Norwegian general practitioners’ collaboration with municipal care providers – a qualitative study of structural conditions

Abstract

Purpose: The purpose of this study was to explore the structural mechanisms that facilitate or counteract collaboration between general practitioners (GPs) and other providers of municipal healthcare. Good collaboration between these actors is crucial for high-quality care, especially for persons in need of coordinated services.

Material and methods: The study is based on semistructured interviews with 12 healthcare providers in four Norwegian municipalities: four GPs, six nurses and two physiotherapists.

Results: GPs are key collaborating partners in the healthcare system. Their ability to collaborate is affected by a number of structural conditions. Mostly, this leads to GPs being too little involved in potential collaborative efforts: (i) individual GPs prioritize with whom they want to collaborate among many possible collaborative partners, (ii) inter-municipal constraints hamper GPs in contacting collaboration partners and (iii) GPs fall outside the hospital-municipality collaboration.

Conclusions: We argue a common leadership for primary care services is needed. Furthermore, inter-professional work must be a central focus in the planning of primary care services. However, a dedicated staff, sufficient resources, adequate time and proper meeting places are needed to accomplish good collaboration.

Keywords:

• Collaboration
• general practitioner
• municipal healthcare
• structural conditions
• integrated care

Introduction

General practitioners (GPs) in Norway play a key role in the provision of healthcare in the municipalities. Good collaboration between GPs and the care service professionals in the municipalities is crucial for high-quality care, especially for persons in need of comprehensive services, such as fragile elderly persons, patients with chronic illnesses, patients with substance use problems and people needing to be followed up during long-term sick leaves. Previous research has shown that inter-municipal collaboration can be challenging or lacking, which in turn could influence the services offered to patients [1–4].

Healthcare and social services in Norway are based on the classic Scandinavian Welfare model with financing and the provision of universally accessible services to everyone. Specialist healthcare is organized by central state authorities, while the municipalities are responsible for primary medical services (GPs), home-based services, long-term care services and social care. They are also obliged to offer rehabilitation and preventive health work. Most GPs work as self-employed on contract with the municipalities, while nursing homes, home-based services and social services are publicly financed and mainly publicly provided.

One characteristic of primary healthcare is that most services are separately organized. Home-based services are organized in groups responsible for defined geographical areas within the municipality. In larger municipalities, the allocation of home-based services is done according to a purchaser–provider model. An application for services is sent to the purchaser office, who decides on the amount of services needed. If the application is granted, an order for the service consented is sent to the provider. GPs are, since 2001, organized according to a list patient system, where patients may register themselves as regular users of the doctor they prefer. On average, the patient list of a GP is approximately 1200 patients. The GP reimbursement system is based on a combination of capitation and fees for service. It may seem as if GPs changed their working practices after the introduction of the list patient system. GPs experiencing competition for patients seem to have more consultations and carry out more tests than other doctors [5]. Kann et al. [6] found that GPs practicing in high competition areas prescribe more drugs to the elderly than GPs facing low competition, and GPs experiencing patient shortages tend to prescribe more reimbursable and more addictive drugs than do their unconstrained colleagues.

As part of the Coordination Reform implemented in 2012, increasing responsibilities were transferred from the central to municipal government, with the key aim of reducing length of stay in hospitals and widening access to health services within the municipality [7]. An important element in the reform is that municipalities are financially responsible for patients who have to stay in hospital and wait for transfer to municipal care, after having been considered ready for discharge. From many municipalities, it has been reported that patients are discharged at an earlier stage and in worse condition than before [8]. Earlier discharges of patients in poor condition make collaboration between GPs and other parts of local healthcare even more decisive. The white paper ‘The primary health and care services of tomorrow – localised and integrated’ [9] can be seen as a continuation of the Coordination Reform, placing emphasis on user involvement, prevention, proactive follow-up, collaboration and good pathways.

