- Email: [email protected]
Overt and covert barriers to the integration of primary and specialist mental health care
ARTICLE IN PRESS
Social Science & Medicine 61 (2005) 1785–1794 www.elsevier.com/locate/socscimed
Overt and covert barriers to the integration of primary and specialist mental health care Linda Gask National Primary Care Research and Development Centre, University of Manchester, Manchester M13 9DL, UK Available online 23 May 2005
Abstract This paper is concerned with the historical attempt over the last 20 years to improve integration between primary and specialist mental health care. Semi-structured interviews were carried out during the period December 2000–March 2001 with primary care workers, specialist medical and nursing staff, managers and other key informants in one large group model Health Maintenance Organization in the USA. Both overt (financial) and covert (attitudinal and conceptual) barriers to the integration of mental health and primary care were identified and the impact of these barriers on organizational development is discussed with reference to Activity Theory. The nature and quality of interprofessional conversation in an organization may be important mediating factors in addressing covert barriers to integration between primary and specialist mental health services. There may be insufficient actual contact between different groups of workers in primary and specialist care to enable these professionals to share ideas, challenge mutual assumptions and understand each others’ viewpoints about the nature of their work, the covert barriers to integration. Workers may differ in the conceptual models of mental health care they utilize, their views about access to services, and the amount of information they require. In order to integrate services effectively, these issues will require discussion. Financial pressures in the system may lead to failure on the part of management to sanction and encourage opportunities for interprofessional conversation and the geographical distance between places of work may also limit opportunities for contact. However, an alternative explanation might be that attitudinal and other covert barriers to integration effectively prevent, in the first place, the development of such a shared space in which these covert barriers might actually be addressed. r 2005 Elsevier Ltd. All rights reserved. Keywords: Primary care; Specialist care; Mental health; Integration; USA
Introduction The management of illness is becoming increasingly complex, and with it, the recognition of the need for improved co-ordination of care (Starfield, 1998). Integrated care has been defined (Donaldson, Yordy, Lohr, & Vanselow, 1996) as not simply ‘co-ordinated’ (ensuring the provision of a combination of health services and Tel.: +44 161 256 3015x220; fax: +44 161 256 1070.
E-mail address: [email protected].
information that meets a patient’s needs and also involving the connection between these services) but also ‘comprehensive’ (addressing any health problem at any given stage of a patient’s life cycle) and ‘continuous’ (care over time by a single individual or team of individuals and effective and timely communication). The research reported in this paper must be understood in the context of efforts over the last 20 years to improve integration between primary and specialist mental health care. Batterham and colleagues (Batterham et al., 2002) note the difference in the dictionary definitions of
0277-9536/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.socscimed.2005.03.038
ARTICLE IN PRESS 1786
L. Gask / Social Science & Medicine 61 (2005) 1785–1794
‘co-ordination’ and ‘integration’. The former is ‘the act of arranging according to a plan especially one drawing together a number of different events, organizations, people etc.’ and ‘working together smoothly in combination’ but the latter is perhaps the more challenging: ‘the act of bringing together parts into a whole’. Using concept mapping in groups of GPs, consumers and other practitioners to develop a model of integration Batterham et al. (2002) found that integration could be conceptualized as containing both elements of process and structure. They drew on Activity Theory (Engestrom, Miettinen, & Punamaki, 1999) to interpret their findings, and commented that there is ‘constant construction and re-negotiation within the activity system’ (Batterham et al., 2002). The health care organization on which this paper will focus has participated in professional discourses into the benefits of integrated mental healthcare over a decade and yet, as we shall see, it has still proved difficult to achieve lasting change on the ground. Activity Theory will be utilized in interpretation of data, which address both structural and process barriers to integration. Champions of integrated mental health care have looked ever optimistically towards the time when effective integration between primary care and mental health services will be achieved- or at least the perceived conditions for it to be achieved will be in place. Statements such as ‘the road to genuine collaboration is not an easy one, but there appears to be much to be to be gained from travelling it: lower costs, better patient care, and a more rewarding work environment’ (Editorial, 1993), are typical. Terminology is confusing in the literature with interchangeable and often undefined use of the terms: integration, collaboration and collaborative, shared care, linkage, co-ordinated care and partnership-working. For simplicity, I will use the term ‘integration’ but define other usage where appropriate. In this case, the authors perceive collaboration between psychosocial providers and biomedical providers as indicative of the need to integrate completely within the primary health care team, a view of care, which places them at the most radical pole of one dimension of mental health care integration. And yet, outside of successful demonstration projects across the world, only some of which have been subject to extensive evaluation (Craven & Bland, 2002), widespread and routine integration between primary care and mental health practitioners remains elusive. Effective integration of care would seem easier to achieve where primary care professionals assume key significance in the health care system through their role as gatekeepers to specialist referral or where registration with a primary care physician (PCP) is necessary as seen, for example, in the UK, Netherlands and Denmark. In the highly competitive but fragmented US healthcare sector, under pressure from increasing costs and the
growth of managed care, integrated healthcare delivery systems, where linkages exist between different levels of the system, have been developed. Such linkages vary from the administrative entity that contracts with a range of employers, insurers and providers to provide services to individuals for a negotiated fee (Preferred Provider Organizations) through to the Health Maintenance Organization (HMO), where, in return for reimbursement, practitioners agree to furnish a package of services as defined in a contract. It is the HMO, most commonly the Staff model (where the physicians work directly for the organization) and Group model (where the HMO contracts with a separate group of physicians) that the US healthcare system has been most successful (indeed more successful than in the NHS (Feacham, Sekhiri, & White, 2002)) in achieving integrated healthcare delivery including integrated delivery of mental healthcare. However, it has been difficult to achieve more widespread recognition of the need to change systems and practices, and to maintain change even when it has occurred (Department of Health and Human Services (DHHS), 2001).
The complex nature of integration From an economic perspective Perry (1989) defined integration as ‘the extent to which market exchange between firms within or across production stages of the value chain is supplanted by internal transfer of goods or services within the boundaries of a single firm.’ Simoens and Scott (1999) have extended our understanding of the complexity of integration as a concept applied to health care delivery systems. Firstly, they note, development of the concept of a health care ‘value chain’ poses numerous problems. ‘In health care, integration relates to the progression of the patient through production stages as determined by the episode of health care, with each stage defined by the organization that provides carey Patients may have multiple episodes of health care during a spell of illness and possibly multiple spells during a lifetime (in the case of, for instance, a chronic condition). For a given episode of health care, the fewer providers involved then the more integrated the service provision. Secondly, primary and specialist care may not be easy to define and there may be overlap between care settings, for example as in outreach clinics run by consultants in primary care. Thirdly, primary care cannot always be considered as an input to secondary care as patients discharged from hospital may continue their treatment in primary care, thus requiring concepts of backward and forward integration, and finally patients undergoing treatment may not progress linearly through the system. They conclude that the receipt of care is therefore more realistically modelled as a cyclical, continuing process, characterized by uncertainty’.
ARTICLE IN PRESS L. Gask / Social Science & Medicine 61 (2005) 1785–1794
It is essential for discussion of the barriers to integration in everyday practice, to recognize and acknowledge this inherent complexity (Plsek & Greenhalgh, 2001) which is often lost in the jargon terms of ‘seamless care’, ‘holistic approach’ or ‘building a team around the client’. Two widely accepted dimensional representations of integration are useful here, horizontal and vertical integration. Horizontal integration refers to the bringing together of professions, services and organizations that operate at similar levels in the care hierarchy. Vertical integration refers to the bringing together of different levels in the hierarchy of care. Rosenberg (1993) has further depicted four dimensions along which organizations might integrate: (1) governance structure; (2) decision making and policy; (3) administration and service delivery; and (4) goal identification and assessment. Peek and Heinrich (1995) have commented, ‘health care systems operate simultaneously in three worlds: the clinical, the operational and the financial worlds’. The resulting tension, Peek and Heinrich believe, dissipates human and financial energy and will potentially work against attempts to achieve changes such as integrate systems They predict that to be successful all action and design has to satisfy the demand of all three worlds.
