Abstract
The co-occurrence of one or more health problems with significant mental illness has been recognized as presenting a major challenge to effective service delivery [1]. Continuity of health service provision is undermined when the patient is required to access a range of specialist services, especially when the effect of the mental illness has been to compromise insight into the illness or motivation for treatment. In recent years various comorbidities spanning conventional service boundaries have been a focus of service development. The four examples are the co-occurrence of Axis I and Axis II mental health conditions, substance misuse problems with mental health conditions (dual diagnosis), mental health conditions with intellectual disabilities (dual disability) or substance misuse problems with acquired brain injuries. However, when three or more disorders coexist the emphasis of dual morbidity itself risks contributing to further system fragmentation.
In the Australian context, addressing service continuity problems associated with complex mental health needs has become a major focus of health policy and funding. Both the First and Second National Mental Health Policies [2, 3], identified the development of ‘intersectorial linkages’ as a priority for mental health services. Programs across the range of youth, adult and aged health services have received funding to undertake coordinating and capacity-building roles in relation to the care of complex health needs. These include: the Primary Care Partnerships [4], Primary Mental Health Teams [5], Dual Diagnosis Teams [6], the Working Together Strategy for troubled youth [7] and the Coordinated Care Trails for complex health needs [8]. All have been tasked with improving the interface between traditional areas of specialization.
Among these areas, dual diagnosis has a growing evidence base on the subject of improving integrative practices between mental health and drug treatment services [9–12]. In their 1998 review article, Drake et al. [13], examined 36 studies, 13 of which were randomized controlled designs and 23 featured quasi-experimental designs. Comprehensive treatment programs that included intensive, stage-wise, motivation-sensitive treatments and assertive outreach components integrated within mental health settings produced the best evidence for improved attraction-to and retention-in treatment and the most enduring improvements in substance usage and mental health symptomatology. More fragile outcomes were associated with group programs or intensive treatments alone. In these studies, the most highly motivated subjects showed improvements while others either dropped out, demonstrated less enduring improvements or no measurable improvements [13].
While the research findings are suggestive, there is some dispute as to their implications for the design of effective services for these populations. Drake et al. [13] argue that creating another layer of specialization, within the fragmented health care system is the best way to facilitate the necessary integrations of practice. However, critics of specialization argue that improving the comprehensiveness of what can be offered in one part of a system, without reference to how that component will interact with the rest of the system, would have a restricted capacity to bring about sustainable improvements in services to people with complex needs [14].
The absence of a clear consensus of opinion as to the service implications of the major research findings has resulted in a continuing interest in qualitative investigations on the views and experiences of key stakeholders. In New Zealand, Todd et al. [15] utilized focus groups of mental health and drug treatment service providers, patients and family members to identify what barriers were operating to reduce the capacity of the system to achieve optimal care for people with dual diagnoses. They identified three broad areas that act as barriers to the optimal care of dually diagnosed people. These were: (i) fragmentation in the system of service providers; (ii) lack of capacity within the various clinical areas to work more comprehensively and cooperatively; and (iii) unproductive attitudes that act to maintain things as they are.
From a systems integration perspective the study provides important formative information about what needs to be addressed in order to bring about improvements in the care of people with dual diagnoses. However, like Drake et al. [13], Todd et al. [15] limited their examination to one area of complexity and thereby placed constraints on their capacity to effectively examine system level issues and how they interact with individual, interpersonal and institutional factors. If each area works in isolation in the absence of a coordinating model each is vulnerable to repeating the fragmentation and simply rearranging the metaphorical ‘deckchairs on the Titanic’. Rather than continuing to examine the various forms of comorbidity in isolation, this paper presents a study that has undertaken a more expansive systematic examination of how the health care system deals with complexity. The social ecological model provides a framework for undertaking such an examination.
