Submitted in total fulfilment of the requirements of the degree of Doctor of ... the framework in Indigenous health and potential criteria for evaluating its ...... The second issue, strengthening the evidence base to guide priority setting, requires ..... (Donaldson 1995; Mitton and Donaldson 2003; Mitton and Donaldson 2004; ...
Thesis Title
Priority Setting in Health: Contextualising Program Budgeting and Marginal Analysis (PBMA) in Indigenous Health.
Michael Ekubu Otim Bachelor of Science (Honours) (Makerere University, Uganda) Graduate Diploma in Economics Studies (Newcastle University) Post Graduate Diploma in Health Economics and Evaluation (Monash University) Master of Economics (Honours) Monash University
Submitted in total fulfilment of the requirements of the degree of Doctor of Philosophy
14 December 2010
School of Population Health The University of Melbourne
I
Abstract The Australian government recently pledged to close the health gap between Indigenous and non-Indigenous Australians within a generation. It is argued in this thesis that explicit priority setting can potentially aid the “closing the gap” Australian government agenda. The key challenge, however, is how to carry out explicit priority setting in order to maximise the health outcomes of Indigenous Australians. This is the key issue addressed in this thesis: 1. How explicit priority setting ought to be done in Indigenous health; and 2. Whether and, if so, how an economic approach to priority setting called Program Budgeting and Marginal Analysis (PBMA) should be contextualised in Indigenous health. Explicit guidelines were developed for priority setting in Indigenous health using normative economic and social justice1 theory, and lessons from the literature and an empirical study. The empirical study was a survey of key decision-makers in regional Victoria involved in Indigenous-specific health care services. The results of the survey indicated that the size of the health burden, feasibility, sustainability, equity and acceptability are important criteria for setting priorities. These guidelines were then used to assess a selection of priority setting frameworks, and the PBMA framework emerged as the most appropriate framework because of its ability to uphold both economic and social justice principles in priority setting. In particular, the due process aspect of PBMA was consistent with Aboriginal values.
Four empirical studies were then carried out to understand how PBMA could be applied in an Indigenous health context: The survey of decision-makers to gauge the acceptability of the framework in Indigenous health and potential criteria for evaluating its performance:
1
Social justice is a virtue, an attribute of individuals. It is the habit of putting the principle of association into daily practice. It allows for people of good will to reach different—even opposing—practical judgments about the material content of the common good (ends) and how to get there (means) Novak, M. (2000). "Defining social justice." First Things December 2000.
II
1. An expenditure study in the Loddon Mallee Region of Victoria to demonstrate how the ‘program budget’ should be done; 2. The costing of a dry areas alcohol intervention to highlight costing and data issues that should be faced and how to resolve them; and 3. The concept of benefit study to demonstrate how to expand the concept of benefit from interventions to reflect what Indigenous Australians want from a health service.
Results from these studies indicated that, although PBMA can be resource and data intensive, it is a potentially acceptable and useful approach for Aboriginal Community Controlled Health Services (ACCHS) to use for setting priorities. For example, the challenges of undertaking the program budget in Indigenous health such as, lack of data with Indigenous identifiers and the specification of the program are potentially resolvable. A checklist to guide ACCHS in implementing PBMA in Indigenous health was developed. This checklist included: a conducive atmosphere for PBMA application; and practical steps for undertaking PBMA exercise. A conducive atmosphere involves ownership of the process and the results, organisational stability of ACCHS and organisational decision-making culture to focus on the costs and benefits. The practical steps for the implementation of PBMA as a routine management tool would include activities such as setting up information systems which produce information (preferably on costs and outcomes) for service planning and decision-making, and setting up of the PBMA management committee.
As a routine decision-making tool, the PBMA framework would assist decision-makers to shift the decision-making focus to costs and benefits of every course of action taken. Furthermore, the information systems in ACCHS may have to be aligned with programs, to better data extraction, or the use of expert advice where such data may not exist. However, most of the above empirical studies were demonstration studies, requiring caution in the interpretation of these findings. In particular, the concept of benefit instrument, which reflects Indigenous health construct, needs to be developed further and trialled. Furthermore, there is need to trial PBMA checklist in an ACCHS setting and then evaluate its performance.
III
In conclusion, it is important for the decision-making process to be transparent (explicit with the nature and kind of values/principles/criteria used for decision-making) and make use of economic evaluation evidence. This would ensure that the choices made reflect Aboriginal social values, and offer value for money, thereby maximising the health of Indigenous Australians. However, given the resource needs of setting up systems and training for the adoption of PBMA, ACCHS would need government support and investment in the implementation of PBMA. Investing in the PBMA exercise is likely to generate more long term benefits for ACCHS and Aboriginal health in general, compared to the initial costs of setting it up and the efforts required in its ongoing application. These benefits include increased transparency, accountability, and improved health care service funding and planning. Adoption of PBMA would aid ACCHS’s efforts to maximise health gain for their constituents.
IV
Declaration: This is to certify that (i)
The thesis comprises only my original work towards the PhD except where indicated in the Preface,
(ii)
Due acknowledgement has been made in the text to all other materials used, and
(iii)
The thesis is less than 100,000 words in length, exclusive of tables, bibliographies and appendices.
Signed:
V
Preface My interest in PBMA application in Indigenous health started in 2001, when the Loddon Mallee Aboriginal Reference Group (LMARG), a loose coalition of Aboriginal Community Controlled Health Services in Loddon Mallee Region, Victoria, asked me to assist them in planning for their services in the region. Ian Anderson, the Director of Onemda, VicHealth Koori Health Research Unit at The University of Melbourne, and I had a meeting with LMARG to discuss what they exactly that they wanted me to do.
It was suggested that PBMA application could assist LMARG members in improving priority setting. However, several Aboriginal community health organisations at that time were either under receivership appointed by government, or facing financial difficulties. It was clear that priority setting was not what the organisations needed at that time to save them from financial and managerial challenges that they faced. It was agreed that, an expenditure study should be undertaken to ascertain the pattern and level of health care service utilisation in the region, and then use the findings in PBMA study at a later date.
The expenditure study, would assist LMARG understand what services Indigenous Australians in the region were utilising and the pattern of such use. Furthermore, the expenditure data would be useful in the program budget component of PBMA study at the latter date. Thus, the expenditure study, reported in Chapter 7, was undertaken prior to the PhD candidature enrolment, and the funding for the data collection came from the Office for Aboriginal and Torres Strait Islander Health (OATSIH) and the Department of Human Services (DHS), Bendigo. The following colleagues undertook the following tasks: 1. Ian Anderson was the overall supervisor of the project and, in particular, he negotiated with the stakeholders such as the DHS Bendigo office, hospitals, pharmacists, the Victorian Division of General Practice, the Aboriginal Community Controlled Health Organisations (ACCHOs) and community health services. 2. Haifa Sekkouah's role included administration support for the project and general research assistance. 3. John Deeble provided expert advice on the methodology and was involved in negotiations with the stakeholders. He was also involved in the writing of the final report.
VI
4. My role included the following: o
Project management – conception, planning and execution;
o
Negotiation with the stakeholders;
o
Literature review;
o
Project methodology development and data collection and analysis; and
o
I wrote the first draft report, and John Deeble edited it.
Immediately after the completion of the expenditure study, I left Onemda and moved to the Centre for Health Policy, Programs & Economics (CHPPE) within the School of Population Health at The University of Melbourne, where I enrolled in a PhD program. Under the auspices of the ACE-Prevention project, I undertook two empirical studies: costing Dry Areas; and the concept of benefit study, to contextualise PBMA in Aboriginal Community Controlled Health Services (ACCHS).
VII
Acknowledgements This thesis has received input from different colleagues who were involved in these research projects in different roles. They are acknowledged below:
The costing of the dry areas (Chapter 7):
This work was done as part of the Assessing Cost Effectiveness (ACE) Prevention project work, with supervisory input from the ACE Prevention Team. The rest of the work was done by me.
Scott Mitchell, NT Police Department; Peter Jones, The NT Liquor Licensing Commission; and Prof Alan Clough provided expert advice on the data for the service activity and unit costs for dry areas intervention.
Concept of benefit (Chapter 8):
ACE-Prevention provided the resources that ensured the success of the workshops.
ACE-Prevention Indigenous Steering Committee and ACE-Prevention research team members (Appendix 3) participated in the workshops.
CRCAH was instrumental in chairing and ensuring that the workshops were a success
Katherine Ong took minutes of some of the workshop meetings.
I would like to thank my primary supervisor, Margaret Kelaher, and co-supervisors Rob Carter, Ian Anderson and Chris Doran. Thank you Margaret for knowing when to push and when to pull back, and for helping me complete my thesis. I admire your generosity and desire to teach me, and am extremely grateful for your investment in me. I would also like to thank Rob Carter for helping me to develop the research question and for allowing me to utilise the ACE-Prevention project resources to undertake two key empirical studies: the costing of Dry Areas; and the Concept of Benefit study. I am thankful for Chris Doran’s insights and contributions to this work, both in terms of broad direction and specific comments. I am grateful to Ian Anderson, firstly for giving me a chance in Aboriginal health, and secondly for guiding me on how things are done in Aboriginal health. On the same note, I am indebted to Onemda, the VicHealth Koori Health Research Unit at The University of Melbourne, where it all began.
VIII
Third party editorial assistance was provided by Rachel Evans of Evans Editing, Sydney. Rachel provided professional copy editing and proofreading of the final copy of this thesis. There are also a number of personal friends who provided comments or assistance on certain aspects of this thesis or initial drafts. These include Libby Smith, Leva Azadi Goodfellow, Haifa Sekkouah, June Perkins, Jeanette Ward, Vita Christie, etc. Thank you one and all. If I have omitted your name, it is probably because I am frantically trying to complete this thesis and hand it in – sorry.
I would like to acknowledge the CRC Aboriginal health for the financial and project support that I received. Specifically, may I thank the CRCAH for the PhD scholarship and the project funding for the survey of decision-makers study that is reported in Chapters 5 and 7. I also offer my thanks to the National Drug and Alcohol Research Centre, University of New South Wales, and the Poche Centre for Indigenous Health at University of Sydney for the logistical support that I received to ensure that I completed my thesis.
Last but not least, profound gratitude to my wonderful family: Shiva; Tara; and Lua, for their support and patience with me. In particular, I must thank Shiva for looking after the home and the kids when I was always away on the PhD thingy. To Tara and Lua, the tickle monster is back! I will watch you doing your homework, and we will play more often! I promise I will tickle you more.
Dedication 1.
To Shiva, Tara and Lua for all their love, support and confidence in me.
2.
To papaka, Silver Opadaka Otim, who started it all, but never lived to see what became of me;
3.
To totoka, Mary Tekit Otim, who supported me when I traversed the planet to realise my dreams; and
4.
To my benefactors, George and Forough Olinga, for their investment in me.
IX
Table of Contents 1.
2.
3.
4.
5.
CHAPTER 1: STATEMENT ON THE RESEARCH QUESTION ......................................................................... 1 1.1
INTRODUCTION .............................................................................................................................................. 1
1.2
AIMS AND OBJECTIVES .................................................................................................................................... 4
1.3
JUSTIFICATION FOR PRIORITY SETTING ................................................................................................................ 5
1.4
THEORETICAL RATIONALE ................................................................................................................................ 7
1.5
PRIORITY SETTING IN INDIGENOUS HEALTH ....................................................................................................... 10
1.6
PBMA AS PRIORITY SETTING TOOL IN INDIGENOUS HEALTH................................................................................. 21
1.7
THESIS SCOPE AND ORGANISATION.................................................................................................................. 23
CHAPTER 2: INDIGENOUS HEALTH CONTEXT ......................................................................................... 28 2.1
INTRODUCTION ............................................................................................................................................ 28
2.2
BACKGROUND ............................................................................................................................................. 29
2.3
HISTORY OF INDIGENOUS POLICY..................................................................................................................... 42
2.4
INDIGENOUS AUSTRALIAN HEALTH CARE SYSTEM ............................................................................................... 48
2.5
SUMMARY: IMPLICATIONS FOR PRIORITY SETTING .............................................................................................. 53
CHAPTER 3: NORMATIVE ECONOMICS FOR PRIORITY SETTING ............................................................. 57 3.1
INTRODUCTION ............................................................................................................................................ 57
3.2
WELFARE ECONOMICS .................................................................................................................................. 59
3.3
NORMATIVE FRAMEWORKS FOR PRIORITY SETTING............................................................................................. 67
3.4
APPLIED WELFARE ECONOMICS IN HEALTH CARE ............................................................................................... 78
3.5
IMPLICATIONS FOR PRIORITY SETTING IN INDIGENOUS HEALTH .............................................................................. 84
CHAPTER 4: NON-ECONOMIC FOUNDATIONS ....................................................................................... 87 4.1
INTRODUCTION ............................................................................................................................................ 87
4.2
IDEOLOGIES IN PRIORITY SETTING .................................................................................................................... 88
4.3
PRINCIPLES OF DISTRIBUTIVE JUSTICE FOR PRIORITY SETTING ................................................................................ 91
4.4
PROCEDURAL JUSTICE ISSUES IN PRIORITY SETTING ............................................................................................. 98
4.5
KEY ISSUES FOR CONSIDERATION IN PRIORITY SETTING ........................................................................................ 99
4.6
SUMMARY ................................................................................................................................................ 102
CHAPTER 5: PERCEPTIONS OF DECISION-MAKERS ............................................................................... 104 5.1
INTRODUCTION .......................................................................................................................................... 104
5.2
METHODS ................................................................................................................................................ 104
5.3
FINDINGS.................................................................................................................................................. 109
5.4
DISCUSSION .............................................................................................................................................. 119
5.5
SUMMARY ................................................................................................................................................ 126
X
6.
CHAPTER 6: GUIDELINES FOR PRIORITY SETTING ................................................................................ 127 6.1
INTRODUCTION .......................................................................................................................................... 127
6.2
METHODS: GUIDELINES FOR PRIORITY SETTING................................................................................................ 127
6.3
THE CHECKLIST FOR PRIORITY SETTING ........................................................................................................... 136
6.4
ASSESSING SELECTED NON-ECONOMIC MODELS .............................................................................................. 139
6.5
ASSESSING SELECTED ECONOMIC APPROACHES ................................................................................................ 143
6.6
DISCUSSION .............................................................................................................................................. 151
7.
CHAPTER 7: PBMA APPRAISAL ............................................................................................................ 158 7.1
INTRODUCTION .......................................................................................................................................... 158
7.2
THE PROGRAM BUDGETING & MARGINAL ANALYSIS APPROACH ......................................................................... 159
7.3
PBMA APPRAISAL: SURVEY OF DECISION-MAKERS ........................................................................................... 167
7.4
EXPENDITURE ON SERVICES FOR INDIGENOUS AUSTRALIANS ............................................................................... 175
7.5
ASSESSING COST EFFECTIVENESS (ACE) APPROACH .......................................................................................... 187
7.6
COSTING DRY AREAS ALCOHOL INTERVENTION ................................................................................................ 189
7.7
IMPLICATIONS FOR PBMA IN INDIGENOUS HEALTH .......................................................................................... 205
8.
CHAPTER 8: EXPANDING THE CONCEPT OF BENEFIT ........................................................................... 210 8.1
INTRODUCTION .......................................................................................................................................... 210
8.2
METHODS AND RESULTS.............................................................................................................................. 224
8.3
DISCUSSION .............................................................................................................................................. 235
8.4
SUMMARY ................................................................................................................................................ 239
9.
CHAPTER 9: PBMA GUIDELINES IN INDIGENOUS HEALTH.................................................................... 242 9.1
INTRODUCTION .......................................................................................................................................... 242
9.2
GENERATING PBMA GUIDELINES .................................................................................................................. 243
9.3
CHECKLIST FOR PBMA IMPLEMENTATION IN INDIGENOUS HEALTH....................................................................... 255
9.4
USING THE CHECKLIST: IMPLEMENTING PBMA IN ACCHS................................................................................. 259
9.5
THESIS SUMMARY ...................................................................................................................................... 269
10.
REFERENCES........................................................................................................................................ 273
11.
APPENDICES ....................................................................................................................................... 293
11.1
APPENDIX 1: DEFINING DIMENSIONS OF BENEFIT ........................................................................................ 293
11.2
APPENDIX 2: DESCRIPTORS OF THE DIMENSIONS OF BENEFIT ......................................................................... 300
11.3
APPENDIX 3: ACE-PREVENTION INDIGENOUS STEERING COMMITTEE MEMBERS ............................................... 307
11.4
APPENDIX 4: EXPENDITURE STUDY............................................................................................................ 308
11.5
APPENDIX 5: SURVEY OF DECISION-MAKERS ............................................................................................... 312
XI
List of Tables Table 1.1: Examples of criteria for priority setting from selected countries ........................... 17 Table 1.2: Summary of the PhD Chapters ................................................................................ 23 Table 2.1: Estimated Resident Population (a), Indigenous–2001 and preliminary-2006 ........ 37 Table 2.2: Highest school year attended, Indigenous -15 years and over (a)–2001 and 2008 39 Table 2.3: Self-assessed health status, Indigenous persons aged 15 years and over, by sex – 1994 and 2004-05 ..................................................................................................................... 41 Table 4.1: Attitudes typically associated with libertarianism and egalitarianism .................... 90 Table 4.2: Some philosophers’ and economists’ concept of equity ........................................ 96 Table 5.1: Background information expressed as both numbers and percentages .............. 107 Table 5.2: Structure of the questionnaire for surveying the key decision-makers ................ 108 Table 5.3: Department of Human Services objectives ........................................................... 110 Table 5.4: Data sources for priority setting in the Department of Human Services .............. 111 Table 5.5: Criteria for priority setting ..................................................................................... 111 Table 5.6: Concept of benefit from health care programs ..................................................... 112 Table 5.7: Assessing the priority setting process ................................................................... 113 Table 5.8: Challenges of priority setting process in Indigenous health ................................. 113 Table 5.9: Improving the priority setting process .................................................................. 114 Table 5.10: ACCHS objectives ................................................................................................. 115 Table 5.11: Data sources for priority setting .......................................................................... 115 Table 5.12: Criteria for priority setting across and within programs ..................................... 116 Table 5.13: Strengths of the current priority setting process ................................................ 117 Table 5.14: Challenges (across and within programs) ............................................................ 117 Table 5.15: Improving the priority setting process ................................................................ 118 Table 6.1: Performance of the league tables approach ......................................................... 145 Table 6.2: Performance of HSW model .................................................................................. 147 Table 6.3: Performance of HBG/HRG approach ..................................................................... 148 Table 6.4: Performance of PBMA against the checklist ......................................................... 150 Table 6.5: Summary of the performance of the selected non-economic frameworks .......... 156 Table 7.1: Stages in PBMA priority setting exercise ............................................................... 161 Table 7.2:PBMA potential ....................................................................................................... 168 Table 7.3: Assessment criteria for PBMA ............................................................................... 170
XII
Table 7.4: Potential challenges for PBMA .............................................................................. 171 Table 7.5: PBMA potential to all respondents ....................................................................... 172 Table 7.6: GPs with addresses in the LMR: Response rates and survey expansion factors, by type of practice and Division .................................................................................................. 178 Table 7.7: Data from respondent GPs, three weeks, April-May, 2004 .................................. 179 Table 7.8: Responses to the pharmacy survey, April-May 2004 ............................................ 179 Table 7.9: Items dispensed by respondent pharmacies, three weeks, April-May, 2004 ....... 179 Table 7.10: Reported in-patient separations, LMR, 2003-04 (including 10% WIES loading for Aboriginality in Victoria) ......................................................................................................... 181 Table 7.11: Reported emergency and outpatient attendances, LMR, 2003-04..................... 181 Table 7.12: Estimated admitted patients (adjusted for under-identification): Total separations; WIES Value, LMR, 2003-04. ............................................................................... 182 Table 7.13: Emergency and out-patient departments, estimated total visits, LMR, 2003-04 ................................................................................................................................................ 182 Table 7.14: Community health services, by type of service, estimated visits and costs, 20032004 ........................................................................................................................................ 183 Table 7.15: Estimated GP services, LMR, 2003-04, by source of service and Medicare benefit cost ......................................................................................................................................... 184 Table 7.16: Estimated prescriptions dispensed, LMR, 2003-04, by PBS category, PBS benefits and total cost .......................................................................................................................... 184 Table 7.17: Estimated expenditures by type by service, total and as averages per Indigenous person, LMR, 2003-04 ............................................................................................................ 184 Table 7.18: List of interventions in alcohol abuse .................................................................. 190 Table 7.19: Social costs resulting alcohol NT (2004/05 prices) .............................................. 200 Table 7.20: Costing one dry area intervention ....................................................................... 201 Table 7.21: Uncertainty in costing one dry areas intervention .............................................. 203 Table 8.1: Summary of the dimensions of health .................................................................. 216 Table 8.2: Weights of the dimensions of benefit from the ACE Prevention ISC members .... 232 Table 8.3: Importance weights for Indigenous services in community health services ........ 233 Table 8.4: Weight of the DALY in the Individual Health Gain ................................................. 233 Table 9.1: Application of PBMA checklist on the LMARG hypothetical PBMA study ............ 267 Table 11.1: The EuroQol 5D+ classification for health status. ............................................... 294
XIII
List of Figures Figure 2.1: Social determinants of good health ....................................................................... 31 Figure 2.2: Historical impacts of colonisation upon Indigenous health ................................... 33 Figure 7.1: Dry areas alcohol intervention pathway .............................................................. 196 Figure 8.1: Definition of Indigenous health ............................................................................ 215 Figure 8.2: Calibration of the descriptors ............................................................................... 230 Figure 8.3: Weighting Using the Global Approach ................................................................. 232
XIV
List of Abbreviations ABS ACCHS ACE ADC AHEC AHMAC
AIHW AQoL ATSIC
Australian Bureau of Statistics Aboriginal Community Controlled Health Service Assessing Cost Effectiveness Aboriginal Development Commission Australian Health Ethics Committee Australian Health Ministers’ Advisory Council Australian Institute of Health and Welfare Assessment of Quality of Life Aboriginal & Torres Strait
CUA CV DAA DALY DHSH
Services and Health DIMIA
Multicultural
DMA DRG ERC
CMA CRS CRC CRCATH
and
Decision-making Approach Diagnostic Related Groups Expenditure Review Committee
EV GP HACC
Islander Services
CEO CGC
of
Indigenous Affairs
ATSIHWIU Aboriginal and Torres Strait Islander Health and Welfare Information Unit ATSIS Aboriginal & Torres Strait
CBA CEA
Department Immigration,
Islander Commission
CAA
Cost Utility Analysis Compensating Variation Department of Aboriginal Affairs Disability-Adjusted Life Year Department of Human
HBG/ HRG HREOC
Council of Indigenous Affairs Cost Benefit Analysis Cost Effectiveness Analysis Chief Executive Officer Commonwealth Grants Commission Cost Minimisation Analysis Category Rating Scale Cooperative Research Centre Cooperative Research Centre for Aboriginal & Tropical Health
HRSCAA
HSM HSW ICER LMR LMARG MSB
XV
Expected Value General Practitioner Home and Community Care Health Benefit Groups/ Health Resources Groups Human Rights and Equal Opportunities Commission House of Representatives Standing Committee on Indigenous Affairs health status measures Health Sector-Wide Incremental Cost Effectiveness Ratio Loddon Mallee Region Loddon Mallee Aboriginal Reference Group Marginal Social Benefits
MSC MAU NACCHO
NAHS NAHSWP
NAIHO
NATSIHC
NATSIHS
NATSISS
NGOs NHMRC NHS NSF
Marginal Social Costs Multi- Attributes Utility National Aboriginal Community Controlled Health Organisation National Aboriginal Health Strategy National Aboriginal Health Strategy Working Party National Aboriginal and Islander Health Organisation National Aboriginal & Torres Strait Islander Health Commission National Aboriginal & Torres Strait Islander Health Survey National Aboriginal & Torres Strait Islander Social Survey Non-Government Organisations Nation Health & Medical Research Council National Health Service National Strategic
Access Program PPBS PTO QALY RCADC
RS SG SMPH SMR SWF SWOT
TTO VAED VAS VEMD VicHealth WHO
Framework OAA OATSIH
WIES
Commonwealth Office of Indigenous Affairs Office of Indigenous and Torres
Strait
YLD YLL
Islander
Health Services PHCAP
Primary
Health
Care
XVI
Planning Programme Budgeting Systems Person Trade-Off Quality-Adjusted Life Years Royal Commission into Aboriginal Deaths in Custody Rating Scale Standard Gamble Summary Measures of Population Health Standardised Mortality Rate Social Welfare Function Strengths, Weaknesses, Opportunities and Threats Time Trade-Off Victorian Admitted Patient Database Visual Analogue Scale Victorian Emergency Minimum Database Victorian Health Promotion Council World Health Organisation Weighted In-lier Equivalent Separation Years Lived with Disability Years of Life Lost
Part A: Background to the Research Question
1. Chapter 1: Statement on the Research Question 1.1 Introduction Aboriginal and Torres Strait Islander2 peoples have significantly poorer health outcomes than non-Indigenous Australians. At the national level for 2005-2007, the gap between Indigenous and non-Indigenous life expectancy was 11.5 years for males and 9.7 years for females (ABS and AIHW 2010). Indigenous Australians often have limited or no access to quality health care services (Brennan and Carter 1998; Anderson, Clarke et al. 2002; Cass, Cunningham et al. 2003; Cass, Cunningham et al. 2003). Where such services exist, Indigenous Australians often face structural problems from the Australian health system such as institutional racism and culturally inappropriate services (Henry, Houston et al. 2004).
2
Aboriginal and Torres Strait Islander peoples, sometimes referred to as Aborigines, Indigenous peoples or Indigenous Australians in this thesis, are a group of Indigenous Australians who have two distinct cultural identities: Aboriginal; and Torres Strait Islander groups. They are the natives or first inhabitants of Australia, and are referred to as ‘peoples’ because of the distinct cultural identities amongst themselves whether they live in urban, regional or remote areas of Australia. The definition for being Indigenous that is adopted in this thesis is the definition approved by NACCHS Board November 2007. It is a three-part definition requiring all three parts to be established for one to be recognised as an Indigenous Australian NACCHO. (2008). "Defintions: Aboriginality." Retrieved 8.4.2010, 2009, from http://www.naccho.org.au/definitions/aboriginality.html. :
One’s parent is of Indigenous or Torres Strait Islander descent; The individual identifies oneself as an Indigenous or Torres Strait Islander; and The individual is accepted as such by the Indigenous or Torres Strait Islander community in which he/she lives.
Page 1
The Australian government assumed primary responsibility for the health care services for Indigenous Australians following the 1967 referendum. Until then, the responsibility for the health of Indigenous Australians resided with State and Territory governments (Anderson 2008). Despite substantial reform, the health status of Indigenous Australians is still well below that of other Australians. The Australian government has acknowledged that the “health gap” between Indigenous Australians and others is inequitable, and has made the pledge of “closing the gap” (COAG 2009). To achieve that pledge requires either more health reform, and/or improved priority setting processes. Priority setting can be defined as a non-market based process of making choices in the face of scarcity (Al et al., 2004; Klein, 1998; Spiegel and Hyman, 1978). Scarcity is the concept on which the discipline of economics is founded, and it is defined as the shortage that arises from the demand for a good or service (Earl-Slater 1999; Culyer 2005).This thesis focuses on improved priority setting and the use of economic evaluation evidence in decision-making. Improved priority setting from an economic perspective, means that the allocation of resources ensures that the objectives being pursued by the decision-makers are maximised or optimised (Ham and Coulter 2000). Priority setting can be used for the following purposes:
Ranking needs and/or health care services in the community;
Deciding which services can be expanded and which ones can be scaled down;
Developing guidelines for decision-making;
Making the dollar go further; and
Deciding whether to purchase expensive treatments or not.
The current priority setting processes used by the government to address Indigenous health are not explicit. Explicit priority setting is when the choices, and the reasons for those choices, are publicly known. The reasons behind the government priority setting process in Indigenous health in which expenditure does not appear to reflect the special health needs of Indigenous Australians, are not clearly known. For example, excess total expenditure per Aboriginal person compared to other Australians was 8% in 1995/96; 22% in 1998/99; 18% in 2001/02; and about 17% in 2004/05 (Deeble,
Page 2
Mathers et al. 1998; AIHW 2001; AIHW 2005; AIHW 2008). It is therefore difficult for government to justify such marginal increases in funding given the health gap.
Other priority setting approaches exist, loosely referred to as implicit priority setting approaches, and this is when the choices and the reasons for such choices may not be publically known (see Section 1.5.1). From an economic point of view, explicit measures can ensure that government efforts aimed at closing the gap can be done efficiently3. The adoption of the explicit approaches in this thesis is because of its ability to embrace economic principles and its potential to allow decision-makers to be held accountable for the decisions that they make.
The key challenge however, is how to undertake explicit priority setting Indigenous health, because it has not been investigated before. Although mainstream frameworks can be adopted, these frameworks need to be modified to incorporate Indigenous Australians’ values and aspirations. The focus of this research is the development of a set of guidelines that may be used to guide explicit priority setting in Indigenous health. It is envisaged that these guidelines may also be used to qualitatively assess the performance of priority setting models in Indigenous health. One model that is potentially acceptable in Indigenous health is Program Budgeting and Marginal Analysis (PBMA).
PBMA is an economic approach to priority setting that has the potential to incorporate social justice4 issues, including fairness, into the decision-making process. It has the potential to increase the meaningful involvement of the Indigenous community in decision-making. However, PBMA is a Western concept. Certain aspects of PBMA, such as the concepts of consultative and collaborative decision-making, are consistent with Indigenous values; other aspects, such as the disregard for non-medical benefits, may not be acceptable or relevant to Indigenous health. Aspects of PBMA that may not be
3
Efficiency (often referred to as value for money) is an economic concept which means achieving the stated objectives in such a way that outcomes are maximised or optimised for the given budget (see Chapter 4 for details). 4 Under social justice, emphasis is on the fairness of the process and equitable distribution of outcomes of allocation of resources. What is fair and what is equitable is discussed in detail in Chapter 4.
Page 3
consistent with Indigenous values will be contextualised to Indigenous health. I will develop a checklist that can be used to guide the successful implementation of PBMA in Indigenous community controlled health care services, based on the contextualisation of PBMA.
1.2 Aims and Objectives This thesis has two aims. The first is to develop guidelines for priority setting in Indigenous health. Developing guidelines for explicit priority setting in Indigenous health will involve addressing the following objectives: 1. To identify the priority setting criteria currently used in Indigenous health; 2. To establish the uptake of economic evaluation evidence in Indigenous health; 3. To identify the key priority setting issues in Indigenous health; 4. To highlight the appropriate place for economic theory in Indigenous health; 5. To identify social justice principles that ought to guide priority setting; and 6. To establish guidelines for priority setting to be used in Indigenous health.
The above objectives will be addressed using an empirical study: a survey of key decision-makers to establish key principles/criteria and other considerations for priority setting. Findings from this study, together with the review of economic and social justice principles, will be used to develop guidelines for explicit priority setting in Indigenous health. These guidelines will form the key elements of an ideal priority setting framework in Indigenous health, and will be used to qualitatively assess PBMA and other existing priority setting models.
The second aim is to contextualise PBMA, the chosen model, in Indigenous health. This will involve addressing the following key objectives: 1. To establish the likely level of acceptability and useability of PBMA in Indigenous health; 2. To identify the issues that ought to be addressed when PBMA program budget is being estimated; 3. To identify potential issues from the costing of Indigenous interventions;
Page 4
4. To identify the appropriate concept of benefit for priority setting in Indigenous health; 5. To establish the most effective way of involving Indigenous Australians in the decision-making process; and 6. To identify the checklist that should guide the successful PBMA application in Indigenous health.
Addressing these objectives will involve undertaking four empirical studies: a survey of decision-makers to appraise The PBMA framework; an expenditure study to demonstrate program budgeting in PBMA; the costing of the dry areas alcohol intervention to demonstrate issues of costing services in Indigenous health; and the concept of benefit study. Results from these empirical studies will be used to create a checklist for contextualising PBMA in Indigenous health. This checklist will provide the key steps and issues for consideration in order to successfully implement PBMA model in Indigenous health.
1.3 Justification for Priority Setting The justification for priority setting in Indigenous health can be traced to the need to improve Indigenous health, and the need to maximise the returns from investments in Indigenous health. Recent epidemiological evidence (Vos, Barker et al. 2007; Vos, Barker et al. 2009) suggests that there is an unacceptable health gap between Indigenous Australians and other Australians. For example, the burden of disease suffered by Indigenous Australians is estimated to be about two-and-a-half times that of the rest of the Australian population (ABS and AIHW 2008). Chronic illnesses, such as cardiovascular disease, diabetes, mental disorders and chronic lung disease, are responsible for over 70% of this burden (Vos, Barker et al. 2007). This is not fair for Indigenous Australians.
The poor health status of Indigenous Australians can be attributed to low socioeconomic status (SES), lack of access to good quality health care services and other factors such as colonisation (see Chapter 2 for a detailed discussion). Majority of
Page 5
Indigenous Australians face poor economic conditions which are characterised by low incomes, high unemployment rates, and low rates of home ownership compared to non-Indigenous Australians (ABS and AIHW 2008). Households are overcrowded with inadequate water, poor sanitation and sewage disposal, and limited food storage, compared to other Australians (ABS and AIHW 2008). Such living conditions affect Indigenous Australians’ ability to lead lifestyles that encourage good health. These poor living conditions, as well as living long distances from health care services and a limited ability to speak the English language also affect Indigenous Australians’ access to health care services, such as specialist care and pharmaceutical services. This situation demonstrates the poor performance of the Australian health system in meeting the needs of Indigenous Australians (CGC 2001).
There are several reasons for the government’s failure to improve the health of Indigenous Australians to comparative levels with other Australians. These reasons range from historical and socio-economic issues that Indigenous Australians face and have faced over the years (Bartlett and Legge 1994; Duckett 2004; ABS and AIHW 2008), to poor priority setting methods that not only fail to deliver value for money services (Ross 1995; McKenzie, Warrior et al. 1998; van Gool, Gallego et al. 2006) to Indigenous Australians, but also perpetuate the existing injustices (Henry and Mooney 2003; Henry, Houston et al. 2004) (discussed in detail in Chapter 2). It is argued in this thesis that government priority setting methods ought to be explicit and should make use of economic evaluation evidence.
In admission of its poor performance, the Australian government recently committed itself to closing the gap between life expectancy and child mortality rates for Indigenous Australians and non-Indigenous Australians within two generations (COAG 2009). However, the current Australian government priority setting process in Indigenous health is not explicit. That is, the choices made by government and the basis for those choices are often not clear and are unpublished. The lack of transparent decision-making around public expenditure has left the government vulnerable to criticism about the way it conducts priority setting. This provides a strong argument for instituting an explicit priority setting process in Indigenous health.
Page 6
Many health economists have been advocating for priority setting in health care to be based on explicit approaches as opposed to implicit ones (Mooney, Gerard et al. 1992; Coast and Donovan 1996; Maynard 1996). This would ensure that the losers and winners in the process are identified, and the reasons for their losses or gains are clearly published, since priority setting can be referred to as “who gets what at whose expense” (Williams 1988). The term ‘priority setting’ is often referred to as rationing5 or resource allocation6. The terms ‘priority setting’, ‘rationing’ and ‘resource allocation’ may be used interchangeably. These terms all refer to the non-market based process of making choices in the face of scarcity in such a way that the benefits from the goods and services are maximised.
In summary, the need to address the health disparities between Indigenous Australians and other Australians constitutes a justification for improved priority setting. The associated levels of government expenditure, given the level of need in Indigenous health, strengthen the need for such a priority setting process. Other reasons that justify the need for priority setting may include: doing more with available resources in Indigenous health, in terms of selection of interventions which offer value for money; and improving the evidence base on what works in Indigenous health.
1.4 Theoretical Rationale The need for priority setting stems from the economic notion that the resources which are required to address all the needs/objectives of a society are scarce. This requires decision-makers to make choices about what to fund/provide given the identified
5
Rationing is the controlled distribution of resources and scarce goods or services: it restricts how much people are allowed to buy or consume. Rationing controls the size of the ration, one's allotted portion of the resources being distributed on a particular day or at a particular time. 6 Resource allocation refers to a condition whereby resources are allocated in a way that maximises the net benefit attained through their use. Allocative efficiency refers to a situation in which the limited resources of a country are allocated in accordance with the wishes of consumers. Allocatively efficient economy produces an ‘optimal mix’ of commodities.
Page 7
objectives of the society in question. Achieving such objectives entails taking two theoretical rationales: social justice; and economic rationales. Economic rationale as a basis for priority setting stems from the need for decision-makers to make choices about what objectives or programs to pursue when faced with a scarcity of resources. Simply put, Indigenous health needs are numerous and yet in any given year, there are limited resources to address all these needs. This requires decision-makers to choose programs to implement and which ones not to implement.
Making such choices involves trade-offs in which the benefits from the programs that are not chosen are forgone. These forgone opportunities are what economists call ‘opportunity costs’. The opportunity cost of mutually exclusive interventions is the benefits forgone from the next best use of the resource or the value of the resource in its alternative highly valued use (Earl-Slater 1999; Culyer 2005). The notion of ‘opportunity cost’ plays a crucial role in marginal analysis. Marginal analysis focuses on incremental changes to resources, programs, cost or outcomes of a service in order to maximise health outcomes to the population targeted, given the allocated budget (Earl-Slater 1999). For example, it challenges the decision-makers to focus on what happens to the policy objective when a little less or more of an activity takes place.
Marginal analysis is a tool used in economic evaluation for assessing evidence of efficiency of the health care interventions (Earl-Slater 1999; Culyer 2005). Economic evaluation of health services has been defined as “the comparative analysis of alternative courses of action in terms of both costs and outcomes” (Drummond, Sculpher et al. 2005). The decision-makers need information on the efficiency of interventions to ensure that the costs are minimised, or the benefits are maximised. Thus, in recognition of the scarcity of resources, given the multitude of health problems Indigenous Australians face, this thesis argues that the choices made about what health services to provide and which ones not to should maximise the health of Indigenous Australians. This implies that economic evaluation evidence should be used when priority setting is being done in Indigenous health, so as to minimise the opportunity costs. Additionally, decision-makers need information on the nature of the
Page 8
distribution of such benefits and/or opportunity costs of such activities, a form of equity concern. Equity7 is a key principle in social justice literature, and it may include fairness/transparency of both the process and outcomes of the allocation of resources (Olsen 1997; Konow 2003), and discussed in Chapter 4. Social justice principles suitable for priority setting can be loosely grouped into distributive and procedural justice principles (discussed in Chapter 4 in detail). Distributive justice principles can vary in what is to be distributed, in the nature of who benefits, and on what basis the distribution should be made, either based on equality, maximisation, according to individual characteristics or according to free transactions (Lamont and Favor 2007). Distributive principles of social justice aim to achieve a fair and just outcome of health maximisation (van Doorslaer, Wagstaff et al. 1993). In Indigenous health, these principles may include efficiency, equality, need and equity.
In procedural justice, the focus may be on the fairness of the processes, how decisions are made, who is involved and what factors were considered. Procedural justice is about carrying out a set of rules in a just and fair manner so that a just outcome might be reached. Fairness ensures that the process of priority setting generates unbiased, consistent, and reliable decisions (Maiese 2003). An unbiased process, which is used to distribute wealth or deliver decisions, can potentially ensure impartiality, as well as consistency in the decision-making process. Thus, procedural justice would ensure the legitimacy of the decisions reached and the community’s acceptance of those decisions. In summary, opportunity cost, marginal analysis, equity and other social justice principles form the key principles that ought to guide priority setting in Indigenous health.
7
Equity involves ethical judgements on the fairness of the distributions of benefits/costs, fairness of the process of the allocation of resources, access, etc Culyer, A. (2005). The Dictionary of Health Economics. Cheltenham, Edward Elgar.. It is more than just the distribution of benefits/costs of the policy actions.
Page 9
1.5 Priority Setting in Indigenous Health 1.5.1 Key Issues in Priority Setting Priority setting spans various disciplines such as: health economics, political science, politics, and bioethics. It is generally agreed in these disciplines that priority setting in health is inevitable, but the preferred approach, explicit or implicit approach, is strongly contested in the international literature. The debate has been fuelled by significant attempts at priority setting in a range of countries and settings, such as the Oregon experiment, the priority setting exercises in the Nordic countries, New Zealand, the United Kingdom, Israel and Australia (Carter 2002). These issues include: 1.
Whether an explicit priority setting approach is preferred to an implicit one;
2.
Whether different approaches should be used at different decision-making levels;
3.
Whether technical rigour is more desirable than due process8; and
4.
Whether the Indigenous community should be involved in priority setting.
(Mooney, Gerard et al. 1992; Mechanic 1995; Mechanic 1997; Daniels 2000; Ham and Coulter 2000).
The first issue debated in the international literature involves the choice between explicit and implicit approaches to priority setting. The proponents of implicit approaches argue that because the decisions for setting priorities themselves are not necessarily known, and the bases for such decisions are not clearly set out, implicit priority setting is “more conducive to stable social relations and a lower level of conflict” (Mechanic 1995).
Critics of the implicit approach, especially economists, describe the implicit priority setting approach as “muddling through elegantly” (Hunter 1996). They argue that the choices made by decision-makers and the guidelines/reasons behind those choices
8
The due process occurs when decision-makers state decisions explicitly, the basis for those decisions, how they arrived at the decisions, and what resources will be required Earl-Slater, A. (1999). Dictionary of Health Economics. Abingdon, Oxon, Radcliffe Medical Press..
Page 10
should be clearly known (Klein, Day et al. 1996; Ham 1997; Ham and Pickard 1998; Klein 1998). This ensures that the priority setting process is transparent and would encourage debate on the appropriateness of the identified criteria. They further argue that such explicit approaches emphasise technical methods, which in turn ensure that evidence, in the form of effectiveness, equity, efficiency or concept of benefit, plays a key role in decision-making. Additionally, in order to ensure the maximisation of social welfare, evidence ought to be constrained by the availability of scarce resources so as to aid decision-makers to make trade-offs. Such evidence-based trade-offs would therefore reduce the potential for government failure in addressing Indigenous health problems. The preferred approach to priority setting in health economics, and adopted in this thesis, is the explicit approach.
However, explicit approaches may not be politically acceptable or feasible at all levels of the health system. The choice between the explicit and implicit priority setting approaches therefore, may be strongly influenced by the different levels decisionmaking, because the needs and institutional dynamics at each level of government may be different (Shiell and Mooney 2002). For example, explicit approaches are suitable in government decision-making levels, while implicit ones may be suitable at doctor-patient. Three major levels of decision-making exist in Australia: macro, meso and micro. At Federal/State/Territory level, decisions are mainly macro level decisionmaking. The meso levels involve the large institutions that deliver health care services, such as regional health authorities, hospitals and ACCHS. Micro level decision-making involves priorities set by clinicians and Indigenous health workers, which may require confidentiality or some other form of sensitivity. The appropriateness of explicit or implicit priority-setting however, should not be the reason for lack of transparency or lack of accountability in the decision-making process. The focus of this thesis is the decision-making process at ACCHS level and at the government level.”
Another debate within the literature on priority setting, involves the choice of technical rigour versus due process as a preferred mechanism undertaking priority setting. The pursuance of technical rigour, such as efficiency or evidence-based medicine, is characterised by a reliance on rational decision-making rules and the
Page 11
development of technical frameworks in which they are applied (Carter 2002; Carter, Vos et al. 2008). Proponents of such methods argue that, provided the criteria for priority setting are acceptable, applying the correct rules makes the decisions legitimate (Carter 2002). Members of the technical school tend to rely on rational decision rules and/or technical frameworks, and the technical school is dominated by health economists, who tend to focus on pursuing the goal of efficiency, and epidemiologists/clinicians who tend to focus on evidence-based medicine and/or needs-based equity (Carter 2002).
The due process school of priority setting questions the possibility of devising rational decision rules. It advocates the role of judgement in making decisions and the use of bargaining to arrive at decisions (Klein 1993; Klein 1995; Klein, Day et al. 1996). The challenge for the due process school, however, is to construct a process that enables proper debate and discussion about the role of judgement, how bargaining can be done, and how to use available evidence in decision-making.
The need for a combined approach that reflects economic technical rigour, empirical experience and due process is therefore appealing. With increasing recognition of the need for informed judgement comes a focus on consultation informed by evidence. This has led to an appreciation that decision-making rules can rarely be followed automatically. In the case of Indigenous health, this thesis argues that both technical and due process are needed. This is reflected in the choice of the PBMA framework which embodies both technical rigour and due process.
The use of a combined technical and due process approach to an explicit priority setting process generates debate on whose judgement ought to be relied upon when priorities in health are being set. The debate centres on the involvement of the public in the priority setting process. Should the views and values of Indigenous Australians as consumers, all Australians or only key decision-makers determine the priorities in Indigenous health? This thesis argues that Indigenous values and principles, together with those of the decision-makers, should guide the priority setting exercise at all levels of the decision-making process, albeit in different ways. This is consistent with
Page 12
the Indigenous concept of ‘community control’ and self-determination. Another issue of contention involves what to consult about and what the appropriate process for consultation should be. These issues will be investigated in Chapter 4 using a survey instrument, and in Chapter 8 using a workshop-based approach to expand the concept of benefit from health interventions to include non-medical benefits.