Models of integration in mental health care delivery A number of writers have attempted to describe how mental health services and primary care can be linked. Some, but not all of these models, are applicable internationally. In the US, Coleman and Patrick (1976) described three models: (1) the primary care practitioners receive limited speciality mental health training; (2) speciality mental health practitioners consult with primary care practitioners on a case-by-case basis; and (3) mental health practitioners serve as part of the primary care team. Later authors, including Pincus (1987), have recognized the need to define more clearly ‘who’ is involved in the development of the relationship between primary and specialist care? ‘What’ is actually involved in terms of activities? And ‘when’ does it take place? To this could be added ‘where’ does this take place? Pincus describes six types of relationship of varying intensity which are echoed in the later comprehensive descriptions by Seaburn, Lorenz, Gunn, Gawinski, and Mauksch (1996) of models where collaboration is defined as ‘recognizing each party’s presence, expertise and special talentsynot a leaderless, amorphous, inefficient group process’. They describe five bands of the ‘collaborative spectrum’, which are: parallel delivery, informal consultation, formal consultation, co-provision of care and expansion of the health care network. However, they go on to define what they consider to be key ingredients for effective collabora-
1787
tion-mutual respect, a common purpose, clear communication and preferably, co-location (my italics). They also point out that the method of payment for services have powerful influences on the nature of collaboration. The interface between mental health and primary care in the USA poses particular problems rooted in its history, structure and context (Druss, 2002). Primary medical care as a specialized professional body developed largely as a reaction to the rapid growth in specialization seen during the mid-twentieth century but ‘family practice’ as a speciality within medicine arguably does still not command the same status as hospital based specialities. In the last two decades, there has been increasing recognition of the central role played by primary care doctors in managing common mental health problems, which has been a force towards integration because of the recognition that care is improved by structured collaboration (see below). Mechanic (1997) has identified six approaches to coordinating behavioural health with general medical care, which specifically reflect the development of managed care in the US during the last decade. These are: mainstreaming- the most common approach in HMOs where a defined comprehensive and physical benefit is provided on a capitated basis, the liaison psychiatry/ collaboration model (see below), independent carve-outs (where mental health care is provider by a separate behavioural healthcare provider), integrated carve-outs (where an HMO has its own mental health speciality service as in this case study), new practitioner models (specialist PCPs and nurse practitioners) and extended care models (integrated health and social care organizations). In the last decade, the rapid increase in the use of ‘carve-out’ organizations to provide behavioural health care has shifted the emphasis of care for mental health problems away from the primary care provider and towards non-physician mental health workers as the first point of contact for the patient (who is given information about how to contact his or her behavioural health provider directly by their health insurance plan). The impact of this trend has been a shift away from integration of care, as ‘carved-out’ mental health services have little economic incentive to offer or participate in integrated treatment because these plans cannot recoup cost-offsets. However, the presence of carve-outs in the wider healthcare arena can also work against the process of integration in an integrated HMO with ‘carved-in’ mental health care, by driving the quality standards agenda towards measures that favour carve-out organizations. The impact of the ‘carve-out’ revolution on the world of a group model HMO providing its own ‘in-house’ or ‘carved-in’ behavioural health care has been considerable. Quirk and his colleagues (Quirk et al., 2000) identify two main issues that might impact on their ability to deliver an
ARTICLE IN PRESS 1788
L. Gask / Social Science & Medicine 61 (2005) 1785–1794
integrated system: the rising costs of care, and the requirements of the NCQA (The National Committee for Quality Assurance, a major accrediting body). Two models of collaboration, which developed in the US have had a significant impact on professional discourse, if not on routine clinical practice. Wayne Katon and colleagues (Katon, Von Korff, Lin, Lipscomb, & Wagner, 1992) developed a model of care working with PCPs in Group Health Cooperative in the late 1980s which was based firmly in the hospital tradition of consultation-liaison psychiatry. A visiting psychiatrist provided both a consultation with the patient and advice to the physician (Katon et al., 1992). In a similar approach, known as Group Health Cooperative Integrated Model (Strosahl, 1998) the behavioural health consultant, who may be from any mental health discipline, does not ‘take over’ care but brings specialized knowledge to bear on problems that require additional expertise and may even engage in limited and strategic follow-up with selected patients. In all cases these activities are designed to support the primary care provider. The last decade has seen a trend towards the establishment of tightly structured models of care delivery, in line with the broader policy drive for ‘evidence-based practice’. The collaborative model developed by Katon and his colleagues (Katon et al., 1996) following disappointing outcomes with their earlier integrated model is much more highly structured and focuses on the care of a specific disorder, depression. This model, theoretically rooted in the Chronic Care Model (Wagner, Austin, & Von Korff, 1996) incorporates substantial re-organization of care including case management, protocols and case-registers, provision of self-care material to patients, training for PCPs, closemonitoring of patients by case-managers, and clearly stated protocols for ‘stepping up’ care to ensure that patients who are not making progress are referred for specialist mental health treatment. Thus far, this tightly defined and highly organized approach to integration has produced impressive outcomes in several randomized trials (see Gilbody, Whitty, Grimshaw, & Thomas, 2003 for a systematic review) but these models have not been widely implemented. The reason generally given for this, in the US setting, is the pressure to reduce costs (Druss, 2002). The more intensive care for mental health problems provided by the Collaborative model both increases treatment costs but also improves costeffectiveness for patients with major depression. A costoffset (in which analysis is confined to direct medical costs (i.e. the perspective of the health-insurer) and excludes the significant impact of mental health problems on employers, families and the social welfare system) was seen for specialty mental health costs but not medical care costs (Von Korff et al., 1998). But this has not dissuaded some services from seeking to
integrate (e.g. Dea, 2000) or dampened the enthusiasm for integration in policy circles (Department of Health and Human Services (DHHS), 2001). Indeed, as Goldman (1999) has suggested, perhaps discussion of costoffset is missing the point- ‘one rarely hears a defense of treatment coverage for infectious disease on the basis that it will reduce spending for cancer, or heart disease or ADM (mental health) treatment’. Potential solutions involving changes in the financial structure of organizations to promote integration of mental health and primary care in the US system have been suggested by Goldberg (1999) and Melek (1999). Models of collaboration described in US settings can also be recognized in the European, Canadian and Australian literature despite the different health care systems in which they have developed (Craven & Bland, 2002, Gask, Sibbald, & Creed, 1997; Meadows, 1998). In the UK, the greatest emphasis has been on the development of specialist out-reach clinics in primary care as a means of developing more effective vertical integration and the addition of mental health workers such as counsellors to the primary care team which can be viewed as horizontal integration. Although there has been considerable interest in development of consultation-liaison models where discussion takes place between professionals before referral (rather than the professional simply seeing the patient in the primary care clinic) this remains uncommon. Despite the apparent lack of ‘hard’ evidence from randomized controlled trials (Craven & Bland, 2002) the enthusiasm for the looser models of integration such as consultation-liaison and shifted out-patient clinics continues and there has been little progress in implementing more structured collaborative models of integration.
Context, study group and methods The study reported in this paper was undertaken as part of an exploration of the impact of quality improvement initiatives for two common disorders in primary care, depression and diabetes, in a large, group model HMO in the United States. The author was a visiting Harkness Fellow with no experience of working in the organization. Semi-structured interviews were carried out during the period December 2000–March 2001 with primary care workers, specialist medical and nursing staff, managers and other key informants. This paper focuses on barriers to the integration between mental health and primary care identified through analysis of data collected during this study. Forty-five interviews were conducted. Twenty-four PCPs were selected for interview on the basis of their performance on quality indicators for management of depression and diabetes. One-third of the PCPs interviewed performed in the top tertile on both of these
ARTICLE IN PRESS L. Gask / Social Science & Medicine 61 (2005) 1785–1794
indicators, one third in the top tertile in diabetes and the bottom in depression and one-third, vice versa. Mental health workers (13) were suggested for interview by the manager of the mental health service on the basis of working in the localities from which PCPs were selected and also to ensure a full range of different models of working with respect to primary care (based on same site, based on different sites, different sites with visiting mental health professional in primary care, different sites with no visiting professional). Other key ‘context’ informants (8) interviewed were an academic PCP, psychiatrist and psychologist with links to the organization, a local independent management consultant with expertise in the field, a primary care nurse and a primary care based social worker both working in the organization, and finally the senior managers of mental health and of primary care. All of these interviews (45) were carried out by the author, recorded and transcribed and form the basis of the data for this study. The data were organized and analysed with the aid of the WinMAX software programme (WinMAX, 1996), utilizing the constant comparative technique described by Glaser and Strauss (1967). In addition to the interview data, field notes recorded during and after visits to the 14 primary care clinics and 6 mental health departments where the interviews were based were also read during the process of data analysis. The data were analysed utilizing the theoretical framework of Activity Theory (Engestrom et al., 1999). The organization was understood to represent the activity system and the analysis explores the internal tensions and contradictions of the system which form ‘the motive force of change and development’ (Engestrom et al., 1999, p. 9). In the text, in order to protect the confidentiality of interviewees, all PCPs, including managers, are referred to by an abbreviation representing their profession (PCP), and mental health workers are not individually distinguished by professional group or management responsibility.
1789
health care policy and the way it’s paid for in America’. (mental health worker 12) Integrated systems must compete with carved-out behaviour health providers not only on the basis of costs but also on agreed quality criteria, which would seem to inherently disfavour integration: ‘If you are a mental health carve-out company you want to make sure that your people are seen quickly and you want to make sure that they go to the appropriate providerythat there’s some, you know appropriateness of care. So if the carve-out companies were doing thatyNCQA, a large accrediting organisation basically formalised those practices in a way of standards that became part of an accreditation process, it just became really clear that we needed to do that as well’ (mental health worker 13) The organization is required to meet quality criteria, which are essentially designed to favour the carved-out large behavioural health provider organizations, whose patients do not access speciality care via primary care. The structure of referral in this organization was therefore changed to promote direct access for the consumer by setting up a single, central referral bureau, in the context of new quality standards driven by carved-out providers. Speed of access and direct access are the measures of quality, which the customers, the employers, favour: ‘that was a marketing decision, you know, one stop shopping, one phone call’ (mental health worker 3).
