Chapter 12. The changing dynamic of policy and practice in Australian healthcare
Christine Jorm, Margaret Banks and Sara Twohill
Introduction
Managing clinical work to achieve the best possible quality of care requires systems to be established and information to be shared. Policy is a mechanism through which these ends can be realised, and while often defined in a formal sense as a blueprint of intentions, health policy can also be described informally as the ‘courses of action that affect that set of institutions, organisations, services and funding arrangements’ in the healthcare system (Palmer & Short 2000:2). Recent policy in Australia has been directed to health system reform, but ‘good policy intentions’ are often hampered by complex funding and governance arrangements, specifically, in the case of Australia, by the politics of federalism where both state and federal levels of government share responsibility for health services. The duplication and division that shared responsibility brings impedes reform (Willis et al 2005), and the funding provided by the Commonwealth to state public hospitals via the Australian Healthcare Agreements (negotiated between both levels of government) provides the Commonwealth with opportunities to influence the shape and direction of reform. There is no easy ‘solution’ to the federal–state issue, although redesign efforts are occurring at multiple levels in the system to better link funding models with incentives for improved outcomes (Swerissen & Duckett 2002). The difficulty remains in determining which outcomes are chosen, how they are measured and the incentives and sanctions offered.
Health reform is made more complex by a range of other structural factors, including a substantial private sector that accounted for 41% of hospital beds in 2003–04, and an increasing emphasis on the responsibility of individuals for their own healthcare needs. There are the pressures from the professions, specifically the medical profession about their changing place in the system, and from consumers about the level and standard of service provided. Together with the life and death nature of healthcare decisions, these complexities in the system all impact to make health a fraught policy arena and, within this contested environment, the concern that health policy may do harm (Spitz & Abramson 2005) is a significant challenge for policymakers. Examining health policy from the recent past can illuminate important modern practices and provide clues for the future. In this chapter we examine two ‘small’ policy directives and conclude that small policy directives struggle to guide and control care, just as large policy initiatives do.
Health Policy 1990
For the purpose of our discussion, two 1990 policy circulars from New South Wales are reproduced and appended to this chapter: Circular 90/29 (Appendix 1) issued in April 1990 and Circular 90/122 (Appendix 2) issued in December 1990. Both are representative of policy issued by Australian state and territory health departments at that time. They are examples of ‘control and command’ statements that presume the organisationally shared understanding of authority, direction and discipline that is a hallmark of classical management theory. These policies were issued to the public health system in the form of a paper-based notification intended to guide the actions of health practitioners and managers.
Circular 90/29 is brief and deals with two separate issues: first, the commencement of labelling of blood units with a warning of possible infectious transmission; and second, the possibility of directed donations of blood from first-degree relatives causing graft versus host disease, and a central office directive that these blood donations be irradiated in future. These issues are separate but related. The discovery of the transmission of HIV (and other infections such as hepatitis C) via blood transfusion resulted in an increased demand for directed donation. This relationship is not made explicit in the document and multiple ambiguities are encoded into this circular: Which infections might be in the blood and undetected by screening? What data would help doctors and patients decide about the relative risks and benefit of transfusion if the blood could be infected? Are directed donations to be discouraged or encouraged? How are arrangements made for blood irradiation? Who is the label for? Is it to ‘protect’ the blood bank? The label warns, but doesn’t inform, nor does the circular.
Circular 90/122 is even briefer. Published later in 1990 it has acquired a letterhead and logo, increasing the appearance of legitimacy. It refers to a new procedure for billing for inter-hospital air transports. A previous fixed charge is to be replaced by full cost recovery. This is to be met within existing budgets. No reasons are given for the directive. It appears to be an attempt to reduce the demand for, and therefore the cost of air transport by introducing a user-pays mechanism. It is ambiguous in terms of the potential size of full cost recovery charges, and the implications for regional areas could be significant, but there is no suggestion of special budget consideration. There is no information about how clinical appropriateness of air transport might be assessed.