In summary, although explicit priority setting is not necessarily and universally more valid and more reliable than implicit judgement, this thesis argues that an explicit priority setting approach that makes use of both technical and due process, has potential to improved the health of Indigenous Australians, for three reasons. First, decisions and the criteria/reasons/values behind such decisions will be transparent. Second, evidence such as economic evaluation is important in assisting decisionmakers select interventions that work and also offer value for money. Third, in line with the Indigenous principle of self-determination and government principle, the due process enables the need for meaningful involvement of Indigenous Australians in the decision-making (NATSIHC 2003). This helps to ensure that Indigenous values are incorporated in the choices made by the decision-makers, to ensure Indigenous ownership of the process and the results.
1.5.2 Explicit Priority Setting in Indigenous Health How explicit priority setting should be undertaken in Indigenous health, given the clearly articulated objectives in Indigenous health, and Indigenous values and principle of self-determination is a key challenge. Several views and approaches exist on how explicit priority setting should be undertaken. The approach taken in this thesis involves those four key issues to be addressed in order for an explicit priority setting process to appropriately take place in Indigenous health (Earl-Slater 1999; Mooney and Wiseman 1999; Shiell and Mooney 2002; Carter, Vos et al. 2008):
The establishment of the Indigenous health ‘constitution’;
Strengthening of the Indigenous evidence base for decision-making;
The adoption of a decision-making mechanism to resource allocation; and
The establishment of an appropriate incentive and institutional structure.
Page 13
The first issue, the establishment of the Indigenous health ‘constitution’, requires ascertaining the broad principles or criteria that should guide the health system and the trade-offs that the Indigenous Australians are prepared to make. These principles/criteria can be used for determining the extent to which programs or interventions should be publicly provided or financed in health, especially in Indigenous health. They can also be used in the selection of the options for change in the decision-making context, and to inform the development of the checklist for an ‘ideal’ priority setting framework in Indigenous health.
The development of these criteria will involve two key sources of information. First, empirical data from the preferences of the decision-makers, including Indigenous Australians or their representatives, investigated in Chapter 3. Secondly, the theoretical basis for decision-making as a normative issue is addressed in Chapters 4 and 5. The normative issue involves information from economics, such as opportunity cost and marginal analysis, and social justice.
The second issue, strengthening the evidence base to guide priority setting, requires improvements in the underlying forms of evidence, such as economic evaluation and epidemiological evidence, and improvement in the quality and the measurement rigour of such evidence. This thesis will investigate ways to improve the measurement rigour of costs (Chapter 7) and benefits (Chapter 8) in Indigenous health care services. The commitment to provide this evidence as an integral component of explicit priority setting will also be discussed. Furthermore, the criteria for assessing the quality of evidence might need to be widened and strengthened, so that there is no penalty for interventions in Indigenous health, where evidence to support those interventions is not easily obtained (Rychetnyk, Frommer et al. 2002; Sibthorpe, Bailie et al. 2002). This issue will be discussed in Chapter 8, to improve measurement rigour in the concept of benefit.
The third issue that will be addressed in an explicit priority setting framework involves the need for practical decision-making mechanisms for setting priorities to ensure that resource-allocation decisions are consistent with the principles embodied in the health
Page 14
‘constitution’. One such mechanism is PBMA, which will be contextualised in Section C of this thesis. Contextualising PBMA will ensure that PBMA methods are informed by Indigenous values and aspirations. Furthermore, this contextualisation will ensure that key issues that may arise from the empirical studies are noted and possible solutions put in place. These issues may include appropriateness of PBMA as a tool in Indigenous health, how to source evidence, how to identify options for change, and what the appropriate concept of benefit should be. Addressing these issues through the involvement of both decision-makers and Indigenous Australians will enhance transparency and accountability, and thereby ensure that the recommendations generated by the priority setting exercises are acceptable and can be acted upon. This will contribute to the establishment of the institutional structures that are needed to support a systematic approach to priority setting (Ham and Coulter 2000). In summary, the contextualised PBMA is the decision-making mechanism that is informed by the Indigenous health constitution; it makes use of economic and other evidence, and allows the Indigenous population meaningful involvement in the decision-making mechanism.
1.5.3 Criteria for Explicit Priority Setting in Health The constitution of health services refers to the statement of the broad principles or criteria that should guide the health system and the trade-offs that the citizens are prepared to make between the principles (Shiell and Mooney 2002). This thesis focuses on identifying only the criteria for decision-making, without establishing the associated trade-offs due to limited time, and lack of resources (undertaken in Chapter 5). An exploratory study on the trade-offs will be undertaken in Chapter 8 as part of the concept of benefit for priority setting. The views of the decision-makers and the Indigenous Australians and/or their representatives were elicited to inform the development of the guidelines for priority setting in Indigenous health.
International studies indicate that different countries use different criteria to guide priority setting and the most common criteria are listed below in Box 5.1 (Kapiriri and Norheim 2004). Most of the debate around these criteria has been carried out internationally, and may be far different from the realities of Indigenous communities
Page 15
in Australia. The fact that Indigenous community needs far outstrip allocated resources and the health construct (cultural values and characteristics) requires the rethinking of the criteria for priority setting, this chapter aims to list some possible criteria that could be discussed in the Australian context.
Box 5.1: Some of the criteria that are most widely used and debates for priority setting
Source: (Kapiriri and Norheim 2004)
Some countries publish the criteria that they use for priority setting (Table 5.1). For example, in New Zealand, the literature indicated that the criteria for priority setting in Maori health included effectiveness, equity, efficiency and acceptability (Ashton, Cumming et al. 2000). In Sweden, like most Nordic countries, the principle of human dignity, need and solidarity feature most in the list of criteria for priority setting. The publication of these criteria served to address political issues in funding and access to health care services, thus ensuring transparency and accountability by the governments. These criteria also provided advice to the development of the guidelines
Page 16
for assessing the different priority setting frameworks (see Chapter 6) employed by decision-makers (Carter 2002; Sabik and Lie 2008).
Table 1.1: Examples of criteria for priority setting from selected countries Countries Oregon Netherlands (filter mode) Sweden Denmark New Zealand (Maori) Uganda
Canada West Australia Torres Strait, Queensland
Criteria Cost-effectiveness; then equity Service necessity; effectiveness; efficiency; and personal responsibility Human dignity; equal rights; need; solidarity; and cost-efficiency Equality, solidarity, security, and autonomy Efficiency; effectiveness; equity; acceptability; and Maori health principle Efficiency/ accountability; effectiveness/ quality care; equity/fairness/justice/ethics; and national solidarity/cultural practices Maximisation; egalitarianism; claims9 such as severity, human rights, etc Need; socioeconomic disadvantage; access Public outrage; burden of disease; cost; preventability; national priorities
Source (Oberlander and Marmor 2001) (Government Committee on Choices in Health Care 1992) (Health Care and Medical Priorities Commission 1993) (Danish Council of Ethics 1997) (Ashton, Cumming et al. 2000; Tito and Kavanagh 2004) (Kapiriri and Norheim 2004)
(Shiell and Mooney 2002) (Mitton and Prout 2004) (McDermott and Mills 2004)
Lessons from these countries suggest that there was a challenge on how to translate the criteria that had been developed by each country into priorities and how to use them to make trade-offs (Shiell and Mooney 2002). The set of criteria ought to be culturally specific or system specific. A number of lessons from the international attempts at listing priority setting criteria have been identified by some authors (Ham and Coulter 2000; Shiell and Mooney 2002):
Priority setting involves different decision-making levels in the health system and ought to recognise the fact that different decision-making levels may require different sets of criteria.
Priority setting involves a combination of evidence (effectiveness, efficiency, etc) and value and so involves a combination of technique and value judgements.
9
‘Claim’ refers to the duty owed to some members of society who are sick. According to Mooney, this claim is part of what he calls ‘communitarian claims’ derived from the community input in the specification of the legitimacy and strength of the various claims that could be made upon collective health care resources Mooney, G. (1998). "Communitarian claims as an ethical basis for allocating health care resources." Social Science & Medicine 47(9): 1171-1180..
Page 17
Whose values should count when priorities are being set – the community’s, decision-makers’ or the users’ values?
De-listing is ineffective and inconsistent with a solidaristic system.
The process of decision-making ought to be fair and just in order to legitimise the decision made.
An assessment of explicit priority setting attempted in several countries such as the United Kingdom, New Zealand and Sweden yielded mixed results (Sabik and Lie 2008). Sabik and Lie used the following three criteria to assess explicit priority setting in different countries: public input and discussion; appropriate principles; and effect on policy and practice. They found that it is important to base priority setting on principles that are of importance to the society affected by the exercise. Such principles may include equality, equity, human dignity, human rights and autonomy. These principles and procedures have helped to ensure that prioritisation is consistent with society's values and goals for the health care system.
To ensure that the appropriate principles were implemented in priority setting, some countries have established priority setting commissions, especially expert-based ones, which are charged with the responsibility of establishing shared values on which priority setting could be based. Sabik and Lie (2008) found that countries that had established public bodies with some decision-making power on what is actually implemented in the health care system to guide priority setting had been relatively successful. Appropriate public involvement ensures transparency in providing reasons for decisions.
Despite the growing level of interest in priority setting world-wide, there is very little published information on what criteria are used by decision-makers to set priorities at the intervention level in Indigenous health. Readily available data included a body of work involving the analysis of funding and funding models in Indigenous health such as
Page 18
Casemix10 (Owens 1995; Fisher, Murray et al. 1998), expenditure and related issues. This work was concentrated in particular, but not exclusively, on New South Wales, Queensland and the NT (Tsey and Scrimgeour 1996; Warchivker 1996; Deeble, Mathers et al. 1998; McDermott 1998; Mooney, Wiseman et al. 1998; AIHW 2001; AIHW 2005). The problems identified included the lack of transparency and accountability in the way the States and Territories fund Indigenous health services. This lack of transparency can be addressed using explicit priority setting frameworks. It was therefore important to investigate priority setting processes in Indigenous health, especially the performance of the process and what criteria are used in setting priorities. This would enhance transparency in the way the Commonwealth and States/Territories set priorities in Indigenous health and avoid cost-shifting. This would also inform the debate on the appropriateness of the criteria used to develop guidelines for decision-making in Indigenous health.
The National Aboriginal and Torres Strait Islander Health Council has identified some principles to guide priority setting in Indigenous health (NATSIHC 2003). These include cultural respect or cultural security, a holistic approach, health sector responsibility, community control of primary health care services, government, non-government and private organisations working together, localised decision-making, promoting good health, building capacities in the Indigenous community controlled sector, and accountability. On close examination, these principles are useful as general guidelines for Federal, State and Territory hospitals. To use these principles for making trade-offs between what services to fund/provide given other competing demands for the limited resources requires specific guidelines for the implementation of the above principles. It is therefore important to empirically elicit such criteria (Section 5.4) from the decision-makers in Indigenous health. Such criteria would highlight the importance that decision-makers attach to role of evidence, such as economic evaluation.
10
Casemix is a funding mechanism used by government to reimburse hospitals for the costs of treating patients based on the mix of cases that come through the hospital. These cases are based on groups of related diagnoses called diagnosis-related groups (DRGs).
Page 19
1.5.4 Strengthening the Evidence Base: Economic Evaluation The number of economic evaluation studies has been increasing world-wide since the 1970s (Gerard and Mooney 1993). In Australia, there have been an increasing number of economic evaluations studies, with several economic appraisal results being adopted for policy purposes. For example, the Australian government uses economic evaluation to make choices on what pharmaceutical products or medical devices/procedures it lists for government subsidies. This evidence has also been used in other areas such as: breast and cervical cancer screening; organ transplantation; coordinated care trials; the Cancer Council National Strategy; and the National Tobacco Control strategy. However, no such studies have been done specifically in Indigenous health and hence the uptake of economic evaluation evidence is not known.
However, the uptake of economic evaluation results has not been commensurate to the size of the health care sector in Australia. For example, between 1978 and 1993 there were only 33 economic evaluation studies funded and published in Australia, with most of these studies being on drug therapy (Salkeld, Davey et al. 1995), as this is the requirement for the listing of the pharmaceutical products in the PBS national scheme for funding. This represented a very modest role played by economic evaluation, given the size of the Australian health care sector. The modest role could be explained by the fact that Australian health policy emphasis during that period was on the re-structure of the financial organisation of the health system, at the expense of economic evaluation (Richardson 1993).
Barriers to using economic evaluation evidence in priority setting may include: lack of economic evidence, and lack of knowledge about economic evidence; distrust of economic evidence in terms of reliability of results and potential for outside influences; and belief that outside and national results are not relevant or transferable to local situations (Ross 1995; McKenzie, Warrior et al. 1998; van Gool, Gallego et al. 2006). Possible solutions include general awareness and education about economic evaluation and more training of health economists. An exploratory study to gauge the perceptions of the decision-makers on several issues, related to the implementation of
Page 20
an explicit priority setting approach and economic evaluation evidence in Indigenous health, was undertaken and the results are reported in this chapter.
1.6 PBMA as Priority Setting Tool in Indigenous Health The PBMA framework is a combination of two techniques: program budgeting; and marginal analysis (Mooney, Gerard et al. 1992; Peacock and Edwards 1997). Program budgeting involves establishing an overview of the pattern and distribution of expenditure, by describing the current pattern of spending or resource use by a particular population or specific disease. Marginal analysis involves examining potential changes to the current mix of services, referred to here as ‘options for change’, using conventional economic evaluation as a means to improve the efficiency of programs. It makes use of the marginal analysis panel, whose responsibility is to establish the criteria for selecting options for change, and making recommendations to the organisation on the areas of expansion and/or contraction. PBMA adheres to key economic principles to address efficiency. It also has the potential to uphold equity principles and process/participatory issues associated with priority setting (see Chapter 6 for further details). Thus, PBMA embodies both technical methods, that use economic evaluation techniques, and the due process methods, which takes into account participatory and social justice principles in general, making it an ideal framework for priority setting in Indigenous health.
However, applying PBMA in Indigenous health may raise some issues, given that it is a Western concept. These issues may include:
What ethical criteria and their associated weights ought to guide the priority setting exercise;
Poor quality of data and lack of evidence from interventions in Indigenous health;
The level of resources required to conduct PBMA in Indigenous health, given manpower problems in Indigenous organisations and their ‘overburdened’ management with contractual arrangements with the funders (Dwyer, Shannon et al. 2007; Dwyer, O'Donnell et al. 2009);
Page 21
Determining the membership of the panel to make choices;
The need for PBMA to be integrated into the management culture and systems, by capturing managerial needs and multiple objectives;
The challenges of disinvesting in certain programs given the competing health needs of Indigenous Australians; and
The need for processes that ensure full participation of the members.
(Donaldson 1995; Mitton and Donaldson 2003; Mitton and Donaldson 2004; Peacock, Richardson et al. 2007).
The extent of PBMA suitability for use in Indigenous health will be qualitatively appraised in Chapter 6. It will then be contextualised in Indigenous health, as a key contribution of this thesis, in Section C. This will involve undertaking case studies under the auspices of the Assessing Cost Effectiveness (ACE) Prevention project. The ACE Prevention project11, a National Health and Medical Research Council (NHMRC) health services research grant, was a collaborative research program between the University of Queensland and the University of Melbourne. It aimed to provide a recommended menu of interventions to allow policy makers to choose an optimal mix of interventions in non-communicable diseases within their budget constraints.
Under the auspices of the ACE Prevention project, two studies will be undertaken using the ACE Prevention resources to aid the contextualisation of the explicit priority setting framework in Indigenous health. The first project involves the costing of alcohol dry areas intervention for the purposes of economic evaluation. The second involves the development of the concept of benefit instrument, also for the purposes of economic evaluation, to suggest ways of improving the application of PBMA in Indigenous health. With the help of the Indigenous Steering Committee (ISC) for the ACE Prevention project, a workshop approach will be used to explore the concept of benefit from health care interventions (Chapter 8). The costing and the concept of
11
Most of the description here is from the ACE Prevention project protocol and NHMRC research grant application.
Page 22
benefit results will be used to develop guidelines for designing and implementing PBMA in Indigenous health.
1.7 Thesis Scope and Organisation The focus of this thesis will be to develop guidelines for priority setting, and to investigate the way in which PBMA can be contextualised in Indigenous health. This thesis is composed of nine chapters and these are loosely grouped into three sections (Table 1.1). Section A discusses the research question. Section B covers the development of the priority setting guidelines in Indigenous health. Section C covers the contextualisation of the PBMA framework in Indigenous health.
Table 1.2: Summary of the PhD Chapters Chapter number Chapter title Section A: Background to the research question Chapter 1 Statement on the research question Chapter 2 Indigenous health context Section B: Guidelines for priority setting Chapter 3 Normative economics for priority setting Chapter 4 Non-economic foundations for priority setting Chapter 5 Perceptions of decision-makers Chapter 6 Guidelines for priority setting Section C: Contextualising PBMA in Indigenous Health Chapter 7 PBMA appraisal Chapter 8 Expanding the concept of benefit Chapter 9 PBMA guidelines in Indigenous health
1.7.1 Section A Section A is made up of Chapters 1 and 2. Chapter 1 provides a brief overview of the research question, its aims and objectives, justification and theoretical rationale. In this chapter, a brief literature review of priority setting is undertaken to highlight the key issues in priority setting and how an explicit approach can be implemented in Indigenous health. This is then followed by a brief introduction to PBMA as a potential approach to priority setting that can be contextualised in Indigenous health and the scope of this thesis.
Page 23
Chapter 2 highlights the inequities and injustices that characterise Indigenous health compared to other Australians as a case for improved priority setting process. Issues such as employment and education indicators, although important in improving Indigenous health, are briefly covered to highlight the need for improved priority setting. This is followed by the Indigenous health policy review, which focuses on the historical changes that have occurred since 1995 when the responsibility of Indigenous health was transferred to the Commonwealth Department of Health. A review of the funding, organisation and delivery of health services to Indigenous Australians to date is then undertaken. These reviews will demonstrate the government’s role in Indigenous health since 1967, and why it is necessary to improve government priority setting.
1.7.2 Section B Section B covers Chapters 3-6. In Chapter 3, the discussion covers key principles in economics that can potentially guide priority setting. Issues associated with the measurement of these principles are discussed. The discussion centres on welfare economics because it is the normative branch of microeconomics that uses the key economic principles in determining efficiency principle in health and the associated distributions of health associated. It also provides guidance on how welfare (health) improvement, however measured, ought to be estimated. Other economic subdisciplines, which have the potential to contribute to the understanding of priority setting, might not have been explored in detail. These sub-disciplines may include institutional economics, public choice; decision theory and/or political economics. Under welfare economics, normative economic foundations for priority setting are covered. This is followed by the applied welfare economics as a means to highlight potential issues that may be encountered in translating the identified principles into practice, and to address them. Implications for priority setting are then discussed, and the summary of the key economic issues for inclusion in the checklist or guidelines for priority setting are then presented.
Chapter 4 involves the discussion on non-economic foundations for priority setting. The focus will be on the role of ethics and social justice in priority setting; in particular,
Page 24
what social justice principles can potentially improve the performance of economic frameworks. This is in recognition that social justice principles have a role in guiding priority setting, either as key objectives or as criteria for making choices. While priority setting is a multi-disciplinary subject, coverage of the non-economic disciplines largely reflects the extent of their potential contribution to priority setting.
In Chapter 5, an empirical study, the survey of key decision-makers in Victoria who are involved in Indigenous-specific health care services, is undertaken to highlight the user considerations in priority setting. This study particularly seeks to establish the basis on which decisions are made in Indigenous health, the uptake of economic evaluation evidence, and what social justice principles are used for routine decision-making. Views from the decision-makers on how the priority setting process currently performs are elicited and potential means of improving it are identified.
Chapter 5 focuses on the decision-makers at the ‘macro’ and ‘meso’ decision-making levels who are involved in Indigenous-specific health care services. These levels of decision-making, involve State/Regional government levels and Aboriginal community controlled health services. Decision-makers at micro decision-making level, such as general practitioners and Indigenous health workers, although important in influencing the behaviour individuals, are not covered. The focus on the Indigenous-specific health care services means that other mainstream services accessed by Indigenous Australians are excluded, although some of the decision-makers in these services may be involved in Indigenous-specific priority setting. Additionally, the Commonwealth Office of Aboriginal and Torres Strait Islander Health (OATSIH) were not interviewed due to logistical reasons, although OATSIH funds the majority of ACCHS programs.
In Chapter 6, key issues that arose from Chapters 3-5 are summarised. These issues are synthesised and used to develop guidelines for priority setting. In summary, the guidelines for explicit priority setting in Indigenous health form the key elements of an ‘ideal’ priority setting framework in Indigenous health. These guidelines are then used to qualitatively assess existing priority setting models. The chosen model, PBMA, is then contextualised in the next section.
Page 25
1.7.3 Section C Section C contextualises PBMA in Indigenous health and covers Chapters 7-9. In Chapter 7, the potential performance of PBMA is assessed using three empirical studies: the survey of key decision-makers in Victoria; the expenditure study; and the costing of alcohol dry areas. The survey of decision-makers elicits views on PBMA potential involved eliciting preferences from the key decision-makers in Victoria who are involved in Indigenous-specific health care services. This implies that views of other mainstream services accessed by Indigenous Australians are excluded. Additionally, the OATSIH are not interviewed due to logistical reasons, although the OATSIH buy-in for PBMA as the funder of nearly 80% of ACCHS programs would be critical for the success of the approach.
The expenditure study surveys the services utilised by Indigenous Australians in regional Victoria over a specific time frame. The purpose is twofold: to demonstrate how PBMA program budgeting component can be undertaken; and to highlight access to services as will be shown by the utilisation data from this study. The focus is on the publicly funded services, excluding privately provided services and those from nongovernment organisations, in regional Victoria. Extrapolating these results for the rest of the Australian Indigenous population needs to be done with caution.
The costing of the dry areas intervention demonstrates key issues for consideration in PBMA. Expert advice is the main source of costing data, since data availability and quality is an issue for dry areas intervention.
Chapter 8 expands the concept of benefit for economic evaluation purposes to include an Indigenous health construct from the decision-makers’ perspective. This will involve including dimensions of benefit not normally captured in priority setting such as empowerment of individuals and their communities, equity and cultural security in the measure of benefit from health care interventions. Expanding the concept of benefit takes the intervention perspective and not the health system perspective. Certain attributes of the dimensions of benefit not related to interventions per se are
Page 26
therefore excluded. For example, cultural security is restricted to those attributes related to interventions only. Attributes related to the service provider such as the training of workers in the hospital may be excluded. Furthermore, the perspectives of the workshop participants when expanding upon the concept of benefit are context specific.
Chapter 9 is the last chapter. Results from the empirical studies in Chapters 7 and 8 are brought together to develop a checklist for the successful implementation of PBMA in Indigenous health. It is anticipated that the results from this thesis may provide the much needed mechanisms for priority setting in Indigenous health. The methods are potentially useful for assessing the concept of benefit, surveying decision-makers for priority setting purposes, and the costing of services. These provide the platform for the extension of these results and also serve to generate research and debate in priority setting in Indigenous health. Adoption of these methods ought to be handled with caution.
Page 27
2. Chapter 2: Indigenous Health Context “Never before, had I seen *Indigenous+ men living under extraordinary terrorism, many of them fine athletic fellows who, could in case of a row, have settled with their terrorisers in a very summary fashion. But many of them had long been treated as the dogs are treated and were scared into a belief that their employers wielded the power of life and death” – Archibald Meston (18511924).
2.1 Introduction In Chapter 1, an overview of the research question was presented. It was noted that one of the tasks of this thesis is to investigate how an explicit priority setting approach could be undertaken in Indigenous health care. It was explained that economic theory12 offers techniques that can ensure explicit priority setting, and social justice offers ethical principles which ought to guide priority setting. An economic approach to explicit priority setting, which has the potential to uphold social justice principles, is PBMA. However, it was noted that because PBMA is a western concept, it may need to be contextualised in Indigenous health. This is the task which will be investigated in Section C of this thesis, in recognition of some unique issues that characterise the Indigenous health sector.
In this chapter, an overview of the state of Indigenous health is presented. It starts with a brief discussion of social determinants of health, access and utilisation of health care services in order to a scene for understanding the socioeconomic and health disparities between Indigenous and non-Indigenous Australians. A brief history of Indigenous policy since 1995 when the Commonwealth assumed responsibility for the health of Indigenous Australians is presented to demonstrate the progress of the
12
Under economic theory, such frameworks would recognise two key principles in economics: opportunity cost; and marginal analysis.
Page 28
reforms that have occurred. This is followed by an overview of the funding and framework agreements for the delivery of services to Indigenous Australians by all levels of the Australian governments. Issues for consideration in priority setting are then discussed to demonstrate the importance of improving the priority setting processes.
2.2 Background Data used in this section is drawn primarily from the latest information published by the Australian Bureau Statistics (ABS) and the Australian Institute of Health and Welfare (AIHW), such as: The ABS 2006 Census of Population and Housing Survey (ABS 2008); the 2004-05 National Aboriginal and Torres Strait Islander Health Survey (NATSIHS) (ABS 2006); the 2002 National Aboriginal and Torres Strait Islander Social Survey (NATSISS) (ABS 2009); and AIHW Health and Welfare Expenditures Series (AIHW 2001; AIHW 2005; AIHW 2008). The ABS and AIHW also used national administrative data collections such as the National Hospital Morbidity Data Collection, the National Mortality Data Collection, and the National Perinatal Data Collection (ABS and AIHW 2008).
These data and their quality in both the Census and administrative centres are poor, partly due to incomplete identification of Indigenous people in the Census and administrative data sets, and partly due to difficulties in surveying the Indigenous population, which is widely scattered in remote areas (ABS 2007; ABS and AIHW 2008). For example, given the trend of increases in Indigenous population numbers over the past three censuses’ data, the actual Indigenous population is likely to be much higher. The Australian Indigenous population has experienced an increase of about 16% (compared to a 6% increase in the total Australian population) from 1996 Census data to the 2001 and about 13% from 2001 to 2006 Census data (ABS 2007; ABS and AIHW 2008). These data issues potentially lead to net undercounting. For example, in 2006, the Indigenous preliminary adjustment for net undercount was estimated to be 11.5%, and this was not uniform across all States and Territories (ABS 2007).
Page 29
Lack of good quality data may impede planning activities of the government and its agencies. In particular, estimation of epidemiological and other forms of data such as economic data is most likely to be problematic. This would in turn have an adverse impact on basic costs and effectiveness data attributable to the burden of disease and related health interventions, leading to poor planning.
Several initiatives to improve data quality in Indigenous health have occurred over the last decade and have improved the quality of the data and its availability in both census and administrative datasets. These include: a co-ordinated approach to information on Indigenous and Torres Strait Islander Australians in the health, community services and housing sectors; and regular reporting on the health and welfare of Indigenous and Torres Strait Islander Australians. Also, the National Aboriginal and Torres Strait Islander Health Data Principles (2006) provides guiding principles to the realisation of a culturally respectful foundation for the collection, storage and use of health and health-related information on Indigenous Australians (NAGATSIHID 2006). These initiatives have been incorporated in a number of national agreements and strategic plans that have been developed in the health, community services and housing sectors. Despite these achievements, there are still varying levels of under-identification within datasets between different datasets and between regions, which may affect strategies to address social determinants of Indigenous health.
2.2.1 Social Determinants of Health The link between health outcomes and socioeconomic disadvantage and a range of other factors will be referred to as ‘social determinants of health’ in this thesis. The relationship between social determinants, disease and health is very complex (Figure 2.1). Mathews explains that knowledge and resources, education and employment (socioeconomic factors), and individual empowerment behavioural choices and lifestyles, are important factors in determining good health (Mathews 1997). However, he says little about the intervening role of priority setting, expenditure and access to health services, in addressing Indigenous Australians involvements in risky health
Page 30
behaviour. Two theories explain why Indigenous Australians engage in health risky behaviour: the stress associated with being poor and low SES. Figure 2.1: Social determinants of good health
Source: (Mathews 1997).
There is strong evidence in the literature which links low SES with poor health and increased exposure to health-risk factors (Turrell and Mathers 2000; Blakely, Hales et al. 2004; Carson, Dunbar et al. 2007). For example, Indigenous Australians who had completed Year 12 were less likely to engage in risky lifestyle activities, such as smoking, alcohol abuse, physical inactivity and/or not eating fruits and vegetables, than those who had low levels of educational attainment in 2004-05 (ABS 2006).
In 2008, the proportion of Indigenous Australians who reported high and very high levels of psychological stress was about 31% (ABS 2009). The stress associated with being poor and people’s consequent engagement in risky behaviour is demonstrated as a theory in the literature (Evans, Barer et al. 1994; Lynch and Kaplan 2000). In general, there is a gradient of health from poorest to wealthiest which indicates that the lower the material standard of living, as measured by indicators such as income,
Page 31
the worse the level of health, whatever measure of health, mortality, morbidity or quality of life is used (ABS and AIHW 2008). This gradient has been observed for most of the major causes of death, so improving certain aspects of the SES, such as income and education, is likely to improve health status, ceteris paribus (Evans, Barer et al. 1994; Lynch and Kaplan 2000).
Improving SES conditions can be done by improving people’s access to high quality health care services, since access can be affected by income, through the inability to travel long distances in search of quality health care, and the lack of the ability to purchase it. Also, the link between income and health is well documented in literature (Berkman and Kawachi 2000; Turrell, Stanley et al. 2006). This evidence suggests that relative income/deprivation cause health inequalities in developed countries such as Australia. Wildman and Jones argue that relative deprivation affects people’s preferences, and hence their utility functions, through lowering the opportunity cost of unhealthy behaviour causing deterioration in health status (Wildman and Jones 2001). They further argue that policy approaches in the relative income hypothesis should therefore target income inequality through government redistributive principles such as progressive taxation and/or progressive social insurance. Such policies would improve people’s access to resources to address their health needs. In Indigenous health, interventions policies should target poverty-related issues, the living environment, education programs and quality health care services, as a means to reduce risky behaviour.
The link between the stress of being poor and poor health outcomes could also be traced to dispossession and alienation (Figure 2.2) caused by colonisation (Bartlett and Legge 1994; Duckett 2004). Colonisation led to spiritual harm to Indigenous Australians due to dispossession and alienation from their land. It also caused a loss of the traditional hunter-gatherer lifestyle which in turn led to Indigenous Australians being exposed to the white Australian social lifestyle (Duckett 2004). This risky lifestyle led to alcoholism, substance abuse, violence and other adverse health effects among Indigenous Australians.
Page 32
Figure 2.2: Historical impacts of colonisation upon Indigenous health
Source: (Duckett 2004)
Thus, settler colonialism, together with the current socio-economic conditions, have exposed Indigenous Australians to significant lifestyle risk factors, such as traditional behavioural health risk factors and ‘other social determinants’ of Indigenous health. The key behavioural risk factors that contribute to the greater burden of ill health experienced by Indigenous Australians may include smoking, alcohol use, illicit drug use, lack of nutrition, and physical inactivity (ABS and AIHW 2008). These factors impact on the onset, maintenance and prognosis of several non-communicable diseases. Addressing these factors and increased utilisation of health care services would most probably lead to large health gains for Indigenous populations, leading to the reduction of health inequalities between Indigenous and other Australians.
There have been some significant inroads in a number of specific areas in Indigenous health. For example, there have been some improvements in the SES of Indigenous Australians in the years 2001 to 2006 (ABS and AIHW 2008). Year 12 completion rates and unemployment rates improved, and rates of home ownership increased in that
Page 33
period. Furthermore, a decline in the number of Aboriginal infants and children admitted with infections has declined over the years (McNeilly, Cicchini et al. 1983; Gracey, Mark et al. 1992). Despite those registered improvements, the socioeconomic outcomes for Indigenous Australians remain below those for non-Indigenous Australians, limiting Indigenous access and utilisation of health care services and putting them at greater risk of ill-health.
2.2.2 Utilisation and Access to Health Care Services The utilisation rates of health care services by Indigenous Australians given their poor health indicators are well below that of other Australians. For example, in the 2004-05 NATSIHS, about 21% of Indigenous Australians who needed to see a dentist in the previous 12 months had not done so. About 15% had not gone to see a doctor when they needed to. Availability of services, cost of the services, distance and transport, language and cultural barriers were cited as the key reasons for remote Indigenous Australians not seeking medical help (ABS 2006).
Using expenditure on services for Indigenous Australians as a proxy for access, national studies have shown that health care spending on Indigenous Australians has not matched their greater need. For example, expenditure per capita on Indigenous Australians was 17% higher than non-Indigenous Australians in 2004-05. Specific areas of low utilisation of health care services compared to non-Indigenous Australians include the private health sector, about 23% of non-Indigenous expenditure, and primary health care such as the MBS and the PBS. High usage of publically funded services was instead on hospital and public health services (AIHW 2005). Specifically, secondary/tertiary care service use was about 38% higher than for non-Indigenous people’s use, that is, public hospital use was nearly twice that of non-Indigenous people’s use, mainly for ambulatory-care sensitive services.
One of the key issues that affect Indigenous Australians’ use of health care services is access to services due to the long distances they have to travel, especially in the rural and remote Australia. Most of the services are located in places that are hard for Indigenous and Torres Strait Islander Australians to reach. Transport becomes an issue.
Page 34
For example, results of the 2002 NATSISS (NATSIHC 2003) showed that approximately 60% of Indigenous people aged 18 years and over had access to and were able to drive a motor vehicle, compared to 85% of non-Indigenous people. This proportion was lower in remote areas, where only 48% of Indigenous people had access to a vehicle. Furthermore, 78% of discrete Indigenous communities were located more than 50km from the nearest hospital in 2001, and 50% were located more than 25km from the nearest community health centre (ABS and AIHW 2008).
Travelling such long distance may mean staying away from their homelands and that requires money for upkeep. Given that most Indigenous Australians in remote areas depend on government welfare payments, money is an issue. In some remote areas, the provision of private and community based health care may include up-front fees for services and medications. For those Indigenous Australians who may have Medicare cards, when they get seen by GPs, they are unlikely to be referred to a specialist, because poor linkages exist between specialist sectors and other health services/programs in Indigenous health. First, the number of health care personnel that Indigenous people had access to in 2002 was 281 medical practitioners per 100,000 population in remote and very remote areas, compared with 312 per 100,000 in major cities (ABS and AIHW 2008). The most significant shortage of health professionals in remote and very remote areas was of specialist clinicians where there were only 29 per 100,000 compared to 114 per 100,000 in major cities. Secondly, while Indigenous Australians were more likely to be hospitalised than other Australians, they were less likely to undergo a procedure once admitted to hospital. It is not clear why Indigenous patients are less likely to undergo a procedure, but some possible factors include lack of cultural appropriateness of the health services, lack of community control of the service, the gender of health service staff, and the degree of proficiency in spoken and written English are some of the issues that affect Indigenous Australians access to mainstream health care services (Ivers, Palmer et al. 1997). For example, the 2008 NATSISS revealed that about 11.5% of the Indigenous Australians did not speak English as a main language at home (ABS 2009). This would then require the presence of translators or the presence of Indigenous workers in the service to make it welcoming and friendly.
Page 35
The workforce issues in Indigenous health include the availability of Indigenous and Torres Strait Islander staff to enable Indigenous Australians to effectively access services (McDonald 2004). Indigenous participation in the health care service delivery of services is generally very low. For example, Indigenous representation in healthrelated occupations was about 1% of the total workforce. In the 2001 Census, Indigenous Australians aged 15 years and over were less likely (1.5%) than other Australians (3.1%) to be employed in health-related occupations. Indigenous Australians were more likely to report being employed in selected welfare and community service-related occupations than health-related occupations. About 2.7% of those employed in community and welfare service-related occupations were Indigenous compared with 1.6% of non-Indigenous adults. Overall in 2003, 167 Indigenous students (aged over 20 years) completed health-related undergraduate courses and 105 completed welfare-related undergraduate courses, representing 1.0% and 1.5% respectively of all students completing undergraduate courses in these fields – well below the 2.4% Indigenous representation in the Australian population.
The strengthening of comprehensive and community controlled primary health care services has been seen as one of the many components of the foundation on which Indigenous Australians’ health outcomes can be improved. According to the Commonwealth Department of Health and Family Services, the strategic approach to achieving good quality primary health care should involve a delivery mechanism that provides comprehensive and population-focused primary health care services (Deeble 1998). Such a strategy should involve partnership arrangements between governments and Indigenous communities with representative bodies including NACCHO and its affiliates. Thus, such a strategy should integrate primary clinical care, population health and health promotion activities, and ACCHS model currently operating throughout. However, ACCHS are faced with several challenges, which include funding shortfalls, workforce planning issues, priority setting and management issues. The following is a summary of the barriers that Indigenous Australians face when accessing mainstream services (ABS and AIHW 2008):
Long distances to health services;
Page 36
Financial barriers and low Medicare enrolment rates;
Poor linkages between specialist sectors and other health services/programs;
Lack of culturally appropriate health care services;
Inappropriate general (mainstream) health promotion approaches; and
Poor performance of the health system in meeting the needs of those with complex and multiple conditions.
2.2.3 Indigenous Demographics As of June 2006, the Australian Indigenous population was estimated to be 517,200 – about 2.5% of the Australian total population. Table 2.1 shows that as of 30 June 2006, New South Wales (NSW) had the largest estimated resident Indigenous population in Australia (148,200 or 29% of the total Indigenous population) followed by Queensland (146,400 or 28% of the total Indigenous population), whereas the Australian Capital Territory (ACT) had the smallest (0.8%). The Northern Territory (NT) had the largest proportion of its total resident population who were Indigenous (31.6%), and Victoria and ACT had the least proportion of their population being Indigenous.
Table 2.1: Estimated Resident Population (a), Indigenous–2001 and preliminary-2006
Source: (ABS 2008)
Page 37
In June 2006, 32% of Indigenous people lived in cities and 43% lived in regional towns, but the Indigenous people are generally much more widely dispersed than nonIndigenous population. Nearly 25% of them live in either remote/very remote areas, compared with 2% for the rest of the Australian population (ABS 2009). The distance to travel to access quality health care services is likely to be an issue for Indigenous Australians.
2.2.4 Socioeconomic Status The SES of Indigenous Australians is very low compared to non-Indigenous Australians13. Individual person-level measures of socioeconomic position include occupation, income, assets and education. Group or area-level measures include occupation, education and economic structure, housing characteristics and indexes, such as incorporating a number of measures, of poverty or deprivation (Lynch and Kaplan 2000).
The measures of SES in Indigenous communities relative to other Australians indicate significant disparities. These disparities generally increase as remoteness from the urban centres increases, and can be attributed to many factors such as: low income, high unemployment, low levels of education and working in unskilled occupations.
Household income levels for Indigenous Australians are about 40% lower than average mainstream household weekly incomes. They face rates of unemployment that are nearly three times higher than the rest of Australians, with lower home ownership than other Australians (ABS 2009). In terms of levels of education, the 2008 NATSISS indicated that Indigenous Australians older than 15 years were half as likely to have completed Year 12 than their non-Indigenous counterparts (ABS 2009). Like all other Indigenous Australian SES data, these proportions worsened as remoteness increased.
13
The term "socioeconomic status" means the social and economic factors that influence the position individuals and groups hold within the structure of society that may have an influence on their health Lynch, J. and G. Kaplan (2000). Socioeconomic position. Social epidemiology. L. Berkman and I. Kawachi. Oxford, Oxford University Press..
Page 38
Table 2.2: Highest school year attended, Indigenous -15 years and over (a)–2001 and 2008 2001 2008 no
%
No
%
Year 12
41,233
19.5
66,700
20.4
Year 10/11
87,303
41.4
144,500
44.2
Year 9 or below
82,481
39.1
115,900
35.4
Not stated
20,133
na
na
na
(a) Excludes persons still attending secondary school (b) Includes persons who never attended school Source: ABS 2001 Census data and 2008 NATSISS in (ABS and AIHW 2008; ABS 2009)
The Indigenous levels of attainment of school completion by 15 years and over improved marginally over the seven-year period between the 2001 and 2008 censuses (Table 2.2), but it was still below those of non-Indigenous Australians. It can be seen that the proportion of Indigenous people aged 15 years and over who had completed school had increased from 19.5% in 2001 to 20.4% in 2008. Those aged 25-64 years who completed non-school qualifications increased from 32% in 2002 to 40% in 2008.
The labour force participation of Indigenous Australians was very low (62%) in 2008 with the unemployment rate at 10.2% (ABS 2009). For example, the proportion of Indigenous Australians aged 18-24 years who were fully engaged in work and/or study in between 2001 and 2006 has not changed, sitting at the rate of 31% (ABS and AIHW 2008). This inevitably has an impact on home ownership, life styles and living conditions. Furthermore, Indigenous households generally have inadequate water, poor sanitation and sewage disposal, and limited food storage. These living conditions are not conducive to good health and worsen as the locations become more remote.
There have been some improvements in the SES of Indigenous Australians over the years between 2001 and 2006 (ABS and AIHW 2008). For example, Year 12 completion rates increased from 20% in 2001 to 23% in 2006, and unemployment rates fell from 20% in 2001 to 16% in 2006. There have also been some marginal increases in the rates of home ownership from 31% in 2001 to 34% in 2006. Despite those registered improvements, the socioeconomic outcomes for Indigenous Australians continue to
Page 39
decline, and remain below those for non-Indigenous Australians, putting Indigenous Australians at greater risk of ill-health.
2.2.5 Indigenous Health Status The level of ill-health in Indigenous communities is reflected in the data on the health status of Indigenous Australians. The health status of Indigenous and Torres Strait Islander people is much worse than that of the non-Indigenous population and their life expectancy is much lower than that of comparative Indigenous people in the United States, Canada or New Zealand (Kunitz and Brady 1995; NHMRC 2003; Marmot 2005; CSDH 2008; Gracey and King 2009). At the Australian national level, Indigenous Australians had a lower life expectancy than non-Indigenous Australians by about 10.6 years (11.5 years for males and 9.7 years for females in 2005-2007 (ABS and AIHW 2010). Life expectancy at birth for males was about 67.2 years, compared with nonIndigenous males at 78.7 years. For females, life expectancy at birth was estimated to be 72.9 years, compared with 82.6 years for non-Indigenous females.
The measure of Indigenous health status using the self-reported surveys indicates that they are more likely to report poorer health. According to the survey of Indigenous Australians aged 15 years and over reported in the 2008 NATSIHS, about 28% of Indigenous Australians self-assessed their health as being fair or poor (ABS 2009). Results from the 1994 and the 2005 surveys indicated that self-assessed Indigenous health status declined. The proportion of Indigenous Australians who reported their health as fair or poor increased from 18% to 22% (Table 2.3). This was more pronounced among females than males, indicating that Indigenous Australians assess their health to be worsening. Standardised Mortality Rates (SMR)14 for Indigenous Australians were two to four times higher than those of non-Indigenous Australians. Diabetes mellitus was
14
The SMR is the ratio between the observed number of deaths in the Indigenous population and the expected number of deaths that would have occurred if the Indigenous population experienced the same age-specific death rates as the non-Indigenous population.
Page 40
responsible for the highest SMR at about eight to 12 times that of other Australians (ABS and AIHW 2008). These occurred in spite of the falling death rates from all causes combined in every age group in the NT since the late 1960s (Thomas, Condon et al. 2006). Furthermore, available data from Western Australia (WA), South Australia (SA) and NT jurisdictions indicate that there were significant falls in Indigenous infant mortality rates from 1991 to 2005 (ABS and AIHW 2008).
Table 2.3: Self-assessed health status, Indigenous persons aged 15 years and over, by sex – 1994 and 2004-05
Source: ABS 1994 NATSIS, 2004-05 NATSIHS in (ABS and AIHW 2008)
In terms of hospitalisations, the rates for Indigenous Australians were also much higher than for other Australians in 2004-05. For example, care involving dialysis were 14 times the rate for other Australians (ABS and AIHW 2008). Most of the Indigenous hospital discharges or separations were in public hospitals (about 94%) compared with 60% for other Australians. After adjusting for age structures, Indigenous people were more than twice as likely to be hospitalised as other Australians. Once admitted to hospital, they had longer hospital stays than their non-Indigenous counterparts, and were less likely to undergo a procedure (AHMAC 2006; Britt, Harrison et al. 2008). In 2005-06, the rate of hospitalisations due to ambulatory care-sensitive15 conditions for Indigenous Australians was five times the rate for other Australians (ABS and AIHW
15
Ambulatory conditions can be defined as those conditions which are thought to be potentially reduced or eliminated from hospital care by high-quality appropriate primary health care services delivered under ideal conditions Stamp, K. M., S. J. Duckett, et al. (1998). "Hospital use for potentially
Page 41
2008), and about eight times the rate of other Australians for chronic conditions. Potentially preventable acute conditions were recorded at twice the rate of other Australians. Addressing ambulatory care-sensitive issues would require improved access by Indigenous Australians to mainstream services and quality primary health care services from ACCHS. Improvement in the quality of primary health care services can be achieved through the use of explicit priority setting frameworks (Chapter 4).