Social ecological thinking in community psychology extends the focus of assessment and treatment beyond the individual and the immediate environment. Community contexts are understood as systems with multiple interdependent layers of influence. (i) Factors within the individual such as biological vulnerabilities, coping styles, motivations and attitudes interact reciprocally with: (ii) interpersonal and group processes; (iii) factors related to the institutions with which individuals and groups interact; (iv) community levels factors; and (v) higher order factors that effect social policy [16, 17]. The social ecological model posits that robust and sustainable change in community contexts requires well-planned and ecologically relevant programs that coordinate multiple intervention strategies targeting multiple social strata.
The present project has also adopted a participatory action research (PAR) model as the means to engage community stakeholders in the process of building the capacity for more integrative practice in service delivery to complex needs. From its inception the processes and principles of PAR were intended to promote community consciousness-raising in relation to social inequalities [18–20]. The approach involves researcher and the researched collaborating in repeated cycles of reflection and action. Community members contribute their unique perspectives, skills and knowledge to discovering mutually meaningful social insights which inform the development, implementation and evaluation of community projects [21–23]. Participatory action research offers a means to engage the community in a genuine attempt to examine, from the bottom up, what constitutes complexity across the health and welfare systems and what can be done to improve this for the parties involved. The methods are also highly compatible with social ecological analysis, allowing the development of an explanatory model that incorporates multiple stakeholders across the multiple levels of social analysis.
The present project utilized PAR methods to investigate community-wide issues relating to service delivery to people with complex mental health needs. The broader project begins with a proactive and formative evaluation study that takes place over three phases focusing on: (i) needs assessment; (ii) program development and piloting; and (iii) implementation of a community development project. This paper reports the findings of the preliminary proactive community needs assessment process. As the first phase of the larger project, the needs assessment had the goal of engaging stakeholders across the communities of the Southern metropolitan and Gippsland regions of Victoria in a collaborative focus groups process in order to: (i) develop a shared understanding of what is happening in regards to people with complex mental health needs; (ii) undertake a social ecological analysis; (iii) identify obstacles to better service provision to people with complex mental health needs; (iv) recommend ways to overcome the obstacles; and (v) identify appropriate evaluation approaches with respect to any programs arising from (iv).
Method
Participants
Letters were sent to 75 agencies identified through a snowballing method as providing direct services to the population with complex mental health needs. This identified providers across the core stakeholder categories of: (i) disability support, rehabilitation and welfare; (ii) primary care; (iii) crisis response and emergency services; (iv) supported accommodation; (v) legal and forensic services; (vi) alcohol and other drug (AOD); (vii) youth AOD; (viii) mental health services; (ix) child and adolescent mental health services; (x) consumer advocacy groups; and (xi) carer groups. The letters invited stake-holding groups and agencies to circulate the letter within their organization in order to identify interested people and to nominate any people who might be interested in participating in the focus groups process. Interested people were provided with a plain language information sheet and asked to make an informed choice about their participation.
Expressions of interest were attracted from 141 possible participants, from 53 different agencies and support groups, of which 98 were selected and given the invitation to take part in one of seven focus groups, 14 participants for each group. A stratified sampling method was used in which participants were randomly selected within the core stakeholder categories identified above in order to achieve the widest possible representation of experiences within manageably sized focus groups. Across the groups 71 of a possible 98 participants attended the first of two sessions giving a participation rate of 72%. The groups comprised 44 (62%) female and 27 (38%) male participants and ages ranged from 22 to 71 years. Numbers declined for the second group sessions with 39 (55%) of those that attended the first session returning for the second.
Procedure
The participation process
In order to minimize the impact of diffuse status characteristics and authority factors on group communication a series of seven focus groups were implemented in the first phase of the project. These involved homogeneous groups of: (i) direct care workers; (ii) managers; (iii) consumers; and (iv) carers representing rural and metropolitan areas. To ensure adequate representation of remote rural areas two groups were delivered in the far north-east one for (v) workers and managers, and the other for (vi) carers and consumers. Finally, a group combining representatives from across the above-mentioned categories of membership was offered in order to bring any themes that may emerge from interactions into the analysis.
