Beyond Needs-Based Health Funding: Resource Allocation and Equity at the State and Area Health Service levels in –

Doris Gatwiri Kirigia

A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy

School of Public Health & Community Medicine, Faculty of Medicine, University of New South Wales, Australia September 2009

ii ______

Author’s Declaration ‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no material previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in this thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project’s design and conception or in style, presentation and linguistic expressing is acknowledged.’

Doris Gatwiri Kirigia

Signed ……………………………………………………

iii ______Table of Contents

Table of Contents ...... iii List of Tables...... x List of Figures ...... xiii Acronyms ...... xviii Abstract ...... xx Acknowledgements ...... xxi Chapter 1 : INTRODUCTION – RESOURCE ALLOCATION AND EQUITY ...... 1 1.1 Introduction ...... 1 1.2 Equity and Financial Resource Allocation ...... 3 1.3 Equity of Resource Allocation in New South Wales: The Research Issue ...... 5 1.4 Aims and Research Questions ...... 7 1.5 Overview of Methods ...... 8 1.6 Outline of Thesis ...... 8 Chapter 2 : THE NEW SOUTH WALES SYSTEM...... 10 2.1 Introduction ...... 10 2.2 Financing and Delivery of ...... 11 2.2.1 Financing of Health Care: Commonwealth Government ...... 13 2.2.2 Financing of Health Care: State, Territory and Local Governments ...... 17 2.2.3 Financing of Health Care: Non-Government Sources ...... 19 2.2.4 Delivery of Health Care in Australia ...... 21 2.3 The New South Wales ...... 23 2.3.1 Geography and Population...... 23 2.3.2 Health Status and Inequalities...... 25 2.3.3 The NSW Department of Health ...... 29 2.3.3.1 Organisational Structure...... 30 2.3.4 Financing of the NSW Health System ...... 33 2.3.5 Delivery of Health Services in NSW ...... 38 2.3.6 Area Health Services and Health Care Delivery ...... 41 2.3.7 Financing of Area Health Services: A Brief Overview ...... 43 Chapter 3 : RESOURCE ALLOCATION, EQUITY AND HEALTH NEEDS ...... 45 iv ______3.1 Definition and Levels of Resource Allocation ...... 45 3.1.1 Resource Allocation and Theories of Justice ...... 47 3.1.2 Distributive versus Procedural Justice ...... 49 3.1.3 The Justice of Resource Allocation in NSW ...... 53 3.2 Resource Allocation and Efficiency ...... 57 3.2.1 Technical Efficiency ...... 57 3.2.2 Allocative Efficiency ...... 58 3.3 Equity: Concepts and Definitions ...... 59 3.3.1 Horizontal and Vertical Equity ...... 64 3.3.2 Equal Treatment for Equal Need ...... 66 3.3.3 Equality of Access to Health Care ...... 67 3.3.4 Equality of Health ...... 69 3.3.5 Capacity to Benefit ...... 71 3.3.6 Measurement of Equity ...... 74 3.4 Health Needs: Definition and Measurement ...... 75 3.4.1 Approaches for Measuring Health Needs ...... 77 3.5 Beyond Needs–Based Resource Allocation Formulas ...... 80 Chapter 4 : RESOURCE ALLOCATION MECHANISMS ...... 83 4.1 Introduction: Health System Financing and Resource Allocation ...... 83 4.1.1 Historically-Based Funding ...... 84 4.1.1.1 United Kingdom – RAWP ...... 86 4.1.2 Population and Needs–Based Resource Allocation Mechanism ...... 89 4.1.3 Market Driven Resource Allocation Mechanisms ...... 95 4.2 Resource Allocation in New South Wales ...... 98 4.2.1 Introduction ...... 98 4.2.2 Cost/Expenditure Base Funding ...... 99 4.2.2.1 Casemix/Output–Based Funding ...... 99 4.2.2.2 Budget Share Formula ...... 101 4.2.2.3 Purchasing Formula ...... 101 4.2.3 Episode–Based (Casemix) Resource Allocation in NSW ...... 102 4.3 The NSW Resource Allocation Formula ...... 107 4.4 Resource Distribution Formula (RDF) ...... 112

v ______4.4.1.1 Effects of the Resource Distribution Formula (RDF) ...... 121 4.5 Beyond Needs–Based Funding in NSW: Introducing the Research Questions ...... 122 Chapter 5 : RESEARCH METHODS ...... 128 5.1 Study Aim and Research Questions ...... 128 5.2 Research Design ...... 129 5.2.1 Research Model: Sequential Exploratory Model ...... 130 5.3 Qualitative Method ...... 136 5.3.1 Sample Selection...... 137 5.3.2 Rationale for Selecting NSCCAHS and GWAHS ...... 142 5.3.3 Data Sources and Recruitment of Participants ...... 144 5.3.4 Data Collection and Management...... 146 5.3.5 Data Analysis ...... 147 5.3.6 Data Interpretation ...... 151 5.4 Quantitative Method ...... 152 5.4.1 Data Sources ...... 155 5.4.2 Data Analysis – Principal Component Analysis ...... 156 5.4.3 The General Health Need Index (GHNIdx) ...... 161 5.4.4 Premature Mortality and Morbidity: The Double–Variable Index (DVIdx) ..... 162 5.4.5 Resource Distribution Formula (RDF) ...... 166 5.5 Mapping of Health Needs and Facilities ...... 167 5.6 Methodological Limitations ...... 168 Chapter 6 : EQUITY IN RESOURCE ALLOCATION AT THE STATE LEVEL: QUALITATIVE AND QUANTITATIVE RESULTS ...... 173 6.1 Introduction: Perspectives on Resource Allocation and Equity ...... 173 6.1.1 Equity and how it should be Advanced in Resource Allocation ...... 174 6.1.2 Attempts to Balance Funding between and Community Health ...... 180 6.1.3 Applicability and Adequacy of the RDF in Promoting Equitable Resource Distribution ...... 183 6.1.4 The Political Influence on Resource Allocation ...... 187 6.1.5 Adequacy of Health System Funding ...... 189 6.1.6 Amalgamation of Area Health Services and Equity ...... 190 6.1.7 Limited Consultations during Resource Allocation...... 191

vi ______6.2 Introduction ...... 193 6.3 Comparison of Actual Allocations to Area Health Services: 2003/04 to 2006/07 .. 194 6.3.1 Area Health Services Actual Funding per Head of Population: 2003/04 to 2006/07 ...... 195 6.3.2 Comparison of NSW Health RDF Target Shares and the Actual Resources Allocated to Area Health Services: 2003/04 to 2006/07 ...... 199 6.4 Inter–Area Health Service Level Equity in Resource Allocation ...... 204 6.4.1 General Health Need Index (GHNIdx) ...... 204 6.4.2 The Double–Variable Index (DVIdx) ...... 207 6.5 Inter–Area Health Services Equity in Resource Allocation Assessment ...... 211 Chapter 7 : RESOURCE DISTRIBUTION AT THE AREA HEALTH SERVICE LEVEL: QUALITATIVE AND QUANTITATIVE FINDINGS ...... 223 7.1 Overview ...... 223 7.2 Introduction ...... 223 7.1.1 Inadequate Funding and Maldistribution of Available Funds...... 224 7.1.2 Lack of Tools to Guide Resource Distribution at the AHS Level ...... 227 7.1.3 Greater Emphasis on Achieving Efficiency Hampers Achievement of Equity ...... 228 7.1.4 Wastage of Resources Due to Lack of Coordinated Approach to Service Delivery and Collaboration between Health Providers ...... 230 7.1.5 Lack of Resource Allocation Transparency and Accountability ...... 232 7.1.6 Lack of Data Collection, Monitoring and Evaluation Systems...... 233 7.3 Introduction ...... 237 7.4 Distribution of Resources within Area Health Service Level ...... 237 7.1.7 Resource allocation to Area Health Services Programs and Services ...... 238 7.5 Area Health Service Programs and Services per Capita Expenditure ...... 242 7.1.8 Per Capita Expenditure of Primary & Community Health Services ...... 243 7.1.9 Per Capita Expenditure of Aboriginal Health Services...... 245 7.1.10 Per Capita Expenditure of Rehabilitation & Extended Care Services ...... 247 7.1.11 Per Capita Expenditure of Population Health Services...... 247 7.6 Per Capita Expenditure of –Based Programs 2004/05 – 2006/07...... 248 7.1.12 Outpatient Services per Capita Expenditure 2004/05 to 2006/07 ...... 248

vii ______7.1.13 Per Capita Expenditure of Emergency Care Services ...... 250 7.1.14 Per Capita Expenditure Overnight Acute Services ...... 250 7.1.15 Per Capita Expenditure Same Day Acute Services ...... 251 7.1.16 Per Capita Expenditure Mental Health Services ...... 251 7.7 Teaching and Research Total Expenditure ...... 252 Chapter 8 : EQUITY IN RESOURCE ALLOCATION WITHIN AREA HEALTH SERVICES ...... 255 8.1 Opinions on Equity and Resource Allocation ...... 256 8.1.1 Confusion regarding Equity Principles Guiding AHS Resource Allocation ..... 256 8.1.2 Lack of Openness in Funds Distribution ...... 259 8.1.3 Limited Capacity to Benefit from Resource Allocation and Health Services ... 262 8.1.4 Need for Custom–Made Models for Distributing Resources in each AHS ...... 264 8.1.5 Lack of Partnership between Service Providers ...... 265 8.1.6 Restricted Access to Health Services ...... 266 8.2 Health Needs among Local Government Areas (LGAs) and DVIdx Shares ...... 271 8.2.1 Northern Sydney Central Coast Area Health Service – LGAs and DVIdx Shares ...... 272 8.2.2 Greater Western Area Health Service – LGAs and DVIdx Shares ...... 274 8.2.3 Movement towards Equity in Health Status, 1987–2006 ...... 279 8.3 DVIdx Equity Shares and Health Services ...... 280 8.3.1 Northern Sydney Central Coast AHS Health Services ...... 281 8.3.2 Greater Western Area Health Services ...... 287 Chapter 9 : DISCUSSION ...... 294 9.1 To What Extent has Resource Allocation to Area Health Services by the NSW Department of Health been Equitable?...... 294 9.1.1 Introduction ...... 294 9.1.2 Resource Distribution Formula and Equity ...... 296 9.1.2.1 Evidence from the Resource Distribution Formula (RDF) ...... 296 9.1.2.2 The RDF and Health Needs ...... 297 9.1.2.3 Towards a Greater Equity ...... 298 9.1.3 Alternative Formula Based Allocation ...... 299 9.1.3.1 SEIFA–IRSD Index Based Resource Allocation ...... 299

viii ______9.1.3.2 GHNIdx Based Resource Allocation ...... 300 9.1.3.3 DVIdx Based Resource Allocation ...... 301 9.1.3.4 Evidence from Other Sources ...... 303 9.1.4 To What Extent has the Resource Distribution Formula been used? ...... 304 9.1.5 Is the RDF a Good Formula for Achieving Equity in Resource Allocation to AHSs? ...... 307 9.1.5.1 Level of Comprehension and Ease of Applicability of the RDF ...... 307 9.1.5.2 Validity of Indicators Used in the RDF ...... 308 9.1.6 Can Other Alternative Indicators of Health Need be used? ...... 310 9.1.6.1 General Health Need Index (GHNIdx) ...... 311 9.1.6.2 Morbidity and Mortality Index (DVIdx) ...... 312 9.2 How are the Financial Resources Received from State Level Distributed at the Area Health Service level? ...... 314 9.2.1 Introduction ...... 314 9.2.2 What are the Current Processes of Allocating Resources within AHSs? ...... 314 9.2.3 What Role does NSW Department of Health Play in determining the within–AHS Allocation of Funds? ...... 315 9.2.4 What Role does Area Health Services Play in Decision Making? ...... 316 9.2.5 Does the Process of Resource Allocation at the AHS Level appear to reflect a Concern for Intra–AHS Equity? ...... 317 9.3 Is Equity Reflected in the Allocation of Resources at the Area Health Service Level? ...... 318 9.3.1 Introduction ...... 318 9.3.2 What Evidence is Available? ...... 318 9.3.2.1 Opinions from the Interviews ...... 319 9.3.2.2 The Chronic Morbidity and Premature Mortality Data ...... 320 9.3.2.3 The Mapping of Health Needs and Services/Programs ...... 320 9.3.2.4 Evidence from Other Sources ...... 322 9.4 What Prospect is there for Allocation of NSW Health Resources both at the State and Area Health Service Levels to Promote Equity? ...... 324 9.5 Study Limitations ...... 336 Chapter 10 : CONCLUSION AND SUGGESTIONS FOR FURTHER RESEARCH ..... 340

ix ______10.1 Policy Implications of the Study ...... 340 10.1.1 Implication for Examining how Equity with regards to Resource Allocation is Assessed ...... 340 10.1.2 Implications for Developing Appropriate Tools to Guide Resource Allocation ...... 342 10.1.3 Resolving the Historical and Political Factors Affecting the Equitable Allocation of Health Funds ...... 343 10.1.4 Encouraging Multi–Sectoral Collaboration for Delivering and Promoting Health ...... 343 10.2 Contributions to Knowledge ...... 344 10.2.1 Contributions to the Method of Assessing Health Needs ...... 344 10.2.2 Contributions to Current Debate on Resource Allocation and Equity...... 346 10.3 Suggestions for Future Research ...... 347 10.3.1 Further Work is required in Developing Better Health Information Systems for Area Health Services ...... 348 Appendix A Socio–Economic Indexes for Areas (SEIFA)...... 350

x ______List of Tables

Table 2–1: Average Recurrent Health Expenditure per Person (in current price) for States and Territories in Australia, 1997–2007 ...... 19 Table 2–2: Gap in Life Expectancy between Indigenous and Non–Indigenous Populations in New South Wales ...... 27 Table 2–3: NSW Total Health Expenditure on Recurrent and Capital and Sources of Funds, 2001/02 to 2006/07 ...... 35 Table 2–4: Selected Demographic, Socio–Economic and Health–Related Indicators of Area Health Services in NSW, 2007 ...... 40 Table 2–5: Hospital Beds in New South Wales by Area Health Services, 2008...... 42 Table 2–6: Health Professionals in Area Health Services in New South Wales, 2008 .. 43 Table 4–1: Comparative Needs–Based Models of Resource Allocation ...... 92 Table 4–2: The New South Wales Resource Allocation Formula (RAF), 1993 Revision ...... 112 Table 4–3: Principles Guiding the Structure of Resource Distribution Formula (RDF) ...... 114 Table 4–4: Resource Distribution Formula (RDF), 2005 Revision ...... 116 Table 4–5: Components of the Resource Distribution Formula and Program Alignment ...... 120 Table 5–1: Index of Relative Socio–economic Disadvantage (IRSD) of Local Government Areas in Northern Sydney Central Coast, 2008 ...... 139 Table 5–2: Index of Relative Socioeconomic Disadvantage (IRSD) of Local Government Areas in Greater Western Area Health Service, 2008 ...... 140 Table 5–3: Initial Variables Defined ...... 158 Table 5–4: Unrotated Component Matrix ...... 159 Table 5–5: Component Score Coefficient Matrix ...... 160 Table 5–6: Development of Double Variable Index (DVIdx) for Eight Area Health Services in NSW...... 164 Table 5–7: Development of Double Variable Index (DVIdx) for Northern Sydney Central Coast AHS ...... 165

xi ______Table 5–8: Development of Double Variable Index (DVIdx) for Greater Western AHS ...... 166 Table 6–1: NSW Health Actual Share of Resources and RDF Target Shares, 1989–2008 ...... 185 Table 6–2: Actual Funding Allocations to Area Health Services, 2003/04 to 2006/07 194 Table 6–3: Area Health Services Premature Death Rates per 100,000 Persons Aged Under 75 Years in NSW, 1987–2006 ...... 198 Table 6–4: NSW Health Resource Distribution Formula Target Shares, 1989–2008 .. 199 Table 6–5: Comparison of Area Health Services Actual Funding Allocation and RDF Target Shares, 2003/04 to 2006/07 ($M) ...... 201 Table 6–6: Actual Allocations and RDF–based Target Shares: Average Distance from RDF, 2003/04 to 2006/07 ...... 201 Table 6–7: AHSs General Health Need Index Principal Component Analysis...... 205 Table 6–8: General Health Need Index–Equity Share (GHNIdx–ES) ...... 205 Table 6–9: General Health Need Index (GHNIdx) Normalised and Ranked according to Level of Health Needs ...... 206 Table 6–10: Area Health Services Ranked according to the General Health Need Index (GHNIdx) Equity–Shares ...... 207 Table 6–11: Development of Double–Variable Index (DVIdx) Equity Shares ...... 208 Table 6–12: Area Health Services simple Double–Variable Index (DVIdx) and GHNIdx Equity Shares ...... 210 Table 6–13: Comparison of AHS Actual Funds Allocation and General Health Need Index (GHNIdx) Equity Shares, 2003/04 to 2006/07 ($M) ...... 213 Table 6–14: Actual Funds and GHNIdx–based Allocations: Average Distance from GHNIdx, 2003/04 to 2006/07 ...... 213 Table 6–15: Funds for Reallocation to Achieve Equity Targets based on GHNIdx, 2003/04 to 2006/07 ...... 217 Table 6–16: Comparison of AHSs Actual Funds Allocation and Double Variable Index (DVIdx) Equity Shares, 2003/04 to 2006/07 ...... 218 Table 6–17: Actual Funds and DVIdx Allocation: Average Distance from DVIdx, 2003/04 to 2006/07 ...... 218

xii ______Table 6–18: Funds for Reallocation to Achieve Equity Targets based on DVIdx, 2003/04 to 2006/07 ...... 220 Table 7–1: AHSs Health Programs and Services Categorised ...... 238 Table 7–2: Area Health Services Programs/Services Funding, 2005 to 2007 ...... 239 Table 7–3: Area Health Services Community Health, Hospital Services and Teaching & Research Expenditure, 2004/05 to 2006/07 ...... 241 Table 7–4: Per Capita Expenditure of Community–based Programmes, 2004/05 to 2006/07 ...... 244 Table 7–5: Per Capita Expenditure of Hospital–based Programmes, 2004/05 to 2006/07 ...... 249 Table 7–6: Area Health Services Teaching and Research Expenditure, 2005 to 2007 ...... 253 Table 8–1: Chronic Disease and Premature Death Statistics, Northern Sydney Central Coast Area Health Service (NSCCAHS) ...... 272 Table 8–2: Health Need Index (DVIdx) Funding Shares for Northern Sydney Central Coast AHS ...... 273 Table 8–3: Chronic Disease and Premature Death Statistics, Greater Western Area Health Services (GWAHS)...... 276 Table 8–4: Health Need Index (DVIdx) Funding Shares for Greater Western Area Health Service ...... 277 Table 8–5: Population, Health Need Index (DVIdx) and No. of Health Services within NSCCAHS ...... 282 Table 8–6: Population, Health Need Index (DVIdx) and Number of Health Services within GWAHS ...... 288 Table 10–1: Comparison of DVIdx, GHNIdx and SEIFA–IRSD Funding Shares ...... 351 Table 10–2: Comparison of Area Health Services Actual Funds Allocation and SEIFA– IRSD Shares, 2003/04 – 2006/07 ...... 352 Table 10–3: Actual Funds and SEIFA–IRSD–based Allocations: Average Distance from SEIFA–IRSD, 2003/04 – 2006/07 ...... 352 Table 10–4: Actual and SEIFA–IRSD–based Allocation by Area Health Services, 2003/04 – 2006/07 ...... 353

xiii ______List of Figures

Figure 1–1: NSW Area Health Services Weighted Average Distance from RDF Target 6 Figure 2–1: Australia Health Expenditure by Sources of Funds, 2006–07 ...... 12 Figure 2–2: Health Expenditure by Sources of Funds in Australia, 1999–00 to 2006–07 .. 13 Figure 2–3: The Structure of the Australian Health Care System and Major Channels of Funding Flows ...... 17 Table 2–1: Average Recurrent Health Expenditure per Person (in current price) for States and Territories in Australia, 1997–2007 ...... 19 Figure 2–4: Health Expenditure of Non–Government Sector by Sources of Funding, 1995– 2006 ...... 20 Figure 2–5: Premature Deaths by Aboriginality, NSW 1998 to 2006 ...... 26 Table 2–2: Gap in Life Expectancy between Indigenous and Non–Indigenous Populations in New South Wales ...... 27 Figure 2–6: Difficulty Accessing Health Care When Needed by Area Health Services for Persons Aged 16 Years and Over, NSW 2007 ...... 29 Figure 2–7: New South Wales Health Organisational Chart, 2008 ...... 31 Figure 2–8: New South Wales State Health Funding by Sources of Funds, 2001/02 to 2006/07 ...... 33 Figure 2–9: Total Health Expenditure ($30.8 billion) in New South Wales by Area of Expenditure and Source of Funds – 2006/07 ...... 34 Table 2–3: NSW Total Health Expenditure on Recurrent and Capital and Sources of Funds, 2001/02 to 2006/07 ...... 35 Figure 2–10: Commonwealth Government Health Expenditure in New South Wales by Area of Expenditure – 2006–07 ...... 36 Figure 2–11: State and Local Government Health Expenditure in New South Wales by Area of Expenditure: 2006–07 ...... 37 Figure 2–12: Non–Government Sector Health Expenditure in New South Wales by Area of Expenditure: 2006–07 ...... 38 Figure 2–13: NSW Health Eight Health Service Areas – 2009 ...... 39 Table 2–4: Selected Demographic, Socio–Economic and Health–Related Indicators of Area Health Services in NSW, 2007...... 40

xiv ______Table 2–5: Hospital Beds in New South Wales by Area Health Services, 2008 ...... 42 Table 2–6: Health Professionals in Area Health Services in New South Wales, 2008 ...... 43 Figure 4–1: Resource Allocation Mechanisms ...... 84 Table 4–1: Comparative Needs–Based Models of Resource Allocation ...... 92 Table 4–2: The New South Wales Resource Allocation Formula (RAF), 1993 Revision 112 Table 4–3: Principles Guiding the Structure of Resource Distribution Formula (RDF) ... 114 Figure 4–2: Steps in Calculating Resource Distribution Formula Funding Targets ...... 115 Table 4–4: Resource Distribution Formula (RDF), 2005 Revision ...... 116 Table 4–5: Components of the Resource Distribution Formula and Program Alignment . 120 Figure 5–1: Approach Taken to Address Research Questions...... 132 Table 5–1: Index of Relative Socio–economic Disadvantage (IRSD) of Local Government Areas in Northern Sydney Central Coast, 2008 ...... 139 Table 5–2: Index of Relative Socioeconomic Disadvantage (IRSD) of Local Government Areas in Greater Western Area Health Service, 2008...... 140 Table 5–3: Initial Variables Defined ...... 158 Table 5–4: Unrotated Component Matrix ...... 159 Table 5–5: Component Score Coefficient Matrix ...... 160 Figure 5–2: Health Needs Formula ...... 160 Figure 5–3: NSW Area Health Services – General Health Need Indices (GHNIdx) ...... 162 Table 5–6: Development of Double Variable Index (DVIdx) for Eight Area Health Services in NSW ...... 164 Table 5–7: Development of Double Variable Index (DVIdx) for Northern Sydney Central Coast AHS ...... 165 Table 5–8: Development of Double Variable Index (DVIdx) for Greater Western AHS . 166 Table 6–1: NSW Health Actual Share of Resources and RDF Target Shares, 1989–2008 185 Table 6–2: Actual Funding Allocations to Area Health Services, 2003/04 to 2006/07 ..... 194 Figure 6–1: Area Health Services Potentially Avoidable Premature Deaths from Causes Amenable to Health Care; Persons Aged Under 75 Years in NSW (Rates per 100,000 Population) – 2006 ...... 195 Figure 6–2: Area Health Services Funds Allocation per Head of Population: 2003/04 to 2006/07 ...... 196