In summary, despite the marginal gains in the health status of Indigenous Australians, it remains considerably below that of other Australians. For example, Indigenous adults are twice as likely as other Australian adults to report ill-health. Also hospitalisation rates are higher for ambulatory-sensitive conditions, such as diabetes and kidney disease, and the mortality rates of Indigenous people are almost three times the rates for other Australians. Furthermore, Indigenous people live in poor conditions, and in remote or very remote Australia where health care services are not adequate and not accessible. Nearly half the population is aged below 18 years and most of these young people have low levels of education, hence the difficulties Indigenous households face in leading lifestyles that would encourage good health.
2.3 History of Indigenous Policy The purpose of this overview is to highlight the link between historical context and the current policy weaknesses. These weaknesses are demonstrated by the low levels of funding for Indigenous health care services relative to the burden of illness. This overview covers historical periods which have been loosely categorised as:
the pre-1995 era;
the Aboriginal & Torres Strait Islander Commission (ATSIC) era; and
the post-ATSIC era (2005 to the present date).
preventable conditions in Aboriginal and Torres Strait Islander and other Australian populations." Australia New Zealand Journal of Public Health 22(6): 673-678..
Page 42
2.3.1 Pre-1995 The pre-ATSIC era covers the time from 1967, when the 1967 referendum was held to amend the Australian Constitution, when the responsibility for Indigenous health was transferred to the Commonwealth Department of Health, under the direct responsibility of OATSIH, and ends in 1995. The 1967 referendum gave the Australian Federal government a clear mandate to take over Indigenous health. It led to all Indigenous Australians being included in the national census which meant that the data needed for planning purposes in Indigenous health could be generated. The two years following the referendum, saw the establishment of the Commonwealth Office of Aboriginal Affairs (OAA) with Indigenous health identified as a priority area. In 1969, the OAA, the Council of Aboriginal Affairs (CAA) and Indigenous health units in State government were established to administer the Australian government funds (NATSIHC 2003).
By 1971 the first ACCHS called the Redfern Aboriginal Medical Service (AMS), had been initiated on a voluntary basis in Redfern, Sydney (Thomson 1984; Altman and Sanders 1991; Anderson 1997). Three years later the national AMS umbrella organisation, the National Aboriginal and Islander Health Organisation (NAIHO), was formed (ATSIC 1993)16. In 1984, the National Aboriginal Community Controlled Health Organisation (NACCHO) was established; it was, in essence, a replacement of NAIHO which had ceased to exist some five years earlier. ACCHS sector was now becoming a strong provider of health care to its people.
By 1984, responsibility for all Commonwealth Indigenous health programs, including the Department of Health's role in the funding of some AMS, was consolidated within
16
The Program Effectiveness Review noted the debate between NAIHO, which was quoted as arguing for a plan to more than double the number of AMSs in order to replace the States’ grants programs, and the States who continued to defend their role in Indigenous health and their claim to the States’ grants. The report also recommended that, rather than the system used then in which responsibility was shared between the Commonwealth Department of Health and the DAA, Indigenous health funds be transferred to the Department of Health DHA (2009). 2006-09 Department of Health and Ageing Corporate Plan. D. o. H. a. Ageing. Canberra, Commonwealth of Australia..
Page 43
the Department of Aboriginal Affairs (Saggers and Gray 1991)17. In the same year, universal health insurance, Medicare, was introduced, and it implied that Indigenous access to health services would be improved.
Meanwhile, the OAA and the CAA had been replaced with the Department of Aboriginal Affairs (DAA) in 1972, whose responsibility was to advise government on Indigenous policy and administer such policy until 1988 when ATSIC was established. Following the establishment of the DAA, the government established the Aboriginal Development Commission (ADC), a statutory authority to manage developmentoriented Indigenous affairs programs such as loans and grants for housing and business enterprises (McCausland 2005). Since 1972, there have been several advisory bodies, such as the National Aboriginal Consultative Committee and the National Aboriginal Congress, set up by different governments, but these have been criticised for not being able to influence government policy, and for being remote from decisionmaking.
In 1976, The Report on the Delivery of Services by the Department of Aboriginal Affairs was published. It assessed the capability of the DAA to fulfil its responsibilities for Indigenous policy development and administration (Hay 1976). Later, the Commonwealth asked the House of Representatives Standing Committee on Aboriginal Affairs (HRSCAA) to conduct a review of Indigenous health which led to a $50 million five-year Aboriginal Public Health Improvement Program in 1981 to address issues identified in the review. The program, administered by the DAA, focused on unsatisfactory environmental conditions associated with inadequate water, sewerage and power systems (Thomson 1984; Gardiner-Garden 1994).
A meeting of Commonwealth, State and Territory Health and Aboriginal Affairs Ministers in 1987 led to the formation of a Joint Ministerial Forum on Indigenous
17
Since 1973, when the Aboriginal Health Branch was established in the Commonwealth Department of Health, the administrative arrangements at Commonwealth level had become increasingly complex with the Departments of Health and the DAA having overlapping responsibilities for Indigenous health. A number of reports, including the 1980 Program Effectiveness Review , had recommended consolidation within one Commonwealth agency.
Page 44
health and the appointment of a National Aboriginal Health Strategy Working Party (NAHSWP). The NAHSWP was to develop a strategy on Indigenous health that would encompass issues pertaining to funding, Indigenous participation, inter-sectoral coordination and monitoring and meet with the approval of all stakeholders (Thomson 1984; Gardiner-Garden 1994). In 1989, the National Aboriginal Health Strategy (NAHS) (Gardiner-Garden 1994; Anderson 1997) report, was presented to the Joint Ministerial Forum (NAHSWP 1989). To date, the NAHS remains one of the most influential documents in the history of Indigenous health. Several other important strategic documents have since been produced by different Inquiries to complement the NAHS. These include the:
Aboriginal and Torres Strait Islander Health Goals and Targets, which was proposed as a means of evaluating the effectiveness of the NAHS (Anderson 1997).
Royal Commission into Aboriginal Deaths in Custody (RCADC) (ATSIC 1993; Gardiner-Garden 1994).
2.3.2 The ATSIC Era This period covers the history of the Indigenous health policy from 1995, when OATSIHS was established and it assumed the direct responsibility for Indigenous health to 2004, when ATSIC was disbanded. In 1995 in particular, the Department of Human Services and Health (DHSH) and ATSIC instituted a Memorandum of Understanding defining their roles and responsibilities (Thomson 1991). DHSH was became responsible for Indigenous health and ATSIC was left to look after non-health programs such as housing (Richardson 1997).
Following the 1992 National Commitment to Improved Outcomes for Aboriginal and Torres Strait Islander People, the various health ministers agreed to a process for the development of multilateral Framework Agreements with States and Territories (Australian National Audit Office 1998). These were agreements on Indigenous and Torres
Strait
Islander
health
between
the
Australian
governments,
the
States/Territories, the Indigenous community controlled health sector and ATSIC (until 30 June 2004). The Framework Agreements provided for the establishment of
Page 45
consultative national and State/Territory forums to provide policy and planning advice on Indigenous health issues (Commonwealth Department of Health and Family Services 1997). They have been the major mechanism for a collaborative approach to resource allocation, planning and priority setting. However, they have been limited by a lack of a clear mandate to lead planning and readily available data for planning purposes.
In order to improve data for planning purposes, the Aboriginal and Torres Strait Islander Health and Welfare Information Unit (ATSIHWIU), a joint program of the AIHW and ABS, undertook a review to develop a National Plan for Aboriginal and Torres Strait Islander Health Information in 1996 (Commonwealth Department of Health and Family Services 1997). The health ministers agreed to the Aboriginal and Torres Strait Islander Health Information Plan for improving the quality of Indigenous health data (Aboriginal and Torres Strait Islander Health and Welfare Information Unit 1997).
More improvements in data quality and availability took place in 1998, when the AIHW released their report, Expenditures on Health Services for Aboriginal and Torres Strait Islander People, 1995/96 (Commonwealth Department of Health and Family Services 1997). After the release of this report, increased attention was given to Indigenous health and so funding for services for Indigenous health significantly increased in the following year (Deeble, Mathers et al. 1998). Subsequent biennial reports have proven to be very useful for the advocacy and planning purposes in Indigenous health (Barnes 2004). These data contributed significantly to the planning and priority setting process in Indigenous health.
The 2002 Memorandum of Understanding between ATSIC and the DHA set out mutual responsibilities for improving the health and environmental issues in Indigenous health. The following were agreed upon: State/Territory level agreements such as the agreement between the WA Government and ATSIC on the delivery of environmental health services to Indigenous communities in WA; and
Page 46
Agreements between funders in each jurisdiction which outline the principles and arrangements for the impact of specific programs, such as the Memoranda of Understanding for the Primary Health Care Access Program.
In 2003, the Australian and State/Territory governments endorsed and signed the National Strategic Framework for Aboriginal and Torres Strait Islander Health: the Framework for Action by Governments (National Strategic Framework or NSF). The National Strategic Framework is a complementary document that builds on the 1989 NAHS and addresses approaches to primary health care within contemporary policy environments (AIHW 2001; AIHW 2005; AIHW 2008).
In summary, the key issues that emerged during ATSIC era include the improvement in strategic planning for Indigenous health through the establishment of the framework agreements, and the establishment of initiatives for the improvement of the data, both health and expenditure data, availability and quality for planning purposes.
2.3.3 Post-ATSIC Era This period covers the period from 2004 after ATSIC was disbanded to the present day. In early 2004, the Australian government announced changes to the way policies, programs and services were developed and delivered to Indigenous people and communities. ATSIC and its service delivery arm, the Aboriginal and Torres Strait Islander Services (ATSIS), were subsequently abolished. To co-ordinate the approach to Indigenous policy at the national level, an Office of Indigenous Policy Coordination was established within the Department of Immigration, Multicultural and Indigenous Affairs (DIMIA) (NATSIHC 2003). Delivery of all Indigenous-specific programs was transferred to mainstream agencies and a ‘whole-of-government’ approach adopted. According to the Australian government at that time, this new approach was based on a process of negotiating agreements with Indigenous families and communities at the local level in accordance with the concepts of mutual obligation and reciprocity for service delivery.
Page 47
In 2007, the new Labour government, headed by the Hon Kevin Rudd MP, committed to close health gap between Indigenous and non-Indigenous Australians. Specifically, the government committed itself to closing the difference in life expectancy between Indigenous and non-Indigenous Australians within a generation, and reducing the mortality gap of children under five years old within a decade (COAG 2009).
In June 2007, the Parliament tabled the Human Rights and Equal Opportunities Commission’s (HREOC) annual Social Justice Report 2006 (Aboriginal & Torres Strait Islander Social Justice Commissioner 2006). The report identified two major problems with the way in which the government dealt with Indigenous reform: 1. The new whole-of-government approach did not adequately include Indigenous people in the decision-making process; and 2. The government had no frameworks or benchmarks to gauge improved access to services. This thesis contributes to the findings of the Social Justice Report 2006 by investigating the above two issues. The first issue was investigated under the non-economic foundations, in Chapter 4, and the role of the ISC of the ACE Prevention project in the empirical studies, demonstrating meaningful involvement of Indigenous Australians in decision-making. Secondly, the development of the framework for priority setting which embodies economic and social justice principles, investigated through the development of the guidelines for priority setting to health services (Section B), and the contextualisation of PBMA in Indigenous health care (Section C), provide the frameworks to improve access to value-for-money interventions in the Indigenous health care system.
2.4 Indigenous Australian Health Care System The Indigenous health care system involves services that are publicly funded, privately funded and those services funded by the non-government sector. The publicly funded services are provided to Indigenous Australians in two ways: through the universally available health care services, referred to here as ‘mainstream services’; and through
Page 48
other services specifically targeting Indigenous Australians due to their special needs, referred to as ‘Indigenous-specific health care services’.
2.4.1 Funding Indigenous Health Care Services The funding of health care services involves the publicly funded health care services through the Medicare, privately funded services through private health insurance, and non-government organisation funded services. Medicare is financed largely from general taxation revenue, which includes a Medicare levy based on a person’s taxable income. Commonwealth funding for Medicare is mainly provided via the Pharmaceuticals Benefits Scheme (PBS), the Medicare Benefits Scheme (MBS), grants and specific purpose funds to State/Territory governments and other bodies.
The MBS subsidises (or sometimes meets all) the costs of treatment by practitioners such as doctors, participating optometrists or dentists (specified services only). Charges for practitioner services are reimbursed by the Australian government based on the published schedule of fees for different levels of consultation. The PBS subsidises the costs of prescribed medicines, with a safety net providing free medicines for the chronically ill.
The State and Territory governments have the primary responsibility for the delivery and management of free public hospital services funded by the Commonwealth through the Australian Health Care Agreements. State and Territory governments supplement Medicare funding with their own revenues, mainly for the funding of public hospitals. Population health activities are funded through specific purpose grants to State/Territory governments and other bodies, and delivered through both the government and non-government sectors. For example, the funding for the expansion of comprehensive primary health care services in Indigenous and Torres Strait Islander communities is done through the Primary Health Care Access Program (PHCAP).
The funding for Indigenous-specific health care services by the Commonwealth government involves portfolio budget submissions to the Expenditure Review
Page 49
Committee (ERC)18 (Fitzpatrick 2004). When prioritising the budget proposals, the ERC uses information from portfolio budget submissions and briefs prepared by the Department of Finance. A typical portfolio submission has to demonstrate that it is first, a priority, and secondly, that it can achieve positive and tangible outcomes. In this case, evidence such as the size of the burden being proposed to be addressed or evidence of effectiveness of the intervention may be used to demonstrate cost effectiveness of such a policy (NATSIHC 2003). It is important for the portfolio minister to demonstrate that the additional funding provided is likely to lead to greater improvements in relative and absolute outcomes, and that the proposal that is detailed is the responsibility of the Commonwealth government. Fitzgerald explains that the ERC is likely to support policy proposals that are backed by detailed and rigorous analysis, including appropriate quantitative analysis. This increases the chances of the proposal being funded (Fitzpatrick 2004).
In Victoria, for example, Indigenous health funding submissions normally go through two key separate funding categories. The first is through the ‘whole of government’ funding such as housing, correction services, employment and others are considered. This submission goes through the Head of the Ministerial Task Force on Indigenous health. The second submission involves the mainstream pool of programs in which Indigenous health is treated as part of the mainstream populations. When such a submission goes through the mainstream programs, it is treated as part of the Departmental structure. In either case, the funding submission has to be backed by evidence. This evidence may be in the form of epidemiology data, economic evaluation evidence, or political incentives.
2.4.2 Framework Agreements Indigenous Australians are expected to utilise the universally available mainstream health care services; however, due to access issues, they often are not able to. Indigenous-specific health care services are meant to target Indigenous Australians as
18
The ERC is a sub-committee of the Cabinet comprising five to six Ministers including the Prime Minister, the Treasurer and the Minister for Finance and Administration.
Page 50
one way to address these access issues. These services are delivered through two avenues: through the mainstream providers who are specifically targeting Indigenous Australians; and through the Indigenous community controlled health sector. The delivery of the Indigenous-specific health care services is underpinned by the Framework Agreements (NATSIHC 2003; NATSIHC 2003).
These agreements are a part of the NSF for addressing Indigenous health issues. The framework agreements form the primary vehicle for ensuring collaboration in resource allocation, joint planning and priority setting for service delivery between key stakeholders in Indigenous health within each State and Territory. The NSF is based on the principle that both Commonwealth and State levels of governments are “jointly responsible for responding to the needs of all Australians (including) Indigenous and Torres Strait Islander Australians” (NATSIHC 2003; NATSIHC 2003).
The NSF complements the 1989 NAHS by addressing contemporary issues and approaches to addressing Indigenous health within the current policy context. It aims to guide Australian governments at all levels through the co-ordinated, collaborative and multi-sectoral approach to Indigenous health and supported by the Aboriginal community Controlled Health Services (ACCHS). The NSF uses two approaches to address Indigenous health: a macro approach applied through the Indigenous Health Framework Agreements; and a local/regional approach implemented through the development and provision of good quality primary health care services to Indigenous and Torres Strait Islander people.
The policy and planning process in Indigenous health is complex. It is aided by three core processes: Partnership forums and regional/local planning; the Australian Health Ministers’ Advisory Council (AHMAC); and the existing reporting mechanisms (NATSIHC 2003). The health partnerships are established in each State/Territory 19 to oversee the collaborative work under the Framework Agreements. The forum is made up of all the partners in Indigenous health service delivery in the State/Territory, and they are
19
This includes all States (NSW, Queensland, SA, Tasmania, Victoria and WA) and all Territories (ACT and NT).
Page 51
responsible for identifying need and capacity for service expansion and advice to government on the impact of programs on Indigenous health. The regional and local area planning processes involve the notions of need and access issues in regions of individual States/Territories, and individual communities respectively. This process enhances Indigenous community control and participation in health service activities.
AHMAC comprises the Chief Executive Officers (CEOs) of all Australian State and Territory health portfolios. AHMAC is responsible for advising its member Ministers on strategic issues relating to the co-ordination of health services across the nation and operates as a national forum for planning, information sharing and innovation. AHMAC receives advice from NATSIHC relating to the Framework Agreements. NATSIHC primarily provides advice to the Commonwealth Minister for Health and Ageing on matters relating to Indigenous health and on the implementation of the Framework Agreements. Its membership includes representatives of the four Framework Agreement partners, the Australian Indigenous Doctors Association and the Congress of Aboriginal and Torres Strait Islander Nurses, the Chairperson of the NHMRC in an ex-officio capacity and ministerial appointees with expertise in Indigenous and Torres Strait Islander health (NATSIHC 2003).
The existing reporting mechanisms include National Performance Indicators for Indigenous & Torres Strait Islander Health20; and Service Agreements developed between ACCHS and OATSIH. The Service Activity Report is the reporting mechanism for the Service Agreements and it is a joint data collection project between NACCHO and the OATSIH. The data includes service level data on health care and health-related activities funded by the Commonwealth in ACCHS. These reports are complemented by the Reports against the Framework Agreements, which details the progress in meeting the objectives of the Framework Agreements in each State and Territory.
20
The National Performance Indicators for Indigenous and Torres Strait Islander Health were developed through a collaboration of the Indigenous Health Units in each State and Territory government and the Commonwealth government on behalf of AHMAC. The Performance Indicators were compiled with advice from technical experts, NHMRC, NACCHS, the Indigenous community controlled health sector and public health agencies, ATSIC, the Ministerial Council on Aboriginal and Torres Strait Islander Affairs, the ABS and the AIHW.
Page 52
The delivery of Indigenous-specific health care services to Indigenous Australians is through a partnership of NACCHO and its affiliated organisations, ACCHS, State/Territory governments, and the Australian government. NACCHO is the national peak Indigenous health body representing ACCHS throughout Australia. ACCHS are sometimes referred to as an AMS or simply as a ‘health service’. An AMS is “a primary health care service initiated and operated by the local Indigenous community to deliver holistic, comprehensive, and culturally appropriate health care to the community which controls it (through a locally elected Board of Management)”. This implies that ACCHS are accountable to the local community through the Board of Directors (who are elected from the Indigenous community) and the funders of the delivery of the health services through annual reports and service agreements. ACCHS are ‘comprehensive primary health care service’ providers and managed the Indigenous community. Community control ensures that the care provided is culturally safe and appropriate, thereby improving access to the services. About 60% of the primary funding for ACCHS comes from the Australian government administered through OATSIH. State/Territory governments contribute about 30% through some specific programs such as alcohol and illicit drug programs. The rest comes from the non-government sector.
2.5 Summary: Implications for Priority Setting The government response to addressing the barriers of access to the universal Medicare for example, has been twofold: the reform of the health system over the years to improve Indigenous access to mainstream services; and the provision of Indigenous-specific health care programs to meet Indigenous unique health care needs. These reforms can be traced to the time when the first Commonwealth assumed the responsibility of Indigenous health in 1967. Subsequent changes to the structures and fiscal flows in the funding of Indigenous health have failed to significantly address the disparities.
Page 53
Historically, mainstream services have failed to provide appropriate and accessible services to Indigenous Australians. Some steps to address the situations have been taken by various governments since the 1970s. One of the steps involves the development of ACCHS providing comprehensive primary health care to Indigenous Australians and some of the changes to the way in which mainstream services and programs operate. Since the establishment of the OATSIH in the 1990s, a more strategic approach to national action across the health system has been developing and focusing on improving the lives of Indigenous Australians. While progress has been made in some aspects, there is still a long way to go if Indigenous and Torres Strait Islander people are going to have equitable access to an effective health care system (DOHA 2009).
Recognising the multiple and inter-related determinants of health would imply that the solutions also ought to be multiple and inter-related. These solutions would need to target education issues, housing, infrastructure, employment and economic development, land ownership, social welfare and community capacity development. Given the myriad of existing strategies to address Indigenous health problems, economic evaluation evidence would allow decision-makers to select efficient strategies – those that work but offer value for money. Such efficient strategies would imply that the existing resources would go a long way to contribute to the closing of the gap between Indigenous and non-Indigenous Australians. In Chapter 7, a survey of key decision-makers will be done to gauge the potential for an economic approach to be acceptable to decision-makers and potential areas for its trial.
Knowing that most of the determinants of Indigenous health lie outside the health sector, it would be important for the concept of benefit from services to include both medical and non-medical benefits. Justification for such a concept of benefit can be found in the Indigenous health construct and in the 2001 Inquiry into Indigenous Funding (CGC 2001).
Three key issues have been demonstrated in this chapter. First, the health disparities between Indigenous and non-Indigenous Australians are closely linked to the social,
Page 54
economic and historical factors that Indigenous Australians face. Second, addressing access to health care services, both geographical and perceived access issues, can potentially improve the health of Indigenous Australians. Third, that despite the government health reforms and priority setting efforts since 1967, when the Commonwealth assumed responsibility of the health of Indigenous Australians, the health outcomes of Indigenous Australians are still poor, hence the need for improved priority setting.
In recognition of the health inequalities and relative deprivation Indigenous Australians face, the Australian government has recently made a commitment to closing the life expectancy gap between Indigenous and mainstream Australian health indicators as one of the key government objectives (COAG 2009). This commitment represents a coordinated, multi-sector approach to addressing Indigenous health. This commitment is reflected in the government objectives that go beyond the health sector to include education, employment and children education. Two key government objectives that form the “Closing the Gap” group of objectives that can be regarded as falling under the direct jurisdiction of the health sector:
Closing the life expectancy gap within a generation; and
Halving the gap in mortality rates for Indigenous children under five years of age within a decade.
(COAG 2009).
It is possible that improving Indigenous access to mainstream health care services, and selecting interventions and/or groups of interventions which offer value for money, can contribute to the closing of the health gap. The challenge is how to choose which intervention or group of interventions work in addressing the Indigenous access issues identified. Most importantly, an assessment of the performance of such interventions in an Indigenous health context needs to recognise Aboriginal health construct. Such an assessment would involve assessing both costs and benefits of such interventions, aimed at addressing Indigenous access to mainstream health care services. The costeffectiveness results (economic evaluation evidence) would then be used to inform priority setting in Indigenous health. Furthermore, improving the performance of the
Page 55
Australian health system in addressing Indigenous health issues would likely benefit from economic evaluation evidence. This is an issue that will be investigated in Section B. This would pave way for a priority setting framework, which incorporates both social justice issues and economic evaluation in its context.
Page 56
Part B: Guidelines for Priority Setting in Indigenous Health “The ideas of economists and political philosophers, both when they are right and when they are wrong, are more powerful than is commonly understood. Indeed, the world is ruled by little else” – (Keynes 1936).
3. Chapter 3: Normative Economics for Priority Setting “Be anxiously concerned with the needs of the age ye live in, and centre your deliberations on its exigencies and requirements” – (Baha'u'llah 2006).
3.1 Introduction In the previous section, a review of the health status of Indigenous Australians indicated that, despite the government effort in solving Indigenous health problems, there was a widening gap in health outcomes between Indigenous and other Australians. Closing the gap requires improving the priority setting processes, among other things. In particular, the use of clear and transparent criteria for priority setting will improve the government’s priority setting efforts.
The purpose of this chapter is to determine the contribution from normative economic theory to guide priority setting. In order to do so, economic principles to guide priority setting are presented, issues surrounding the specification of the SWF are discussed, and the need for a clear concept of benefit that reflects the attributes of the objective being pursued is discussed. The following text is intended to provide a conceptual normative economics framework which the rest of the thesis is based on. It is argued
Page 57
that priority setting is based on the need to make trade-offs on the stated objectives due to the scarcity of resources. The social welfare function (SWF)21 is discussed as one way to aggregate social welfare in order to inform the development of normative guidelines for priority setting. The notion of scarcity was introduced in Chapter 1, and it is restated here. Resources for addressing the numerous Indigenous health problems are scarce, and so decision-makers make choices on what services to provide either in the short term or in the long term (Henderson 2005; Folland, Goodman et al. 2007). Making such choices due to scarcity of funds implies that some opportunities to address Indigenous health are lost – the opportunity cost. This implies that the decision-making ought to minimise the opportunities forgone. Decision-making in practice happens at the margins22, as small incremental adjustments to an existing plan of action, adjustments at the edges of the plan, according to the concept of marginal analysis (Mooney 2002). For example, what happens to the health of Indigenous Australians if a little less or more of an activity takes place? Marginal analysis can be defined as an economic technique that focuses attention on the incremental changes to resources, programs, cost or outcomes of a service to maximise health outcomes to the population targeted, given the allocated budget (Earl-Slater 1999; Culyer 2005). Put simply, marginal analysis is a tool to achieve efficiency23. Two concepts included in marginal analysis are marginal benefit
21
The overall welfare of a society and it is generally assumed by economists to depend upon individual welfares of the individuals who make up that society Culyer, A. (2005). The Dictionary of Health Economics. Cheltenham, Edward Elgar.. 22 The margin is defined as the extra increase or decrease, either in benefits or costs, of a specific activity or intervention. 23 Three main types of efficiency exist: technical, production, and allocative efficiencies Hurley, J. (2000). An Overview of the Normative Economics of the Health Sector. Handbook of Health Economics. A. Culyer and J. Newhouse, Elsevier. Volume 1A.. Technical efficiency refers to minimising factor inputs for a given output (without necessarily taking into account factor prices). It corresponds with being on an isoquant. Productive efficiency is about minimising the cost of producing a given output (taking into account input factor prices). This corresponds to the tangency between an isoquant and an isocost line. Techniques such as: Ordinary Least Squares, Data Envelopment Analysis, and Stochastic Frontiers techniques would be useful for ascertaining productive efficiency Ganley, J. and J. Cubbin (1992). Public Sector Efficiency Measurement: Applications of Data Envelopment Analysis. London, North Holland, Skinner, J. (1994). "What do stochastic frontier cost functions tell us about inefficiency?" Journal of Health Economics 13: 323-328, Zuckerman, S., J. Hadley, et al. (1994). "Measuring hospital efficiency with frontier cost functions." Journal of Health Economics(13): 255-280, Hollingsworth, B., P. Dawson, et
Page 58
and marginal cost. The marginal benefit of a good or service is the additional benefit derived from the consumption or use of the health care service, or the additional benefit from the specific use of the good or service that would be abandoned in response to a given decrease (Gold, Siegel et al. 1996). In general, benefit from an intervention or service should reflect the objective being pursued by the decisionmakers, such as health gain to an individual. In the case Aboriginal health, such an objective should include benefit to an individual and the community in general. How to conceptualise such benefit from an Aboriginal perspective will be attempted in Chapter 8.
The marginal cost is the additional or incremental opportunity cost that comes from an increase in the use of a resource or service. The relationship between marginal costs and marginal benefits yields the efficiency principle. Efficiency involves, in a restrictive sense, maximising the benefits from an activity or minimising the opportunity costs to society of any given resource allocation or expenditure (Phelps 2003; Drummond, Sculpher et al. 2005; Feldstein 2005; Henderson 2005; Folland, Goodman et al. 2007). Allocative efficiency24 incorporates demand side issues with productive efficiency and refers to an ‘optimal’ or ‘maximised’ production of output and distribution of output in line with what consumers ‘value’ (Gold, Siegel et al. 1996; Hurley 2000; Jacobs, Smith et al. 2006; Folland, Goodman et al. 2007). It is the key concept in welfare economics.
3.2 Welfare Economics Welfare economics is the normative branch of economics that uses microeconomic techniques to simultaneously determine allocative efficiency within an economy and the income distribution associated with it (Culyer 2005; Drummond, Sculpher et al. 2005). It links a competitive market mechanism with Pareto optimality, by stipulating a
al. (1998). "Efficiency measurement of health care: a review of non-parametric methods and applications”." Health Care Management Science 2: 161-172.. 24 Allocative efficiency refers to the best outcome to the consumer at minimum cost – thus takes into account technical efficiency and the consumer preferences Pindyck, R. and D. Rubinfeld (1998). Microeconomics. New Jersey, Prentice Hall..
Page 59
social welfare improvement from one inferior social state to a superior one. Social welfare refers to the overall welfare of society, and it can be specified, with strong assumptions, as the summation of the welfare of all the individuals in the society (Culyer 2005). It can be measured from individual preferences25 using ‘utils’, dollars or using Pareto efficiency. Utility relates to the wellbeing, satisfaction or pleasure of an individual, and it is defined as: “that property of any object whereby it tends to produce benefit, advantage, pleasure, good or happiness; or to prevent the happening of mischief, pain, evil or unhappiness of the party whose interest is considered” (Bentham, (1789); see also (Earl-Slater 1999; Suzumura 2002; Culyer 2005; Drummond, Sculpher et al. 2005). There are two mainstream schools of thought in welfare economics: the early Neoclassical (N-C) Welfare or Cardinalist economists; and the New Welfare or Paretian economists.
The N-C approach was developed by Sidgwick, Alfred Marshall, Pigou and others in the early 19th century and the approach assumes the following (Suzumura 2002; Just, Hueth et al. 2004; Feldman and Serrano 2006):
Individual utility is cardinal or measurable;
Individual preferences are stable and exogenously determined;
Utility decreases with additional consumption of the good or service, the law of diminishing marginal utility; and
All individuals have interpersonally comparable utility functions in social planning and ethics.
In N-C welfare approach, it is possible to construct a SWF with these assumptions, simply by summing up all the individual welfare/utility functions, the foundation of Utilitarianism (Earl-Slater 1999; Suzumura 2002; Clarke and Islam 2004; Culyer 2005;
25
Individual preferences are based on the following fundamental assumptions: completeness; transitivity; and rationality. Completeness here implies that people know what they like among the set of things and so are able to choose. Transitivity implies that if x is preferred to y, and y is preferred to z, then x is preferred to z. Rationality involves individuals being consistent with their choices – if they preferred the colour blue yesterday, one would expect them to prefer the same colour today, ceteris paribus. These assumptions provide the basis for establishing a numerical or utility function which represents a person’s preferences, thereby providing the link between preferences and utility derived from the consumption of the goods or services.
Page 60
Drummond, Sculpher et al. 2005). Actions of consumers and producers under Utilitarianism are judged by their contributions to the subjective wealth, happiness, moral worth or satisfaction of individuals – called ‘utility’. The decision criteria involve maximisation of the sum of individual utilities, which lends itself to mathematical elegance using laws of marginal utility (Culyer 2005).
The New Welfare (Paretian) economics approach, on the other hand, arose due to the recognition that utility is neither interpersonally comparable nor cardinal. This approach assumes that utility derived from consuming a service can be ranked into worse, equal or better but is not measurable or interpersonally comparable. It is based on the work of Pareto, Edgeworth, Hicks, Slutsky, Kaldor, and others (Suzumura 2002; Clarke and Islam 2004; Just, Hueth et al. 2004; Feldman and Serrano 2006). Efficiency is assessed using the Pareto efficiency criteria and the Kaldor-Hicks compensation tests. Issues of equity such as income distribution are specified in the SWF. Many economists use Pareto efficiency as their efficiency goal (Earl-Slater 1999; Suzumura 2002; Culyer 2005; Drummond, Sculpher et al. 2005).
Some key terms frequently used in new welfare economics, especially when the topic of social welfare is being discussed, include Pareto efficiency, optimality and improvement (Feldman and Serrano 2006). Pareto efficiency is the state of the world in which no one can have his or her welfare improved without having someone else’s welfare reduced (Earl-Slater 1999; Culyer 2005; Drummond, Sculpher et al. 2005). Pareto optimal/optimum is an allocation in which it is not possible to improve the welfare of someone/people without making at least someone else worse off. An allocation X is Pareto superior to an allocation Y; if all people like X and Y and some people actually prefer X to Y. Movement from an allocation Y to X is a Pareto improvement – a change such that the welfare of at least one person can be improved without the welfare of anyone else being reduced. To achieve such allocations under welfare economics could be done using the competitive market mechanism.
Page 61
3.2.1 Role of the Competitive Market in Resource Allocation Under welfare economics, fundamental relationships exist between the competitive market mechanism to allocate resources and achievement of efficiency. When a competitive market exists, the competitive equilibrium26, exists ensuring that maximisation or Pareto optimality is achieved (Rice 2003; Just, Hueth et al. 2004; Donaldson, Gerard et al. 2005; Henderson 2005). Thus, with an appropriate initial income distribution and price vector, the invisible hand of competition transforms private greed into social welfare. This constitutes the first fundamental theorem of welfare economics.
3.2.1.1 Theorems of Welfare Economics
The first fundamental theorem of welfare economics states that if all the market participants have monotonic self-interested utility functions, the competitive market mechanism automatically distributes goods among them in an optimal manner (Hammond 1992; Rice 2003; Just, Hueth et al. 2004; Donaldson, Gerard et al. 2005; Henderson 2005; Feldman and Serrano 2006). The competitive market system works automatically – prices rise in response to excess demand and prices fall in times of excess supply. The competitive market, with an infinite number of self-interested agents, brings about allocatively efficient distribution of resources automatically without government intervention. The first theorem fails to address the distributive issues associated with the market allocation of resources. The allocation may be efficient but not necessarily socially desirable, especially if the allocation is dependent on the initial endowments of the market participants – potentially exacerbating inequalities (Mishan 1981; Hammond 1992). In recognising the equity problems with the theorem, the proponents introduced a second theorem to try to address the weaknesses of the first theorem.
26
A competitive equilibrium is an economic concept in which a set of prices that make the markets to clear. This is the situation in which everyone chooses the bundle of goods and services that maximises its utility subject to its budget constraints. This is the case where the supply is equal to the demand of the commodity at a given set of prices.
Page 62
The second theorem is almost the converse of the first theorem – it states that the modified competitive market mechanism (with its price system) can be used to get the allocation to any desired Pareto optimal allocation (Rice 2003; Just, Hueth et al. 2004; Donaldson, Gerard et al. 2005; Henderson 2005; Feldman and Serrano 2006). It is assumed that the price system is distributionally neutral, and that an appropriate money transfer system is in place that can address equity issues. Such a transfer mechanism is considered the responsibility of the government. Thus, when a perfectly competitive economy exists, the consumers in such an economy will always seek to optimise their consumption of the goods and services, and the producers respond to the demand for such commodities, without government intervention (Hammond 1992). The optimal points of consumption for each individual in such an economy generate the loci of points of efficiency in consumption or utility functions (Mishan 1981; Rice 2003; Just, Hueth et al. 2004; Donaldson, Gerard et al. 2005; Henderson 2005; Feldman and Serrano 2006). Choosing between the points of loci as the socially desirable allocation employs some compensation principle.
3.2.1.2 The Compensation Principle
The compensation principle defines improvement from a resource allocation such that the winners could potentially compensate the losers and still be at least as well off. An example of such a compensation principle is the Pareto criterion (Mishan 1981; Just, Hueth et al. 2004): 1.
The strong Pareto criterion, which states that an allocation, X, is Pareto preferred to (or Pareto dominates) another, Y, if at least one person is better off, and no one is worse off, with X than with Y.
2.
The simple Pareto principle, allocation X, is preferred to another, Y, if it is potentially Pareto preferred, that is, if it is hypothetically possible to undertake lump-sum redistribution from X to achieve an allocation that Pareto dominates Y.
Page 63
Another compensation criterion is the Kaldor-Hicks compensation criterion where both gainers and losers theoretically agree that the proposed activity will move the economy toward Pareto optimality (Mishan 1981; Just, Hueth et al. 2004). The social welfare change is considered a Pareto improvement if the assessed winners’ gain is greater than the assessed losers’ loss, regardless of whether the change when implemented would actually involve the payment of any compensation. This raises equity issues.
The key issues with the Paretian criteria is that they focus on the preferences of the specific group of participants involved while ignoring the interests of individuals whose very existence were affected by the outcome of the decision-making process (Mishan 1981; Clarke and Islam 2004; Just, Hueth et al. 2004). Secondly, the compensation test does not lead to a ranking of policy sets containing the first best set of allocations, because allocations are assumed not to be comparable. This renders it practically useless because it becomes difficult to pinpoint the level of an economic activity which is efficient (Culyer and Evans 1996; Just, Hueth et al. 2004). Furthermore, the criteria also assume that economic activity involves costless transactions, and actual compensation of those who are hurt by an economic activity does not necessarily take place, as it is assumed that the overall sum of social welfare would be much higher than before the economic activity. Lastly, the criteria do not address equity issues associated with an economic activity. Due to these weaknesses with the compensation criteria, SWFs were used as a means to choose which allocations were socially desirable.
3.2.2 When Markets Fail in Resource Allocation27 Under the N-C welfare economic theory, perfectly competitive markets alone can always achieve optimal (efficient) allocation of resources in the health without government involvement (Mishan 1981; Getzen 1997; Phelps 2003; Rice 2003; Feldstein 2005) if certain conditions hold. These conditions include the existence of
27
Market failure refers to a case where markets or systems of markets are unable to provide goods and services at all or at an economically optimal level to society (Pyndick and Rubinfeld 1998).
Page 64
infinitely many suppliers and consumers in the market28, availability of full information pertaining to benefits and costs, such as risks, and that consumers are rational enough to make decisions in their best interests and aim to maximise consumption with any given resources. Furthermore, the theory assumes that consumers know what is best for them, the notion of consumer sovereignty. Thus, under perfectly competitive markets, privately determined consumption leads to the efficient allocation of resources.
However, when prerequisites for ‘perfect competition’ do not hold, the markets fail to efficiently allocate resources. This could be due to: information asymmetry, externalities29, market power, and public goods in the market (Mishan 1981; Pindyck and Rubinfeld 1998; Phelps 2003; Rice 2003; Feldstein 2005; Folland, Goodman et al. 2007). Information asymmetry occurs when one party to a transaction, such as the health provider, has more or better information than the other party (the patient) (Donaldson, Gerard et al. 2005). In Indigenous health, health care providers such as general practitioners (GPs), have more information about health, health care and risk factors, than Indigenous consumers. The providers can potentially encourage Indigenous Australians to use more services which are financially beneficial to the providers, violating the consumer sovereignty assumption in health economics. So, privately determined consumption may lead to inefficient allocations (Mishan 1981; Donaldson, Gerard et al. 2005).
Another important implicit assumption under the N-C welfarist approach is that each consumer makes consumption decisions without worrying about what other
28
Sufficient market players would ensure ‘price taking’ – no single market player can determine the price of the commodity, and market players are free to enter and exit the market. In this case, welfare economic theory suggests that prices of the goods and services would reflect the equilibrium between marginal costs of the producers and the marginal benefits to consumers. 29 Externalities are abbreviations for external economies and external diseconomies or what Pigou called “uncompensated services or disservices” Rice, T. (2003). The Economics of Health Reconsidered. Chicago, Health Administration Press.. These are third-party effects of an economic activity, and may be positive or negative. Positive externalities occur when a third party benefits from an economic activity in which it is not involved, whereas negative externalities occur when the costs are borne by a third party. In most cases, we are concerned with negative externalities which are regarded as socially undesirable products of economic activity.
Page 65
consumers are doing (Varian 1999; Wallis and Dollery 1999; Phelps 2003; Feldstein 2005). It is assumed that consumers pay the full prices for the commodities that they consume. However, when one consumer’s actions affect others, externalities occur and the prices consumers would pay for their actions would not reflect the full costs and benefits involved, resulting in market failure (Mishan 1981; Wallis and Dollery 1999; Phelps 2003). The issue of externalities in health is inherent in the health care good but this is even a bigger issue in Indigenous health, given the low socio-economic conditions Indigenous Australians face (see Chapter 2). In order to correct these externalities, policy options may include: 1.
The use of an optimal (Pigouvian) tax to raise the consumer’s private cost function to be equal to the social cost function – internalising the externality; and
2.
Subsidies to those who are affected by externality as a form of compensation if it is not possible to levy the Pigouvian tax.
(Phelps 2003; Rice 2003; Folland, Goodman et al. 2007).
Another cause of market failure in Indigenous health is the concept of ‘pure public goods’30, a class of goods which private providers may not be able to supply adequately (Feldstein 2005; Folland, Goodman et al. 2007). These goods are characterised by non-excludability31 and non-rivalry32 in their consumption, and so the competitive market may not bring about the desirable distribution of resources automatically without government intervention. The Cooperative Research Centre for Aboriginal and Tropical Health (CRCATH) argued that Indigenous health should be treated as a pure public good issue (CRCATH 2002). This is because improvements in Indigenous health outcomes would not only benefit the Indigenous individuals but the whole Indigenous and the Australian community as a whole. This is a non-excludable
30
Pure public goods are characterised by non-rivalry (one ship’s use of a lighthouse does not affect another ship’s use of the same lighthouse) and non-exclusivity (it is not easy to exclude a person from consuming clean air) in consumption, which is why it is very difficult for the private market to provide these goods (Getzen, 1997). 31 Non-excludability means that no one can be effectively excluded from using the good. 32 Non-rivalry in consumption means that the consumption of the good by one individual does not reduce availability of the good to others.
Page 66
benefit. Governments could do this through direct funding of research or through subsidies on research, and then use it for public education on how to address lifestyle risk factors. Indeed, the public good argument provides an appropriate platform for government involvement in Indigenous health from the theoretical viewpoint.
Equity involves the ethical judgements about the fairness of the distribution of the costs and benefits of the health outcomes in the community, the accessibility of health care services, the fairness of the process of allocation or identification of the gainers and losers (Mishan 1981; Hurley 1998; Hurley 2000; Culyer 2005; Folland, Goodman et al. 2007). It is generally defined in economics as equality in the distribution of some phenomena in a socially desirable way such as wealth, rights, etc but with some added qualification such as ‘equality of need’ (Wolf 1989; Mooney 1996; Mooney 1998; Williams and Cookson 2000). The other type of equity that is often not considered in economics is the procedural equity or ‘due process’. Equity issues in Indigenous health could constitute key objectives of the health care system.
In summary, if the issue of Indigenous health was left to the private market alone, Indigenous Australians would likely face issues of access to and utilisation of healthcare services, due to low private health insurance and low Medicare enrolments. Under the different normative economic frameworks, justification for government intervention in the health care market is dealt with differently (see the next section).
3.3 Normative Frameworks for Priority Setting Since society is collection of individuals, the challenge is how to aggregate individual preferences/welfare into social preferences. Could the societal preferences be similar to the preferences of individuals? If that is the case, then there must be a way of judging which Pareto optimal allocations are better – is allocation A socially better than allocation B? Using the assumptions under individual preferences, such as completeness and transitivity, it can be assumed that there exists a bundle of goods/services which can be consumed and from which continuous utility can be derived (Clarke and Islam 2004; Feldman and Serrano 2006). Thus, there exists a utility
Page 67
function for an individual in which the set of alternatives for the consumer becomes the set of distributions or allocations of services in the economy. At societal level, it implies that the societal utility function can be defined by aggregating the utility functions of the individuals in society using SWFs.
3.3.1 The Social Welfare Function Due to the inadequacies of the previously proposed criteria for choosing between alternative options for the achievement of the social optimum, the focus shifted to the SWF. The SWF can be defined as an expression of the society’s tastes for different economic states, and it is normally a function of individual welfares of the members of the community (Earl-Slater 1999; Varian 1999; Culyer 2005; Drummond, Sculpher et al. 2005). In theory, the individual consumer preferences would be elicited from the individuals and then aggregated to form one social preference or health constitution (Shiell and Mooney 2002).
Whatever the functional form, when a SWF is being formulated, it is possible that individual preferences may be defined over the entire allocations rather than over an individual’s bundle of goods. For example, the SWF can be defined as a function of only individual utilities (Earl-Slater 1999; Varian 1999; Culyer 2005; Drummond, Sculpher et al. 2005): SWF = f (u1, u2… ui). Where ui is the utility of the ith individual.
There may also be interdependencies in people’s utility functions leading to consumption externalities. Varian has listed three properties that would need to be fulfilled when individual preferences are being aggregated to form a SWF that ranks outcomes consistently (Mishan 1981; Varian 1999):
Given any set of complete, reflexive and transitive individual preferences, the social decision-making mechanism should result in social preferences that satisfy the same properties.
If everybody prefers X to Y, then the social preference should rank X above Y.
Page 68
The preference between X and Y should depend only on how people rank X versus Y, and not on how they rank other alternatives. This is a bit restrictive since it excludes horizontal equity and rank-order voting (Slesnick 1998).