Informants reflected the widely held concern over the cost of delivering health care in the US context:
However, the system is no longer set to favour referral from primary care or even communication with primary care. The potential gain here is that employers will feel satisfied that people with mental health difficulties will be seen rapidly and (hopefully) be restored to full productivity quickly. The downside is that the advocacy role played by the PCP in ensuring that patients with the greatest need are seen earlier and by the most appropriate professional is devalued and the needs of patients with somatic presentation of emotional problems, who are unlikely to directly approach mental health, will not be met. A frequently expressed belief has been that integrating mental health care in primary care would ‘pay for itself’ because of the ‘cost-offset’ effect. Some physicians clearly believed it was possible to achieve savings through improved treatment and decreased utilization of other medical service, but others did not and therefore could not support the deploying of personnel to primary care:
‘With the economic crunch it’s a very tough business to be in and it’s hard to imagine things changing in the next ten years without some significant changes in
‘It’s costingya lot of money not to deal with all these people who are depressed so we want to do a better job, because it would be better for the patient
Results Two major themes arising from the data, concerning evidence for overt and covert barriers to integration, will be described with examples taken from interview texts to illustrate key concepts. Overt barriers to integration: contextual and structural factors
ARTICLE IN PRESS 1790
L. Gask / Social Science & Medicine 61 (2005) 1785–1794
care and it would save us money. We don’t want these people in here having X-rays of this and that and somatizing all the timeywe’d rather just deal with the depression and get them better and then they’ll feel better and that’ll be better for them and it’s certainly a lot cheaper for us’. (PCP 21) ‘If you put a consultant in the primary clinic and look at sort of the total number of people seen by specialists it will go up. Which is probably good but it’s not as if you can say that we’re putting someone in the primary care clinic and it really won’t cost us any more money because all they do is see the same people earlier on. There’ll be some y as we say new businessyright now I can’t with a clear conscience, push to divert more and more resources to that’. (mental health worker 4) Clinics where behavioural health and primary care staff were permanently physically co-located were in the minority, although this had historically been more common, before the development of professional ‘departments’. In order to be able to regularly communicate with each other, people need to have opportunities for contact. The nature of professional development within an organization could however make this difficult: ‘Mental health providers, as in any speciality, need to circulate with their own and discuss cases and share new information and so there may have been a greater need for them to collaborate more with each other rather than being ‘isolated’ in a primary care setting’. (PCP 25) This ‘critical mass’ concept of professional power and influence which can be seen to provide ‘buffering’ (Thompson, 1967) against the effects of change within the organization is also played out in UK mental health politics, in the placement or otherwise of psychologists and psychiatrists in community mental health teams rather than large central departments which may be perceived as more attractive to recruiting staff. Yet this leaves no opportunity for professional interaction: ‘We don’t have a relationship with the mental health service that we know who we’re dealing with. You know we’ve never met any of the psychiatrists or the mental health specialists. We don’t even know what they look like’. (PCP 13) There was very little opportunity for interaction between mental health and primary care staff on a day-to-day basis: Interviewer: What would you like to see happeningideally how would you like it to operate? Mental Health Worker: I think it would be nice to be co-located with the mental health staff so that it’s a
lot easier to pop down to their office and vice versa. And I think I’d like to see some regular scheduled time with them. Interviewer: That simply doesn’t happen at all? Mental Health Worker: It doesn’t happen- I think nobody’s got the time to do that. (mental health worker 2). Covert barriers: attitudes and conceptual models Lack of contact and opportunities for interprofessional communication makes it difficult for professionals to challenge each others’ attitudes- to each other and to the nature of the task of dealing with mental health problems. There was a sense in the primary care doctors of confusion and bewilderment when faced with difficult mental health problems for which they could not easily access support from specialist care. One mental health worker, who had worked closely with primary care in the organization, described the attitude of the PCPs as one of ‘learned helplessness’: ‘They try for help for a period of time and if they can’t get it, and they can’t get it the way they need it, which is immediately, they give up’ (mental health worker 3) The way in which mental health professionals work, eschewing interruption and maintaining the psychoanalytic tradition of the ‘50 minute hour’ even today, militates against fast access to specialists: ‘We had a therapist here that was on call one night and a PA who gets more awards than anyone else in this clinicyknocked on her door and she said go away don’t bother me I’m with a patient. We can’t do that. I think you just poison your relationships’ (mental health worker 1) ‘It requires I think a kind of flexibility that a lot of practitioners in our field don’t have. I think many people want to sit in their office and see people on the hour and go home’ (mental health worker 3) Two conceptual differences between primary care and specialist mental health care in how they view their roles in helping people with problems were identified. First, there was tension between the ‘medical’ model of distress shared by most of the PCPs (though not all) and the psychiatrists, particularly when it came to managing people with moderately severe mental health problems and the ‘personal responsibility model’ favoured by nonmedical therapists: ‘‘The medical staff say ‘I’m the doctor, I’m calling and I want this person seen, see him, you know the therapists take more of a stance that it’s their (the patient’s) responsibility.. If they want it, they need to want to come’ (mental health worker 13)
ARTICLE IN PRESS L. Gask / Social Science & Medicine 61 (2005) 1785–1794
This tension bubbled under the surface not just between primary care and mental health but also between the mental health professionals—the psychiatrists and master’s level therapists (usually with a social work background). Second, was the mutual misunderstanding about the nature of primary and specialist care. Primary care has little time to pay attention to detail, because, unlike specialist care, it has highly permeable boundaries and easy access for patients: ‘Ask a psychiatrist if they’re going to have same day accessyof course they’re not that’s not going to workybut we’re taking care of all-comers wherever they areySomeone comes in having dizzy spells and lightheadedness, and this horrible back pain, and also ‘my marriage is breaking up and I’m depressed and I’m seeing spots and my left eye went blind yesterday for 15 min, does that mean anything doctor?’’ (PCP 1) The expertise of the PCP is in managing the breadth of problems. He or she is a ‘specialist generalist’. A mental health worker expressed exasperation that his special expertise in understanding human relationships in depth was not valued by primary care staff: ‘When you’re around people who are talking about the human condition and its psychological implications and y your conversations are in detail about the intricacies of people’s livesy then you walk into a primary care setting where that kind of information really isn’t valued very much and they really don’t want to know it, they haven’t got the time to listen to it’. (mental health worker 5). A further complication is the nature of specialist training in the US, which prepares psychiatrists for a health system that outside of the HMOs, encourages them to compete with PCPs for the management of mental health problems which in gatekeeper controlled systems would be managed by primary care. This may lead to negative views about the capability of primary care to manage mental health problems and also result in an overload of the mental health services because of a failure to negotiate discharge back to primary care: ‘Turnover is not goodyit seems frequently once the patients come to us they’re ours for everyI think the patients like it here better. I mean we spend more time with them. I think we have more accessibility. Frequently they’ll see a physician’s assistant or someone else that will work the family doctor’s office and you know they’ll say, well can we say here?yI hate to say no you’ve got to go back when they don’t want to so it’s one of the ongoing issues that we have’. (mental health worker 2)
1791
Discussion: barriers to organizational change The integration ‘project’ has been on the professional and management agenda in this organization for many years. Progress has been intermittent but the ‘project’ still survives at a ‘symbolic’ level. Some mental health workers continue to spend time providing both off and on-site consultation to primary care workers despite the lack of central support for this activity. This is not surprising. Batterham et al. (2002), with reference to Activity Theory (Engestrom et al., 1999), note that when a new way of working is introduced into a system: Tasks are reassigned, and redivided, rules are set and reinterpreted. Beyond this contradictions may sometimes occur between different parts of the system and other systems. This introduces disequilibrium and change into the activity system. In all these different circumstances co-ordination between different subjects and their versions of the object must be achieved to ensure continuous operation. In this system, continued liaison working in primary care in several sites was tolerated by the system even though it ran counter to current management goals. It seemed to serve an almost ‘symbolic’ purpose (mental health worker 13), thus maintaining a vision of a way of working, which many viewed as the unattainable or impractical ideal. Progressing further organizational change in the present financial climate was perceived by the champions of integration to be difficult. This was because, in their view, furthering the integration project required senior management to set integration as a clear priority. There is no doubt that leadership of the organization must be ‘signed up’ to the task if effective integration across the interface is to take place (Kvamme, Olesen, & Samuelsson, 2001). Financial reasons were cited for the failure to maintain an integrated model of care delivery or to implement the collaborative care model. The importance of addressing not only structure but also process when attempting to achieve interprofessional integration in healthcare has been insufficiently addressed (Burns, 1999; Iles & Sutherland, 2001), a point echoed by one of the context informants: ‘we really have to deal with the process first and the financing second and I think that in most systems particularly in the US, we’ve dealt with the financing first and hoped that it would solve the problems’. (Independent healthcare consultant) If change is to occur within an organization it would seem to be essential to acknowledge the covert barriers to change thus allowing them to become overt. It is difficult to see how this could happen without space and time for interprofessional communication within an
ARTICLE IN PRESS 1792
L. Gask / Social Science & Medicine 61 (2005) 1785–1794
organization. There are at least two possible reasons for this. The overt barrier to effective communication would be the financial pressures in the organization leading to failure on the part of management to sanction and encourage opportunities for interaction. However, an alternative explanation might be that attitudinal and other covert barriers to integration effectively prevent the development of such a shared space in which attitudinal barriers might be addressed. The exact part played by each might only be determined by further research.