Factors that promote and hampers collaboration

Efficient collaboration between personnel in municipal health and social care is considered important – both for the quality of services given and for the patients’ well-being and experience of a seamless service [10,11]. The GP holds the overall medical responsibility for patients in the municipalities and consequently plays a key role in patient care. Because many patients of GPs are in need of comprehensive services, GPs need to interact and coordinate treatment and care with most other parts of municipal care. Health and social service personnel in primary healthcare often claim that collaboration – especially between GPs and allied personnel – is weak or non-existent [1–4]. Measures to improve collaboration do not seem to have had the desired effect [12–14].

Collaboration can be conceptualized in different ways and it includes a wide spectrum of activities from simple electronically conveyed information exchange via face-to-face encounters to comprehensive integrated collaboration in meetings. Integration necessitates a certain degree of collaboration among the parties caring for the same patient [15–16]. On a practical level, it requires effort to integrate and translate the themes and schemes shared by different professional groups and the shared ownership of common goals, decision-making processes, and the integration of specialized professional knowledge and expertise. The barriers that must be overcome in terms of successful integration between healthcare professionals include differences in professional cultures, competence, knowledge and ways of thinking and a blurring or misunderstanding of professional identities, roles and responsibilities [16–20]. Insight into each other’s work, a culture of mutual respect and recognition of each other’s areas of expertise and competence, the free and open exchange of information, good relationships between providers, and having sufficient time and resources for ongoing relationship building are also key elements. The institutional, organizational, and economical contexts constitute important preconditions for collaborative functioning, either as facilitating factors or as barriers [16,21,22]. The key significance of effective organizational leadership is highlighted. There is consequently much knowledge of collaboration and the lack of collaboration in health and social care, as well as many explanatory models for understanding collaboration problems at the individual level. This study aimed to explore how structural conditions facilitate or restrain collaboration in practice between GPs and other providers in the municipalities.

Materials and methods

This study is based on a series of semistructured interviews with health personnel in four Norwegian municipalities during spring 2011. Two cities and two rural municipalities were included. Informants were recruited by letters to administrative leaders of the health service sector in the various municipalities, asking them to pass our request on our call for an interview to representatives of the health services in question. In total, 12 persons were interviewed. Informants are presented in Table 1.

Table 1. Providers interviewed.

	City	Rural
GPs	3	1
District nurses	1	1
Physiotherapists	1	1
Public health nurses	1	1
Psychiatric nurses	1	1
Total	7	5

The construction of interview guides was inspired by the Critical Incident Technique [23]. Informants were asked to describe an ordinary day of work, focusing on tasks in which some sort of interprofessional collaboration was considered favourable, seen from either the GP side or the other side. When collaboration was considered favourable but not possible, informants were asked to describe situations and processes wherein collaborative initiatives were unsuccessful. Interviews were recorded and transcribed in full length by a secretary. Analysis was performed according to the principles of grounded theory [24]. By reading and re-reading the interviews, comparing important similarities and differences in the informants’ stories, we aimed to conceptualize the general and common elements in our informants’ experiences.

Results

In the analysis, we identified three main themes that seem to affect GPs’ collaboration with municipal actors:

• GPs’ key role: a matter of priority and individual choice

• Inter-municipal organizational constraints prevent the GPs from contacting collaboration partners

• GPs fall outside the hospital–municipality collaboration

GPs’ key role: a matter of priority and individual choice

The GPs interviewed were concerned with their key role in primary healthcare with the great number of different patients seen daily, leaving them with a high number of possible collaborative partners. They described their workday as busy and unpredictable, with tasks of various urgencies. GPs’ explained that they had to prioritize with whom they would collaborate:

Collaboration and meetings: We have to make priorities on time – because then we have less time for….other tasks. And I think many of our collaborators, they kind of don’t understand. I have 1,400 patients on my list, and I have 18 patients in common with district nurses in home-based care. Of course, maybe these are very important patients because they are old and need a lot of care, but I work with so very many others. While district nurses, they have only those [patients].

