There has been a growing international evolution of the role and purpose of quality improvement in primary care; particularly in the United Kingdom, Europe, Australia and New Zealand. Research has focused on improving clinical care programs, with a corresponding focus on the identification and development of clinical measures and indicators of quality care. Subsequently, attention has been given to the role of clinical governance in the delivery of high-quality care in general practice, and exploration of the relationship between clinical management and patient health outcomes.1–5

The past 12 years also witnessed an international movement from funders paying for activity to paying for outcomes. The UK introduced “pay for performance” in 2002, resulting in general practices receiving up to 25% of their funding from measuring and reporting against 134 quality benchmarks.6 The United States debated options including significant bundled payments to family practitioners for quality targets.7 New Zealand developed a framework to guide clinical quality in primary care.8 Australia’s quality measures, including the Practice Incentives Program and Service Incentive Payments, were introduced in 2002, but their funding has been progressively decreased over recent years. The Royal Australian College of General Practitioners standards form the basis of Australian general practice accreditation, and include elements such as infrastructure and clinical management.9

There is a growing international consensus regarding the impact of organisational elements on the delivery of high-quality care and as enablers of successful continuous quality improvement.10–12 Elements include leadership; practical and human resources; active engagement of all staff; and attention to multifaceted interventions and coordinated action at all levels of the health system, such as investment in training and development of robust and timely data through supported information technology.13–15 The relationship between practice assessment, organisational development and assessment, and quality improvement is complicated. An understanding of the elements of high-performing practices may contribute to the development of organisational and cultural assessment processes, which in turn lead to organisational development relevant to primary care as part of the broader concept of successful continuous quality improvement. Conversely, undertaking successful continuous quality improvement through activities associated with organisational development, and using organisational assessment approaches may, in turn, indicate elements integral to high-quality practice performance (Box 1). Indeed, there is dearth of evidence related to the impact of organisational components of general practice and patient outcomes.16

In this article, we sought to identify elements, contained in the literature, which have been demonstrated as integral to high-quality practice performance. For the purpose of this study, practice organisation was defined as “‘systems, structures and processes aimed to enable the delivery of good quality patient care’, excluding clinical processes and clinical outcomes”.17

We defined the term tools as surveys, questionnaires or assessment instruments designed to measure overall or specific elements related to practice organisation, as defined above.

Methods

We undertook a narrative systematic literature review. The specific questions addressed were:

• What elements (attributes or characteristics) are integral to high-quality primary care practice organisation?
• What are the current key considerations relating to organisational performance in primary care?

Key methods are summarised below; full details can be found in Appendix 1.

We searched several electronic databases: PubMed, the Cochrane Library, CINAHL, Embase, Emerald Insight, PsycInfo, the Primary Health Care Research and Information Service (PHCRIS) website and Google Scholar. Additional articles and reports were identified through the reference lists of all reviewed studies. A complete description of the search strategy, including search terms and study eligibility criteria, is found in Appendix 2.

Data were extracted from eligible articles and reports using a data extraction form guided by the template used by Dunbar and colleagues.18

A qualitative inductive thematic approach was used to explore the discussion sections of all articles (surveys; trials; frameworks, models or approaches to organisational improvement or assessment) to explore information about the application and perceived impacts of tools and frameworks. Incorporating this information added further clarification, particularly of the relevance of identified elements to general practice settings.

Data were configured at the study level to allow for the inclusion of findings from a broad range of study types (systematic literature reviews, trials, frameworks, descriptive, knowledge-building papers and key informant discussions). Results were compared and combined to identify the elements of organisational assessment in primary health care that are integral to high-quality practice performance.