xv ______Table 6–3: Area Health Services Premature Death Rates per 100,000 Persons Aged Under 75 Years in NSW, 1987–2006 ...... 198 Figure 6–3: Area Health Services Premature Deaths for Persons Aged Under 75 Years in NSW 1987 to 2006 (Rates per 100,000 Populations) ...... 198 Table 6–4: NSW Health Resource Distribution Formula Target Shares, 1989–2008 ...... 199 Table 6–5: Comparison of Area Health Services Actual Funding Allocation and RDF Target Shares, 2003/04 to 2006/07 ($M) ...... 201 Table 6–6: Actual Allocations and RDF–based Target Shares: Average Distance from RDF, 2003/04 to 2006/07 ...... 201 Figure 6–4: Actual & Resource Distribution Formula (RDF) by Area Health Services: 2003/04 – 2006/07 ...... 203 Table 6–7: AHSs General Health Need Index Principal Component Analysis ...... 205 Table 6–8: General Health Need Index–Equity Share (GHNIdx–ES)...... 205 Table 6–9: General Health Need Index (GHNIdx) Normalised and Ranked according to Level of Health Needs ...... 206 Table 6–10: Area Health Services Ranked according to the General Health Need Index (GHNIdx) Equity–Shares ...... 207 Table 6–11: Development of Double–Variable Index (DVIdx) Equity Shares ...... 208 Table 6–12: Area Health Services simple Double–Variable Index (DVIdx) and GHNIdx Equity Shares ...... 210 Figure 6–5: GHNIdx and DVIdx Equity–Shares Compared ...... 211 Table 6–13: Comparison of AHS Actual Funds Allocation and General Health Need Index (GHNIdx) Equity Shares, 2003/04 to 2006/07 ($M) ...... 213 Table 6–14: Actual Funds and GHNIdx–based Allocations: Average Distance from GHNIdx, 2003/04 to 2006/07 ...... 213 Figure 6–6: Actual & GHNIdx–based Allocation by Area Health Service: 2003/04 – 2006/07 ...... 215 Table 6–15: Funds for Reallocation to Achieve Equity Targets based on GHNIdx, 2003/04 to 2006/07 ...... 217 Table 6–16: Comparison of AHSs Actual Funds Allocation and Double Variable Index (DVIdx) Equity Shares, 2003/04 to 2006/07 ...... 218

xvi ______Table 6–17: Actual Funds and DVIdx Allocation: Average Distance from DVIdx, 2003/04 to 2006/07 ...... 218 Figure 6–7: Actual & DVIdx–Based Allocation by Area Health Services: 2003/04 – 2006/07 ...... 219 Table 6–18: Funds for Reallocation to Achieve Equity Targets based on DVIdx, 2003/04 to 2006/07 ...... 220 Table 7–1: AHSs Health Programs and Services Categorised...... 238 Table 7–2: Area Health Services Programs/Services Funding, 2005 to 2007 ...... 239 Table 7–3: Area Health Services Community Health, Hospital Services and Teaching & Research Expenditure, 2004/05 to 2006/07 ...... 241 Figure 7–1: Area Health Services Community Health, Hospital Services and Teaching & Research Expenditure: 2004/05 – 2006/07 ...... 241 Table 7–4: Per Capita Expenditure of Community–based Programmes, 2004/05 to 2006/07 ...... 244 Figure 7–2: Per Capita Expenditure of Primary and Community Health–based Health Services: 2004/05 – 2006/07 ...... 244 Table 7–5: Per Capita Expenditure of Hospital–based Programmes, 2004/05 to 2006/07 249 Figure 7–3: Per Capita Expenditure of Hospital–based Health Programs: 2004/05 – 2006/07 ...... 249 Table 7–6: Area Health Services Teaching and Research Expenditure, 2005 to 2007...... 253 Figure 7–4: Comparison of Area Health Services Teaching and Research Expenditure, 2004/5 – 2006/07 ...... 253 Table 8–1: Chronic Disease and Premature Death Statistics, Northern Sydney Central Coast Area Health Service (NSCCAHS) ...... 272 Table 8–2: Health Need Index (DVIdx) Funding Shares for Northern Sydney Central Coast AHS ...... 273 Figure 8–1: Population & Index of Health Need (DVIdx) by Local Government Areas: Northern Sydney Central Coast Area Health Service ...... 274 Table 8–3: Chronic Disease and Premature Death Statistics, Greater Western Area Health Services (GWAHS) ...... 276 Table 8–4: Health Need Index (DVIdx) Funding Shares for Greater Western Area Health Service ...... 277

xvii ______Figure 8–2: Population & Index of Health Need (DVIdx) by Local Government Areas: Greater Western Area Health Service ...... 278 Figure 8–3: Northern Sydney Central Coast and Greater Western Area Health Services Premature Mortality Rates per 100,000 Population, 1987 to 2006 ...... 279 Figure 8–4: Difficulty Accessing Health Care when Needed in NSCCAHS and GWAHS for Persons Aged 16 Years and Over, NSW 2007 ...... 280 Table 8–5: Population, Health Need Index (DVIdx) and No. of Health Services within NSCCAHS ...... 282 Figure 8–5: Health Services and Health Needs (DVIdx) within Northern Sydney Central Coast Area Health Service Compared...... 283 Figure 8–6: Population and Health Needs – Northern Sydney Central Coast AHS ...... 285 Figure 8–7: Mapped Health Services – Northern Sydney Central Coast AHS ...... 286 Table 8–6: Population, Health Need Index (DVIdx) and Number of Health Services within GWAHS ...... 288 Figure 8–8: Premature Mortality and Chronic Morbidity Health Needs Index (DVIdx) – Greater Western AHS ...... 289 Figure 8–9: Mapped Population and Health Needs – Greater Western AHS ...... 291 Figure 8–10: Mapped Health Services – Greater Western AHS...... 292 Table 10–1: Comparison of DVIdx, GHNIdx and SEIFA–IRSD Funding Shares...... 351 Table 10–2: Comparison of Area Health Services Actual Funds Allocation and SEIFA– IRSD Shares, 2003/04 – 2006/07 ...... 352 Table 10–3: Actual Funds and SEIFA–IRSD–based Allocations: Average Distance from SEIFA–IRSD, 2003/04 – 2006/07 ...... 352 Table 10–4: Actual and SEIFA–IRSD–based Allocation by Area Health Services, 2003/04 – 2006/07 ...... 353 Figure 10–1: Actual and SEIFA–IRSD–based Allocation by Area Health Services: 2003/04 – 2006/07 ...... 354

xviii ______Acronyms

ABS Australian Bureau of Statistics ACT Australian Capital Territory ACAC Australia (Independent) Commission Against Corruption AHS Area Health Services AHCA Australian Health Care Agreements AIHW Australian Institute of Health and Welfare DoH Department of Health DVIdx Double Variable Index EDOCU Index of Education and Occupation GHNIdx General Health Need Index GWAHS Greater Western Area Health Service GSAHS Greater Southern Area Health Service HFC Health Care Financing HIC Commission NHS National Health Insurance (British) H&NEAHS Hunter & New England Area Health Services LGA Local Government Areas IRSD Index of Relative Socioeconomic Disadvantage MBS Benefits Schedule NCAHS North Coast Area Health Service NSCCAHS Northern Sydney Central Coast Area Health Service NHS National Health Survey NSW New South Wales NSW Health NSW Department of Health NT PBS Pharmaceutical Benefits Scheme PCA Principal Component Analysis PHIDU Public Health Information Development Unit QLD RAF Resource Allocation Formula xix ______RDF Resource Distribution Formula RAWP Resource Allocation Working Party SEIFA Socioeconomic Index for Areas SES&IAHS South Eastern Sydney & Illawarra Area Health Service SSWAHS Sydney South West Area Health Service SWAHS Sydney West Area Health Service SA TAS UNDP United Nations Development Programme VIC WA

WHO World Health Organisation

NSW Health has been used in some places interchangeably with the NSW Department of Health, both terms referring to the State Government’s central health authority, but ‘NSW Health’ can also include other entities such as Area Health Services.

Indigenous : The Aboriginal and Torres Strait Islander people of Australia are also referred to as ‘’.

Medicare is Australia‘s publicly–funded, universal health scheme, providing affordable treatment by doctors and in public hospitals for all citizens and permanent residents (as well as tourists from countries which have reciprocal arrangements with Australia).

xx ______Abstract

Addressing inequities in health both within and between countries has attracted considerable global attention in recent years. In theory, equity remains one of the key policy objectives of health systems and underpins the allocation of health sector resources in many countries. In practice, however, current evidence demonstrates that only limited progress has been made in terms of bridging the health inequity gap and improving the health of the least advantaged. The persistence of inequities in health and health outcomes raises concerns about how governments and health authorities distribute limited health resources to improve the health of the poor and most vulnerable and thereby promote equity. This thesis is about equity and allocation of financial resources in the health system of New South Wales, one of the eight states of Australia. It investigated the extent to which there has been a movement towards equity in resource allocation to Area Health Services under the NSW Health Resource Distribution Formula and whether this has been reflected in equitable resource allocation within Area Health Services. It considered only resources allocated through the NSW Department of Health.

The study employed a combination of qualitative and quantitative methods to gather and analyse data. The qualitative component analysed data gathered through semi–structured interviews with policy makers, health executives, managers, and other stakeholders to establish the resource allocation processes and the factors upon which the allocation decisions were based. The quantitative component analysed health expenditure and health needs data to assess the extent to which allocation of resources from the State to Area Health Service levels has been equitable in terms of reflecting the level of health needs. Two indices were constructed and used as proxies for health needs. Principal component analysis was used in the construction of one of the indices, using demographic, socio– economic and health-related data. The other index was developed using a combination of premature mortality and morbidity data. The quantitative study spans the two decades 1989/90 to 2006/07, with a more detailed analysis of material for the years 2003/04 to 2006/07.

The findings of the study show a considerable degree of inequity in resource allocation with several Area Health Services (AHSs) receiving less than a fair share of funding for the years analysed, although some movements towards equity were evident. This contradicts the general impression that the introduction of the resource distribution formula in NSW has significantly improved equity in resource allocation. In general, funding allocation at the State level correlated significantly with population size but not with health needs of the eight AHSs in NSW. Similarly, within the AHSs, allocation of funds was based on programs and services and not on health needs. Key issues that emerged from the qualitative data as affecting the equity with which health funds are allocated in the NSW health system include limited use of the resource distribution formula at the state level, lack of an effective resource allocation tool to guide the distribution of funds within AHSs, and insufficient emphasis on equity at the AHS level. It is crucial that these and several other issues identifies in the study are addressed if current inequities in funding and in health outcomes generally are to be effectively reduced.

xxi ______Acknowledgements

First, I would like to acknowledge the University of New South Wales and School of Public Health Community Medicine for granting me the opportunity to pursue this PhD through the award of a scholarship. I also acknowledge those who participated in this study, mainly from the NSW Department of Health, and the Northern Sydney Central Coast and Greater Western Area Health Services, for their informative contributions. This project could not have been successfully completed without the support, important suggestions and feedback from many individuals. Special thanks go to Associate Professor Paul McNeill for all his support during the conception of this project. I also acknowledge the support and contributions of Associate Professor Anna Whelan, Associate Professor Rosemary Knight and Mr Kevin Forde.

Most of all, I am indebted to Dr John Dewdney and Dr Augustine Asante for being enthusiastic supervisors and a great source of inspiration. Their constant critical and very constructive feedback, support and guidance on a complex research issue and under difficult circumstances is very much appreciated. Their positive attitude and encouragement, especially during those times when my health was proving a real hurdle, have led to timely completion of this project.

Sincere thanks go to all my colleagues in the Samuels Building and the PhD room for their support, laughter and for organising many social events to balance studies and life generally. Special thanks must also go to my great friends Dr Yomi Ojitunde, Pauline and Tony Gatome, Dr Muli and Ted for their encouragement and support throughout the PhD journey, including Keith, and Drs Newall, Razee and Demirkol for their valuable feedback.

I am extremely grateful to Neil, Rosie, Mwenda and Munene Linwood for their exceptional support throughout my stay in Sydney and for their familial role in the absence of family in Australia.

Special thanks go to my sisters: Judy, Mercy, Lenity, Jane, Berth and brothers: Mwirigi, Koome, Laban, and Moses for their encouragement and most of all to my brother, Dr Joses xxii ______Kirigia, for being a wonderful role model and for his unwavering exceptional support and counsel throughout the my life’s journey.

Many thanks go to Mum and Dad for teaching us the principles of ‘fairness’ at a very early age. They might not have known about the implied meaning of that term but their attitude in sharing whatever possessions they have with the less fortunate is one just act. This altruism has indeed continued to guide our existence. If the collective world pursued similar acts of benevolence towards our less fortunate members of society and paid a little attention to their needs, this world would be a better place for us all. Thank you also for instilling in us the best foundations in life; your passion and investment in education has indeed empowered us all.

This thesis is dedicated to my lovely nieces: Kathure, Karimi, Tina and nephews: Gitonga, Jason, Jayden, and Mutuma. I hope you can forgive me for being an absent Auntie and for missing out on all your birthdays and other special occasions. My wish is that this piece of hard work will be an inspiration and a testimony that there is no dream unattainable if one strives to achieve it. May this encourage you to always challenge yourself outside your ‘comfort–zone’ and to never settle for second best!

1 ______

Chapter 1: INTRODUCTION – RESOURCE ALLOCATION AND EQUITY

“Equity in health is not about eliminating all health differences so that everyone has the same level of health, but rather to reduce or eliminate those which result from factors which are considered to be both avoidable and unfair…” Whitehead 1992

Overview This chapter provides a broad introduction to the study, including the motivation for it and its objectives. It conceptualises the research issue from a global health system financing and equity perspective and narrows down to Australia and New South Wales demographics and health funding. It reviews briefly the approaches that have been used for resource allocation in NSW and then follows with a statement of purpose that articulates the research issue and the main questions addressed by the study. The final section provides the outline of the thesis, briefly describing what each of the 10 chapters focuses on.

1.1 Introduction The reduction of inequities in health both within and between countries is an important aspect of the present global agenda to reduce poverty and improve the health of the least advantaged. Equity is considered to be one of the guiding principles for and allocation of health sector resources in many countries (Carr–Hill et al., 1994, Gwatkin et al., 2004, Mooney, 2003, McIntyre and Mooney, 2007). However, while some progress has been made in relation to improving health and health care access worldwide, there are still significant disparities in both developed and developing countries that are seen as unfair, unnecessary and avoidable (Whitehead, 1992, Whitehead et al., 2001). These disparities raise concerns about how governments and health authorities allocate limited health resources to meet the health needs of their populations and the degree to which these allocations benefit the most needy populations and health jurisdictions.

There has been intensive debate over the last few decades about improving equity of health systems (Williams, 2005, Zere et al., 2007, Asada and Kephart, 2007, McIntyre and Mooney, 2007). Governments have come under pressure to reorient health priorities 2 ______and redistribute health care resources to promote equity1. Many industrialised countries have made some efforts to address these issues by introducing resource allocation policies that focus on improving the equity of distribution of health care funds (Asante et al., 2006, Black and Mooney, 2002, Carr–Hill et al., 1994) with varying degrees of success. In many settings, however, these (equity–focused resource allocation) policies are implemented only at the national level of the health system to guide inter–regional, provincial resource allocation; they are not implemented at sub–national level to guide resource allocation within regions, provinces or areas (Mitton and Donaldson, 2003, McIntyre et al., 2002).

In fact, equity policies are not always clearly set out in many health systems (McIntyre and Gilson, 2002, McIntyre and Mooney, 2007) or communicated well to the lower levels of the health system. Health managers and other stakeholders at sub–national (Area Health Service) levels may have limited knowledge or understanding of what equity and resource allocation policies the health system is pursuing. This thesis argues that while it is essential that equity and resource allocation policies are implemented vigorously at the highest levels of health systems to reduce geographical funding inequities, it is crucial that these policies are also promoted and adopted at lower levels to facilitate broader achievement of equity and equitable resource distribution within the entire health system.

The motivation for this work is two–fold: the first relates to my interest in health equity in general and equity of resource allocation in particular. There are genuine concerns among health care analysts that the growing problem of inequities in health cannot be adequately addressed without a change in the way resources are distributed and used. Available evidence suggests that, in many countries, increases in financial resources for health are significantly outpaced by the growing demand for health services. At the same time, historic inequities in health outcomes between different population groups persist. I share the view that these difficult challenges cannot be effectively addressed without a conscientious redistribution of current resources.

1 Equity is defined in this study as equality of access to financial resources for health care on the basis of health need. Health need is defined in terms of population health status.

3 ______The second motivation for this study stems from the apparent ambiguities surrounding resource allocation and equity policies in the health sector. Since health equity has become an issue of global concern, policy makers have devoted considerable time and energy to developing policies, including sophisticated resource allocation formulas designed to improve equity. However, these policies are often implemented in at the top level of the health system with local health authorities having little or no idea about what these policies seek to achieve. In general, this affects the degree to which health equity and equity of resource allocation in particular is promoted in the health system. I believe that, if any significant improvements in health equity, including equity of resource allocation are to be achieved, equity–oriented policies have to be clearly established and implemented at all levels of the health care system (national and sub– national).

This dissertation focuses specifically on equity of resource allocation policies and their implementation at different health system levels using the NSW health system as a case study. It highlights the lack of clear articulation of these policies and the disjointed manner in which they are often implemented. The study contributes to filling the current gap in knowledge about how the issue of resource allocation and equity are dealt with at different levels of the health system. Overall, it serves the purpose of providing empirical data to inform and support health policy reforms that seek to promote equity, especially equity of resource allocation in the NSW health system. The next section places the study in the global context of equity and financial resource allocation in the health sector.

1.2 Equity and Financial Resource Allocation Equity remains an important health policy goal despite the conceptual difficulties surrounding the term. Pursuing health equity is widely seen as reflecting efforts to reduce unequal opportunities to be healthy often associated with membership of less privileged social groups (World Bank, 2006, Braveman, 2003, Braveman et al., 2005, Gwatkin, 2005). Worldwide, there is growing evidence of inequalities in health that are ‘unfair, unnecessary and avoidable’ (Marmot, 2001, Leeder, 2003, Gwatkin, 2005, Korda et al., 2007, Macintyre et al., 2003, Whitehead, 1992).

4 ______These inequalities are usually exemplified by the well–off receiving more and better quality health services, while the poor and the most disadvantaged facing severe access restrictions (Gwatkin, 2004, van Doorslaer et al., 2006). Apart from the rich having access to better quality health care, the allocation of health dollars tends to favour the rich. Studies have shown that in many countries a disproportionate amount of the health budget is spent on hospital–based curative services that are urban–biased and frequently used by the rich (Castro–Leal et al., 2000). WHO (2001) estimates that most health systems spend between 60 to 80% of their resources on hospital–based , leaving a small proportion for basic primary health services often used by the poor (WHO Commission on Macroeconomics Health, 2001). Across population groups, analysis of health expenditure patterns shows that in several countries spending on disadvantaged and minority populations falls far short of what is required to bridge the gap of inequalities in health outcomes.

Debate about the best approach for overcoming health inequities has escalated in the last decade (Gwatkin et al., 2004, Sheldon and Smith, 2000, Leeder, 2003, McIntyre and Gilson, 2002, Mooney, 2003, Sen, 2002). This has subjected resource allocation mechanisms to intense scrutiny in several countries. Many of these mechanisms are seen as either inherently inequitable and thus contributing to the current problem of inequities, or not properly designed and thus not able to shift funds around to enhance equity (Gilson and McIntyre, 2005, Gwatkin et al., 2004, McIntyre and Gilson, 2002, Mooney and Houston, 2004, Asante et al., 2006).

In the last 30 years, following the publication of the Resource Allocation Working Party’s (RAWP) formula in England in 1976 (Department of Health, 1976), several countries have sought better ways of allocating health sector resources to enhance equity and improve health outcomes of their populations (for a full review of the RAWP and other models for resource allocation in the health sector see Chapter 4). Much of their effort has been directed at developing resource allocation mechanisms that will ensure that poorer jurisdictions and disadvantaged population groups have access to a fair share of funding and basic health services commensurate with their health needs (Gilson, 1994, Gwatkin, 2003, Gwatkin, 2005). However, significant questions remain as to whether such mechanisms have had any impact on equity. This study examines the

5 ______degree to which equity in the distribution of health sector resources has been achieved in Australia’s most populous state, New South Wales (NSW).

Australia has a population of about 21 million spread over six states: New South Wales, Victoria, Queensland, Western Australia, South Australia, Tasmania and two territories, the Northern Territory and the Australian Capital Territory (Australian Bureau of Statistics, 2007b, Australian Bureau of Statistics, 2004). New South Wales (NSW) has the largest population, about 7 million inhabitants (see Chapter 2 for further details on New South Wales).

1.3 Equity of Resource Allocation in New South Wales: The Research Issue The NSW Health Department has sought to address geographical inequities in funding through the application of a Resource Distribution Formula (RDF) designed to eliminate geographical inequities in funding among Area Health Services (NSW Health, 2005, NSW Health, 1996). The RDF distributes funding to the eight Area Health Services (AHSs) on the basis of several indicators that reflect the health needs of populations in each area. Once these funds are received, the AHS authorities are empowered to determine local priorities and distribute funds according to these locally defined priorities. According to NSW Health, this arrangement ensures that AHSs ‘have the maximum flexibility to be able to move funds to provide a continuum of care to patients and to fully integrate services’ (NSW Health, 2006c). A few programs (such as mental health) receive from NSW Health earmarked fund allocations which cannot be redistributed by AHS authorities.

After nearly two decades of utilising the RDF, the NSW Health Department contends that it has been largely successful in reducing geographical inequities in funding. According to the Health Department, this is illustrated by the reduction in the proportion of the health budget that needs to be reallocated to achieve geographical equity as captured in the following policy statement and Figure 1–1:

6 ______“Since the adoption of the RDF approach in the late 1980s, considerable progress has been made in reducing the disparities in funding across NSW. In 1989–90 approximately 13.8% of the health budget needed to be reallocated to achieve equity in funding. By 1994–95 this figure was reduced to 11.0% and by 1998–99, it was 5.0%.... in 2004–05 about 3% of the health budget was required to be reallocated to achieve equity in funding…the weighted average distance from target for the 8 Area Health Services is around 1.8% as previously under– resourced Area Health Services have been combined with better resourced Areas” (NSW Health 2005: p 5).

Figure 1–1: NSW Area Health Services Weighted Average Distance from RDF Target

16.0% 13.8% 14.0% 11.0% 12.0%

10.0%

8.0% 5.0% 6.0%

4.0% 3.0% 1.8% 2.0%

% Average Distance From RDF Targe RDF From Distance Average % 0.0% 1989-90 1994-95 1999-00 2004-05 2006-07 Years

Weighted Average Distance From RDF Target

Data Source: NSW Health, RDF Technical Paper 2005 Revision

The NSW Health Annual Report 2006/07 further reaffirms that in 1989/90, AHSs were on average 14% away from their RDF target and, with a greater share of growth funding allocated to historically under–funded population growth areas, the average distance from target was less than 2% in 2006/07 (NSW Health, 2007a). A search of the literature has revealed only two articles that also confirm that the RDF has achieved its objective of reducing geographical inequities in funding (Gilbert et al., 1992, Gibbs et al., 2002). Despite the claims that the RDF has achieved its objective by reducing

7 ______existing inequities in geographical funding, no independent research has been conducted, to date, to assess the extent to which geographical equity in the allocation of health funds to AHSs has indeed been achieved.

Another key issue is how the achievement of equity at inter–AHS level (if indeed equity has been achieved) has impacted on the distribution of funds at the AHS level. Although AHSs report annually to the NSW Health Department on how the funds received are utilised, the reporting system does not give any indication of how equity has been pursued in the distribution/utilisation process and who has benefited from the resources.

Indeed, it remains largely unknown how the Area Health Services’ (AHSs) authorities distribute resources and how much credence they give to equity in the process. This dissertation is designed to fill the current gaps in knowledge regarding the extent to which equity in resource allocation under the NSW Health Resource Distribution Formula has been achieved at the inter–AHS level and how far the pursuit of equity is reflected in allocation of resources at the Area Health Service (intra–AHS) level.

1.4 Aims and Research Questions The overall aim of this study was to examine whether there has been movement towards equity in resource allocation to Area Health Services under the NSW Health Resource Distribution Formula (RDF) and, if there has, whether this has been reflected in equitable resource allocation within Area Health Services. Equity is defined in this study as equality of access to financial resources for health care on the basis of health need. Health need is defined in terms of population health status.

The research questions addressed by the study were:

 To what extent has resource allocation to Area Health Services from the NSW Department of Health been equitable?  How are the financial resources received from the State level distributed at the Area Health Service level?

8 ______ Is equity reflected in the allocation of NSW Health resources at the Area Health Service level?  What prospect is there for allocation of NSW Health resources both at the State and Area Health Service levels to promote equity?