Doctors shared that they have to prioritize with whom they want to collaborate based on what they see as most important in the actual situation. The quality of personal relations with collaborative partners is also reported to be of significance. A GP with a strong engagement in working for psychiatric patients and drug addicts experienced personnel discontinuity among his collaborative partners in other parts of health and social security as a frustrating obstacle to good collaborative relations:

There are maybe too many actors… I like to think that having met—at least once—that is a minimum needed to oil the collaboration between us. But when you have established…have had some good meetings and developed a good understanding of the patient’s situation, then you may be pretty sure that within a month or two, that person is heading for a new role….And so, I’ll be sitting here and feel the impact of the hell of reorganizing—don’t tell me!

Most of the GPs’ collaborative partners interviewed emphasized the great variation among GPs. In general, they reported it easier to identify and get in contact with the patient’s (responsible) doctor with the patient list system; nevertheless, there were great individual differences. Some doctors were reported to participate in meetings and engage in common routines to follow-up with patients. Others did not show up to meetings or they did not respond to requests from local health personnel concerning common patients. Patients with mental health problems and patients with complex problems needing long-term care are (by law) entitled to a follow-up plan, linked with a team of responsible actors in healthcare. In most cases, a GP is required to take part. GP participation is dependent on personal choices made by the individual GPs themselves. Some GPs participate actively when asked, but others do not. A psychiatric nurse in a smaller municipality described his experience with GPs’ participation in patient-related teams:

Some of them have a genuine interest for the patients. Some of them are sincerely interested. Others consider them a pest…troublesome. These GPs don’t bother to show up to team meetings. They don’t have time. Those who are interested, however, they show up, and we have a very good collaboration with them’.

Inter-municipal organizational constraints prevent GPs from contacting collaboration partners

The informants – both GPs and their collaborative partners – reported that different organizational systems make it difficult to initiate collaboration. The combination of the patient list system- and the home-based services’ geographic organization might for example hinder collaboration. Patients on a GP patient list may be recruited from all over town, as choice of doctor often relates to where people work and not to where they live. Home-based care, on the other hand, serves patients living in separate districts within the city. An example of collaboration on a practical level is the routines available for collecting blood tests. A blood test may give the GP important information, and it is easy to administer by a nurse in a patient’s home. The incompatibility between district-organized services and list-organized GPs, however, was said to impose practical obstacles for even this kind of collaboration. A home-based care nurse’s experience was that GP offices near to where a patient lived insisted that blood samples be delivered to the patient’s own doctor for analysis. Giving this kind of service to another doctor’s patients was considered contrary to the internal logic of the ‘system:’

Collecting blood for testing takes five minutes, but then we may have to drive to the other side of town to deliver it. We have to deliver it to the [patient’s] GP office, which may be situated anywhere. We have tried to get appointments with GP offices nearby…Their medical secretaries did not want additional work [by serving other GPs’ patients].

Another collaborative hindrance described by the GPs was the purchaser provider organization of home-based services. This organization implies that GPs have to send an application to the office of the purchaser, who decides on the amount of services needed. If the GP considers the patient in need of help, for example, a follow-up after the GP’s home visit or a blood test, he/she must first send an application to the purchaser’s office. Sometimes the help will come too late. A GP said:

We do ask home-based care to take a home visit, but this is a troublesome process, because we have to make an application to the purchaser office. They have to make a formal decision. We cannot always wait for that…In earlier days, we simply phoned them, but this doesn’t work any longer…The district units in home-based care have no budget of their own—they get their money according to what they have to do.

A nurse in home-based care thought that the incompatible financial platforms, GPs as private doctors and home-based services as publicly financed created barriers for practical collaboration:

There may be enquiries [from GPs] about following up patients in various ways. I, personally, mean that this should not be our responsibility. If so, public service personnel would be doing unpaid service for private practitioners, and that is a question….