Results

A total of 241 manuscripts were obtained for review. Of these, 210 manuscripts were excluded due to no or insufficient information about the tool and elements of practice performance. Tools that focused on elements of patient safety (eg, the Frankfurt Patient Safety Climate Questionnaire for General Practices; the UK National Health Service [NHS] Manchester Patient Safety Framework – Primary Care; NHS Education for Scotland’s SafeQuest) or on the developing patient role in health care decision making (including the General Practice Assessment Questionnaire and Europep) were excluded from data extraction.19–25 A full description of the systematic review screening process is presented in Box 2.

All citations were imported to Endnote. A total of 31 articles were included for data extraction, including one systematic review; 25 developmental trials or surveys of organisational assessment and tools or guides and six descriptive articles or reports. Of these, three were Australian; 13 were from the US; 14 from the UK and Europe; and two were Canadian. These articles included information on 10 organisational assessment tools or methods.

Elements identified as integral to high-quality practice performance

We identified 10 elements that were integral to high-quality organisational performance in general practice. Appendix 3 lists these elements according to the tools that include them. It should be noted that this table highlights elements that were specifically contained in the tools. However it should also be noted that several externally facilitated tools could draw out broader issues in practice management during facilitation.

The following 10 elements were most frequently included in existing organisational assessment tools:

Patient–centred approaches: This element included tailoring service delivery to the context of family and the broader local community; the importance of a community focus; use of community resources; and cultural competence in relation to a knowledge and understanding of the local community the practice serves.

Leadership and leading: This included the concepts of knowledge of, and attitudes to, practice leaders in practice organisation, administration and clinical care. The element also related to individual practice members who may lead in relation to innovation and change; it addressed concepts of effective leadership and “leading” behaviours, regardless of individual positions and roles.

Focus on staff: This element contained the concepts of staff satisfaction, autonomy, skills and professionalism. It also addressed monitoring of staff workloads and job stresses.

Clinical governance: Closely associated with the element of communication and collaboration, the element of clinical governance related to the formal systems and structures in place to ensure effective care delivery and clinical safety, such as patient complaints procedures; patient call-back processes and medicine alerts; with a specific emphasis on clinical care structures and risk management.

Multiprofessional teams: This element contained concepts of the ease of forming multiprofessional teams in practice; effective teamwork in relation to important diseases; and relationships between clinical and non-clinical staff, including understanding each other’s roles and the ability to learn through conflict.

Communication: This element encompassed a range of concepts — formal and informal processes of communication internal to the practice, including delegation; environmental and cultural factors that support effective collaboration and linkage between the practice and outside services; methods of timely referral; and sharing of patient information. It also encompassed environmental factors that supported sharing information with patients.

Education and training: This element was seen as a fundamentally important part of quality improvement. It relates to provision of, and access to, appropriate training for staff, as part of existing roles. It also included education and training tailored to changes undertaken in the practice.

Process improvement: Closely linked to the element of performance results, this element also most commonly related to clinical processes of care, the systems in place for monitoring the process of practice health care delivery, and internal improvements to the practice.

Performance results: This encompassed processes to support reporting of results of performance measures internally and externally. It included benchmarking against other services.

Information and information technology: Finally, the use of information and information technology included aspects such as the effective collection and use of information and the effective use of technology (eg, practice clinical software). This element most commonly focused on the use of patient clinical information, less common was the inclusion of information and data on practice finances and billing, and data related to human resource management. It also encompassed the sharing of patient clinical information internally, and between the practice and external services.

Three less commonly included characteristics were identified from the discussion sections of descriptive articles and trials of some tools. During the inductive thematic review, phrases were grouped into the most commonly occurring themes or attributes associated with high-quality practice performance, namely:

Incentives and rewards: This attribute included the capacity for all staff to receive recognition and reward for their work — not just financially, but also in recognition from peers, ability to attend conferences or join professional groups. Financial incentives as part of organisational development facilitated and sustained effective team-based approaches to care. However, this element was not specifically included in existing tools but was linked to discussions in organisational governance and change management.

Organisational governance: This element included the definition of a shared direction, mission and values; strategic planning and implementation; and collection and inclusion of administrative data.