Once aggregation of individual welfares has taken place and the SWF established, social welfare is then maximised. Theoretically, social welfare is maximised when the following conditions have been fulfilled (Mishan 1981; Feldstein 2005; Feldman and Serrano 2006): 1.
First condition – pure consumption: When the marginal rate of substitution between any two goods is the same for all consumers;
2.
Second condition – Pure production: When the rate of technical substitution between any two inputs is the same for all industries that use the said inputs;
3.
Third condition – equilibrium: The marginal rate of transformation must be equal to the marginal rates of substitution for all consumers; and
4.
Fourth condition – societal level: Marginal social significance of consumption must be equal to all individuals for all each commodity.
In practice, the SWF needs to be established before maximisation occurs. To do so, two approaches are commonly used to elicit arguments: using a democratic process (using economic criteria, market criteria or voting in a referendum) to elicit individual preferences and then aggregating them; or imposing it on society (using government, experts, or community/opinion leaders). Under the democratic process, three major approaches to specifying the SWF may be used: subjective approach; basic axiomatic approach; and moral justice approach (Mishan 1981; Just, Hueth et al. 2004).
The subjective approach completes the functional form of the SWF on subjective ethical grounds. Bentham (1961) and colleagues argued that, social welfare ought to be the sum of individual utilities with a weighting of 1 – utilitarian functions (Mishan 1981): W = U1 + U2 + ... +Un Where Un is utility for individual n, and the policy qualifies for implementation when W > 0.
Page 69
There are, however, disagreements amongst the proponents of the subjective approach on the specification of the SWF from the democratic process (Just, Hueth et al. 2004). The disagreements stem from the fact that others have advocated for ‘normal’ distributional judgements and yet what is ‘normal’ is not defined. Some have advocated for inequality aversion and yet there was not an agreement on the appropriate level of inequality aversion (Just, Hueth et al. 2004).
The basic axiomatic approach, on the other hand, used mathematics to investigate the existence and form of the welfare function (Just, Hueth et al. 2004). Under this approach, the SWF is based on a set of plausible underlying axioms about individual preferences:
The domain of decisions is unrestricted;
The Pareto principle applies; and
Rankings are independent of irrelevant alternatives.
The Arrow Impossibility Theorem33, however, proved that it is difficult to find a rule that satisfies all of the above properties when aggregating preferences into a SWF, unless the decision-making process is a dictatorship. In response, other authors have suggested that it was the intensity and not the simple rankings that mattered when individual preferences were being aggregated into a SWF (Kemp and Ng 1977). Still others have suggested that the axiom on “rankings are independent of irrelevant alternatives” ought to be dropped34 and certain kinds of rank-order voting should be employed (Fishburn 1973; Sen and Williams 1982). This would provide the analyst with the ability to weigh the gains of the winners against the losses of the losers (Slesnick 1998; Varian 1999). Just et al (2004) argued that the major practical problem with the axiomatic approach was that, even under weaker conditions where voting works, transactions costs of voting and compiling the ranking would be prohibitive.
33
Arrow’s Impossibility Theorem states that if a social decision mechanism satisfies all the three assumptions of the SWF, then it must be a dictatorship: all social rankings are the rankings of one individual (Varian, 1999). 34 Slesnick (1998) explained that it was Amartya Sen who demonstrated that relaxing the interpersonal comparability assumption expands the spectrum of possible SWFs dramatically.
Page 70
The moral justice approach argues that the Arrow impossibility theorem happens as a result of the majority groups acting selfishly – preferring to eliminate consideration for the minority groups. To address this problem requires admitting moral considerations such as impartiality35 and economic justice in the SWF. In these SWFs, a criterion of distributional optimality is suggested which tends to advocate for equality or equal weighting. Champions of the moral justice SWFs include Rawls, Harsanyi, Benthamite, Arrow and others (Mishan 1981; Just, Hueth et al. 2004).
Efforts to develop a generally accepted SWF have not been successful because there is no objective way of making interpersonal comparisons of individual utilities. In practice, a SWF requires the individual utilities to be cardinally measurable so that intensities of preferences can be compared. In contrast, the Pareto and compensation criterions, in which utility is measured in ordinal units, has many applications but is not useful for identifying a unique social optimum. Mishan explains that although one can always claim that a SWF can provide useful insights from attempts to construct them theoretically, they cannot necessarily be translated into practical solutions. At the same time, lay understanding of welfare criteria, analysed in abstract terms can be translated into practical propositions (Mishan 1973). Under different normative frameworks, such as the N-C welfarist approach or the non-welfarist approach, arguments in the SWF can be defined in different ways. Furthermore, the aggregation of these welfares can take different forms.
3.3.2 Normative Neoclassical Welfarist Approach Hurley summarises the key features of the N-C welfarist approach to normative economics to include the following (Hurley 2000): welfarism; consumer sovereignty; utility maximisation (behavioural assumption); and consequentialism. Welfarism is the normative framework that relies on the notion that social welfare is only a function of individual utilities as a foundation of analysis (Hurley 2000; Ng 2000).
35
Moral concerns for equal treatment of individuals.
Page 71
Individual utility/happiness is assumed to be the function of only goods and services. Utility is defined by revealed preferences (through people’s choices of goods and services) which are not distorted by ignorance, imperfect foresight, or misinformation because the individuals are ‘rational36’ and ‘responsible’ in their choice making (Mishan 1969; Mishan 1981). The revealed preferences are determined by consequences, and the process may be conceptualised as ‘consequentialism’37 (Culyer 1998; Mooney and Russell 2003). Consumers are rational or consistent over time and reveal their preferences through their choices of goods and services. Rationality occurs because the preferences are assumed to fulfil the following axioms (Varian 1999):
Completeness axiom: Any two bundles of goods or services are comparable, and so it ensures that a consumer can make a choice between them;
Reflexive axiom: Any bundle would be considered at least as good as itself; and
Transitive axiom: People’s choice behaviour ensures that consumers are able to make ‘best’ choices.
Three more assumptions underpin welfarism: continuity; non-satiability; and exhibition of diminishing marginal rates of substitution, and these could be considered technical assumptions (Varian 1999).
The N-C welfarist framework further assumes that individuals are the best judges of their own welfare (consumer sovereignty) because the market is perfectly competitive with an infinite number of consumers and producers with full market information. The assumption of ‘consumer sovereignty’38 ensures that the free market mechanism under the N-C welfarist framework is accorded the supremacy to allocate resources efficiently. The policy implication is that with full information and no consumption
36
Ng defined rationalism as individual preferences which are free from correctable ignorance, irrationality and non-affective altruism or malice, and irrationality is the case where an individual prefers a bundle of a good which reduces his welfare due to the fact that he may not know all the risks associated with consuming such a good (as in the case of smoking) Ng, Y. (2000). Efficiency, Equality, and Public Policy: With a Case for Higher Public Spending. London, Macmillan.. 37 Consequentialism holds that any policy should be judged on its outcome or consequences, because consumer preferences are based on outcomes Richardson, J. (2000). Empirical Ethics Verses Analytical Orthodoxy: Two Contrasting Bases For The Reallocation of Resources. Twenty Second Australian Conference of Health Economists. A. H. E. Society. Gold Coast.. 38 The consumer sovereignty assumption means that the consumers ultimately dictate that the producers are price takers.
Page 72
externalities, consumers maximise their utility because they are the best judges of their own welfare. Under the N-C welfarist approach, the objective function, reflected in the SWF, involves improving or maximising utility/welfare and it can be reflected in the consumer’s willingness to pay for the service in competitive markets without government involvement (Cowen 2000; Donaldson, Gerard et al. 2005).
Several SWFs have been proposed under the N-C welfarist framework and may include the Bergson-Samuelson SWF, and the axiomatic SWF of Ng, Fleming, Harsanyi and Nash. One popular N-C welfarist SWF is the utilitarian welfare function, also called a Benthamite welfare function. The utilitarian SWF sums the utility of each individual in order to obtain society's overall welfare. All people are treated the same, regardless of their initial level of utility. One extra unit of utility for a starving person is not seen to be of any greater value than an extra unit of utility for a millionaire. The SamuelsonBergson SWF says that social welfare is determined by utilities of all the individuals such that the higher value of the function is preferred. The properties in this SWF with respect to the individual utilities are the same as those in the individual utility function with respect to the goods and services consumed.
However, the N-C welfarist framework for allocating resources is rejected by the nonwelfarist approaches on ethical grounds. For example, the term ‘utility’ itself in welfarism is subject to debate. Richardson refers to at least four types of utility: pleasure/pain as in the Hedonism tradition; psychological (strength of preference) as in anger/satisfaction; behavioural as in the von Neuman-Morgenstein expected utility axioms; and ordinal ranking of preferences as in organisational framework for positive analyses (Richardson 1994). The exact form of utility that should be maximised is not specified in the N-C welfarist frameworks.
Secondly, utility maximisation as the sole or even a primary objective under the N-C welfarist approach is questionable due to the likelihood of preference failures and endogeneity of these preferences (Richardson 2000). It is argued that there is a possibility of preference failure due to the following issues: poor information or misunderstanding of the questions presented; the context in which the questions may
Page 73
be presented; or lack of will/motivation on the part of the respondents. Further, there may be interdependencies in people’s utility functions leading to externalities, and consumer tastes are not predetermined; they are a result of past and/or present consumption experience (Rice 2003). Also, the utility concept used in the N-C welfarist approach focuses on emotional responses to goods’ and services’ characteristics, which can be influenced by social status or physical disabilities (Culyer 1990).
Furthermore, there is contention as to the primacy of individual utilities as opposed to including community benefits in the SWF. Some authors (Shiell and Hawe 1996; Mooney 1998) have argued that in programs such as public health and Indigenous Australian health programs, program theory is based on the community as opposed to the individual. When the community is viewed ecologically, the sense of community and community competencies are properties of the community. It is therefore questionable whether aggregating the effects on individuals can capture the full benefits of community action (Sheill and Carter 1998). Efforts to address the above criticisms led to the development of the non-welfarist approaches.
3.3.3 Non-welfarist Approaches: Extensions beyond Welfarism The non-welfarist approaches include all those, excluding the N-C welfarist approaches, which allow non-utility attributes to define the SWF. These approaches may include extra-welfarist, communitarianism, capability approaches, decisionmaking approach (DMA) and others. These approaches themselves differ markedly from one another, but all share one thing in common: that they are not welfarist (Drummond and McGuire 2001).
Development of the non-welfarist frameworks (Drummond and McGuire 2001) arose as it became clear that the ‘merit goods’ argument, initially raised by Musgrave in his Theory of Public Finance, could not fit in the N-C welfare economics theory, necessitating departure from N-C welfarism (Musgrave 1959). Attempts had been made by some economists such as Culyer (1971) to include merit goods in the N-C welfarist approach, but these efforts were generally regarded as unsuccessful (Carter 2002).
Page 74
Additionally, the capability approach offered an alternative framework for assessing social welfare, which solved the Arrow impossibility theorem, using functional ‘capabilities’ instead of utility. Sen explained that ‘functionings’ might include basic human functions such as “moving, being well-nourished, being in good health, being socially respected” (Sen 1985). He then described ‘capability’ as the extent to which a person is able to function in a particular way, whether or not he or she chooses to do so. Emphasis was placed on how human beings actually function and on their capabilities. For example, poverty, lack of access to services, low levels of education, institutional racism or unhealthy environments can be seen as capability deprivation (Sen 1982; Sen 1992).
The influence of the capabilities approach on the health care sector started in the mid to late 1980s, when Culyer attempted to bring a more theoretical basis to the nonwelfarist perspective by building upon Sen’s notions of functioning and capabilities (Coast, Smith et al. 2008). The result was the development of the extra-welfarist approach within health economics. Extra-welfarism is defined as a normative framework which supplements traditional welfare in the SWF with other ‘non-goods characteristics’ of individuals (Culyer 1989; Culyer 1990). In health care, the relevant characteristic is health and implies that the health status information directly influences the social states that individuals prefer, contrasting sharply with welfarism (Hurley 1998). The function may be defined as: SWF = f (u1, u2… ui, v1, v2… vi) Where ui, & vi are the utility and non-utility attributes of the ith individual’s consumption of goods and services respectively.
For example, according to Culyer, two pivotal concepts that emerge from the characteristics39 of people (capabilities approach) are ‘deprivation’ and ‘need’ (Culyer 1990). He stated:
39
Characteristics of people refers to attributes that describe a person and these may include their genetic endowment of health; his or her SES; moral-worth and deservingness; utility; severity of pain; and/or equity, fairness and social justice (Culyer, 1989).
Page 75
“If the characteristics of people are a way of describing deprivation, desired states, or significant changes in people’s characteristics, then commodities and their characteristics are what are often needed to remove their deprivation”.
It is the non-utility characteristics such as deprivation and need that create the demand for healthcare. The focus on deprivation and need implies that need and equity determine the demand for a health intervention, and not necessarily the willingness to pay as advocated in the N-C welfarist approach. Culyer advocated the use of Quality Adjusted Life Years (QALYs) as the unit of measure of health, and the weights40 used in the QALYs are not necessarily derived from utility values (Culyer 1990). Other authors have argued that other important considerations other than QALYs (such as illness severity or the concern for the welfare of others) should be included as attributes of the SWF (Menzel, Gold et al. 1999; Nord 1999; Mooney and Russell 2003). The policy implication is that health is the key outcome in the SWF as opposed to utility, and equity attributes based on the key characteristics of people should be included in these outcomes as non-utility attributes of the SWF.
Mooney and Russell (2003) criticised extra-welfarism for being stuck with the use of individual preferences to determine the values or weights attached to the arguments in the SWF, for being consequentialist, and for not allowing differential relevance of valuations or preferences in different contexts. They argue that communitarianism or ‘communitarian claims’ approach addresses these limitations. Communitarianism advocates for the use of community preferences in the SWF, and that these same members are the most appropriate to determine what weights to attach to these preferences (Mooney 1998; Mooney and Russell 2003). Further, the approach emphasises the need to differentiate individual from community preferences, and proposes that the preferences of the disadvantaged groups should be used to address their disadvantage.
40
The weights could be based on any factor (such as age, social/occupational roles/family responsibilities, initial health status, etc) that affects how individuals value producing health for a person with the characteristic as opposed to a person without (Hurley, 1998).
Page 76
In theory, it aims to attain some degree of coherence and transparency in the assessment of equity from an economics perspective. It classifies reasons for a person’s claim for a good into two sectors: one being ‘claims’41, and the second being ‘other reasons’ (Mooney 1998). Equity is seen to be consistent with community perspectives on how individuals should be treated relative to one another. Communitarianism is especially appealing Indigenous health. This is because the sense of the community and community competencies are properties of the community, and it is therefore questionable whether aggregating the effects on individuals can capture the full benefits of community action (Shiell, Seymour et al. 1998). Thus, equity is defined by how claims are established, and then how different claims are weighted.
The influence of the capability approach in health care decision-making today continues to grow and it is being accepted as the more appropriate normative framework for the evaluation of health care programs and interventions. For example, Coast and colleagues argued that the significance of the health status in the SWF was a direct result of the capability approach (Coast, Smith et al. 2008). This perceived importance of health status in the SWF led to the development of DMA with a focus on the societal objectives as determined by the decision-makers (Weinstein and Stason 1977; Sugden and Williams 1978; Culyer 1980). Sugden and Williams advocated for the use of directly obtained or inferred values of decision-makers in the specification of the SWF (Sugden and Williams 1978). Two methods exist for determining the decisionmakers’ objectives: government policy documents /guidelines; and carefully elicited/ inferred objectives from surveys, for example, see the surveys in Chapter 5 and Chapter 7.
41
‘Claims’ are defined as reasons backed by a notion of duty for one’s claim for a good.
Page 77
3.4 Applied Welfare Economics in Health Care Given the disagreements within and between the proponents of different normative frameworks on what constitutes individual and social welfare, the key challenge in welfare economics is how to measure welfare improvements in practice. Since welfare is not directly observable, welfare change would be inferred from what an individual gets before and after the implementation of change. The next task would be to postulate the relationship between the observed item and individual welfare. In other words, how can an improvement in Indigenous health be captured or measured given the implementation of the government policy?
Theoretically, individual welfare would be measured before and after the allocation has taken place using appropriate units. For example, the N-C welfarist approach tends to use consumer surplus or compensating variations as measures of individual welfare. Individual welfare scores would then be aggregated using an appropriate welfare function depending on the normative framework guiding the evaluation to estimate social welfare. In this case, crude SWF can be constructed from the results of the measurement of social welfare.
3.4.1 Assessment of Social Welfare Change In theory, the subjective valuation of the different combinations of the goods to assess social welfare can be conceptualised using the ordinal or cardinal approaches. As already indicated, the cardinal approach assumes that individual utility is measurable and individually comparable, while the ordinal approach disagrees and instead assumes that utility ought to be measured on an ordinal scale. Ordinal approach also argues that utility cannot be comparable but can be ranked in the order of preference because utility is a subjective phenomenon. However, when individual choice is being assessed under uncertainty, it requires that utility should be cardinally measurable (Drummond and McGuire 2001; Clarke and Islam 2004). This is because the concepts of expected utility and risk attitudes are second derivatives of individual utility
Page 78
functions. For an individual to be able to see a meaningful difference between two utility levels, utility ought to cardinally measurable.
The commonly used method to observe individual utility is to convert it into monetary value (Brent 2003; Just, Hueth et al. 2004; Feldman and Serrano 2006). For example, if an individual prefers social state A to social state B, then that individual pays a monetary sum to get the preferred option B. This is the compensating variation (CV), a measure of how much a consumer is hurt by a price increase. It is a cardinal measure although the change in utility itself may not be cardinal. CV is the metric behind Kaldor-Hicks compensation criterion; if the winners can potentially compensate the losers from a particular policy change, then that policy is efficient, even if the compensation does not take place (Just, Hueth et al. 2004; Culyer 2005). Mathematically, CV can be written as an expenditure function f (Px, U). CV = new expenditure less expenditure on the old bundle at current prices = f (p1,u1) − f(p1,u0) where p0 and p1 are the old and new prices respectively, and u0 and u1 are the old and new utility levels respectively.
The equation can be interpreted as: under the new price regime, the consumer would accept CV in exchange for allowing the change to occur (Just, Hueth et al. 2004; Feldman and Serrano 2006). This interpretation of the CV is closely related to another measure of consumer surplus or welfare called equivalent variation (EV), which measures the amount of money a consumer would be willing to pay to avoid a policy change, before it happens if he/she reckons it is not to his/her best interest for the policy to take place. Actual amounts of CV can be elicited subjectively using a revealed preference approach which is based on the actual information on people’s expenditures. It can also be elicited using stated or expressed preferences, also known as contingent valuation, based on hypothetical market-like situations. Estimating CV is also known as the WTP approach to valuation of utility (Culyer 2005).
Eliciting subjective values can also be done using experienced utility (Kahneman, Wakker et al. 1997). Experienced utility is a direct measure of utility and does not
Page 79
require individuals to somehow place themselves in a particular hypothetical situation. Cost and colleagues argued that there are two advantages of measuring experienced utility: it takes account of an individual’s adaptation to changed circumstances; and it takes away the need for monetary valuation of benefits (Coast, Smith et al. 2008).
3.4.2 Implications for Economic Evaluation in Health Care The common approach to social welfare assessment in the new welfare economics is the use of the Kaldor-Hicks efficiency criterion to assess costs and benefits of various options. These benefits and costs are then combined to ascertain the overall benefit relative to the costs (economic evaluation) using Cost Benefit Analysis (CBA) (Drummond and McGuire 2001; Brent 2003; Rice 2003). Economic evaluation of health care stems initially from standard Paretian welfarist views and it follows the basic principles of consumer sovereignty and Pareto optimality (Boadway and Bruce 1984; Rice 2003; Folland, Goodman et al. 2007). In CBA, for example, benefits are measured in monetary units and are based on individuals’ willingness to pay for a service and it is theoretically relevant to the compensation principle as the means to achieve both absolute and relative efficiency (Mishan 1988; Brent 2003; Zerbe Jr and Bellas 2006).
Economic evaluation measures the relative efficiency of action or intervention undertaken by the decision-maker in health care. Theoretically, economic evaluation assesses efficiency or an ‘optimal’ allocation of resources. The decision context occurs when the marginal social benefit (MSB) of consuming a commodity is equal to its marginal social cost (MSC) to society. Mathematically (Gold, Siegel et al. 1996; Drummond, Sculpher et al. 2005; Folland, Goodman et al. 2007): MSB1/MSC1 = MSB2/MSC2 = …. = MSBi/MSCi. Where 1, 2, 3, …i represent various interventions.
If the MSB1/MSC1 is higher than MSB2/MSC2 then it is preferable to increase resources in intervention 1 up to a point where the two ratios are equalised to achieve allocative efficiency.
Page 80
The key differences between the existing approaches to economic evaluation, such as CBA, cost minimisation analysis (CMA), cost effectiveness analysis (CEA), and cost utility analysis (CUA), may arise in different ways. One school of thought bases its nomenclature of the approaches to economic evaluation on the units of outcome measures (Johannesson 1996; Drummond, Sculpher et al. 2005). For example, if the outcome units are in natural units, then the approach is CEA. If the units are in subjective valuations such as QALYs or Disability Adjusted Life Years (DALYs)42, then it is CUA. When the units of measure are in monetary units, the approach is a CBA. Another group of economists have argued that the differences are due to the type of efficiency the techniques actually measure – technical or allocative efficiency. For example, Donaldson and others argue that CBA measures allocative efficiency and CEA measures technical efficiency (Donaldson 1998; Currie, Donaldson et al. 2000). Gold and colleagues instead argue that CUA is a subset of CEA but from a societal perspective (Gold, Siegel et al. 1996). These differences have been the subjects of debate in the literature (Drummond and McGuire 2001) but resolution of this argument is not the intent of this thesis.
All these frameworks, however, have limitations because of the units employed for valuing benefits/outcomes. For example, when the natural units (such as lives saved, injuries prevented or cases screened) are used as measures, the results cannot be used across different disease interventions. The use of QALYs overcomes the problem of comparative analysis across different disease interventions and can therefore be used for comparative analysis across the whole health sector, but cannot be used outside the health sector. Moreover, issues of measuring the quality of life remain unresolved.
42
DALY is the measure of the burden of disease and injury, and involves a summation of the years lived with disability (YLD) and years of life lost (YLL) for a specific disease or health conditions, where YLD refers to the healthy years lost due to a disability for an incident case of the health condition, and YLL refers to the years lost due to a specific disease for premature death and modelled for the life Zhao, Y., S. Guthridge, et al. (2004). "Burden of disease and injury in Aboriginal and non-Aboriginal populations in the Northern Territory." Medical Journal of Australia 180: 498-502, Culyer, A. (2005). The Dictionary of Health Economics. Cheltenham, Edward Elgar..
Page 81
The monetary unit of measure such as in CBA overcomes all the weaknesses of the preceding units of outcome measures (Brent 2003; Zerbe Jr and Bellas 2006). It can be used for comparative purposes across all sectors in the economy. Additionally, CBA can be used to determine both absolute and relative efficiencies. However, for health outcomes or benefits valuation in health care – especially in life and death situations, two clear difficulties arise with the use of money metric for measuring welfare in CBA. The first issue is related to the distribution of benefits – equity concerns. Knowing that measuring welfare using the willingness to pay concept is determined by the ability to pay, the allocation of health care resources would be ‘skewed’ towards the wealthy. Empirically, van Doorslaer, et al. found that among 10 OECD43 countries included in their analysis, there was strong support for equity in health care as indicated in their policy documents (van Doorslaer, Wagstaff et al. 1993). Also people have been consistently found to be prepared to sacrifice total benefits to achieve equity, even when they knew that they would be in the group which would be hurt by the redistribution (Yaari and Bar-Hillel 1985; Kahneman and Varey 1991). There are several processes for empirically determining the population ethics of resource allocation including: ‘communitarian claims’ by Mooney44 (Mooney 1998); and ‘empirical ethics’ by Richardson (Richardson 2000).
The second issue involves the discomfort society has in valuing life and death issues in monetary terms, resulting in its reluctance to participate in such exercises (Tsuchiya and Williams 2001; Feldman and Serrano 2006). This has led to the emphasis on the use of the natural units such as numbers and/or percentages of the outcomes of interest, and the focus on subjective valuations of individual welfare/benefit. Subjective valuations are undertaken using instruments such as rating scales, standard gamble and time trade-offs to assess the relative severity of the consequences of the diseases. In these valuations, characteristics of severity of diseases such as pain and
43
OECD stands for the Organisation for Economic Co-operation and Development – a loose organisation of 30 countries such as Australia, the United Kingdom, the United States and others sharing a commitment to democratic government and the market economy. 44 The concept of ‘communitarianism’ as a process for building equity into the SWF under extrawelfarism is discussed here despite the fact that Hurley (2000) considers Gavin Mooney to be a welfarist.
Page 82
disability are determined through repeated assessments over time to establish the durations. The data derived may be expressed over an interval scale to estimate the intensity (see Chapter 8 for further discussion).
An example of such valuations is the QALY. QALYs express the sum of individual utility gains from medical interventions elicited using subjective valuation techniques such as time trade-offs. As such, they are potentially a useful measure in deciding how to allocate resources. They should not, however, be equated with the overall value that society places on different health care programmes. The QALYs that society places on different health care interventions is determined by aggregating the number of QALYs gained, but also by a series of distributional and ethical considerations (Drummond and McGuire 2001). In most cases, the QALYs may be weighted to account of the distributional and other ethical issues before the aggregation takes place (Coast, Smith et al. 2008). This should be made clear in the presentation of QALY-calculations, particularly in cost-per-QALY league tables. Thus, benefit from an allocation could be measured in economic evaluation using one, or a combination, of the three broad units: the natural unit, subjective values, or monetary values of the services distributed in the community.
Page 83
3.5 Implications for Priority Setting in Indigenous Health In Indigenous communities where health problems seem to outstrip allocated resources, it is important therefore for any priority setting framework to be able to capture the opportunities lost when choices are made. Such lost opportunities ought to be reflected in the selection of interventions for review or the options for change in ACCHS. The decision context under any economic framework is efficiency, whether defined by a cardinal- or ordinal-based measurement of efficiency. Economic evaluation evidence assists government decision-making to identify options for change which offer value-for-money from investments. The decision context is to maximise benefit for Indigenous Australians from interventions when choices are being made.
The challenge however exists on how to translate the SWF into the concept of benefit for the purposes of economic evaluation to guide priority setting. Some authors favour the use of monetary units as in CBA and others prefer the use of QALYs or DALYs. Given the Indigenous health values and strategies to address their health, this issue will be investigated in Chapter 8, in order to ascertain what ought to capture benefit from health care intervention for priority setting purposes.
It was argued that, in estimating the concept of benefit, the non-welfarist frameworks, the preferred frameworks in this thesis, the SWF includes utility and non-utility attributes such as health and/or issues of equity in addition to individual utility. The challenge, however, is to determine what attributes should enter the SWF. It is recognised and acknowledged in this thesis that individual preferences are important for computing social welfare, but community preferences such as community control, access to land for Indigenous Australians and community development are equally important, and should be incorporated in the Indigenous SWF. The next challenge is to determine what weights should be attached to the arguments in the SWF, and who should be determining them. The key contribution from economic evaluation involves estimating value for money from investments in health. Such estimates are often expressed as marginal social cost for additional social units of health benefit. Such units ought to reflect the social objectives that are being pursued, hence the need for
Page 84
an appropriate concept of social benefit. Thus, any priority setting framework in Indigenous health ought to use the concept of benefit, which is flexible enough to capture both health and non-health benefits. Such a priority setting framework ought to embody the following regardless of the level of rigour:
Recognition of the opportunity costs of the choices made. Making such choices due to scarcity of funds implies that some opportunities to address Indigenous health are lost – the opportunity cost. This implies that the selection of options in decision-making should focus on minimise the opportunities forgone.
Decision-making in practice, requiring marginal analysis, should focus on the incremental changes to resources, programs, cost or outcomes of a service to maximise health outcomes to Indigenous Australians. This yields efficiency, the maximisation of the benefits from an activity or minimisation of the opportunity costs to society of any given resource allocation. A priority setting framework should therefore recognise the role of marginal analysis in the decision context.
N-C welfarist approaches are rejected as a conceptual framework for analysing economic issues in Indigenous health. Non-welfarist frameworks were preferred because these frameworks allow utility, health and other important attributes to enter the Indigenous SWF. It was argued that the SWF should be specified empirically. Specifying the SWF should not be left to the decisionmakers alone, but should include Indigenous community representatives.
Indigenous Australians should be recognised as legitimate claimants of the resources determined, and they should be allowed to specify the arguments in the SWF and what weights should be attached to the stated arguments. The specification of the SWF should not be left to the decision-maker alone or to the Indigenous consumers alone, but should be done empirically by both the decision-makers and the Indigenous representatives.
The concept of benefit should reflect the extent to which the strategy being employed captures all the pertinent attributes of the objectives being pursued, and attributes of the SWF in general. Such attributes may include health, equity, access and cultural sensitivities.
Page 85
A demonstration concept of benefit study will be undertaken in Chapter 8 on how the decision-maker and the Indigenous community representatives can specify the attributes of the SWF in Indigenous health and their associated weights. The study will aim to demonstrate that health and other non-utility attributes such as access to health care services are important attributes that ought to be represented in the SWF.
Page 86
4. Chapter 4: Non-Economic Foundations "Tread ye the path of justice and equity in all things … Justice and equity are twin Guardians that watch over men. From them are revealed such blessed and perspicuous words as are the cause of the well-being of the world and the protection of the nations” – (Bahá'u'lláh 1892).
“An important part of the economist’s task is to find out how well the production and distribution of goods and services conform to the public’s wishes. The first thing to ascertain in this connection is what the public’s wishes are” – (Scitovsky 1986).
“Social justice is what faces you in the morning. It is awakening in a house with adequate water supply, cooking facilities and sanitation. It is the ability to nourish your children and send them to school where their education not only equips them for employment but reinforces their knowledge and understanding of their cultural inheritance. It is the prospect of genuine employment and good health: a life of choices and opportunity, free from discrimination” – (Dodson 1993).
4.1 Introduction In Chapter 3, it was demonstrated that the free market alone cannot yield efficient outcomes in resource allocation, let alone the socially desirable ones, for the following reasons. First, it is nearly impossible to realise a ‘perfectly competitive’ market in Indigenous health. This is because the Indigenous health care market is characterised by market failure, information asymmetry, and incomplete markets, which render the free market system alone practically useless to address efficiency in Indigenous health. Secondly, the N-C economic frameworks fail to adequately address equity and procedural issues that affect people’s access and consumption of health care. This is because these frameworks assume that all consumers face level playing fields in health
Page 87
care markets, by excluding value judgements and initial resource endowments. There is therefore a need for the exploration of the social justice principles that ought to guide priority setting. These principles include: efficiency (utilitarianism); equality (strict egalitarianism); equity; and need (Olsen 1997; Konow 2003; Donaldson, Gerard et al. 2005). These principles will be used, together with the principles from economic theory to develop the normative guidelines for assessing priority setting models in Indigenous health.
This chapter is organised into four key sections: ideologies in priority setting; principles of distributive justice; procedural justice; and key issues for consideration in priority setting. Ideologies in government priority setting are presented in the context of egalitarianism and libertarianism. This is followed by a discussion of distributive and procedural justice principles relevant to priority setting in Indigenous health. The last section of this chapter involves a discussion on the issues for consideration in priority setting.
4.2 Ideologies in Priority Setting In policy areas, there are different interpretations of what constitutes fair treatment and a just share of resources, and these interpretations by governments are often influenced by ideologies. Different ideologies have different conceptions of what constitutes a 'just and fair society'. For example, libertarianism views access to health care (Table 5.1) as a reward for one’s personal responsibility, and it is very consistent with the N-C welfarist approach but somewhat narrower (Williams 1988). In a ‘perfectly competitive’ market, it is assumed that people will choose their utility maximising strategy if government leaves them alone to make their own decisions, discussed in Chapter 3. For those who are disadvantaged or unable to fend for themselves, privately administered charity can help them maximise their utility. The policy implication under the N-C welfarist approach is that people who can afford and are willing-to-pay for their own health care services should be allowed to use their income and wealth to access more health services than their counterparts. Thus, health care should be privately provided and accessed according to one’s ability to
Page 88
pay. The N-C welfarist framework as an ideology for use in Indigenous health was rejected in Chapter 3 as a guiding normative framework for priority setting.
Egalitarianism, on the other hand, advocates for equality of opportunities, such as access to health services, as a right for every citizen regardless of wealth (Table 5.1). Gosepath argued that equality of access should be the criterion for resource allocation and should be equated to the many freedoms people actually enjoy (Gosepath 2007). This provides the platform for the human rights-based approaches to priority setting – the right to equal health status for Indigenous Australians. Given the scarcity of resources, each claim for resources has to be judged in relation to the claims of all other members of society and all available resources. For example, to bring the Indigenous health status to a level comparable with the rest of the Australian population, Indigenous Australians ought to make more claims on the resources.
The policy implication of this position is that government should have a dominant role in the funding and provision of health services, since it has more ability to compare the claims of every one in society, ceteris paribus. Priority should therefore be accorded to those in higher ‘need’ and should reflect social values. An egalitarian health system may include the following specific objectives (WHO 2000):
Maximising population health;
Reducing inequalities in population health;
Maximising health system responsiveness;
Reducing inequalities in responsiveness; and
Financing health care equitably.
The challenge is to establish which social justice principles are appropriate for priority setting in Indigenous health. These challenges are discussed in the next section.
Page 89
Table 4.1: Attitudes typically associated with libertarianism and egalitarianism
Source: (Williams 1988)
Page 90
4.3 Principles of Distributive Justice for Priority Setting 4.3.1 Efficiency Principle The efficiency principle is a moral doctrine which states that a human being should act so as to produce the greatest possible happiness or pleasure (Konow 2003). It is the main principle in the welfare economics family of theories, such as Utilitarianism. For Utilitarians, everyone counts as one and no one is more than one, the interests of all are treated equally without consideration of contents of interest or an individual's material situation (Gosepath 2007). Efficiency principle as a type of justice is a key criterion for allocating resources within the N-C welfarist framework. The N-C welfarist framework is deeply rooted in consequentialism45, and advocates the reward of the effort through the market system. It ignores the issue of the overall distribution of resources, in spite of the importance that consumers may attach to the distribution of outcomes at a societal level. Richardson argues that the only case where N-C welfarist approach acknowledges to the existence of equity and social justice is when society might prefer a certain distribution of well-being (Richardson 2000). In such a case, in order to retain efficiency, these distributions must be achieved through lump sum redistribution of resources to increase the initial endowments of the less well-off individuals, thus ignoring the final distributions of outcomes after an allocation has taken place.
Evidence in the literature demonstrates that consumers care about outcomes at the social level, and not just at the individual level, and sometimes at a personal cost. (Konow 2003). The efficiency principle figures prominently in popular conceptions of a fair economic system. For example, at the micro level of the health system, support for the Pareto principles is sensitive to the size of benefits, and the results (Konow 2001). This suggests that efficiency in this sense is a type of justice and not necessarily at odds with social justice.
45
Consequentialism refers to those moral theories which hold that a morally right actions, rules, institutions, etc are to be evaluated in terms of their good outcomes or consequences, where the standard of goodness employed is a standard of non-moral goodness Culyer, A. (2005). The Dictionary of Health Economics. Cheltenham, Edward Elgar..
Page 91
Thus, accepting efficiency principle may mean violating some other form of social justice such as equality of outcomes (Olsen 1997).
4.3.2 Equality Principle The equality principle signifies a relationship between groups of persons that have similar qualities in at least one respect, such as equality of health outcomes. It may be expressed as ‘equality in outcomes’, ‘equality in opportunities’, ‘equality in access or use’, ‘equal human rights’, or ‘equal treatment for equal need’. It is the key principle in egalitarianism and in the Rawlsian theories of social justice (Konow 2003). The equality principle is treated as the central principle for social justice and Rawls stated that it is “a standard whereby the distributive aspects of the basic structure of society are to be assessed” (Rawls 1971). According to Rawls, the principles of justice are manifested as part of the social contract that is chosen by free and rational human beings who are behind the ‘veil of ignorance’ of their own places in society. Under Rawls’s theory, the equality principle may be expressed as equal rights, liberties, health and/or opportunities.
Egalitarianism focuses on the equality of outcomes and states that the fairest allocation of resources in the Australian health system is when benefits and costs from the allocation are distributed equally among all Australians (Gosepath 2007). This theory, however, ignores differences in effort, talent, and productivity when resources are being allocated. When all of society receives the same opportunities, then the rights of various groups within it, for example equality of men and women, can be realised (National Spiritual Assembly of the Bahais of Ecuador 1988). This is because equal outcomes are virtually impossible to achieve in practice due to differences in natural endowments, differences in people’s capacities to benefit, and differences in people’s willingness to participate because their tastes and preferences are not homogenous (Baha'i International Community 1998; Konow 2003). In practice, however, equality of opportunity should be pursued along with other social justice objectives. “*E+qual opportunity, rights and privileges for both sexes, advocates compulsory education, abolishes extremes of poverty and wealth, exalts work performed in the spirit of service to the rank of worship, recommends the adoption of an auxiliary
Page 92
international language, and provides the necessary agencies for the establishment and safeguarding of a permanent and universal peace" – (Shoghi Effendi 1934).
Thus, social justice policies should be judged by the equality of opportunities and/or equality of access accorded to the members of society. These policies should be pursued in conjunction with other social justice objectives such as human rights and the basic health needs principle.
4.3.3 Needs Principle The needs principle states that socially just or desirable allocations provide all the members of society with the basic minimum of services such as health care services, which are essential for the tolerable living of everyone in society (Raphael 1980; Konow 2003). In Indigenous health, this implies that more resources are needed to raise the health status of Indigenous Australians to a level comparable with non-Indigenous Australians. This would be consistent with the government’s stated effort in closing the gap between Indigenous and other Australians ,and consistent with higher expenditure on Indigenous health .
The needs principle features in the Rawl’s second principle of justice, called the Difference Principle, involves the ‘maximin’ rule for the distribution of goods and services equally, unless an unequal distribution of any, or all of these goods is to the advantage of the least favoured in society (Konow 2003). The minimum standard in Indigenous health could involve closing the gap in health outcomes between Indigenous and non-Indigenous Australians.
However, the needs principle is sometimes criticised because it does not recognise differences in productive contributions, or distinguish between real needs and manifested needs. For example, Rawls’ difference principle has been criticised by many authors such as Arrow (1973) and Harsanyi (1975) for failing to recognise that people may also seek to maximise other outcomes and not just the distribution of goods and services (van Doorslaer, Wagstaff et al. 1993; Konow 2003). Additionally, the needs principle does not provide an explanation for resource allocation beyond achieving the minimum standards required for human existence (Jan and Mooney 1997). Indeed the evidence in the literature suggests
Page 93
that although people care about need and show concern for the least disadvantaged, they also care about adverse effects from basing allocations solely on need, thus rejecting need as the sole foundation for a system of distribution (Konow 2003).
In economic evaluation, the needs principle is the dominant ethical principle considered to be the basis for assessing the concept of benefit from health interventions. In health care, "need" can be defined as health status; disease severity; or capacity to benefit from the existence of an effective intervention (Culyer and Wagstaff 1993; Mooney 1994; Olsen 1997; Hurley 2000). Defining need using the health status would reflect the size of the health problem. In this case, need is commonly defined and measured using standardised mortality rates (SMR). SMR is data collected from death rates, and adjusted for differences such as sex, age and location. A higher SMR indicates a greater need for the intervention to solve the cause of deaths reflected in these rates. The epidemiological assessments of the mortality and the incidence and prevalence of the health problem is used to assess need. Olsen argued that this approach is egalitarian because it emphasises an equitable distribution of health profiles (Olsen 1997).
However, the use of SMR fails to reflect the full extent of ‘need’ in Indigenous communities because it excludes the many debilitating illnesses that are not necessarily fatal but may be chronic. These include diabetes, and kidney, lung, and heart problems. Research on methods that incorporate standardised morbidity (illness) rates and combine them with SMR would improve the epidemiological approach for measuring need, as would further research into the value of interventions (McDermott, Plant et al. 1996). The main criticism of the health status approach to defining need is that it does not provide a link between the existence of the burden of illness and the effectiveness or value for money of the intervention, which is likely to generate inefficiencies.
When need is defined by the severity of the disease alone, rather than the existence of the disease, it tends to give priority to those who are severely affected by an illness over those less severely affected, a maximin principle under Rawlsian theory. In this case, those clinically worse off, would get priority treatment, irrespective of the forgone improvements to the less ill. Thus, people with the worst initial health level would be given priority in the allocation of resources in this model.
Page 94
The third form of need is defined by the existence of an effective intervention, often referred to as the ‘capacity to benefit’ (McDermott, Plant et al. 1996; Culyer 2001) by those targeted by the intervention approach (Jan and Mooney 1997; Hurley 2000). It measures the extent to which those people affected by a health intervention itself would benefit from that intervention, by focusing on health gain as opposed to pre or post-treatment profiles of the populations. The non-welfarist normative economic framework advocates this approach because it involves an epidemiological assessment of a health problem, and the existence of an effective intervention targeting such a problem (Olsen 1997). It is therefore a preferred concept of benefit for use in Indigenous health, since it has potential to improve the selection of interventions that work in Indigenous health. However, the capacity to benefit (effectiveness) is not well described for many interventions (Mooney and Houston 2004) and totally lacking for Indigenous-specific behavioural interventions, and so it is not widely used. It is therefore necessary for increased research to be undertaken in the ‘measurement of the notion of need’, in particular, in the measurement of ‘need’ in Indigenous health.
4.3.4 Equity Principle The equity principle explains the role of justice in social interactions that may be motivated by both self-interest and the desire to address perceived inequities. According to the principle of equity, a fair Indigenous economic system is one that distributes goods to Indigenous individuals in proportion to their efforts (Konow 2003). Effort typically comes in the form of productivity, ability or talent. More effort in an economic activity ought to be rewarded more, and vice versa.
A range of economists have offered several concepts to explain the notion of equity in health care and these have been summarised in Table 4.2 (Donaldson, Gerard et al. 2005). These explanations emerged from the work done in the United Kingdom’s National Health Service (NHS), and they all lead to the concept of ‘equal access for equal need’ as a guiding principle for resource allocation. The Australian government uses a dual policy instrument on need: equal access for equal need for all Australians with respect to mainstream health care services; and equal utilisation of health care services by all Australians for equal need. Equal access for equal need is implemented through the universal Medicare system. Equal
Page 95
utilisation for equal need is implemented through the Australian government’s provision of Indigenous-specific health care services, which in essence adopts utilisation as a measure of equity.
The debate about the notion of equity mirrors the wider debate within moral philosophy over what the appropriate criterion ought to be for fairness as a principle (Mooney 1998; Mooney 1998; Williams and Cookson 2000). Principles of equity in moral philosophy, for example, can be classified in terms of how the consequences that arise from the exercise of personal preferences are accounted for in the debate. For example, equity principles in the Australian health system can be defined as (Wolf 1989): 1.
Equality of opportunity;
2.
Equality of health outcome;
3.
Perfect equality of outcome, unless departure helps the Indigenous Australians;
4.
Treating equals equally – horizontal equity;
5.
Treating unequals unequally – vertical equity (allocation according to need); and
6.
Marxian equity – from each according to ability and to each according to need.
Table 4.2: Some philosophers’ and economists’ concept of equity Economist/Theory Musgrave (1969): Merit Good Theory
Concept Individuals will choose to consume less health care than they should. Government intervention aims to ensure appropriate consumption by increasing availability, and is thus concerned with equality of consumption. Lees (1962) The crux of the UK NHS was the provision of free health care at the point of use to ensure equal consumption of health care Titmuss (1970) Social duty means the UK NHS was concerned with providing the opportunity to promote health and altruism; thus the closest interpretation is equal opportunity to use health services (i.e. access) because this enables the promotion of both health and altruism. Sen (1977) Individuals are committed to the idea that others’ opportunity to consume health care may be just as, if not more, important than their own; thus the appropriate definition is equal access for equal need. Margolis (1982) Individuals are interested to do their fair share for the community, which leads them to want to contribute to making services available; thus the appropriate definition is equal access for equal need Some Philosophers’ Concept of Social Justice for Priority Setting Efficiency Misplaced theory of social justice. Maximising the greatest happiness for Utilitarianism the greatest number, but ignores distributional issues. Need A basic set of primary social Rawlsian theory goods are distributed (behind Equity the ‘veil of ignorance’) so that the position of the least well off in society is Rawlsian (Rawls, maximised. Health care is not one of these goods but may be seen as a 1971) primary institution to achieve fair distribution of basic liberties and so on (Daniels, 1981). Even so, it is not clear if equity is about maximising consumption of primary social good of health care. Therefore, it is possible
Page 96
Equality (Egalitarianism)
that equality of utilisation is desirable. Equal shares of a distribution of a commodity. Could be taken to mean equality of health or equality of health care.
