When colocation is not enough: a case study of General Practitioner Super Clinics in Australia

Governments in developed nations are experimenting with new models for delivering primary care services. The Patient Centred Medical Home in the USA, Family Health Teams in Canada and Integrated Family Health Centres in New Zealand represent attempts to meet evolving community needs for safe, effective and affordable services (Jackson 2012; Brown et al. 2013).

General Practice Super Clinics (GPSCs) were introduced in 2010 as a key component of Australia’s then National Primary Health Care Strategy, with stated principles aligned to these new US, New Zealand and Canadian models (Department of Health and Ageing 2010). The Strategy aimed to improve primary care infrastructure and ease impending challenges of chronic disease, clinical complexity and increasing demand for clinical placements for future health professionals. GPSCs were intended to provide integrated and multidisciplinary services, enhance links with community organisations and optimise culturally appropriate preventive care and chronic disease management (Department of Health and Ageing 2010). The GPSC program supported construction of purpose-built facilities, which were intended to facilitate integrated care through coordination and colocation of multiple disciplines, with shared clinical governance and care protocols. GPSCs could colocate GPs, a range of allied health, mental health, visiting medical specialists, chronic disease nurses and community education. These services could be delivered on GPSC premises by a range of providers, including out-posted staff from hospital networks and community health services.

Unlike the international models, GPSCs were introduced without reform to practice funding models and lacked external support for additional health professionals. The GPSC scheme only funded construction, fit out including IT and limited start-up costs. Organisations submitted tenders to the Federal Department of Health and Ageing (now Department of Health) to construct GPSCs in over 60 predetermined locations. Once established, GPSCs had a 20-year reporting responsibility against program objectives (Table 1).

Table 1.  GP Super Clinic: national program objectives (Department of Health and Ageing 2010)

Despite an initially warm reception, the GPSC program attracted criticism from professional bodies and the medical press, with questions raised as to whether funding ($419 million 2008–12; Australian National Audit Office 2013) could be better directed to supporting existing practices (Van Der Weyden 2011). The program was discontinued following a change of government in 2013. Nevertheless, from 64 contracts awarded, 61 GPSCs were constructed, 60 were operational as at February 2016 and three were cancelled.

Given the novelty of the GPSC program and its stated objective of ‘well-integrated multidisciplinary patient-centred care’, we aimed to illustrate how the process of transitioning into a GPSC influences the development of organisational and clinical routines, particularly relating to the collaborative care of persons living with chronic illness.

The case study methodology used a rapid ethnographic approach informed by the sociological concept of routines as the unit of analysis for understanding organisational change in primary care (Becker 2004; Pentland and Feldman 2005; Greenhalgh 2008). Key concepts of routines are represented in Table 2.

Table 2.  Key concepts of routines (Becker 2004; Pentland and Feldman 2005)

As interviews often gather information mainly on ‘ostensive’ or explicit aspects of routines, we also employed observational techniques to access information on ‘performative’ or implicit aspects (Pentland and Feldman 2005). Our case study approach allowed a detailed, intensive exploration of individuals and organisations in context (Patton 2002).

Two GPSCs, sited in the Australian states of New South Wales and Victoria.

A typical case sampling strategy (Patton 2002) identified potential GPSCs that shared key characteristics from publicly available sources. Eligible cases (~10) had GPSC program contracts with the Federal Government, incorporated non-GP health professionals from two different professions and were situated within 100 km of a major population centre. Principal investigators selected two sites in different states that were in reasonable proximity to researchers, and organised visits to explain the study. Clinics were offered $1000 in recognition of inconvenience associated with data collection. Both clinics approached agreed to participate.

A field worker visited both GPSCs twice, separated by 12 months, for 3 days on each occasion. During visits, we used ethnographic techniques of non-participant, direct observation and in-depth interviews with practice members and clinic leaders. The field worker and two investigators observed board meetings in one clinic. Data collection focused particularly on the routines associated with multidisciplinary care.

Observational data for each site was organised using a validated practice environment template (Ohman Strickland and Crabtree 2007). Templates, interview transcripts, practice documents and field notes were coded using NVivo9 (QSR International, Melbourne, Vic., Australia; Richards 2002).