Change and change management: This attribute included the ability, willingness and flexibility of the practice staff to adapt to new standards and procedures; the ability and willingness of the practice staff to make, manage and sustain change. A practice with a history of change was also identified as an important structural part of successful and ongoing performance improvement.

Synthesis of results

Combined, these 13 elements represent the characteristics of organisational context that are integral to high-performing practices (Box 3). Rather than being discrete elements, they are also interlinked. For example, the element of communication — information availability, defined as the sharing and communication of information internally and externally, is closely linked to the development of multiprofessional team-based care approaches. The element of education and training for all staff and incentives is integral to successful change management and ongoing readiness for change.

Discussion

Most existing quality improvement tools are designed for tertiary care settings or for non-health care organisations.5 Those developed for general practice and primary health care are single strategies such as audit, professional education (continuing medical education) and frameworks for patient safety and risk management. A review of organisational assessment tools used in the UK concluded that while externally led quality improvement approaches, such as accreditation, are reasonably well developed, internally led approaches are much less so.46

There are several considerations relating to organisational performance. First, a primary health care tool should define elements essential to primary health care, be sensitive to clinical management, enable the users to better understand their own practice context, and include staff in the process of identifying and addressing improvement activities of relevance. Second, as health care delivery becomes more complex and technology-driven, the organisational context in which qualitative improvement initiatives take place becomes a crucial determinant of their effectiveness.10,11 Research on primary care practice development identified these contextual elements as “adaptive reserves”, or features that represent a practice’s internal capability.47 Finally, many of the elements identified here are also organisational attributes that enable the success of practice-based improvement initiatives.13 Thus, the challenges are to develop a tool bespoke to primary care with a whole-of-practice approach; that can fit a wide range of contexts; that is relevant to the clinical management aspects of general practice and effectively takes into account organisational aspects of care; that includes elements relating to organisational culture; that is low or no cost; and that can be used effectively as part of an internally led quality improvement process.

The literature in relation to quality improvement is large and diverse; however, studies relating specifically to practice management performance are poorly defined in relation to specific search terms. We deliberately used broad search terms to capture the totality of elements, domains and characteristics included across all tools, frameworks, models and approaches.

There are presently no validated organisational improvement tools, designed specifically for primary care, which can be used internally in a best-practice approach that facilitates both organisational improvement and also raises awareness about the present culture of a practice. Based on the findings from this systematic review, we have now undertaken the development and pilot of a new internally led Primary Care Practice Improvement Tool (PC-PIT), incorporating the elements identified here.49 A national trial is now underway, and future work will explore the link between improved organisational performance and patient outcomes at the practice level. With the continued focus on the important place of general practice and primary health care in the broader health care context, and a refocus on the importance of organisational aspects of practice in relation to quality care delivery, the time is now right to focus on a standardised internally led approach to improving practice performance, designed for the dynamic context of primary health care.

1 Relationship between key terms

2 Flow diagram depicting selection process of studies for analysis

Accreditation of health services and its potential role towards improving health care has been described previously.1–3 Improving patient safety through the process of accreditation has been noted in the acute care setting.4–6 Further research is needed to evaluate patient safety in accreditation of primary care, its impact, and how to improve it.

The Royal Australian College of General Practitioners (RACGP) has developed a set of standards to protect and improve the safety and quality of health care provided in general practices. These standards serve as a “template for safe and high quality care”, for the general practices that have joined an accreditation program (about 80% of practices in Australia).7 These standards are used by accreditation agencies to carry out the accreditation process through experienced surveyors who are involved in primary care. GP surveyors and co-surveyors, such as practice managers or practice nurses, carry out the accreditation process, which is approved by the agency accreditation review committee.