Source: (Donaldson, Gerard et al. 2005)
Two approaches are being used by the Australian government to translate equity principles into practice: horizontal equity and vertical equity (Mooney 1982). Horizontal equity assumes that the groups or individuals being addressed have equal economic and social standards, or their health needs are the same. When the term ‘equity’ is used in the Australian health policy context, it often means horizontal equity (McDermott and Beaver 1996; Jan and Mooney 1998). This notion is clearly embedded in the Australian Medicare principle of ‘universal access’ to health care by all Australian citizens. The application of horizontal equity is more suited, to a limited extent46, to the allocation of resources between Indigenous Australian communities themselves, but not between Indigenous and non-Indigenous Australians. Horizontal equity criterion between Indigenous Australian communities vis-à-vis the rest of the country, would be inequitable and unfair.
Vertical equity, on the other hand, refers to the notion of unequal but equitable treatment (Mooney 1982). It implies that, in order to be fair to two ‘unequal’ groups, which have different resource endowments or an unequal number of problems, it is fair to treat them unequally or differently. This would be positive discrimination of unequal groups (Mooney 1999), as is the case with Indigenous Australians who, compared with other Australians, have an unequal number of health problems and poor health indicators such as higher mortality and morbidity rates (ABS and AIHW 1999). Vertical equity implies that any significant improvement in Indigenous Australian health problems can only occur if more resources are given over to Indigenous health intervention programs than to nonIndigenous programs. It is suitable for allocation between Indigenous and non-Indigenous Australians.
46
Health problems between Indigenous communities themselves are highly disparate, such as between rural and remote communities in central Australia and an urban community in Melbourne. It would not seem appropriate to assume that these two groups have the same health needs.
Page 97
In conclusion, the equity principle contrasts with the equality principle, in that the equality principle seeks to satisfy all needs regardless of effort and SES. In whatever way it is defined, the equity principle has been a dominant reason for government intervention in the health sector to reduce inequalities in population health, achieve a fair and just process of health maximisation or both (van Doorslaer, Wagstaff et al. 1993). The priority setting approach, which focuses only on health outcomes equalisation in the policy context (as in welfarism), may emphasise distributive justice at the expense of procedural justice issues (Wallis and Dollery 1999).
4.4 Procedural Justice Issues in Priority Setting Procedural justice in Indigenous health could be viewed as: the level of Indigenous involvement in the priority setting process; or the extent to which Indigenous Australians exercise autonomy in determining how resources are spent, which is an objective in its own right (Mooney, Jan et al. 2002; Konow 2003). The extent to which Indigenous Australians exercise autonomy in determining how resources are allocated probably reflects the N-C welfarist view of the autonomy of the consumer. It is also consistent with the notion of selfdetermination or community control, as advocated by Indigenous Australians and implemented through ACCHS. This concept of autonomy in the input to resource allocation is also justified in the sense that it ensures that the community is appropriately informed of the potential health consequences of its choices (Jan and Mooney 1997). This is reflected in following extract (Coombs 1994): “Indigenous decisions about their preferred lifestyles are, in part, based on the Indigenous ethic of accountability to others. Social responsibility is required by their commitment that autonomy, at personal and a group level, will be exercised to as to ensure that what is done contributes to the care and nurture of others with whom they are related; so that personal behaviour remains socially grounded … This sense of mutual obligation is not due to an external agent, like that sought by the white authority, but the group as a whole.”
This indicates that Indigenous involvement in the decision-making process comes in different forms. It could be in the form of Indigenous representation in a meaningful
Page 98
manner to address Indigenous Australians’ access to mainstream services. This could also be in the form of the Indigenous community liaison workers at mainstream health care services who have a tangible influence in the priority setting process, and who are involved at all stages of that process. Current government approaches to Indigenous Australians’ involvement in the priority setting process include ministerial committee appointments, and other invitations to serve in advisory roles. At the State level, Victoria has a policy encouraging mainstream health care services to employ Indigenous liaison officers, whose roles are to communicate with the community and the health care service.
4.5 Key Issues for Consideration in Priority Setting Principles of social justice which are suitable for priority setting include equality, equity, need and procedural justice. Health resources are not infinite, and there is evidence in the literature to support the role of these principles in priority setting. The challenge is to translate these principles into practice in Indigenous health. These principles can be identified and measured either as health objectives/criteria, in the case of need and equity principles, and/or as a means to an end, in the case of procedural justice or due process in priority setting.
The current Australian government policy objectives of the Australian health care system, with respect to Indigenous health, may be interpreted as aiming to achieve ‘equality of health status’ or minimisation of health inequalities, a form of the distributive principle. Equalisation of a health status objective can be translated into the ‘equalisation of health’ principle, and is not at odds with the concept of vertical equity. To achieve the ‘equalisation of health’ objective in Indigenous health, two perspectives need to be addressed: allocations between Indigenous communities themselves; and those between Indigenous and non-Indigenous Australians.
When allocations are between remote and urban Indigenous communities or between Indigenous communities and non-Indigenous Australians, the appropriate objective should be to pursue vertical equity. Otherwise, pursuing horizontal equity would be the key
Page 99
objective when the allocation is between Indigenous communities with similar socioeconomic conditions, ceteris paribus.
The social justice principles discussed in bioethics involves access to health care services, especially for low-income households such as those found in Indigenous communities. The discussion often involves questions such as whether society is willing to bear health care costs for low-income families and the less fortunate members of society. Programs such as national health insurance, which guarantee universal coverage, have been found to be consistent with prevailing population ethical views in nearly all developed countries except in the United States, where the support for national health insurance was very low (van Doorslaer, Wagstaff et al. 1993). Additionally, because good health is a vehicle for economic growth and individuals’ full potential to express their capabilities, improving the health of citizens would potentially improve economic growth.
The Australian government’s universal health insurance policy, Medicare, embraces the idea of publicly funding health care services. This is in recognition of the socio-economic disadvantage that some Australians may be facing. Implicitly, it assumes that all Australians have equal access to the MBS and the PBS through their GPs and other medical services, and the pharmaceutical services respectively. However, Indigenous Australians utilise less health care services than other Australians (see Chapter 2). To improve Indigenous access to health care services, the government funds Indigenous-specific health care services. Equalisation of health as an objective could also be viewed as a human right, which may include an “equal ‘right’ to health care” (Thurow 1977; Anderson 2008).
Indigenous health policy objectives should be informed by the Indigenous values and aspirations identified through empirical research. These objectives should be subjected to ethical debate so as to reflect the views of Indigenous Australians (Mooney 1998; Mooney 1998; Richardson 2000). Two approaches to eliciting public views on health objectives are preferred: empirical ethics and communitarianism.
For priority setting purposes, Richardson argues that social ethics ought to guide the government health policy objectives, and these ethical issues should be captured using the ‘empirical ethics’ approach (Richardson, 2000; Richardson, 2001). This would involve ethical
Page 100
values being subjectively derived from a sample of Indigenous individuals within the population and these values being subjected to ethical analysis and criticism. The sample could be made up of randomly selected members of the Indigenous community and their subjective values elicited using an appropriate instrument. These values can then be quantified and converted into a scale that can be used to weight the importance of QALYs or some other chosen measure of health (Richardson 2000; 2001). In the event that the social value is shown to lead to abhorrent outcomes, such as racism, sexism or discrimination, then preferences should be ‘laundered’, and an iterative process undertaken between the ethicist and the public.
The communitarianism approach seeks to attain some degree of coherence and transparency in the assessment of equity from an economics perspective. This approach is built on the concept by Broome47 (1989), and it classifies reasons for a person’s claim for a good into two groups: ‘communitarian claims’48, and ‘other reasons’ (Mooney et al, 1998). To understand individuals in society, especially when undertaking priority setting, one must look at the community from where the individual comes and their communal relationships (Avineri and de-Shalit 1992). The communitarians therefore argue that equity is consistent with social/community perspectives on how individuals/or communities should be treated relative to one another. This is because the sense of the community and community competencies are properties of the community, and it is therefore questionable when benefits of a community action are equated to the sum of the effects on individuals (Shiell and Hawe 1996; Mooney 1998; Shiell and Carter 1998). Thus, equity under communitarianism is defined by the way in which claims are established, and how the different claims can be weighted. The key challenge is how to subjectively elicit communitarian preferences.
Subjective values (pleasure, utility or health) of Indigenous Australians or their representatives would be elicited using the empirical approaches, since subjective valuations are widely accepted and well established (Konow 2003). The cardinal measurement of the elicited values would be undertaken to specify the intensity of
47 48
John Broome (1989) in Mooney, Jan & Wiseman (1998). ‘Claims’ are defined as reasons backed by a notion of duty for one’s claim for a good.
Page 101
preferences, assumed to be the same finite end points, 0-1, on the cardinal health scale (see Mooney and Olsen, 1991; Wagstaff, 1991). The cardinal measurement would also allow for the elicited preferences to be interpersonally compared to avoid the Arrow impossibility theorem (Olsen 1997; Mooney 1998). Under these assumptions, the preferred distribution of health care may take the form of the ‘maximised’ sum of individual health or community benefits.
4.6 Summary The Australian government adopts a mix of the libertarian and egalitarian ideologies in priority setting. These ideologies influence the ethical values that guide priority setting. For Indigenous health, the government has acknowledged and pledged to reduce the health inequalities between mainstream and Indigenous Australians. To do so requires a normative framework and health care policies that have regard for both efficiency and social justice. This is because efficiency policies may not necessarily be just or equitable to minority and marginalised groups since these groups lack the social power to lobby for their share of health resources, and to gain access to them (Jan and Mooney 1997).
Distributive principles have been the dominant reasons for government intervention in the health sector as a way to achieve the reduction of inequalities in population health (van Doorslaer, Wagstaff et al. 1993). In Indigenous health, allocation through the mainstream health system could best be addressed using vertical equity, and allocations within Indigenous population groups could best be achieved through horizontal equity processes. In procedural justice, the focus is on the fairness of the processes by which priorities are set, and how decisions are taken.
In practice, social justice principles overlap greatly and so categorising them as distributive or procedural is a challenge in itself. The choice of one form of justice, such as equality of access, can mean sacrificing another form of justice, such as unequal treatment of unequals. The presumption of equality provides an elegant procedure for constructing a theory of distributive justice using the following questions, which can be used as a contribution from social justice, for the development of a checklist for priority setting (Gosepath 2007):
Page 102
Which social goods comprise the object of distributive justice?
What goods and burdens are to be justly distributed or should be distributed?
What are the spheres of justice into which these resources have to be grouped?
Who are the recipients of the distribution? Who has a prima facie claim to a fair share?
What are the commonly cited, yet in reality unjustified, exceptions to equal distribution?
Which approach, conception or theory of egalitarian distributive justice is therefore the best?
Do the costs of the exercise bear undue pressure on the survival and protection of Indigenous values and culture?
In summary, social justice issues, which ought to inform priority setting, include equity (both vertical and/or horizontal equity); need (in whatever form it is defined); fairness (in whatever form it ought to be defined, either as distribution of outcomes, autonomy or community control); and/or access (both physical and perceived forms). The use of explicit criteria in priority setting provides a link between ethical principles and health care resource allocation. This link would be manifested in the form of the health objectives and/or the due process of achieving those objectives because such a process ought to yield socially just/desirable outcomes. Incorporating ethical principles can potentially inform the efficiency principle inherent in cost-effectiveness analysis. Need would be reflected in the selection of interventions/options for change, and then utilised in the assessment of the effectiveness of the interventions, to inform economic evaluation. The equity principle could also be treated as an objective in itself, and could potentially be reflected as a weighting that can be attached to the concept of benefit. This issue will be investigated in Chapter 8 in detail.
Page 103
5. Chapter 5: Perceptions of Decision-makers Some men and women glory in their exalted thoughts, but if these thoughts never reach the plane of action they remain useless: the power of thought is dependent on its manifestation in deeds (Abdu'l Baha 1912).
5.1 Introduction In this chapter, the results from the survey of decision-makers are discussed to establish the criteria used for making choices, given their organisational objectives such as access to services and maximising the health of the population. Two categories of issues to inform the development of the priority setting guidelines in Indigenous health are discussed:
Pragmatic issues from the literature; and
User considerations from the survey.
In order to do so, the following key objectives will be undertaken in the survey:
Identifying the criteria being used to set priorities in Indigenous-specific health care services;
Assessing the current priority setting process to highlight the level of uptake of economic evaluation evidence; and
Identifying how the priority setting process can be improved from the decisionmakers’ perspective.
5.2 Methods The study was a descriptive cross-sectional survey of the key decision-makers involved in Indigenous-specific health care services to elicit their views and values on a range of issues associated with priority setting. The instrument was adopted from the work done by Mitton
Page 104
and colleagues (Mitton 2001; Mitton and Donaldson 2002; Mitton and Donaldson 2003; Mitton and Prout 2004; Teng, Mitton et al. 2007).
The survey was a written paper questionnaire which had both open-ended and closedended questions. It consisted of an interviewer-administered questionnaire presented at formal face-to-face semi-structured interviews. Respondents were required to tick the relevant responses from the list of the possible responses. Where a response was not listed, respondents were required to write it down under ‘other’. The glossary of the key terms used in the survey was included in the plain language statement which accompanied the interview questions. At the time of the interview, respondents were reminded to ask for explanations in cases where the questions were unclear.
The interview questions were piloted on staff and students in the School of Population Health, The University of Melbourne, before the data collection started, in order to finetune the questions and the interviewing process. Ethics approval was obtained from the University of Melbourne before the interviews took place.
5.2.1 The Sample Two groups of key decision-makers were interviewed: those involved in Indigenous-specific health care services in the Victorian DHS; and the CEOs and/or their representatives in ACCHS. Only these two groups were chosen to be interviewed for financial and logistical reasons. The rest of the funding bodies and providers of Indigenous health care services such as hospitals, GPs, community pharmacists, community health services and ACCHS workers, could not be interviewed during the available time frame for this thesis.
The DHS sample was primarily involved in the funding of the services for both Indigenous and non-Indigenous Australians in Victoria. They are also responsible for the budget submissions to the State government ERC for funding of the services under their jurisdictions. Once the funding allocations have been approved, each portfolio or unit funds its own activities in Indigenous health, often in a prescriptive manner. Making decisions across programs rarely happens. ACCHS get funding from the DHS, the OATSIH and from NGOs. Resource shifts between programs are very restricted, except in a few areas such as
Page 105
funding from ‘Home and Community Access’ program and the Medicare funding for GP services. These are the main areas where the organisations have liberty to move resources across programs, implying that certain priority setting criteria may be applicable to the DHS but not to ACCHS.
ACCHS are comprehensive primary health care service providers whose focus is on Indigenous community members. Equity is not an issue since the communities they may be responsible for could be considered fairly homogenous. Efficiency and impact of the services on the community members are priority issues for ACCHS management. They would also be concerned with mechanisms of making the funds go the furthest, since the communities under their jurisdictions have multiple needs, and the funders demand extensive accountability.
To identify and select the potential participants in this study, a ‘snowballing’ sampling process was used to identify all the key informants. Snowballing is a tool for selecting participants from extended associations, through previous acquaintances, to find people involved in Aboriginal-specific health care services (Goodman 1961; Berg 1988). The key reason for choosing the snowball sampling involved its ability to allow us to identify other members within the DHS and ACCHS involved in Aboriginal-specific health care services who could be interviewed.
For the two groups of respondents in this study, the initial contacts were the Director of Public Health in the DHS and the CEO of Victorian Aboriginal Controlled Health Organisations (VACCHO). After the interviews, they identified their colleagues whom we then approached. Those respondents were later, in turn, asked for other potential interviewees. This ensured that potential participants were identified and the identified participants felt encouraged to participate in the study.
Overall, 39 potential participants from the DHS were approached for the interviews, and 26 of these participated (67% participation rate). For ACCHS sample, the participation rate was about 65% (24/37). The participation rates were not that great, which implies that generalising these results has to be done with caution. This partly because those identified
Page 106
using the snow-balling process revealed that they were not directly involved in priority setting. However, the purpose of the survey was to demonstrate how such a survey can be done and generate indicative results for further research, and possibly for nation survey of decision makers in Aboriginal-specific health care services.
58% of the DHS respondents were females, and 50% of ACCHS were females. Most of the DHS
participants
interviewed
were
senior
members
of
their
respective
organisations/departments and 46% of them had served for more than two years in their respective positions. For ACCHS respondents, nearly 50% had served for more than two years in their respective positions. In terms of exposure to health economics, more than 80% of the DHS respondents had been exposed to health economics, either through work or university courses, and others were aware of the presence of health economists. 58% of all ACCHS respondents were of Indigenous or Torres Strait Islander descent, and half of them had been exposed to health economics either through work or university courses, and others were aware of the presence of health economists (Table 5.2). The two groups of respondents were interviewed using the same interview instrument, and reported as two separate samples.
Table 5.1: Background information expressed as both numbers and percentages
DHS ACCHS
Indigenous
Gender
Yes = (0) 0% No = (26) 100% Yes = (14) 58% No = (10) 42%
F = (15) 58% M = (11) 42% F = (12) 50% M = (12) 50%
Exposure to health economics Yes = (21) 81% No = (5) 19% Yes = (12) 50% No = (12) 50%
Time (yrs) in current position > 1 yr = (17) 65% > 2 yr = (12) 46% > 1yr = (16) 67% > 2yr = (12) 50%
Health experience Yes = (16) 62% No = (10) 38% Yes = (13) 54% No = (11) 46%
5.2.2 Data Collection Instrument Interview questions sought to elicit the interviewees’ personal understanding of the issues related to priority setting in their organisations, and not the priority setting issues outlined in the official documents such as the annual reports or business plans. The administered questionnaire was divided into four broad themes: the current process of priority setting; an assessment of the current process; the way in which the current priority setting process can be improved; and the potential for PBMA (Mitton 2001) as shown in Table 5.2.
Page 107
Table 5.2: Structure of the questionnaire for surveying the key decision-makers Part I: Current process of setting priorities and allocation of resources What do you consider to be the key objectives of this organisation? What sources of information are used in determining short-term and long-term priorities? On what basis or criteria are decisions made in resource allocation and program/service activity? How is the ‘benefit’ from programs/interventions viewed in your organisation? Part II: Assessment of the current priority setting process How does the process of priority setting perform? What are the strengths and challenges of the current process? Part III: Improving the process of priority setting How could the current process of setting priorities be improved? Do you think that with the current process there are challenges to using evidence from economic evaluation? How could we improve the uptake of economic evaluation evidence?
5.2.3 Data Collection The interviews took place between December 2007 and March 2008 (see Appendix 5 for the letters of invitation and data collection instruments). Interview questions were sent to the participants prior to the scheduled interview to ensure that the interviewees had given thought to the questions since some of the questions were abstract in nature. At the time of the interview, participants were asked if they consented to the interview being recorded for accuracy purposes. At the conclusion of the interview, participants were advised that all the transcripts would be sent to them for review and were also asked if they could be quoted by name and/or organisation in this thesis and subsequent publications.
The face-to-face approach was chosen because it allowed clarification of the concepts in priority setting and the discussion with the respondents on an individual basis to enhance accurate elicitation of the views. Some interviewees filled the questionnaires and others preferred the interviewer to do so on their behalf. When the respondents filled out the questionnaire themselves, notes on responses to the interview questions (closed and openended questions) were taken down during the interviews. At the same time, the responses were audio-taped and then transcribed later to ensure the accuracy of the information. The interview transcripts were later sent back to the respondents for confirmation of their responses.
Page 108
5.2.4 Data Analysis Descriptive statistics (quantitative analysis) was used to analyse responses from closedended questions and the results were then reported separately for each sample. Qualitative data from open-ended questions were analysed quantitatively using content analysis, and supplemented with explanatory quotes. Content analysis can be defined as "a research technique for the objective, systematic, and quantitative description of manifest content of communications" (Barelson 1952). In general, this method summarises content from the interviews by counting various responses which had been coded. Responses may get coded under major themes (e.g. data, reasons; etc) and then under the sub-themes such as ‘data accessibility’ and ‘data validity’ (Busha and Harter 1980; Miles and Huberman 1984). The criteria for coding and subsequent reporting of themes included the number of respondents made specific comments and the relevance of such responses to the particular questions.
The final score for each sample involved the number of times (n) a particular response appeared or the number of respondents who gave a particular response. In some cases, these numbers were expressed as the percentage of the total number of respondents (%) who gave a particular response. Using content analysis in this study allowed qualitative data to be analysed quantitatively. Thus, the results from the survey (both open-ended and close-ended responses) were reported as numbers and percentages of respondents.
5.3 Findings The results from Parts I-III of the questionnaire (Table 5.3) are presented in the next section. First, results from the survey of DHS respondents are reported, and then the results from ACCHS respondents follow.
Page 109
5.3.1 Department of Human Services 5.3.1.1 Current process of setting priorities and allocation of resources
The DHS respondents identified effectiveness, equity, sustainability, acceptability and access to services as the top five key objectives in Indigenous-specific health care services (see Table 5.4). Based on the number of respondents who mentioned efficiency as a key objective, efficiency was ranked the sixth objective with affordability ranked in the bottom of the list.
Table 5.3: Department of Human Services objectives Objectives Effectiveness Equity Sustainability Acceptability Access to services Efficiency Impact Feasibility Affordability
Responses 20 18 18 15 13 11 11 8 7
Percentage 77% 69% 69% 58% 50% 42% 42% 31% 27%
The results also revealed that the main sources of data for short-term priority setting were very similar to the sources of data for long-term planning (Table 5.5). These data were from State/Federal policy documents, evidence, Indigenous communities, and the key organisation objectives, and funds available/financial directives. These results were consistent with the WA study, which had found that decision-makers used evidence in the form of epidemiological data and State/Federal policy in setting priorities (Mitton and Prout 2004). However, differences between WA and Victoria occurred in the use of regional/State reports, medical opinion and needs assessment as the main sources of data; the Victorian survey indicated that these sources of data were not commonly used.
Page 110
Table 5.4: Data sources for priority setting in the Department of Human Services Short term Sources of data Evidence: economic evaluation, epidemiology, State/Federal policy documents
Responses (16) 62%
Long term Sources of data State/Federal policy documents
Responses (19) 73%
(15) 58%
(19) 73%
Indigenous community Key organisation objectives Funds available/financial directive Needs assessment Mainstream public opinion
(13) 50% (12) 46% (10) 38%
Evidence from economic evaluation, epidemiology, etc Key organisation objectives Indigenous community Needs assessment
(8) 31% (7) 27%
Funds available/financial directive Three-year business plan
(8) 31% (6) 23%
(17) 65% (14) 54% (10) 38%
The top five criteria for priority setting across and within programs49 identified by the DHS respondents included the size of the burden, feasibility/sustainability, equity and acceptability as important criteria (see Table 5.6). In fact, the Director for Public Health said: “There are no criteria at all (currently) for dividing resources across programs at the DHS … for me the most important issues would be the equity … and the size of the health burden would be an issue but not the overriding factor … to guide the priority setting process.”
The importance of feasibility/sustainability could be seen to be consistent with the need for programs that would be easy to implement and continue to be delivered to the Indigenous Australians. The use of the size of the health burden and equity as criteria for resource allocation across programs reflects the importance of egalitarian principles in Victoria.
Table 5.5: Criteria for priority setting Across programs Size of the health problem
Respondents (18) 69%
Within programs Feasibility/sustainability
Respondents (14) 54%
Feasibility/sustainability Equity Political ‘hot spots’ Acceptability Access to services Historical trends/patterns Efficiency
(13) 50% (11) 42% (11) 42% (10) 38% (10) 38% (9) 35% (3) 12%
Size of the health problem Equity Acceptability Efficiency Access to services Historical trends/patterns Political ‘hot spots’
(13) 50% (12) 46% (10) 38% (7) 27% (6) 23% (5) 19% (4) 15%
49
A program here may mean a collection of interventions such as diabetes, maternal health or HIV/AIDS services. Within a program may mean resources being moved from cervical to breast cancer, say, within women’s health.
Page 111
There was a unanimous agreement among those surveyed from the DHS that benefit from health care services to Indigenous Australians should not only include individual medical benefit but also include non-individual and non-medical benefits such as community benefit, equity and cultural security (see Table 5.7). These results provided some support and useful information on the concept of benefit study, reported in Chapter 8.
Table 5.6: Concept of benefit from health care programs Criteria within programs Community benefit Individual benefits Cultural security Equity
Respondents (21) 81% (19) 73% (10) 38% (9) 35%
5.3.1.2 Assessment of the Priority Setting Process
When the DHS respondents were asked to assess the way resources were being allocated (across and within programs areas – such as drugs and alcohol, child health or physiotherapy) in their departments/units, they reported that their current process did not work well. Only 27% reported that the process worked well, with 15% being unsure. For example, the Director of Public Health again said: “The priority setting process is quite patchy … I think in some cases it is quite good and in some cases, it leaves a lot to be desired..”
On assessment of the priority setting process, about 20% of the DHS respondents reported efficiency as one of the key strengths of the process across programs (see Table 5.8). This information strengthened the case for the DHS to list efficiency as a key objective for the department in order to raise the profile of health economic evaluation evidence. The DHS respondents thought that the process was transparent, in that it involved the stakeholders and that it was widely accepted in the Indigenous community. However, the DHS respondents reported that the process was not systematic and not explicit.
Page 112
Table 5.7: Assessing the priority setting process Across programs Involves stakeholders in the process
Responses (13) 50%
Responses (9) 35%
(8) 31%
Within programs Involves stakeholders in the priority setting process Widely accepted/supported in the Indigenous community widely accepted/supported in the organisation Equitable
Widely accepted/supported in the Indigenous community Transparent
(9) 35%
Widely accepted/supported in the organisation Based on evidence Equitable Efficient
(8) 31% (6) 23% (5) 19%
Based on evidence Transparent Efficient
(7) 27% (6) 23% (5) 19%
(8) 31%
(7) 27% (7) 27% (7) 27%
In terms of the current process (see Table 5.9), the DHS respondents reported that the key challenges were lack of good quality data (58% of the respondents mentioned it) and lack of expertise (54%). These issues would clearly have an impact on the role of economic evidence in priority setting in the DHS, given that there is lack of economic evaluation evidence in about Indigenous health for priority setting, and not many economists are employed in Indigenous health. Other challenges included lack of resources and lack of public or stakeholder involvement.
Table 5.8: Challenges of priority setting process in Indigenous health Challenges Lack of good quality data Lack of expertise Crisis oriented Lack of resources Not systematic No/little public/stakeholder involvement Implicit decision-making Lack of time
Responses (15) 58% (14) 54% (7) 27% (7) 27% (6) 23% (5) 19% (5) 19% (5) 19%
5.3.1.3 Improving the Priority Setting Process
In order to identify how the priority setting process could be improved, respondents were asked to rank their choices by relative importance. The data from the DHS respondents revealed that about 81% of the respondents indicated that better data and strength of evidence was needed to improve the priority setting process. They also indicated that having a longer-term view (54%) when the resources are being allocated would improve the
Page 113
priority setting process (see Table 5.10). Others identified the need to have a mechanism that can be used to systematically shift resources across programs (35%), while being as transparent as possible, with less political influence, to ensure explicit priority setting.
On the role of economic evaluation evidence in priority setting, one respondent understood economic evidence being one of the many issues needed for priority setting. He said: “There are two issues in the use of economic evaluation evidence. *First+, economic evaluation is not an entire evaluation. There may be other reasons apart from economic reasons why things may be done especially in terms of efficiency. There are reasons for using expensive interventions. Secondly, economic evidence is not the only form of evidence to use in decision-making.”
Table 5.9: Improving the priority setting process Improving the priority setting process Better data/strength of evidence Longer term view
Responses
Mechanism for shifting resources across programs required Systematic More explicit/transparent
(9) 35%
Examining the margin
(2) 8%
Not total need and less political influence
(2) 8%
(21) 81% (14) 54%
(8) 31% (7) 27%
Improving the uptake of economic evaluation Up-skilling of staff Better data from commissioned studies Longer term view taken
Responses
Better communication strategy More education on the credibility of economic evaluation Timely data provision from decision-makers Transparent economic models
(10) 38% (10) 38%
(15) 58% (14) 54% (10) 38%
(10) 38% (9) 35%
The DHS respondents indicated that upskilling of staff in health economics and/or more training of health economists (58%) would improve the uptake of economic evaluation evidence (see Table 5.10). Good quality and timely data from commissioned studies (54%) was also much needed to improve the uptake of economic evaluation evidence. One respondent lamented that: “The trouble with academics is that they take too long to get results to us. When you ask them for some information, they tell you that results for an economic evaluation will take two years to be produced. By then priorities will have changed.”
Page 114
5.3.2 Aboriginal Community Controlled Health Services 5.3.2.1 Current process of setting priorities and allocation of resources
Results from ACCHS respondents identified impact, sustainability, access to services, acceptability, effectiveness, efficiency, feasibility, affordability and equity as the key objectives in their community organisations (see Table 5.11).
Table 5.10: ACCHS objectives Objectives Impact Sustainability Access to services Acceptability Effectiveness Efficiency Feasibility Affordability Equity
Responses 19 18 16 15 14 12 9 9 7
Percentage 79% 75% 67% 63% 58% 50% 38% 38% 29%
In order to achieve the objectives identified above, the main sources of both short-term and long-term data for priority setting reported by most of the respondents were the Indigenous community and needs assessment (see Table 5.12). Other sources of short-term data included mainstream public opinion, key organisation objectives, and funds available or financial directives. In general, short-term data sources, in terms of rankings and percentages of respondents who mentioned these sources, were very similar to the longterm data sources, except for mainstream public opinion, which was not considered the main source of long-term data and was only mentioned by two respondents.
Table 5.11: Data sources for priority setting Short-term sources of data Indigenous community
Responses (17) 71%
Long-term sources of data Indigenous community
Responses (16) 67%
Needs assessment Mainstream public opinion Key organisation objectives Funds/financial directives Evidence State/Federal policy documents Medical opinion
(13) 54% (12) 50% (12) 50% (12) 50% (9) 38% (9) 38% (8)33%
Needs assessment State/Federal policy documents Funds/financial directives The three-year business plan Key organisation objectives Medical opinion Evidence
(13) 54% (12) 50% (11) 46% (11) 46% (10) 42% (9) 38% (6) 25%
Page 115
The size of the health burden and the acceptability of the priority setting process were the key criteria for priority setting both across and within the programs (see Table 5.13). Other top-five criteria included historical trends/patterns, feasibility and efficiency. The importance of feasibility/sustainability could be seen to be consistent with the need for programs/services that can be implemented and can continue to be delivered to the Indigenous Australians.
Table 5.12: Criteria for priority setting across and within programs Criteria across programs Size of the health problem Acceptability Efficiency Historical trends/patterns Feasibility/sustainability Access to services Top-down decision-making Whoever yells loudest
Responses (15) 63% (13) 54% (11) 46% (10) 42% (10) 42% (8) 33% (5) 21% (5) 21%
Criteria within programs Size of the health problem Acceptability Historical trends/patterns Feasibility/sustainability Efficiency Access to services Equity Whoever yells loudest
Responses (17) 71% (13) 54% (12) 50% (12) 50% (10) 42% (10) 42% (4) 17% (3) 13%
5.3.2.2 Assessment of the Priority Setting Process
When the respondents were asked to assess the priority setting process, most of them felt that the process in their organisations worked well for them. Over half (58%) of the respondents reported that the priority setting process worked well with 8% who were unsure (see Table 5.14). One respondent from ACCHS commented that: “Many things work in the current priority setting process in our organisation but much is uncaptured largely due to the restrictive nature of the reporting requirements of the funders. Of course we know that there is always room for improvement.”
ACCHS felt that the priority setting process was efficient across programs, but not within programs. One ACCHS respondent explained that: “Yes, I would be happy to know what reasons the funders use to select programs/services to fund. Most importantly, I would like to know the reasons why they have not funded an activity. Most of the times the funding body does not
Page 116
provide reasons for refusing to fund an activity. It is not helpful for us and very frustrating.”
Table 5.13: Strengths of the current priority setting process Strengths (across programs) Widely accepted/supported in the Indigenous community Involves stakeholders in the priority setting process Transparent Based on evidence
Responses (18) 75% (17) 71% (9) 38% (7) 29%
Strengths (within programs) Widely accepted/supported in the Indigenous community Involves stakeholders in priority setting process Efficient Transparent
Responses (13) 54% (12) 50% (11) 46% (9) 38%
ACCHS respondents’ reported lack of resources (79%) and the tendency of the priority setting process to be crisis oriented (63%) as the key challenges for them (see Table 5.15). These two issues were quite different from those faced by the DHS. Other issues that ACCHS respondents raised included a lack of expertise, lack of good quality data, and lack of time. In addition, they identified an overreliance on providers, pervasion of individual interests, political influence being too great, and implicit and unsystematic decision-making as challenges.
Table 5.14: Challenges (across and within programs) Challenges Lack of resources Crisis oriented Lack of expertise Lack of good quality data Over reliance on providers Not systematic Lack of time Individual interests pervade
Responses (19) 79% (15) 63% (12) 50% (10) 42% (7) 29% (6) 25% (6) 25% (5) 21%
5.3.2.3 Improving the Priority Setting Process
In order to assess the way that the priority setting process could be improved, respondents were asked to rank their choices. ACCHS ranked better data/strength of evidence (63%) and longer-term view (63%) as the key issues to be addressed in order to improve priority setting (see Table 5.16). The next issue which needed to be addressed was the need to be systematic (38%) in process. This information vindicates the need for an explicit priority
Page 117
setting process which is backed by technical rigour especially in the form of economic evaluation. One ACCHS interviewee responded that: “It would be good to see a College of ACCHS CEOs established. Such a college would be responsible for training CEOs and/or accrediting them. This would ensure that the CEO is equipped with the skills and training to management skills. This would be most useful to the new CEOs who may have not previously worked in Indigenous communities or existing CEOs who may just need the skills to run their organisation.”
Table 5.15: Improving the priority setting process Improving the priority setting process Better data of evidence Longer-term view Systematic More explicit/transparent Mechanism for shifting resources across programs Less political influence
Responses
Uptake of economic evaluation
Responses
(15) 63% (15) 63% (9) 38% (8) 33% (4) 17%
Up-skilling of staff Data from commissioned studies Longer-term view taken Better communication strategy Education on the credibility of economic evaluation Timely data to decision-makers
(12) 50% (10) 42% (9) 38% (8) 33% (7) 29%
(3) 13%
(7) 29%
In particular, the need to upskill staff in health economics and/or more training of health economists (50%) featured significantly from ACCHS respondents as a means to increase the use of economic evaluation evidence: “Short course or training would enable us to use some of this economic stuff or undertake economic evaluation ourselves.”
The longer-term view taken (42%) in priority setting would mean that economic evaluation evidence would be needed in ACCHS. With regards to availability of in-house data for priority setting, one respondent from an ACCHS argued that the funding agencies ought to recognise that ACCHS have the data on their records. One respondent remarked: “It is just a matter of someone with expertise coming to comb through our records or collect the data through our organisation, rather than borrowing inappropriate data from elsewhere to use for priority setting.”
Page 118
5.4 Discussion The primary aims of this chapter were to identify some of the criteria decision-makers often use for setting priorities in health, and to elicit the preferences of the decision-makers that would be used to inform the development of the guidelines for priority setting in Indigenous health. These guidelines will contribute to the making of the priority setting process explicit. Making the priority setting process transparent will encourage debate on the appropriateness of the identified priority setting criteria, and improve accountability of the decision-makers to Indigenous Australians.
The literature indicated that there is a need for: explicit criteria for decision-making; employment of technical methods; and an examination of the role of economic evaluation evidence. Thus, an explicit priority setting approach in Indigenous health ought to embody both technical rigour and due process to ensure that more efficient, equitable and democratic decisions are implemented in the Indigenous health care sector.
The survey of decision-makers in Victoria served as a means to “listen to bureaucrats” (Mooney and Wiseman 1999) in order to understand what criteria the decision-makers use for priority setting. It was also important to gauge the level of the uptake of economic evaluation by the decision-makers, and the kind of barriers to the use of economic evaluation evidence, in order to understand what can be done to increase the uptake of economic evaluation evidence in decision-making.
5.4.1 Criteria for Priority Setting DHS respondents in the study were commonly involved in the funding of health care services. The priority setting perspective often taken by the DHS staff is that of the whole population, raising issues of the distribution of the resources and outcomes. The interest of Indigenous organisations in Victoria, on the other hand, lies in the provision of health care services that have an impact on the lives and families of Indigenous Australians. To ACCHS, what happens in mainstream health is important for them, mainly for reasons of access to those services. Their primary responsibility does not include equity considerations between Indigenous and non-Indigenous Australians. For these reasons, priority setting criteria and
Page 119
needs may differ somewhat. The common denominator for both groups of respondents is the improvement of the health of the Indigenous Australians.
Effectiveness, sustainability, access and efficiency were among the six key objectives common to both samples. Differences were noticeable on the ranking of effectiveness, equity and impact by the DHS or ACCHS alone. Effectiveness, which was at the bottom of ACCHS objectives, was ranked on the top of the DHS objectives. Impact, which was ranked top of the list for ACCHS, was ranked on the bottom of the DHS list. It was interesting to note that ACCHSs were interested in impact, large outcomes for their communities, whereas the DHS was interested in effectiveness - interventions can potentially work.
Most of the respondents from both organisations used the size of the health burden, acceptability, access and feasibility/sustainability as key criteria for priority setting. It was disappointing find that the DHS did not consider efficiency as a criterion for priority setting especially since they had reported efficiency as one of the key objectives for the DHS. ACCHS considered efficiency as a key criterion. The probable explanation for the divergence between the DHS and ACCHS in the use of efficiency as a criterion lies in the direct role that OATSIH has with ACCHS. The OATSIH is responsible for more than 80% of ACCHS’s core funding (Dwyer, O'Donnell et al. 2009), leading to emphasis for ACCHS to consider efficiency as a criterion for priority setting.
The ratings of the criteria were similar between the two organisations. However, there were differences around the rating of equity. Within programs, equity featured among the top five criteria in the DHS; however, it was not considered by many decision-makers in ACCHS as a key reason for the setting of priorities. The DHS responses highlighted the DHS’s role as a member of the funding agency, and its view on equity, the distribution and access issues that population face, since it takes the State government’s perspective. ACCHS were primarily service providers for Indigenous Australians and their families, and so equity across populations is not part of their responsibility.
The key shortcoming of the study is that the criteria identified did not indicate the intensity of the preferences of the respondents, since the respondents did not rank the criteria. The identification of the criteria for priority setting was based on the number of respondents
Page 120
who cited a specific criterion being the reason for decision-making. These results may not be useful for making trade-offs between criteria when choices are being made. This issue will be addressed in Chapter 6.
5.4.2 Improving the Uptake of Economic Evaluation Evidence Efficiency was not listed among the DHS core objectives in the literature (DHS 2007), so it was not entirely clear why efficiency featured as an objective in the survey results from the DHS respondents. One possible explanation for efficiency featuring as an objective is that efficiency is a criterion that is used in the context of other political imperatives. One respondent from the DHS (the Director for Public Health) said: “You will never hear me say something to do with efficiency, because I am not interested in it – it is not to say that it is not important but it is just a tool and not an issue to guide priority setting.”
The two organisations both identified ‘key organisation objectives’ as the main source of data for both short-term and long-term priority setting. The rest of the data sources identified by each sample were different. Interestingly, compared to the DHS, ACCHS were very responsive to mainstream public opinion when setting short-term priorities. This may reflect on the role of the mainstream public media in priority setting, especially in Indigenous communities, an issue that was not pursued in this thesis. Historical trends for setting priorities have been found to be the predominant approaches to priority setting in Canada and Australia (Mitton and Donaldson 2002; Mitton, Donaldson et al. 2003; Mitton and Patten 2004; Mitton and Prout 2004). It was encouraging to learn from the data that historical trends were not the main reason used by most of the respondents for the setting of priorities in Indigenous health in Victoria.
In terms of the assessment of the priority setting process, most ACCHS respondents felt that the resource allocation process in their organisations worked well and that it was efficient. The DHS bureaucrats considered the priority setting process in their organisations as not working well, with indication that the State government priority setting process in Indigenous-specific health services was not efficient. The bureaucrats were, however, not very consistent in their responses to the strengths and challenges of their priority setting
Page 121
process. For example, although the DHS respondents thought that the process was transparent, it involved the stakeholders and was widely accepted in the Indigenous community, there was a contradiction when respondents were asked about the challenges of the process currently. The respondents mentioned that there was little or no involvement by the stakeholders in the process, and that Indigenous values were ignored.
The survey also asked the respondents to assess the uptake of economic evaluation results in priority setting, and both organisations (77% DHS; 75% ACCHS) thought that there were challenges in using evidence from economic evaluation to improve the priority setting process. Just like ACCHS respondents, the DHS respondents indicated that upskilling of staff in health economics and/or more training of health economists would improve the uptake of economic evaluation evidence. Better data from commissioned studies was also much needed to improve the uptake of economic evaluation evidence. The size of the health burden was consistently used to guide priority setting. This finding contrasted with the findings from the survey of the WA Health Department, which had indicated that historical trends dictated the allocation of resources (Mitton and Prout 2004).
5.4.3 Issues for Consideration in Improving Priority Setting Respondents from both organisations ranked ‘better data/strength of evidence’ and the need for a ‘longer-term view’ equally as the key issues to be addressed in order to improve priority setting (Table 4.8). Preferences of the two samples were very similar. This information vindicates the need for an explicit priority setting process that is backed by technical rigour, especially in the form of economic evaluation.
The findings in this chapter indicate that the priority setting process in Indigenous health is not performing as well as the decision-makers would like it to be. Respondents indicated that there is a need to have an approach to priority setting which has an ability to shift resources matched with health gain, and to be able to have an impact on the health of Indigenous Australians. Such an approach would also need to demonstrate an overall, longterm increased health benefit to Indigenous Australians and the approach’s flexibility with the concept of benefit. Thus, such an approach would need to use available data/evidence
Page 122
(such as economic evaluation evidence), be pragmatic, and be based on both economic and social justice principles
For the priority setting process in Indigenous health to perform well, the respondents from both the DHS and ACCHS identified four key issues that need to be addressed in order for the priority setting process to improve. First, when priorities are being set, the key needs of decision-makers should include the need for a systematic and explicit transparent process, and a longer-term view to be taken. Issues of access, equity (especially for the DHS) and the concept of benefit (investigated farther in Chapter 8) were considered important issues for consideration in priority setting. Invoking equity and access implied that social justice issues were important in priority setting, issues which were discussed in Chapter 4.
Secondly, there is a need for timely and better data, and a better strength of evidence base, especially in the form of economic evaluation results for use in priority setting. This is the challenge that the funding agencies and health economists need to jointly address, and it was discussed in Chapter 3. Chapters 7 and 8 contribute to the strengthening of economic evaluation evidence, while focusing on contextualising PBMA in Indigenous health. Using the empirical studies on costing the health care services and the expansion of the concept of benefit to include non-medical benefits, this thesis will significantly contribute to the improvement of the quality of the data and strength of economic evaluation evidence.
Thirdly, the majority of the respondents from both the DHS and ACCHS indicated that they needed a mechanism for shifting resources (a decision rule) within programs which aids decision-making at the margin, and not just the use of the concept of total need.
Fourthly, the need to involve the Indigenous representatives in the priority setting process would enhance acceptability, support and adoption of the results. The effective involvement of the Indigenous Australians in the process can be done in a participatory manner using the ‘marginal analysis panel’ in PBMA, and the existing ACCHS infrastructure to identify representation in the panel. However, depending on the nature of the problem being addressed, the views of the community might be better taken into consideration rather than relying on only those of the key decision-makers. For example, in the expansion of the concept of benefits from health interventions to include non-medical benefits (Chapter 8),
Page 123
both Indigenous community representatives and government representatives actively participated in the project.
The use of an interview questionnaire for eliciting preferences from decision-makers meant that the benefits from alternative approaches, such as group-based approaches, were forgone. Group techniques require posing a specific question to the group of individuals and then getting their responses, which can then be ranked according their level of importance. Group techniques can include: nominal group techniques; Delphi techniques; single or multiple voting; constrained ranking; citizen’s jury; focus groups; and options appraisal technique (Hasman 2003). These approaches can be expensive and potentially taxing to organise (Jones and Hunter 1995; Kellogg 1998).
Individual approaches for eliciting public values include the following: individual assessment of preferences using written questionnaires or surveys; ranking or rating the importance of specific health care services, distribution principles or health states; assessing willingness to pay; grounded theory techniques; experimental survey designs such as discreet choice experiments (Penelope 1999; Mullen and Spurgeon 2000). These approaches can complement already existing data through observations and cross-checking whether previously collected information was correct (Kellogg 1998).
However, these techniques have several weaknesses. For example, it is impossible to furnish respondents with sufficient information about issues involved in the interview questions for them to have an adequate understanding of the issues, leading to inconsistent and unreliable responses (Massialos and King 1999). To overcome this, the surveys for this study were administered face-to-face and explanations of key concepts were undertaken in order to furnish respondents with more information on the interview question. This issue is specifically important in the interview questions that are closed ended.