Data analysis was adapted from an approach developed in a larger Canadian study (Russell et al. 2009). Data were first analysed by the field worker using a constant comparative approach (Strauss and Corbin 1998), then refined at regular investigator meetings and at a face-to-face data retreat with all investigators, including senior social science, academic GP and organisational behaviour researchers. Site visits in the later stages of the study allowed practice members to check presentations of summary data and interpretations made using the routines framework.

The study was approved by the Royal Australian College of General Practitioners’ National Research and Evaluation Ethics Committee (approval number NREEC 10–010).

Both GPSCs approached agreed to participate, and their names are fictionalised here. Table 3 summarises the characteristics of each site. The field worker interviewed 15 practice members at the ‘Outertown’ GPSC and 16 at the ‘Hillside’ GPSC, including board members, managers, GPs, allied health and nurses. They observed board and practice meetings, informal interactions among staff and between staff and patients, GP and nurse consultations and reception procedures. Boxes 1 and 2 summarise the ethnographic data.

Table 3.  Key characteristics of GPSCs
GPSC, General Practice Super Clinic; FTE, full-time equivalent

In both GPSCs, we mostly observed independent serial care by different disciplines in fairly typical general practices, under significant external pressure. The sites had ongoing problems in attracting either patients or clinicians, and both were preoccupied with the demands of construction, strongly influenced by the contractual requirements of the funding agreement with the Department of Health and Ageing.

Outertown prioritised the need to generate an adequate new patient base. Keen to offer a more comprehensive service, and aware of the need to augment multidisciplinary activities, the clinic rented clinical space to a physiotherapist, a psychologist, a psychiatrist and members of a community mental health team. However, these services worked in parallel with little interaction, structured by GPSC–hospital organisational agreements for specialist clinics or the use of MBS Team Care Arrangement items for private allied health. Coming from different organisations, clinicians maintained discipline-specific routines that were codified in different operating procedures and clinical information, such as hospital network systems for the mental health team.

Hillside’s multi-site, hub and spoke model had few problems attracting new patients, given its location in a region of health workforce shortage. However, construction delays were profound, and provider recruitment proved difficult, especially given the preference for UK GPs, perceived as being trained in routines of multidisciplinary teamwork. Hillside attempted to form a clinical coalition of previously independent, geographically dispersed providers of varied disciplines. Where some collaborative care was achieved, it relied on established relationships between organisations or individual practitioners, allowing efficient use of MBS items. Again, multidisciplinary collaboration was sporadic, reflecting differences in disciplinary frameworks, as demonstrated by the osteopath’s uncertainty about GPs accepting her clinical judgement.

Despite these contextual challenges, both GPSCs generated examples of more integrated models of primary care delivery, driven by an awareness within both practices of the importance of generating organisational and individual routines distinct from traditional general practice care. Such routines were generated in response to differing local needs, organisational structures and history.

Both GPSCs developed a condition-specific clinic, which flourished. Each initiative arose from a desire for a ‘whole of practice’ approach to clinical care. Each clinic required additional enablers: leadership, clear role definition, external protocols and incentives, and identification of a distinct patient cohort. Transfer of clinical responsibility for Clozapine patients to Outertown clinic staff was initiated and supported by the regional Hospital Network, whereas the Hillside Diabetes clinic was championed by the GPSC’s leadership. These features laid the groundwork for an agreed ‘truce’ about how practices allocated staff roles, which involved changes to performative routines. Evidence-based protocols gave a clear base for new collaborative routines to evolve; each clinic had specific requirements regarding data collection, monitoring, patient reminder systems and, to some degree, patient self-management.

General Practice Super Clinics entered a landscape of minimal change in the infrastructure of primary care in Australia. Apart from the rise of large GP ‘corporatised practices’ from the late 1990s, general practice care in Australia has mainly been delivered through small, privately owned general practices. The GPSC initiative was the first nation-wide attempt to implement a new model of primary care delivery. Our in-depth investigation of the evolution of two different GPSCs shows that the constraints inherent within the GPSC program, and the lack of effective incentives for collaborative care in fee-for-service MBS items, meant that program objectives for integrated multi-disciplinary care were largely unattainable. Despite diligent and, at times, creative approaches to the implementation of new models of care, neither GPSC could embed meaningful changes in team function and chronic disease management over the 12-month period of our observations. It was difficult to embed new routines at the individual health practitioner level when the organisations lacked the structure or capacity to prioritise these changes.