GPs fall outside the hospital–municipality collaboration

The informants described several challenges to collaboration between hospital and municipal services. When a frail elderly person is transferred from hospital to municipal care, the practical preparation and arranging of the transfer are dependent on good collaboration between the nurses in the hospital and the nurses in municipal services. After the transfer, the patient’s GP is responsible for medical follow-up. Because the discharge process involves collaboration between hospital nurses and municipal care nurses, GPs are not included. A municipal nurse related an example: when a hospital-initiated project was established, aimed at better collaboration for chronic obstructive pulmonary disease (COPD) patients, the GPs had been put further on the sideline:

A very important effect of that project was that the GPs were pushed to the sideline. The dialogue was homecare nurses, patients, and the hospital. And so, in fact, we had a meeting with them (counterparts in hospital), telling them that we want them to continue with their recommendations on individual patients, but we want to relate to the GP when a patient is no longer in hospital. Our experience is that COPD patients usually have other diseases in addition, and so, parking the GPs is wrong.

Both GPs and their collaborative partners described examples of good collaboration between professionals in the hospital and professionals with the same background in the municipality, and this could be crucial for collaboration in the municipality. A GP, describing his non-collaborative relation to psychiatric nurses, had experienced that the strong collaborative lines between specialist psychiatric care and psychiatric nurses in municipal care made his GP role difficult:

It seems as if—and so they tell me—when they get a patient from a psychiatrist who tells them what to do, it is easier for them to follow-up a patient than if they get it from the GP. The psychiatric nurses in the municipality, they are more oriented towards hospital…yes, some of them have worked in psychiatric hospitals beforehand, having their relations.

Discussion

It is well known that collaboration between GPs and their colleagues in primary care is weak [1–4,7]. The results of this study show how the structural context within which primary health personnel work influences collaborative patterns between GPs and other important personnel in the municipality, and how where organizational, geographical and financial conditions affect this collaboration. Nonetheless, collaborative partners have experienced great individual differences between GPs in terms of to what extent they participate in inter-professional collaboration, indicating that individual conditions also influence collaboration.

Collaborative motives – the distribution of patient-related tasks, attitudes towards collaboration and GPs’ individual choices

GPs share their patients with many possible collaborative partners whom they must prioritize. Primarily, the GP informants in this study identified time pressures as the reason for the lack of collaboration with other providers in primary care, which is in accordance with other studies [1,2]. The GPs’ time pressure may be due to an increased workload, to new duties because of health reforms [7,9], to patients who are older and who have more health complaints and a greater need for healthcare, as well as to a population with increased expectations of healthcare services.

The task distribution between the various actors in primary care seems to affect their motivation for collaboration. Thus, collaborative needs and motives are not necessarily reciprocal. GPs see many patients daily, most of whom present with quite common problems that can be solved without inter-professional collaboration. However, the list patient system makes the GP responsible for all patients on the list, and patients in the need of coordinated services have the right to adequate treatment from the GP. GPs must prioritise their time between the many patients with simple problems and the few patients in need of coordinated services.

The GPs’ collaborative partners reported in interviews the striking individual differences between GPs in terms of willingness to enter into binding inter-professional collaboration, which is in accordance with other studies [25,26], engaging in collaboration for patients in this respect can be seen as an individual matter. GPs are self-employed on contract with the municipalities, and they assess and prioritize patients based on their interests and values. Hafting and Garløv’s study [26] on GPs’ collaboration with mental health services for children and adolescents showed considerable differences between GPs’ willingness to collaborate. This was interpreted partially as differences in the understanding of the GP’s role and partially as differences between generations, where younger GPs were more willing to collaborate than older GPs. Another interview-based study [27] showed that GPs differed greatly in their styles of communicating with patients with common mental disorders or emotional problems, and this was interpreted as differences in professional identities and roles. Some GPs in our study prioritize patients with simple problems instead of spending time on patients with needs that are more complex, and this may be due to different professional identities and roles. Competence may also influence GPs’ willingness to collaborate [28].