One of the RACGP standards is to provide a systematic approach for clinical risk management, to recognise and avoid near misses, slips, lapses or mistakes.7,8 Risk management is defined by the RACGP as “the culture, processes and structures that are directed toward the effective management of potential opportunities for adverse events”.7

Patient safety in primary care is underestimated due to poor use of available data and difficulties in measuring safety incidents,9,10 although several studies have described the errors and harms that occur in primary care in Australia11,12 and internationally.13–17

The Australian Safety and Quality Framework for Health Care,18 endorsed by heath ministers in 2010 as an important driver of quality, has highlighted three key domains for creating safe, high-quality care. One of these domains requires health services to be “organised for safety”, which further emphasises the importance of accreditation in reducing harm in general practice.

This study aimed to explore Australian General Practice Accreditation Limited (AGPAL) surveyors’ perceptions of the impact of accreditation on patient safety, and to elicit suggestions for improving patient safety in Australian general practices.

Methods

We asked AGPAL to recruit a national purposive sample of their surveyors on our behalf to represent most Australian states and territories, their experience in the use of different editions of the RACGP standards for accreditation, and number of practices they had accredited. AGPAL sent an email invitation with the study information and consent form to surveyors. Interview questions centred on the process of accreditation, general practice performance and patient safety (Box 1).

We conducted semi-structured telephone interviews (40–60 min) with participant surveyors from 2 July to 14 December 2012. All interviews were audio recorded, transcribed and summarised using the interview schedule as a guiding framework to identify participant perspectives of the impact of accreditation on safety in general practice.

This project was granted ethics approval from Flinders University Social and Behavioural Research Ethics Committee (project no. 5609).

Results

Eleven surveyors consented to participate; one declined at a later time due to time constraints. The 10 AGPAL surveyors who participated in the study were involved in the accreditation of 2022 general practices over 15 years across the Australian states of New South Wales, Victoria, Queensland, Western Australia, South Australia and Tasmania, and the Australian Capital Territory. Seven were GP surveyors and three were co-surveyors. The number of practices accredited by each surveyor varied from 100 to more than 400 practices, except for two surveyors who had accredited only five practices (Box 2). Several editions of RACGP standards had been used by eight surveyors.

Accreditation and overall general practice performance

Overall, participants were of the view that accreditation improved general practice performance. This was particularly noted as an improvement over time and as a result of the accreditation process. Examples of improvement centred on the physical environment of general practice, such as equipment safety, appointment systems, patient records and electronic records.

Some participants mentioned that the systematic approach of accreditation provided a method for improving general practice systems, such as through establishing business thinking and activity in general practice.

However, doubts were also expressed about accreditation and its impact on improved practice performance, such as the “one-off” nature of accreditation and sustainability of improvement efforts.

General practice staff awareness of patient safety

There were mixed responses regarding general practice staff awareness of patient safety, evident in the two quotes below, from concerns about an overall lack of awareness to views that awareness was related to staff roles and responsibilities; especially for practice managers and practice nurses, who were regarded as having high awareness because they were mainly responsible for the accreditation process and activity.

The following participant noted that safety awareness was discipline-specific, with GPs being aware of the clinical components of safety and other staff being aware of the physical safety elements of general practice.

Patient safety as a component of the accreditation process

Most participants stated that patient safety indicators are included in the accreditation process. The indicators primarily mentioned were from section five of the RACGP standards,7,8 which covers the physical environment of the practice such as infection control and sterilisation, cold chain, vaccination and physical access. Here, it was noted that accreditation lacks the ability to effectively assess safety in clinical practice as distinct from this focus on safety processes in the physical environment, which are easily recognised as being achieved or not.

Participants suggested that the accreditation process could be improved through the inclusion of tighter clinical safety indicators and the requirement of verifiable evidence of a working clinical risk management system.

Evidence of clinical risk management systems in general practice

Participants confirmed that most general practices did not have sufficient evidence to demonstrate the existence of a clinical risk management system.

The general practices considered to be high performing on patient safety indicators were described as having evidence that verified an active and effective clinical risk management system. These included having a working clinical risk management document, incident register or policy, and staff participating in meetings and discussion about slips, lapses and mistakes, and producing documented outcomes and actions.