The circulars detail the authors’ contact details and file references available within central office. While the files may have contained information on the circular development, including the issues that precipitated the development of the policy statement (and possibly the evidence base and development methodology), this information does not appear on the circular. Policy statements at this time rarely referenced published literature or evidence, and it was infrequent that evidence informed their development. The methodology used to develop such a circular varied and included brief stakeholder consultation through to central composition by a single author.
Policy papers in 1990 needed to be printed, generally off site, and mailed to recipients in health services from the central health department in accordance with a pre-determined distribution list. Inclusion, or indeed exclusion, from the distribution list was determined by the policy officer. Their level of understanding of the operations of health services varied, therefore distribution could exclude groups vital to the policy’s implementation. For Circular 90/29 the print run was 136 copies and these were distributed to directors within the health department (25 copies), public health units (21 copies), health services (29 copies), health professional associations and related organisations, which would have included the 12 medical colleges (19 copies), public hospitals (27 copies) and private hospitals and day procedure centres (15 copies). Further distribution was reliant on these organisations having effective dissemination mechanisms.
The size and fragmentation of the private hospital sector, which at that time consisted of over 33% of available hospital beds (National Health Strategy Working Group 1991), meant this sector was unlikely to consistently receive new policy directives. Copies could be accessed directly from the health department if it was known they existed. The system did not have a reliable mechanism for dissemination of circulars to the 18,000 doctors registered in NSW in 1990 (Australian Institute of Health and Welfare 2001) nor to the approximately 30% of these doctors practising in a specialty area where they were likely to order and administer blood.
When viewed, the documents look ancient (a manual typewriter has been used) although they are comparatively recent. We ask what might these directives look like if they were developed now?
Health policy today
Outlined in Box 12.1 are the elements that would be expected to be taken into account when health policy is being formulated today.
Box 12.1
A policy directive developed today would be expected to contain the following elements:
▪ definition of the problem in terms of service provision for clinical stakeholders e.g. ‘health services have had n instances of this issue, which has had x effects on patient outcome (e.g. four deaths) and resulted in the following resource requirements’
▪ detailed reference to the current evidence
▪ a methodology – the process of consultation and development for the policy would be described e.g. a reference group consisting of participants/representatives of a number of specified groups and the outcome of their deliberations
▪ consideration of risks of implementation
▪ specification of expectations around implementation, resourcing and standardisation, e.g.
▪ state-based standardisation of a process and protocol
▪ quarterly reporting against specified indicators
▪ to be funded either within existing budget, via seeding funding or new recurrent program funding
▪ identification of whose responsibility implementation was – if it was sufficiently high level it would be included in managers’ and practitioners’ performance contracts (e.g. that of the blood bank director for a health area)
▪ links to multiple other relevant policy documents
▪ often inclusion of implementation tools, mechanisms for monitoring, and at times evaluation requirements, e.g. the safety and quality manager’s tool kit, use of the state-wide incident monitoring system.
The boxed elements reflect a number of significant developments in Australian health policy in the 17 years since the 1990 documents were created. We highlight four:
1. explicit use of best available evidence
2. significant participation by stakeholders
3. improved accessibility of information
4. introduction of accountability mechanisms.
Each element seems both compelling and essential, indeed a clear improvement over earlier policy processes. Yet when we explore them, their circumstances are complex, their value is contested and the quantity and quality of evidence supporting the benefits of the changes is variable. We discuss each element in turn, examining the current versions of exemplar policy circulars before moving on to the fourth element, accountability mechanisms.
Elements in health policy development and implementation
Explicit use of best available evidence
Policy statements in 1990 rarely if ever referenced published literature or evidence and indeed evidence infrequently informed circular development. Circular 90/29 existed as policy until it was superseded by Circular 92 issued in 2002 – well behind the systems and technology that addressed safety and quality assurance issues (although more likely in a clinical domain).
The term evidence-based medicine emerged less than 20 years ago and is no longer the province of clinical practice alone, with public health practice and policymaking increasingly looking to evidence to support effective health policy development and implementation (Sheldon 2005, Shortell et al 2007). As the healthcare sector becomes inherently more complex, expensive and demanding of greater attention to safety and quality, there is pressure to apply evidence to practice. Yet numerous barriers to its application exist (Canadian Heath Services Research Foundation 2007) and evidence informs policy to a lesser degree than might be expected. As an example, when issuing advice in areas where evidence is available, the World Health Organization rarely uses systematic reviews to inform this advice, relying on expert consensus (Oxman et al 2007).