Focus groups process
Each of the focus groups met for two 90-minute sessions in which they were led through a qualitative inquiry process. The process was conducted by two experienced facilitators who were not affiliated with the auspicing agency and therefore not a party to any biases that might have been present in the goals and objectives of the agency. In the first session all the groups were tasked with generating statements in response to an identical series of focusing questions. These were intended to orientate the attention of the focus groups to the issues, opportunities, problems, concerns or needs relating to providing effective health services to patients with complex mental health needs. The questions were as follows:
On the basis of your experiences what are the kinds of difficulties that identify/define complex patients, complex needs? What in your experience helps? Can you suggest reasons why the people you identified may not be getting adequate services? Can you think of anything that could make a difference so as to enable these patients to get better services?
All groups were audiotaped and tapes were transcribed for analysis. Between the first and second set of focus groups sessions, the transcripts were condensed into two sets of representative statements by two independent raters. The four sets of statements produced by the two raters were highly consistent and the final two sets were reached by consensus. The first set involved the 74 statements that related to defining, ‘what constituted complex patients or complex needs’, drawn from the discussion generated in relation to the first focusing question, presented above. The second set comprised 108 statements that related to the theme of ‘what needs to be done to overcome barriers to better services for complex needs’, generated in relation to the remaining focusing questions presented above. The sets of statements were then processed to produce two concept clarification instruments using the following procedures:
The first set of 74 ‘problem defining’ statements were incorporated into a Likert type survey in which each statement can be rated on the 5-point scale (Not at all, Sort of, Moderately, Clearly, Extremely) in relation to how central each idea is to a respondent's thoughts and experiences with complex needs or complex patients. The second set of 108 ‘solutions’ statements were individually printed onto separate 75 mm × 40 mm cards. Each statement card was produced with a unique identifying number from 1 to 108.
The second 90-minute focus group sessions were held 2 weeks after the first sessions. Participants were presented with three activities designed to clarify the themes present in the statements that had been produced in the previous sessions:
The sessions commenced with an open-ended invitation for participants to explore any themes that arose for them in the previous session.
Defining features of complex needs – Participants completed the Lickert type survey in order to assist the clarification of the key features of complex mental health needs.
Overcoming the barriers to service delivery – Using a methodology based loosely on Trochim's [24] ‘Concept Mapping’ procedure, participants produced clusters of related themes through a ‘card sorting’ procedure, and generated construct names for each of the themes. Participants were presented with the card set and given the instruction to: ‘Sort the cards into piles of related ideas. There must be more than one pile and there cannot be 108 single card piles. Then, give each pile of cards a name which describes the idea the statements have in common. Finally give each card a rating of how closely it is related to the central idea of the pile on a four point scale: 1 = sort of related, 2 = moderately related, 3 = clearly related, 4 = extremely related’. Participants were encouraged to complete the tasks independently, although some struggled alone and were formed into pairs. The card piles and ratings were then recorded onto a summery sheet using the unique identifying numbers printed on each card. Thirty-two card sorts were produced by the 39 participants.
Summary of findings
The two lists of themes represented in the ‘problem defining’ and ‘solutions’ lists mirrored each other closely. For this reason only the themes and relationships identified in the ‘solutions’ card set will be presented in detail. The responses to the ‘solutions’ card sorting instrument were plotted manually to track the frequency with which each card was sorted together with each other card.
The process revealed four strong clusters of cards representing conceptual themes that were named by the participants themselves as part of the card sorting process [24]. The conceptual themes became the thematic core by which to organize the qualitative analysis of the transcripts produced in the second focus group sessions. These four ‘solution’ clustered will now be described briefly.
Reduce fragmentation in interagency service delivery in complex cases
With reference to the social ecological model, fragmentation was described as occurring within the system from the top down and the bottom up. From the top down, funding practices were seen as continuing to favour reactive strategies without integrated planning. In addition, standard forms of funding accountability were identified as focusing on key performance indicators such as bed occupancy, service throughput and episodes of care. Many service providers, consumers and carers were of the opinion that dealing with less complicated and time-consuming cases yielded positive productivity indicators, whereas, working with more complex people resulted in the appearance of lower productivity. The processes were believed to encourage the over-servicing of straightforward ‘core business’ needs and discourage service providers from working with more complex and challenging multi-problem patients.