Participants were asked to estimate the number or proportion of practices performing well in patient safety indicators and in clinical risk management. There were mixed estimates; five of the participants estimated that 5%–10% of general practices they had accredited were high performers in patient safety, three participants estimated 30%, and one participant estimated less than 1%. Only one participant gave a high estimate of around 90%. It was noteworthy that the more experienced surveyors (accredited more practices) provided lower proportions of general practice performing well on patient safety indicators.

Improving patient safety culture in general practice

Participants were asked to recommend approaches to improve the safety culture in general practice. Their responses included further education and training, and novel ways of thinking, emphasising the roles of practice managers and practice nurses, instilling an interest and inclination towards improvement and safety, establishing open, honest, and sharing communication practices along with reflection and documentation. Also mentioned were introducing a business model to practices, creating systematic and “multi-pronged” interventions to introduce change, and funding and incentives for change. While the above recommendations were viewed as enabling change, some also acknowledged the difficulties and challenges required when attempting to change practice behaviour and activities.

Following on with discussions around possible ways to improve patient safety culture in general practice, participants were asked about the Australian Primary Care Collaboratives (APCC) Program. One surveyor who had recently joined AGPAL was not aware of the program. Nine surveyors believed that the APCC Program is commendable, with potential to promote and improve patient safety culture in Australian primary care because of the systems that it introduces to practices.

Discussion

This is the first study in Australia to examine AGPAL surveyors’ perceptions of the impact of accreditation on patient safety in general practice. The findings suggest that accreditation has improved the safety and quality of Australian general practice, but there is still room for improvement, particularly concerning clinical safety and providing verifiable indicators that require practices to demonstrate evidence around clinical risk management.

Participants affirmed that the physical factors of the environment affecting safety, such as infection control and sterilisation, cold chain, vaccination and physical access, are well addressed during the accreditation process. Conversely, clinical risk management indicators lacked sufficient verifiable evidence needed to demonstrate acceptable levels of safety. Furthermore, this type of evidence was thought to be provided by only 5%–10% of Australian general practices.

One solution could be the addition of extra procedures in the accreditation process that require evidence of clinical risk management. Those suggested by participants included: a) having a significant incidents register; b) providing documentation of near misses, slips, lapses or mistakes; and c) engaging in regular clinical meetings to discuss incidents and how to avoid them in the future.

Accreditation could advance the use of clinical risk management in general practice, as outlined in the RACGP standards with the proposed recommendations listed above. Additionally, these recommendations are a step forward to meeting the proposed actions to prevent or minimise harm from health care errors reported in the Australian Safety and Quality Framework for Health Care.18

Incident registers with documentation of significant safety incidents could be the foundation to developing a national register for anonymous reporting of errors and near misses.

The APCC Program’s success in improving quality is evident,19,20 and was well endorsed by participants in our study to be a program that could promote and improve the patient safety culture in Australian primary care. With this in mind, we recommend that the APCC Program add patient safety to its agenda for improvement.

Limitations to this study include the possibility of selection bias and the use of self-report. The number of participants was small and surveyors not sampled for this study may have different responses and hence different recommendations and conclusions to study participants. Finally, the proportion of practices who were perceived as high performers for clinical risk management activities are based on self-reported estimates from participants, and therefore cannot be considered to be an actual estimate.

Our study suggests the current “softness” around patient safety indicators in the accreditation process for general practice needs to be improved. The inclusion of tighter indicators that require verifiable evidence is a step forward to retaining the positive role of accreditation in improving general practice performance in quality and safety. Furthermore, given the APCC’s previous success in general practice quality improvement, it would be an appropriate organisation to further improve patient safety in general practice.

Is Australia ready for evidence into policy?