1.5 Overview of Methods The study used a combination of qualitative and quantitative methods to address the research questions. The qualitative method involved key informant interviews with a range of stakeholders including policy makers, AHS executives and health program managers involved in resource allocation decision–making. The interviews explored the processes for resource allocation, the consideration given to equity, and the factors that limit the promotion of equity among other things.

The quantitative method involved the development and use of different approach to assess the extent of equity in resource allocation at both the State and AHS levels. This included the use of socio–economic and health–related data to develop indicators for assessing health needs and equity, the analysis of expenditure data from the State Health Department (2003/04 to 2006/07) and AHS authorities, and the mapping of health services available in local government areas.

1.6 Outline of Thesis The thesis is presented in 10 chapters. Chapter 1 provides an overview of the research issue and the rationale for the study, including the study aims and the research questions explored. Chapter 2 focuses on the NSW health care system, starting with organisation and financing of health care in Australia generally before narrowing down to New South Wales. Chapter 3 is a comprehensive review of the literature relating to concepts and definitions of equity and health needs in the context of resource allocation. Chapter 4 presents a review of the literature concerned with resource allocation mechanisms. It includes a review of the various approaches used for resource allocation in the NSW health system. Chapter 5 presents the research methods employed in this study including approaches used to address each of the research questions and detailed

9 ______information on the tools used for assessing equity of resource allocation at both the State and Area Health Service levels.

Chapter 6 relates to the first research question, presenting the findings of an examination of the extent of equity in the allocation of resources from the State level to the eight Area Health Services in New South Wales. It is divided into two sections: section 1 presents the qualitative findings and section 2 the quantitative. Chapter 7 presents the findings of the inquiry relating to the second research question, exploring how Area Health Services allocate the funds they receive from the state level. In Chapter 8 are the findings relating to the third research question, reporting on the case studies of equity of resource allocation within two Area Health Services. Chapter 9 discusses the findings of the study in the light of the last research question – the prospect for resources to be allocated to further promote equity in the NSW health system. The final chapter (Chapter 10) highlights the key policy implications of the findings and makes recommendations for improving the current situation in NSW, and concludes the study by reflecting on the contributions it makes in filling current gaps in knowledge on resource allocation and equity in the health sector.

10 ______

Chapter 2: THE NEW SOUTH WALES HEALTH CARE SYSTEM

Overview The goal of health care systems worldwide is to provide services to improve the health status of the populations they serve. Achieving this goal, however, depends on several factors, including the way the health system is organised, how it is managed and financed, and the mechanism by which resources are allocated to meet the health needs of the population. This chapter reviews some of the important contextual issues that influence the financing and delivery of health care in New South Wales, including some of the recent reforms implemented to promote equity of resource allocation. It begins with a brief review of health care financing and delivery in Australia.

2.1 Introduction The Australian population is among the healthiest in the world with life expectancy at birth estimated at 79 years for men and 84 years for women (United Nations Development Programme, 2007). However, considerable inequities in health status persist with the life expectancy of the country’s indigenous population lagging about 17 years behind that of non–indigenous Australians (Australian Institute of Health and Welfare, 2007b). There are also considerable differences between rural and urban populations, particularly in terms of access to health services. People living in rural and remote areas generally have limited access to health services when compared to their metropolitan counterparts. This is reflected in rural and remote residents having higher rates of premature mortality from all causes, lower life expectancy, higher hospitalisation rates for some causes of ill health, and lower survival rates for cardiovascular diseases and cancer (Australian Institute of Health and Welfare, 2006).

In general, the high life expectancy of Australians has contributed significantly to the country’s ageing population. According to the Australian Bureau of Statistics, the proportion of the population aged over 65 years is projected to increase from the current 11 ______13.3% to around 25% or 6.2 million people by 2042 (Australian Bureau of Statistics, 2006a). The ageing population has significant economic and social implications, including implications for financing and delivery of health care.

2.2 Financing and Delivery of Health Care in Australia Responsibility for Australia’s health care financing and delivery is divided between the different levels of governments (Commonwealth, State and Local governments) and the private sector. Australia’s national public health insurance scheme, Medicare, is financed and administered by the Commonwealth government and consists of three health care components: medical services, which include visits to the general practitioners (GPs) and other medical practitioners; prescription pharmaceuticals; and hospital treatment as a public patient. The Commonwealth, State and Territory governments jointly fund public patient hospital treatments. They also fund and deliver a range of other health services, including population health programs, community health services, health and medical research and high–level residential aged care (Australian Bureau of Statistics, 2008a).

In terms of health expenditure the Commonwealth government contributes the largest share, followed by the Non–government sector and State and Local governments. In 2006/07 for example, the Commonwealth government’s contributions were estimated to be around 42%, while the State and Territory governments’ input was 26% and the Non– government sector (Insurance funds, Individuals and other) contributions were 31% (see Figure 2–1).

12 ______Figure 2–1: Australia Health Expenditure by Sources of Funds, 2006–07

**Other, 7%

Individuals, 17% Commonwealth Government, 42%

Health Insurance Funds, 7%

State and Local Governments, 26%

Commonwealth Government State and Local Governments Health Insurance Funds Individuals Other**

Note: ‘**Other’ includes expenditure on health goods and services by Workers’ Compensation and compulsory third–party motor vehicle, as well as other sources of income (for example, interest earned) for service providers. Source: Australian Institute of Health and Welfare health expenditure report, 2008

Figure 2–2 shows Commonwealth, State and Local governments and non–government sector (private health funds, individuals and other) health expenditure from 1999/00 to 2006/07 (Australian Institute of Health and Welfare, 2008). However, Commonwealth government contributions were much higher throughout the eight years compared to the States/Territories and Local governments and the Non–government sector.

In 2006/07, the total health expenditure was about $94 billion or $4,507 per person. This accounted for about 8.98% of GDP and was an increase of around 1.3% from the 7.72% spent ten years before in 1996/97 (Australian Institute of Health and Welfare, 2007a). The 9% of GDP Australia spent on health in 2006/07 was almost at par with the proportions of GDP spent by many OECD countries, including Italy, New Zealand and Norway. It was, however, higher than the proportion spent in the UK and lower than health spending in the United States which accounted for 15% of GDP in 2006/07 (Australian Institute of Health and Welfare, 2008).

13 ______Figure 2–2: Health Expenditure by Sources of Funds in Australia, 1999–00 to 2006–07

45,000

40,000

35,000

30,000

25,000

20,000

15,000

Health Expenditure $M Expenditure Health 10,000

5,000

0 1999–00 2000–01 2001–02 2002–03 2003–04 2004–05 2005–06 2006–07 Years

Australian Government State/territory and Local Government Non-government Sector

Data Source: Australian Institute of Health and Welfare health expenditure report, 2008

2.2.1 Financing of Health Care: Commonwealth Government The Commonwealth government is primarily responsible for health service funding, regulation of health products, services and workforce and national health policy leadership. The responsibilities of the Commonwealth government include the following:

 Medicare Benefits Schedules (MBS) component of Medicare that provides rebates to private patients for medical services provided by privately practising doctors, optometrists and other allied health practitioners;  Pharmaceutical Benefits Scheme (PBS) component of Medicare that provides rebates to private patients for a wide range of prescription pharmaceuticals;  Australian Health Care Agreements component of Medicare that provides grants to state and territory governments for the provision of free hospital treatment as a public patient;

14 ______ Public Health Outcomes Funding Agreements that provide broad–banded and specific purpose funding from the to the states and territories for a range of public health programs;  Rebates for private health insurance premiums that subsidise access to a range of ancillary health services and treatments as a private patient in hospital;  Grants and payments to government and non–government health service providers for a range of health services (e.g. radiation oncology, pathology and medical services) to improve services, and to improve the quality of service and health outcomes; and  Health services for war and defence service veterans provided under a number of schemes administered through the Department of Veterans’ Affairs, including the Local Medical Officer Scheme, the Repatriation Pharmaceutical Benefits Scheme, and the Repatriation Private Patients Scheme (for treatment as a private patient in hospital).

Medicare is financed from taxation through a 1.5 percent levy on taxable income introduced in 1984. In 2004/05, for example, the Australian Taxation Office estimated that the revenue raised from the Medicare levy was around $6.1b, which represents about 17.2% of total Australian Government health expenditure for that year (Australian Bureau of Statistics, 2008a). The principal objective of Medicare is to remove or reduce financial barriers to accessing health care by all Australians. Thus, under Medicare, the Commonwealth has a responsibility to ensure universal access to health care (Peacock and Segal, 2000). Medicare benefits include access to public hospital care at no charge and cash benefits related to costs of private medical services (Duckett, 2004). Other Commonwealth benefit programs provide subsidies for nursing home care and prescription medicines supplied by private pharmacists. Medicare and PBS are administered through the Health Insurance Commission (HIC) (Australian Institute of Health and Welfare, 2007a).

Apart from the direct expenditure on health programs such as Medicare and PBS, the Commonwealth government provides Specific Purpose Payments (SPPs) to the states and territories for health activities. These are grants to states and territories provided under the

15 ______Australian Health Care Agreements (AHCAs) between the two levels of government (Commonwealth and State). The payments are primarily directed to expenditure on public hospital services in the states and territories (Australian Institute of Health and Welfare, 2008). In return for this funding, state and territory governments abide by a number of conditions. For example, they have the duty to provide a network of hospital services and to allow all patients/consumers to be able to access inpatient services in public hospitals free of charge. Thus AHCAs are important from the point of view of equity in hospital service provision. The state and territory governments are also responsible for the full marginal cost of any increase in hospital budgets during the term of the agreement (Australian Institute of Health and Welfare, 2007a).

The overall funding arrangement between the Commonwealth and State/Territory governments is complex, and to a great extent, characterised by political buck–passing and blame–shifting (Buckmaster and Pratt, 2005). The Medicare funding, for example, is based on a formula unrelated to actual hospital budgets and is adjusted only for exogenous factors such as population growth and ageing. This places strong incentives on states to achieve efficiency improvements such as through the introduction of case–mix funding of hospitals with a consequent reduction in demand, or reducing the provision of hospital facilities (Duckett, 2004).

There have been growing calls for health financing reforms in Australia in the last decade, including reform to Medicare and the PBS (National Health and Hospital Reform Commission, 2008). Although Medicare enjoys widespread political support, it has come under increasing pressure as a result of the changing demographics, the growing burden of chronic diseases, rising health care costs associated with medical advances and better technology, higher public expectations (Duckett, 2008a, Butler) and increases in patients’ co–payments/out–of–pocket payments (costs to patients that are not rebated by Medicare, private health insurance or other sources).

In 2009, for example, Doggett noted that over the last decade there has been a trend in increasing co–payments for health services in Australia and that policy makers have

16 ______neglected the co–payment increases despite the fact that co-payments are ‘inequitable, discriminate against patients with certain types of health care needs or living in particular geographic areas, creates barriers to accessing cost–effective health care, is confusing to both patients and providers, results in perverse incentives in the use of health care, is complex and expensive to administer, typically imposes the highest costs on consumers when they have the least ability to pay and does not support the diversity of patient health care needs or promote patient choice’ (Doggett, 2009). She concludes that with rising advanced and expensive technologies and treatments, the burden of health care costs to individuals is likely to continue to increase, threatening the overall equity and efficiency of Australian health system.

A House of Representatives Standing Committee on Health and Ageing inquiry into health funding in Australia detailed some of the challenges the health system faces in its recent report ‘The Blame Game’. The challenges include duplication of services and waste of resources as a result of the division of funding responsibility between Commonwealth and State governments, cost shifting, the interdependence of the public and private sectors (which leads to some confusion about the role of private insurance in the context of a compulsory, universal health insurance scheme such as Medicare), funding arrangements that have a bias towards treating illness instead of preventing it, lack of financial incentives to deliver high quality and safe health care, and poor continuity of care when multiple health care providers are involved (House of Representatives Standing Committee on Health and Ageing, 2006). Figure 2–3 illustrates the complex structure of the Australian health care system flow of funding between the government and private sectors, and the providers of health goods and services. The next section looks at the States’ and Territories’ contributions to health care funding.

17 ______Figure 2–3: The Structure of the Australian Health Care System and Major Channels of Funding Flows

Source: Australian Institute of Health and Welfare, ‘Health Expenditure in Australia 2003–04, AIHW Cat No. HWE 32, Canberra, with author adjustments

2.2.2 Financing of Health Care: State, Territory and Local Governments States and territories in Australia have a core responsibility for funding and delivery of public health services (public hospitals, community health and public dental care) and the regulation of health care providers and private health facilities. The health portfolio is one of the most important State government portfolios in both political and fiscal terms. It accounts for about one–third of State recurrent budgets. Each State has a minister responsible for health, as either a major portfolio in its own right, or the largest component

18 ______of a broader “human services” portfolio, which might include related areas such as aged and community care (Mooney and Scotton, 1998, Australian Institute of Health and Welfare, 2006). The States essentially are autonomous in administering health services within the constraints of their own legislation and agreements with the Commonwealth government. Their responsibilities include the following:

 Funding and administering public acute care and psychiatric hospitals;  Funding and providing a wide range of community and public health services, including school health, limited dental services, maternal and child health, occupational health and disease control activities, and health promotion;  Registration of health professionals;  Licensing of public and private hospitals; and  Making health–specific payments to local governments.

The tradition of “federalism” in Australia means that the health care system has developed somewhat differently in each State, with variations in size, organisational structures, and utilisation rates (Healy et al., 2006).

Table 2–1 shows the average per capita allocation of recurrent health expenditure for 10 years from 1997/98 to 2006/07 for the states and Australia in general. In 2006/07 for example, allocations in three states, Victoria, Queensland and Tasmania, were slightly lower than the national average (of $4,185), while those of New South Wales, the Northern Territory, Western and South Australia were higher than the national average. Although the differences were not great, they highlight the variations in health per capita spending across the states in Australia.

19 ______Table 2–1: Average Recurrent Health Expenditure per Person (in current price) for States and Territories in Australia, 1997–2007

Year NSW Victoria Qld WA SA Tas NT Australia 1997–98 2,258 2,276 2,190 2,199 2,143 2,353 2,685 2,243 1998–99 2,470 2,383 2,290 2,257 2,349 2,409 2,731 2,391 1999–00 2,571 2,577 2,450 2,428 2,564 2,614 3,040 2,552 2000–01 2,772 2,898 2,723 2,702 2,773 2,839 3,215 2,796 2001–02 2,992 3,149 2,850 2,907 2,994 3,336 3,423 3,013 2002–03 3,197 3,405 3,032 3,241 3,319 3,168 3,832 3,243 2003–04 3,480 3,461 3,178 3,470 3,582 3,221 4,208 3,432 2004–05 3,778 3,756 3,406 3,763 3,866 3,444 4,532 3,713 2005–06 3,975 3,943 3,683 3,899 4,070 3,707 4,916 3,921 2006–07 4,225 4,156 4,025 4,212 4,267 3,988 5,282 4,185 NSW = New South Wales; Vic = Victoria; Qld = Queensland; WA = Western Australia; SA = South Australia; Tas = Tasmania and NT = Northern Territory Source: Australian Institute of Health and Welfare health expenditure database 2006/07 published in 2008

2.2.3 Financing of Health Care: Non-Government Sources In Australia, non–government funding accounts for around one–third of all health expenditure. It includes private health insurance expenditures and out–of–pocket2 payments by individuals. Figure 2–4 shows non–government health expenditure from 1995 to 2006. As is evident from the figure, the bulk of the expenditure over the 10 years came from out– of–pocket payments. For most of the years, the out–of–pocket contributions to health system finance were estimated to be over $15 billion a year.

2 Out–of–pocket (co–payments) are the ‘gap’ payments that patients pay directly for health and medical care which are not rebated by Medicare, private health insurance or other sources. An out–of–pocket payment can be a small amount of the overall cost of accessing services to purchasing medications or the absolute cost of services and medications (Doggett, 2009).

20 ______Figure 2–4: Health Expenditure of Non–Government Sector by Sources of Funding, 1995–2006

20.0

18.0

16.0

14.0

12.0

10.0

8.0

6.0

4.0

2.0 Private Sector Health Expenditure (%) 0.0 1995–96 1996–97 1997–98 1998–99 1999–00 2000–01 2001–02 2002–03 2003–04 2004–05 2005–06 Years

Health Insurance Funds (%) Individuals (%) Other Non-Government (%)

Data Source: Australian Institute of Health and Welfare, 2006–07

The distinction between the public, private not–for–profit and private for–profit sector hospitals is increasingly blurred. For example, many public hospitals contract out tasks to private providers and admit private patients. Other privatisation permutations include a State government contracting with the private sector to finance the construction of a new hospital, contracting with a private hospital to run a public hospital on behalf of the State or contracting with a private hospital to provide some services for public patients.

The majority of doctors in Australia are engaged in some level of private practice (Segal and Bolton, 2009, Australian Institute of Health and Welfare, 2008). Private General Practitioners (GPs) provide most primary care while private medical specialists provide most ambulatory secondary health care, but they also may contract their services to public and to private hospitals. Private doctors are key stakeholders in health sector governance and have a major influence upon health care policies (Australian Institute of Health and Welfare, 2006).

21 ______Private health insurance funds are significant players in the health –financing environment. The current Commonwealth policy is to support the private health insurance industry, which is heavily subsidised by a tax rebate: 30% for general population, 35% for people aged 65–69 years, and 40% for people aged 70 years and over on premiums (Australian Bureau of Statistics, 2008a). As at 30 June 2007, there were 41 registered private health fund organisations, but six funds dominate the health insurance industry with a combined share of 76% of the market (measured by premium income) (Private Health Insurance Administration Council, 2007). The two largest are Private (which was separated from the Health Insurance Commission in 1997 to become a government business enterprise) and the Medical Benefits Fund of Australia. The private health insurance industry is regulated by a statutory authority, the Private Health Insurance Administration Council, under the regulatory framework set out in the National Health Act 1953 and the Health Insurance Act 1973. A private insurance fund must be a Registered Health Benefit Organisation and its activities are tightly controlled; for example, insurers must accept all applicants and must not discriminate in setting premiums and paying benefits (Australian Institute of Health and Welfare, 2008).

As of June 2007, almost half of all Australians had private health insurance. Forty–four per cent of the population (over nine million Australians) were covered by hospital insurance for treatment as private patients in both public and private hospitals and had ancillary cover for non–medical services provided out of hospital, such as physiotherapy, dental treatment and the purchase of spectacles (Australian Bureau of Statistics, 2008a).

2.2.4 Delivery of Health Care in Australia The State and Territory governments are the main providers of health services in the public sector in Australia. These health services are financed by a combination of Special Purpose Payments from the Commonwealth government, funding by the States and Territories out of their own fiscal resources, and funding from non–government sources (usually in the form of user fees). State and Territory governments also provide or purchase ambulance services, public dental care and community health services, for which they provide most of

22 ______the funding. Further, they are a major source of public health activities such as prevention and control of infectious diseases, including health promotion campaigns (Australian Institute of Health and Welfare, 2008, Australian Bureau of Statistics, 2008a).

It has been argued that health service structures and patterns in the States are converging, given the common pressures of cost containment, indigenous health, quality of services, access and equity issues, including access in remote and rural Australia, and health workforce shortages (Segal and Bolton, 2009). State health departments have been reorganised many times, depending upon the governing political party. During the 1970s, the separate administrations for hospitals, community health and mental health generally were amalgamated into “health” and “community” functions (Duckett, 2004). In the next phase of restructuring, “super departments” were created that incorporated most aspects of health and community services, but recently some States, for example South Australia, returned these functions to separate departments.

All State health departments have undertaken major reviews of their policies, structures and programmes in the last few years. Some of these reviews were triggered by inquiries into allegations of clinical incompetence in hospitals (Wilson and Van Der Weyden, 2005), while others were motivated by the desire to reduce the growth of health spending and to improve quality and safety for patients.

In all states, local governments are responsible for some public health services, including public health surveillance, but not clinical medical services. They also undertake environmental health activities such as collecting rubbish and monitoring food standards. In addition, LGAs are often involved in disease prevention efforts such as immunisation programmes, supporting maternal and child health screening centres, and some health promotion activities (Duckett, 2004). The level of involvement of local governments in health delivery, nonetheless, varies across states. Victorian local governments, for example, are the most active in health and welfare services delivery.

23 ______The private sector plays an active part in health care delivery in all states and territories. Although doctors in private practice deliver GP and specialist medical services, this activity is in fact heavily subsidised by the Commonwealth government through Medicare. Private medical practitioners provide most out–of–hospital medical services, while specialist doctors deliver a large proportion of hospital services. Private practitioners provide most dental services and allied health services such as physiotherapy. Private hospitals are significant players in the hospital field, with a total of 547 private hospitals providing about 30% of the bed supply in Australia. The number of private hospitals grew after the introduction of Medicare in 1984, remained fairly constant in terms of hospitals and beds in the first half of the 1990s, and expanded their capacity in the late 1990s (Australian Bureau of Statistics, 2008a). Private hospitals generally are smaller than public hospitals. Therefore, they deal with a more limited range of cases, rarely offer emergency services, and undertake a substantial amount of elective surgery. The growth of larger corporate players has given the private hospital sector greater negotiating power, with over two– thirds of all private hospital beds now owned by large for–profit chains and the Catholic Church (Healy et al., 2006, Duckett, 2004). The diagnostic services industry is one of the private venture areas that grew considerably during the 1990s, with the expansion of pathology services and diagnostic imaging, and includes increased corporatisation, with mergers between companies and public listings on the Australian stock exchange (Foley, 2000).

2.3 The New South Wales Health System

2.3.1 Geography and Population

New South Wales is located on the east coast of Australia bordering the state of Queensland to the north, Victoria to the south, South Australia to the west and the Pacific Ocean to the east. It covers a total land area of 800,725 square kilometres and has three large urban centres. The population is close to 7 million people (2% constitutes Indigenous people) making New South Wales (NSW) Australia’s most populous state. The state’s population has been growing at about 71,900 people (1.1%) since June 2006. The largest component of this population growth was due to net overseas migration estimated to be

24 ______around 54,900 people. Natural increase (births minus deaths) accounted for a further increase of 44,300 people. During the same period, NSW recorded a net loss of 27,300 people to other states and territories (Australian Bureau of Statistics, 2008a). Approximately three–quarters of the state’s population live in the three main cities of Sydney, Newcastle and Wollongong and along the coast. The rest of the population is sparsely distributed across many small country towns, few of which exceed 20,000 in population (Australian Bureau of Statistics, 2007a).

In terms of demographic attributes, New South Wales is closer to the national average than most other States and Territories in Australia. For example, in 2006, about 19.8% of the population were children under the age of 15 compared to the national average of around 19.6%, while over 15% were above 65 years compared to the 14.6% national average (Australian Bureau of Statistics, 2007a). The population aged 65 years or older continues to expand as more ‘baby boomers’ enter retirement years. The median age of the NSW population has also continued to increase estimated at around 37.0 years in 2006, an increase of 2.2 years since 1997 (Australian Bureau of Statistics, 2008b).

Of the estimated 7 million people in New South Wales, 148,200 identify themselves as Aboriginal and . This is about 2% of the total population (Australian Bureau of Statistics, 2006b). The Indigenous population in NSW, as in other states, is younger than the rest of the population. In 2006, for example, over one–third (38%) of the Indigenous population were aged 0–14 years, twice the proportion recorded for non– Indigenous children (19%). In contrast, just over 3% of Indigenous people were aged 65 years or older compared with 14% of non–Indigenous people. As a result, there is a large difference in the median ages of these population groups. The median age for Indigenous people in NSW was 20.7 years in 2006, an increase of 0.8 years since 1996. This is significantly younger than the 37.2 years recorded in 2006 for non–Indigenous people (Australian Bureau of Statistics, 2008b).

25 ______2.3.2 Health Status and Inequalities The health status of the people in New South Wales, as in other parts of Australia, has improved markedly in recent decades. This is reflected in the high rate of life expectancy at birth of many NSW residents. Recent estimates put the average life expectancy rate for men at 79.3 years and 84.2 years for women. This is an improvement of 6.1 years for males and 4.6 years for females over the past 20 years (NSW Health, 2008, Australian Bureau of Statistics, 2009). The high health status, however, does not diminish the seriousness of the health challenges faced by the state. In line with Australian generally, cardiovascular diseases including coronary heart disease and stroke are a major problem, accounting for about 40 percent of deaths in 2006. Other diseases such as cancer (27%), chronic respiratory diseases (7%), nervous system diseases (5%) and diseases of the digestive system remain a key problem in NSW (NSW Health, 2008, NSW Health, 2006a). The high incidence of these diseases and other problems such as unintentional injuries and pregnancy and childbirth–related conditions have led to an increase in hospitalisation in the past decades. In 2002–03, for example, hospitalisation cases were over 2 million, representing a 34 percent increase in the age–adjusted rate of hospitalisation since 1989–90 (NSW Health, 2006a, Centre for Epidemiology and Research, 2007).

