The prevalence of chronic disease in the Australian population is rising. Diabetes is a major cause of mortality, morbidity and disability and an important risk factor for several other chronic diseases . Diabetes is the second most frequent chronic condition managed in Australian general practice and the most frequent reason for referral to other health care providers , reinforcing the importance of the primary health care sector in diabetes care. General practice guidelines for Type 2 Diabetes Mellitus (T2DM) highlight the need for access to well coordinated health care from a range of medical and allied health care professionals, including GPs, medical specialists, diabetes educators, dieticians, optometrists, and podiatrists . Referrals to allied health professionals such as diabetes educators or dieticians are low, even among patients who are overweight or obese , indicating that there is a need to improve collaborative care across professional and organisational boundaries.
Previous research has examined collaboration between GPs and other community-based health professionals. Disagreements and conflicts over roles and role boundaries and a lack of shared decision-making suggest that issues of power and authority are important factors in these relationships and influence the patterns of collaboration [5–7]. Trust and respect are also important enablers of collaboration and mistrust and perceived lack of respect are barriers [5, 8–10].
Resource dependency theory and transaction cost analysis have been used to explore collaboration amongst healthcare organisations [11, 12]. These theories propose that collaboration is a function of the need for resources. The need for resources creates uncertainties and dependencies for organisations that they strive to reduce while maintaining their autonomy and pursuing their interests. Resource dependency theory emphasises the importance of resources to the organisation and contends that their concentration by other organisations determines the nature of the interdependency and power relations [13, 14]. Transaction cost analysis emphasises the role of governance arrangements to regulate relationships and address the uncertainties about the behaviour of others, particularly their trustworthiness . Van de Ven and Walker  argue that, in the health context, relationships established for the purpose of referring clients tend to develop on a case-by-case basis, are less formalised, and rely more on personal knowledge and trust among interacting parties.
Power has been described according to: (a) who has formal authority to make decisions and who controls the resources; and (b) who has less tangible aspects of symbolic power or the ability to control ideas and meaning . Power differences based on unequal professional status are an example of the latter. Hardy and Phillips  argue that it is the distribution of these tangible and intangible resources in interorganisational relationships which determines the strategies of engagement; namely the choice between strategies based on cooperation or conflict. The same could also be said for interprofessional relationships where the sources of power differentials, including the broader social, cultural and professional systems, produce and reinforce the power imbalances . In the hierarchy of health professions, doctors have traditionally defended their professional autonomy and independence and professional status in their relationships with other health care workers. As Hudson found, these 'turf wars' maybe intra-professional as well as inter-professional . Other research points to the situational context of power, with relationships between health professionals in hospital settings mediated by the exercise of medical dominance as opposed to the use of more collegiate approaches found in community settings .
Trust is a way of handling uncertainty and risk in the delivery of collaborative healthcare that crosses organisational and professional boundaries. It involves the expectation that other parties will behave in ways that are predictable and fair, that they are competent and will refrain from opportunistic behaviour . As a feature of interprofessional relationships, trust is often related to concepts of competence, professional identify and respect [18, 22]. Trust is also viewed as an earned characteristic that develops over time . Where opportunities to develop personal relationships are limited, the use of rules and norms to govern behaviour can substitute for interpersonal trust . Referral arrangements between health professionals are an example of such norms. However, there is a paradox: while rules and norms can substitute for trust, the transaction costs associated with these strategies could be reduced if there was mutual trust between the parties .
The concepts of power and trust from the perspective of resource dependency theory and transaction cost analysis provide a lens for investigating interprofessional relationships among primary and other community-based health services. To our knowledge this perspective has not been used to examine interprofessional relationships that cross organisational boundaries. We use the example of collaboration for the management of T2DM to address the following research question: how do power dynamics and trust influence interprofessional relationships, and how do these factors impact on the experiences of patients receiving care from multiple providers.
Setting and context
The structure of the Australian health system sets some important challenges for achieving more collaborative service delivery. Two levels of government are responsible for policy, planning and service delivery and there is a lack of integration between their various initiatives . The primary health care sector comprises a diverse range of health professions, disciplines and practitioners working in private and public sector organisations of varying size and complexity and under different funding arrangements. GPs and allied health professionals in private practice operate on a fee-for-service basis, and public sector health professionals are funded through block funding arrangements. These structural differences make it difficult to collaborate as there are few opportunities for communication and for the development of personal relationships.
Policy initiatives by the Commonwealth government to improve the management of T2DM are part of a broader focus aimed at improving chronic disease management in Australia. Since 1999/2000 financial incentives have been introduced for GPs to support more systematic and coordinated care. In 2005 with the introduction of the Chronic Disease Management (CDM) program, the financial incentives were extended to allied health professionals to support access to more comprehensive and affordable care for patients referred by a GP. GPs are required to develop a multidisciplinary care (MDC) plan which identifies the care goals and allied health professionals involved in the delivery of services to achieve these goals. The care plan is shared with the patient and the relevant allied health professionals who are also required to report back to the referring GP. The agreement is known as a Team Care Arrangement (TCA). The allied health incentives are restricted to practitioners working in private practice and a total of five occasions of service per year. No additional funding has been provided for public sector health professionals, although they can be included in the MDC plan and TCA. Financial incentives and other policy initiatives to support practice nurses have seen a significant increase in their numbers, with approximately 57% of general practices employing at least one practice nurse . Their roles include prevention and chronic disease management, quality and integration of patient care, and liaison with external providers. These initiatives have not, however, been matched by policies regarding their relationships and roles as regards other community-based nurses who have similar chronic care roles. Both groups have developed in isolation from one another.