Conclusion: barriers, conversations and change In this organization, which has been attempting for many years to address integrated system of mental health care across the primary-specialist interface, islets of integration activity exist with an apparent ‘symbolic’ purpose but without the perceived support of senior levels of the organization for further development. Economic factors are cited as the major reason for this hiatus, yet it is apparent that, within other similar organizations (see for example Dea, 2000) the enthusiasm for integration apparently continues unabated. I use the term advisedly, because experience from this study would suggest that it is difficult to ascertain the success or otherwise of such a project without interviewing workers in the organization ‘on the ground’ rather than simply reviewing the published literature. The major limitation of this research is that coding of transcripts was done by one person. Emerging themes were cross-checked with another researcher (MEP) who cross-read a sample of the transcripts. However, the data were only collected from one organization so cannot be generalized to other organizations. Nevertheless, the results are supported by theory and may shed some light on the reasons why integration is difficult to achieve across organizational interfaces in healthcare settings. A project cannot survive without senior management support. If managers and professionals are going to become engaged in an enterprise, conceptual differences and attitudinal barriers need to be overtly addressed. Senior managers in the organization are themselves primary care and mental health professionals who carry with them the mental models they developed at the frontline. Organizations that seek to challenge covert barriers to change, and which aspire to the vision of the ‘learning organization’ (Senge, 1990, need to create to time and space for a ‘conversation’ to take place—both between managers and workers and between professional interest groups (Peek & Heinrich, 2000). This particularly applies to the key task of ‘developing shared mental models’: the discipline of working with mental models allowing individuals to unearth the assumptions and general-
izations that influence their understanding of the world and shape how action is taken; and ‘building shared vision’ or unearthing ‘shared pictures of the future’ that foster genuine commitment and enrolment rather than simply compliance. Conversation as a medium for change is not a new concept to mental health services, forming as it does the central tenet of psychological therapies. But the most effective conversational therapies appear to be those with a clear structure in which the participants develop mutual trust and shared understanding (Frank & Frank, 1993). This conversation, may be both informal (ad hoc between professionals as they meet and work together) or formal (in meetings and ‘workshops’ (Fitzsimmons & White, 1997)). An activity system is always heterogeneous, and, from the perspective of the author, a mental health professional, the medium of constructive negotiation that was missing in this organization was the ‘conversation’ or medium of effective communication, an observation which is conceptually consistent with Activity Theory. Davydov (1999), examining the relationship between activity and communication, concludes that: ‘collective and individual activity is realized in the form of material and spiritual social relationships. Communication is a processual expression of these relationships,’ (p. 47). Such communication as it occurs at the micro level between professional groups in an organization may be effective in producing change at the macro level. But the very attitudinal and conceptual barriers that it can be successful in addressing will also simultaneously seek to prevent it taking place. In this organization, the absence of conceptual, geographical and temporal space in which to locate such communication seemed to contribute to the failure to challenge covert boundaries. The nature and quality of any interprofessional conversation that takes place may be important mediating factor in addressing covert barriers to integration between primary and specialist mental health services. Further research might address the respective parts played by overt and covert barriers over the passage of time in an organization, and most particularly the impact of these barriers on the effectiveness of organizational interventions designed to challenge attitudes and behaviour of professional staff (for example quality improvement interventions). The impact of such interventions should address not only organizational level ‘measures’, but also seek to understand the views and day-to-day experiences of those who work in the organization and examine what the impact is on outcomes for patients.
Acknowledgements This study was carried out during a Harkness Fellowship in Health Care Policy with the support of
ARTICLE IN PRESS L. Gask / Social Science & Medicine 61 (2005) 1785–1794
the Commonwealth Fund of New York. With thanks to all the professionals who kindly agreed to be interviewed, Adele Clark for transcribing the interviews, Mary Ellen Purkis for cross-reading the transcripts and Anne Rogers and Carl May for their comments on earlier drafts of the manuscript.
References Batterham, R., Southern, N., Appleby, N., Elsworth, G., Fabris, S., Dunt, D., & Yound, D. (2002). Construction of a GP integration model. Social Science & Medicine, 54, 1225–1241. Burns, L. R. (1999). Polarity management: The key challenge for integrated healthcare systems. Journal of Healthcare Management, 44, 14–31. Coleman, J. V., & Patrick, D. L. (1976). Integrating mental health services into primary medical care. Medical Care, 14(8), 654–661. Craven, M. A., & Bland. R. (Eds.). (2002). Shared Mental Health Care: A bibliography and overview. Canadian Journal of Psychiatry, 47(Suppl. 1), 1–103S. Davydov, V. V. (1999). The content and unsolved problems of activity theory. In Y. Engestrom, R. Miettinen, & R.-L. Punamaki (Eds.), Perspectives on activity theory (Learning in doing). Cambridge: Cambridge University Press. Dea, R. A. (2000). The integration of primary care and behavioural healthcare in Northern California KaiserPermanente. Psychiatric Quarterly, 71(1), 17–29. Department of Health and Human Services (DHHS). (2001). Report of a Surgeon General’s working meeting on the integration of mental health services and primary care; 2000 Nov 30–Dec 1; Atlanta, Georgia. Rockville, MD: Office of the Surgeon General. Donaldson, M., Yordy, K., Lohr, K., & Vanselow, N. (Eds.). (1996). Primary care: America’s Health in a New Era. Institute of Medicine. Washington, DC: National Academy Press. Druss, B. G. (2002). The mental health/primary care interface in the United States: History, structure and context. General Hospital Psychiatry, 24(4), 197–202. Editorial. (1993). The generic health care team. Family Systems Medicine, 11(2), 119–120. Engestrom, Y., Miettinen, R., & Punamaki, R.-L. (1999). Perspectives on activity theory (Learning in doing). Cambridge: Cambridge University Press. Feacham, R. G. A., Sekhiri, N. K., & White, K. L. (2002). Getting more for their dollar: A comparison of the NHS with California’s Kaiser Permanente. British Medical Journal, 324, 135–143. Fitzsimmons, P., & White, P. (1997). Crossing boundaries: Communication between professional groups. Journal of Managed Medicine, 11, 96–101. Frank, J., & Frank, J. (1993). Persuasion and Healing: A comparative study of psychotherapy. Baltimore: Johns Hopkins University Press. Gask, L., Sibbald, B., & Creed, F. (1997). Evaluating models of working at the interface between mental health services and primary care. British Journal of Psychiatry, 170, 6–11.
1793
Gilbody, S., Whitty, P., Grimshaw, J., & Thomas, R. (2003). Education and organization interventions to improve the management of depression in primary care: A systematic review. Journal of the American Medical Association, 289, 145–151. Glaser, B. G., & Strauss, A. (1967). The discovery of grounded theory. Chicago: Aldine. Goldberg, R. J. (1999). Financial incentives influencing the integration of mental health and primary care. Psychiatric Services, 50(8), 1071–1075. Goldman, H. (1999). Justify mental health care costs. Health Affairs, 18(2), 94–95. Iles, V. & Sutherland, K. (2001). Organisational Change: A review for health care managers, professionals and researchers. National Co-ordinating Centre for NHS Services Delivery and Organisation R &D: London School of Hygiene and Tropical Medicine. Katon, W., Robinson, P., Von Korff, M., Lin, E., Bush, T., Ludman, E., Simon, G., & Walker, E. (1996). A multifaceted intervention to improve treatment of depression in primary care. Archives of General Psychiatry, 53, 924–932. Katon, W., Von Korff, M., Lin, E., Lipscomb, P., & Wagner, E. (1992). A randomised controlled trial of psychiatric consultation with distressed high utilizers. General Hospital Psychiatry, 14, 86–98. Kvamme, O. J., Olesen, F., & Samuelsson, M. (2001). Improving the interface between primary and secondary care: A statement from the European Working Party on Quality in Family Practice (EQuiP). Quality in Health Care, 10, 33–39. Meadows, G. N. (1998). Establishing a collaborative service model for primary mental health care. Medical Journal of Australia, 168, 162–165. Mechanic, D. (1997). Approaches for coordinating primary and specialty care for persons with mental illness. General Hospital Psychiatry, 19, 395–402. Melek, S. P. (1999). Financial, risk and structural issues related to the integration of behavioural healthcare in primary care settings under managed care. Milliman and Robertson Inc. Peek, C. J., & Heinrich, R. (1995). Building a collaborative healthcare organisation: From idea to invention to innovation. Family Systems Medicine, 13, 327–342. Peek, C. J., & Heinrich, R. (2000). Integrating behavioural health and primary care. In M. E. Maruish (Ed.), Handbook of psychological assessment in primary care settings. London: Erlbaum & Associates. Perry, M. K. (1989). Vertical Integration: Determination and effects. In R. Schmalensee, & R. Willig (Eds.), Handbook of industrial organisation, Vol. 1 (pp. 185–255). Amsterdam: North-Holland. Pincus, H. A. (1987). Patient-orientated models for linking primary care and mental health care. General Hospital Psychiatry, 9, 95–101. Plsek, P. E., & Greenhalgh, T. (2001). The importance of complexity in health care. British Medical Journal, 323, 625–628. Quirk, M. P., Simon, G., Todd, J., Horst, T., Crosier, M., Ekorenrud, B., Goepfert, R., Baker, N., Steinfeld, B., Rodenberg, M., & Strosahl, K. (2000). A look to the
ARTICLE IN PRESS 1794
L. Gask / Social Science & Medicine 61 (2005) 1785–1794
past, directions for the future. Psychiatric Quarterly, 71(1), 79–94. Rosenberg, S. (1993). Rural models for integrating primary care and mental health services. Administration and Policy in Mental Health, 25(3), 287–308 (Cited in Bird, D. C., Lambert, D., Hartley, D., Beeson, P. G., & Coburn, A. F. (1998). Seaburn, D. A., Lorenz, A., Gunn, W. B., Jr., Gawinski, B. A., & Mauksch, L. B. (1996). Model of collaboration: A guide for mental health professionals working with health care practitioners. New York: Basic Books. Senge, P. (1990). The fifth discipline: The art and practice of the learning organisation. London: Doubleday/Century Business. Simoens, S., & Scott, A. (1999). What determines integration in primary care? Theory and evidence. HERO Discussion Paper 04/99. University of Aberdeen: Health Economics Research Unit.
Starfield, B. (1998). Primary care: Balancing health needs services and technology. Oxford: Oxford University Press. Strosahl, K. (1998). Integrating behavioural health and primary care: The primary mental health care model. In A. Blount (Ed.), Integrated primary care—The future of medical and mental health collaboration. New York: Norton. Thompson, J. D. (1967). Organizations in action. New York: McGraw-Hill. Von Korff, M., Katon, W., Bush, T., Lin, E. B., Simon, G. E., Saunders, K., Ludman, E., Walker, E., & Unutzer, J. (1998). Treatment costs, cost-offset, and cost-effectiveness of collaborative management of depression. Psychosomatic Medicine, 60, 143–149. Wagner, E., Austin, B., & Von Korff, M. (1996). Organising care for patients with chronic illness. The Millbank Quarterly, 74, 511–543. WinMAX. (1996). London: Sage Publications.