The fact that the GPs and their collaborative partners in our study often have different expectations of and needs for collaboration may be due to their different attitudes towards collaboration. Several studies address different professional groups’ attitudes towards other professional groups and towards collaboration with them. Physicians are often reported as dominating inter-professional collaboration [4,29–31]. Kharicha et al. [31] describe how GPs in collaboration with social workers often want them to re-structure the delivery of social care, with little change to their own working arrangements. Insight into each other’s work, a culture of mutual respect, the recognition of each other’s areas of expertise and competence and the free and open exchange of information are key elements of successful inter-professional collaboration. Physicians describe ambiguity in their new role as equal and democratic members of the team, and they seem insecure with this position. However, in Hansson et al.’s study, GPs considered collaboration to save time in the long run [2]. The GPs interviewed in our study expressed the same ambiguity concerning a need for collaboration, and at the same time, frustration because collaboration is time-consuming.

Service-internal organization

It seems that each of the services included in primary care is organized according to its own service-specific rationale, and this may suppress collaboration with other personnel in primary healthcare. A service-internal organizational rationale is grounded in which patients to serve, what kinds of services to deliver and a funding system designed to optimize the offering of the service in question. The list patient system gives the patients on the list well-defined rights and a more stable patient–doctor relationship. The system also gives each GP a well-defined population to serve, and it determines income possibilities. For the collaborative partners, one benefit is that each patient has a responsible doctor and therefore a defined potential collaborative partner. However, the patient recruitment mechanism among GPs – patients’ choice of doctors – is incongruent with that of other important services in local care, such as home-based nursing, physiotherapy and mental health. While GPs’ patients are recruited from all parts of a municipality, patients in home-based care services are recruited from a local district in the municipality. District-organised service units may reduce travelling distances for personnel working with people in their own homes and may reduce the actual number of different persons working with the same patients.

Many municipalities have bureaucratic routines, such as the purchaser–provider organisation of home-based services, and the results show how this hampers requests for collaboration initiatives from GPs. This model is set up to secure both the effective use of the municipal nursing capacity and prioritization among patients who need care [32]. The GPs’ listing system, the district orientation and the purchaser–provider system in home-based care are each designed carefully according to the different service-internal organising rationales to ensure the efficient fulfilment of their respective primary welfare goals. Seemingly, however, sensible service-internal rationales can turn into anti-collaborative organisational mechanisms when studied from a joint perspective.

Cross-level collaborative patterns between primary and secondary health services may block collaboration within the primary care system itself

Effective collaboration with specialist healthcare personnel is crucial to realising high-quality primary healthcare. GPs refer their patients to hospitals when specialized care is needed, and they follow up with their patients after hospital discharge. Municipal nurses preparing for the discharge of frail elderly persons from hospital are dependent on good information and joint planning with the hospital nurses. Mental health service nurses in the municipality depend on both professional support and continuous joint planning with higher-level services, as many patients have both a continuous and fluctuating need for help in the form of both primary and specialist care. However, the introduction of routines to secure good coordination and collaboration between specialised care and a municipal service may disturb and set aside collaborative patterns within primary healthcare itself. An example in our study is the collaborative set-up between hospital and municipal nurses in the COPD project, considered a necessary and important initiative to raise the quality of care available to a vulnerable patient group. However, GPs were not included in information exchanges and discussions, and the municipal nurses involved considered this a threat to the local collaborative climate. Other studies show similar results. Bambling [1] found that GPs were seen as resistant to working with allied mental health nursing staff, preferring (or insisting on) contact with psychiatrists. In de Stampa et al.’s [3] study, the geriatric team in the municipality was becoming accustomed to working with the geriatrician, and the GPs felt excluded.

Implications

As we have pointed out, motives for collaboration are not necessarily reciprocal between possible partners. Instead of focusing on collaboration as a superior goal, one ought to ask which patient-related tasks are dependent on collaboration. What routines are necessary to engage relevant personnel in collaboration when needed, which adjustments in reimbursement and funding systems are necessary and how may infrastructural measures ease collaborative work in primary healthcare?