The influential role that research should play within health policy is hard to dispute. Research has the potential to strengthen and improve healthcare systems when used as the basis for policy decisions. Nonetheless, policymakers presently view research and the evidence base that it produces as limited and problematic. There is a real need for systematic, rigorous and global methods for identifying, interpreting and applying evidence (Dobrow et al 2006). While research in the areas of dissemination and implementation of research into clinical practice is increasing in frequency, there is still a paucity of research into the effectiveness of interventions intended to improve the delivery of care. For instance, despite the sizeable investments made in patient safety, there is little evidence to support links between organisational factors in healthcare, medical errors and patient safety (Hoff et al 2004). Quality improvement strategies also lack a strong evidence base (Shojania & Grimshaw 2005) and there is a deficiency in quality scientific evidence produced through appropriate clinical trials (Tunis et al 2007), for instance for the appropriate use of medications.
Two major arguments have been advanced for this research policy gap: the differing nature of policy and research, and communication challenges in the relationship between researchers and policymakers. We briefly discuss each in turn.
The differing nature of policy and research – compromise versus clarity
If policymaking is to establish agreement between competing claims, this can only be done via compromise (Booth 1988). This agreement may be general and ambiguous, enabling the resulting policy to be flexible in scope and thus more likely to being open to interpretations (as was our exemplar 1990 blood policy). Policy derived from this methodology is adaptable to varied aims and interests and therefore more likely to be accepted.
On the flipside, research is precise in its attempt to provide clarity to inform policy decisions. It does this by taking a number of variables and examining them in a controlled and systematic manner but ‘rarely provides the breadth of vision policymakers require’ (Booth 1988:230). Obtaining an increased understanding of an issue may increase the difficulty and challenge in arriving at an agreed policy decision (especially where experience and opinion are major influences upon health policy development) (Bowen et al 2005). Hence policy and research derive from very different streams of thinking and methodological frameworks. The fluid and ever-evolving environment in which policy is created is a challenge for researchers. The appropriate entry points for evidence-based research into a policy process are rarely controlled, and they are not the same in all instances. Evidence rarely enters into the genesis of a policy proposal because there is a lack of empirical evidence outlining the best methods for effective knowledge transfer into the policy domain (Almeida & Bascolo 2006). There are also apparent or real inadequacies in the timeliness, quality and relevance of research (Innvaer et al 2002). Thus the current utilisation of research in both policy and practice is subtle and indirect.
Relationship between researchers and policymakers – communication challenges
Another theory contends that a lack of communication between researchers and policymakers is the major reason for the policy research gap (Meyer et al 2006). The ‘two communities’ theory maintains that research and policy players are two factions unable to realise the perspectives and realities of one another’s sphere (Innvaer et al 2002). Moreover, with communication lines weak, any endeavour to ‘pair up’ these spheres remains a significant challenge in part because of an absence of personal contact between key players within these domains and because of their mistrust (Meyer et al 2006). Other critics simply advocate the use of ‘[policy] user-friendly formats’ (Laupacis & Straus 2007) for research reviews. By this is meant that research should be brief and focused on validity, applicability and implementation.
A further concern stems from the recently developed pejorative phrase ‘policy-based evidence making’ (House of Commons Science and Technology Committee 2006:para 89) referring to government use or commissioning of research that supports an already-determined policy approach. Doctors in the UK, for example, disbelieve that the National Institute of Clinical Evidence (NICE), the organisation responsible for developing national guidelines, acts independently; 85% of UK doctors state that they would ignore NICE guidance if they thought it ‘was wrong’ – presumably based on the perception that tainted or biased evidence was used (McLoughlin & Leatherman 2003). Consequently, Culyer & Lomas (2006) advocate for transparent deliberative processes allowing policymakers to develop ‘evidence-informed’ decisions. While this might seem an ideal solution, NICE is considered an exemplar of deliberative processes (Culyer & Lomas 2006) but it is still not always considered credible (Milewa 2006).