Social Science & Medicine 61 (2005) 1785–1794 www.elsevier.com/locate/socscimed
Overt and covert barriers to the integration of primary and specialist mental health care Linda Gask National Primary Care Research and Development Centre, University of Manchester, Manchester M13 9DL, UK Available online 23 May 2005
Abstract This paper is concerned with the historical attempt over the last 20 years to improve integration between primary and specialist mental health care. Semi-structured interviews were carried out during the period December 2000–March 2001 with primary care workers, specialist medical and nursing staff, managers and other key informants in one large group model Health Maintenance Organization in the USA. Both overt (financial) and covert (attitudinal and conceptual) barriers to the integration of mental health and primary care were identified and the impact of these barriers on organizational development is discussed with reference to Activity Theory. The nature and quality of interprofessional conversation in an organization may be important mediating factors in addressing covert barriers to integration between primary and specialist mental health services. There may be insufficient actual contact between different groups of workers in primary and specialist care to enable these professionals to share ideas, challenge mutual assumptions and understand each others’ viewpoints about the nature of their work, the covert barriers to integration. Workers may differ in the conceptual models of mental health care they utilize, their views about access to services, and the amount of information they require. In order to integrate services effectively, these issues will require discussion. Financial pressures in the system may lead to failure on the part of management to sanction and encourage opportunities for interprofessional conversation and the geographical distance between places of work may also limit opportunities for contact. However, an alternative explanation might be that attitudinal and other covert barriers to integration effectively prevent, in the first place, the development of such a shared space in which these covert barriers might actually be addressed. r 2005 Elsevier Ltd. All rights reserved. Keywords: Primary care; Specialist care; Mental health; Integration; USA
Introduction The management of illness is becoming increasingly complex, and with it, the recognition of the need for improved co-ordination of care (Starfield, 1998). Integrated care has been defined (Donaldson, Yordy, Lohr, & Vanselow, 1996) as not simply ‘co-ordinated’ (ensuring the provision of a combination of health services and Tel.: +44 161 256 3015x220; fax: +44 161 256 1070.
E-mail address: [email protected].
information that meets a patient’s needs and also involving the connection between these services) but also ‘comprehensive’ (addressing any health problem at any given stage of a patient’s life cycle) and ‘continuous’ (care over time by a single individual or team of individuals and effective and timely communication). The research reported in this paper must be understood in the context of efforts over the last 20 years to improve integration between primary and specialist mental health care. Batterham and colleagues (Batterham et al., 2002) note the difference in the dictionary definitions of
0277-9536/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.socscimed.2005.03.038
ARTICLE IN PRESS 1786
L. Gask / Social Science & Medicine 61 (2005) 1785–1794
‘co-ordination’ and ‘integration’. The former is ‘the act of arranging according to a plan especially one drawing together a number of different events, organizations, people etc.’ and ‘working together smoothly in combination’ but the latter is perhaps the more challenging: ‘the act of bringing together parts into a whole’. Using concept mapping in groups of GPs, consumers and other practitioners to develop a model of integration Batterham et al. (2002) found that integration could be conceptualized as containing both elements of process and structure. They drew on Activity Theory (Engestrom, Miettinen, & Punamaki, 1999) to interpret their findings, and commented that there is ‘constant construction and re-negotiation within the activity system’ (Batterham et al., 2002). The health care organization on which this paper will focus has participated in professional discourses into the benefits of integrated mental healthcare over a decade and yet, as we shall see, it has still proved difficult to achieve lasting change on the ground. Activity Theory will be utilized in interpretation of data, which address both structural and process barriers to integration. Champions of integrated mental health care have looked ever optimistically towards the time when effective integration between primary care and mental health services will be achieved- or at least the perceived conditions for it to be achieved will be in place. Statements such as ‘the road to genuine collaboration is not an easy one, but there appears to be much to be to be gained from travelling it: lower costs, better patient care, and a more rewarding work environment’ (Editorial, 1993), are typical. Terminology is confusing in the literature with interchangeable and often undefined use of the terms: integration, collaboration and collaborative, shared care, linkage, co-ordinated care and partnership-working. For simplicity, I will use the term ‘integration’ but define other usage where appropriate. In this case, the authors perceive collaboration between psychosocial providers and biomedical providers as indicative of the need to integrate completely within the primary health care team, a view of care, which places them at the most radical pole of one dimension of mental health care integration. And yet, outside of successful demonstration projects across the world, only some of which have been subject to extensive evaluation (Craven & Bland, 2002), widespread and routine integration between primary care and mental health practitioners remains elusive. Effective integration of care would seem easier to achieve where primary care professionals assume key significance in the health care system through their role as gatekeepers to specialist referral or where registration with a primary care physician (PCP) is necessary as seen, for example, in the UK, Netherlands and Denmark. In the highly competitive but fragmented US healthcare sector, under pressure from increasing costs and the
growth of managed care, integrated healthcare delivery systems, where linkages exist between different levels of the system, have been developed. Such linkages vary from the administrative entity that contracts with a range of employers, insurers and providers to provide services to individuals for a negotiated fee (Preferred Provider Organizations) through to the Health Maintenance Organization (HMO), where, in return for reimbursement, practitioners agree to furnish a package of services as defined in a contract. It is the HMO, most commonly the Staff model (where the physicians work directly for the organization) and Group model (where the HMO contracts with a separate group of physicians) that the US healthcare system has been most successful (indeed more successful than in the NHS (Feacham, Sekhiri, & White, 2002)) in achieving integrated healthcare delivery including integrated delivery of mental healthcare. However, it has been difficult to achieve more widespread recognition of the need to change systems and practices, and to maintain change even when it has occurred (Department of Health and Human Services (DHHS), 2001).
The complex nature of integration From an economic perspective Perry (1989) defined integration as ‘the extent to which market exchange between firms within or across production stages of the value chain is supplanted by internal transfer of goods or services within the boundaries of a single firm.’ Simoens and Scott (1999) have extended our understanding of the complexity of integration as a concept applied to health care delivery systems. Firstly, they note, development of the concept of a health care ‘value chain’ poses numerous problems. ‘In health care, integration relates to the progression of the patient through production stages as determined by the episode of health care, with each stage defined by the organization that provides carey Patients may have multiple episodes of health care during a spell of illness and possibly multiple spells during a lifetime (in the case of, for instance, a chronic condition). For a given episode of health care, the fewer providers involved then the more integrated the service provision. Secondly, primary and specialist care may not be easy to define and there may be overlap between care settings, for example as in outreach clinics run by consultants in primary care. Thirdly, primary care cannot always be considered as an input to secondary care as patients discharged from hospital may continue their treatment in primary care, thus requiring concepts of backward and forward integration, and finally patients undergoing treatment may not progress linearly through the system. They conclude that the receipt of care is therefore more realistically modelled as a cyclical, continuing process, characterized by uncertainty’.
ARTICLE IN PRESS L. Gask / Social Science & Medicine 61 (2005) 1785–1794
It is essential for discussion of the barriers to integration in everyday practice, to recognize and acknowledge this inherent complexity (Plsek & Greenhalgh, 2001) which is often lost in the jargon terms of ‘seamless care’, ‘holistic approach’ or ‘building a team around the client’. Two widely accepted dimensional representations of integration are useful here, horizontal and vertical integration. Horizontal integration refers to the bringing together of professions, services and organizations that operate at similar levels in the care hierarchy. Vertical integration refers to the bringing together of different levels in the hierarchy of care. Rosenberg (1993) has further depicted four dimensions along which organizations might integrate: (1) governance structure; (2) decision making and policy; (3) administration and service delivery; and (4) goal identification and assessment. Peek and Heinrich (1995) have commented, ‘health care systems operate simultaneously in three worlds: the clinical, the operational and the financial worlds’. The resulting tension, Peek and Heinrich believe, dissipates human and financial energy and will potentially work against attempts to achieve changes such as integrate systems They predict that to be successful all action and design has to satisfy the demand of all three worlds.