From the bottom up, the potential effectiveness of human service integration activities were seen as being compromised by a myriad of factors that impact on an agency's ability to sustain itself as a business within an economic rationalist political environment. Economic imperatives, workload constraints, workforce issues, clinical accountability, philosophical differences and resource limitations left service providers with restricted capacities and incentives to work beyond a strict interpretation of role definitions. Services were seen as having withdrawn to rigid notions of ‘core business’ that create numerous gaps for patients to fall through. Almost without exception representatives from government-funded clinical agencies described practice imperatives that were organized around crisis response or short-term focused interventions.
In regards to the question of what can be done to improve the situation the groups identified six broad principles focusing on: (i) the accessibility of services and treatments; (ii) structures that promote interagency cooperation towards the goal of integrated service delivery; (iii) assisting clinicians to maintain their involvement with complex patients; (iv) holistic models of care; (v) the scope and inclusiveness of the ‘continuum of care’; and (vi) continuous improvement.
A comprehensive ‘continuum of care’ was identified as being crucial in providing the scope of services to address complex needs. A necessary range of appropriate services were identified including treatment, rehabilitation, supported accommodation, recreation, extended care, case management and the capacity for workers to remain involved over time.
Enhance integrative case management practices across the system
This theme emphasizes working towards the enhancement of existing assessment, case formulation and case planning practices to promote improved integrative efforts across agencies. Case management was seen as the most accessible means to promote continuity and coordination and to minimize the risk of diffusion of responsibility across multiple providers. Participants endorsed the principle of engaging patients collaboratively as the foundation for undertaking comprehensive assessment and case formulation processes. Communication barriers need to be overcome so that these formulations can be used to inform the inputs of multiple service providers. In addition, case formulation processes were seen as needing to be broader than clinical diagnosis processes. These also needed to be sensitive to the difference between state and trait issues across the immediate and longitudinal aspects of the patient's presentations.
Issues in the wider social contexts of the family, health and welfare institutions and the broader community were also identified as impacting on individuals in ways that need to be incorporated into actual case planning. Commonality of language and transparency of purpose across the elements of the system were identified as crucial foundation goals that need to be addressed in the interests of improving the sys-tem's capacity to communicate and coordinate more effectively.
Community attitudes
A number of attitudes and attribution tendencies were identified by participating consumers, carers and workers that were believed to act as barriers to integrative practice. These were thought to be destructive to the accessibility of services generally, and particularly prejudicial towards people with complex needs. The attitudes and attributions included: (i) beliefs about the relative deservingness of certain types of problems in comparison to others; (ii) beliefs about the relative worth or moral legitimacy of certain coping choices; and (iii) the tendency for difficulties engaging the cooperation of patients to be attributed to character-logical resistance or opposition even when psychopathology is recognized as being present that can itself account for the behaviour; and (iv) value judgements about abstinence as opposed to empower-ment-based or harm-minimization models. The groups also referred to professional ‘blind spots’ and rivalries as contributing to service access difficulties and to communication problems across service providers. For example, a patient may not be exposed to certain indicated service or treatment options because a worker was either unaware of, or distrustful of, those options. The provision of training, education and information across the community at large, which is informed by best practice, was identified as the key means of addressing these issues in the short to medium term.
Address the lack of money and resources
The groups identified higher level social processes impacting on resource distribution as contributing to system fragmentation. It was thought that community service providers and funding agencies are often caught in patterns of miscommunication, which leave both sides ill-informed for effective integrative innovations to develop in a mutually supportive way at the institutional, community and policy levels. The groups identified the need to promote more effective, mutually accountable communication processes between funding agencies, service providers and other stakeholders that make transparent the differences between what is conceptualized and what is implemented in the field.