Primary healthcare in Norway is not a very well-integrated organization, but rather a conglomerate of loosely coupled units with a multitude of in-built restraints on collaboration. Enhanced communication and improved inter-sectorial collaboration between GPs and collaborative partners in the municipality are needed, and there is no current mechanism to drive this development, as each sector has its own imperatives. Usually, there is no superior management responsible for coordinating primary care in the community. Primary healthcare services must be organized in such a way that every service’s needs in completing their principal work are ensured, while conditions for collaboration between providers in the different services are attended to as well. The Norwegian government points out that services are fragmented in part due to healthcare's organization in a silo structure, and it will lay the foundation for more team-based health and care services [9]. In the realisation of team-based care, it is necessary to consider the challenges related to teamwork [2]. The government argues that placing services at the same site is a good first step towards improving collaboration and coordination across today’s subservices. It seems, however, that colocation without integrated leadership and an adequate strategy, support and resources for building an integrated identity and practices does not improve collaboration [33,34].

Cross-level collaborative patterns, crucial for the fulfilment of the political health goals of primary care, must be designed and implemented to strengthen internal collaboration in primary care itself. For health authorities, responsible for the development and planning of primary healthcare, an important challenge is developing organizational and economical task-related measures that may contribute to a more integrated primary healthcare system, despite the centripetal forces of the mixed models. According to Ahgren [35], there is neither inter-organisational integration in the Norwegian healthcare system nor any opportunity for collaborative synergy. We argue that a common leadership for primary healthcare services is needed. The managers should be responsible for the providers’ collaboration and must ensure that individual differences between the GPs’ participation in collaboration are as few as possible. Inter-professional work must be a central focus in the planning of primary care services. A dedicated staff, sufficient resources, adequate time and proper meeting places are needed. Health system integration requires management practices that support relation building and information sharing across organizational and professional boundaries. Developing and implementing such measures presuppose thorough insight into the practical accomplishments of daily work and concrete bottom-up change work must take place by means of engaging the healthcare staff concerned [35].

Methodological considerations

Qualitative interviews seemed suitable for exploring the informants’ experiences and personal perceptions of collaboration between GPs and their colleagues in primary care. The informants told stories about collaboration and the lack thereof in their daily work, highlighting challenges in teamwork and inter-professional collaboration. Through the systematic collection and organisation of data, as well as the interpretation of the interview transcripts in the light of theories on organisation, professional identity and attitudes towards collaboration, we have developed knowledge on structural conditions facilitating and restraining collaboration in practice. To verify that the findings were founded in the data, we have aimed to describe the research process in such a way that the reader can follow the process.

Informants were recruited through the administrative managers of the health service sector in the municipalities. It is likely that the managers did not know anything about the practitioners’ collaboration patterns and preferences; consequently, this recruiting method hardly constituted bias. To obtain the best possible variation in the material, we chose to include GPs and four different professional groups of collaboration partners in both cities and rural municipalities. Our informants’ stories about challenges in collaboration correspond to other findings, and we argue that our perspectives on the structural conditions facilitating and restraining collaboration in primary care can be relevant for primary care in other cities and municipalities.

Ethics

The project’s protocol was submitted to the Regional Committee for Medical and Health Research Ethics but was found to fall outside the Committee’s mandate, as it did not aim for new knowledge on persons’ illness or health. Principles of confidentiality and anonymity have been applied in the conduct, reporting and storage of data in accordance with the Act on Processing of Personal Information.

Notes on contributors

Sissel Steihaug has worked as a general practitioner for many years. She holds a PhD in medicine and she has worked as a senior research scientist in SINTEF since 2004.

Bård Paulsen has a master’s in political science and he has worked within the field of integrated care and health policy since the 1980s. He worked as a senior research scientist in SINTEF and is now retired.

Line Melby holds a PhD in sociology. She is a senior research scientist in SINTEF. Her main expertise is within the field of collaboration across healthcare, integrated care, and qualitative methods.

Disclosure statement

The authors declare there is no conflict of interest.

Additional information

Funding

The study was supported by the Research Council of Norway, grant number 196426/V50.