Models of integration in mental health care delivery A number of writers have attempted to describe how mental health services and primary care can be linked. Some, but not all of these models, are applicable internationally. In the US, Coleman and Patrick (1976) described three models: (1) the primary care practitioners receive limited speciality mental health training; (2) speciality mental health practitioners consult with primary care practitioners on a case-by-case basis; and (3) mental health practitioners serve as part of the primary care team. Later authors, including Pincus (1987), have recognized the need to define more clearly ‘who’ is involved in the development of the relationship between primary and specialist care? ‘What’ is actually involved in terms of activities? And ‘when’ does it take place? To this could be added ‘where’ does this take place? Pincus describes six types of relationship of varying intensity which are echoed in the later comprehensive descriptions by Seaburn, Lorenz, Gunn, Gawinski, and Mauksch (1996) of models where collaboration is defined as ‘recognizing each party’s presence, expertise and special talentsynot a leaderless, amorphous, inefficient group process’. They describe five bands of the ‘collaborative spectrum’, which are: parallel delivery, informal consultation, formal consultation, co-provision of care and expansion of the health care network. However, they go on to define what they consider to be key ingredients for effective collabora-
1787
tion-mutual respect, a common purpose, clear communication and preferably, co-location (my italics). They also point out that the method of payment for services have powerful influences on the nature of collaboration. The interface between mental health and primary care in the USA poses particular problems rooted in its history, structure and context (Druss, 2002). Primary medical care as a specialized professional body developed largely as a reaction to the rapid growth in specialization seen during the mid-twentieth century but ‘family practice’ as a speciality within medicine arguably does still not command the same status as hospital based specialities. In the last two decades, there has been increasing recognition of the central role played by primary care doctors in managing common mental health problems, which has been a force towards integration because of the recognition that care is improved by structured collaboration (see below). Mechanic (1997) has identified six approaches to coordinating behavioural health with general medical care, which specifically reflect the development of managed care in the US during the last decade. These are: mainstreaming- the most common approach in HMOs where a defined comprehensive and physical benefit is provided on a capitated basis, the liaison psychiatry/ collaboration model (see below), independent carve-outs (where mental health care is provider by a separate behavioural healthcare provider), integrated carve-outs (where an HMO has its own mental health speciality service as in this case study), new practitioner models (specialist PCPs and nurse practitioners) and extended care models (integrated health and social care organizations). In the last decade, the rapid increase in the use of ‘carve-out’ organizations to provide behavioural health care has shifted the emphasis of care for mental health problems away from the primary care provider and towards non-physician mental health workers as the first point of contact for the patient (who is given information about how to contact his or her behavioural health provider directly by their health insurance plan). The impact of this trend has been a shift away from integration of care, as ‘carved-out’ mental health services have little economic incentive to offer or participate in integrated treatment because these plans cannot recoup cost-offsets. However, the presence of carve-outs in the wider healthcare arena can also work against the process of integration in an integrated HMO with ‘carved-in’ mental health care, by driving the quality standards agenda towards measures that favour carve-out organizations. The impact of the ‘carve-out’ revolution on the world of a group model HMO providing its own ‘in-house’ or ‘carved-in’ behavioural health care has been considerable. Quirk and his colleagues (Quirk et al., 2000) identify two main issues that might impact on their ability to deliver an
ARTICLE IN PRESS 1788
L. Gask / Social Science & Medicine 61 (2005) 1785–1794
integrated system: the rising costs of care, and the requirements of the NCQA (The National Committee for Quality Assurance, a major accrediting body). Two models of collaboration, which developed in the US have had a significant impact on professional discourse, if not on routine clinical practice. Wayne Katon and colleagues (Katon, Von Korff, Lin, Lipscomb, & Wagner, 1992) developed a model of care working with PCPs in Group Health Cooperative in the late 1980s which was based firmly in the hospital tradition of consultation-liaison psychiatry. A visiting psychiatrist provided both a consultation with the patient and advice to the physician (Katon et al., 1992). In a similar approach, known as Group Health Cooperative Integrated Model (Strosahl, 1998) the behavioural health consultant, who may be from any mental health discipline, does not ‘take over’ care but brings specialized knowledge to bear on problems that require additional expertise and may even engage in limited and strategic follow-up with selected patients. In all cases these activities are designed to support the primary care provider. The last decade has seen a trend towards the establishment of tightly structured models of care delivery, in line with the broader policy drive for ‘evidence-based practice’. The collaborative model developed by Katon and his colleagues (Katon et al., 1996) following disappointing outcomes with their earlier integrated model is much more highly structured and focuses on the care of a specific disorder, depression. This model, theoretically rooted in the Chronic Care Model (Wagner, Austin, & Von Korff, 1996) incorporates substantial re-organization of care including case management, protocols and case-registers, provision of self-care material to patients, training for PCPs, closemonitoring of patients by case-managers, and clearly stated protocols for ‘stepping up’ care to ensure that patients who are not making progress are referred for specialist mental health treatment. Thus far, this tightly defined and highly organized approach to integration has produced impressive outcomes in several randomized trials (see Gilbody, Whitty, Grimshaw, & Thomas, 2003 for a systematic review) but these models have not been widely implemented. The reason generally given for this, in the US setting, is the pressure to reduce costs (Druss, 2002). The more intensive care for mental health problems provided by the Collaborative model both increases treatment costs but also improves costeffectiveness for patients with major depression. A costoffset (in which analysis is confined to direct medical costs (i.e. the perspective of the health-insurer) and excludes the significant impact of mental health problems on employers, families and the social welfare system) was seen for specialty mental health costs but not medical care costs (Von Korff et al., 1998). But this has not dissuaded some services from seeking to
integrate (e.g. Dea, 2000) or dampened the enthusiasm for integration in policy circles (Department of Health and Human Services (DHHS), 2001). Indeed, as Goldman (1999) has suggested, perhaps discussion of costoffset is missing the point- ‘one rarely hears a defense of treatment coverage for infectious disease on the basis that it will reduce spending for cancer, or heart disease or ADM (mental health) treatment’. Potential solutions involving changes in the financial structure of organizations to promote integration of mental health and primary care in the US system have been suggested by Goldberg (1999) and Melek (1999). Models of collaboration described in US settings can also be recognized in the European, Canadian and Australian literature despite the different health care systems in which they have developed (Craven & Bland, 2002, Gask, Sibbald, & Creed, 1997; Meadows, 1998). In the UK, the greatest emphasis has been on the development of specialist out-reach clinics in primary care as a means of developing more effective vertical integration and the addition of mental health workers such as counsellors to the primary care team which can be viewed as horizontal integration. Although there has been considerable interest in development of consultation-liaison models where discussion takes place between professionals before referral (rather than the professional simply seeing the patient in the primary care clinic) this remains uncommon. Despite the apparent lack of ‘hard’ evidence from randomized controlled trials (Craven & Bland, 2002) the enthusiasm for the looser models of integration such as consultation-liaison and shifted out-patient clinics continues and there has been little progress in implementing more structured collaborative models of integration.
Context, study group and methods The study reported in this paper was undertaken as part of an exploration of the impact of quality improvement initiatives for two common disorders in primary care, depression and diabetes, in a large, group model HMO in the United States. The author was a visiting Harkness Fellow with no experience of working in the organization. Semi-structured interviews were carried out during the period December 2000–March 2001 with primary care workers, specialist medical and nursing staff, managers and other key informants. This paper focuses on barriers to the integration between mental health and primary care identified through analysis of data collected during this study. Forty-five interviews were conducted. Twenty-four PCPs were selected for interview on the basis of their performance on quality indicators for management of depression and diabetes. One-third of the PCPs interviewed performed in the top tertile on both of these
ARTICLE IN PRESS L. Gask / Social Science & Medicine 61 (2005) 1785–1794
indicators, one third in the top tertile in diabetes and the bottom in depression and one-third, vice versa. Mental health workers (13) were suggested for interview by the manager of the mental health service on the basis of working in the localities from which PCPs were selected and also to ensure a full range of different models of working with respect to primary care (based on same site, based on different sites, different sites with visiting mental health professional in primary care, different sites with no visiting professional). Other key ‘context’ informants (8) interviewed were an academic PCP, psychiatrist and psychologist with links to the organization, a local independent management consultant with expertise in the field, a primary care nurse and a primary care based social worker both working in the organization, and finally the senior managers of mental health and of primary care. All of these interviews (45) were carried out by the author, recorded and transcribed and form the basis of the data for this study. The data were organized and analysed with the aid of the WinMAX software programme (WinMAX, 1996), utilizing the constant comparative technique described by Glaser and Strauss (1967). In addition to the interview data, field notes recorded during and after visits to the 14 primary care clinics and 6 mental health departments where the interviews were based were also read during the process of data analysis. The data were analysed utilizing the theoretical framework of Activity Theory (Engestrom et al., 1999). The organization was understood to represent the activity system and the analysis explores the internal tensions and contradictions of the system which form ‘the motive force of change and development’ (Engestrom et al., 1999, p. 9). In the text, in order to protect the confidentiality of interviewees, all PCPs, including managers, are referred to by an abbreviation representing their profession (PCP), and mental health workers are not individually distinguished by professional group or management responsibility.
1789
health care policy and the way it’s paid for in America’. (mental health worker 12) Integrated systems must compete with carved-out behaviour health providers not only on the basis of costs but also on agreed quality criteria, which would seem to inherently disfavour integration: ‘If you are a mental health carve-out company you want to make sure that your people are seen quickly and you want to make sure that they go to the appropriate providerythat there’s some, you know appropriateness of care. So if the carve-out companies were doing thatyNCQA, a large accrediting organisation basically formalised those practices in a way of standards that became part of an accreditation process, it just became really clear that we needed to do that as well’ (mental health worker 13) The organization is required to meet quality criteria, which are essentially designed to favour the carved-out large behavioural health provider organizations, whose patients do not access speciality care via primary care. The structure of referral in this organization was therefore changed to promote direct access for the consumer by setting up a single, central referral bureau, in the context of new quality standards driven by carved-out providers. Speed of access and direct access are the measures of quality, which the customers, the employers, favour: ‘that was a marketing decision, you know, one stop shopping, one phone call’ (mental health worker 3).