Discussion
The qualitative themes that have emerged through this process yield a shared representation generated by a broad group of participants who have been interacting with the system from a range of roles and perspectives [22]. They provide clues as to how complexity is constructed and interpreted across the system of involved agencies and interest groups. The current model represents a work-in-progress that identifies influential factors across the five levels of social ecological analysis and a set of priority areas for addressing the obstacles to better practice. The priority areas were organized in terms of four conceptual themes: (i) system fragmentation; (ii) integrative case management; (iii) community attitudes; and (iv) money and resources. The implications of these findings for the development and evaluation of this project will be examined in relation to top-down and bottom-up processes before going on to consider the next step for the project.
Top-down considerations
Much has been written in the program planning and evaluation literature about the fragmentation that is inherent in complex open systems [14, 19, 20, 25]. An aspect of this focuses on the problems associated with organizing the health system around conventional specialist programs and services [1]. Conventional specialist service configurations have been described as being unnecessarily problematic to patients, duplicative, fragmented, expensive, inefficient, unresponsive, crisis orientated and compromised by placing professional interests above those of patients [25, 26]. The proponents of system integration models argue that increasing the capacity of the system for comprehensively addressing the needs of multi-problem patients, requires an array of interventions involving efforts to: (i) improve the accessibility and continuity of the diffuse range of services encompassing the continuum of prevention, early intervention, treatment, rehabilitation and extended care; (ii) reduce unnecessary duplication, waste and inefficiency; and (iii) improve communication and accountability [1]. All of these themes were echoed by the focus groups in this project.
In Victoria, the growing emphasis on funding new integrative initiatives has taken up the challenge of working to overcome the drawbacks of organizing programs around specialization without sacrificing the benefits of specialist knowledge and expertise. This endorsement in policy, of the principles of comprehensive practice is an important development in itself. However, implementing these principles takes on the dimensions of having to orchestrate a revolution in the form of a Kuhnian paradigm shift [27]. In the USA where human service integration has been on the agenda since the early 1970s, numerous difficulties have constrained the implementation of these ideas. The principles of integration have been challenged at the legislative and service provision levels by groups invested in the traditional distribution of power and resources [1]. It has been argued that even though advancements have taken place in policy and practice it remains the case that the working parts of the system are predominantly aligned around traditional categorical thinking [1].
Bottom-up considerations
The themes produced by the focus groups are consistent with many of the issues identified by Todd et al. [15] in relation to the system of dual diagnosis primary stakeholders in New Zealand. Fragmentation across the system of service delivery agencies was seen as being a major impediment to dealing more effectively with complex mental health needs. Like Todd et al., the case management model that identifies a single clinician with responsibility for coordinating a range of inputs was identified as a potential ‘focal point’ for interagency cooperation [16]. If properly supported, the role was seen as having the potential to coordinate inputs from a broad range of service providers to minimize the ‘diffusion of responsibility’ that occurs when multiple providers become involved. Beyond this, a more comprehensive approach to assessment and case planning encompassing bio-psycho-social factors was seen as necessary for achieving better coverage of complex needs. For example, issues in the wider social contexts of the family, health and welfare institutions and the broader community were identified as also impacting on individuals in ways that are not consistently incorporated into actual case planning. Moreover, incorporating such elements into case planning efforts was seen as crucial to the task of engaging a diverse service system meaningfully and willingly into cooperative service delivery.
Once again consistently with Todd et al. [15], a number of attitudes and attribution tendencies were identified that were believed to act as barriers to the efforts of integrative practice. Many of the attitudes encountered in service providers and consumers appear to mirror those of the community at large where issues such as substance misuse, self-harming behaviours, homelessness, unemployment, sexual assault, sexual abuse and chronic pain are often associated with the blaming of the victim 28– [31]. It was thought that such processes may serve to promote reluctance in service provision personnel and agencies to commit resources to longer-term work with complex patients.