Despite emerging international literature on the varied iterations of the patient-centred medical home (Jackson 2012; Wagner et al. 2012; Quinn et al. 2013), little has been published on the GPSCs. The literature is limited to several articles that use GPSCs as a setting (Akter et al. 2014; Nancarrow et al. 2014; Bajorek et al. 2015), an outline of one GPSC’s model (Dart et al. 2010) and an early evaluation of the program sponsored by the Department of Health and Ageing.

That evaluation found concerns about financial viability, disciplinary colocation rather than collaboration and the lack of tools to support multidisciplinary care (Considine et al. 2012). Our study confirmed that these issues were ongoing, and adds to understanding the difficulty in changing existing routines without adequate program focus. Some key issues are: the need to establish an agreed division of roles between GPs, nurses and allied health; clinicians under pressure may resist a shift from existing routines of working in a single discipline; it takes time to change routines in busy primary care practices; clinical information systems and standard operating procedures vary between organisations and disciplines.

The GPSC program can be considered in the context of worldwide initiatives to reform the organisation and delivery of primary care services. Jurisdictions have implemented new delivery models (Rosser et al. 2010), trained new primary care providers (DiCenso et al. 2007) and embedded frameworks of prevention, integration and team-based chronic disease management (Coleman et al. 2009). An associated literature is emerging on how primary care practices can best be transformed to accommodate these systemic changes (Wagner et al. 2012).

International experiences indicate that practice transformation towards multidisciplinary care takes time, as clinicians and other practice staff adjust established practices and routines (Rämgård et al. 2015). Success requires: embedding principles of evidence-based care, relational continuity and patient-centred care; systematic approaches to delivery of comprehensive clinical care (Wagner et al. 2012); and active, visible leaders, explicit staff training and quality data (Quinn et al. 2013). A supportive medical neighbourhood, quality improvement collaborative groups and outreach facilitation seem to assist. Several authors have emphasised the importance of alignment or re-organisation of financial incentives and how meaningful reform may require 3–5 years (Nutting et al. 2011).

Although both GPSCs had strong clinical leaders, willing staff and an organisational vision of a broader model of care, they made minimal progress towards establishing routines supporting integrated multidisciplinary patient-centred care. As the lack of institutional support left practices to fend for themselves, any moves to the system objective of integrated multidisciplinary patient-centred care were serendipitous or as the result of individual passion, interest and of the occasional opportunities provided by the local context.

Transferability of findings is limited by the low number of participating clinics, although they were in different states and with different governance structures.

Data collection had a short timeframe, therefore limiting assessment of their evolution, although we could make longitudinal comparisons over 12 months.

Use of a single observer could bias interpretation. However, principal investigators also visited each site several times and emerging findings were regularly reviewed by the research team, including two in-depth data analysis retreats. Case analyses were presented to each clinic.

Our ethnographic approach was well suited for capturing clinic routines and behaviours. Epidemiologic methods would be required to examine the influence of practitioner orientation, practice structure and local context.

Collaborative inter-professional routines develop slowly and require individual, practice and system support (Rämgård et al. 2015). Our data showed how preoccupation with financial viability at two new GPSCs could conflict with other objectives, significantly slowing development of collaborative routines for chronic care.

Our data has implications for policymakers and researchers interested in primary care transformation. We suggest that, given the lack of meaningful support for real transformation of the model of care, the GPSC initiative should be considered a program for infrastructure development rather than one of practice-based primary care reform. Future Australian attempts to modify similar aspects of the delivery of general practice care should be mindful of the emerging evidence from other nations where primary care transformation is viewed as a complex domain (Bodenheimer et al. 2014) requiring enduring external supports for integrated multidisciplinary patient-centred management of chronic disease (Lebrun-Harris et al. 2013).

G. Russell conceptualised the study. R. Lane carried out the field visit, analysis in NVivo and was primarily responsible for drafting the manuscript. G. Russell, E. A. Bardoel, N. Zwar, J. Advocat, P. G. Powell Davies and M. F. Harris, participated in its design, coordination, analysis and helped to draft the manuscript. All authors read and approved the final manuscript.