Informants reflected the widely held concern over the cost of delivering health care in the US context:
However, the system is no longer set to favour referral from primary care or even communication with primary care. The potential gain here is that employers will feel satisfied that people with mental health difficulties will be seen rapidly and (hopefully) be restored to full productivity quickly. The downside is that the advocacy role played by the PCP in ensuring that patients with the greatest need are seen earlier and by the most appropriate professional is devalued and the needs of patients with somatic presentation of emotional problems, who are unlikely to directly approach mental health, will not be met. A frequently expressed belief has been that integrating mental health care in primary care would ‘pay for itself’ because of the ‘cost-offset’ effect. Some physicians clearly believed it was possible to achieve savings through improved treatment and decreased utilization of other medical service, but others did not and therefore could not support the deploying of personnel to primary care:
‘With the economic crunch it’s a very tough business to be in and it’s hard to imagine things changing in the next ten years without some significant changes in
‘It’s costingya lot of money not to deal with all these people who are depressed so we want to do a better job, because it would be better for the patient
Results Two major themes arising from the data, concerning evidence for overt and covert barriers to integration, will be described with examples taken from interview texts to illustrate key concepts. Overt barriers to integration: contextual and structural factors
ARTICLE IN PRESS 1790
L. Gask / Social Science & Medicine 61 (2005) 1785–1794
care and it would save us money. We don’t want these people in here having X-rays of this and that and somatizing all the timeywe’d rather just deal with the depression and get them better and then they’ll feel better and that’ll be better for them and it’s certainly a lot cheaper for us’. (PCP 21) ‘If you put a consultant in the primary clinic and look at sort of the total number of people seen by specialists it will go up. Which is probably good but it’s not as if you can say that we’re putting someone in the primary care clinic and it really won’t cost us any more money because all they do is see the same people earlier on. There’ll be some y as we say new businessyright now I can’t with a clear conscience, push to divert more and more resources to that’. (mental health worker 4) Clinics where behavioural health and primary care staff were permanently physically co-located were in the minority, although this had historically been more common, before the development of professional ‘departments’. In order to be able to regularly communicate with each other, people need to have opportunities for contact. The nature of professional development within an organization could however make this difficult: ‘Mental health providers, as in any speciality, need to circulate with their own and discuss cases and share new information and so there may have been a greater need for them to collaborate more with each other rather than being ‘isolated’ in a primary care setting’. (PCP 25) This ‘critical mass’ concept of professional power and influence which can be seen to provide ‘buffering’ (Thompson, 1967) against the effects of change within the organization is also played out in UK mental health politics, in the placement or otherwise of psychologists and psychiatrists in community mental health teams rather than large central departments which may be perceived as more attractive to recruiting staff. Yet this leaves no opportunity for professional interaction: ‘We don’t have a relationship with the mental health service that we know who we’re dealing with. You know we’ve never met any of the psychiatrists or the mental health specialists. We don’t even know what they look like’. (PCP 13) There was very little opportunity for interaction between mental health and primary care staff on a day-to-day basis: Interviewer: What would you like to see happeningideally how would you like it to operate? Mental Health Worker: I think it would be nice to be co-located with the mental health staff so that it’s a
lot easier to pop down to their office and vice versa. And I think I’d like to see some regular scheduled time with them. Interviewer: That simply doesn’t happen at all? Mental Health Worker: It doesn’t happen- I think nobody’s got the time to do that. (mental health worker 2). Covert barriers: attitudes and conceptual models Lack of contact and opportunities for interprofessional communication makes it difficult for professionals to challenge each others’ attitudes- to each other and to the nature of the task of dealing with mental health problems. There was a sense in the primary care doctors of confusion and bewilderment when faced with difficult mental health problems for which they could not easily access support from specialist care. One mental health worker, who had worked closely with primary care in the organization, described the attitude of the PCPs as one of ‘learned helplessness’: ‘They try for help for a period of time and if they can’t get it, and they can’t get it the way they need it, which is immediately, they give up’ (mental health worker 3) The way in which mental health professionals work, eschewing interruption and maintaining the psychoanalytic tradition of the ‘50 minute hour’ even today, militates against fast access to specialists: ‘We had a therapist here that was on call one night and a PA who gets more awards than anyone else in this clinicyknocked on her door and she said go away don’t bother me I’m with a patient. We can’t do that. I think you just poison your relationships’ (mental health worker 1) ‘It requires I think a kind of flexibility that a lot of practitioners in our field don’t have. I think many people want to sit in their office and see people on the hour and go home’ (mental health worker 3) Two conceptual differences between primary care and specialist mental health care in how they view their roles in helping people with problems were identified. First, there was tension between the ‘medical’ model of distress shared by most of the PCPs (though not all) and the psychiatrists, particularly when it came to managing people with moderately severe mental health problems and the ‘personal responsibility model’ favoured by nonmedical therapists: ‘‘The medical staff say ‘I’m the doctor, I’m calling and I want this person seen, see him, you know the therapists take more of a stance that it’s their (the patient’s) responsibility.. If they want it, they need to want to come’ (mental health worker 13)
ARTICLE IN PRESS L. Gask / Social Science & Medicine 61 (2005) 1785–1794
This tension bubbled under the surface not just between primary care and mental health but also between the mental health professionals—the psychiatrists and master’s level therapists (usually with a social work background). Second, was the mutual misunderstanding about the nature of primary and specialist care. Primary care has little time to pay attention to detail, because, unlike specialist care, it has highly permeable boundaries and easy access for patients: ‘Ask a psychiatrist if they’re going to have same day accessyof course they’re not that’s not going to workybut we’re taking care of all-comers wherever they areySomeone comes in having dizzy spells and lightheadedness, and this horrible back pain, and also ‘my marriage is breaking up and I’m depressed and I’m seeing spots and my left eye went blind yesterday for 15 min, does that mean anything doctor?’’ (PCP 1) The expertise of the PCP is in managing the breadth of problems. He or she is a ‘specialist generalist’. A mental health worker expressed exasperation that his special expertise in understanding human relationships in depth was not valued by primary care staff: ‘When you’re around people who are talking about the human condition and its psychological implications and y your conversations are in detail about the intricacies of people’s livesy then you walk into a primary care setting where that kind of information really isn’t valued very much and they really don’t want to know it, they haven’t got the time to listen to it’. (mental health worker 5). A further complication is the nature of specialist training in the US, which prepares psychiatrists for a health system that outside of the HMOs, encourages them to compete with PCPs for the management of mental health problems which in gatekeeper controlled systems would be managed by primary care. This may lead to negative views about the capability of primary care to manage mental health problems and also result in an overload of the mental health services because of a failure to negotiate discharge back to primary care: ‘Turnover is not goodyit seems frequently once the patients come to us they’re ours for everyI think the patients like it here better. I mean we spend more time with them. I think we have more accessibility. Frequently they’ll see a physician’s assistant or someone else that will work the family doctor’s office and you know they’ll say, well can we say here?yI hate to say no you’ve got to go back when they don’t want to so it’s one of the ongoing issues that we have’. (mental health worker 2)
1791
Discussion: barriers to organizational change The integration ‘project’ has been on the professional and management agenda in this organization for many years. Progress has been intermittent but the ‘project’ still survives at a ‘symbolic’ level. Some mental health workers continue to spend time providing both off and on-site consultation to primary care workers despite the lack of central support for this activity. This is not surprising. Batterham et al. (2002), with reference to Activity Theory (Engestrom et al., 1999), note that when a new way of working is introduced into a system: Tasks are reassigned, and redivided, rules are set and reinterpreted. Beyond this contradictions may sometimes occur between different parts of the system and other systems. This introduces disequilibrium and change into the activity system. In all these different circumstances co-ordination between different subjects and their versions of the object must be achieved to ensure continuous operation. In this system, continued liaison working in primary care in several sites was tolerated by the system even though it ran counter to current management goals. It seemed to serve an almost ‘symbolic’ purpose (mental health worker 13), thus maintaining a vision of a way of working, which many viewed as the unattainable or impractical ideal. Progressing further organizational change in the present financial climate was perceived by the champions of integration to be difficult. This was because, in their view, furthering the integration project required senior management to set integration as a clear priority. There is no doubt that leadership of the organization must be ‘signed up’ to the task if effective integration across the interface is to take place (Kvamme, Olesen, & Samuelsson, 2001). Financial reasons were cited for the failure to maintain an integrated model of care delivery or to implement the collaborative care model. The importance of addressing not only structure but also process when attempting to achieve interprofessional integration in healthcare has been insufficiently addressed (Burns, 1999; Iles & Sutherland, 2001), a point echoed by one of the context informants: ‘we really have to deal with the process first and the financing second and I think that in most systems particularly in the US, we’ve dealt with the financing first and hoped that it would solve the problems’. (Independent healthcare consultant) If change is to occur within an organization it would seem to be essential to acknowledge the covert barriers to change thus allowing them to become overt. It is difficult to see how this could happen without space and time for interprofessional communication within an
ARTICLE IN PRESS 1792
L. Gask / Social Science & Medicine 61 (2005) 1785–1794
organization. There are at least two possible reasons for this. The overt barrier to effective communication would be the financial pressures in the organization leading to failure on the part of management to sanction and encourage opportunities for interaction. However, an alternative explanation might be that attitudinal and other covert barriers to integration effectively prevent the development of such a shared space in which attitudinal barriers might be addressed. The exact part played by each might only be determined by further research.