Another fundamental issue with surveys involves the presumption that the specific moral theory underlies the questions, and that the respondents are bound by that theory (Holm, Gjersoe et al. 1996; Massialos and King 1999). This may not necessarily be the case. For example, respondents may be tempted to ignore social values in their responses and instead choose to concentrate on their personal interests by focusing on the official
Page 124
government line. To overcome this in this study, respondents were continually reminded during the interviews that their responses would be kept confidential.
However, content analysis may suffer from several disadvantages, both theoretical and procedural (Palmquist 2009), for example, when content analysis proves to be extremely time consuming and automating or computerising the data proves to be difficult – as was the case in this survey. Data was instead manually coded and manually counted. Content analysis does not always reflect the context that produced the text, as well as the state of things after the text was produced. To address this issue, quotes from the participants were included to ‘clothe’ numbers with meaning (Palmquist 2009).
Lastly, the use of the snowball sampling enabled the inclusion of decision-makers in the survey that would not have been known. It also helped in finding potential interviewees quickly they are known to each other. However, snowball sampling is strongly biased toward inclusion of those who have many interrelationships with or are coupled to, a large number of individuals (Berg 1988). It would not possible to know whether the sample was representative of the target population. To address the risk, identification of the respondents started from the top management which served to: identify subordinates involved in Aboriginal-specific health care services; and to provide indirect permission for the sub-ordinates to participate. In spite of the efforts made to address all the potential methodological weaknesses in this chapter, it is possible that some weaknesses still exist, which implies that these results ought to be treated with caution. The criteria that decisionmakers use as key criteria for priority setting include: the size of the health burden; acceptability; feasibility; sustainability; access; and equity. These criteria were, however, not ranked or weighted for relative importance.
Page 125
5.5 Summary In summary, these findings represent the first time that values of the decision-makers in Indigenous-specific health care services have been empirically elicited. These findings reveal the need for economic evaluation evidence from Indigenous health care interventions. This highlights the need for a set of guidelines that should be used for priority setting in Indigenous health, a task that is addressed in the next chapter. The respondents indicated that any priority setting framework in Indigenous health ought to consider the following:
The task of priority setting to reflect the needs of the community and organisation;
Clear concept of benefit that allows other dimensions of benefits to be captured;
Technical issues such as the use of evidence;
Clear ethical principles, such as efficiency and equity, to guide priority setting;
The importance of the acceptability of the process to the Indigenous population;
Involvement of the Indigenous Australians in the decision-making process;
The need for the due process to be followed in the decision-making process;
Clear decision rules such as shift resources backed by health gain;
The use of marginal analysis to maximise the health of the population; and
Considerations for resource issues and the implementation of the recommendations.
Page 126
6. Chapter 6: Guidelines for Priority Setting “The time has come to formulate guidelines for the ethical conduct of scientist, perhaps in the form of a voluntary Hippocratic Oath” – Joseph Rotblat (1908-2005).
6.1 Introduction In Section A of this thesis, it was argued that despite the government efforts in addressing Indigenous health problems since 1967, the health problems and the health gap between Indigenous and non-Indigenous Australians remains an issue, constituting social injustice50 (Anand and Peter 2000). To address this injustice, Section B of this thesis, which comprised Chapters 3-5, aimed to develop a checklist that would be used to guide priority setting in Indigenous health. The primary purpose of this chapter is to summarise and synthesise the findings from Chapters 3-5, in order to develop a set of guidelines for priority setting in Indigenous health. These guidelines will then be used to assess selected economic and noneconomic frameworks for priority setting in Indigenous health. The chosen framework will then be contextualised in Indigenous health in the next section (Section C).
6.2 Methods: Guidelines for Priority Setting The starting point for the development of priority setting guidelines in this chapter involved studies undertaken by Carter (Carter 2002; Carter, Vos et al. 2008), Segal (Segal 2001), Gold and colleagues’ study (Gold, Siegel et al. 1996) and from other studies in the literature (Mooney and Wiseman 1998; Mooney 2002; Mitton, Patten et al. 2003; NHMRC 2003; NHMRC 2003; Mitton and Donaldson 2004; Drummond, Sculpher et al. 2005; Mooney 2010). Gold and colleagues proposed the following as elements of an ideal priority setting framework: theory; ethics; empirical evidence; user considerations; and conventions (Gold, Siegel et al. 1996). Segal proposed another set which included: societal perspective; range
50
Justice has been construed as equality of original positions, opportunity, proportions and rights.
Page 127
of service options; clear objectives; marginal analysis; and evidence (Segal 2001). Carter developed more comprehensive guidelines which could be viewed as an extension of what Gold et al (1996) and Segal et al (2002) each developed separately (Carter 2002; Carter, Vos et al. 2008). Carter’s checklist included lessons from economic theory, ethical disciplines, pragmatism and user considerations, and comprised research questions; concept of benefit; range of service options; marginal analysis; decision criteria; the role of judgment; data needs tractability; due process; measurement rigour; and reporting/implementation (Carter 2002; Carter, Vos et al. 2008).
Lessons drawn from four key areas covered in Chapters 3-5 were synthesised to inform the development of a set of guidelines which can constitute the key elements of an ideal priority setting framework in Indigenous health. These lessons include: contribution from normative economic (E); contribution from social justice (J); contribution from the literature (L); and issues from the needs of decision-makers (N) in both Indigenous and nonIndigenous settings. These guidelines constitute an ideal framework which can be expressed as a checklist, and will be used to briefly appraise a selection of priority setting approaches in Indigenous health. The preferred approach will then be contextualised within Indigenous health in Section C.
6.2.1 Contribution from Economic Theory (E) Key issues for inclusion in priority setting from normative economics were discussed and summarised in Section 3.6. The key contribution involves economic principles such as efficiency principle and marginal analysis techniques. Other contributions include need for a clear and answerable research question, the need for the concept of benefit to reflect the objectives being pursued, the options being chosen to reflect opportunities forgone, and robust methods for measurement of costs and benefits.
Efficiency is the key principle that captures economic value of the policy objective. It is a type of justice which is a prominent criterion for allocating resources within economic frameworks and it is deeply rooted in consequentialism (Culyer 2005). It emphasises the need to reward the efforts of individuals regardless of the overall distribution of the
Page 128
outcomes and/or the community preferences, at least as defined in the N-C welfarist framework.
Estimating efficiency requires a clear specification of the objectives being pursued and social welfare function. The discussion in Chapter 3 revealed that the appropriate normative economic framework for specifying the social welfare function in Indigenous health is the non-welfarist framework. Such a framework allows attributes of the SWF to reflect government objectives (Musgrave 1959; Sugden and Williams 1978) and the ethical values of the Indigenous Australian community being targeted by the policy objectives (Culyer 1998; Mooney and Russell 2003). This ensures that social welfare is not governed by ideology but by stable ethical values of the Indigenous Australians as a society. Thus, the non-welfarist framework takes a societal perspective in order to identify all the pertinent issues in Indigenous health.
Under the non-welfarist framework, efficiency is assessed using utility and other non-utility based attributes of welfare such as health. The non-utility based attributes allow for social justice principles to be captured in the SWF as key objectives (Weinstein and Stason 1977; Sugden and Williams 1978; Culyer 1980). Given that the Indigenous values are centred on the community and the quest for self-determination, the non-welfarist framework allows for the concept of benefit for economic evaluation to be broadened beyond individual utility, and to reflect the objectives being pursued.
The QALY is the commonly preferred concept of benefit, based on both utility and nonutility attributes, by health economists (Tsuchiya and Williams 2001; Feldman and Serrano 2006). The QALY incorporates both the quality of life and the extension of life associated with health care intervention. However, the QALY sits uncomfortably in Indigenous health because it assumes that the SWF is made up of only health arguments, regardless of other community benefits (Shiell, Seymour et al. 1998). As a concept of benefit, the QALY is a limited measure for economic evaluation of Indigenous health care interventions because it does not explicitly take into consideration social justice issues, such as access or the concern for the welfare of others.
Page 129
Economic evaluation utilises marginal analysis as a mechanism to estimate the efficiency of the strategy to achieve the key objectives identified, using a clear concept of benefit and rigorous methods of measurement. The use of a clear concept of benefit, which is flexible enough to capture the non-health benefits improves economic evaluation results. The discussion in Chapter 3 further revealed that any priority setting framework in Indigenous health should embody an efficiency criterion, and should take into consideration Indigenous values and principles, such as community health gain and cultural security.
6.2.2 Contribution from Social Justice (J) The discussion on Indigenous values and principle can be addressed through social justice theory. The contribution from the social justice to priority setting (sometimes referred to as ‘ethical considerations’) was discussed in Chapter 4. It was argued that social justice can effectively inform priority setting through the use of justice principles such as equality, needs, equity, and procedural justice. These principles can be adopted as key social objectives in themselves and thus enter the SWF, or can be used as a checklist to guide the priority setting process. Such objectives and procedures of justice and equality may include fairness or equity, human rights, equal access (both physical and perceived forms) to basic health resources, sensitivity to the needs of minority and marginalised groups, intergenerational justice, respect of the views and values of targeted group, and consideration for efficiency.
For example, access may be addressed through the concept of autonomy and/or community control of the services. This can be demonstrated by the Indigenous community control of health care services for their communities, the right of people to control individually and as a community, as far as is feasible, their health care and treatment. This involves the whole community participating or being effectively represented in the decisionmaking process, or having tangible input into the setting of health care goals and targets and in evaluating the latter. It also involves setting up structures and processes to enable such community inputs, which should be something more than one-off ‘community consultations’, to be made effectively.
Page 130
Access to health care services may also be improved using strategies that ensure cultural security of the services utilised by Indigenous Australians. Cultural security would be demonstrated by the respect of the individuals, human dignity and social worth, and sanctity of the human body and life. This would be a characteristic of relationships between people, and the way in which individuals behave, being fundamental to the functioning of the moral society. Improving access can potentially have a positive impact on equity issues.
The notion of equity in moral philosophy can be classified in terms of how the consequences of personal preferences are accounted for in priority setting (Konow 1996; Konow 2003). In practice, it could be captured using vertical or horizontal equity. Horizontal equity may be captured using the objectives such as equality of access/outcomes for equal needs. Vertical equity, which implies unequal levels of access/utilisation due to unequal need as in the ‘closing the gap’ government objective, would be the key equity and/or equality objective.
Equality principles can be expressed as human rights principles, such as cultural diversity, human dignity, sanctity of human body and life, and social worth in the access to health care resources. Procedural justice is about fairness in the process of priority setting, ensuring unbiased, consistent, impartial and reliable outcomes (Maiese 2003). This would further ensure the acceptability of the priority setting methods and decisions. The challenge however, is to learn from the lessons from countries which have attempted priority setting.
In summary, selecting options for review should be guided by explicitly stated ethical principles which are empirically determined from Indigenous participants and the decisionmakers. The concept of benefit should also reflect the ethical objectives being pursued. The decision-making process should allow ethical principles to inform value judgement in priority setting.
6.2.3 Lessons from the Literature (L) Several lessons can be drawn from the experiences of countries (such as the United States, the Netherlands, Sweden, New Zealand and the United Kingdom) that have made attempts at systematising priority setting in health (Ham and Coulter 2000; Shiell and Mooney 2002; Mitton and Donaldson 2004; Carter, Vos et al. 2008). For example, Sabik and Lie identified
Page 131
three key criteria for judging priority setting frameworks: Indigenous involvement in the priority setting process; establishment of the appropriate principles; and potential to address the issues identified (Sabik and Lie 2008). Other lessons involve the recognition that the different levels of government may influence the criteria and needs for priority setting. For example, the priority setting needs of the decision-makers at the Federal and/or State levels would be different to those of the decision-makers at the clinical or Indigenous community levels. For example, it may not be appropriate for a doctor to tell older patients that they cannot receive organ transplants because organs are reserved for the young. There is no single solution or method for priority setting and any approach requires flexibility.
Priority setting requires the best available technical evidence to inform the priority setting process. Strengthening the evidence base to guide decision-making must not only include economic evaluation evidence, but also requires a combination of technical rigour and value judgments about what constitutes benefits or costs (Mooney 2006). Technical judgements may rely on the availability and the quality of data, whereas value judgements may rely on the principles and values of the decision-makers and the target population. Equity, for example, can be addressed using some existing explicit formulae for allocating resources based on some assessment of the differential needs of different populations (e.g. degree of sickness, remoteness, etc.). This could be done by weighting benefits to some groups (Indigenous Australians, for example) more highly than the benefits to other Australians. Weighting these needs or the use of any other mechanism to address equity issues ought to be informed by the community values and other explicit criteria. Where data is not readily available, the expert advice has to be part of the process for assessing the quality of evidence in Indigenous health.
De-listing services is not an effective approach to priority setting, and it may be inconsistent with solidarity in the community (Shiell and Mooney 2002). Shiell and Mooney explained that very few services are always ineffective or inappropriate, and those that are tend to be insignificant in terms of overall health care (Shiell and Mooney 2002). The benefits of delisting may be slight, but the harms that can be generated in terms of galvanising opposition to rational planning can be great.
Page 132
The decision-making mechanism should ensure that Indigenous values and aspirations, as well as the needs of the decision-makers, guide the process of priority setting. This would legitimise the decisions made and increase the potential acceptance and adoption of the results. The establishment of appropriate incentives and institutional structures would ensure that the priorities identified are acted upon and the chosen priority setting framework is adopted as routine decision-making tool (Shiell and Mooney 2002).
The lessons above can be translated into practical guidelines for explicit priority setting. One key issue involves ascertaining the broad principles/criteria that should guide the priority setting process and the trade-offs that Indigenous Australians are prepared to make between the principles. In essence, these criteria should include the Indigenous values and aspirations, or what Mooney and colleagues refer to as the Indigenous ‘health constitution’ (Mooney and Wiseman 1999; Mooney 2000; Mooney 2002; Shiell and Mooney 2002; Mooney 2003; Mooney and Houston 2004; Mooney 2010). Furthermore, these criteria ought to reflect the social justice principles and the due process of priority setting.
Explicit criteria for priority setting from the literature indicate that both technical and social justice principles should be used to guide resource allocation. In New Zealand, for example, as in the United Kingdom, the criteria for priority setting include: effectiveness; equity; efficiency; and acceptability (Ashton, Cumming et al. 2000; Tito and Kavanagh 2004). In the Nordic countries, solidarity and human dignity, effectiveness and need are commonly used as criteria to guide explicit priority setting (Government Committee on Choices in Health Care 1992; Mooney, Gerard et al. 1992; Health Care and Medical Priorities Commission 1993; Mechanic 1995; Danish Council of Ethics 1997; Mechanic 1997; Office for Oregon Health Plan Priority and Research 1999; Daniels 2000; Ham and Coulter 2000; Oberlander and Marmor 2001; Carter 2002; Kapiriri and Norheim 2004; Sabik and Lie 2008). Emphasis is placed on the accountability and transparency of the funding and facilitation of the improved access to health care services for those who are at a socioeconomic disadvantage. Other countries have efficiency and need, such as severity, effectiveness, quality care and burden of disease, as criteria and emphasise the fairness of the process. Respecting people’s human rights, human dignity, autonomy, solidarity, and cultural security ensures accountability and acceptability of the process.
Page 133
There have been several attempts to produce guidelines to guide resource allocation in Australian health. For example, the Australian Health Ethics Committee (AHEC) established a subcommittee in 1992 to promote discussion of the ethical issues to guide health care resource allocation in mainstream health care sector (NHMRC 1993). The subcommittee held a workshop in June 1992 and identified the following ethical issues to guide resource allocation: justice; autonomy; respect for human life; quality of life; respecting the integrity of the human body; importance of human disabilities; and social worth.
Ten years later, the NHMRC undertook another workshop-based approach to identify values to guide research in Indigenous health (NHMRC 2003). Six values were identified and subsequently adopted as the basis of the NHMRC ethical conduct of research in Indigenous and Torres Strait Islander health. These included the following: spirit and integrity; reciprocity; responsibility; respect; survival and protection; and equality. Around the same time, NATSIHC identified nine principles or values (although from a different context) that ought to guide explicit priority setting in Indigenous health (NATSIHC 2003). These principles include: cultural respect or cultural security; a holistic approach; health sector responsibility for the health of Indigenous Australians; Indigenous community control of primary health care services; government, NGOs and private organisations working together; localised decision-making; promoting good health; building capacities in the Indigenous community controlled sector; and accountability. The challenge is how to translate these broad principles into specific guidelines for priority setting in Indigenous health. In summary, lessons from the literature indicate that the following should be addressed in priority setting:
Specification of well-specified and answerable questions;
Needs assessment/options selection mechanism (E, J, L, N);
Clear ethical principles to guide the process such as in the selection of interventions/options, and to guide value judgement;
The concept of benefit to reflect the objectives;
Explicit criteria for technical judgements;
The use of marginal analysis and technical rigour in the measurement process;
A fair process for decision-making which involves Indigenous Australians; and
Relevance of the results to ensure implementation.
Page 134
6.2.4 Needs of Decision-makers (N) The findings from the survey reported in Chapter 5 indicated that the decision-makers use the size of the health burden, acceptability, feasibility/sustainability, access, and equity as key criteria for priority setting. They expressed the need for a systematic and explicit transparent priority setting approach that can assist them to shift resources matched with health gain (the need for marginal analysis. This implies that such an approach would need to be able to capture both costs and benefits, be rigorous in its assessment of benefit and to have decision rules on how to choose between alternative policy objectives.
The decision-makers also expressed the need for a priority setting framework that would be flexible in its assessment of the concept of benefit, so as to allow non-medical benefits to be captured in the assessment of the objectives. The non-medical benefits include community health gain, the benefit from a health care intervention which goes far beyond the benefits accrued by an individual. Cultural security, equity and access as dimensions of benefit not traditionally captured in the concept of benefit for economic evaluation in mainstream health care, were also considered to be of significance to Indigenous Australians.
Most importantly, such an approach should demonstrate its ability to have a positive impact on the health of Indigenous Australians by having an ability to make long-term health gain a reality, not just intellectual elegance. These requirements are consistent with the need for an economic approach to priority setting. Closely related to this is the need for timely and better data, and better strength of evidence (especially in the form of economic evaluation results) for use in priority setting.
Acceptability of such a priority setting approach within Indigenous organisations and the community at large was also considered to be crucial for its success. The involvement of the stakeholders (especially Indigenous representatives) in the priority setting process was considered to be one way to ensure the acceptability, support and adoption of the results of the stakeholders in the priority setting process. These issues constitute the due process issues in priority setting, and ensure acceptability and ownership of the process and of the results. In summary, the contribution from the survey indicated the following: the
Page 135
specification of a clear and answerable question; explicitness in criteria for selecting options for change and for making judgments; emphasis on technical rigour in measurement of costs and the concept of benefit, and in marginal analysis; and incorporation of the principles of social justice such as human rights and equity into the priority setting process.
6.3 The Checklist for Priority Setting The above lessons (lessons from normative economics, social justice, the literature and empirical studies in priority setting) were synthesised and used to generate a set of guidelines for priority setting. An attempt is made to indicate the key areas from which they were developed, from economic theory - E, social justice - J, literature – L, and/or needs of decision-makers - N. The following constitute guidelines for priority setting and can constitute the key elements of an ideal priority setting framework in Indigenous health:.
Well-specified and answerable questions (L, E, N):
Needs assessment/options selection mechanism (E, J, L, N);
Ethical principles to guide the process (J, L, N);
The clear concept of benefit that reflects the objectives being pursued (E, J, N);
Clear criteria, both technically and value-based, for making judgements (E, J, L, N);
The use of marginal analysis in the framework E, (L, N);
Technical rigour: the robustness of the measurement process (E, L);
Due process, the fair process of decision-making which involves Indigenous Australians (E, J, N); and
Relevance and implementation of the results (L, N).
The above guidelines were then converted into a checklist for priority setting (Box 6.1). The following questions are designed to guide priority setting in Indigenous health.
Page 136
Box 6.1: Checklist for priority setting in Indigenous health 1) Well-specified and Answerable Questions (L, E, N) Are the objectives of the exercise clearly specified? Do objectives include equity, efficiency, access, etc? Are human rights ‘claims’ part of the objectives? Is the perspective for decision-making specified? Are comparators specified? Is the question adaptable to different decision contexts and issues? Is the framework appropriate to the question and context that is specified? 2) Needs Assessment/Options for Change (L, E, J, N) Is needs assessment undertaken? Is there an explicit method for generating options for change? How comprehensive are the options, given the perspective? Are the pathways for the options clearly defined and specified? Does needs analysis involve a sufficient evidence base? Does the method embody ‘opportunity cost’, e.g. cost of illness study? Does the method involve both increments and decrements? Does the method involve an inventory of the services currently provided? Does the inventory of the services include the current expenditure? 3) Ethical Principles (L, J, N) Is the framework capable of addressing ethical principles relevant to Indigenous health? Are Indigenous values embodied in the framework as guiding principles? Is the respect for human life appropriately taken into account? Is the respect for the integrity of the human body appropriately recognised? Is the importance of human disabilities appropriately evaluated? Is accountability a guiding principle in the priority setting framework? a) Equity (L, J, N) Does the framework address issues of fairness or equity? Are considerations of efficiency being allowed to outweigh justice and fairness issues? Are health care recipients being positively discriminated on ethically relevant grounds? Has consideration been given to the impact on future generations? b) Access (J, N)
Page 137
Does the framework recognise access issues to appropriate and effective health care? Is access viewed as both a physical and a perceived issue? Are Indigenous Australians being assisted to access health care resources? Are institutional barriers to Indigenous access to services appropriately addressed? Are Indigenous liaison officers employed to enable Indigenous access to services? Is access to health care being determined by appropriate ‘social worth’ criteria? c) Community Control (J, N) Is the right of Indigenous Australians to control their health care being respected? Is the Indigenous community’s decision-making, participation and control supported? Are there effective structures and processes to ensure effective Indigenous participation? 4) The Clear Concept of Benefit (E, J, N) Is there a clear and appropriate concept of benefit? Does the concept of benefit appropriately link objectives and perspective of the exercise? Are quality of life considerations appropriately recognised? Are benefits from the community, cultural security, spirituality and access considered? 5) Criteria for Making Judgements (L, E, J, N) Are the decision rules for ranking options for change clearly specified? Is efficiency an integral rule for making judgement? Is the role of judgement recognised in the specification and interpretation of the results? Are criteria, for making decisions clearly stated and ranked for relative importance? 6) Marginal Analysis (L, E, N) Is marginal analysis an integral component of the framework? Are resource shifts in ‘bite size chunks’ or gradual movement? 7) Technical Rigour (L, E) Are the data needs manageable or tractable? Do the measurement methods, especially for benefit, demonstrate appropriate rigour? Are standardised protocols, such as economic evaluation protocols, used? Is expert advice used where data may not available or is considered to be of poor quality? 8) Due Process – Involving Indigenous Australians (J, N) Is the need for due process recognised? Does the framework allow for a fair, transparent and open process to priority setting?
Page 138
Does it provide for active Indigenous community involvement/consultation? Does it make use of Indigenous representatives in the priority setting process? Does it allow for the Indigenous advice to be embraced in the priority setting process? 9) Relevance and implementation of the results (L, N) Does the model produce and report results that address priority setting needs? Is there capacity for the implementation of the results? Does the framework contribute to capacity building in Indigenous health services? Do the results from the framework support localised decision-making? Are the transaction costs of the exercise prohibitive?
6.4 Assessing Selected Non-Economic Models The above checklist was used to qualitatively assess a selection of economic and noneconomic frameworks in order to identify the frameworks that can potentially be useful for priority setting in Indigenous health. The non-economic frameworks include: historically based allocations; needs assessments (burden of disease, cost of illness, community surveys, goals and targets); clinical guidelines and protocols (defining core services, best practice guidelines, and shared protocols); and ethically based approaches (consensusbased approach, accountability for reasonableness, and rights-based approaches).
6.4.1 Historically based allocations Historically based allocation is the most common approach to priority setting and resource allocation in which decisions to fund or provide services are based on what has traditionally been done (Mitton and Donaldson 2004). In Victoria, for example, Casemix allocations in hospitals are based on the throughput of the nominated year. The decision rule involves indexing the previous year’s allocation plus or minus a nominated percentage for changes in costs and/or demographic profiles. In essence, it assumes that the previous years’ allocations and priorities are optimal, and so reductions or increases in allocations are equally optimal.
Page 139
It is a less controversial approach, and easy to implement (Mitton and Donaldson 2004). However, this approach is not appropriate for use in Indigenous health because it does not provide a mechanism for maximising health benefit within a given budget, thus failing to address the efficiency principle. Furthermore, where inequalities exist and the current allocations do not seem to reflect the level of health need, it has the potential to perpetuate the health inequalities and injustice that already exist. Furthermore, the historically based allocations do not involve the Indigenous Australians in the decision-making, which is not in line with Indigenous concept of community control/self-determination.
6.4.2 Needs-based Assessments Needs assessments define health problems in a population by assessing the comparative health need over time, and across regions (Mitton and Donaldson 2004). These approaches are mainly epidemiologically based studies, and may include the burden of disease studies (such as cost of illness studies51, and avoidable mortality and morbidity), community surveys52, and setting goals and targets53 to be achieved either at an individual or a societal level. The prevalence and incidence parameters are often used to estimate the levels of unmet need. The decision rule under the needs-based approaches involves increased funding for the greatest un-met need (Mitton and Donaldson 2004). These approaches provide good insight into the nature of inequalities in health and to the nature of the contributions to the inequalities. They also have potential to address the due process issues in Indigenous health.
However, these approaches do not provide information on how much and in what manner resources should be spent or shifted to match the identified burden of disease (Mitton and
51
Cost of illness studies involve estimating the impact of the disease or illness based on its epidemiological profile, and the health service use and cost. It includes direct costs of the illness associated with the management of disease, indirect costs incurred by the rest of the sectors due to lost productivity or premature deaths, and intangible costs or loss in wellbeing that people experience. 52 Community surveys involve assessing community views about certain issues and conditions using survey instruments. They may involve identifying issues of concern in the community with or without ranking them as priority areas for health care investment. 53 Burden of disease studies are often used to set goals and targets for health planning. They begin with the description of the health problem, and then proceed to specify targets for the incidence and prevalence of the selected problems.
Page 140
Donaldson 2004; Segal and Mortimer 2006). Furthermore, although these approaches may be rigorous in measurement, they often ignore the opportunity costs of the services to be provided. In conclusion, these approaches are best used in combination with economic frameworks which provide mechanisms to assess efficiency.
6.4.3 Human Rights-based Approach In general, the rights-based approaches to priority setting use human rights values and principles to systematically focus attention on underlying power dynamics that deny people access to health services. These values and principles may include full and comparable health
status,
universality
and
inalienability,
indivisibility,
interdependence
and
interrelatedness, non-discrimination and equality, participation and inclusion, accountability and the rule of law (Office of the High Commisioner for Human Rights 2003; Gruskin and Daniels 2008). In Indigenous health care specifically, these values and principles may include full and comparable health status, equity in resource utilisation, access issues, accountability of the decision-makers, procedural issues, and capacity developments, as the bases for priority setting.
However, human rights are indivisible and all have equal status in principle, and cannot be ranked in any hierarchical order (Gruskin and Daniels 2008). When choices have to be made on what to provide given the scarcity of the resources, it is important to be able to rank or weight these principles. Failure to do so renders the human rights-based approaches difficult to use for priority setting (Dolan, Shaw et al. 2005).
The theory supports the idea of giving some priority to those who are worse off, but is silent on how much it should be and thus, fails to provide the decision criteria for choosing between alternative courses of action for a particular policy decision, given scarcity of resources, generating unresolved priorities. Unfortunately, in both the literature and in practice, when human rights are invoked, it is often assumed that simply by invoking its key components, the appropriate solution to setting priorities will present itself (Asher 2004). Neither the human rights nor the appeal to general principles of distributive justice suffices to resolve disagreements about how to decide among these claims in setting priorities. The theory does not provide a mechanism of making trade-offs between achieving best
Page 141
outcomes against giving people fair chances at some benefit (Kamm 1993) because the winners and losers created by priority setting choices all have rights and claims of distributive fairness on their side.
In summary, the human rights-based approaches set out good guidance by highlighting government accountability in improving the health of the Indigenous Australians. However, taken on their own, human rights-based approaches are too general to yield determinate answers to the key distributive questions required in setting priorities. These approaches are not suitable for maximising the health of Indigenous Australians.
6.4.4 Accountability for Reasonableness The accountability for reasonableness approach places emphasis on the fairness of the chosen process for priority setting, thus emphasising procedural justice. It is a deliberative process grounded in the theory of democracy (Daniels and Sabin 2002). The approach assumes that people are ‘fair-minded’ and seek mutually justifiable terms of co-operation, and that they are able to agree on, and justify the reasons for the priorities necessary to meet health needs fairly, even when resources are constrained. For accountability for reasonableness approach to work, four conditions have to be fulfilled (Mitton and Donaldson 2004):
Publicity: Decisions that establish priorities in meeting health needs and their bases must be publicly accessible. Forming a public record of decisions and reasons behind them enshrines transparency and accountability in the priority setting process;
Relevance: Decision-makers should aim to provide the reasons for the choices in priority setting. Such reasons may include evidence, and community values and principles. Closely linked to this condition is the inclusion of a broad range of stakeholders in decision-making, to ensure that a range of relevant arguments and interests are considered but also allows for buy-in and enhances legitimacy, even as the difficulties inherent in ensuring vulnerable groups are heard is acknowledged;
Revision and appeals: There must be mechanisms for the revision and improvement of policies in light of new evidence or arguments; and
Regulation: There must be public regulation of the process to ensure that conditions 1, 2, and 3 are met.
Page 142
Accountability for reasonableness provides a systematic rationale for the application of the key elements of a human rights-based approach. It emphasises transparency in selecting the interventions, such as the rationale used to give priority to a particular area of focus, the need to directly inform affected communities of the criteria used in selecting goals and targets, and the mechanisms in place to ensure government accountability over time (Chapman 1997; Office of the United Nations High Commissioner for Human Rights 2006). Researchers in Canada and elsewhere in the world have embraced this approach as a framework for evaluating priority setting in various countries (Ham and Robert 2003; Martin, Shulman et al. 2003; WHO and UNAIDS 2004; Rawlins 2005; Miller, Robinson et al. 2006). Accountability for reasonableness exposes government incapacity or unwillingness to improve Indigenous health. By making explicit the process through which priorities are set and decisions are made, it becomes much harder to disguise unwillingness. Even as resource constraints may not permit a completely deliberative and equitable process in every instance, accountability for reasonableness helps eliminate much of this vagueness by yielding an explicit record of why and how alternatives are chosen. Over time, such a record may reveal the true commitments to change in Indigenous health that a government is able and willing to make. The key weakness with this approach involves its inability to incorporate marginal analysis and other technical methods in the decision-making process. It is therefore not a preferred approach to maximising health in priority setting.
6.5 Assessing Selected Economic Approaches The selected commonly used economic-based approaches to priority setting were also qualitatively assessed using the checklist in Section 6.3. The purpose of the assessment was to identify which economic approach can best address priority setting in Indigenous health. These approaches included: the league table approach; the Health Sector-Wide Model (HSW); Health Benefit Groups/Health Resources Groups (HBG/HRG) and PBMA.
Page 143
6.5.1 League Table Approach The league table approach involves bringing together the results from economic evaluation and then ranking them in the order of marginal cost per QALY or life years in a single priority setting exercise. In such an exercise, the key objective of the health sector is the QALY. The Oregon experiment is one of the most commonly cited examples of the league table approach to priority setting. This experiment took place in the state of Oregon in the United States (Brown 1991; Haddorn 1991; Fox and Leichter 1993; Office for Oregon Health Plan Priority and Research 1999; Oregon Health Services Commission 2005). It was initially based on the cost per QALY but later revised to reflect the wishes of the community in which cost per QALY became a secondary criterion for priority setting.
The league tables provide the decision-makers with information on two key areas: what the forgone opportunities of different individual interventions are; and information on the bundle of interventions, and various options for change, for a given context. The forgone opportunities are reflected in the relative cost effectiveness results for the given context. In practice, the concept of benefit may not reflect all the attributes of the SWF. In Indigenous health, for example, the use of DALY may under-value the benefits from interventions, because the concept of benefit may be broader than the DALY (this is investigated in Chapter 8). In the case of the bundle of interventions the decision-maker may choose interventions from the least cost per DALY until the budget runs out. The other option is to choose all interventions with cost per DALY below a certain minimum or critical value of cost effectiveness, a value which may have been derived from the literature, past decisions, or rule of thumb (Weinstein 1995). However, the league table does not provide appropriate concept of benefit or values attached on the weights placed on the ratios. In most cases, QALYs or DALYs are assumed to be of equal value – the utilitarian principle, ignoring both distributive and procedural justice (Mooney 1992; Carter 2002).
Another issue with the league table as a priority setting tool is that it would be a very demanding exercise if it were to be used for the entire Indigenous health sector. It would mean estimating the cost per QALY for all possible interventions/services within Indigenous health programs and across all programs, diseases and different locations. This would constitute a significant research exercise that an Indigenous community organisation would
Page 144
most probably be unable to undertake or interpret the results on their own. Such an exercise would require large quantities of data and amount of resources, yet the league tables approach does not explain how to manage such data. It is therefore not an appropriate tool on its own. The performance of the league tables approach is summarised in Table 6.1.
Table 6.1: Performance of the league tables approach Criterion Well-defined question: What is/are the scope, goals or objectives to be achieved Needs assessment Ethical principles in selection of interventions or options for change Concept of benefit
Criteria for making judgements Marginal analysis Technical rigour in measurement
Due process and Indigenous involvement Relevance or implementation of the results Overall assessment
Performance The research question is often well defined and potentially achievable as most of the studies follow standard economic evaluation protocols. The scope of the questions is often dependent on the perspective and context taken, and so varies from study to study. Needs assessment often involves quality of life studies and the existence of such evidence to support analysis. Ethical principles from social justice (such as equity, community control, cultural security and access) are ignored when options for change are being considered. No explicit criteria for selecting interventions are in place. The concept of benefit is often reported using DALYs or QALYs. These studies often ignore non-medical benefits as reflected in the Indigenous health construct. The criterion commonly employed is efficiency – which recognises opportunity costs. Has potential to utilise marginal analysis. The current practice is using average cost effectiveness results. Rigorous approach to measurement is potentially achievable. However, in cases where the league table involves analysis of economic evaluation literature, the level of rigour becomes difficult to trace. Not normally considered. Issues of relevance to priority setting and implementation of the results are not explicitly considered. Interest groups cannot easily influence the results. The league table can potentially provide the decision-maker with information on costs and benefit (however narrow it may be) and the average cost effectiveness results. However, its failure to incorporate due process in the analysis constitutes the key weakness for use in Indigenous health. It is best used with a non-economic framework such as accountability for reasonableness.
6.5.2 Health Sector-wide Model The HSW model is a sequential approach to disease planning and offers comprehensiveness of scope in identifying intervention options (Segal and Mortimer 2006). It focuses on the patient needs that are structured via the disease pathway and it is centred on specific
Page 145
interventions. The HSW model seeks to establish desirable resource shifts that would minimise morbidity and mortality for all diseases and health issues, given the total current allocation of resources in the sector.
The model categorises the entire health sector into diseases and health problems in order to provide a framework that allows the HSW priority setting task to be staged. It then employs standard economic evaluation with the aim of maximising QALYs contingent upon the global budget constraint for the set of diseases/health problems under consideration. The decision context is marginal analysis, beginning with a disease continuum and then across stages for the population sub-groups (Segal and Mortimer 2006). It recognises nonhealth arguments in its specification of the objective function, and that political constraints frequently limit degrees of freedom.
The HSW’s strengths are its ability to cover the whole health sector, its focus on health interventions, and requirement for comprehensiveness in the range of interventions. It tends to separate technical issues from implementation issues. The HSW model relies on the use of objective evidence, with a limited role for clinical experts and consumers (Segal and Mortimer 2006), which renders it weak in terms of addressing the due process issues which are important in Indigenous health. The model fails to provide a mechanism for specifying non-health and political issues in the concept of benefit. Thus, it does not entrench community control in the priority setting process and so its use may not be appropriate in Indigenous health.
Furthermore, the model is data ‘hungry’, requiring massive amounts of data to undertake economic evaluation of different interventions across the disease spectrums, diseases and across populations. It also fails to provide a mechanism of handling the massive data requirements or what to do in case of the absence of such data. Applying it in Indigenous health may be a challenge due to the absence of data and poor quality of the existing data in Indigenous health. Overall, the HSW model is similar to the league tables approach, but with well-developed technical methods. It is weak on due process and other issues and so it is not a preferred approach to priority setting in Indigenous health on its own. Its performance against the checklist developed earlier is summarised in Table 6.2 indicating
Page 146
that it is best used in conjunction with a non-economic model which addresses due process issues.
Table 6.2: Performance of HSW model Criterion Well-defined question
Performance The research question is often well defined and a macro-based model which takes a societal perspective. It is not designed for use at Indigenous community organisation level. Needs assessment/options Needs assessment often involves quality of life studies and the selection existence of such evidence to support analysis. Has a comprehensive and well-developed mechanism for options generation and selection. Ethical principles Ethical principles from social justice (such as equity, community control, cultural security and access) are ignored when options for change are being considered. Concept of benefit The concept of benefit is the QALY. Equity and other community views are recognised but it is silent on how these are built into the framework. Criteria for making The criterion employed is efficiency – which recognises opportunity judgements costs. Minimises value judgement by concentrating on technical criteria for making decisions. Marginal analysis Utilises marginal analysis and opportunity cost. Relies on the literature to generate results. Technical rigour in It is rigorous in its approach to measurement. It has huge data demands measurement which are similar to league tables without a clear mechanism to manage data demands. Due process and Not considered and no mechanism to allow procedural issues to be Indigenous involvement taken into consideration. Relevance/implementation Issues of relevance to priority setting and implementation of the results of the results are heavily skewed to the researcher. No mechanism for explicit consideration of the decision-maker’s views is in place. Interest groups cannot easily influence the results. Overall assessment Very similar to the league table with highly developed technical analysis. It does not provide for an explicit involvement of Indigenous Australians. It emphasises the technical rigour at the expense of the due process, so it is best suited to be used with a non-economic model.
6.5.3 Health Benefit Groups/Health Resource Group HBG/HRG framework, just like the HSW framework, seeks to establish desirable resource shifts that would minimise morbidity and mortality for all diseases for all the populations given the total current allocation of resources in the sector (Sanderson and Mountney 1997; Beaver, Williams et al. 1999; Segal and Chen 2001). Resource shifts are based on marginal analysis beginning with a disease continuum and then across stages for the population subgroups. The HRG/HRG groups the population into the following categories: population not at risk; population at risk; population with the disease; and population with ongoing
Page 147
consequences. It then maps those groups onto the services available to estimate the resources needed to address the problems as a basis for resource allocation. The rows show HRG and the columns show HBGs. The aim is to identify where the resources can be best invested to achieve the highest returns in terms of health gain. The following constitute key steps in this approach:
Selection of disease for analysis;
Allocation of population into health benefit groups, HBG, as those not at risk, those at risk, those who have presentations, acute and chronic disease;
Gathering of evidence on costs and effectiveness of alternative treatment options; and
Analysis to identify those groups that will benefit most from treatment, and those with minimal capacity to benefit to be excluded from purchasing arrangement.
Table 6.3: Performance of HBG/HRG approach Criterion Well-defined question Needs assessment/ options selection Ethical principles Concept of benefit
Criteria for making judgements Marginal analysis
Technical rigour in measurement Due process and Indigenous involvement Relevance/ implementation of the results
Overall assessment
Performance The research question can potentially contribute to priority setting as a financial planning tool. It allows for different perspectives to be taken – ranging from societal to provider level. It tends to use burden of disease data to assess the level of need at the disease level. No mechanism has yet been developed to explicitly deal with the selection of interventions. Ethical principles, such as equity, community control, cultural security and access, are ignored when options for change are being considered. Benefit is taken to be health gain and equity. There is no inbuilt mechanism to clarify the concept of benefit and equity considerations are not canvassed. The decision context uses standard optimisations rules using health gains, cost savings or cost per DALY. The judgement role is not explicitly specified. The approach uses incremental analysis to address allocative efficiency, using the status quo as the base comparator. It ignores technical efficiency at intervention level. Technical rigour in the measurement of outcomes and costs is achievable. It has large data needs that are potentially unmanageable, which renders it inapplicable to Indigenous health where data is an issue. No consideration for the due process, since it is a purely technical method for priority setting. Indigenous involvement is not provided for. Intervention level analysis of the options for change is not considered which renders implementation of interventions to address the chosen areas for investment problematic. Furthermore, pragmatic and ethical issues such as feasibility or acceptability to the Indigenous community is not taken into consideration It is a potentially useful tool for health planning and priority setting. The challenge is for the framework to evolve to such an extent that it can
Page 148
assess efficiency of interventions and be able to address due process issues in Indigenous health. Furthermore, the cost and outcomes assumptions are not yet very clear.
It is a conceptually appealing framework for purchasing services since it matches health need with service delivery and resource use. However, it assumes there is a linear relationship between investment and returns, and that the groups in the HBG or HRG are homogenous, opening itself up to criticism (Carter 2002). In its potential application in Indigenous health, those assumptions are likely to fail due to multiple issues and heterogeneity of the groups. The Australian application was undertaken in the NT, where a computer program was developed based on the model (Beaver, Williams et al. 1999). In North Queensland, HBG/HRG was applied in Indigenous health, but not evaluated for its usefulness (McDermott 2004; McDermott and Mills 2004). HBG/HRG approach needs a large amount of data, which may prove an issue in Indigenous health (Deeble 1999; Segal and Chen 2001). The strengths and weaknesses of HBG/HRG against the checklist above are summarised in Table 6.3. It performance indicates that this approach would best be used in conjunction with a non-economic model; even then, it still needs refinement of its assumptions and costing approach.
6.5.4 Program Budgeting and Marginal Analysis PBMA is a relatively common approach to priority setting in the health sector, and tends to take the agency perspective. Recent modifications to PBMA have led to its health sectorwide applications, like the exclusion of the program budget component, as in ACE approach (Carter, Vos et al. 2008) and macro marginal analysis (Mitton, Donaldson et al. 2003). It utilises the best available evidence in an explicit and systematic planning process, and has the
potential
to
uphold
social
justice
principles,
such
as
distributive
and
process/participatory issues associated with priority setting.
One of the strengths of PBMA involves its responsiveness to local objectives and constraints. This implies that PBMA has potential to assist local decision-making in the Indigenous community organisations. However, the use of the subjective views of the panel may potentially not reflect the full range of arguments in the SWF.
Page 149
Table 6.4: Performance of PBMA against the checklist Criterion Well-defined question Needs assessment Ethical principles
Concept of benefit
Criteria for making judgements Marginal analysis
Technical rigour in measurement Due process and Indigenous involvement Relevance/ implementation of the results Overall assessment
Performance The research question is often well defined and achievable. It is suited to organisational or regional level planning. The scope of the questions is often dependent on the perspective and context taken. Needs assessment often involves quality of life studies and the existence of such evidence to support analysis. Ethical principles from social justice (such as equity, community control, cultural security and access) can be used to assess the options for change. Explicit criteria for selecting interventions can be used. The concept of benefit can be DALYs, QALYs or any other metric which appropriately captures the key objectives being pursued. It is capable of including non-medical benefits as reflected in the Indigenous health construct. The criteria commonly employed include ‘value for money’, and opportunity cost. It is capable of including other social justice criteria such as equity, access and human rights principles. It is based on marginal analysis and other economic concepts of opportunity cost. The extent to which marginal analysis can be achieved varies from study to study. Rigorous approach to measurement is potentially achievable. It has been criticised for leaning too much towards expert judgement at the expense of the use of technical data. Addresses the due process using the working panel. It is advocated that the methods be as explicit as possible. Can be adopted as a routine decision-making tool. There may be issues of relevance to priority setting. Implementation of the results is not explicitly considered. It has potential for interest groups to highjack the results using the marginal analysis panel. The transaction costs may be significant. It is very flexible and capable of being a valid and practical approach to priority setting. Recognises and addresses due process issues. As a framework, it has potential to be adopted as macro framework (ACE method or as macro marginal analysis), and is proven to work in a vertical program management.
The possible solution is the representation of the members of the Indigenous community in the panel so as to capture the community preferences in the SWF. This issue will be investigated in Chapter 8. Furthermore, the existence of the empirically determined criteria for priority setting can guide the panel members on how to make choices and trade-offs. This work was investigated in Chapter 5 and it was found that evidence and the impact of such evidence were considered important criteria for priority setting. To Indigenous health, elements of technical rigour and due process are critical for priority setting, and this constitutes one reason for the choice of PBMA approach to be contextualised in Indigenous health. It is also possible to complete PBMA exercise at a relatively low cost and with few data inputs, making it a useable tool in Indigenous health where data and financial issues
Page 150
exist. Overall, PBMA performance against the checklist developed in Section 6.3 was best overall compared to the rest of the models assessed here (Table 6.5). PBMA is capable of being a valid and practical approach to priority setting in Indigenous health especially because it makes use of evidence, upholds equity issues and can actively involve Indigenous Australians in the process. It is therefore a preferred approach to priority setting in Indigenous health.