The issue of the proper support for the case management role is an important one. A prevalent perspective among the clinicians, managers, consumers and carers in the focus groups was that case managers find themselves caught in ‘parallel processes’ of disempowerment along with the people they are supposed to be helping. The mental health system particularly is argued to be filled beyond capacity. Case managers have heavy patient loads, distributed across large distances, a strong duty of care, crisis-focused mandate, limited opportunities for shared care and little time to spend on coordination and liaison activities. In rural areas the problem is further amplified by even greater distances, greater dispersion of resources, over-stretched supervision and support structures, a lack of specialist services and limited public transport, creating a further burden on patients, carers, workers and services to facilitate service accessibility. Across the system there is little capacity to undertake psychosocial and therapeutic interventions that have established efficacy in overcoming or preventing the escalation of problems. These resource issues were argued to make it logistically impractical at the present time for case managers to consistently follow through with the responsibilities of the idealized role in the community.
Where to from here?
The present study lends encouragement to the principles of integrative practice. The preliminary explanatory model is sensitive to the social ecological landscape of the areas in which it was generated and thus supplies meaningful conceptual foundations for undertaking the second phase of the project. This diverse group of service users and providers have found agreement in identifying the need for improving practice across the diverse elements of the health and welfare systems through the development of integrative models, more supportive case planning structures and overcoming unproductive attitudes.
In regards to the construction of the present project, there is an endorsement for the principles of integrative practice despite the acknowledgement of many barriers to its implementation across the levels of social analysis. At the level of policy and funding the present strategy of targeting specific problematic interfaces, such as dual diagnoses, is a reasonable place to start. However, such targeted programs are unlikely to infiltrate the system and effect change across the levels of social analysis unless these too are embedded within a model that can conceptualize and organize more expansive efforts at integration [17]. The challenge is to find ways of expanding on the practices and rhetoric of the dominant paradigm of traditional categorical approaches to health system organization and service delivery.
Any attempts to improve the system's capacities for dealing with complex needs will require the promotion of principles and attitudes that can address both the topdown and the bottom-up influences. With the key goal of facilitating integrative practices, the project will need to incorporate education and training across the stakeholding agencies addressing the elements of comprehensive assessment, case formulation, case planning and review. Moreover, fragmentation across the health and welfare systems was also partially attributed to higher level social processes that impinge on resource distribution. The project would aim to influence and inform policy-makers through its contribution to evidence-based practice and through community participation to contribute to the wider community's momentum for change.
Ongoing participation
Much of the literature relating to PAR methodologies has examined the significant benefits and problems associated with seeking to achieve the representation of a range of stakeholder groups in the design and administration of projects [32]. Community participation is acknowledged to be a challenging goal and is itself an emergent property resulting from the characteristics of the problems and environmental conditions under study, and the aims, resources and capacities of the research team [33]. Projects are often examined from the point of view of how group processes have suffered from power imbalances, group inertia or conflict [14], or have fallen short of achieving the desired participation and empowerment [19, 20]. This study, likewise, was vulnerable to selection biases and other social factors that shaped the composition of focus groups. Where the study succeeded in achieving a broadness of representation, the process has also identified stakeholders and interest groups that need additional representation in the developing project. As the project continues, the representation of children's and youth services, schools, general medical practitioners, emergency services and supported accommodation agencies needs to be improved.
The next phase will continue the development of the explanatory model through an ongoing participatory process involving: (i) a series of local community advisory groups made up of representatives from the agencies identified as being primary stakeholders in the project; and (ii) regular consultative forums that will seek to continue and further develop the engagement of a wider body of informants and stakeholders. This process acts in the service of the major goal of utilizing the evolving explanatory model to inform the parallel development of action strategies and evaluation instruments addressing the priority principles that have been identified in this first phase of the project. This goal will involve developing interventions that are achievable using the available resources and identifying courses of action for highlighting or addressing issues that lie outside the project's immediate influence. Ultimately, the project aims to bring about improvements in the capacity of the system to work more effectively with complex mental health needs. The parameters of which will continue to emerge as the project continues.