Internationally, health reform initiatives have identified improved integration between community and acute care delivery as key to sustainability.1,2 Australian reform initiatives have been no exception. In 2009, Australia’s National Health and Hospitals Reform Commission laid the “blueprint” for Australia’s health care future.3 It stressed the importance of “connecting and integrating health and aged care services for people over their lives”. However, it also observed that “each level of government formulates policy in relation to its own responsibilities, not necessarily taking account of the health system as a whole”, and that “current governance arrangements are contributing directly to weaknesses in the quality, effectiveness and efficiency of the Australian health system”.3

Three years later, the National Healthcare Agreement 2012 committed all federal, state and territory governments to support an integrated approach to the promotion of healthy lifestyle, and prevention, diagnosis and treatment of illness across the continuum of care.4 Despite the critical nature of joint community and acute care cooperation in delivering on the plethora of mutually agreed objectives, there was no mention of any commitment to the integrated governance arrangements pivotal to such success.

Concurrently, Australia’s first National Primary Health Care Strategy established a network of 61 primary health care organisations, Medicare Locals (MLs), across Australia.5 While the Strategy stated that MLs “will be an integral component of the National Health and Hospitals Network” and “have some common governance membership with the Local Hospital Networks [LHNs] in their region”, an integrated governance model was never developed.5

Given the priority placed on effective governance frameworks to deliver clear roles and responsibilities to both funders and providers of health care, what is the governance vehicle best suited to achieving our national reform outcomes, and how is it best crafted in the current Australian health care reality?

In 2013, we conducted a systematic review to explore international peer-reviewed articles and relevant websites for effective and sustainable integrated primary–secondary health governance models. Ten key elements were identified, many interdependent, from 21 articles that met the inclusion criteria.6 The evidence suggests the following specific governance elements are important to support integrated care across the primary–secondary care continuum:

• Joint planning was identified as key in 18 of the 21 articles. Governance arrangements included formal agreements such as memoranda of understanding (MOUs), joint board memberships and multilevel partnerships in the planning process.
• Integrated information communication technologies were noted in 17 articles, particularly, a shared electronic health record, and systems that link clinical and financial measures.
• Effective change management was noted in 17 articles, requiring a shared vision, leadership, time and committed resources to support implementation.
• Sixteen studies agreed on the importance of shared clinical priorities, including the use of multidisciplinary clinician networks, a team-based approach and pathways across the continuum to optimise care.
• Aligning incentives to support the clinical integration strategy, noted in 15 studies, includes pooling multiple funding streams and creating equitable incentive structures.
• Providing care across organisations for a geographical population, noted in 13 articles, required a form of enrolment, maximised patient accessibility and minimised duplication.
• Use of data as a measurement tool across the continuum for quality improvement and redesign, found in 12 studies, requires agreement to share relevant data.
• Professional development supporting joint working, supported by 11 articles, allowed alignment of differing cultures and agreement on clinical guidelines.
• An identified need for consumer/patient engagement, noted in eight studies, is achieved by encouraging community participation at multiple governance levels.
• One-third of articles acknowledged the need for adequate resources to support innovation to allow adaptation of evidence into care delivery.6

Governance elements identifiable in Australia’s current health care reform environment

We reviewed key statements regarding integrated care delivery from federal, state and regional perspectives to identify evidence of integrated primary–secondary governance support. Although several elements above are noted as goals, formal documents mostly relate to silos of sector activity and not the interface (Box).