6.6 Discussion The primary purpose of this chapter was to develop a set of guidelines for priority setting in Indigenous health, using the findings from Chapters 3-5. These guidelines were then used to qualitatively appraise a selection of priority setting approaches. It was found that noneconomic based frameworks would best be used in conjunction with the economic approaches. The PBMA framework, which combines technical and the due process, potentially addressed all the key elements of the checklist. However, being a western concept, PBMA may need to be contextualised in Indigenous health, a task that will be undertaken in Section C.
6.6.1 Guidelines for Priority Setting in Indigenous Health The key challenges from Chapters 3 to 5 involved how to translate the key principles identified into measurable and practical concepts, either as objectives for the health system or as criteria to guide the priority setting process. For example, in Chapter 3, it was explained that individuals aim to optimise their individual welfare under a set of axioms developed in welfare economics. However, optimising social welfare cannot be achieved by using the same axioms of choice, which are based on individual utility theory alone, because of Arrow’s impossibility theorem. Using the concept of ‘capabilities’ as a way in which social welfare could be computed, non-welfarist frameworks offered a solution to Arrow’s impossibility theorem in health care. This allowed non-utility based attributes, such as health, to be included in the SWF.
Page 151
Furthermore, what is being optimised needs to be clearly defined. The objectives of Indigenous health care need to be clearly articulated. Do the objectives include optimising or maximising health, improving access, ensuring cultural security of the services, or maximising Indigenous involvement in the decision-making process? Economic theory suggests that maximising health should be one of the key objectives, and social justice suggests equity, equality, fairness and needs-based allocation as the key objectives. These objectives should be clearly known and publicised so that any attempt at priority setting is as explicit as possible. The challenge is how to make the objectives as discrete as possible due to the potential for overlap in certain areas, especially in cases where more than one objective may be stated. For example, ‘equality of health’ overlaps with ‘equality of outcomes’ or ‘equality of use’, but differs significantly. The similarities and differences ought to be reflected in the approach used for the maximisation of health.
Maximising health gain from a given budget can be done using Cost Benefit Analysis, an approach which potentially addresses both technical and allocative efficiency. In this case, health gain would be estimated using the willingness to pay approach. However, this is contentious due to the potential for utility from the consumption of health care failing to capture the true value of health. The use of willingness to pay approach is further complicated by the fact that monetary valuation of life and death issues is rejected by some health economists (discussed in Section 3.4.2), because society is often reluctant to participate in such valuation (Tsuchiya and Williams 2001; Feldman and Serrano 2006). QALYs and DALYs, although considered as the most appropriate means of capturing benefit from health care by health economists, also fail to capture equity and other social justice issues in Indigenous health. To highlight this issue in the guidelines for priority setting, a separate set of questions related to the concept of benefit were listed. This issue will be investigated in Chapter 8.
DMA would appear to be the most appropriate framework for specifying the arguments in the SWF since the government is responsible for the biggest portion of the health budget. DMA, however, fails to explicitly recognise the role of the Indigenous community as the consumers and the role of the due process. In the case of ACCHS, with their management structure which involves Indigenous community members as the Board of Directors, DMA fits well. In the mainstream health care services, which Indigenous Australians are entitled
Page 152
to access, leaving the specification of the SWF to the decision-maker, without Indigenous representation, may not appropriately include Indigenous values and aspirations in the health objectives and may downgrade Indigenous self-determination and control of the priority setting process. For example, the Australian health care system systemically discriminates against Indigenous Australians (Henry and Mooney, 2002). This discrimination is in part due to constitutional and institutional failures to protect Indigenous Australians from discrimination. It is therefore important for further investigation on how DMA can be used with other approaches such as empirical ethics (Richardson 2000) and the ‘communitarian claims’ approach (Mooney, 2002) in Indigenous health.
The empirical and the communitarian approaches can assist to clarify the population values and preferences for making judgements. These approaches can also elicit preferences for weighting criteria or guidelines for their relative importance, which can then be used in making trade-offs. Publishing such criteria and the associated weights can allow the public to debate on the appropriateness of such criteria and the associated weights. It was not possible to elicit the weights of the criteria from the respondents in Chapter 5, so results used to develop the guidelines implicitly assumed that the weights all had the same relative importance. This could be a potential area for research in future.
The checklist developed above should therefore be treated with caution when assessing existing priority setting models. For example, the ethical principles in Section 6.3.3 may need to be investigated further to validate the findings from Chapters 3 and 5. These ethical principles not only guide the decision-makers in the choice of objectives to pursue, but also guide the value judgement exercise. They also guide the nature and conduct of the due process component of the priority setting exercise. For example, PBMA approach requires that the criteria for decision-making should be established before the priority setting exercise begins. These criteria would be used to inform the selection of the options for change, the weighting of the scores, and interpretation of the findings. Thus, the use of ethical values or principles which are not carefully inferred from the respondents, may not reflect the true values of the Indigenous Australians. The use of such criteria may not lead to desirable outcomes from the priority setting process, however ‘ideal’ the framework used may be.
Page 153
The ideal priority setting framework should be able to shift resources matched with health gain, and to be able to have an impact on the health of Indigenous Australians. Importantly, such a framework should potentially address Indigenous community needs and aspirations. Thus, any priority setting model in Indigenous health should recognise the following:
Past and present Indigenous disadvantages;
Existence of structural barriers which are embedded in the Australian social, economic and political system that perpetuate systemic discrimination;
Situations where the application of the same rules to unequal groups can generate unequal and unequitable results; and
Impartial justice concerns normative and policy analysis in economics and it forms a compelling foundation for an ethical theory.
Shifting resources matched by health gain requires the use of marginal analysis, and this involves the uptake of economic evaluation results. However, challenges to the uptake of economic evaluation results in Indigenous health have not been addressed, and the current level of uptake of economic evaluation evidence is not known. These challenges include:
Lack of data from quality economic evaluation studies in Indigenous health;
Lack of in-house expertise and knowledge in economic evaluation – more training of health economists is needed;
Lack of data and poor quality of the data from census and administrative datasets;
Lack of good evidence from health outcomes for policy purposes; and
Lack of general awareness and education about economic evaluation.
There is therefore a need for more investigation into the uptake of economic evaluation results. This task is indirectly addressed in Chapter 7, where PBMA potential is gauged from the survey of key decision-makers in Victoria.
6.6.2 Assessment of the Priority Setting Models The purpose of the assessment of the approaches to priority setting was to highlight the potential use of the guidelines that were developed for priority setting, and to demonstrate the credentials of PBMA. The approaches to priority setting that were assessed were a selection of the commonly used models from both non-economic and economic disciplines. The focus was on the approaches that had been already applied in Indigenous health
Page 154
(Beaver, Williams et al. 1999; Mitton 2001; McDermott 2004; McDermott and Mills 2004; Gruskin and Daniels 2008), such as PBMA, HBG/HRG, and the historically based approaches. Additionally, a selection of the dominant approaches to priority setting, such as the needsbased approaches and league tables (Mitton 2001; Segal and Chen 2001; Carter 2002; Mitton and Donaldson 2004; Segal and Mortimer 2006; Carter, Vos et al. 2008) was included. Another set of models that was included involved those approaches which were potentially useful in Indigenous health, either for their non-economic credentials, such as the accountability for reasonableness approach and human rights-based approach (Daniels 2000; Martin, Giacomini et al. 2002; Mitton and Donaldson 2004; Kapiriri, Norheim et al. 2007; Gruskin and Daniels 2008), or for being variants of the dominant approaches, such as the HSW (Segal and Chen 2001; Carter 2002; Segal and Mortimer 2006). These models were grouped as economic and non-economic based priority setting models and then assessed.
The assessment of non-economic based approaches demonstrated that these approaches have merit and can significantly contribute to the priority setting process in Indigenous health. The key contribution involves needs assessment, the due process and value judgement in priority setting. These approaches emphasise explicitness of the values and criteria that ought to guide priority setting, and the active involvement of the Indigenous stakeholders in the process. For example, the accountability for reasonableness approach in particular has potential to perform well in Indigenous health because it embraces human rights and other social justice issues addressed by other non-economic approaches, and has been widely used in Canada.
The key weakness of the non-economic approaches (Table 6.5) involves their failure to incorporate opportunity costs in the selection of interventions/options for change. These models also lack explicit decision rules, such as marginal analysis, for allocating resources between competing policy objectives given the scarcity of resources (Segal and Chen 2001; Carter 2002; Mitton and Donaldson 2004; Segal and Mortimer 2006). In conclusion, these models are best utilised in combination with economic models.
Economic-based frameworks on the other hand have potential to aid the decision-maker to shift resources matched with health gain because economic models recognise scarcity and employ marginal analysis as a decision criterion. Assessment of these models indicated that,
Page 155
although these models can aid the decision-makers in the identification of strategies that maximise health, they do not provide information on distributive and procedural justice. It may therefore be advisable to use these models in conjunction with broader frameworks which involve social justice (Mooney 1992; Carter 2002). Specifically, the Oregon experiment and the HSW model, which rank interventions based on cost per QALYs alone should be rejected, because social justice issues which are important in Indigenous health care are ignored. Furthermore, both approaches fail to provide a mechanism for specifying non-health and political issues in the concept of benefit to reflect objectives in Indigenous health. Thus, both approaches do not entrench community control in the priority setting process.
Table 6.5: Summary of the performance of the selected non-economic frameworks Criterion Well-defined question
Needs assessment/ options for change Ethical principles
Concept of benefit
Criteria for making judgements
Marginal analysis Technical rigour Due process and Indigenous involvement Relevance of the results Overall assessment
Performance The well-defined question is potentially achievable but varies from model to model. Some models take societal perspectives while others focus on disease and/or specific sectors. Needs assessment often involves quality of life studies and can potentially include the existence of evidence to support analysis. Criteria for selecting options are not always explicit and may involve voting. Ethical principles from social justice (such as equity, community control, cultural security and access) can potentially be taken into consideration when options for the selection of interventions are being considered. The concept of benefit in these approaches does not reflect the key objectives being pursued. It often ignores broader objectives, and does not provide a mechanism for discussion and clarification. These approaches generally fail to include explicit decision rules for making choices in the face of scarcity. The role of judgement is potentially achievable in some models, especially the consensus-based models, but not explicitly specified in others such as needs-based ones. Not utilised. Measurement rigour is achievable. The due process is achievable in some models. For example, in human rights-based approaches, the due process is considered but in needsbased approaches, the due process is ignored. Issues of concern to the decision-makers are achievable, except for efficiency issues. Consensus is achievable. Best utilised in combination with economic frameworks.
PBMA model appeared to address both economic and the social justice principles. It potentially provides a valid and practical way of applying economic principles behind the achievement of allocative and technical efficiency in priority setting, with due consideration for social justice. Its engagement of the key players through membership in the marginal
Page 156
analysis panel increases the likelihood of the acceptance of the recommendations by the Indigenous community. The emphasis that PBMA puts on the role of the panels offers a means for the Indigenous Australians to be effectively involved in the priority setting process – that is, to have community control.
However, the process of local decision-making can potentially be captured by vested interests (Mortimer and Segal, 2006). The selection of the models for assessment did not follow any systematic procedure, so potentially more useful models in Indigenous health could have not been assessed. Furthermore, the results do not rank the performance of the models against the criteria, and they do not provide comprehensive assessment of the models.
PBMA employs both technical rigour and due process in its analysis of options for change in priority setting. Technical rigour involves standard economic evaluation with option to rigorously assess the concept of benefit to include Indigenous values, principles and aspirations. The due process component allows for the inclusion of Indigenous Australians and other stakeholders in the priority setting process, and the mechanism of deliberative consensus. However, PBMA is a western concept and needs to be Aboriginalised, the task undertaken in the next section.
Page 157
Part C: Contextualising Indigenous Health
PBMA
in
7. Chapter 7: PBMA Appraisal 7.1 Introduction In the previous section (Section B) of this thesis, guidelines for priority setting in Indigenous health were developed and used to assess a selection of priority setting models. The PBMA framework was chosen to be the most appropriate tool to be used in Indigenous health due to its economic credentials and its ability to incorporate social justice issues in the priority setting process. However, PBMA is a western concept and needs to be contextualised in Indigenous health. In this section (Section C), PBMA is contextualised in Indigenous health using four empirical studies, reported in Chapters 7 and 8. Lessons from these projects will be synthesised and used to develop a checklist for guiding the successful application of The PBMA framework in Indigenous health. This checklist, reported in Chapter 9, is designed to be used in ACCHS.
In this chapter, results from three empirical studies are reported: the survey of decisionmakers; an expenditure study in regional Victoria; and the costing of dry areas intervention. This chapter starts with the description of PBMA and its applications, and followed by the discussion on the survey findings from the survey of key decision-makers in Victoria. Next, an expenditure study to demonstrate issues related to the estimation of the program budget in PBMA will be discussed. The ACE approach is then briefly described, followed by of the costing of the alcohol dry areas intervention as part of the ACE Prevention project. Implications from the three projects on PBMA are then discussed. The issues for discussion involve PBMA’s potential to aid decision-making, and issues for consideration in estimating the program budget, and the costing of options.
Page 158
7.2 The Program Budgeting & Marginal Analysis Approach 7.2.1 Description The PBMA framework is a pragmatic, explicit and output-oriented economic approach for priority setting which also has the potential to uphold equity principles and process/participatory issues associated with priority setting. The framework has been fully described in the literature (Mooney, Gerard et al. 1992; Peacock and Edwards 1997). Essentially, it is a combination of two techniques: program budgeting and marginal analysis. The first technique is output oriented and involves the establishment of the program budgeting to provide an overview of the pattern of expenditure and its distribution in the organisation in question (Earl-Slater 1999; Culyer 2005). This involves describing the current pattern of spending or resource use by a particular population or specific disease. The second part is the marginal analysis54. Marginal analysis involves examining potential changes to the current mix of services, referred to here as options for change, provided using conventional economic evaluation as a means to improve the efficiency of programs (Donaldson 1995; Earl-Slater 1999; Culyer 2005). It provides a practical way to achieve allocative efficiency by utilising the best available costing and outcomes evidence in an explicit and systematic planning process, referred to in this thesis as ‘technical process’. This evidence may be sourced from the published literature or the local databases to ‘model’ the probable future streams of changes in costs and benefits of the programs under review. Where routine data is poor, primary data collection may be required or expert advice (subjective data) can be utilised while acknowledging the deficiencies in the data (Henderson and Scott 2001).
Assessing options for change using marginal analysis normally involves examining incremental gains and incremental costs of changes in what is spent through comparisons across and within programs (Donaldson 1995). For example, within a program, the potential for shifting resources between two types of services for the population, such as cervical
54
Marginal analysis is an economic technique that looks at the incremental costs and consequences of potential changes across and within programs that might enable the organisation to better meet its objectives.
Page 159
cancer versus breast cancer within a cancer program or within women’s health, can be examined. As well as across programs, marginal analysis can be used to look at whether a shift of resources from a particular program would assist the organisation realise ‘value for money,' such as investments in mothers’ health as opposed to mental health programs.
At the core of PBMA approach is a working group or expert panel, which is charged with making recommendations for resource re-allocation. The process can be supported by the range of 'hard' and 'soft' evidence, and by the clearly defined and weighted decision-making criteria, such as from the health constitution, on which decisions can be based. The ‘hard’ and ‘soft’ evidence may include economic evaluation evidence from the published and/or unpublished literature, and other evidence, including expert advice on sustainability, equity and affordability.
The relevant criteria may be locally or nationally relevant-based and can be adopted from the literature or from the funding agency guidelines. These criteria can also be empirically determined through direct elicitation from the decision-makers or the community as reported in Chapter 3. Such criteria may need to include a method of defining what constitutes benefit, undertaken in Chapter 9, from the services being assessed and the trade-offs between those benefits. Regardless of the source of the decision-making criteria, it is important that the criteria be specified before the marginal exercise commences. In summary, most PBMA studies involve seven key steps (Figure 7.1) and make use of the expert panel or management in its process as means to ensure that the due process takes place (Mitton and Donaldson 2004).
PBMA technique is capable of providing both a valid and practical approach to priority setting. PBMA is perhaps best understood as a pragmatic tool, which feeds into a complex decision-making process (Peacock, 1998b). It utilises best available evidence in an explicit and systematic planning process. In practice, however, the acceptable quality of evidence depends largely on the research question, the setting and the stakeholders. Most PBMA studies have relied on local subjective judgements for assessing the effectiveness of options for change, and this has had the potential to undermine PBMA credibility (Coast and Donovan 1996; Posnett and Street 1996; Peacock 1998; Segal and Chen 2001).
Page 160
Table 7.1: Stages in PBMA priority setting exercise 1. Determine the aim and scope of the priority setting exercise 2. Compile a program budget, being a map of current activity and expenditure 3. Form marginal analysis working group 4. Determine relevant decision-making criteria a. Decision-maker’s input b. Board of Director’s input c. Public input 5. Advisory panel to identify options in terms of: a. Areas for service growth b. Areas for resource release through producing same level of output (or outcomes) but with less resources c. Areas for resource release through scaling back or stopping some services 6. Advisory panel to make recommendations in terms of: a. Funding growth areas with new resources b. Decisions to move resources from 5(b) to 5(a) c. Trade-off decisions to move resources from 5(c) to 5(a) if relative value in 5(c) is greater than in 5(a) 7. Validity checks with additional stakeholders and final decisions to inform budget planning process Source: Priority Setting Toolkit (Mitton and Donaldson 2004)
Furthermore, PBMA has the ability to address process and other non-process issues, referred to as ‘due process’, in priority setting in Indigenous health. It does this by empowering local decision-making in identifying, measuring and valuing benefits from options identified to address the stated objectives. These issues may include Indigenous community control, to ensue acceptability, and cultural security of the services, in the design and delivery of the interventions. This can be done through the involvement of Indigenous Australians in the decision-making process.
7.2.2 PBMA Assessment in the Literature PBMA has its origins in the attempts by the United States government to apply systems theory (using the Planning Programme Budgeting Systems – PPBS approach) to public policies in the 1950s (Wohlstetter and Mann, 1957) cited in (Lockett, Raftery et al. 1995). The PPBS approach involved identification of the objectives, definition of time horizons, development of indicators for program efficiency, and development and comparison of alternative means or programs of achieving the stated objectives. In 1965, it was expanded
Page 161
to be applied to all United States federal agencies and later assessed as a failure (Wildavsky, 1975, cited in (Lockett, Raftery et al. 1995). In particular, the PPBS was found not to be a cost-effective approach to priority setting. Other problems included the lack of principles for defining programs, lack of a single rationality which could be used for decision-making, the difficulty in linking inputs to outputs in public services, and its mission of politics. Several other assessments both in defence and critical of the PPBS approach took place. In response to the criticisms, Gavin Mooney introduced the options appraisal to the PPBS approach and renamed it PBMA (Mooney, Gerard et al. 1992; Lockett, Raftery et al. 1995). Since then, PBMA has had several modifications to try to address its limitations. These modifications have included variants of PBMA such as ACE approach discussed in Section 7.5, macro marginal analysis (Mitton, Donaldson et al. 2003; Mitton, Patten et al. 2003) and specific aspects of PBMA being improved to address specific contexts (Peacock 1998; Peacock, Richardson et al. 2007).
Most developments on PBMA over the years have focused on better understanding the decision-makers’ objectives/utility functions and the understanding the political and institutional constraints that decision-makers face in priority setting (Jan 2000). Some researchers have sought to improve the evidence base (Carter, Stone et al. 2000) and others have sought to improve the level of rigour and the options appraisal (Peacock, Richardson et al. 1997; Edwards, Peacock et al. 1998; Peacock 1998) so that multiple objectives can be specified and a single benefit score used. These developments have enabled the PBMA framework to be used as a decision-making tool in several settings.
PBMA has been widely used to inform strategic decision-making in Canada, New Zealand and the United Kingdom to decide on investments/disinvestments in health services (Cohen 1995; Ratcliffe, Donaldson et al. 1996; Mitton and Donaldson 2001 Oct). International experience with PBMA over the last three decades would indicate that this approach has the potential to make substantial improvements to the priority setting process (Mitton and Donaldson 2004).
In Australia, the PBMA framework has been applied by several State/Territory governments, and the non-government sector such as the Cancer Council and the National Tobacco Control Program (Carter 2000; Carter and al. 2000; Mitton and Donaldson 2004; Carter, Vos
Page 162
et al. 2008). In NSW Health, PBMA exercise was done in almost 13 area health services in the areas of dental health services, mental health services, diabetes, dentistry, asthma, family services, etc (Newberry 1996). WA and NT have each undertaken PBMA exercises (Mitton and Donaldson 2004). In Victoria, PBMA has been applied to several disease-specific programs such as heart disease, mental health, obesity, and prevention of noncommunicable disease programs, as part of ACE approach (Carter 2000; Haby, Carter et al. 2004; Magnus, Carr et al. 2005; Vos, Haby et al. 2005; Haby, Vos et al. 2006; Carter, Moodie et al. 2009). In SA, a PBMA exercise was conducted under the auspices of the South Australian Health Commission in five hospitals and several community health services between 1996 and 1998 (Peacock and Edwards 1997; Peacock, Richardson et al. 1997; Edwards, Peacock et al. 1998; Peacock 1998; Peacock, Richardson et al. 2007).
The PBMA framework has also been trialled in Indigenous Australian health sectors. The SA community health sector PBMA application involved an Indigenous component and it was considered a resounding success (Peacock 1998). Two Torres Strait Islander health services undertook demonstration studies which were not evaluated (Mitton and Donaldson 2001 Oct). Another PBMA application was undertaken in an ACCHS in Sydney and was well received by the organisation (Jan 2002). In general, published assessments of the impact of PBMA studies on decision-making are still few but growing (Peacock, Richardson et al. 1997; Carter 2000; Carter, Stone et al. 2000; Mitton, Donaldson et al. 2000; Viney, Haas et al. 2000; Haby, Carter et al. 2004; Magnus, Carr et al. 2005; Vos, Haby et al. 2005; Haby, Vos et al. 2006; Peacock, Richardson et al. 2007; Carter, Vos et al. 2008; Moodie, Haby et al. 2008; Carter, Moodie et al. 2009; Mihalopoulos, Harris et al. 2009; Moodie, Haby et al. 2009).
The published assessments of PBMA cite lack of measurement rigour, inadequacies in option selection, narrow perspectives and poorly developed marginal analysis as weaknesses of PBMA (Segal and Chen 2001). Some of these criticisms are resolvable within PBMA approach, as evidenced by recent developments (Peacock, Richardson et al. 1997; Carter, Stone et al. 2000; Peacock, Richardson et al. 2007; Carter, Vos et al. 2008; Carter, Moodie et al. 2009). However, there are challenges in designing and applying PBMA or its variants to Indigenous health. Possible challenges may include the following: 1. Acceptability of the framework in Indigenous health; 2. Availability and the quality of data;
Page 163
3. The challenge of forming the program budget; 4. Challenges of disinvesting in certain programs; 5. Challenge of measuring and valuing benefits; 6. Conducting PBMA may require trained staff and resources; 7. Determining the membership of the panel could be a challenge; 8. How to involve Indigenous Australians in the process; and 9. Integrating PBMA into the management culture and systems. (Jan 1998; Peacock 1998; Mitton 2001; Mitton and Donaldson 2004).
One key challenge is that the PBMA frameworks are based on western values and processes which may not be entirely appropriate to Indigenous health. Acceptability of PBMA approach as tool for routine decision-making therefore needs to be ascertained within Indigenous health. There is potential for community backlash if certain services are contracted or deleted because they are not cost-effective or inefficient. The Indigenous community might find it difficult to see why a service should be deleted when there are many people in their community with significant health care needs, however ineffective the deleted service may be. The challenge is to educate and involve the Indigenous community in the process of PBMA studies.
For the marginal analysis component of PBMA, evidence on both costs and benefits is required, yet the Indigenous data and its quality in both the census and administrative centres are very poor (AHMAC & AIHW, 1997; ABS and AIHW, 2003). For example, data on acute health care is often readily available but finding data on services such as community health and other primary care services may prove very difficult. This is because most community and primary health care services do not have Indigenous identifiers that would help estimate the costs attributable to Indigenous Australians. This affects both the costing of interventions or services for economic evaluation purposes and the estimates of the program budget in PBMA study, investigated in Section 7.6. The challenge is how to estimate the program budgets. This issue is investigated in Section 7.4 and guidance on how to estimate the program budget in Indigenous health is developed.
Closely related to this is the issue of effectiveness of interventions/programs, due to the absence of data and the poor quality of the available data/evidence. This is mainly because
Page 164
there are very few clinical trials specifically done in Indigenous health (Morris 1998; Mitton 2001; Sibthorpe, Bailie et al. 2002). Ineffectiveness is a common issue when the interventions/programs have not been reviewed to ascertain their effectiveness, when there is not enough solid evidence to demonstrate its effectiveness or a much more effective program has come to light. Ineffectiveness55 may also be viewed from an epidemiological perspective and not from the ‘capacity to benefit’ perspective and this would not be fair to Indigenous Australians or would be plainly a wrong concept (Wiseman and Mooney 1998; Mooney 2000). Furthermore, effective programs are likely to end up being ineffective when significant access and cost issues exist or when institutional racism exists in the system as in the case of Indigenous Australians (Henry, Houston et al. 2004). It is possible to model uncertainty attributed to the effectiveness of the interventions.
The challenge here involves the decision-maker investigating other issues not just costs and benefits to aid decision-making; otherwise, the decision-maker can delete/contract a program that is actually needed when there may not be any alternative in sight. The concept of benefit often used in mainstream health, may not be appropriate to use in evaluating Indigenous health. For example, the DALY ignores community benefits and so it could be categorised as a ‘sickness’ measure of benefit. The challenge here would be how to measure and value non-DALY benefits from interventions in Indigenous health, a task attempted in Chapter 8.
Another issue in the challenge of disinvesting in certain programs within the organisation itself is how much of the program to scale down, and whether the scaling down necessarily translates to efficiency savings (Lockett, Raftery et al. 1995). There could be a possibility of resistance by the program managers since disinvesting may means that their funding goes to their colleagues or it is a sign of failure on their part. One respondent revealed in an interview that "... we come from a scarcity mentality... where you protect your resources... you don't share those resources. And I think that's a challenge" (Dionne, Mitton et al. 2008).
55
Effectiveness from an epidemiological perspective may be computed using the following general conceptual equation: Effectiveness = (% of those who accept to participate in the intervention once offered) x (sensitivity of the intervention) x (adherence to intervention) x (efficacy) (Marciosek et al, 2001) – methods for priority setting among clinical preventive services.
Page 165
Contracting certain services within an organisation implies that the released funding may have to be transferred to another a manager in the organisation. This may be perceived as indirectly suggesting that the losing manager has not done a good job.
One of the key components of the PBMA framework is the marginal analysis panel. The key issues here involve how to determine the membership of the panel and how to ensure that the panel members effectively participate in the priority setting process. In some communities, determining panel membership may prove difficult since each community member may want to be on the panel. Furthermore, the lack of guidelines or processes that ensure full participation of the members may prove challenging, since certain members may have their own agendas. These issues are investigated in Chapter 8 and the method to ensure full participation of the members in the process is trialled.
Another challenge involves the resource intensity of conducting PBMA. Conducting PBMA may require trained staff and additional resources, which may not be readily available in Indigenous community organisations. Most community organisations already face manpower problems, limited resources and competing health needs/demands for resources (Dwyer, O'Donnell et al. 2009). This would affect the success of integrating PBMA into the management culture and systems. In summary, four key empirical studies, a survey of decision-makers on PBMA potential; a survey of expenditure on services for Indigenous Australians; costing of dry areas; and the concept of benefit, will be undertaken to investigate how some of the above key issues can be addressed in order to contextualise PBMA in Indigenous health. The first three projects are reported here and the fourth project is reported in Chapter 8.
Page 166
7.3 PBMA Appraisal: Survey of Decision-makers The purpose of this study involved ascertaining what the decision-makers in Indigenous health think of the PBMA framework in terms of its potential to meet their expectations, its acceptability and its chances of being adopted as a routine decision-making tool. Specifically, it aimed to identify the key features that ought to be addressed so as to aid its potential adoption for routine decision-making in Indigenous health. This survey was administered as part of the survey that was reported in Chapter 5. Please refer to the full description of the methods in Chapter 5.
7.3.1 Methods The study involved the survey of key decision-makers in Victoria using face-to-face interviews (see Appendix 5 for the letters of invitation and data collection instruments). The key decision-makers in Victoria involved in Indigenous-specific health care services were recruited through the snowballing process. The respondents included bureaucrats from the Victorian DHS who are involved in Indigenous-specific health care services, and the senior executives in ACCHS in Victoria. About 50 decision-makers from both the DHS and ACCHS participated in the survey. Whilst the respondents were not a randomised sample, it was a comprehensive and seasonally complete sample. About 66% of the potential respondents participated in the survey.
The PBMA framework was introduced to the respondents and following hypothetical questions were asked: 1. Do you think PBMA has the potential to aid in priority setting? 2. What specific potential challenges do you foresee with PBMA? 3. Would expanding benefits from services to include non-health benefits improve the priority setting process? 4. What criteria would you use to evaluate the success of PBMA? The survey instrument involved both closed- and open-ended questions, administered faceto-face interviews. The survey results were analysed using content analysis and are reported here. Results for each question are presented as the number of respondents who
Page 167
mentioned it and reported as numbers and/or a percentage. In some cases, the respondents were asked to rank their responses for the relative importance of each issue mentioned.
7.3.2 Results: Potential for Using PBMA in Priority Setting The PBMA framework was well received by the respondents as the potential means to improve the priority setting process in Indigenous health. The number of respondents from the DHS who felt that PBMA was acceptable as a routine decision-making tool was 69% – slightly lower than the number of respondents from ACCHS. Nearly 80% of ACCHS respondents thought that PBMA was intuitively appealing and would most probably be an acceptable approach to priority setting in their organisations (Table 7.2). ACCHS respondents identified with PBMA approach and most them were excited to learn that there existed a formal approach to priority setting which was similar to the way business is done in the Indigenous community organisations. Such responses demonstrated that the PBMA framework would be acceptable to the Indigenous organisations and therefore more likely to be adopted. The CEO of VACCHO said: “I cannot believe that someone else somewhere around the world formalised the way we do things in our community organisations! We emphasise collective and community input in our decision-making processes and consciously undertake program budgets and juggle the funding and amount of activities to be implemented given our meagre resources.”
The number of respondents who were familiar with PBMA was the same (17) between the two samples – ACCHS and DHS respondents (Table 7.2). An attempt to unpack the level of knowledge of the respondents on PBMA approach was not undertaken. Subsequent responses from other questions on marginal analysis indicated that most respondents – particularly those from ACCHS – were familiar with the program budgeting aspect of the approach but not the marginal analysis component.
Table 7.2:PBMA potential PBMA potential PBMA acceptability PBMA knowledge Explicit costs and benefits Concept of benefits
DHS responses (Yes) (18) 69% (17) 65% (16) 62% (18) 69%
Page 168
ACCHS responses (Yes) (19) 79% (17) 71% (16) 67% (21) 88%
Respondents were then advised about one of the strengths of PBMA – its potential to be used even in the absence of data for technical analysis. In such cases, it was a way of thinking and doing things in an organisation that focuses people on comparing, in an explicit sense, the marginal costs and marginal benefits of a particular action. When all the respondents were asked if this was how things were done in their respective organisations, whether decisions were explicitly compared using marginal costs and marginal benefits, 67% of ACCHS respondents and 62% of the DHS respondents responded ‘yes’ (Table 7.2). This signified that PBMA would potentially provide a means to demonstrate what kind of data an organisation would need and what for, providing a platform for a transition from the lack of data to the need for data.
PBMA’s ability to allow for the concept of benefit in the outcome measure to be expanded to include other possible non-medical dimensions of health benefit, such as community benefit or cultural security, was presented. Again, respondents were asked if expanding the concept of benefit would in their opinion potentially improve the priority setting process. Eighty-eight per cent of ACCHS respondents answered ‘yes’ compared to the 69% from the DHS (Table 7.2). This question is further explored in Chapter 8 using the options appraisal method to elicit the views of the decision-makers on what the dimensions of benefit should be for priority setting.
One of the key issues addressed in this study involved eliciting the potential criteria for assessing the performance of PBMA if it were to be implemented in an Indigenous health care setting. This information was meant to inform the development of guidelines/checklist of PBMA implementation in Indigenous health. Results indicated that the key criteria for assessing PBMA potential in Indigenous health involve its ability to shift resources matched by health gain. This serves as confirmation for the need for the marginal analysis – an economic approach to priority setting. Seventy-five per cent of the respondents from ACCHS considered the “ability of PBMA to shift resources matched with health gain” as one of the key criterion for evaluating whether the implementation of PBMA within a specific program was a success (Table 7.3). Other criteria identified by ACCHS included the ease of use of the approach (58%), its ability to ensure long-term increased health benefit (54%), and the buy-
Page 169
in by the funders (46%). It could be deduced that ACCHS decision-makers strive to maximise the improvement in the health of their constituents with what resources that they have.
Table 7.3: Assessment criteria for PBMA DHS Shifting of resources matched with health gain Easy for managers to use Long-term increased health benefit Buy-in by funders
Responses (n=16, 62%)
Continued use within the organisation Buy-in by medical staff/nursing and uptake by other organisations
(n=6, 23%)
(n=15, 58%) (n=14, 54%) (n=11, 42%)
(n=3, 12%)
ACCHS Shifting of resources matched with health gain Buy-in by funders Easy for managers to use
Responses (n=18, 75%)
Long-term increased health benefit Continued use within the organisation Buy-in by medical staff/nursing
(n=11, 46%)
(n=14, 58%) (n=13, 54%);
(n=8, 33%) (n=7, 29%)
Just like ACCHS respondents, the majority of the DHS respondents (62%) identified the ability of PBMA to shift resources matched with health gain as the key criterion for evaluating whether the implementation of PBMA within a specific program was a success (Table 7.3). The other criteria were also similar to ACCHS responses and included the following: easy for managers to use (n=15, 58%); long-term increased health benefit (n=14, 54%); and buy-in by funders (n=11, 42%). The reason why the ‘buy-in by the funders’ (although the respondents were not asked why) featured among the DHS respondents could mean that the DHS was keen for the OATSIH (its partner funder of ACCHS) to support ACCHS’s efforts to use of PBMA.
In order to get an idea for the potential of piloting PBMA approach, respondents were asked to identify what areas in their organisations that would benefit most from PBMA application. Almost all the respondents identified disease and/or program areas. This suggested that PBMA budgeting aspect would benefit most from taking the disease continuum perspective. However, they foresaw several specific challenges with PBMA approach in their organisations or departments. First and foremost, both the DHS and ACCHS respondents either combined (68%) or individually (62% of ACCHS; 65% of the DHS respondents) thought that PBMA would probably not be understood in their organisations (Table 7.4). Secondly, the respondents also expressed concern that the model would probably need more data, which is already scarce. Furthermore, the respondents felt that
Page 170
PBMA could easily prove to be more work for the decision-makers who were already busy with many other tasks, especially in ACCHS. These issues would make it difficult to adopt PBMA given the already tight budgets. Further, when the respondents were asked if there was any other model they would rather see being used if it was not PBMA, needs assessment was the preferred option especially by ACCHS. Identifying the respondents’ fallback position, the needs assessment approach, which tends to prioritise problems but not the solutions, further demonstrates the need for the organisations to adopt PBMA so as to shift their focus from the problems to the solutions.
Table 7.4: Potential challenges for PBMA PBMA challenges DHS Not well understood in the organisation Won’t have data Extra resources required
Responses (n=17, 65%)
Responses (n=17, 71%)
(n=8, 31%)
PBMA challenges ACCHS Not well understood in the organisation Extra resources required Will be more work for decisionmakers Won’t have data
Resources cannot be readily moved Will be more work for decision-makers Lack of timelines
(n=6, 23%)
Lack of timelines
(n=5, 21%)
(n=5, 19%)
Resources cannot be readily moved
(n=4, 17%)
(n=12, 46%) (n=9, 35%)
(n=9, 38%) (n=8, 33%) (n=8, 33%)
7.3.3 Discussion the PBMA framework was intuitively appealing to all the respondents, although most of the respondents were not familiar with the marginal analysis component of PBMA. Adoption of PBMA is potentially feasible and would therefore enable the decision-makers to build upon the existing approaches similar to the program budgeting, which they are already familiar with. This would ensure that the acceptance and the probable adoption of PBMA as a priority setting tool would be increased. Indeed all the respondents (from both the DHS and ACCHS) reported that PBMA would be acceptable in their organisations/departments (Table 7.5). ACCHS respondents particularly identified with PBMA and judged the approach to be acceptable due to its ability to address the issue of the program budget.
Page 171
Table 7.5: PBMA potential to all respondents PBMA knowledge PBMA acceptability Explicit costs and benefits Concept of benefits
All the responses (Yes) (34) 68% (37) 74% (32) 64% (39) 78%
Respondents from the DHS revealed that there is no single coherent funding approach to services for Indigenous Australians in the DHS. This reflected the disparate elements amalgamated into one Department. The health element, which covers health, mental health and senior Australians, has since been curved out of the DHS to form a new Department of Health. This has left the community services and housing making up the new DHS. The funding mechanisms for Indigenous-specific health care services were either based on throughput, such as Casemix funding approach for hospital services, inputs such as number of employees in population health services, or population numbers such as in home and community care (HACC) services. Thus, the DHS funding mechanisms may need to be revised in order to be outcomes based. This would ensure that the funding frameworks are compatible with PBMA priority setting framework which is outcomes based. This might prove to be the key obstacle to the adoption and continued use of the PBMA framework in the DHS.
The other important attribute of PBMA that was appealing to the decision-makers, especially those from ACCHS, was its flexibility with respect to the concept of benefit. It offered decision-makers an opportunity to use a formal approach (options appraisal) to measure other diverse benefits, such as equity, cultural security and access, from the activities that their organisations undertake, which cannot be captured by the traditional or conventional health economics outcomes measures such as DALYs or QALYs. The willingness to pay approach can potentially capture measure these activities and expressed in monetary terms, a challenge that is not discussed here. the PBMA framework also has the potential to offer ACCHS the means to formally report their organisational activities to the funding agencies in terms of outputs or outcomes, as opposed to the existing reporting mechanism which involves using inputs. The outcome reporting mechanism would allow the organisations to demonstrate the program impact on the health of their constituents, and thus encourage funders to review their funding mechanisms to reflect outcomes.
Page 172
The use of hypothetical questions in the interview design was an important limitation of the study, since the respondents had not applied the PBMA framework in their workplaces. This meant that the respondents were limited in their ability to appropriately assess PBMA. Further, the respondents would not understand the components of PBMA, weaknesses and possible strengths. However, explanations were provided to clarify concepts, key issues and there was every indication that participants understood the questions after explanations. The interviewees gave responses which were accompanied by explanations that demonstrated their understanding of the questions.
Explanation of economic concepts and other issues was required with ACCHS interviewees and it seemed unlikely that relevant responses would have resulted had these questionnaires been administered without face-to-face meetings. Despite every effort being made to ensure that the explanations were not leading the respondents to specific responses, it is possible that some respondents could have been led to pick certain responses and hence there existed the potential for some interview bias. On average, 66% of the potential participants participated in the interviews. Most respondents who declined to be interviewed cited ‘little involvement in the priority setting process’ or ‘not being qualified to participate in the interviews’ being the reason for declining to participate. From ACCHS, those who failed to participate in the interviews cited lack of time to participate.
7.3.4 Summary Findings from this survey have implications for decision-makers involved in Indigenous health because it presents a list of potential criteria for priority setting. These findings can be used to generate debate on the appropriateness of any existing criteria for priority setting in Indigenous health and could be used as benchmark against which current practices can be assessed. These findings also have implications for further research in the development of guidelines for priority setting.
PBMA is a pragmatic, explicit and outcome-oriented economic approach for priority setting, and has the potential to uphold equity principles and process/participatory issues associated with priority setting. Several PBMA exercises have been conducted in the United Kingdom, Canada, New Zealand and Australia (Mitton and Donaldson 2001 Oct). Differences
Page 173
have emerged in the level of rigour in the following areas: selection of options; consultation or bargaining process; measurement of costs and benefits; and in the level of evidence demanded (Segal and Chen 2001; Carter 2002). PBMA also allows the concept of benefit to be related to program objectives and the local context. Furthermore, it allows the key economic principles of marginal analysis and opportunity cost to be applied in a practical way (Mooney, Gerard et al. 1992).
Criticisms levelled against the PBMA frameworks include: lack of evidence base, lack of measurement rigour, inadequacies in option selection, game playing and the danger of ‘vested interests’ taking over the process, narrow perspective and poorly developed marginal analysis (Segal 2001; Segal and Chen 2001). In response to such criticisms, several researchers have used the PBMA framework to develop priority setting methods that addressed the stated weaknesses and this has led to the emergence of different evaluation approaches within PBMA’s overarching framework to suit specific levels of decision-making. One such approach is ACE approach (Carter, Vos et al. 2008; Carter, Moodie et al. 2009) which is discussed in Section 7.5.
Page 174
7.4 Expenditure on Services for Indigenous Australians 7.4.1 Introduction This study demonstrated how the program budget in the PBMA framework can be undertaken in Indigenous health. Such a program budget is normally part of a regional PBMA study. The study provided the pattern of expenditure in the region and distribution of the resources (Earl-Slater 1999; Culyer 2005). It involved describing the current pattern of spending or resource use by Indigenous Australians in the Loddon Mallee Region of Victoria. The methods of estimating the expenditure and key issues addressed contributed to the establishment of a framework within which consistent information on program budgeting for priority setting can be generated routinely by both spending and funding agencies. This study also served to highlight the Indigenous Australians’ levels of access to and utilisation of publicly funded health care services, such as hospital, medical, pharmaceutical and community health services, as measured by expenditures. It is one of the purposes of the program budget in PBMA, to highlight where the current pattern of expenditure in the region.
This study was undertaken at the request of the Loddon Mallee Region (LMR) Indigenous Reference Group, a peak Indigenous community reference group comprising six Indigenous co-operatives (Bendigo, Mildura, Echuca, Robinvale, Swan Hill and Kerang) in regional Victoria. Such baseline data are essential for any informed discussion of the adequacy of services and the setting of priorities for future expansion through either the State or Commonwealth funding authorities.
The Indigenous population of the LMR was recorded at 3,850 people in the 2001 Census. That was 13.8% out of the Victorian Indigenous population of 27,846. Both the Loddon LMR and Victorian figures were significantly higher than had previously been estimated. Indigenous birth rates are much higher than for non-Indigenous people and there also appears to have been significant under-identification in the past. It is therefore difficult to project the numbers in 2004. The resident population was estimated at about 3,950 in that year, an increase of 2.6% over the 2001 Census figure.
Page 175
The LMR is geographically one of the largest of the Victorian government’s health regions. It covers approximately one-quarter of the State from the Macedon ranges in mid-Victoria to the Murray River borders with NSW and SA. Its residential population was 280,447 in 2001, and estimated about 283,400 in 2004. However some NSW residents use health services on the Victorian side of the border, and for Indigenous Australians, the main such group come from Euston in south-western NSW. Adding these to the Victorian numbers gives an estimated total of about 4,000 Indigenous Australians served by the mainstream health services of the region. The region’s geographic spread raises issues of access to health services, transport and cost that are particularly important for low-income people, which includes the majority of the Indigenous population. The LMR population is also more mobile than elsewhere in Victoria, both within the region, intrastate and interstate (AIHW 2001; AIHW 2005; AIHW 2008).
As in other regions, the Indigenous component has a much more youthful age profile than for non-Indigenous people (in the LMR, 58.8% of all Indigenous Australians are aged 0-24 years compared with 36.3% of the non-Indigenous population) and they are underrepresented as clients of the mainstream services in health and education, particularly those that rely on people referring themselves (DHS 1998).
There are five ACCHS in the region. They are at Mildura, Robinvale, Swan Hill, Echuca and Bendigo, although the Swan Hill service was not operating in 2004 and the Robinvale service had no doctor at the time of the survey. There are 18 hospitals in the LMR. Four of them have accident and emergency departments and one also provides some scheduled outpatient services. The rest of the hospitals are categorised as either level C, D or E, which mainly add only allied and community health programs to their in-patient activities. Ambulatory patients are seen by the local GPs. In most of these hospitals, specialist services are not run by the hospitals but by visiting specialists from centres such as the Bendigo Health Care Group. Mental health services for the region are provided by Bendigo Health Care Group and Mildura Base Hospital. At the time of the survey, there were 257 GPs with practice addresses in the region: five of them worked in Indigenous Health Services and one had a close association with an ACCHS. The remainder were in independent private practice. There were 71 pharmacies operating at the time of the survey.
