. Health care provided by both county councils/regions and municipalities. The county councils are responsible for all health care except for older people or people with disabilities living in special housing. Within the county councils the regional/local hospitals are responsible for acute medical care that requires hospital admission, and primary health care centres are responsible for outpatient care. Municipal healthcare is in charge of health care for those living in special housing, and after local agreements for home nursing for those living in their own homes. This calls for integration and collaboration between all involved care providers. In Sweden, ‘Health care chains’, understood as “coordinated activities in the health care systems linked together to achieve a final result of good quality for the patient”, have become an important part of integrated care [19, 20, 21]. Different initiatives have been performed at the local level, however, national evaluations point to deficits in the provision of integrated care [4–6].

When introducing new initiatives to accomplish integrated care, Glendinning [8] emphasises the importance of distinguishing between short-term implementation problems and their solutions, and fundamental flaws that may impose barriers to the performance of integrated care for older people. She concludes that organisational integration therefore needs to be supplemented with extensive engagement in supporting change among those involved in performing integrated care, in order to relinquish traditional professional and organisational roles and status conditions. Åhgren and Axelsson [22] report that there are three major determinants of successful performance of integrated health care chains. These are professional dedication, legitimacy and confidence, and space to develop trust and motivation among the involved actors. Both Glendinning, Åhgren and Axelsson recommend a move towards bottom-up approaches to integration [8, 22]. Still, there is a need for further studies concerning the process of implementing new initiatives aimed at providing a continuum of care for older people.

This article is based on 44 qualitative interviews performed on four occasions between 2008 and 2010 with 26 key actors at three organisational levels within the organisations involved (Table 1). The respondents were interviewed on one to four occasions each. At the upper management level, the interviewees were the heads of geriatric, medical and emergency care departments at the regional hospital, the head of primary health care, the heads of home care, the social care office and the health care and rehabilitation department at the municipality. The interviewees at the management level were the department director at the emergency ward of the hospital, the managers for health care, and rehabilitation and social workers. The respondents at the operative level were the geriatric nurses at the emergency ward of the hospital and nurses from other hospital wards involved in the project, the CMs, and other members of the municipal team, e.g., social workers, and occupational and physiotherapists.

The interviews each lasted 30 minutes to 1 hour. Interviews were audio-recorded and transcribed verbatim. The analysis proceeded from open to focused coding in several steps [30]. Discussions among the authors took place in each step of the analysis. In the first step, we took an empirically oriented approach using the concepts understanding, commitment and ability as a general sense of reference, or as sensitising concepts [31]. Meaning units relevant to the aim of the study were selected from the interviews and condensed. In the second step, we identified and organised themes according to the sensitising concepts and the identified phases of the implementation process (Tables 2, ​,3,3, and ​and4).4). In the conclusive discussion we related the findings to our analytical framework and existing knowledge. We emphasise that neither the sensitising concepts nor the empirical themes should be regarded as mutually exclusive; they may overlap. The analysis was interwoven with other steps of the research process, as characteristic of most qualitative research [32].

This research forms part of the randomised control trial (RCT) study ‘Continuum of Care for Frail Elderly People’ evaluating the effect and implementation of a continuum of care model targeting frail older people [3, 33]. It is performed by an interdisciplinary research group at the Vårdal Institute, a national environment for research and development in the field of health care and social service. The research group consists of both researchers studying the effects of the intervention and researchers studying the implementation process (the authors of this paper). The results will be used to formulate a local programme for a continuum of care for this population group. The study was approved by the Regional Ethics Board in Gothenburg, Sweden (study code 413-08).

In the interviews, it became evident that the organisations involved had different traditions and a different understanding of project work. Consequently, the roles of the actors involved in designing and implementing the model have to be clearly defined, including who the senders and receivers are, as well as whether the process is to work top down or bottom up, or both [23].

Understanding the general aims of the model, and the ability to connect these to the operative performance, emerged as important in the interviews with the operative staff involved. Grasping the general idea contributed to the understanding of the weight of one’s own work and positively affected the response to the model. However, not only the staff who are directly involved needed to be informed about the general aims of the model. In order for the model to have an impact, the staff in organisational units that are indirectly involved are just as important. In addition to earlier identified components of successful implementation [24], it is important to elaborate a strategy for spreading and maintaining the information about both the aims of the model and how to work according to the model within the organisations.

When introducing a new model, the senders have to be aware of recent and parallel projects in the organisations, and how these fit in with the new one. Fundamental for the commitment of the actors involved in our study was that the new model corresponded to experienced problems and was perceived as having the capacity to respond to them. Furthermore, all organisations and professions involved have to see a potential benefit of the model for their own organisation and/or profession. These findings are also supported by results from previous research [34, 35].

Different aspects of what operative staffs need to be able to put the model into practice emerged in the interviews. It emerged as important for the senders of a new model to have sufficient knowledge about the organisations and professional roles involved to secure adequate competence of the operative staff and anticipate possible problems. Furthermore, they have to obtain resources to make implementation of the model possible. This encompasses both basic material resources, and time for the operative staff to perform their work properly, as well as time and other resources to work out and overcome obstacles in the organisations. Consequently, realistic interventions that fit within the existing conditions have been showed to be more successful [34].

The importance of involving all organisations concerned with the issue at hand have been pointed out in the literature [34, 35]. In our study the interviewees reported of occurring problems related to the non-involvement of primary care. Furthermore, an inter-organisational infra-structure has to be evolved at different organisational levels to promote trust and confidence among the actors [22]. In the project, this was done through the steering group and the project group. Both these groups had participants from all organisations involved.¹

In our study the model was revised and transformed several times during the course of the project. This points to the importance of enabling the operative staff to elaborate the detailed components during the implementation process and adapt the model to the specific circumstances at hand, i.e., to use their discretion [26] guided by mutual aims. Complex interventions have been found to be difficult to implement without allowing some flexibility, but the quality of the implementation relies on the understanding of the underlying ideas and core components of the implemented model [24, 36]. However, staffs operate in a professional landscape of conflicting and competing interests regarding professional status, discretion and jurisdiction [37]. As shown in other studies, our results reveal some conflicts and problems in the implementation process based on professional self-interest [8].

The commitment of the participants is crucial to the impact of the model, and depends on the model’s potential to solve critical problems experienced by the actors involved, professional groups and organisational units as well as whole organisations [34]. What the problems are, and how they are apprehended, is closely related to the demands of the organisation’s environment, and differs according to the organisational level [25, 38]. For the operative staff, the demands came mainly from the older people, their next of kin and the personnel carrying out their health/social care and rehabilitation, and concerned the ideology of care. For the managers, the demands came from both the top levels in their organisation and from the organisation’s environment, and concerned cost and work effectiveness to promote the legitimacy of the organisation [9, 25, 38]. Towards the end of this project, however, new competing ideas and policies regarding user choice caught the municipal managers’ attention and the project was marginalised within the organisation.

The new model comprised changing the traditional working modes and professional and organisational roles, as well as power and status conditions. Organisations, which have been described as a continuously changing and negotiated order [28], consist of segmented social units with contradictory interests and goals. Consequently, the new model imposed a threat to the negotiated order and was partially counteracted and fought by professional as well as organisational actors at different levels [8, 22].

Bibi Hølge-Hazelton, Director of Research Region Zealand DK; Honorary Teaching Fellow, Faculty of Health and Life Sciences, Coventry University UK; Associate Research Professor. The Research Unit for General Practice and Section of General Practice, Department of Public Health, University of Copenhagen, Øster Farimagsgade 5 Postboks 2099, 1014 København, Denmark