A range of health behaviours has been identified as affecting the health of the population in New South Wales. These include smoking, excessive alcohol consumption and lack of adequate physical activity. There is evidence, for example, that in 2004, 42 percent of the population aged 16 and over were smokers. Similarly, the overweight and obese population aged 16 years and over was 48.3 percent and those not engaging in adequate physical activity were 52.2 percent in 2004 (NSW Health, 2006a, Centre for Epidemiology and Research, 2007).

Another issue of major concern in NSW and Australia as a whole is the inequalities in health status. Despite the improvement in health status and life expectancy, significant gaps exist among population groups. In general, factors such as gender, geographic region, socioeconomic disadvantage, occupation and country of birth underpin the variations in health status in the State (Draper et al., 2004, Centre for Epidemiology and Research,

26 ______2007). People who experience social and economic disadvantages tend to be sicker and die younger than others. This is most starkly evident for Indigenous Australians, who are likely to die prematurely from avoidable conditions (see Figure 2–5). In 2006, for example, 75% of all premature Aboriginal deaths were potentially avoidable3, compared with 65% of non–Aboriginal premature deaths in NSW (NSW Health, 2006a). As demonstrated by Figure 2–5, the rates of premature deaths among Aboriginal people decreased between 2000 and 2005 but increased again in 2006. The rate of premature deaths among non– Aboriginal people has declined over time.

Figure 2–5: Premature Deaths by Aboriginality, NSW 1998 to 2006

700.0 637.2 601.7 610.0 565.8 600.0 559.3 540.7 547.3 510.6 498.7 500.0

400.0 308.1 301.3 288.2 273.5 300.0 269.8 262.1 253.7 241.2 233.7 200.0

Premature Deaths/100,000 Pop. Deaths/100,000 Premature 100.0

0.0 1998 1999 2000 2001 2002 2003 2004 2005 2006 Years

Aboriginal Persons Non-Aboriginal Persons

Data Source: NSW Health, Chief Health Officer Report, 2006

Aboriginal people are also disadvantaged across a range of socioeconomic factors that impact on their health, including living in rural or remote NSW (Australian Institute of Health and Welfare, 2007b). Although the majority of NSW Indigenous people live in

3 According to Page et al (2006), deaths classified as potentially avoidable are divided into those causes of premature death that are potentially preventable by preventing the condition occurring in the first place through primary prevention and those deaths that can be avoided by health care interventions and hence are amenable to these interventions.

27 ______cities and towns, they still account for a high proportion of the population in some rural and remote areas. According to the Australian Bureau of Statistics, around a quarter of the Indigenous population in NSW live in areas classified as “remote” or “very remote” (Australian Bureau of Statistics, 2006b).

As Table 2–2 shows, the average life expectancy for Indigenous males born between 2000 and 2006 was 60 years and 65.1 years for females compared to about 79 years and 84 years respectively for non–Indigenous males and females. This leaves an average gap of about 19 years between Indigenous and non–Indigenous Australians in NSW (higher than the national gap of about 17 years) (Australian Bureau of Statistics, 2006b). The Indigenous life expectancy is comparable to several developing African countries, such as Ghana with an average life expectancy of 58.5 years, and worse than Senegal with 61.6 years on average (United Nations Development Programme, 2007). It highlights the vast inequalities in health and other social determinants of health among the New South Wales population.

Table 2–2: Gap in Life Expectancy between Indigenous and Non–Indigenous Populations in New South Wales Population groups in NSW – 2008) Male Female

Non–Indigenous 79.3 84.2 Indigenous 60.0 65.1 Life expectancy gap 19.3 19.1 Data Source: NSW Health Chief Health Officer Report, 2006

While the poor health status of Indigenous people in NSW is not disputed, the precise extent of the health disadvantages and whether they are improving has been hard to measure. This is partly because Indigenous people are not well accounted for in census and/or administrative records in Australia due to the practical and statistical challenges of surveying a small population with a relatively high remote area presence and the historical discriminatory policies which made it difficult for Indigenous people to identify or register themselves as Indigenous Australians (Australian Institute of Health and Welfare, 2007b). Although these policies have been reversed and the current government has apologised to

28 ______the Indigenous people for past mistreatments, the general picture of poor health status persists and many believe it will take decades to ameliorate.

Finally, inequalities in access to health care when needed are also a major concern for the NSW population, reflecting not only the pattern of illness but also differences in the availability and accessibility of health care services. In 2007, for example, 17% of the NSW population reported having difficulties accessing health care, according to the NSW Population Health Survey published in 2007. Compared with the State average (17%), urban –based Area Health Services (AHSs) residents were less likely to report difficulties accessing health care. For example, Sydney South West AHS reported (12.2%), South Eastern Sydney & Illawarra (13.0%), Sydney West (13.4%), and Northern Sydney Central Coast (11.3%).

On the other hand, rural and remote Area Health Services: Hunter and New England, North Coast, Greater Southern and Greater Western residents were, not surprisingly, more likely to report difficulties accessing health care, ranging from 24.5% to 29.9%. The inequalities in access to health care as reported by NSW residents increased with remoteness. Approximately one–third (33.1%) of people living in outer regional and 38.3% of people living in remote and very remote Area Health Services reported difficulties accessing health care when needing it. Only 12.1% of people living in major cities reported such difficulties (see Figure 2–6). The most frequently reported difficulties were waiting times for a (GP) appointment, difficulty in accessing specialists, and transport problems (Centre for Epidemiology and Research, 2007)

29 ______Figure 2–6: Difficulty Accessing Health Care When Needed by Area Health Services for Persons Aged 16 Years and Over, NSW 2007

45.0% 40.0% 38.3% 35.0% 33.1% 29.4% 29.8% 29.9% 30.0% 24.5% 25.0% 23.6%

20.0% 17.0% 15.0% 13.0% 13.4% 12.2% 11.3% 12.1% 10.0% 5.0% 0.0%

NSW NCAHS GSAHS SSWAHS SWAHS GWAHS SES&IAHS NSCCAHSH&NEAHS Major cities Pop. having difficulty accessing health care (%) Inner regionalOuter regional

Area Health Services Remote and very remote

Persons Having Difficulty Accessing Health Care (% )

Data Source: NSW Health Chief Health Officer Report, 2008

2.3.3 The NSW Department of Health The NSW Department of Health has the objective of working to provide the people of NSW with the best possible healthcare, including monitoring the performance of the public health system (NSW Health, 2007a). A key vision of the Department as stated in their ‘Healthy People 2010 – The Population Health Plan for NSW’ policy document is to ‘ensure that all employees of the Department of Health work together to achieve better health for the people of New South Wales – now and in the future’. The goals of the Department include the following: keeping people healthy, providing the health care that people need, delivering high quality services and managing health services well (NSW Health, 2007b).

The key responsibilities of the Health Department include formulating policies aimed at promoting, protecting, maintaining, developing and improving the health and well–being of

30 ______the people of NSW and allocation of funding to the eight Area Health Services (AHSs) in charge of the day–to–day delivery of health care in the state (NSW Health, 2007a). The Department also exercises oversight responsibilities over the health delivery organisations in NSW, including the AHSs, the Corrections Health Service (which mainly provides health care to inmates of correctional facilities in the State), the Ambulance Service of NSW, and the Westmead Children’s Hospital (NSW Health, 2007a).

2.3.3.1 Organisational Structure The New South Wales government exercises control over health care through the NSW Minister for Health. The Minister’s responsibilities are specified in the Health Administration Act 1982, which defines the structure and functions of the Health Department, and in the Health Services Act 1995, which defines the service delivery elements, including the Area Health Services and various affiliated bodies operated by charitable and religious organisations (NSW Health, 1996).

The executive and statutory roles of the NSW Health Minister are supported by two assisting Ministers; they have the core responsibility for the Cancer and Mental Health Services portfolios. Under the two assisting Ministers is the Director–General who chairs the NSW Department of Health Management Board. The role of the Management Board is to consider and make decisions on issues of department and health system–wide interest, including budgetary decisions, development of health policy, and monitoring of health system performance (NSW Health, 2009).

As illustrated by Figure 2–7, the office of the Director General provides high–level executive and administrative support to the Director General across a broad range of issues and functions. It works closely with the Deputy Director General and members of the NSW Health Executive to ensure the Director General receives advice that is accurate, timely and reflects a cross agency view on critical policy and operational issues (NSW Health, 2009).

31 ______Another key unit in the organisational structure of the NSW Health Department is the office of Internal Audit. This office provides financial and compliance auditing as well as assurance services to Branches and key functions of the Department of Health. It undertakes special investigations of matters within the Department as referred by the Minister, the Director General, the NSW Auditor–General, the Ombudsman or the Independent Commission Against Corruption (ICAC). In addition, it provides specific audit, review and advisory services on information systems across NSW Health (NSW Health, 2007a).

Figure 2–7: New South Wales Health Organisational Chart, 2008

Source: NSW Health 2009, with adjustments by author

32 ______According to NSW Health (2007), the Department has values that govern the entire health care system. These include achieving an equitable health system, and accountability and effectiveness through pursuit of quality outcomes. For example, by the year 2010 NSW Health aims to have achieved a health system that:

 has a greater focus on and investment in improving health and preventing illness while continuing to treat illness effectively, paying particular attention to reducing the health gap for communities that experience multiple disadvantages;  is focused on quality and safety, providing patients with ready access to safe and satisfactory journeys through the NSW health services and ensuring that patients and their carers are informed and involved in health care decisions and treated with respect;  helps people to access the health care they need through a network of integrated primary health and community care services across the public and private health systems;  has a greater focus on healthy ageing strategies, integrating services across different levels of government and the private sector;  engages more effectively with other government and non–government agencies, and the broader community to provide a more integrated approach to planning, funding and delivering health services to local communities and regions;  makes the most effective use of the finite resources available and manages costs, services and infrastructure effectively to meet the State’s health care needs, while maintaining financial sustainability;  has a valued skilled workforce that is available in sufficient numbers, is well trained, organised and deployed creatively to focus on the changing needs of health consumers, carers and the wider population; and  is alert and capable of readily adapting to the changing needs of the community and is quick to anticipate and respond to new issues as they emerge (NSW Health, 2007b).

33 ______2.3.4 Financing of the NSW Health System Local health expenditure in New South Wales has increased from $22 billion in 2001/02 to nearly $31 billion in 2006/07. The estimated $30.8 billion spent on health in New South Wales in 2006/07 was sourced from the (i) the Commonwealth government, (ii) the NSW State, Territory and Local governments, and (iii) the Non–government sector as illustrated by Figure 2–8. The NSW state and local government contributions were less than that of either the Commonwealth or the non–government sector as shown in Figure 2–9.

Figure 2–8: New South Wales State Health Funding by Sources of Funds, 2001/02 to 2006/07

16,000

14,000

12,000

10,000

8,000

Amount ($M) 6,000

4,000

2,000

0 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 Years

Federal Government State & Local Government Non-Government sources

Data Source: Australia Institute of Health and Welfare: 2001–2007

In 2006/07, the Commonwealth government was responsible for nearly half (43.9%) of the total health expenditure for the state of NSW; the private sector (including the health insurance funds, individuals and other non–government actors) contributed about 30.4% and the remaining 25.7% came from the state and local governments (see Figure 2–9) (Australian Institute of Health and Welfare, 2008). Thus, of the three sources of funds to meet the costs of health service provision in NSW, the NSW government contributes only about a quarter of the total health expenditure. The state, nonetheless, retains coordination

34 ______and policy oversight responsibilities as well as resource allocation, regulatory and performance monitoring roles through the NSW Department of Health. However, the NSW Department of Health does not handle a significant part of the funding that is administered by general practitioners (GPs), specialist doctors, pharmacists and many allied health workers as mentioned earlier. NSW Health also does not control major areas of health activities (see Figure 2–11, page 36).

Figure 2–9: Total Health Expenditure ($30.8 billion) in New South Wales by Area of Expenditure and Source of Funds – 2006/07

State & Local Government, 25.7%

Federal Government, 43.9%

Non-Government sources, 30.4%

Federal Government State & Local Government Non-Government sources

Data Source: Australian Institute of Health and Welfare (AIHW), 2008

Table 2–3 shows the total health expenditure by source and key expenditure items from 2001/02 to 2006/07. The health budget has been increasing gradually from around $22,428 millions in 2001/02 to $30,778 millions in 2006/07. Most of the Health budget goes to recurrent expenditure which includes salaries and employee related expenses, food, drugs, medical and surgical supplies and visiting medical staff. Other funds are spent on capital appropriation. Similarly, capital expenditure experienced growth in funding from $976 millions in both 2001/02 and 2002/03 to $1,815 millions in 2006/07 (see Table 2–3).

35 ______

As illustrated by Table 2–3, the Federal government is the main contributor followed by the non–government sector. The Commonwealth (Federal) government’s contribution to total health expenditure was around 47.7% in 2001/02 and by 2006/07, it was estimated to have decreased to 43.9%. The state and local government contributions have increased progressively from 21.3% in 2001/02 to 25.7%, while the non–government health funding contributions have remained consistent at around 31%, with minor increases or decreases, for the six years (Australian Institute of Health and Welfare, 2007a).

Table 2–3: NSW Total Health Expenditure on Recurrent and Capital and Sources of Funds, 2001/02 to 2006/07

Data Source: Australian Institute of Health and Welfare, 2001–2007

As illustrated by Figure 2–10, hospital services are allocated the largest share of the total health expenditure in NSW. In 2006/07, about 30% of the Commonwealth government’s health expenditure in NSW was allocated to public hospitals. This was the second largest expenditure item after medical services, which were allocated 33% (Figure 2–10).

36 ______Figure 2–10: Commonwealth Government Health Expenditure in New South Wales by Area of Expenditure – 2006–07

Administration, 3% Research, 4% Aids and appliances, 1.2%

Medications, 15% Public Hospitals, 30%

Public health , 3%

Community health, 0.8% Private hospitals, 4% Other health practitioners, 2% Patient transport Dental services, 2% services, 0.3%

Medical services, 33%

Public Hospitals Private hospitals Patient transport services Medical services Dental services Other health practitioners Community health Public health Medications Aids and appliances Administration Research

Data Source: Australian Institute of Health and Welfare 2007

A significantly larger proportion of funding from state and local government sources (which represent about a quarter of the health expenditure of NSW) is allocated to public hospitals. In 2006/07, about 73% of the total health funding from state and local government sources was allocated to public hospitals (see Figure 2–11).

37 ______Figure 2–11: State and Local Government Health Expenditure in New South Wales by Area of Expenditure: 2006–07

Public health , 2% Research, 1%

Community health and other, 17%

Dental services, 2%

Patient transport services, 4%

Public Hospitals, 73%

Public Hospitals Patient transport services Dental services Community health and other Public health Research

Data Source: Australian Institute of Health and Welfare, 2007

Finding the right balance for expenditure allocations between hospital and community– based services has been a concern for many years in NSW (Australian Institute of Health and Welfare, 2008). Expenditure on public health programs and services remains considerably lower despite the perceived benefits. In 2006/07, for example, the Commonwealth government expenditure on public health programs in New South Wales was a meagre 3%, but it was still slightly higher than what the NSW and local governments spent (2%) while public health expenditure in the private sector was 0.1% (see Figures 2– 11 to 2–12). The aims of public health spending in NSW are clear: to protect and promote the population’s health by controlling the determinants of health influences, social, economic, environmental, and behavioural. The question, however, is: how much funding should be spent on public health initiatives?

38 ______Figure 2–12: Non–Government Sector Health Expenditure in New South Wales by Area of Expenditure: 2006–07

Research, 1.6% Administration, 2.7% Public Hospitals, 8.5% Aids and appliances, 6.8% Private hospitals, 14.0%

Medications, 20.5%

Patient transport services, 1.7%

Public health , 0.1%

Community health, Medical services, 1.0% 16.0%

Other health practitioners, 9.3% Dental services, 17.8%

Public Hospitals Private hospitals Patient transport services Medical services Dental services Other health practitioners Community health Public health Medications Aids and appliances Administration Research

Data Source: Australian Institute of Health and Welfare, 2007

Funding from the private sector (representing nearly 30% of the total health expenditure of NSW) is largely allocated to four main expenditure items: medications, dental services, medical services and private hospitals. In 2006/07, about 70% of the total funding from the private sector was allocated to these four expenditure items with 21% going to medications, 18% to medical services, 16% to dental services and 14% to private hospitals (Figure 2– 12).

2.3.5 Delivery of Health Services in NSW Health services are delivered in NSW through NSW Health, affiliated health organisations managed by religious and/or charitable groups, a large private sector (much of which receives funding from the Commonwealth government) and a number of other non– government agencies which may receive some subsidy from the government sources. NSW Health Services are administered through eight Area Health Service authorities appointed by the NSW government. The eight AHSs are Sydney South West, South Eastern Sydney

39 ______& Illawarra, Sydney West, Northern Sydney Central Coast, Hunter & New England, North Coast, Greater Southern and Greater Western. Each AHS embraces a number of Local Government Areas (LGAs) and some AHSs have clustered LGAs for purposes of resource allocation and service provision (NSW Health, 2007a). Figure 2–13 is a map of NSW showing the eight geographically defined AHSs.

Figure 2–13: NSW Health Eight Health Service Areas – 2009

Source: NSW Health, 2009

As shown in Table 2–4, the Area Health Services (AHSs) have diverse demographic and socioeconomic characteristics. AHS populations range between 300,000 and 1,400,000 with great differences in land area. Greater Western Area Health Service (GWAHS), for example, has the smallest population of approximately 305,440 distributed across 444,586 square kilometres of land area. That gives a population density of about 0.69 persons per square kilometre of land. In contrast, Sydney South West Area Health Service (SSWAHS),

40 ______which has the largest population of over 1.3 million, has a total landmass of just 6,380 square kilometres, about 70 times smaller than the landmass of GWAHS. The population density of SSWAHS stands at about 209 persons per square kilometre (NSW Health, 2007a) .

Table 2–4: Selected Demographic, Socio–Economic and Health–Related Indicators of Area Health Services in NSW, 2007

**SEIFA–IRSD Index is the Socio–Economic Index for Areas – Index of Relative Socioeconomic Disadvantage constructed by the Australian Bureau of Statistics from the 2001 Census of Population and Housing data. SEIFA scores are derived from multiple–weighted variables, with the reference score for the whole of Australia or NSW set to 1,000. Lower scores indicate lower socioeconomic status (Adhikari, 2006).

The NSW health care system currently includes about 220 public hospitals and 15 multipurpose small facilities, mainly in rural areas, that provide a mix of hospital and community services. The community health services include public health, health promotion and some aspects of long–term community care. In addition, there are 280 community health centres, 500 early childhood health centres, and 15 nursing homes. There are 86 currently licensed private hospitals in NSW. The regulatory responsibilities with respect to the operation of private hospitals and nursing homes, and public and environmental health are undertaken by the NSW Department of Health. The Department is also responsible for state–wide policy development, planning, performance monitoring, and resource allocation as indicated earlier.

41 ______2.3.6 Area Health Services and Health Care Delivery The Area Health Services (AHSs) are responsible for the actual planning, delivery and coordination of local health services. They are also responsible for resource management and maintaining a balance between treatment and prevention services within their geographical areas. They provide public health services through community health centres and public hospitals, including community support services, domiciliary nursing and other outreach programs (NSW Health, 2007a). Other responsibilities include providing services to ‘cross–border’ or ‘inter–Area’ patients who access services outside their respective AHSs. Area Health Services are also required to provide state–wide services such as infant screening–hearing programs, severe burns services, spinal cord injury services or rescue helicopter services. Area Health Services develop their own individual service and capital development plans which are linked to state–wide health objectives and strategic directions. To a larger extent, AHSs determine the range, quantum and distribution of acute and non– acute services that they will provide to their local communities in response to the health needs of their populations (NSW Health, 2006a).

The provision of health services by AHSs is guided by rules set by the NSW Department of Health. Compared to regional authorities in other states and territories, AHSs in NSW theoretically enjoy a considerable degree of autonomy in terms of planning and delivery of services. For example, they have the flexibility to shift funding between hospital and non– hospital services, although, in practice, this might be difficult due to the method used in the distribution of resources. Area Health Services’ expenditures and outputs are reported annually against the following health programs: population health, oral health services, primary and community based services, outpatient services, emergency department services, acute inpatient services, mental health services, and rehabilitation and extended care (NSW Health, 2005, NSW Health, 2006c). These health programs are discussed in detail in Chapter 4.

The health care provided by the AHSs is complemented by a broad range of preventive and curative services provided by the private sector, including a number of non–governmental organisations. However, not all AHSs have a range of private hospitals or non–

42 ______governmental providers. Other service providers who complement the efforts of AHSs include GPs. General Practitioners (GPs) are usually the first point of contact and largely funded by the Commonwealth, not the NSW Department of Health. Since this study is mainly concerned with resources allocated through the NSW Department of Health, the funding of General Practitioners has not been reviewed.

The Area Health Services are characterised by an uneven distribution of hospital beds per head of population and of health professionals. In general there is a concentration of public hospital beds in inner Sydney, thus more doctors per 1000 of population. For example, in 2008 inner Sydney and urban based AHSs (Sydney South West, South Eastern Sydney & Illawarra, Sydney West and Northern Sydney Central Coast) had the highest number of public and private hospital beds compared to beds located in outer Sydney and remote AHSs (Hunter & New England, North Coast, Greater Southern and Greater Western). However, based on beds per 1000 population, rural and remote AHSs attracted a higher number of beds due to their low population (see Table 2–6). Similarly, urban based AHSs with large teaching and research hospitals tended to attract more medical doctors, while rural and remote AHSs had fewer doctors but more nurses (see Table 2–7).

Table 2–5: Hospital Beds in New South Wales by Area Health Services, 2008

Data Source: NSW Health Annual Report, 2007/08

43 ______Table 2–6: Health Professionals in Area Health Services in New South Wales, 2008

Note: Health professionals does not include other Professionals & Para professionals: Oral Health Practitioners & Therapists, Corporate Services, Scientific & Technical clinical support staff, Hotel Services, Maintenance & Trades and Hospital support workers. Data Source: Area Health Services Annual Reports, 2008

2.3.7 Financing of Area Health Services: A Brief Overview This section provides a brief overview of how funding of health service providers in New South Wales has evolved. A detailed discussion of this evolution is presented in Chapter 4. Prior to the establishment of Area Health Services (AHSs) in the late 1980s, health service providers in NSW were funded largely on the basis of historical expenditures (Hindle, 2002). However, ‘historical–based funding’4 was seen as ineffective and inequitable and this led to the introduction of a Resource Allocation Formula (RAF) in the late 1980s. The RAF, though innovative at that time, presented Area Health Service managers with several challenges, including how to establish an effective administrative process that suited the new reforms. In 1995, NSW Health introduced a revised formula, the Resource Distribution Formula (RDF), as a guide to inter–AHS allocation of NSW Health controlled financial resources. The development and application of the RDF are discussed at length in Chapter 4 and later chapters of this thesis.

4 Historical funding is based on traditional development based on incrementing the previous year’s allocation and adjusting for growth (NSW Health, 2008).

44 ______

Chapter 2: Summary Chapter 2 has addressed the following: Presented some contextual information relevant to understanding the Australian health system in general and the New South Wales system specifically for the purposes of this study.

Illustrated the complexity of the Australian health system within which health financing decisions are made, including the division of responsibilities between the Commonwealth, State/Territory and Local governments and the Non–government sector in health services financing and health care delivery.

Shown that the Commonwealth government and the non–government sectors contribute the majority of funds for health expenditure. In terms of health financing, the State/Territory governments contribute less than one third of total health service funding.

Focussed attention on NSW (home to this study). The state is the largest among the six states and two territories in Australia in terms of population size (approximately 7 million), thus attracting a larger share of health expenditure. Although the state receives a greater share of health expenditure compared to other states and territories, it plays a minor role in health financing and delivery of health services.

Noted that NSW population health status, as in other parts of Australia, is reflected in the high rate of life expectancy (around 79.3 years for men and 84.2 years for women). However, this high life expectancy is not shared equally between Indigenous and non– Indigenous Australians. The life expectancy gap between the two population groups is estimated to be over 19 years, highlighting the grim inequities in life expectancy in NSW.