Conclusion: barriers, conversations and change In this organization, which has been attempting for many years to address integrated system of mental health care across the primary-specialist interface, islets of integration activity exist with an apparent ‘symbolic’ purpose but without the perceived support of senior levels of the organization for further development. Economic factors are cited as the major reason for this hiatus, yet it is apparent that, within other similar organizations (see for example Dea, 2000) the enthusiasm for integration apparently continues unabated. I use the term advisedly, because experience from this study would suggest that it is difficult to ascertain the success or otherwise of such a project without interviewing workers in the organization ‘on the ground’ rather than simply reviewing the published literature. The major limitation of this research is that coding of transcripts was done by one person. Emerging themes were cross-checked with another researcher (MEP) who cross-read a sample of the transcripts. However, the data were only collected from one organization so cannot be generalized to other organizations. Nevertheless, the results are supported by theory and may shed some light on the reasons why integration is difficult to achieve across organizational interfaces in healthcare settings. A project cannot survive without senior management support. If managers and professionals are going to become engaged in an enterprise, conceptual differences and attitudinal barriers need to be overtly addressed. Senior managers in the organization are themselves primary care and mental health professionals who carry with them the mental models they developed at the frontline. Organizations that seek to challenge covert barriers to change, and which aspire to the vision of the ‘learning organization’ (Senge, 1990, need to create to time and space for a ‘conversation’ to take place—both between managers and workers and between professional interest groups (Peek & Heinrich, 2000). This particularly applies to the key task of ‘developing shared mental models’: the discipline of working with mental models allowing individuals to unearth the assumptions and general-
izations that influence their understanding of the world and shape how action is taken; and ‘building shared vision’ or unearthing ‘shared pictures of the future’ that foster genuine commitment and enrolment rather than simply compliance. Conversation as a medium for change is not a new concept to mental health services, forming as it does the central tenet of psychological therapies. But the most effective conversational therapies appear to be those with a clear structure in which the participants develop mutual trust and shared understanding (Frank & Frank, 1993). This conversation, may be both informal (ad hoc between professionals as they meet and work together) or formal (in meetings and ‘workshops’ (Fitzsimmons & White, 1997)). An activity system is always heterogeneous, and, from the perspective of the author, a mental health professional, the medium of constructive negotiation that was missing in this organization was the ‘conversation’ or medium of effective communication, an observation which is conceptually consistent with Activity Theory. Davydov (1999), examining the relationship between activity and communication, concludes that: ‘collective and individual activity is realized in the form of material and spiritual social relationships. Communication is a processual expression of these relationships,’ (p. 47). Such communication as it occurs at the micro level between professional groups in an organization may be effective in producing change at the macro level. But the very attitudinal and conceptual barriers that it can be successful in addressing will also simultaneously seek to prevent it taking place. In this organization, the absence of conceptual, geographical and temporal space in which to locate such communication seemed to contribute to the failure to challenge covert boundaries. The nature and quality of any interprofessional conversation that takes place may be important mediating factor in addressing covert barriers to integration between primary and specialist mental health services. Further research might address the respective parts played by overt and covert barriers over the passage of time in an organization, and most particularly the impact of these barriers on the effectiveness of organizational interventions designed to challenge attitudes and behaviour of professional staff (for example quality improvement interventions). The impact of such interventions should address not only organizational level ‘measures’, but also seek to understand the views and day-to-day experiences of those who work in the organization and examine what the impact is on outcomes for patients.
Acknowledgements This study was carried out during a Harkness Fellowship in Health Care Policy with the support of
ARTICLE IN PRESS L. Gask / Social Science & Medicine 61 (2005) 1785–1794
the Commonwealth Fund of New York. With thanks to all the professionals who kindly agreed to be interviewed, Adele Clark for transcribing the interviews, Mary Ellen Purkis for cross-reading the transcripts and Anne Rogers and Carl May for their comments on earlier drafts of the manuscript.
References Batterham, R., Southern, N., Appleby, N., Elsworth, G., Fabris, S., Dunt, D., & Yound, D. (2002). Construction of a GP integration model. Social Science & Medicine, 54, 1225–1241. Burns, L. R. (1999). Polarity management: The key challenge for integrated healthcare systems. Journal of Healthcare Management, 44, 14–31. Coleman, J. V., & Patrick, D. L. (1976). Integrating mental health services into primary medical care. Medical Care, 14(8), 654–661. Craven, M. A., & Bland. R. (Eds.). (2002). Shared Mental Health Care: A bibliography and overview. Canadian Journal of Psychiatry, 47(Suppl. 1), 1–103S. Davydov, V. V. (1999). The content and unsolved problems of activity theory. In Y. Engestrom, R. Miettinen, & R.-L. Punamaki (Eds.), Perspectives on activity theory (Learning in doing). Cambridge: Cambridge University Press. Dea, R. A. (2000). The integration of primary care and behavioural healthcare in Northern California KaiserPermanente. Psychiatric Quarterly, 71(1), 17–29. Department of Health and Human Services (DHHS). (2001). Report of a Surgeon General’s working meeting on the integration of mental health services and primary care; 2000 Nov 30–Dec 1; Atlanta, Georgia. Rockville, MD: Office of the Surgeon General. Donaldson, M., Yordy, K., Lohr, K., & Vanselow, N. (Eds.). (1996). Primary care: America’s Health in a New Era. Institute of Medicine. Washington, DC: National Academy Press. Druss, B. G. (2002). The mental health/primary care interface in the United States: History, structure and context. General Hospital Psychiatry, 24(4), 197–202. Editorial. (1993). The generic health care team. Family Systems Medicine, 11(2), 119–120. Engestrom, Y., Miettinen, R., & Punamaki, R.-L. (1999). Perspectives on activity theory (Learning in doing). Cambridge: Cambridge University Press. Feacham, R. G. A., Sekhiri, N. K., & White, K. L. (2002). Getting more for their dollar: A comparison of the NHS with California’s Kaiser Permanente. British Medical Journal, 324, 135–143. Fitzsimmons, P., & White, P. (1997). Crossing boundaries: Communication between professional groups. Journal of Managed Medicine, 11, 96–101. Frank, J., & Frank, J. (1993). Persuasion and Healing: A comparative study of psychotherapy. Baltimore: Johns Hopkins University Press. Gask, L., Sibbald, B., & Creed, F. (1997). Evaluating models of working at the interface between mental health services and primary care. British Journal of Psychiatry, 170, 6–11.
1793
Gilbody, S., Whitty, P., Grimshaw, J., & Thomas, R. (2003). Education and organization interventions to improve the management of depression in primary care: A systematic review. Journal of the American Medical Association, 289, 145–151. Glaser, B. G., & Strauss, A. (1967). The discovery of grounded theory. Chicago: Aldine. Goldberg, R. J. (1999). Financial incentives influencing the integration of mental health and primary care. Psychiatric Services, 50(8), 1071–1075. Goldman, H. (1999). Justify mental health care costs. Health Affairs, 18(2), 94–95. Iles, V. & Sutherland, K. (2001). Organisational Change: A review for health care managers, professionals and researchers. National Co-ordinating Centre for NHS Services Delivery and Organisation R &D: London School of Hygiene and Tropical Medicine. Katon, W., Robinson, P., Von Korff, M., Lin, E., Bush, T., Ludman, E., Simon, G., & Walker, E. (1996). A multifaceted intervention to improve treatment of depression in primary care. Archives of General Psychiatry, 53, 924–932. Katon, W., Von Korff, M., Lin, E., Lipscomb, P., & Wagner, E. (1992). A randomised controlled trial of psychiatric consultation with distressed high utilizers. General Hospital Psychiatry, 14, 86–98. Kvamme, O. J., Olesen, F., & Samuelsson, M. (2001). Improving the interface between primary and secondary care: A statement from the European Working Party on Quality in Family Practice (EQuiP). Quality in Health Care, 10, 33–39. Meadows, G. N. (1998). Establishing a collaborative service model for primary mental health care. Medical Journal of Australia, 168, 162–165. Mechanic, D. (1997). Approaches for coordinating primary and specialty care for persons with mental illness. General Hospital Psychiatry, 19, 395–402. Melek, S. P. (1999). Financial, risk and structural issues related to the integration of behavioural healthcare in primary care settings under managed care. Milliman and Robertson Inc. Peek, C. J., & Heinrich, R. (1995). Building a collaborative healthcare organisation: From idea to invention to innovation. Family Systems Medicine, 13, 327–342. Peek, C. J., & Heinrich, R. (2000). Integrating behavioural health and primary care. In M. E. Maruish (Ed.), Handbook of psychological assessment in primary care settings. London: Erlbaum & Associates. Perry, M. K. (1989). Vertical Integration: Determination and effects. In R. Schmalensee, & R. Willig (Eds.), Handbook of industrial organisation, Vol. 1 (pp. 185–255). Amsterdam: North-Holland. Pincus, H. A. (1987). Patient-orientated models for linking primary care and mental health care. General Hospital Psychiatry, 9, 95–101. Plsek, P. E., & Greenhalgh, T. (2001). The importance of complexity in health care. British Medical Journal, 323, 625–628. Quirk, M. P., Simon, G., Todd, J., Horst, T., Crosier, M., Ekorenrud, B., Goepfert, R., Baker, N., Steinfeld, B., Rodenberg, M., & Strosahl, K. (2000). A look to the
ARTICLE IN PRESS 1794
L. Gask / Social Science & Medicine 61 (2005) 1785–1794
past, directions for the future. Psychiatric Quarterly, 71(1), 79–94. Rosenberg, S. (1993). Rural models for integrating primary care and mental health services. Administration and Policy in Mental Health, 25(3), 287–308 (Cited in Bird, D. C., Lambert, D., Hartley, D., Beeson, P. G., & Coburn, A. F. (1998). Seaburn, D. A., Lorenz, A., Gunn, W. B., Jr., Gawinski, B. A., & Mauksch, L. B. (1996). Model of collaboration: A guide for mental health professionals working with health care practitioners. New York: Basic Books. Senge, P. (1990). The fifth discipline: The art and practice of the learning organisation. London: Doubleday/Century Business. Simoens, S., & Scott, A. (1999). What determines integration in primary care? Theory and evidence. HERO Discussion Paper 04/99. University of Aberdeen: Health Economics Research Unit.
Starfield, B. (1998). Primary care: Balancing health needs services and technology. Oxford: Oxford University Press. Strosahl, K. (1998). Integrating behavioural health and primary care: The primary mental health care model. In A. Blount (Ed.), Integrated primary care—The future of medical and mental health collaboration. New York: Norton. Thompson, J. D. (1967). Organizations in action. New York: McGraw-Hill. Von Korff, M., Katon, W., Bush, T., Lin, E. B., Simon, G. E., Saunders, K., Ludman, E., Walker, E., & Unutzer, J. (1998). Treatment costs, cost-offset, and cost-effectiveness of collaborative management of depression. Psychosomatic Medicine, 60, 143–149. Wagner, E., Austin, B., & Von Korff, M. (1996). Organising care for patients with chronic illness. The Millbank Quarterly, 74, 511–543. WinMAX. (1996). London: Sage Publications.