The following integrated governance elements are currently well documented:

• Joint planning is documented in agreements at both federal and state level. It is identified as a key role for MLs and LHNs, and there is evidence to support it as a key objective in both strategic plans.9,11 Some MLs and LHNs have created joint board positions and local MOUs, demonstrating a commitment to managing deliverables, risk and processes through a collaborative approach.12,13
• A key objective of MLs7,8 and federal–state agreements4 is to work in partnership to document and together address shared clinical priorities for action. ML–LHN MOUs identify agreed shared clinical priority areas,12,13 some based on local need (eg, mental health),13 others on national benchmarks set by the National Health Performance Authority (NHPA), Close the Gap and national priority areas.12
• At local level, ML–LHN MOUs have documented evidence of commitment to patient, consumer and community engagement.12,13 This includes joint planning fora and input into informed patient choice regarding options for care. A joint consumer engagement approach including co-hosting of events is supported in one ML–LHN MOU.13
• Federal and state health departments and authorities produce documentation on the health of their populations, and MLs and LHNs must focus on population health service planning as part of their reporting. MLs and LHNs have collaborated to provide population health reports for regions. Lack of enrolment is currently a limitation to accurate geographical population health data in Australia.

Areas still to operationalise:

• Policy direction requires “e-health tools to link providers and improve quality of care”.4 However, as an integrated information communication technologies tool, the personally controlled electronic health record (PCEHR) has some way to go before it becomes a shared electronic health record capable of providing “access to more health information, creating a more efficient system, making continuity of care easier and improving treatment decisions”.14 This remains a significant barrier to integrated health care going forward. Federal support for the rollout of the PCEHR is articulated,7,9 although this is not reflected in state plans.10,11 The PCEHR review was completed on 20 December 2013, but the findings have yet to be released.15
• Federal and state governments “will … look to improve quality and accessibility of data to inform planning and service delivery with a ‘whole of system’ view”.7 One ML and LHN have agreed to use shared data as a measurement tool to meet performance requirements articulated by NHPA.12

Areas still to evolve:

• There is little in current policy documents to incentivise integrated care. Primary care functions on a largely fee-for-service model, moving patients to emergency departments or hospitals when more comprehensive care is required. State-funded acute care has few current funding or governance levers to link with private or federal-funded care. New models in New Zealand employ governance frameworks that create the funding and business rules to better incentivise care models across the interface. Instead of input-defined, competitive, fee-for-service contracts with penalties for underperformance, it has moved to “alliance” contracting to create joint incentives to manage cost.16,17 Preliminary evaluation shows wins in patient acceptability, quality of care and hospital avoidance.16
• There is no documented evidence that LHNs or MLs have currently committed resources to jointly manage the change required to working collectively across the interface. This can be complex, challenging and resource intensive.18
• There is no documented evidence of organisational commitment and resourcing to deliver interprofessional training across the continuum or to develop training programs that align differing cultures and integrated ways of working.19
• The Council of Australian Governments supports the need to “invest in research that promotes evidence based practice and innovation”.4 Although one ML noted the importance of building a culture of innovation and the need to invest in and demonstrate leadership in innovation, evidence of this occurring is lacking.

The Australian reform environment has made steady progress in building integrated governance arrangements around joint planning, shared clinical priorities, consumer involvement and population health service planning. However, other areas, such as integrated information communication technologies, using shared data as a measurement tool, shared resources to support change, and interprofessional or interorganisational training and innovation remain ad hoc or non-existent. Despite their importance, incentives for integrated care still fall predominantly into short-term programs, rather than robust governance arrangements at federal, state or local level.

If we are to apply important evidence to health care policy, and maximise reform success, we must review current governance frameworks to address the gaps identified in this paper. While it is challenging to bring historically disparate partners together into formal agreements, they are essential to creating the “business rules” and sustainable environment required to achieve the new care models we seek.

Multidisciplinary leadership teams and flexible approaches are helping streamline New Zealand’s health care system

While the search continues for governance arrangements that support health system and service integration,1,2 developments in New Zealand provide useful new insights. New Zealand presently has 20 district health boards (DHBs) planning and funding regional hospital and other services, and around 30 primary health organisations (PHOs) that plan and fund elements of general practice and primary care for enrolled patients. These two sets of arrangements have functioned largely separately from one another, despite DHBs funding PHOs and both having common populations.3 New Zealand’s policymakers and health care providers have concluded that it is no longer acceptable nor sustainable to operate a health system with parallel structures that lack coordination or a governance model that supports this.