Identified the principal providers of health services in NSW: the NSW Department of Health, General Practitioners (GPs), specialist doctors, allied health professionals and a range of other non–government agencies. The public sector is dominated by the NSW Department of Health delivering services through public hospitals and community health centres.

Described how NSW Health delivers health care through eight decentralised administrative Area Health Services and other entities. Health funding to these AHSs is allocated by the NSW Department of Health using the Resource Distribution Formula as a guide.

Recognises that private provision of health care in NSW is an important source of health care. The non–government sector constitutes the second largest source of health care financing after the Commonwealth government.

Notes that, although NSW Health seeks to achieve equity in the provision of health services throughout NSW, there are constraints that limit the possibility of NSW Health achieving this goal.

45 ______

Chapter 3: RESOURCE ALLOCATION, EQUITY AND HEALTH NEEDS

Overview Allocation of healthcare resources (financial, human and others) remains one of the difficult challenges confronting policy makers. For the optimal performance of health systems, resources must be allocated not only efficiently to ensure value for money but also equitably to facilitate access to the worst–off. Policy makers have a duty to strike a good balance between financial resources and the many inputs used to deliver health services. In addition, they have the responsibility to promote equity by ensuring that resources are allocated fairly such that the health needs of the most disadvantaged get the top priority. The lack of conceptual clarity around the terms ‘resource allocation’ and ‘equity’ complicates these issues further. This chapter reviews some of the key theories and concepts underpinning resource allocation and equity in the health sector and analyses the extent to which they explain how resource allocation is dealt with in New South Wales.

3.1 Definition and Levels of Resource Allocation Resource allocation is the process of sharing resources (financial or other) among competing needs. In the context of healthcare, it involves the movement of resources from the funders of healthcare to the providers and the sharing among individual patients consuming healthcare (Eager et al., 2001, Mitton and Donaldson, 2003). Resource allocation decision–making occurs at different levels of health systems and involves various actors. The two basic levels where resource allocation decisions are often made in the health sector are the macro (the national or state) and micro (population or individual) levels.

Macro level resource allocation decisions are made largely by policy makers at the national or state level and are intended to allocate the available resources to providers at different levels of the health system (Denier, 2007). They (macro level resource allocation decisions) usually involve determining the amount of funding that should be allocated to a particular jurisdiction or provider and the services to be provided using those funds. Sometimes the decisions include the methods for distributing the funds to programs and services, 46 ______especially where the national level allocation is made in the form of a block grant. Overall, the resource allocation decisions made at the macro level determine the kinds of health– care services that will exist in a society, who will get access to these services, on what basis and how equitable the system of financing will be (Anderson, 1999).

Micro level decisions, on the other hand, are concerned with distributional problems relating to individual patients or population groups. In many settings, micro distribution is designed by relatively autonomous health–care providers sometimes in consultation with the patient. Although they may be constrained by guidelines laid down by the central level, they have some autonomy to design and implement their preferred method of allocating resources (Eager et al., 2001, McKneally et al., 1997). Micro decisions mainly determine which particular person will receive what specific health services. For example, within the NSW health care system, Area Health Services are responsible for some micro resource allocation decisions including which particular individuals will benefit from the resources and services they provide, prior budgeting practices, budgeting guidelines or policies and interests of organisations. Other criteria deemed crucial in resource allocation decision– making are the “communitarian claims” (Mooney, 2006) and “weighted capacity to benefit” (Mooney and Houston, 2004) from resources.

In New South Wales State, the macro level resource allocation decisions are made on two levels: at the State parliament treasury level which decides how much funding should be allocated to the health portfolio and at the NSW Health Department level where decisions on how the health–care budget is allocated by Area Health Services and programs/services are made. Therefore, on these levels, policy makers shape the general disposition of the health–care system by their control over the general resources to be allocated in society, thus determining the framework within which micro issues arise (McKneally et al., 1997, Denier, 2007). The macro level decisions are in most cases influenced by public demands or interest, the specific country/state Health Act, current legislation and politics. However, in most medical literature five important criteria are often cited as ethical considerations in health care resource allocation. These include the following: likelihood of benefit, change in quality of life, duration of benefit, urgency of need, and amount of resources required. The inappropriate criteria for allocating resources are as follows: the ability to pay, social

47 ______worth, perceived obstacles to treatment, patient contribution to disease and past use of resources (Buchanan, 1997).

This study is concerned with resource allocation decisions made at the macro level (that is, from the NSW Department of Health to the geographical Area Health Services) and the micro level (population or Area Health Service level). According to Daniels (1985), micro level issues are real dramas and can become ‘modern morality plays’, with typical characters, dilemmas, dramatic choices and weighing of interests, often performed and replayed in case studies, jurisdiction and the media, where they have a highly sensational character (Daniels, 1985a). But, although macro decisions are less dramatic, they have a greater impact on the general health status of its members since they determine the structure and extent of basic health–care institutions within which the micro problems arise (Buchanan, 1984).

The macro level determines whether particular actions, policies or priorities in resource allocation decisions reflected at the micro individual patient level are just or unjust. Since healthcare systems involve a diverse and constantly evolving set of different institutions with actors taking decisions at the different levels of society, all the decisions made, especially resource allocation decisions, have a major impact on the level and distribution of a population’s well–being. The difficult choices involving population groups at the micro level may sometimes induce society to modify its macro resource allocation policies to increase supply of the resources if they are deemed insufficient (Denier, 2007). Because of this, issues of social justice are involved. The next section discusses theories of social justice in the context of equitable resource allocation.

3.1.1 Resource Allocation and Theories of Justice Equity with regards to health care resource allocation is a concern for many people particularly as public health spending comes under increasing pressure and demand for better health care rises. Decisions about how best to allocate scarce health resources among competing priorities to ensure equity ought to be based on some moral principle assessment (Gillon, 1986, Phillips, 1979, Denier, 2007). Although there is no consensus as to which

48 ______particular moral principle should take precedence in resource allocation decision–making, such decisions are explicitly or implicitly based on some theory of social justice (Asante, 2006). According to Barry (1989), a theory of justice should, in principle, answer three questions: Firstly, what is justice? Secondly, why should one behave justly? And thirdly, how can we determine what justice demands?

Philosophers for decades have made attempts to explain justice using various concepts such as ‘fairness’, ‘equity’, ‘desert’ and ‘entitlement’ (Buchanan, 1981, Rawls, 1985). In most of these accounts, justice is interpreted as fair, equitable and appropriate treatment in the light of what is due or owed to an individual. In this context, one who has a valid claim based in justice has a right, and hence is due something. So an injustice can be described as a wrongful act or omission that denies people benefits to which they have a right, or which fails to distribute burdens fairly (Mooney, 1992, Denier, 2007, Culyer, 1991). Thus, as noted by Whitehead (1994), social justice would be deemed to exist when conditions allow associations and individuals to obtain what they are entitled to according to their nature and vocation. Consequently, social justice theories contain principles for evaluating laws and institutions from a moral standpoint that are separate from those laws and institutions (Phillips, 1979, Sen, 1998). They provide a criterion against which existing institutions and social arrangements that allow people to receive a fair share of societal resources and burden (Sen, 2002) can be judged.

Thus, it is vital to examine the objectives and impacts of spending limited health resources, specifically social spending, to determine whether they are producing visible successes in bridging injustices in their various aspects among the different groupings in society. Indeed, decisions about how best to allocate scarce health resources among competing priorities ought to be based on some moral assessment (Phillips, 1979, Gillon, 1986). Although there is no agreement on the preferred ethical principle in resource allocation decision–making, most of the decisions made are overtly or implicitly based on some theory of justice (Whitehead, 1994). However, social justice is not simply about relative sharing of scarce resources. Social justice is different from distributive and procedural justice. The term ‘resources’ or ‘health resources’ as used in the study refer predominantly to financial resources. Although this study notes the importance of other forms of health care resources,

49 ______such as human resources, the study has deliberately adopted financial resources as the primary area of concentration to give the analysis a clearer and more concise focus. The next section looks at distributive and procedural justice in the context of resource allocation.

3.1.2 Distributive versus Procedural Justice Distributive justice is about the notion of fair share of social goods or income (Bayer et al., 1983). According to Aristotle, the distribution of social goods ‘that have to be shared among members of the political community’ is just as far as it complies with the principles of equality in ‘geometrical proportion’ (Aristotle, 1996). This implies that for a distribution to be fair, there is the need to treat equals equally and unequals unequally. In other words, one cannot render the same amount of whatever is to be shared to everyone regardless of their circumstances. People are different and thus, efforts to promote equality would take into account the relevant differences between individuals. The key issue, however, is how to determine who is equal and who is unequal? There are no clear guidelines on the particular respects in which equals ought to be treated equally and no specific criteria for determining whether two or more individuals are in fact equals (Culyer, 1991). Aristotle merely asserts that, whatever respects are under consideration as relevant, persons equal in those respects should be treated equally and those unequal should be treated unequally in the same respect. How relevant is distributive justice in the context of health care resources?

Distributive justice, in the context of health resources, deals with “how a society or group should allocate its scarce resources among individuals with competing needs or claims” (Roemer, 1996). It is concerned with the allocation of resources in a fair and appropriate manner (Mooney, 2006) in the light of what is due, or owed, to people. The problems of distributive justice are particularly manifested in healthcare because of the universal mismatch between demands and resources available. It is one of the most debated topics in bioethics. However, there is no consensus on a single theory of justice when it comes to the best approach for distributing health resources.

50 ______Culyer (2001) observed that the absence of an agreed theory of social justice arises from the lack of a “monist theory of morality”. According to Crowder (2003), a ‘monist theory of morality’ is the view that all ethical questions have a single correct answer and that all these answers fit together in a single coherent moral system. Moral theory attempts to capture intuitions about whether an action is right or wrong and may give a deeper understanding of moral claims as it gives an account as to why an action is good or bad (Crowder, 2003). With a lack of agreed justice theory on which to base decisions, healthcare workers, researchers, and policy makers often find themselves in challenging positions, being required to make decisions in the face of novel ethical dilemmas at both macro and micro health system levels regarding what should be just allocation of health resources (Kluge and Tomasson, 2002). In the debate over distributive justice for healthcare and health resources, there are various moral theories that feature prominently and are predicated on utilitarianism, egalitarianism or libertarianism (Beauchamp and Childress, 1989).

Classical utilitarianism is concerned with maximising happiness (a concept which is not concerned solely with health), although some (Gillon, 1994a) prefer to describe the approach in healthcare as the ‘maximising principle’, seeking to maximise the overall good which can be determined by population health outcomes. In its purest form, the utilitarian approach considers distributive justice to be best served by maximising social utility. It is often expressed as ‘‘the greatest good for the greatest number of sentient beings’’ or “the greatest happiness principle”. The theory suggests that resource allocation decisions should create the greatest good in terms of health outcomes for the greatest number of people as part of maximising value.

Distributive justice is an obligation of the utilitarianism that involves tradeoffs between risks and benefits (Beauchamp and Childress, 1989, Macklin, 1987). To put this into practice in the health care setting, the health system must be able to decide what good it should be maximising (fair distribution of resources, health outcome, access to health care or peoples’ health care needs), whether it can be measured and how to balance the quality of a good with the quantity of that good (Mooney, 2000a). According to Norman Daniels,

51 ______the role of the health care system is to protect an individual’s share of the normal opportunity range, both by curing disease and preventing disease (Daniels, 1985b), meaning it is the range of opportunities that are being maximised.

Therefore, decisions about social value judgements that aim to be either implicit or explicit within a healthcare system should focus on the acceptable implications of each theory of distributive justice without necessarily, or invariably, choosing one over another. Inevitably, some people will be dissatisfied (Gillon, 1994b), for not everyone’s claims will be met. Procedural justice, however, places a premium on ensuring that the processes by which decisions are reached have legitimacy (Daniels and Sabin J.E, 2002, Daniels and Sabin, 1997, Daniels, 2000, Daniels and Sabin, 1998, Gutmann and Thompson, 1996). But, what is procedural justice?

Procedural justice is about fairness in the processes that resolve disputes and allocate resources. According to Daniels and Sabin (1997), procedural justice concerns the fairness and the transparency of the processes by which decisions are made (Daniels and Sabin, 1997). Some theories of procedural justice hold that fair procedure leads to equitable outcomes, even if the requirements of distributive justice are not met. Procedural justice deals with the perceptions of fairness regarding outcomes and the extent to which the perceived outcome allocation decisions have been justly made (Bone, 2003, Hay, 1997).

The philosopher John Rawls (1972), in his theory of justice, distinguished between three types of procedural justice: perfect procedural justice, imperfect procedural justice, and pure procedural justice. Perfect procedural justice has an independent criterion for what constitutes a fair or just outcome of the procedure, and a criterion that guarantees that the fair outcome will be achieved. Imperfect procedural justice shares the first characteristic of perfect procedural justice (an independent criterion for a fair outcome) but has no method that guarantees that the fair outcome will be achieved. Pure procedural justice describes situations in which there is no criterion for what constitutes a just outcome other than the procedure itself (Rawls, 1999). In general, therefore, procedural justice of a distribution is entirely dependent on the path used to reach an outcome, meaning the process or the

52 ______procedure (Westphal, 1996), applied to achieve the outcome should be fair (Beauchamp and Childress, 1989, Daniels and Sabin, 1997).

Daniels (2000) and other bioethics and political philosophers are of the opinion that, if there is to be confidence in the legitimacy of decisions, the procedures adopted should have four characteristics: publicity, relevance, revision and appeals including regulation. This means that decisions about limits on the allocation of resources must be public, including the grounds for making those decisions. ‘Relevance’ means that the grounds for decisions are ones that fair–minded people would agree are relevant to meeting healthcare needs, especially when there are constraints on resources. In particular, ‘relevance’ focuses on the importance of deliberation about the limits of the common good and acknowledges that such ‘deliberate democracy’ should involve the decision–makers themselves and those whom the decisions may affect (e.g. patients) (Daniels, 2000).

On the other hand, ‘revision and appeals’ means that there must be opportunities for challenging decisions and mechanisms for resolving disputes. There should be systems in place for revising decisions when new, or additional, evidence becomes available or new arguments are put forward. Equally, there should be voluntary or public regulation of the process of decision–making to ensure that it has all the other three characteristics (publicity, relevance and opportunities for revision and appeals). Ensuring that procedures have all four characteristics makes decision–makers ‘accountable for their reasonableness’ (Daniels, 2000, Daniels and Sabin, 1997, Daniels and Sabin, 1998, Gutmann and Thompson, 1996) in distribution of healthcare resources.

In the NSW health care system, different distributive and procedural theories of justice have different levels of applicability. For example, the Labor Party advocates an egalitarian approach to resource allocation through implementation of strategic policies in support of equality of access to health services (NSW Health, 1990, NSW Health, 2005). In contrast, the Liberal and National Party coalition supports a reduction in public sector involvement in favour of the private sector and market mechanisms. The next section discusses the theories of justice in the context of health resource allocation in NSW.

53 ______3.1.3 The Justice of Resource Allocation in NSW According to the utilitarian theory, greater equality of health can be achieved by directing health resources where they will have the greatest long–term effect for the greatest number of people. In NSW, given the level of health inequalities between Area Health Services and population groups (NSW Health, 2008), this theory as a guiding principle for health resource allocation will have limited applicability. One may argue that allocation of resources to maximise the greatest possible utility for the greatest number of people essentially means resources can be efficiently allocated to maximise the health of the entire population. In the NSW context, such an argument, on the surface, may have some merit: take, for example, chronic diseases, which are the leading cause of morbidity and mortality in the state. Allocating resources on the basis of the utilitarian principle to combat chronic diseases could lead to the greatest possible improvement in population health by saving many lives, something public health policies strive to achieve. However, in allocating resources to maximise aggregate population health or the sum total of individual utilities, the utilitarian approach suggests that personal circumstances do not matter. Inequalities across population groups, for example, are not significant as long as the overall utility of the general population is maximised. In health care settings with deep–rooted inequalities among population groups (Australian Indigenous and non–Indigenous), it is difficult to justify allocating resources without due regard for such inequalities which have been regarded as a “national disgrace” (Sydney Morning Herald, 2008) and often been compared to those prevalent in developing countries.

The other difficulty of applying utilitarianism in resource allocation in NSW would be the measurement of health maximisation; this requires a composite measure of health gains capable of being estimated across illnesses and health care interventions or programs/services (Segall, 2003), as well as across different population groups. What is of great concern with regards to measurement is the utilitarian approach of assigning equal weight to individual utilities; therefore, adopting this measure in resource allocation would mean that no area or individual or population group deserves any special attention irrespective of the level of health needs. For example, all Area Health Services in NSW either rural or urban, and population groups (Indigenous or non–Indigenous) (Mooney et

54 ______al., 2002), would be weighted equally when resources are distributed. One may even argue that the utilitarian approach would require that resources be allocated towards rich Area Health Services with relatively well functioning health infrastructures to obtain as great health benefits as possible in order to maximise the total health gains.

By contrast, more fundamental egalitarian principles focus attention on how individual utilities are aggregated to yield social welfare. Rather than simply summing utilities up, as classical utilitarianism does, these essential egalitarian principles adopt a social welfare function which gives greater weight to individuals with lower utility (Sen, 1995). There is no fairness in assuming that individuals or groups have the same utility function in a state where inequalities between areas and population groups still exist and everything points to the fact that without extra attention or ‘positive discrimination’ (Mooney and Houston, 2004) to certain groups, unjust and unnecessary disparities will persist. These difficulties would be overwhelming when it came to the implementation of utilitarian theory.

The libertarian theory of justice has some fundamental equity implications for allocation of health care resources in NSW, where current inequalities in health may be linked to unfair distribution of resources in the past. Historically, allocation of health resources in NSW has tended to favour areas with large hospitals and those that were urban based (see Chapters 2). According to Nozick (1974), in his principle of ‘rectification of historical injustices’ in distribution, the current inequities in health have to be rectified (Nozick, 1974). His third principle has the potential to redistribute resources in favour of disadvantaged areas or population groups (i.e. Indigenous people) in NSW who did not receive a fair share of resources in the past. In this regard, the entitlement theory has some potential, particularly ease of implementation to improve equity by shifting resources to the worst off areas and the most disadvantaged population groups.

Nozick’s proposition is that, if an injustice has occurred, we should look at the historical pattern of past distribution and determine what the expected outcome would have been if the injustice had not occurred and use that to correct the situation. The challenge here is that most inequities in health result from a combination of factors such as the political,

55 ______economic and social determinants of health (Daniels et al., 1999). They are not only the result of unjust distribution of health care resources, but also unequal distribution of other resources that control the political and social determinants of health (i.e. education, employment, environment, housing and politics). In such a situation it would be difficult to isolate the injustice that might have occurred as a result of unfair health resource distribution from those resulting from unequal distribution of other social determinants of health sector resources in order to remedy past injustices. In many countries, justice in distribution of health resources is not the only answer (Varian, 1975) to rectifying past injustices in distribution of resources. Therefore, a concerted effort would be required from the different sectors concerned with the social and economic determinants of health and from politicians to repair past injustices in allocation especially within the NSW health care system. Neither of the theories takes into consideration how such problems might be resolved nor does the entitlement theory provide clear guidance as to how to rectify past injustices. So what would be the best theory of justice to guide resource allocation within NSW health system?

Rawls (1972), when discussing specifically his difference principle, writes: “…now the difference principle is not of course the principle of redress. It does not require society to try to even out handicaps as if all were expected to compete on a fair basis in the same race. But the difference principle would allocate resources in education, say, so as to improve the long–term expectation of the least favoured. If this end is attained by giving more attention to the better endowed, it is permissible; otherwise not”. This provides an alternative approach for maximising greater benefits for the least advantaged. Although the focus is not health, the principle could be used to guide the distribution of health resources to Area Health Services or disadvantaged population groups such as the Indigenous population to reduce the existing inequalities in health within the NSW health sector.

Daniels’ (1999) linking of the Rawlsian theory to the health sector demonstrates that disease and ill–health impair people’s capacity to function normally as human beings and thus unable to also contribute to the economy. Therefore, there is a need to address health care needs if a fair equality of opportunity is to be guaranteed with the available scarce

56 ______health resources. As Daniels (1999) noted, addressing the health needs of the most disadvantaged means that substantial resources must flow and not “just a mere trickle down principle but one that requires maximal flow in the direction of helping the worst–off groups”(Daniels, 1999). Thus, the difference principle demands more than simply small incremental increases of resources but rather substantial increases that would guarantee some improvements in the lives of the worst off Area Health Service population groups, especially Indigenous Australians. Although the Labor Party advocates the egalitarian principle, it has not practiced egalitarianism strictly in distributing health resources to help the worst–off Area Health Services (AHSs) and population groups, at the expense of the better off. For example, even with the introduction of a population and needs–based resource distribution formula in NSW Health and recommendations that resources need to be redistributed from the over–funded AHSs to the historically under–funded AHSs, the Department of Health found this difficult to implement (NSW Health, 1999, NSW Health, 2005). But how egalitarian should allocation of resources be in a non–egalitarian society where historical differences in funding exist between Area Health Services or even more so where population groups have not been justly treated in terms of receiving a fair share of health and other types of resources?

The reality is that, in Australia and more specifically in New South Wales, as in many other countries, some inequalities in health are the result of past injustices and bad national, state and local policies, including poor distribution of state resources, which have resulted in poor health outcomes for Indigenous Australians. In such instances, seeking to remedy injustices becomes crucial and one can draw some guidance from Robert Nozick’s principle of rectification of past injustices to support the justification for why there is the need to give the health needs of the worst–off greater priority.

Investing in social determinants of health such as education, as Rawls proposes, or employment, environment, housing and politics as other studies have shown (Whitehead, 1992, CSDH, 2008, National Health Strategy, 1992) would no doubt improve the long– term health outcomes of the general and worst–off population. But, a multi–sectoral collaboration is required from all relevant sectors (health and non–health sectors) since the

57 ______health finances objective is to meet the health needs of the population and not socioeconomic determinants of health. Equally, partnership is required from politicians and officials who have a responsibility to redirect health and other resources to meet the health and non–health needs of the worst–off. In the short–term, health policy makers have to work on improving the efficiency of health systems’ finances and equity in resource allocation. The next section discusses efficiency in the context of resource allocation.

3.2 Resource Allocation and Efficiency While health equity has gained considerable attention in recent years and is central to this thesis, the concept of efficiency has long been applied to the allocation of scarce health resources; it is therefore important that a brief review/discussion of the concept and how it has been applied in health care resource allocation is carried out.

3.2.1 Technical Efficiency Technical efficiency describes the physical relation between resources (capital and labour) and health outcome. It refers to the production process and the use of least cost combinations of resource inputs for the production of any scheduled goods and services (Segal and Richardson, 1995). Technical efficiency is achieved when the maximum possible improvement in outcome is obtained from a set of resource inputs. Thus, the production of a given good or service is technically inefficient when the actual costs of production are greater than the price of the output (Evans, 1979). For example, an intervention would be technically inefficient if the same (or greater) outcome could be produced with less of one type of input (i.e. number of nurses or doctors).

Closely associated with technical efficiency is the concept of productive efficiency which, in the health field, refers to the maximisation of health outcomes for a given cost, or the minimisation of cost for a given outcome. Since different combinations of inputs are being used, the choice between interventions is based on the relative costs of these different inputs (Haddix et al., 2002, Drummond et al., 2005). In healthcare, productive efficiency

58 ______enables assessment of the relative value for money of interventions with directly comparable outcomes. It cannot address the impact of reallocating resources at a broader level, for example from geriatric care to mental illness, because the health outcomes are incommensurate. Productive efficiency involves choosing different combinations of resources to achieve the maximum health benefit for a given cost (Greenhalgh and Peacock, 2005).

3.2.2 Allocative Efficiency Allocative efficiency refers to the optimal choice of input mix, given their respective prices. In health care, it is achieved when it is no longer possible to increase the overall benefits produced by the health system by reallocating resources between programs/services. Duckett (2008) notes that allocative efficiency is about ensuring the best allocation of available resources so that the inputs allocated to the health care system yield the best possible outcomes. Thus, the concept of allocative efficiency takes account not only of the productive efficiency with which healthcare resources are used to produce health outcomes but also the efficiency with which these outcomes are distributed among the community (Drummond, 1991). Such a societal perspective is rooted in welfare economics and has implications for the opportunity costs5 (Palmer and Raftery, 1999, Gold et al., 1996, Russell, 1992).