Page 176
7.4.2 Methods The study used the ‘National Accounts’ method used in the AIHW Health Expenditure Australia series, which is based on the Organisation for Economic Cooperation and Development’s (OECD) System of Health Accounts (OECD 2000), to estimate expenditure on services for Indigenous Australians in regional Victoria. Health expenditure includes all expenditures on goods and services that have the main objective of improving or maintaining health. All GPs, ACCHS, pharmacies, hospitals and community health services in the region were asked to participate (see Appendix 4 for the letters of invitation and data collection forms). The medical and pharmaceutical surveys were conducted through mail questionnaires which sought information on all services or items provided to Indigenous patients or clients over a three-week period in April-May 2004. The period was selected to minimise the impact of winter illnesses. The hospital and community health service data were collected somewhat later, from September-October 2004, but also at a time when use was less likely to have been affected by obvious seasonal factors.
As far as participation of GPs was concerned, assistance of the local Divisions of General Practice was sought through their Victorian umbrella organisation and of the Victorian branch of the Pharmacy Guild of Australia. The Divisions and the Guild agreed to encourage their members to participate and to forward and collect the questionnaires on our behalf. Non-members were contacted with the assistance of the Divisions and Pharmacy Board of Victoria.
For the pharmacies, these arrangements achieved a participation rate of 70%. One pharmacy that has a “special relationship” with ACCHS in an area with a number of Indigenous Australians did not respond but it was possible to obtain sufficient information on dispensing from ACCHS itself. This relationship involved the pharmacy providing free medicines to all Aboriginal patients and then billing the ACCHS for the co-payments. It was possible to use the information sent by the pharmacy to the ACCHS to estimate Aboriginal use of prescription medicine. For GPs, the numerical response rate was only 38% – still higher than the 29% rate achieved in the most recent BEACH survey – but because all of the doctors in ACCHS participated, as well as those doctors who were known to have an association with them, the coverage of significant providers is believed to have been much
Page 177
greater. Details are shown below, together with the methodology used to estimate hospital and community health service use.
7.4.3 Results 7.4.3.1 The GP survey
Table 7.6 shows, by GP division, the number of doctors surveyed, the numbers responding, response rates and the multipliers used to expand the service data to an estimated total for the region. The multiplier or expansion factor was derived from the participation rates to represent 100% participation rate. It also included a factor to use to expand the 3 week survey period into a 12 month period. For the GP data, the relevant expansion factors for participation were as shown earlier in Table 7.6. They were applied separately to each category of doctor and each of the areas (Divisions) in which they practised. Five doctors worked in or for the Indigenous health services and one was covering for ACCHS doctor pending a new appointment. The remainder were private although there were several who saw more Indigenous patients than the average, about five to 10 patients per week compared with two to three or none, and had a relationship with the local ACCHS. Table 7.7 shows the number of GP visits over a three-week period, the associated MBS values, and the average cost per visit.
Table 7.6: GPs with addresses in the LMR: Response rates and survey expansion factors, by type of practice and Division Division/Type of practice Private GPs Bendigo Western Victoria Central Highlands Murray Plains Mallee Sub-total AMS doctors Total
Total 88 12 41 59 51 251 6 257
Respondents 32 4 18 16 22 92 6 98
Page 178
Response rate (%) 36 33 44 27 43 37 100 38
Expansion factor 2.78 3.03 2.27 3.70 2.33 2.70 1.00 2.63
Table 7.7: Data from respondent GPs, three weeks, April-May, 2004 Practices Private ACCHS Total
Visits (no)
Medicare benefit value ($) Per GP Visit ($) 132 4,944 286 8,976 418 13,920
37.45 31.38 33.30
7.4.3.2 Pharmacies
Table 7.8 shows the number of pharmacies surveyed, the number of respondents, the response rate and the multiple used to expand the survey data. Table 7.9 shows the number of prescribed items dispensed over a three-week period, their value at PBS levels or private script payments, the average cost per item and the factor used to expand the survey data into regional estimates.
Table 7.8: Responses to the pharmacy survey, April-May 2004 Total (no) 71
Respondents (no) 50
Response rate (%) 70
Expansion factor (multiple) 1.43
Table 7.9: Items dispensed by respondent pharmacies, three weeks, April-May, 2004 Category PBS items Non-PBS Total
Items (no) PBS benefits ($) Patient cost ($) Total cost ($) Per item ($) 523 12,776 2,140 14,916 28.52 33 498 498 15.09 556 12,776 2,638 15,414 27.73
The PBS covered 94% of all dispensed items. From the survey data, nearly 82% of these were in the ‘concessional’ category, 12% were free under the safety net and just under 5% were classified as ‘general’. The average PBS cost per item was almost exactly the national average for patients in the concessional group. Patients paid, on average, $4.09 per PBS item. Payments for non-PBS scripts were much higher but the numbers were small.
In addition, aggregate information was available for another pharmacy with Indigenous dispensing worth about $22,300 for the year, or $1,287 in a three-week period. In the final estimates, its dispensing was added to the expanded figures for the responding pharmacies. The composition of its work was assumed to be the same.
Page 179
7.4.3.3 Other services
The survey was introduced and outlined during the visits to each hospital and community health service in the region. All hospitals in indicated their willingness to participate in the survey, except for one. Of these that responded, only seven reported ever seeing Indigenous patients and conducted a formal survey of patient separations over a threeweek period. The Mildura Base Hospital and the Bendigo Health Care Group also cooperated by testing the proposition that identification would be higher if the question on Aboriginality was asked at discharge rather than admission (Britt, Miller et al. 2003).
However, for the hospitals, data from the hospital records as collected by the DHS in the Victorian Admitted Patient Database (VAED) and Victorian Emergency Minimum Database (VEMD) rather than the surveys, were used for this study. The DHS pays a substantial loading for admitted Indigenous patients. In 2004, it was 10% over the normal payment per Weighted Inlier Equivalent Separation (WIES), rising to 30% in the next financial year. The hospitals argued that they had a strong incentive to identify all Indigenous peoples and that their reported data were therefore accurate. As will be shown, that conclusion may be questionable even for inpatients and although the hospitals were asked to take particular care in identifying emergency cases and outpatients over the survey period, some underidentification in those areas was very likely.
7.4.3.3.1 Admitted patients
Table 7.10 shows the estimated annual figures for admitted patients. The methodology involved costing inpatient admissions at their individual case WIES costs plus the DHS loading of 10% for the estimated additional cost of greater co-morbidity in Indigenous patients. Despite the hospitals’ confidence, an expansion factor of 10% was then applied. The basis for this was the result from Mildura Base Hospital. As mentioned earlier, Mildura Base Hospital agreed to test a proposition developed from the ABS, AIHW study of 12 hospitals nationally in 1998 which showed that although Indigenous identification rates varied considerably, they were on average about 11% higher when the question of Aboriginality was asked at discharge rather than on admission. The Mildura Base Hospital
Page 180
followed the same procedures and then matched the responses for patients admitted over a three-week period. The result was an increase of just under 10% in the number of Indigenous patients, very similar to that in the national study. The DHS suggested a 25% expansion but that figure appears to have been drawn from the 1998-99 Health Expenditure report (Report I) and has no documented support. 7.4.3.3.2 Emergency and outpatients
Emergency and outpatient attendances were valued at the average costs per visit reported by the hospitals in the region to the DHS. Again, only two of the four hospitals with emergency departments participated in a formal three-week survey. However, comparison of their survey results with the information contained in the VEMD gave almost complete agreement, so using the central record system for the others was justified. The same 10% expansion factor as for in-patients was applied on the ground that the factors which might inhibit identification at the time of patient presentation should apply equally to both admitted and non-admitted patients. That may well be an underestimation though, because the hospitals did not have the same monetary incentives to identify Indigenous patients in emergency departments as they did for inpatients. As mentioned earlier, only one hospital reported any scheduled outpatient services and its data were derived by the expansion of a three-week survey. The reported figures are shown in Tables 7.10 and 7.11, and adjusted estimates for an assumed 10% under-identification are shown in Tables 7.12 and 7.13.
Table 7.10: Reported in-patient separations, LMR, 2003-04 (including 10% WIES loading for Aboriginality in Victoria) Category Same day Multi-day Total
Indigenous Non-Indigenous Total (no) Indigenous WIES Value (no) (no) (no) ($000) 1,583 36,646 38,229 252 795.1 847 39,689 40,536 1,023 3,228.8 2,430 76,335 78,765 1,275 4,023.9
Table 7.11: Reported emergency and outpatient attendances, LMR, 2003-04 Category Emergency Outpatient Total
Indigenous (no)
Non-Indigenous (no)
Total (no)
3,376 247
72,209 na
75,585 na
4.47 Na
471.3 28.3
3,623
na
na
Na
499.6
Page 181
% Indigenous
Value ($000)
Table 7.12: Estimated admitted patients (adjusted for under-identification): Total separations; WIES Value, LMR, 2003-04. Category Same day Multi-day Total
Indigenous (no)
Non-Indigenous Total (no) Indigenous Value (no) WIES (no) ($000) 1,741 36,488 38,229 277 873.9 932 39,604 40,536 1,125 3,549.4 2,673 76,092 78,765 1,402 4,423.3
Table 7.13: Emergency and out-patient departments, estimated total visits, LMR, 2003-04 Category Indigenous (no) Non-Indigenous (no) Total (no) % Indigenous Value ($000) Emergency 3,714 71,871 75,585 4.91 518.3 Out-patient 272 na na Na 31.1 Total 3,986 na na Na 549.4
Total admissions for Indigenous Australians were 3.4% of the regional total. That was about 2.45 times the admission rate for non-Indigenous residents. In cost-weighted separations (WIES numbers), the Indigenous share of total WIES was just under 3% (1,402 out of 47,031). It was 2.1 times the Indigenous percentage of the population, although the average cost weight for Indigenous multi-day admissions was significantly higher than for other people. A very high proportion (64%) of Indigenous admissions was mainly for relatively inexpensive same-day treatments, particularly dialysis, compared with only 48% of nonIndigenous admissions in the same-day category.
For emergency attendances, the estimated Indigenous rate was 0.929 visits per person per year, 3.6 times the non-Indigenous average of 0.25 per person. The mix of emergency services was not significantly different to that for non-Indigenous people.
7.4.3.3.3 Community health services Community health services comprise a wide range of allied health, prevention, promotion, health education and community-based mental health care services delivered in a variety of settings, including hospitals and free-standing community health centres. As in the national expenditure figures, they include all of the non-doctor activities of ACCHS. ACCHS expenditures were adjusted to exclude the estimated proportion of non-Indigenous people attending ACCHS – about 8% on average. The national health expenditure reports have also
Page 182
excluded some of their outlays as being primarily on welfare, rather than health, but this distinction was not made in this study.
Excluding ACCHS, there were seven community health services that reported seeing Indigenous patients and ran separate community health programs, mainly in the area of drugs and alcohol. Six of the seven participated in a three-week survey of their patients in September-October 2004 and their data have been expanded to provide some full year estimates. Table 7.14 shows the results. As can be seen, ACCHS were by far the largest providers – over 80% of the identified total. However, the other figures are probably understated. As pointed out earlier, many of the small hospitals (which did not participate) included elements of community health, mental health and aged care in their services that they were unable to separate or attribute to Indigenous patients. The extent of the underestimation is unknown, although the organisations involved were relatively small. Table 7.14: Community health services, by type of service, estimated visits and costs, 2003-2004 Services through Hospitals Health centres ACCHS School dental services Total
Visits (no)
Cost ($000) 11,116 1,048 na 418 na
354.2 38.7 2,707.3 53.1 3,153.3
7.4.3.4 Expansion of the Survey Data
All of the GP and pharmacy data that were derived from three-week surveys required expansion for two factors – participation rates and the survey period. For the GP data, the relevant expansion factors for participation were as shown earlier in Table 7.6. They were applied separately to each category of doctor and each of the areas (Divisions) in which they practised. For responding pharmacies, a single multiple was applied (1.43 or 100/70) and the annual data from the additional pharmacy was then added. All of the adjusted figures were then multiplied by 17.33 to expand them from three-week data to a 52-week year. The expanded estimates of services volumes and costs are shown in Tables 7.15 and 7.16.
Page 183
Table 7.15: Estimated GP services, LMR, 2003-04, by source of service and Medicare benefit cost Services by: ACCHS Private GPs Total
Three-week survey (no) 286 132 418
Estimated total ($) 8,976 4,944 13,920
(no) 4,956 5,926 10,882
($000) 177.4 226.6 404.0
Table 7.16: Estimated prescriptions dispensed, LMR, 2003-04, by PBS category, PBS benefits and total cost
PBS items Non-PBS Total
Items (no) PBS benefits ($000) Patient cost ($000) Total cost ($000) 13,720 335.1 56.1 391.2 862 13.1 13.1 14,582 335.1 69.2 404.3
7.4.3.5 Estimated Total Expenditures
Table 7.17 shows estimated total expenditures on hospital, medical, pharmaceutical and community health services provided to Indigenous people in the LMR in 2004, in total and as expenditures per person.
Table 7.17: Estimated expenditures by type by service, total and as averages per Indigenous person, LMR, 2003-04 Service Hospital – in-patients – emergency/out-patients GP visits Prescribed drugs ACCHS Other Total Services covered
Total ($000) 4,424 548 404 404 2,708 444 8,932
%
Per person($) 49.53 6.14 4.52 4.52 30.32 4.97 100.0
1,106 137 101 101 677 111 2,233
7.4.4 Discussion This study aimed to demonstrate how the program budget in the regional the PBMA framework can be undertaken in Indigenous health. Such a program budget is normally part of a regional PBMA study. It identified the services utilised by Indigenous Australians and estimated the levels of service use, provided consistent information on the regional program budget for priority setting purposes, and highlighted the current pattern of
Page 184
expenditure in the region. The program budget was reported as expenditure estimates per person in the region.
As in all studies of spending on health services for Indigenous and Torres Strait Islander people, the main features of the LMR’s Indigenous population were, first, its reliance on publicly provided services and, secondly, its disproportionate use of hospitals as both admitted patients and outpatients. Expenditures on public hospital inpatient care were nearly 50% of the estimated total. Per person, the figures for Indigenous people were just over twice the non-Indigenous average and spending on emergency treatment was over three and a half times the non-Indigenous figure for the region. The overall difference in inhospital care was, however, not that great. The least expenditure was in GP and prescribed medicines, points of interest for ACCHS, for priority setting purposes. This sort of information would be useful for the program budget in PBMA.
Three issues should be made about the expenditure study. Firstly, even for the services covered, these figures did not measure all health expenditures on Indigenous Australians living in the LMR. They related only to those services provided within the region. Patients referred to the metropolitan teaching and specialist hospitals were therefore excluded – for example, children admitted to the Royal Children’s Hospital. It would be possible to ascertain and add the costs incurred by Indigenous residents of the LMR in hospitals outside the region but it would not be particularly relevant to the project’s purpose. In PBMA exercise, it would be important for the scope of the exercise to be set in order to determine what services to include and which ones not to. In other words, PBMA task needs to be achievable.
Secondly, the coverage of services was limited. Only those that provide health care directly to individuals were included and not all of those were covered. Specialist medical care services outside the hospitals were not included; nor were nursing homes, patient transport and a range of public health and state-wide community based programs which are administered centrally and for which there are no direct indicators of Indigenous use. PBMA, being a pragmatic tool for priority setting, can be selective on what is achievable and where data may not be collectable, expert advice may be utilised.
Page 185
Finally, with one small exception, costs covered public outlays only. Any spending on private hospitals was therefore excluded together with over-the-counter medicines, private dentistry and privately provided ancillary services. GP visits were valued at the Medicare benefit level on the assumption that patient charges were minimal but information was available on patient payments for both PBS prescribed drugs and private prescriptions. However, the amounts involved were small and some of the charges may have been met by ACCHS. In the 2001-02 national expenditure estimates, the services covered accounted for about 78% of all the expenditures attributed to Indigenous people.
However, a few key issues are unique to Indigenous health. One such issue is the availability of utilisation data for primary health care. The data is generally poor since Indigenous identification in the datasets is not compulsory (AIHW 2009). Analysts will have to be fill in the data gaps through primary data collection (AIHW 2004). Secondly, the analyst has to be conscious of the extra costs associated with the co-payment that ACCHS pay pharmacists on behalf of Aboriginal patients. Thirdly, ACCHS incur significant costs in transporting and following up of the Aboriginal patients, whose medical appointments may be with mainstream services. Such costs can easily be missed in an analysis.”
In summary, the expenditure estimates have been prepared from data collected directly from health care providers in the LMR. The same methodology as in the Australia-wide Indigenous health expenditure studies was used and they are broadly comparable. The key question here was how these results could inform PBMA application in Indigenous health. The key answer is in the methods, which involves identifying the type of service provided, by whom and for how long the service took to be delivered, and tracking who is likely to meet the costs. The hospital sector had good data systems with reasonable Indigenous identifiers. Other services did not have Indigenous identifiers which necessitated running surveys of the services provided. These methods can easily be replicated by ACCHS to estimate the program budget for PBMA. Where such surveys are not possible to run, expert advice can be utilised to estimate expenditure on such services.
Page 186
7.5 Assessing Cost Effectiveness (ACE) Approach ACE method, a variant of PBMA, is an economic approach to priority setting that combines rigorous technical analysis, in the form of economic evaluation and epidemiological modelling, and due process in priority setting by involving stakeholders in a working group. The decision context for ACE approach is efficiency of health care services and the best bundle of services that offer value for money given budgetary constraints (Carter, Vos et al. 2008; Carter, Moodie et al. 2009). Like its parent framework, PBMA, ACE approach uses marginal analysis to maximise benefit from interventions. They both use explicit criteria for the selection of options for change, and for making trade-offs. Furthermore, both approaches use marginal advisory panels as a means to allow value judgement in the analysis. The key difference between ACE method and PBMA is that ACE method does not include the program budget in its analysis, although due consideration for the level of budget to spend is taken into consideration, whereas PBMA has a program budget.
The key steps of ACE approach include the following: 1. Specification of the research question 2. Creation of the working group 3. Selection of interventions 4. Confirmation of the technical methods 5. Undertaking technical analysis and 2nd stage filters 6. Agreement on the findings and then dissemination. (Vos, Carter et al. 2010).
The essence of ACE approach is to clearly define and maximise ‘benefit’ from the interventions used in order to realise the objectives being pursued with the resources available. Like its parent framework, PBMA, ACE approach attempts to realise the maximisation of benefit through the application of marginal analysis. Such analysis essentially asks: ‘If resources are taken from program X and given to program Y, will the net benefit be higher or lower?’ If the net benefit is higher, there is an economic case for making that change.
Page 187
The approach uses ACE working group, which involves stakeholders recruited from the community, policy-makers, topic experts and clinicians/providers, to undertake a range of tasks. They are often involved in the selection of interventions, methodological issues, and value judgements, such as the interpretation of data and findings. Their value judgements ensure that broader considerations that are less amenable to quantification are taken into account. The working group also contributes to the due process, ensuring that stake holder and policy issues are duly discussed and there is consensus in the decision-making process (Carter, Vos et al. 2008; Carter, Moodie et al. 2009).
The ACE method involves two key steps in the assessment of benefit: examining options for change through the use of individual cost-effectiveness using DALY; and then subjecting those options to second-stage filters such as: capacity of the intervention to reduce inequities or equity56, feasibility and sustainability of the intervention, cultural safety, strength of evidence57, and acceptability to stakeholders58 to assess other important benefit issues (Carter, Vos et al. 2008; Carter, Moodie et al. 2009). This stage involves stating, explicitly, the level of reliance on judgement and the notion of due process using a Steering Committee. Alternatively, options are available using decision theory for the more formal integration into an index score, investigated in Chapter 8. The ACE approach is used to cost the dry areas intervention in Indigenous homelands (see Section 7.6) in this chapter as means of contextualising PBMA in Indigenous health.
56
Equity may be viewed as inequality in the distribution of health outcomes/problems or inequality in access/utilisation of health care by SES, gender, rural/remote/urban residence and/or ethnic origin. 57 Strength of evidence is due to the nature of the evidence for different kinds of health interventions being inevitably variable. If the evidence on effectiveness is weak, there would be great uncertainty as to the size of the impact measure entered into the simulation modelling of uncertainty. 58 Acceptability of the intervention refers to the anticipated acceptability of proposed interventions to the various stakeholders affected by the intervention (i.e. patients, the general community, third-party funders, health service providers, and politicians and their advisors). This may require a judgement call from the Indigenous Steering Committee as to what may be acceptable to policy-makers. Under the criterion of feasibility, issues such as the availability of appropriate expertise to implement the intervention on a national scale, the potential size of the financial commitment, and the time scale for implementation, are considered. As with ‘acceptability’, it is a criterion that involves judgement rather than technical decision rules. For this criterion, the Indigenous Steering Committee has also restricted itself to flagging issues that require attention and to presenting descriptive information to assist policy-makers.
Page 188
7.6 Costing Dry Areas Alcohol Intervention This analysis involves ascertaining the costs and cost off-sets of the dry area intervention to reduce alcohol consumption in Indigenous Australian communities for the purposes of economic evaluation. These costs will be used, in combination with the concept of benefit information that will be derived in Chapter 8, in the marginal analysis exercise of PBMA. Issues of identification, measurement and valuation of costs for economic evaluation purposes in PBMA will be the focus of the discussion. How to obtain costs data when such data is not readily available, will be highlighted, and used to inform the contextualisation of PBMA in Indigenous health.
7.6.1 Background The background information on the alcohol use in Indigenous Australian communities indicates that, in 2004-05, nearly half of all Indigenous adults (49%) reported having consumed alcohol in the week prior to the interview. In the same year, about 16% reported drinking at risky/high-risk levels in the long term, and this was similar to that for nonIndigenous adults (ABS 2006). Furthermore, the Indigenous drinking at risky/high-risk levels was statistically similar between remote (15%) and non-remote (17%) Indigenous Australians (ABS 2006).
Rates of alcohol consumption are significantly higher in the NT than in the rest of Australia, with adults consuming alcohol at a rate 53% above the national average in 2004/05 (Whetton, Hancock et al. 2009). Indigenous consumption at the same time was estimated to have been 16.9 litres (or 14.2% more than non-Indigenous consumption) and nonIndigenous consumption 14.5 litres (Skov et al. 2009 cited by Whettoon, et al 2009).
7.6.1.1 Burden of Disease
Vos and colleagues estimated that alcohol was linked to 7% of all deaths in Indigenous Australians and 6% of the total burden of disease in 2003 (Vos, Barker et al. 2007). Alcohol
Page 189
abuse was associated with 3.4% of all Australian deaths in 1996 and 4.8% of the total burden of disease (Mathers 2002).
In 2001-02, total expenditure on alcohol and drug services by State and Territory governments for Indigenous people was $48.2 million or 20% of the total expenditure (AIHW 2005). In terms of per capita expenditure, this translated into about 10 times the amount spent on other Australians. How much of that proportion could be attributed to alcohol specifically is not clear and these figures do not include all expenditure undertaken in community health services, most notably alcohol and drug education.
Hazardous alcohol use is associated with adverse health and social consequences, and it is a major risk factor for (ABS and AIHW 2008):
Health problems such as cardio-vascular disease, mental disorders, foetal alcohol syndrome or foetal alcohol spectrum disorder, liver disease, diabetes, and some types of cancers;
Injury in the form of car accidents, falls, burns, and suicide;
Domestic violence, family breakdowns, child abuse, and anti-social behaviour;
Financial problems, unemployment, absenteeism and low productivity at work; and
Community breakdown, theft or crime.
This implies that the prevention of alcohol abuse would be desirable based on the associated size of the health problem. Indeed, preventing alcohol abuse in Indigenous communities featured as a high priority area in the 2003 National Strategic Framework and was agreed to by the Australian Health Ministers’ Advisory Council (NATSIHC 2003). It is also identified as a priority program area for research by the CRCATH – a public good research organisation that brings together the Indigenous, academic and industry community. Preventing alcohol abuse in Indigenous Australian communities takes different forms and a summary of these interventions are presented in Table 7.18 (Gray, Saggers et al. 1995).
Table 7.18: List of interventions in alcohol abuse Health promotion (demand reduction) Alcohol education, changing behaviour
Policy, legislative and enforcement interventions (supply reduction) Fiscal policy (e.g. taxation), drink-driving legislation, advertising controls,
Page 190
Harm reduction Night patrols to reduce alcohol-related conflict and
through communication, training and public awareness
restricted areas of alcohol use, liquor licensing legislations
harm, sobering-up shelters (a detoxification program)
7.6.1.2 Restricted Areas Interventions
This study focused on ‘restricted’ areas in indigenous communities (as opposed to private homes) as a strategy to reduce hazardous alcohol consumption. These interventions aim to reduce alcohol consumption behaviour and associated harm by forbidding or restricting the supply and consumption of alcohol within certain areas (d'Abbs 1990) and may embody all or some of the following elements:
It may be statutory, with little or no community involvement, such as the SA model, the WA model, and/or a mix of statutory and community control (NT model);
Delivery of the intervention is by the local communities, government, or both;
Enforcement is by police, community enforcement agencies, liquor licensing authorities, or all of the above;
They target areas of excessive alcohol drinking;
Restrictions may be continuous or for certain periods;
May include all or certain types of liquor;
May or may not exclude certain people in the area;
May or may not be initiated and/or supported by the local community; and
May range from a total ban to imposition of per capita limit of alcohol consumption.
Restricted areas interventions can be grouped into alcohol free zones and dry areas. In alcohol free zones, the consumption of alcohol is prohibited in a particular zone, creating a liquor free area such as a public park, entertainment venue or sports ground. The prohibition may be confined to certain events or time of the day. In contrast, dry areas (sometimes referred to as ‘dry communities’) necessitate an entire area or community to be declared alcohol free, that is, no consumption, possession or intoxication unless an individual has a license of exemption. Some limited benefits have been found from banning alcohol altogether from isolated indigenous communities (d'Abbs 1990; Bourbon, Sahggers et al. 1999; Berman, Hull et al. 2000; Berman and Hull 2001).
Page 191
The dry areas intervention limits the availability of alcohol in Indigenous lands. It is a major approach to reducing availability of alcohol beverages in areas where Indigenous Australians have obtained land rights and State or Territory legislation has been passed which empowers them, or an external body on their behalf, to limit the availability of alcoholic beverages on their lands. Currently, dry areas have existed in SA, WA and the NT since 1978, although each of these States has a slightly different dry areas model implemented (d'Abbs 1990). These include: the ‘Statutory Model’ (the common model in SA); the ‘Community Controlled Model’ (as it appears in WA); and the ‘NT Complementary Control Model’ (which is a mix of statutory and community-controlled approaches and is prevalent in the NT). The implication for the choice of the model involves:
Community control issues; and
Enforcement issues.
The statutory model is based on little or no community involvement in the declaration process for the dry area. The State government empowers local councils to apply to the Liquor Licensing Commission (LLC) to declare particular places within their jurisdictions ‘dry’. Importantly, the LLC may declare an area to be dry with or without Indigenous community involvement, since it is not obliged to consult the Indigenous community. The policing is the responsibility of the State government and a number of public places in SA , such as Port Augusta and Ceduna, in which Indigenous people formerly met to drink have been declared dry (d'Abbs 1990). This could have implications for racism and Indigenous selfdetermination and community control.
The community-controlled model involves minimal or no statutory involvement in the implementation, monitoring and enforcement of the intervention. In this model, councils in WA make use of the Indigenous Communities Act 1979 (WA), which empowers them to pass by-laws to control the supply and consumption of alcohol. The council can declare an area in their jurisdiction to be a dry area through an absolute majority of council members. It is then up to the council to decide how best to police the policy. The policing is through the community justice system, comprising Indigenous members of the community appointed as Justice of the Peace, bench clerks, probation officers and rangers. State police are only involved when the by-laws have been breached. These breaches are punishable by a fine of up to $100 or up to three months imprisonment.
Page 192
The NT Complementary Control model is a combination of the statutory and community controlled models. In this model, residents of the community or local government may apply to LLC to have an area or home declared dry. The Chairman of the LLC is obliged to hold a public meeting in the community to elicit public views. If there is enough community support for such restrictions then the chairman passes them into law. The LLC may (depending on the circumstances), however, declare the community dry even if there is little community support for the application.
It is evident that each of the above models entails different approaches to the degree of Indigenous community involvement. Importantly, the National Strategic Framework states that community control must constitute a significant component in the challenge of addressing Indigenous health (NATSIHC 2003). This would imply that a dry area model should involve the Indigenous community when it is being enacted to ensure community control59. Furthermore, the enforcement of the intervention should probably be the responsibility of the State/Territory government to ensure that enforcement is sustainable, but working closely with the community justice system to ensure acceptability (d'Abbs 1990). The NT model (referred to here as the ‘dry areas model’) best fitted the above requirements and was therefore the recommended model for economic costing by the ACE Prevention project ISC.
7.6.1.3 The Dry Areas Intervention
The available Australian study of the dry areas model containing both quantitative and qualitative evidence is an evaluation of dry area legislation in the NT (d'Abbs 1987). The d’Abbs (1987) study found that the prevalence of alcohol consumption in those communities which had declared themselves dry fell significantly (p5 YEARS
Prior to this, have you worked in any other organisations or in the health sector more broadly? [ ] WORKED IN ANOTHER ORGANISATION (Specify:
)
[ ] WORKED IN HEALTH SECTOR (Specify: [ ] NO (Specify where from:
) )
What is the official title of your current position? ________________________________
How long have you been in your current position? _______________________________ [ ] 0-6 MONTHS [ ] 7-12 MONTHS [ ] 1-2 YEARS
Page 318
[ ] 2-5 YEARS [ ]>5 YEARS
In
this
position,
what
is
your
primary
role
in
the
organisation?
___________________________________________________________________________
What is your specific role in terms of setting priorities in the organisation? ___________________________________________________________________________ How many people in your organisation are involved in setting priorities? ___________________________________________________________________________
What is your specific role in terms of allocating resources in the organisation? ________________________________________________________________________
Setting Priorities In answering these questions, please state your own opinion on the current priority setting process. We are trying to find out the criteria/values/basis for decision-making in your organisation, what sources of data are used, how resources are allocated and what constitutes ‘benefit’ from services funded/provided by your organisation. There are no right or wrong answers.
What do you consider to be the key objectives of this organization? (Please state your own opinion, not necessarily those objectives which appear in the organisation’s documents) ______________________________________________________________________ [ ] ACCEPTABILITY OF PROGRAMS TO STAKEHOLDERS [ ] EFFECTIVENESS [ ] FEASIBILITY [ ] EFFICIENCY [ ] EQUITY ISSUES [ ] AFFORDABILITY (Is the service/intervention affordable?) [ ] IMPACT (of services on the health of Aboriginal people) [ ] SUSTAINABILITY (Is the service/intervention sustainable over a long time)
Page 319
[ ] ACCESS TO SERVICES ISSUES
What sources of information are used in determining short-term priorities (i.e. < 1 year) in this organisation?... note that in a moment I will ask you about longer-term priorities. [ ] EVIDENCE FROM ECONOMIC EVALUATION/EPIDEMOLOGICAL/OTHER RESEARCH STUDIES (Specify) [ ] KEY ORGANISATION OBJECTIVES (Which ones:
)
[ ] OTHER ORGANISATION REPORTS (Specify:
)
[ ] MEDICAL OPINION (Specify:
)
[ ] NEEDS ASSESSMENT (Specify:
)
[ ] STATE OR FEDERAL POLICY DOCUMENTS (Specify:
)
[ ] MAINSTREAM PUBLIC OPINION [ ] ABORIGINAL COMMUNITY [ ] THE 3 YEAR BUSINESS PLAN (How used:
)
[ ] FUNDS AVAILABLE/FINANCIAL DIRECTIVES [ ] OTHER (Specify) ____________________________________________________
What sources of information are used in determining longer-term priorities (i.e. > 1 year) in this organisation? [ ] EVIDENCE FROM ECONOMIC EVALUATION/EPIDEMIOLOGICAL OR OTHER RESEARCH STUDIES (Specify:
)
[ ] KEY ORGANISATION OBJECTIVES (Which ones:
)
[ ] OTHER ORGANISATION REPORTS (Specify:
)
[ ] MEDICAL OPINION (Specify:
)
[ ] NEEDS ASSESSMENT (Specify:
)
[ ] STATE OR FEDERAL POLICY DOCUMENTS (Specify:
)
[ ] MAINSTREAM PUBLIC OPINION [ ] ABORIGINAL COMMUNITY [ ] THE 3 YEAR BUSSINESS PLAN (How used:
)
[ ] FUNDS AVAILABLE/FINANCIAL DIRECTIVES [ ] OTHER (Specify:
)
Page 320
Once priorities are defined, how are decisions made to divide up the resources across programs?.....i.e. on what basis are decisions made in determining how much money goes where? [ ] SIZE OF THE HEALTH BURDEN [ ] ACCEPTABILITY OF PROGRAMS TO STAKEHOLDERS [ ] FEASIBILITY/SUSTAINABILITY ISSUES [ ] HISTORICAL TRENDS/PATTERNS [ ] EFFICIENCY * + POLITICAL ‘HOT SPOTS’ [ ] EQUITY ISSUES [ ] ACCESS TO SERVICESS ISSUES [ ] TOP DOWN DECISION-MAKING (i.e. influence from the board of directors for the organisation) [ ] OTHER POLITICAL INFLUENCE * + WHOEVER ‘YELLS LOUDEST’ [ ] OTHER (Specify) ___________________________________________________
Once resources are broadly divided up, on what basis are decisions made about the amount of activity to provide within each program? [ ] SIZE OF THE HEALTH BURDEN [ ] ACCEPTABILITY OF PROGRAMS TO STAKEHOLDERS [ ] FEASIBILITY/SUSTAINABILITY ISSUES [ ] HISTORICAL TRENDS/PATTERNS [ ] EFFICIENCY * + POLITICAL ‘HOT SPOTS’ [ ] EQUITY ISSUES [ ] ACCESS TO SERVICESS ISSUES [ ] TOP DOWN DECISION-MAKING (i.e. influence from the board of directors for the organisation) [ ] OTHER POLITICAL INFLUENCE * + WHOEVER ‘YELLS LOUDEST’ [ ] OTHER (Specify) ___________________________________________________________
Page 321
How is the ‘benefit’ from programs/interventions viewed in your organisation (what constitutes benefit from a program/service to Aboriginal people when resources are being allocated)? [ ] Benefits to individuals in the Aboriginal community [ ] Benefits to the Aboriginal community as a whole [ ] Equity [ ] Culturally secure services [ ] Others (specify) _____________________________________________________________________ _____________________________________________________________________
Feedback on Current Process In this section we would like your personal assessment of the priority setting process in terms of the strengths and challenges of the current process employed in your organisation. Again, there are no right or wrong answers.
In your opinion, does the current process of allocating resources across and within Aboriginal-specific health programs work well?. .that is, do you have a sense that, objectives of your organisation are well achieved? For example, if efficiency is an objective, that the benefit to the Aboriginal population is being maximized? [ ] WELL_______________________________________________________________ [ ] NOT WELL______________________________________________________________ [ ] OTHER __________________________________________________________________
Thinking in terms of allocating resources across different programs (e.g. alcohol, diabetes, maternal health, etc), what are the specific strengths of the current approach?
[ ] SIMPLE TO DO AND FAST [ ] GOOD MECHANISM FOR CONSTRAINING COSTS [ ] UNTHREATENING
Page 322
[ ] EQUITABLE [ ] EFFICIENT [ ] INVOLVES STAKEHOLDERS IN THE PRIORITY SETTING PROCESS [ ] TRANSPARENT [ ] WIDELY ACCEPTED/SUPPORTED WITHIN THE ORGANISATION [ ] WIDELY ACCEPTED/SUPPORTED IN THE ABORIGINAL COMMUNITY [ ] BASED ON BEST AVAILABLE EVIDENCE [ ] OTHER (Specify) _________________________________________________________
Thinking in terms of allocating resources within specific programs (eg diabetes), what are the specific strengths of the current approach?
[ ] SIMPLE TO DO AND FAST [ ] GOOD MECHANISM FOR CONSTRAINING COSTS [ ] UNTHREATENING [ ] EQUITABLE [ ] EFFICIENT [ ] INVOLVES STAKEHOLDERS IN THE PRIORITY SETTING PROCESS [ ] TRANSPARENT [ ] WIDELY ACCEPTED/SUPPORTED WITHIN THE ORGANISATION [ ] WIDELY ACCEPTED/SUPPORTED IN THE ABORIGINAL COMMUNITY [ ] BASED ON BEST AVAILABLE EVIDENCE [ ] OTHER (Specify) ___________________________________________________________
What are the challenges of the current approach? (and note. I will ask you about how the process could potentially be improved very shortly ...) [ ] ARBITRARY [ ] CRISIS ORIENTED [ ] IMPLICIT DECISION-MAKING (OBJECTIVES AND CRITERIA ARE NOT CLEAR – NOT TRANSPARENT) [ ] INDIVIDUAL INTERESTS PERVADE [ ] NO/LITTLE PUBLIC/STAKEHOLDER INVOLVEMENT
Page 323
[ ] NOT EVIDENCE-BASED [ ] NOT NEEDS BASED (OUTCOMES NOT CONSIDERED) [ ] NOT SYSTEMATIC [ ] NO FRAMEWORK (LARGELY ANECDOTAL) [ ] OVER RELIANCE ON PROVIDERS [ ] POLITICAL INFLUENCE TOO GREAT [ ] FEE FOR SERVICE (INCENTIVES NOT RIGHT) [ ] NOT EFFICIENT [ ] IGNORES ABORIGINAL VALUES [ ] INEQUITABLE [ ] LACK OF RESOURCES [ ] LACK OF GOOD QUALITY DATA [ ] LACK OF TIME [ ] LACK OF EXPERTISE [ ] OTHER (Specify) ___________________________________________________________
For a given year, once the allocation of resources within and across program(s) is set, can this be readily changed? [ ] YES____________________________________________________________________ [ ] NO _________________________________________________________________ [ ] OTHER ___________________________________________________________________
Part II: Improving the priority setting process In this section, we are interested in how you think the priority setting process can be improved, particularly if you think there is a place for an economic approach to priority setting. There are no right or wrong answers.
In your opinion, how could the current process of setting priorities and allocating resources be improved? If more than one, would you rank them? (e.g. if you chose 3 things, rank them from 1 to 3) [ ] MORE EXPLICIT/TRANSPARENT (Objectives are clear) [ ] SYSTEMATIC
Page 324
[ ] BETTER DATA/STRENGTH OF EVIDENCE [ ] LONGER-TERM VIEW TAKEN [ ] LESS P0LITICAL INFLUENCE [ ] MECHANISM FOR SHIFTING RESOURCES ACROSS PROGRAMS REQUIRED [ ] EXAMINING THE MARGIN, NOT TOTAL NEED [ ] OTHER (Specify) _________________________________________________________
Do you think that with the current process there are challenges to using evidence from economic evaluation (such as evidence from cost-effectiveness; cost benefit; value for money; or costing studies?) [ ] NO [ ] YES (Specify some) __________________________________________________
How could challenges from question 23 above be addressed? [ ] MORE EDUCATION ON THE CREDIBILITY OF ECONOMIC EVALUATION [ ] BETTER DATA FROM COMMISSIONED STUDIES [ ] TIMELY DATA PROVISION FROM DECISION-MAKERS [ ] TRANSPARENT MANAGEMENT STRATEGIES [ ] LONGER-TERM VIEW TAKEN [ ] TRANSPARENT ECONOMIC MODELS [ ] LESS P0LITICAL INFLUENCE [ ] BETTER COMMUNICATION STRATEGY TO RAISE ITS PROFILE AND DEMYSTEFY ECONOMIC JARGON [ ] Upskilling of staff in health economics and/or MORE TRAINING OF HEALTH ECONOMISTS [ ] OTHER (Specify) ______________________________________________________ ______________________________________________________________________
Part III: An Introduction to Program Budgeting & Marginal Analysis (PBMA)
Do you know much about PBMA approach to priority setting?... If so, from where? ___________________________________________________________________________ ___________________________________________________________________________
Page 325
To introduce to you what PBMA is; It is a practical and user-friendly economic approach for priority setting which also has the potential to uphold equity and process/participatory principles. PBMA is a combination of two techniques – ‘Program Budgeting’ and ‘Marginal Analysis’’.
Program budgeting describes the current pattern of spending within an organisation and is a logical starting point for considering options for change. Marginal analysis is an economic technique that looks at the costs and consequences of potential changes that might enable the organisation to better meet its objectives.
It has been used in the UK, Canada and New Zealand. In Australia, it has been applied in Tobacco control; Cancer programs; NSW Area Health Services; and SA Community Health Services. One of the strengths of PBMA is its adaptability to the local context. Seven basic steps can be identified as set out in figure 1 below.
Stages in a PBMA priority setting exercise. 1 2 3 4
5.
6
7
Determine the aim and scope of the priority setting exercise Compile a “program budget”, being a map of current activity and expenditure Form “marginal analysis” advisory panel Determine locally relevant decision-making criteria (a) decision-maker input (b) board of Director input (c) public input Advisory panel to identify options in terms of: (a) areas for service growth (b) areas for resource release through producing same level of output (or outcomes) but with less resources (c) areas for resource release through scaling back or stopping some services Advisory panel to make recommendations in terms of: (a) funding growth areas with new resources (b) decisions to move resources from 5(b) to 5(a) (c) trade-off decisions to move resources from 5(c) to 5(a) if relative value in 5(c) is greater than in 5(a) Validity checks with additional stakeholders and final decisions to inform budget planning process
Source: Priority Setting Toolkit (Mitton and Donaldson 2004).
PBMA can be used by the organisation to suggest ways of improving the efficiency of its programs through examining potential changes to the current mix of services provided (to identify those services which offer value for money for the organisation or the Aboriginal community). For example, within a program, the potential for shifting resources between two types of services for the same group of patients (e.g. outpatient vs. inpatient surgery) could
be
examined.
PBMA can
also
provide
an
opportunity for members of the
organisation/community to fully participate in the priority setting process. Furthermore, it can clarify the concept of benefit (by relating ‘benefit’ to organisational objectives), by utilising best available evidence, and by involving stakeholders in an explicit and systematic planning process. However, PBMA takes more time and effort than other approaches, such as prorated increases/decreases of last year’s budget.
Page 326
From what you may have known about PBMA previously, and from the limited description given above. Do you think PBMA would be an acceptable approach to priority setting in this organisation'?..In other words, are there any characteristics of PBMA that you are attracted to? [ ] YES (Explain:
)
[ ] NO (Why not:
)
[ ] UNSURE
Arguably, the strength of PBMA is that even though data will not always be present, it is at least a way of thinking and doing things in an organisation which focuses people on comparing, in an explicit sense, the marginal costs and benefits of a particular action. Is this how things are done in your organisation? [ ] YES (Explain:
)
[ ] NO (Explain:
)
[ ] UNSURE
PBMA approach allows for the health outcome measure to be expanded to include nonmedical benefits from health services such as community empowerment, equity or cultural security. Do you think that expanding benefits from services/programs could improve the priority setting process? [ ] YES (Explain:
)
[ ] NO (Explain:
)
[ ] UNSURE/OTHER
On what criteria would you evaluate whether the implementation of PBMA within a specific program was a success? [ ] EASY FOR MANAGERS TO USE [ ] BUY IN BY MEDICAL STAFF/NURSING [ ] buy in by funders [ ] SHIFTING OF RESOURCES MATCHED WITH HEALTH GAIN (Increases health gain from the same budget) [ ] LONG-TERM INCREASED HEALTH BENEFIT
Page 327
[ ] CONTINUED USE WITHIN THE ORGANISATION [ ] UPTAKE BY OTHER ORGANISATIONS [ ] OTHER (Specify)
If you were to think that PBMA is not an appropriate tool for your organisation, what other (perhaps conceptual) tool might help in the priority setting process? [ ] NEEDS ASSESSMENT [ ] CONSENSUS-BASED APPROACHES [ ] OTHER (Specify)
Part IV: Potential for Implementing PBMA In this section, we are gauging the potential of piloting and uptake of PBMA approach in your organisation as a tool for routine decision-making. Again, there are no right or wrong answers.
In order to get an idea for the potential of piloting PBMA approach in your organisation, what areas might benefit from initial PBMA application?... in answering, perhaps consider the organisation’s objectives, other known priorities and a priori strategies... your answer might be a particular disease group, a speciality, a care setting, risk factor area, etc. ___________________________________________________________________________ ___________________________________________________________________________ ___________________________________________________________________________ ___________________________________________________________________________ ___________________________________________________________________________
What specific challenges would you foresee with PBMA approach in this organisation? [ ] WON'T HAVE THE DATA [ ] LACK OF TIMELINESS [ ] NOT WELL UNDERSTOOD IN THE ORGANISATION – NEEDS TRAINING OF PERSONNEL [ ] RESOURCES CANNOT BE READILY MOVED [ ] EXTRA RESOURCES REQUIRED [ ] WILL BE MORE WORK FOR DECISION-MAKERS
Page 328
[ ] OTHER (Specify)
Do you have any other general comments that you would like to make at this time? ___________________________________________________________________________ ___________________________________________________________________________ ___________________________________________________________________________ ___________________________________________________________________________ ___________________________________________________________________________
Thank you
Postscript Would you allow Michael Otim to use your organisation’s name and/or your words in this thesis in the form of quotes or anecdotes? [ ] Yes [ ] No [ ] UNSURE AT THIS TIME
Page 329