In response, from mid 2013, New Zealand moved to implement a governance model across the entire country, aimed at integration by requiring an alliance between each DHB and corresponding PHOs. This followed investment in 2010 in nine pilots. The alliance concept derives from the construction industry, where independent companies collaborate, rather than compete, to ensure that large, complex projects are delivered on time and within budget. While the health alliances are forced by policy, they are an example of an experimental governance model4 that, evaluations of the pilots suggest, provide considerable promise.5 For example, alliances have helped drive important new initiatives that provide better support for complex patients in primary care settings by enabling general practitioners to work together with hospital specialists and other providers. While early days, there is some evidence of reductions in emergency department admissions and of more services traditionally provided in hospital settings being delivered in the community, such as specialist outpatient consultations, older people’s health, and emergency response services that might otherwise require a hospital visit. Importantly, those involved in alliances believe it is a model that helps steer health system and service design in an important new direction.5,6

Some important factors underpin the alliances. Members should

• be clinical leaders from across the health system, with influence and respect among colleagues;
• have capacity to bring resources to the alliance table so decisions can be implemented; and
• very importantly, cast aside sectoral interests, work to assist one another, and take a whole-of-system approach to planning and decision making based on what is best for the patient and health system.

Alliance goals variously include shifting services from hospitals to primary care or creating new arrangements combining elements of both service domains to, for example, reduce avoidable hospitalisation or improve chronic condition management. The key, as noted, is to focus on and work towards what makes best sense in the context of integration to the players in the local health system.

All DHBs now have an alliance leadership team (ALT), membership of which is determined by the DHB and PHO and evolves as an ALT sees fit. Members are likely to include doctors, nurses, allied health professionals, others from hospital and primary care settings, and those with resources, such as the chief executives of the DHB and respective PHOs and consumer representatives. Each member signs a charter spelling out the rules of engagement and focus of the ALT, which then sets local priorities and plans how these will be met.

There is flexibility for how an alliance goes about its activities. Many ALTs are focused on developing service-level alliance teams (SLATs). These are work streams that include, again, a combination of clinical leaders and management. The Southern Health Alliance Leadership Team, of which I am Independent Chair, has chosen initially to create SLATs for acute service demand management; outpatient services; diagnostics; rural health; community and hospital pharmaceuticals; frail older people; and respiratory conditions. To illustrate how a SLAT functions, initial respiratory SLAT work involved a workshop including hospital emergency department and respiratory physicians, GPs, nurses and ambulance services. Resulting actions include identifying frequently hospitalised patients, providing nurse-led care plans for them and ensuring that the patient’s GP and, where necessary, hospital services are involved in this, and developing primary care-based options for ambulance services. Development of web-based clinical pathways aimed at integration, involving health professionals from the primary and hospital sectors, is also governed by the ALT.

In the Canterbury region, where alliance development is more established, dozens of people from different parts of the health system are involved. With care design decided on advice of a SLAT, it is then up to the ALT and its member organisations to pool or shift resources to support new configurations. This process is being propelled by new flexible funding arrangements, whereby the PHO can use existing ring-fenced allocations in new ways as decided by the ALT. The DHB is expected to contribute to this pool which will grow with time, along with the level of joint risk sharing, as an alliance work program advances.

How alliance performance will be measured is an important question that the government is tackling. An impending Integrated Performance and Incentive Framework incorporates a range of system-wide measures, including patient experiences with the system, requiring an effective alliance and integration in order to perform well.7

Like Australia and other countries, New Zealand’s public hospitals and GPs work with quite different incentives and business models.8 Yet, alliances have provided a powerful method of bringing health professionals together from different parts of the system and motivating them to work collaboratively on what services should look like from a patient and clinical perspective. Given their relatively embryonic state, the challenge now is to monitor closely how the alliances perform over time and consider lessons for policymakers elsewhere.