Key elements of allocative efficiency include priority setting among diseases and within diseases (i.e. preventative versus curative investments), and minimising preventable hospitalisations, for example, through better (Duckett, 2008a). Priority setting is one of the important health planning tasks that face governments in seeking to enhance the distribution of health care resources through the way health funds are allocated to populations (primarily driven by supply or allocated through a needs adjusted funding formula to meet equity objectives) or between programs and services which can be

5 According to Gold et al (1996), opportunity cost is the benefits forgone by a particular use of resources. Since health resources are scarce relative to needs, the use of resources in one way prevents their use in other ways. Therefore, the opportunity cost of investing in a healthcare intervention is best measured by the health

59 ______essentially adhoc or based on a formal priority setting mechanism. In the NSW health system where resource allocations are decided based on historical budgets, number of beds, or utilisation rates of the services, the outcome is often the same, of an increasingly inequitable circular process of high utilisation, more resources used to create more supply, which generates more demand and use of hospital services than primary health care by some population groups and less by the most in need. So is the allocation in NSW aimed at achieving efficiency or equity?

Allocative efficiency might be improved through several measures. For example, it is widely believed that providing better primary health care services such as general practitioners (GPs) services is a more appropriate and cost–effective alternative to hospital emergency services (for those patients not needing emergency treatments). Similarly, providing residential aged care for elderly people (not acutely ill) might improve allocative efficiency better than would providing them with hospital acute care (National Health and Hospital Reform Commission, 2009, Haddix et al., 2002). The next section discusses the concept of equity, which has been one of the commonly invoked concepts in the distribution of health care resources but is also very contentious.

3.3 Equity: Concepts and Definitions The work of the philosopher John Rawls (1972) on the theory of justice, although not directly related to health, has provided some fundamental underpinnings for the concepts of equity and resource allocation in health policy. The concept of equity in health care has been widely debated for many years as a guiding principle for health care systems around the world and has produced an enormous literature; however, it still remains an ethical concept that eludes precise definition not least of all because equity is interpreted differently by individuals and in each society depending on the dominant ideology within that society (Culyer, 1976b, Williams, 2005, van Doorslaer et al., 1993, Oliver, 2002, Mooney, 1986, Le Grand, 1987, Evans et al., 2001, Culyer, 2001, Mooney, 1983). In benefits [life years saved, quality adjusted life years (QALYs) gained] that could have been achieved had the money been spent on the next best alternative intervention or healthcare programme (Russell, 1992) .

60 ______practice, the pursuit of ‘equity’ in health care has been difficult and, as Birch and Abelson (1993) put it, “notoriously difficult to interpret and operationalise” (Birch and Abelson, 1993). Due to its varying interpretations, health care systems rarely specify fully the equity goals they seek to achieve (McIntyre and Gilson, 2000, Mooney, 1987, Culyer, 2001) or the means of achieving it.

Equity is a concept equated with justice based on fairness in distribution of something. It is a significant factor that determines whether or not the concept can generate an allocation that meets a population’s obligation for justice. It requires a normative assessment, so that decisions made depend upon people’s values (Mooney, 2006). Equity is an essential concern for many people with regard to the allocation of health care resources, considering the ever increasing constraints on public spending with increased demand for better services (Mooney, 1983). Thus, it is vital to examine the objectives and impacts of this spending, specifically social spending, to determine if it is producing visible successes in bridging inequality in its various aspects among the different groupings in society.

The notion of equity in health (healthcare) has been widely debated over the years. Definitions of equity flourish. Mooney, among others, has been a persistent advocate of equity and proposes seven different ways of defining equity (Mooney, 1992, Mooney, 1987, Mooney, 1994a, Mooney, 1986, Mooney, 1983):

 Equality of expenditure per capita;  Equality of inputs per capita;  Equality of marginally met need;  Equality of access for equal need;  Equality of inputs for equal need;  Equality of utilisation for equal need; and  Equality of health.

Some of the above equity definitions by Mooney have provided valuable foundations. Equality of expenditure per capita signifies an equal distribution of health funds among the

61 ______different geographical areas based on the population size. However, this definition does not consider the differences in need since the main component being considered is the population.

Equality of inputs per capita means that physical resources are divided equally among the different geographical health areas based on the population size. This may, for example, be reflected in ratios such as number of doctors per population and number of beds per population. Although it is progress over the previous definition in the sense that it takes into account price differentials, it suffers from the same shortcoming because it does not accommodate differences in need (Mooney, 1987).

Equality of marginally met need assumes that regions categorise needs in order of priority to be met in an identical manner (Mooney, 1986). Based on this definition, equity will be achieved when each region is able to meet the same last need with its available resources. Culyer and Wagstaff (1993), however, argued that this is better viewed as an efficiency principle with the objective of maximising health rather than an equity principle.

Equality of access for equal need, on the other hand, entails a fair geographical distribution of resources based on health care needs and ease of access. Equal access is mostly understood to mean equal access to health care or equal treatment for equal need regardless of the financial, cultural or geographical barriers. Unequal access results if people are denied health care because of their socio–economic status, or other factors unrelated to the need for care. If people have access to the same information, tastes and preferences for health care, equality of access for equal need becomes identical to equal utilisation for equal need. But what is the difference between equity and equality?

The terms equity and equality are closely related and have often been used to refer to almost the same thing. However, even though equity and equality are interrelated, they both have significant and noteworthy differences in meaning (Asante et al, 2006). As observed by Le Grand (1987), equality is fundamentally a ‘descriptive’ term while equity is basically

62 ______a ‘normative’ term (Le Grand, 1987). Thus, a particular distribution can be observed and, on the basis of the observation, one can conclude whether the distribution is equal or unequal. However, the “presence or otherwise of equity cannot be established by reference solely to an observed distribution”. One has to make value judgements combined with the observed facts about the distribution (Mooney, 2000a). In effect, while “equity statements are statements of value; equality statements are statements of fact” (Le Grand, 1987). On the other hand, Sen (2002) argued that, “equality is an abstract idea, with restricted cutting power unless one starts to specify what is it that is to be equalised”. Furthermore, the central issue begins with the specification of the space in which equality is to be sought, and the equitable accounting rules that must be followed to arrive at the aggregative concern as well as the distributive ones.

Others saw equality as related to what people are entitled to and equity as dealing with rewarding fairly (Asante et al, 2006, Wilson, 1993, Wolfson and Rowe, 2001, Whitehead, 1993, Culyer and Wagstaff, 1993). They maintained that both equality and equity are components of our ideas about fairness, and that these components may be difficult to resolve. According to Brian Barry (1990), equity is a ‘comparative’ principle, a judgement about how a person or group of people is situated relative to others. Therefore, equity required that “equals be treated equally and unequals be treated unequally”(Mooney and Jan, 1997, Culyer and Wagstaff, 1993). This is the analogous to the concepts of horizontal and vertical equity (discussed in the next section) in health care analysis (Culyer and Wagstaff (1993) which ensure that the principle of equality is preserved, while the diverse health needs of individuals are recognised. Thus, it seems that while the concept of equity deals with fairness and is based on value judgements, equality is about entitlements to the same level of access, in this case to health care irrespective of socio-economic status or geographical location.

Clearly, there are complexities inherent in how different people view ‘equity’ and ‘equality’. However, there is agreement that inequity and inequality have vital differences. Inequity has a moral dimension. According to Braveman (2003), inequity could have an accusatory, judgemental or morally charged tone. A similar reasoning underpins Margaret

63 ______Whitehead’s most cited proposal (1992), which made an effort to distinguish between inequities and social inequalities: “that inequalities which are avoidable and unfair constitute health inequities”. In her previous work Whitehead (1985), also provided definitions of health inequity as “differences in the distribution of health” that are “unnecessary and avoidable, and in addition, are also considered unfair and unjust”. This means that, for a situation to be deemed inequitable, the cause has to be examined and judged to be unfair in the context of the social setting with which one is dealing.

According to Whitehead (1992), not all differences in health can be labelled as inequitable and therefore it is not feasible to achieve equality of health. She provided a list of seven6 determinants of health differentials and suggests that those related to biological variation and free informed choice are more likely to be inevitable/unavoidable or fair inequalities. Those differentials arising from determinants where individuals have less choice in lifestyle, work conditions, or access to health care and other public services are more likely to be considered avoidable and unfair and thus inequities.

However, others believe that with current progress in medicine and technology, it is possible that such factors as genetic variations will become amenable to change in the future (Gakidou et al., 2000). Inequalities in health describe the differences in health between groups independent of any assessment of their fairness or variation in distribution of access to health care. The fairness or unfairness qualification invokes assessments of whether the inequalities are avoidable or unavoidable. It makes no moral or intrinsic judgement about the differences. Health inequality is defined simply as the variations in health status across individuals in a population (Murray et al., 1999, Evans et al., 2001). Inequities, on the other hand, refer to a subset of health inequalities that are deemed unfair

6 The seven main determinants of health differentials are as follows: 1) Natural, biological variation; 2) Health–damaging behaviour if freely chosen, such as participation in certain sports and pastimes; 3) The transient health advantage of one group over another when that group is first to adopt a health–promoting behaviour (as long as other groups have the means to catch up fairly soon); 4) Health–damaging behaviour where the degree of choice of lifestyles is severely restricted; 5) Exposure to unhealthy, stressful living and working conditions; 6) Inadequate access to essential health and other public services; and 7) Natural selection or health–related social mobility involving the tendency for sick people to move down the social scale (Whitehead, 1992).

64 ______and unjust and therefore unnecessary and avoidable (Whitehead, 1992, Marchand, 1998, Williams, 2005).

According to Whitehead (1992), health differences that can be termed as unfair, unnecessary and avoidable are classified into four categories: first, those which are a result of health–damaging behaviour where the degree of choice of lifestyles is severely restricted; second, those that are a result of exposure to unhealthy, stressful living and working conditions; third, those that are a result of inadequate access to essential health and other public services; and finally, those that are a result of natural selection or health– related social mobility involving the tendency for sick people to move down the social scale. However, with the fourth category she noted that, although the original ill health in question may have been unavoidable, the low income of sick people seems both preventable and unjust. The next section discusses horizontal and vertical equity.

3.3.1 Horizontal and Vertical Equity All the interpretations of equity examined so far and others in the literature predominantly deal with horizontal equity. However, for the past years, authors such as Mooney, Cuyler, McIntyre, Gilson, Wiseman, Jan and others have highlighted the need for more emphasis on vertical equity in health and health resource distribution, particularly in countries where there are substantial differences in health status between different population groups (Mooney et al., 2002, McIntyre and Gilson, 2000). In the introduction of Mooney’s article on vertical equity, he observed that much has been done to stimulate debate about equity in health, but “the practical outcomes are disappointing; the poor in income remain poor in health, as do various minority groups and indigenous peoples” (Mooney, 2000b, Mooney and Jan, 1997, Mooney et al., 2002, Mooney and Houston, 2004). Indeed, Mooney has been a great advocate for recognizing the poor health status and inequities in health that exist between the Indigenous and non–Indigenous population in Australia, including inadequacies in distribution of health resources between the two population groups (Mooney et al., 2002, Mooney, 1998b).

65 ______According to the proponents of vertical equity such as Mooney, Culyer, Gilson and McIntyre, the health needs of the most disadvantaged should be given priority over those of other people for equity to be served in the allocation of health care resources. The emphasis is on differential allocation of resources in favour of the most disadvantaged with poor ill– health in society. The analysis in this study is built on the ground–breaking works of Mooney, Culyer, Gilson, McIntyre and others that campaign for vertical equity objectives for allocation of health care resources. In this study equity means access to financial resources on the basis of health needs. Health needs means the health status of the population, while health status is an indicator for health need.

Horizontal and vertical equity are two terms that have been pursued and clearly distinguished in terms of definition. In the context of equity, they both have important distinctions. In general, horizontal equity refers to equal treatment for equal need while vertical equity is defined as unequal but equitable treatment of unequals (Mooney, 1994b). Policymakers globally are faced with the dilemma of making a choice between horizontal or vertical equity objectives in order to guide health decision–making, particularly with regards to resource allocation and service delivery. Although some health care systems pursue horizontal equity, recently it has been losing popularity. This is due to the realisation that the existing inequities in health and health resource allocation can only be improved through policies that focus on vertical equity since it appropriately considers relative differences in the health needs of different people. These two concepts of equity have their roots in the Aristotelian formal principle, which states that “equals should be treated equally and unequals unequally in proportion to the relevant inequalities” (Gillon, 1986).

Horizontal equity, on the other hand, calls for equal treatment for equal need in the delivery of health care and it requires that individuals who manifest the same level of illness should receive the same amount of resources (Wagstaff et al., 1991). In contrast, vertical equity means individuals with unequal health needs are treated in an appropriately dissimilar manner in proportion to their level of inequality compared to those with lesser health needs. Horizontal equity, as interpreted and generally agreed on by policy–makers and researchers in Europe and North America, means that health care should be distributed according to

66 ______health needs and financed according to the ability to pay for healthcare (Wagstaff, 1989). This means that individuals or households with the same ability to pay make the same contributions. Vertical equity, by contrast, refers to a situation where individuals or households of unequal ability to pay make appropriately dissimilar payments (Wagstaff and van Doorslaer, 1998, Wagstaff and Van Doorslaer, 1993, Braveman and Gruskin, 2003). In terms of financing of health care, horizontal equity means allocation of equal or equivalent resources for equal health needs, while vertical equity refers to differential allocation of resources for different levels of health needs.

Although the nature and importance of the equity concept (either horizontal or vertical) employed within a health care system is ultimately a political judgement, the differences between horizontal and vertical equity are important as they have different implications for policy and decision makers. In recent years, attempts have been made to elicit equity concepts from the general public (Wiseman et al., 2003, Mooney, 2006) and Dolan et al. (2000) and Shaw et al. (2001) found that many people in the United Kingdom are keen to support vertical equity by redirecting considerable resources toward people of low social class or with adverse health prospects, at the expense of other National Health Service (NHS) activity, in line with the government’s public health policy. Indeed, Bambas and Casas (2001) observed that “a plan might appeal to proponents of horizontal equity on the basis that everyone needs health care at some point” while “targeted programs for the poor would appeal to vertical equity”. The choice of horizontal or vertical equity presents several issues that must be resolved. In the case of horizontal equity, policymakers might have the challenge of how to determine the ‘equal ability to pay’ (Asante et al, 2006). And with vertical equity, Wagstaff noted that, there is always the concern about the precise form that differential treatment should take. The next section looks in more detail at the flourishing equity interpretations and definitions (equal treatment for equal need, equality of access, and equality of health).

3.3.2 Equal Treatment for Equal Need Interpretation of equity as equal treatment for equal needs means individuals with the same health care needs should receive the same quantity and quality of treatment. According to

67 ______Le Grand (1987), the underlying assumption is that distribution of health care “should be independent of the distribution of income, wealth, or any other form of economic or political power” (Le Grand, 1987).

The two practical difficulties that have been identified in relation to this definition are how to measure ‘equal treatment’ and how to define ‘equal need’. Critics often argue that two persons with the same level of self–reported sickness may receive the same treatment but the response to treatment might not be the same for both individuals. The opponents conclude that equity cannot be deemed to have been achieved in such an instance. The issue is whether equal treatment should result in an equal outcome and, as noted by Whitehead (1992), ‘there is bound to be some natural variations between one individual and another’ in every attribute including health. This implies that if variations in genetic make–up are accepted as unavoidable, then different health outcomes from equal treatment may not be inequitable. The definition of ‘need’ is discussed later in the chapter.

3.3.3 Equality of Access to Health Care The principle of equality of access to health care is a central objective of many health care systems (Goddard and Smith, 2001, Olsen and Rodgers, 1991, NSW Health, 2007a). It implies that individuals should be given equal opportunity to use health services without regard to other characteristics, such as their income, education, ability to pay, ethnicity, or area of residence. Equality of access is derived from a notion of equal opportunity and implies the willingness to devote resources to improving access to health and health care for some population groups. The justification for equal access does not derive from its effect on the distribution of health care or health, although the implicit assumption on the part of policymakers and researchers alike is that equal access may improve inequalities in health. This assumption may not hold if health care is ineffective in improving the health of the unhealthy. The concept of equality of access is often coupled with some notion of need, so that access to services should be equal for individuals in equal need, and unequal for those in unequal need (Mooney, 1983, Mooney et al., 1991, Olsen and Rodgers, 1991). According to Le Grand, equality of access is “the requirement that people should face the

68 ______same personal cost of receiving medical treatment” (Le Grand, 1987). Clearly, the principle focuses exclusively on health care, ignoring all other possible ways of improving health, and so differs markedly from the principle of equality of opportunity for health outcomes.

However, the proponents of this definition believe that equity is not served if some people are made to pay more to access health care than others, or have to travel further than others to access health care, or are required to wait longer than others in order to receive health care. The obvious difficulties with such an interpretation is the definition of ‘access’ in terms of ‘cost’ of receiving treatment and how ‘personal cost’ is measured, whether it should be measured in monetary terms, or in terms of utility or satisfaction forgone (Le Grand, 1987, Mooney, 1987). Norman Daniels (1985) observed that the literature on equity of access is complex and confusing, partly because “access is itself a complicated notion, a composite of many factors”. Several authors have argued for more attention to the equality of access interpretation (Mooney, 1983, van Doorslaer et al., 2006, Van Doorslaer and al., 1998).

The equality of access interpretation based on cost implicitly assumes that two people facing the same cost for health care would enjoy the same access. In other words, it is irrelevant whether people have money or not to pay for the same cost they may incur, or whether a given health care facility is actually being utilised or not. Thus, the definition emphasises the opportunity to use health care rather than actual usage. Culyer, van Doorslaer and Wagstaff have rejected the equality of access interpretation as a whole, and the cost–based definition of access in particular. They challenge the idea that depending on income and other factors, people may have differential access even if they face the same cost (Culyer et al., 1992a, Culyer et al., 1992b). Culyer and Wagstaff have further argued that “irrespective of how one interprets access, and irrespective of whether equality of access is applied to those in equal need, application of the principle of equality of access to health care will not yield an equal distribution of health” (Culyer and Wagstaff, 1993).

69 ______Equality of access for equal need might pose two problems: the relationship between access and utilisation and how to define ‘need’. Mooney has distinguished between equality of access and equality of utilisation, arguing that the former is wholly a supply–side phenomenon, while the latter is a function of both supply and demand (Mooney, 1992). Despite the ambiguity, equity as equality of access is generally perceived to be popular among policymakers as noted earlier. Mooney and associates have argued that equality of access is “what policymakers understand to be equity”, noting that equality of access “provides individuals with the opportunity to use health services” (Mooney et al., 1991, McIntyre and Mooney, 2007). Culyer et al. (1992a; 1992b) argued that many policy documents not only contain references to ‘equality of access’ but also to ‘distribution according to need’ and ‘equality of health’. Disagreement is rife on the interpretation of need, which makes the meaning of “equal access for equal need” unclear as well. But what does equality of health mean?

3.3.4 Equality of Health Equality of health as a definition of equity does not imply that people will enjoy the same life expectancy or state of health. According to Culyer and Wagstaff (1993), the relevant equity principle is equality of health. Good health is necessary for individuals to “flourish”, and any position but one in which everyone has the same opportunity to flourish is hard to defend (Culyer and Wagstaff, 1993). Therefore, a just distribution of health is an equal one, and an equitable allocation of health care is one that gives rise to equality in health. However, Culyer and Wagstaff (1993) point out that health care is not the only determinant of health; health care alone is not expected to lead to an equal distribution of health.

As noted by Whitehead (1992), there are certain health conditions that are inevitable and unavoidable and which may, for example, include health conditions resulting from biological and generic variations (Whitehead, 1992). Therefore, the meaning of equality of health is self–evident – equal health status for everyone. The interpretation of equity as equality of health clearly has fewer advocates, including Culyer and Wagstaff (1993). However, commentators recognise that some differences in health are simply inevitable. Thus, irrespective of what is done, there would still be some variations in health status

70 ______among individuals. Whitehead’s definition of inequity as “differences in health which are unnecessary and avoidable” and also “unfair and unjust”, supports this assumption (Whitehead, 1992).

There is overwhelming evidence that natural and biological factors such as genetic make– up and the aging process have some irreversible consequences on health that cannot be labelled as unfair or unjust. This makes it unrealistic for everyone to have equal health outcomes or equality of health. To interpret equity as equality of health has often been dismissed as too expensive to be an appropriate interpretation of equity (Mooney, 1986). Le Grand (1987) has observed that “since individuals’ health is located within themselves, it is impossible to take away someone’s health and give to someone else; that is, it is impossible to redistribute health”. For him, a more rational alternative is to focus on health care rather than health itself, because “health care can be redistributed by act of policy” (Le Grand, 1987). Culyer and Wagstaff (1993), on the other hand, qualify the idea that equalising the distribution of health is not to be achieved by deliberately (as an act of policy) reducing the health of some members of society. This suggests that equality of health can only be achieved by an increase in expenditure. The money spent on the health of healthy individuals should stay at the present level, and spending on the health of unhealthy individuals should be increased according to their health needs. Thus, the concept of equality in health implies an unspecified increase in budget (Culyer and Wagstaff, 1993) based on systematically assessed health needs.

Obviously there are difficulties with the way equity as equality of health is interpreted. As noted earlier, Culyer and Wafstaff (1993) think that health is a necessity for human flourishing; therefore, the appropriate equity principle should be a just distribution of health or equality of health. Culyer and Wagstaff reject need–based interpretations of equity such as equal distribution according to ‘need’, arguing that “irrespective of how one interprets ‘need’, equality of health will not be attained if persons in equal need are treated the same and persons in unequal need are treated in proportion to the relevant inequalities”. They equally reject equality of access, as earlier noted, on the grounds that it would not yield an equal distribution of health. The rejection of both interpretations does not mean that the

71 ______concepts of ‘need’ and ‘access’ are not important in resource allocation decisions in health. The notion of ‘need’, they claimed, has a role to play by determining which resources are to be distributed but it does not “indicate the appropriate distribution of these resources” (Culyer and Wagstaff, 1993). ‘Need’ has been an elusive term to define although many interpret it as ‘ill–health’ and the ‘capacity to benefit’. But why is ‘capacity to benefit’ associated with equity?

3.3.5 Capacity to Benefit The ‘capacity to benefit’ notion of need has been debated extensively among health economists (Culyer, 1995, Crampton and Laugesen, 1995, Culyer and Wagstaff, 1993, Normand, 1991, Mooney and Houston, 2004). Capacity to benefit as interpretation of need sees health need as determined by the amount of health care resources required to exhaust an individual’s capacity to benefit from health care (Culyer, 2001, Culyer, 1995). Proponents have argued that a need for health care can exist only when there is a capacity to benefit (Culyer, 1976a, Williams, 1974). That is, an individual’s need for a particular health service is related to his/her potential to benefit from it. Within this framework, “health need is not an absolute but a measure for rationing scarce resources” (Eager et al., 2001). This sharply contrasts with the ill–health notion that ignores limitations on resources. Culyer and Wagstaff (1993) observed that a person might be ill but not need health care, particularly where effective treatment of that ailment may be non–existent. Indeed, they propose an alternative definition of need as the expenditure required to exhaust the capacity to benefit. This assumes that treatments will eventually meet their medical limits, and beyond this limit there is no need for health care. The key question would be what about those illnesses that, although not treatable, may require patients to be supported with expensive long term therapies, thus requiring ongoing resources?

The definition of need as “capacity to benefit” allows quantification of the amount of expenditure a person needs. The definition means that if resources are too limited to completely exhaust each person’s capacity to benefit, then, in that case, some needs would have to remain unmet. Proponents of the definition might argue that the degree of need could be measured by the amount of resources required to reduce them, if it is not possible

72 ______to exhaust them completely. However, an individual may also have the capacity to benefit from health care but may not be ill as in the case of preventive interventions (Culyer, 1995). Therefore, not everybody endorses the ‘capacity to benefit’ interpretation without reservation. As observed by McIntyre et al. (2000), “while the notion of ‘capacity to benefit’ is conceptually appealing, it is difficult to operationalise in some instances, particularly where health data are poor”. According to others, there is no point in focusing resources on health care if there is little chance that jurisdictions or people will benefit from them; however, they also acknowledged that there are problems with defining need as ‘capacity to benefit’ (Crampton and Laugesen, 1995, Mooney and Houston, 2004).