In summary, while it is easy to suggest that closing the current gap between research and policy is essential for effective health policy development, there are barriers. Research and policy are two very different entities that struggle to align. A number of reasons may account for this difference: there is an overall paucity of policy relevant research; research is often considered to fail policymakers in its appropriateness and timeliness; and there is a lack of mechanisms to appropriately review, disseminate and implement research.
Significant participation by stakeholders
In Australia in 1990 few substantial stakeholder structures existed and the roles and expectations for stakeholders in the policymaking process were ill defined: the current peak national consumer body, the Consumers Health Forum of Australia, had only been founded in 1987; the Commonwealth Department of Health established the Australian General Practice Network and the National Rural Health Alliance in late 1992.
In the absence of defined structures, policymakers met with individuals or representatives of very small groups. Engaging in a process of multiple negotiations with stakeholders was thus impractical, time consuming and expensive. Not surprisingly, the policymaker assumed a position of authority, justified from the perspective that policy legitimacy derived from its enactment by a professional bureaucracy accountable to an elected government (Nagel 2006). The published documents reflect this sense of the policymaker ‘knowing best’. Stakeholders were often brought into the process late, such as at ‘final draft’ stage, and Harvey (1991:14) commented that to establish an efficient and effective healthcare system ‘the three rules are: (1) find out what works; (2) choose what to do – after considering what value is obtained for each additional dollar spent;(3) make sure it happens’. This simple process formulation did not reflect the need for stakeholder engagement or its value.
In the case of the 1990 circular concerning transfusion practice, it is not known who might have been involved in the development of this directive. Such a directive now would involve haematologists, the blood bank, the Therapeutic Goods Administration (TGA), medical managers, medical colleges, associations and foundations, accreditation bodies, nurses, doctors and consumers. Extensive discussion would be expected with consumers about directed donations (including issues of patient and family choice and rights) and about how to appropriately involve patients in the discussion about newly recognised risks of blood transfusions. Educational material for patients would be developed during the process.
Modern policy practice sees stakeholder consultation as part of the enactment of policy (see Box 12.2). The multiple expectations of the stakeholder process include education, influence, lobbying and listening. Policy is communicative work. Via these communications, policymakers can have three aims: to manage stakeholders, to seek opportunities for creative problem solving, or to build legitimacy for the process of change (Nagel 2006). These aims are not always compatible, and there is little evidence to support which is more effective or important (Nagel 2006). Sometimes the very substance of the policy is devised by communicative work with stakeholders – the deliberative democracy perspective (see Mooney, Chapter 13). Such an approach allows for flux and transformation and accepts that organisation (here content and agreement) may emerge from chaos (Morgan 1997). However, if the substance of policy is determined by the process, the definition of success and measurement or evaluation of success factors becomes difficult.
Box 12.2
A detailed documentation of the stakeholder process now forms a regular part of policy development. A good example is the recently released national Australian Palliative Care Needs Assessment Guidelines (Girgis et al 2006), which have as appendices a list of the 58 organisations and groups involved in the review, and details of the individuals that represented these groups at a national consensus meeting. However the investment of time and resources required can be substantial, and may not always seem appropriate for the issue. For instance, one recent improvement initiative in a single London sexual health clinic had representatives of approximately 65 stakeholder groups involved in the oversight group (Greenhalgh, 2007 and Greenhalgh, 2007). This caused process and practical difficulties including obtaining sufficiently large venues. The Australian Commission’s review of national safety and quality accreditation had involved 125 meetings (from one-on-one to focus group), reviewed over 140 written submissions, published four discussion papers and presented at multiple conferences to develop an alternative accreditation model. However, this is a potential policy reform that will affect the whole healthcare system and this investment in stakeholder consultation is critical to its success.
Enhanced stakeholder engagement remains problematic. Harvey (1991:14)
Buy Membership for Internal Medicine Category to continue reading. Learn more here