Indeed Mooney and Houston (2004) observed that ‘improving individuals or communities capacity to benefit will involve the addressing what they called Management Economic Social and Human (MESH) infrastructure of communities’. Management capability not only deals with how finances are managed but also the technicalities of how services are being provided. This would mean addressing leadership issues, processes of eliciting community perspectives on preferences and health care needs, including planning infrastructural improvement in order to facilitate how health services are provided. Similarly, building economic infrastructure would eliminate geographical barriers that hinder access to health services, including broader issues such as the ‘economic development of amenities, services and employment in a specific area’ (Mooney and Houston, 2004).

Social cohesion, on the other hand, is about how communities are organised in agreement and how they interact between themselves and across different government sectors. The human component of the MESH package relates to the effective deployment of relevant skilled human resources in underserved areas and ensures that appropriate incentives are in place to help attract and retain personnel. According to Mooney and Houston (2004), incorporating a MESH framework in any funding mechanism would require available funds to be separated into two components: the first is the funds targeted to various programs to meet the operating costs; the second is the funds for overseeing the implementation of the MESH infrastructure. Mooney and Houston are of the idea that the MESH approach to

73 ______‘resource allocation differs from more conventional approaches which normally concentrate on allocating resources according to primarily the size of the problem (often called health need and measured in terms of some assessment of the amount of sickness in a population)’ (Mooney and Houston, 2004).

However, the approach emphasises the importance of directing funds towards the development of a MESH infrastructure as a prerequisite for efficient utilisation of resources or rather improving the ‘capacity to benefit’ from resources (Mooney and Houston, 2004). This acknowledges the need for inter–sectoral collaboration in developing a foundation for effective financing of providers of Health Care services. The approach is based on the idea that resources should be allocated to reflect the good or benefit that policy seeks to pursue. This good is ideally that which is sought by and perceived by the community affected (Mooney, 2006). The preferences of local communities, however, need to be informed. In turn, these informed preferences must be set against realistic or prevailing resource constraints. It is thus important to understand what can be done to build a MESH infrastructure in relation to such issues of governance. A range of different requirements exists across different Area Health Services with different associated cost implications, such as distance to health facilities and health professionals’ travel costs amongst others. Hence each ‘community may need a tailored approach according to its needs: one size will not fit all’ (Mooney and Houston, 2004).

According to Mooney and Houston (2004), to invest in building up such infrastructure in communities where it is deficient is seen as a way to make communities’ capacity to benefit more equal. Thus building a MESH infrastructure relates to equity. Communities that currently have a low capacity to benefit from program monies because a MESH infrastructure is lacking might require investments made in MESH for this approach to be effective and so that in time their capacity to benefit from monies made available for health programs would rise (Mooney and Houston, 2004). The next section looks at the measurement of equity.

74 ______3.3.6 Measurement of Equity Since the definition of equity lacks consensus, it is obvious that there is no apparent agreement as to how it should be measured in practice. In most cases, equity has been measured from the perspectives of health care delivery or financing by examining payments and allocation of resources. Most of the empirical work on equity in the delivery of health care mainly concentrates on differences in utilisation of health care by different socio– economic groups (Le Grand, 1978, O’Donnell and Propper, 1991, Wagstaff et al., 1991a). On the other hand, the empirical literature on financing has focused significantly on households’ ability to pay for treatment (Wagstaff and van Doorslaer, 1998, van Doorslaer et al., 1993). Empirical works on equity in the delivery of health care are limited and the very first to be carried out was in the United Kingdom by Julian Le Grand. He used a ‘range measure of inequality’ to analyse the ratio between public expenditure per person ill for the highest socio–economic group defined in terms of occupation and that for the lowest (Le Grand, 1978, Le Grand, 1991). The work involved a two–level analysis by estimating the costs of seeking health care in the National Health Service (NHS) per person reporting illness in each socio–economic group.

To achieve this, each socio–economic group’s total imputed expenditure was divided by the number of persons reporting either ‘limiting long–standing illness’ (chronic) or ‘acute illness’ in the group. Second, he computed the share of expenditure received by each socio– economic group and compared this with the group’s share of ill–health (Le Grand, 1978). Using his study findings as a basis, he concluded that horizontal equity (equal treatment for equal need) has not been achieved in the NHS (Le Grand, 1982). However, his approach has attracted criticisms from O’Donnel and Propper (1991) and Wagstaff et al. (1991) and one of the main weaknesses often highlighted is Le Grand’s use of an aggregate morbidity measure as a proxy of need. Collins and Klein (1980) criticised his implicit assumption that only sick people receive treatment.

Other proponents of equity measurement like Wagstaff et al. (1991) used a concentration curve to measure equity. Like the other approaches, this approach focuses on horizontal equity, which is about equal utilisation for equal need. It attempts to quantify the degree of

75 ______inequity in the distribution of health care by first categorising and ranking individuals according to income, and then constructing and comparing an ‘illness concentration curve’ with an ‘expenditure concentration curve’. The authors measured utilisation by calculating the average number of services used over time or by imputed expenditure (Wagstaff, 1991b). Expenditure comprises the imputed resource costs of utilisation of primary health care facilities, as well as hospital in–and outpatient facilities. Self–reported illness was used as an indicator of health care needs.

As noted by Le Grand (1991), one of the weaknesses of this approach is the assumption that all persons reporting acute sickness are in equal need of health care and those reporting chronic illness also have equal need of health care (Le Grand, 1991). Other methods of measuring equity in distribution of health care, including the regression method (Mathers, 1994, Puffer, 1986), have been proposed and used. The next section examines the concept of ‘need’, which is central to both the interpretation and measurement of equity.

3.4 Health Needs: Definition and Measurement Explicit definition of need is crucial to make the equity principles of “allocation according to need” operational. The concept of ‘need’ has been debated vigorously but still remains controversial. According to Hurley (2000), the need definition equates need for health care with ill–health. The level of need, as defined by the severity of illness, means the sickest have the greatest need (this concept of need underlies the equity principle of “rule of rescue”7). Some authors argue that this definition of need is problematic because it ignores the limits of what is medically possible (Culyer and Wagstaff, 1993) or, as discussed earlier, where capacity to benefit is diminished with no returns. The argument is that if there is no effective treatment, no matter how ill a person is, it is questionable whether

7 According to Williams and Cookson (2000), most equity concepts are proposed in the context of decisions about population groups rather than individuals. Hadorn (1991) noted that, at the individual level, one principle for decision-making is the “rule of rescue”. Society and each individual have an ethical duty to do everything possible to help those in immediate life–threatening distress. This means that the patient with the most serious condition is treated first irrespective of the costs of treatment. The rule of rescue is identical to the concept of “allocation according to need”, where need is defined as severity of illness. Thus, severity of illness of a patient or patient group establishes priority for health care expenditure irrespective of capacity to benefit from treatment.

76 ______health care is needed. According to William (1974), ‘need’ for health care in its basic form implies that someone is better off with the ‘needed’ treatment than without it (Williams, 1974) and that the ‘better offness’ has to do with the persons’ health (Culyer et al., 1971).

On the other hand, Bradshaw (1972), in Bradshaw’s Taxonomy of social need began the debate over the interpretation of need as having four (normative, felt, expressed and comparative) dimensions. Normative need exists when an expert, professional administrator or scientist defines need by laying down their desired standard and comparing it with the standard that actually exists. Felt need is identified by equating need with want and is assessed by simply asking people if they feel they need a service. Expressed need is where the individual’s felt need is turned into action. Finally, comparative need exists where the characteristics of a population that receive a service are ascertained, and where people with similar characteristics that do not receive services are judged to be in need (Bradshaw, 1972). The debate since Bradshaw’s work was presented has done little to reduce the disagreement between scholars over the appropriate interpretation of need for health care, which is an even bigger challenge for policymakers as it still remains a contentious and disputed territory.

In the medical and philosophical literature, health need is equated with ill–health (Donabedian, 1973, Gillon, 1986). The interpretation, which deals largely with the individual’s pre–treatment health, has much appeal among clinicians. Indeed, Nord and Nord et al. (1996; 1999) defined ‘need’ as severity of illness and went as far as incorporating it into an economic evaluation tool they developed. The WHO definition of health needs is also geared towards ‘need’ as ill–health, although public health measures are incorporated. According to the World Health Organisation (1974), “health needs are to be understood as the deficiencies in health or social well–being that call for preventive, curative, rehabilitative or public welfare measures”. The interpretation of need in terms of ill–health generally suggests that people who are sicker than others have greater need for health care and should, therefore, receive special attention in allocation of resources. This has been criticised on several grounds including the fact that not all health problems are

77 ______amenable to health care intervention and also that such an interpretation ignores resource constraints (Mooney, 2000a, Acheson, 1978).

3.4.1 Approaches for Measuring Health Needs Measuring the needs of the population remains a problem despite the availability of a number of measurement models (Shannon, 1976, Anderson, 1978). Approaches for measuring health needs vary considerably; there is no consensus on any single method. For decades the use of mortality data as a proxy for health needs has enjoyed wide application (Mays, 1987, Birch, 1993, Pain, 1996). Mortality measures are largely based on correlations between age and sex standardised mortality ratios and various aspects of morbidity and risk of ill–health, particularly those conditions associated with the need for secondary care (Newbold et al., 1995, Forster, 1977, Bennett and Holland, 1977, Diderichsen, 2004).

However, the use of mortality data to measure health needs has several limitations. The first is the unclear relationship between mortality and need for health care. For example, the obvious question often asked is, is there a relationship where, say, a 10 percent higher mortality should motivate a 10 percent higher capitation? The second limitation is the limited focus of defining health need as the ‘capacity to benefit’. Most mortality measures fail to reflect the full extent of non–fatal morbidity (Gibson, 2002) and also ignore issues such as underlying determinants of health/ill–health and other aspects of social deprivation with respect to health care needs (Blane et al., 1997, Townsend and Davidson, 1988, Fox, 1978).

Recent empirical work on equity uses morbidity data as a proxy for health need. Le Grand (1978), as noted earlier, used aggregate morbidity measures as an indicator of need. He defined need as “the number in each socio–economic group reporting acute and/or limiting long standing illness” (O’Donnell and Propper, 1991). He calculated the age and sex standardised percentage of persons reporting ill (either chronic and/or acute illness) by occupational group to represent the proportion of the population in need of health care.

78 ______O’Donnell and Propper (1991) also used disaggregated morbidity measures covering non– limiting long–standing illness; limiting long standing illness; persons reporting neither restricted activity nor long standing illness; and persons reporting ‘not good health’, as an indicator of need. Equally, Collins and Klein (1980) used morbidity (specific) measures as proxies of need. They divided health need into different categories including the non–sick, the acutely sick, and the chronically sick, and compared resources received by each socio– economic group within each category of health need (Wagstaff et al., 1991a).

Similarly, Wagstaff et al. (1991) measured need for health care based on persons reporting either acute sickness or limiting long–standing illness, long standing illness alone or persons reporting ‘not good’ health (Wagstaff et al., 1991a). All the morbidity measures mentioned above, use self–reported ill–heath as an indicator of health care need. Apart from the approach of Le Grand, the others attempt to use disaggregated morbidity data. However, none completely manages to overcome the problem of collective measure of health need. Their heavy reliance on income and occupation to define a socio–economic group underestimates the link between other non–economic variables and health. It is also important to note that the use of self–reported ill–health as an indicator of health care need leads to the same problem of under–reporting of illness especially in countries where data is not available (McIntyre and Gilson, 2000).

According to Diderichsen (2004), if ‘need’ is proportional to morbidity, epidemiological data on the incidence and prevalence of disease would be appropriate indicators of relative need. Insofar as variations in mortality rates are proportional to morbidity rates, even mortality statistics would be useful. Although he acknowledges that in most countries, both developed and developing, there is epidemiological data that shows a declining mortality and increasing or unchanged morbidity in non–fatal conditions often with a different demographic and socioeconomic pattern. Therefore, mortality might be decreasing in relevance as a reflection of need. Equally, using mortality rates alone in such a situation might introduce ‘perverse incentive’, in the sense that effective care that lowers mortality rates would be punished by means of a declining budget. However, where both morbidity and mortality data exist within a health care system and it is judged as unbiased by

79 ______variations in record–keeping efficiency across institutions and regions, then it would be good to use both as an indicator of health needs and an instrument for making the allocation of health resources more equitable (Diderichsen, 2004, Zere and McIntyre, 2003).

In most countries, morbidity or mortality data are either generated within the health care system by registration of discharges and visits (hospitals, General Practitioners, births and deaths registration), and through other vital registration systems, or through population surveys. If mortality data are available from vital registration systems with good coverage data on age–specific mortality rates, such as infant– or under–five–mortality and other premature mortality, they might be good indicators. In Britain for example, age– standardised mortality rates were used for some years as the main indicator of need but currently they also utilise specific morbidity and some socioeconomic data (Smith et al., 1994).

Many countries use a combination of socioeconomic data (often called need factors) but the problem is that they are limited in terms of their ability to predict variations in health across individuals. Studies from countries with low–mortality show that even a combination of several demographic and socioeconomic determinants (e.g., age, gender, ethnicity, marital status, education, income and employment) rarely explains more than a few percentage points of the variation, and even with the introduction of data on earlier morbidity, the figures rarely reach beyond 20 percent (Kapur and Webb, 2000, Kaplan et al., 1996, Rice and Smith, 2001a). However, variations across geographical areas in settings where morbidity or mortality does not exist may be better explained by socio–demographic need factors (Diderichsen, 2004).

This study defines health needs as population health status as defined by premature mortality and chronic morbidity (two indicators of health status in this study). The study argues that, in order to improve equity in resource allocation, available financial resources should be in proportion to the greater morbidity and premature mortality within the population, especially the worst–off in society. Evidence has overwhelmingly shown the strong link between poor ill–health and all levels of socioeconomic disadvantage.

80 ______Therefore, to reduce inequities in health outcomes, health resources not only need to be re- focussed to target the actual health needs, but other sectors associated with social and economic determinants of health responsible for providing services that have positive effects on socioeconomic disadvantages also urgently need to be engaged. This study supports the aspects of vertical equity (the unequal treatment of unequals) meaning disadvantaged groups bearing a heavy burden of disease should receive preferential treatment when it comes to allocation of health care resources so as achieve more rapid improvements in their health outcomes.

3.5 Beyond Needs–Based Resource Allocation Formulas Despite the disagreements over interpretation and measurement of equity and health needs, there is consensus that equitable access to health services and improvements in health outcomes will be difficult to achieve without some form of equity in resource allocation (Eaves, 1998, Culyer, 2001, Braveman and Gruskin, 2003) ideally based on health needs. For many years, health resources have been allocated using various population and needs– based mechanisms; however, inequities still exist within those health care systems globally. Needs-based formulas fail because they are not based on measureable indicators that fully reflect the actual health needs of the population. For example, Sweden is one of the countries operationalising a resource distribution formula and although life expectancy is higher than in most developed countries, health equity is still a major issue in Swedish health policy (SOU2000:91, 2001) as it is in many other countries such as England (Department of Health, 2009, House of Commons Health Committee, 2009), Canada (Health Canada, 2008), and New South Wales (Garling, 2008, NSW Health, 2008).

Therefore, the question is: what are these countries and others doing wrong and what is the most equitable and fair mechanism for allocating resources? In many countries, previous budget allocations, current service or facility locations, capital development or political factors heavily influence resource allocation (Green et al., 2000). In the last half–century, health reforms in many countries have seen a shift from historically–based systems of resource allocation to the development of different funding formulas based on population

81 ______health needs and others driven by market mechanisms. The next chapter (Chapter 4) reviews some of the prominent mechanisms that have evolved around the world and their implications for equity in resource allocation.

82 ______Chapter 3: Summary

Chapter 3 has reviewed the definitions and main levels of resource allocation, including key concepts underpinning health resource allocation and theories of social justice that shed light on how health and health care resources ought to be distributed in general and in New South Wales in particular.

The major theories (Robert Nozick’s theory of entitlement, the utilitarian perspective, and John Rawls and egalitarian theory) of justice as fairness were reviewed. Other theories examined were Norman Daniels’ extension of Rawls’ theory to health care and Amartya Sen’s basic capability theory.

The theory of justice as fairness adopted as the basic framework for this study was Rawls’ ‘Difference Principle’ which states that 1) inequalities in society must be organised to benefit the worst–off members, 2) that societal resources be used to counter advantages conferred on some people by accidents of birth such as race, ethnicity, gender etc., and 3) that society should enhance equality of opportunity by improving the long–term expectations of the least favoured, particularly by investing in education. This theory was preferred because of its compatibility with the objectives of this study.

Although Rawls’ primary goods did not include health, Norman Daniels’ extension of the theory has shown that health fits well in the Rawlsian analysis, given that disease and ill– health affect the normal functioning of human beings, and hence, impact on the range of opportunities available to individuals.

However, this study does not share Rawls’ view of not seeking redress for past injustices but rather adopts a stance drawn from Robert Nozick’s principle of rectification of past injustices to support the need to address the imbalances in the resource allocation system in the short and medium term, including improving long–term expectations of the most disadvantaged population groups, especially the Indigenous population in Australia.

The resource allocation goals (efficiency and equity) pursued by policy makers in health care systems have been discussed, including the various economic evaluation appraisal techniques applied in health services and programs to determine their costs and effectiveness.

This chapter has also covered the concept of equity and provided a detailed review of the debate over interpretation of equity. Key issues considered were the differentiation of equity from equality and the discussion of vertical and horizontal forms of equity. Horizontal equity refers to equal treatment for equal need while vertical equity is defined as unequal but equitable treatment of unequals.

The working definition of ‘equity’ developed in this study is equality of access to financial resources for health care on the basis of health needs. Health needs are defined in terms of population health status. This implies that for health resources to be deemed as equitable, they have to be distributed according to assessed health needs (health status as defined by premature mortality and morbidity) of the population.

Population health status in this study is measured by premature mortality and chronic morbidity.

83 ______

Chapter 4: RESOURCE ALLOCATION MECHANISMS

Overview Health care systems have the duty of mobilising resources and allocating these effectively across competing needs. Resource allocation decisions are made at different levels of the health system, especially in health systems that are decentralised. The New South Wales Health Department uses several mechanisms for allocating resources to different health jurisdictions and to programs and services. This chapter reviews the mechanisms for allocating resources at different levels of the NSW health system. It begins with a review of the international literature on resource allocation mechanisms that have relevance for the development of resource allocation mechanisms in New South Wales.

4.1 Introduction: Health System Financing and Resource Allocation Allocation of resources to competing health needs is an important facet of health care financing. The performance of health care systems, including the equity and/or efficiency with which they function, depends largely on the way in which their inputs are allocated. The term health care financing (HCF) is broadly seen as encompassing mobilisation of funds and the allocation of these funds to various regions, population groups and specific types of health care activities (programs and services) (Hsiao, 2007, Hsiao and Liu, 2001). The challenge of effective HCF is largely seen as dealing with how health care resources are organised and the means by which these resources are allocated through the health system to maximise population health outcomes (Ensor and Witter, 2001, Carr–Hill et al., 1994). The financing mechanism defines, for example, the level of financial protection offered against unexpected costs of illness, especially to disadvantaged population groups with limited ability to pay for health care (Hsiao and Liu, 2001, Wagstaff and van Doorslaer, 1998, Williams, 1993).

The term resource allocation is defined primarily as the distribution of health care financial resources among competing health needs (Eager et al., 2001). Different countries have different mechanisms for distributing health care resources. These range from historical–

84 ______based funding models to population needs–based approaches and market–driven mechanisms such as episode or casemix funding (Midwinter, 2002, Mooney, 2006, Birch and Chambers, 1993, McMillan, 2002, Rice et al., 2000, NSW Health, 2005). A common purpose for developing and using mechanisms for resource allocation is to ensure that resources are allocated in the most efficient and equitable manner possible. It is noteworthy that in many countries not all aspects of resource allocation are controlled by governments since consumers of health services allocate significant finances directly (and/or indirectly) into the system; the UK is the exception in many respects since it is a universal fully tax funded health system. The next section reviews various resource allocation mechanisms (see Figure 4–1) as developed and applied in other parts of the world.

Figure 4–1: Resource Allocation Mechanisms

4.1.1 Historically-Based Funding Prior to the mid –1970s, funding distribution based on the previous year’s expenditure, generally referred to as ‘historical–based funding’, was the original approach to funding health care systems in most countries. The historically–based funding model distributes resources on the basis of the previous year’s expenditure (Eager et al., 2001, Pearson, 2002). In this type of funding distribution, the funder or policy makers made decisions to provide a little more in form of adjustments or ‘growth’ in funding to reflect the current

85 ______financial year. According to Maynard and Ludbrook (1980), this system is about “What you got last year, plus an allowance for growth, plus an allowance for scandals” (Maynard and Ludbrook, 1980a) equals the budget allocation for that year.

In the context of the British National Health Service (NHS), May and Bevan (1986) noted that the increments that funders or policy makers provide at their own political discretion tend to go to the “noisiest rather than the neediest” (Mays and Bevan, 1986). Indeed, Rice and Smith (2001) also noted that resources are often distributed as they have always been in the past for administrative reasons rather than according to any independent formula. They were of the opinion that, in the short run, it is realistic to provide funds for efficient use of a service structure that is already there, but warn that, for historical reasons, existing infrastructure is often poorly distributed; therefore, in practice this method of resource allocation will perpetuate inequitable patterns of resource distribution (Rice and Smith, 2001b).

Historical–based funding has been criticised for failing to address equity and efficiency objectives (Green et al., 2000, Pearson, 2002, Eager et al., 2001). Eager et al. (2001) noted that this type of funding system, especially in New South Wales, Australia, as in many other countries, fails to match the health needs of the local population and the available resources and services. The authors also observed that, although new population growth has taken place on the outskirts of most capital cities over the last several decades, the use of a historical funding model has ensured that funds are continually channelled to existing facilities (Eager et al., 2001). Therefore, the use of existing patterns of expenditure as the basis of current allocation ignores the degree of unmet need in more disadvantaged population groups (Sheldon and Smith, 2000).

Furthermore, Hanson et al (2001) observed that whether allocations are decided according to historical budgets, number of beds, or utilisation rates of the services, the result is often the same: an increasingly inequitable, circular process of high utilisation – more resources used to create more supply – which, in turn, generates more demand and use by some groups and correspondingly less by others, especially the most disadvantaged. For example,

86 ______evidence from both developed and developing countries has demonstrated that budget share for providers in urban and affluent areas grows, often contrary to an explicit political ambition to move in the opposite direction (Hanson et al., 2001).

The other limitation as far as historical funding models are concerned is that they provide incentives to health authorities to spend their allocations in order to guarantee the following year’s allocation, thus leading to inefficiency in funding use (Eager et al., 2001). Irrespective of the historical funding model limitations, some countries still allocate health care resources across geographical regions, districts and health plans on a historical basis, with some adjustments achieved through political negotiation (Belli, 2004). However, other countries and states such as Sweden, New Zealand, Scotland, South Africa, New South Wales (Australia), and Alberta (Canada) have developed a population and needs–based resource allocation system modelled on the basis of the Resource Allocation Working Party (RAWP) approach (Bedard et al., 2000, McIntyre and Gilson, 2000, Lake, 2000, Rice et al., 2000, Sheldon, 1997, Diderichsen, 1997) introduced in 1976 in the UK (Department of Health, 1976). The next section reviews the RAWP as the innovator of all other population and needs–based resource allocation models.

4.1.1.1 United Kingdom – RAWP The UK Resource Allocation Working Party (RAWP) was the innovator in developing the technical population and needs–based approach to allocating health care resources. Since the 1976 appointment of the RAWP, health care equity has been an explicit goal of resource allocation within the NHS. The RAWP and the NHS led the world in pioneering scientific methods of equitable allocation of health resources (Carr–Hill et al., 1994, Shaw and Smith, 2001). According to the health care equity principle, resources were to be geographically distributed to ensure ‘equal opportunity of access to health care for people at equal risk’. As noted by Mays and Bevan (1986), geographical inequalities in resource allocation were evident since the creation of the NHS in 1948, as its creation was influenced by the desire to achieve a more equitable distribution of Hospital and Community Health Service (HCHS) resources. Prior to the RAWP, resources were

87 ______allocated on a historical basis and were largely thought to be inequitable (Maynard and Ludbrook, 1980a).

Therefore, the aim of the commissioning of the RAWP was to recommend a system of allocating resources within the NHS which would be responsive to the health needs of the population, and to identify and correct inequalities in the existing pattern of resource distribution (Department of Health, 1976). The equity criterion adopted by the Working Party was equal opportunity of access to patients in equal need, regardless of where they live (Paton, 1985, Shaw and Smith, 2001). In 1976, the Working Party reported its findings (Mays and Bevan, 1986), recommending that revenue resources for hospital and community health services (HCHS) should be distributed on the basis of population, weighted according to the ‘need for health care’ and the costs of providing services to different geographical areas (Carr–Hill et al., 1994). The formula was to account for geographical variations in the need for health care by conducting a detailed assessment of the comparative health needs of different populations to determine their appropriate shares of resources; however, it proved to be one of the major challenges faced by the Working Party. In the design of the formula, four key variables were used – population size, age and sex structure, morbidity, and cross–boundary flows. The formula also made adjustments for service increments for teaching hospitals (Smith et al., 1994, Gibson et al., 2002).

Among the indicators used in the calculation of allocation targets, population size was the only indicator that attracted less attention compared to other variables in the formula (Mays and Bevan, 1986), possibly because the use of population figures in resource allocation is not uncommon. However, this is not without controversy as the choice of a basic population figure involves a choice of an appropriate base date, particularly in areas where population change is more rapid than elsewhere (Buxton and Klein, 1978). For revenue targets, the Resource Allocation Working Party used the latest available mid–year population estimates and for capital targets, a five–year forward projection of population was recommended (Department of Health, 1976).

88 ______In the intervening period, much of the technical effort that has gone into formula determination has focused on the measurement of the ‘need for health care’. This is conceptualised as comprising two elements, ‘age–related need’ and ‘additional need’. The latter concerns that part of a population’s need for health care that is over and above that which is due to its demographic composition. In other words, this addresses the effect of socio–economic deprivation on a population’s health care needs (Asthana et al., 2004). Unfortunately, whilst the link between deprivation and health status is well established, it is difficult to quantify this relationship in terms of resource needs since socioeconomic determinants of health often fall outside the health budget portfolio. The variety of different approaches used since 1976 to measure additional need is symptomatic of this difficulty (Carr–Hill and Dixon, 2006).

The Working Party acknowledged that, in addition to population size, demographic characteristics also influence the need for health care (Mays and Bevan, 1986). Consequently, in developing a measure for relative need, the population of each area was weighted according to its age and sex structure. By weighting the population for age and sex structure, the RAWP took into account the variations in the use of health resources by the different age and sex groups. The age and sex structure was accounted for in the form of national utilisation rates for each age and sex group (Smith et al., 1994). In addition, morbidity was also accounted for in the RAWP formula. However, the difficulty was to find an appropriate measure of morbidity. In the absence of any adequate proxies of morbidity, the Working Party adopted standardised mortality ratios (SMR) as the best index of morbidity and a proxy of health needs (Department of Health, 1976). The SMR is a single index number which compares the mortality experience of a given region’s population to that of a reference population (Asthana et al., 2004, Bedard, 1999). In 1986, the RAWP formula was revised to incorporate deprivation as an indicator of need in order to address some of the problems arising from the use of standardised mortality ratios (SMRs).

The RAWP formula was used from 1977 to 1990 (Diderichsen et al. 1997) and, according to earlier evaluations, it was successful in redistributing resources from metropolitan

89 ______English regions to the poorer regions in the north (Holland, 1986, Shaw and Smith, 2001). However, the RAWP evolved and, in 1995, it was superseded by the Resource Allocation Group (RAG), which was in turn replaced by the independent Advisory Committee on Resource Allocation (ACRA) in 1997. The ACRA makes recommendations to Ministers on possible changes to the formula, prior to each round of Primary Care Trusts (PCTs) revenue allocation (Department of Health, 2005a). Throughout the evolution of the RAWP, ‘equal opportunity of access for equal needs’ has remained a key objective of resource allocation within the NHS. In 1999, however, the ACRA introduced an additional requirement, that resource allocation should ‘contribute to the reduction of avoidable inequalities in health’ within England (Department of Health, 2005b).

However, over the thirty years since the introduction of the first needs–based funding mechanism and almost ten years since the commitment to reduce avoidable inequalities in health, England, like many other countries, is still struggling with inequalities in health (House of Commons Health Committee, 2009, Buxton and Klein, 1978, Department of Health, 2009), especially between the well–off and the most disadvantaged members of the society. Debate about the distribution of resources in the health system has been renewed and the government has recommitted to tackling health inequalities by shifting NHS resources which would otherwise be spent on general services and interventions of benefit to the entire population towards those people who have the worst health outcomes (Department of Health, 2009). The funding formula used by the Department of Health to allocate resources to Primary Care Trusts (PCTs) is ‘weighted’ to allow for the extra health needs faced by disadvantaged areas and population groups (House of Commons Health Committee, 2009). But how are population and needs–based resource allocation mechanisms derived?

4.1.2 Population and Needs–Based Resource Allocation Mechanism Following the work of the UK Resource Allocation Working Party (RAWP), there has been a growing body of literature discussing the need to reorient health care resource allocation to make health systems more responsive to local health needs (Birch and Chambers, 1993, Bourne, 1990, Gilbert et al., 1992, Mays, 1995, Lake, 2000, Resource Allocation Working

90 ______Party, 1976). Needs–based formulas of resource allocation attempt to distribute resources on the basis of the health needs of a population as opposed to historical patterns (Diderichsen, 1997, Gilbert and Bennett, 1991). These resource allocation models generally operate at the level of the individual and the population at large. At the individual level, the resource distribution authority provides funds to meet the identified needs of the individual consumer or patient.

At the population level, needs–based models recognise that funding should be distributed according to the size and needs of a defined population (Eager et al., 2001). Models of resource allocation based on population and needs create the opportunity for equalising distribution of resources with the level of health needs across geographical regions. They also have the advantage of facilitating a gradual, smooth transition process from historical allocation to a new regime where redistribution occurs over time, thereby giving the historically under–funded or ‘losers’ in the redistribution process the opportunity to adjust and the over–funded or ‘winners’ enough time to develop a new planning capacity (Belli, 2004). Most population and needs–based allocation models have their origins in the work of the RAWP (1976) as discussed earlier. Since the Working Party’s report, many countries, predominantly in the industrialised world, have developed their own needs–based models for distributing resources within the health sector. Most of these formulas are designed based on the premise of achieving equal access to health care for equal needs (Culyer, 1976a, Culyer, 1995). ‘Need’ is discussed in detail in Chapter 3.

The concept of needs–based resource allocation across the different public services provides scope for intensive debate. In most cases, resource allocation relies on selection and weighting of various indicators that are often politically influenced (Midwinter, 2002). However, since the 1970s, needs–based capitation systems of health care resource allocation have become the influential approach for financing health care needs within many countries. The central problem to date is how to design a reliable methodology and practical tools that can correctly estimate population health needs. In a needs–based system, resource allocation is based on the central principle of horizontal and vertical equity which advocates a policy of “equal treatment of equals”, while those who are different in relevant

91 ______respects are treated proportionally differently (Rice and Smith, 2001b). Nevertheless, the allocation of resources in most health systems worldwide remains a complicated procedure and consists of direct needs–based, historical, health program resource allocation alongside more recently developed market mechanisms of allocating health care resources (Talbot– Smith and Pollock, 2006), such as the casemix (episode funding).

Globally, the most influential and universally acceptable code is the notion that health care resources should be distributed on an equitable basis according to the health needs of the population. However, a central problem that hinders this approach is that there is no universal definition of the term “need” as discussed in Chapter 3. Despite the lack of consensus in defining “need”, many alternative approaches have been developed since the 1970s aimed at improving needs–based methods of resource allocation. The British National Health Service (NHS), an innovator in many aspects of health system arrangements, introduced the first new method of allocating health resources in 1976, initially to geographically–defined planners and then to health authorities, primary health trusts and other health care purchasers (Talbot–Smith and Pollock, 2006).

According to Coast et al (1996), as a result of the initial Resource Allocation Working Party (RAWP), various integrated health systems have developed needs–based resource allocation mechanisms (see Table 4–1).

92 ______Table 4–1: Comparative Needs–Based Models of Resource Allocation

Source: Smith, 2001 and various policy documents

Most of these needs–based approaches are based on two broad conceptual approaches: firstly, that resources are allocated taking into account the general burden of disease in the population measured by healthcare utilisation data, registration of cases, or population self– reports of morbidity or, indirectly, using data derived from censuses or surveys (Petrou and Wolstenholme, 2000); and secondly, that the mechanisms simultaneously seek to consider the clinical concerns, ethical norms, and economic considerations of need and determine “priorities which incorporate clinical and cost–effectiveness and patients’ perspectives” including ‘capacity to benefit’ from resources (Mooney and Houston, 2004, Wright et al., 1998, Mooney, 1998a, Mooney, 2006). The two different mechanisms primarily employ a

93 ______block contractual system with a little uncertainty about the expected level of available sources of health care intervention and restricted prospects for opportunistic behaviour by medical staff (Marini and Street, 2006).

In many countries, almost all direct needs–based resource allocation methods use formulas of weighted capitation which, theoretically, should secure distribution of resources according to the principle of equal opportunity of access to health care for people of equal need. This is because they try to identify and meet health care needs by taking into account the most relevant socioeconomic and demographic factors that determine health status and healthcare needs. However, there is much debate about statistical techniques and explanatory indicators that must be included in the formula (Rice et al., 2000). Among others, Carr–Hill and Sheldon (1992) warned that proper allocation is not just a technical problem of application of statistical techniques. However, the fundamental challenge is to “find a combination of factors reflecting dimensions of need and then appropriate weights with which to combine them, within the context of the guiding principles of equity and efficiency” (Carr–Hill and Sheldon, 1992).

Petrou and Wolsternholme (2000) point out that unavailable “time, resources and commitment are required to collect the necessary information”. In addition, to take into account the specifications of different health care services, distinct formulas must be applied in different services (Talbot–Smith and Pollock, 2006). The geography of an area must also be taken into consideration since it plays a role in terms of an area’s social capital, and in differences in costs of providing health care due to the geographical distribution of population and facilities, which, in turn, creates different opportunities of access to services. Lack of consideration of an area’s features may lead to a biased picture of the true needs of a population (Rice and Smith, 2001b). Carr–Hill et al (2002) and Judge et al (1994) suggested that employing relatively sophisticated measures of socioeconomic conditions and more advanced statistical methods with strong theoretical justification would be a far better tool for resource allocation purposes than using indirect census based proxies (Carr–Hill et al., 2002).

94 ______Similarly, in addition to technical models for distributing resources, communitarian and individual approaches to the problem must also be distinguished. The opponents of the model are concerned that relying on the concept of “technically derived needs” as the basis for allocating countries’ resources in health care systems does not sufficiently take into account “the values of the community whose needs are to be served by that health service” because medical professionals, health service planners and politicians, rather than the community, have the final word in the allocative decisions (Mooney, 2006). Nevertheless, even if it is assumed that health systems manage to allocate resources equally to all communities according to their needs and interests, as Miller and Stokes (1978) explained, increasing the supply of health care to the population is not always directly associated with improved health. This notion leads to a distinct, individualistic approach to the problem of resource allocation that is concerned with inequalities in access to health care services, “a circumstance that exists when race, income, or the availability of insurance predicts utilisation rather than need” (Litaker and Love, 2005). In other words, various socioeconomic determinants of health factors (unemployment, education, income, insurance status, environment, etc) that determine access to health service are not affected by policies of effective resource allocation (Grumbach et al., 1997).

Equally, integrated health systems do not always manage to distinguish between legitimate and illegitimate health care needs. The former assumes that “the average level of care as historically delivered to each type of person equates to the desired package of health care”, while the later emphasises the unmet needs of a disadvantaged population that “is a difference between the desired package of care and the package actually received” (Rice and Smith, 2001b). In reality, a formula approach unilaterally rarely determines actual allocations. In the United Kingdom, real financial flows are based on the earlier allocations adjusted over the current financial year; for example, primary care prescribing budgets have largely been determined by spending in previous periods (Talbot–Smith and Pollock, 2006). Consequently Oliver (2005) has argued that this system secures the opportunity for provision according to the principle of equal access to equal needs rather than the actual provision of this principle (Oliver, 2005). At the same time, an allocation based on a previous year’s budget (historical –based system) is inflexible and can lead to unmet needs

95 ______when demand does not match the purchaser’s expectations. In addition, these arrangements create little incentive for providers to exceed activities over already contracted resources or to reduce their costs (Marini and Street, 2006). But are market driven approaches to resource allocation better in capturing the health needs of the population?

4.1.3 Market Driven Resource Allocation Mechanisms Globally, many countries have started to evolve from integrated models of population and needs–based funding formulas to funding arrangements defined by the market, although the outcomes of the changes are controversial, and the rational factors that drive such a transformation are not clearly explained. According to van de Ven (1996), the main justification of the reforms is to make resource allocation “more efficient, more innovative and more responsive to consumers’ preferences” than centrally integrated health systems (van de Ven, 1996). In contradiction to this statement, Besley and Gouveia (1994) argue that the adjustment of health care resource allocation, is in large part caused by innovations because technology that is widely employed in all sectors of health systems is continuously changing and advancing, affecting costs and methods of health care delivery (Besley et al., 1994). However, other amendments are based on the need to improve universal access to some health care services.

Proponents of private sector involvement argue that such a model is capable of improving needs–based resource allocation through an explicit account of disadvantaged groups and communities. Further, they argue, it can enhance the degree of democratisation and accountability of the medical profession because resource allocation would be determined by purchasers (i.e. Area Health Services, Primary Care Trust, Regional Health Authority, an insurer) who are close to and represent the interests of their populations (Saltman and Figueras, 1997). In addition, in comparison with the old integrated health care system, the new model potentially has an advantage in its flexibility in adapting to continuously changing health care needs (van de Ven, 1996). The private market is also more innovative and better informed and thus is capable of encouraging cheaper per capita and cost effective

96 ______decisions for a given level of needs (Rice et al., 2000). However, these arguments have not been carefully evaluated to determine their practicability.

A key point for understanding market–oriented reforms is that the rise of markets involves an incremental application of various market tools to different aspects of the health care system rather than the complete alteration of one kind of a system to another (Ranade, 1998). Van de Ven (1996) identified three major trends observed in health systems that have major implications for health care resource allocation: 1) the depletion of vertically integrated systems and the separation of the health care purchasers and the providers, but instead establishing contractual relationships with each other; 2) the tendency towards managed competition among health care providers; and 3) the introduction of policies on patient mobility and the free choice of providers (van de Ven, 1996).

The most important implication of these reforms is the introduction of activity–based or episode–based (casemix) resource allocation and funding: a system of paying hospital and other health care providers on the basis of episodes of care provided rather than previously applied to defined budgets based on needs assessment methodology (Marini and Street, 2006, Duckett, 2008b). The contracting process between suppliers and providers is also undergoing modifications. Unlike the integrated systems where budgets were defined by t bloc contracts, the new system relies on cost–and–volume and cost–per–episode or case contractual relationships, in which payments are closely linked with the services offered. Although new contracting practices differ widely across countries and the types of contracts vary along with contracting parties and mechanisms, legal status and comprehensiveness, which are the central characteristics, still remain the same (Saltman and Figueras, 1997, Duckett, 2004).

Therefore, the market oriented health care systems either introduce a new or modify the old mechanisms of health care provider reimbursement and, along with direct regulations, reimbursement incentives are the mechanisms by which governments attempt to influence providers’ performance (Duckett et al., 2008). Generally, the most common reimbursement practices across health care services tend to be direct payments to providers based on Diagnostic–Related Group (DRG) casemix cost evaluation (Oliver, 2005, Eagar and

97 ______Hindle, 1994). In primary and preventative care, different forms of capitation are intensively applied, typically adjusted to the main socioeconomic and morbidity indicators in order to promote equity and/or encourage efficiency, whereas for out–patient health care the dominant financing method is the fee–for–service payment (Szende and Magyorosy, 2004). Nonetheless, most often the different methods of reimbursement coexist within the same health care system.

According to Gay and Kronedfeld (1990), the gradual evolution of an activity or episode– based resource allocation can be traced to the United States where, from 1983, most reimbursement for health care providers was based upon the Diagnostic–Related Group (DRG) where patients within different categories were classified as clinically similar and were expected to use the same level of hospital resources8 (Lowe, 2000). However, the DRG represented only the first attempt to employ a market approach to productivity and to control scarce resources through enacting expenditure limits, and had substantial effects on improving the utilisation of resources rather than on their reallocation (Bardsley et al., 1989). Nonetheless, the role of fixed DRG prices is indispensable and its scope is further enhanced for the New South Wales (NSW Health, 2008) and United Kingdom’s (Kimberly et al., 2008) versions of comprehensive activity –based payment policy which is “Payment by Results” (expected to be fully operational by 2009). Within such a system, hospitals are reimbursed with fixed prices after the treatments or operations take place. The disadvantages associated with this kind of resource allocation are significant as well. Providers can gain if they manage to bring down the costs that might be achieved purely at the expense of reduced quality of services. They can also reclassify the status of treatment to a more expensive one in order to get more money (King’s Fund, 2005, Viney et al., 1991).

Resource allocation is also substantially affected by rising transaction costs, which are not mainly and only associated with insurance–based health care systems, but also with

8 The Diagnostic–Related Group (DRG) was almost immediately transferred to Australia (1988), Italy (1988), Norway (1987), West Germany (1986) and France (1985). More recently, Lowe (2000), after evaluation of DRG coding in large hospitals in New Zealand, also found that the new casemix budgeting was instrumental “to make a difference within organisation” (Lowe, 2000).

98 ______national health systems’ emerging contractual arrangements. Replacing the practice of block contracting by episode–based resource allocation is increasing transactional costs associated with control of volume, collection of data monitoring and enforcement of contracts (Marini and Street, 2006). However, the extremely high transaction costs of private insurance systems still remain one of the main barriers for entrance into insurance coverage for a significant proportion of the population with a high degree of health care problems. For example, in the United States, the transaction costs can reach 30 per cent of all money spent on health care (Pollock et al., 2007). As a consequence, resource allocation becomes less determined by a population’s health care needs.

As observed by Pollock (2007), activity–based resource allocation may result in the abolition of less profitable services, while more resources are unreasonably gravitated to fewer financially attractive treatments and localities, deteriorating the principle of equity in health care (Pollock et al., 2007). Moreover, according to Koivasulo (2003) in recent European experience there is also little evidence that market instruments of health care provider reimbursement lead to improved and lower cost health care systems with better resource allocation. Indeed, Koivasulo has argued that episode–based resource allocation brings new problems in terms of cost–containment and equity considerations (Koivasulo, 2003). The next section discusses financial resource allocation in New South Wales, Australia.

4.2 Resource Allocation in New South Wales

4.2.1 Introduction The NSW Department of Health is one of many players in health resource allocation. Currently, the Department allocates the health budget to the eight Area Health Services, using three main funding mechanisms: cost or expenditure–based funding, casemix or episode funding, and needs–based funding (known as the Resource Distribution Formula) (NSW Health, 2008). The NSW Health Resource Distribution Formula (RDF) is concerned with “health care equity” rather than “health equity”. According to NSW Health (2005) “RDF provides an important way the Government can be accountable to the people of

99 ______NSW in terms of achieving a key objective of health – equity of access to services”. However, NSW Health’s ability to influence health care equity is limited in that it does not “control” the funding of private hospitals and nursing homes, pharmaceuticals (except those supplied through public hospitals), general practitioners (GPs), specialist medical practitioners (except those who work in public hospitals), private laboratory and medical imaging services, and much allied health activity and research. As noted in Chapter 2, the NSW State health funding contribution is only around 24%, while the rest of the funds are from the Federal government (46%) and the non–government sector (30%). The next section discusses the various funding mechanisms used in NSW.

4.2.2 Cost/Expenditure Base Funding Cost/expenditure–based (also known as input–based) funding involves the use of cost information from the Area Health Services (historical or expected) to estimate how much should be allocated to a particular area. Information on cost of service provision may relate to single episodes of care, or even components of episodes such as actual radiology or operating room procedures. These costs may also be estimates of total cost of care provided over a prolonged period such as a financial year. The main advantage of cost–based models is their operational simplicity; they are easy to design and implement. However, there are many serious weaknesses with these models; in particular, they create incentives for spending in order to secure more funding. For example, a hospital can generate additional revenues by providing more billable services such as more diagnostic tests and procedures and keeping patients for longer than normal periods (Hindle, 2002). In short, cost–based models offer few rewards for cost containment.

4.2.2.1 Casemix/Output–Based Funding Casemix (or output–based) funding is based on hospital episodes of patient care. Under this type of funding hospital products are defined by a classification which is used as a basis for funding (NSW Health, 2006c, Duckett, 1998). The number of episodes is usually adjusted to take account of the relative costs of each type. There are three main variants of casemix funding: the payment model, budget share formula and the purchasing formula. Payment

100 ______models involve the setting of payment rates in advance for each type of episode of care. The hospital receives payment every time an episode occurs (NSW Health., 2003a, NSW Health, 2006c). However, payment may be adjusted depending on changes in casemix and the number of episodes.

Australian states and territories are independent and are free to determine their own state– funded and state–managed health services. In 1990, Scotton and Owens, in a National Health Strategy report, presented different payment options for paying hospitals based on output (Scotton and Owens, 1990). The aim was to provide a broadly based system of payment across public and private hospital sectors, acute and non–acute services, Commonwealth and state funding authorities and the health insurance system. Some of the proposed funding options in their report were various methods of using casemix, either minimally or comprehensively, whereby the Commonwealth government would pay hospitals directly based on hospital output. Following this report, in 1993, Victoria was the first state in Australia to introduce payment of public hospitals based on output classified by DRGs. The introduction of casemix–based hospital funding in Victoria and subsequent modifications to the funding formula was followed with close interest by the federal authorities and other state health departments (Grant and Lapsley, 1993).

After Victoria became the innovator of casemix funding in Australia, other states also implemented the funding mechanism. South Australia was the first to follow, introducing the model in 1994–95 and Western Australia, Tasmania and Queensland in 1996–97. And as observed by Duckett (2000), New South Wales “is the only Australian state that has not started a formal casemix funding approach for its hospitals”, but even though its not officially introduced, NSW applies casemix policy to ‘inform’ budget–setting processes and distribute budget cuts (Duckett, 2000). Compared to the other states, NSW has adopted the view of “approach with caution” and this state has considered casemix as a management tool rather than a funding tool (Stoelwinder and Viney, 2000). Some officials from NSW Health Department expressed particular concern that the casemix based payments would cut across the equity objectives of the existing resource allocation process, using the then Resource Allocation Formula (RAF) (Viney et al., 1991). Other concerns expressed

101 ______elsewhere regarding episodes of patient care using DRGs were that the method of payment would increase incentives for patients to be moved through the hospital more quickly, out of the inpatient section to other forms of care not covered by the DRG process, resulting in reduced quality of care. Although the aim of the per episode-based DRG mode of payment was to reduce hospital inpatient costs, these concerns warrant special attention.

In NSW, although a casemix approach was not fully implemented, Viney et al (1991) observed that casemix information has been an important component of the NSW Resource Allocation Formula (RAF) and the Resource Distribution Formula (RDF), and that casemix has assumed a growing role in the allocation of funds in NSW. For example, Area Health Services (AHSs) in NSW are encouraged to determine hospital budgets by casemix. Overall, NSW “has not ignored casemix but has relegated it to a lesser role” (Stoelwinder and Viney, 2000).

4.2.2.2 Budget Share Formula The budget share formula is most relevant where a large insurer, such as a state health authority, is responsible for providing most of the revenue to a group of hospitals (NSW Health, 2004b). The insurer has a fixed (or capped) total budget that is shared appropriately among the hospitals. Usually the budget is divided among the hospitals in proportion to their casemix–adjusted workload. A hospital that has more episodes of care would receive a larger share of the budget, taking into account episode types and costs. The key distinctive feature here is that, since the total budget is fixed in advance, only the shares for individual hospitals would be affected by changes in volume or casemix. If, for example, more patients were processed or casemix became more complicated, hospitals would normally have to do the extra work without additional compensation.

4.2.2.3 Purchasing Formula The purchasing formula variant of casemix has a distinctive feature involving negotiation of contracts between multiple funders (or purchasers) and providers with respect to groups

102 ______of patient care episodes (Hindle, 2002). All forms of casemix–based models present technical difficulties associated with the requirements to determine a comprehensive classification of products and to update it regularly to take account of changes in clinical practice. They also do not offer any direct encouragement to promote equity of access to services since payments are based on episodes of care irrespective of the socio–economic status of the patient and are made in advance. However, they enhance equity of funding across hospitals and other health care units by shifting budgetary savings from one hospital to another and offer the best incentives for cost containment. Episode –based funding is a form of “yardstick competition” designed to stimulate greater efficiency in contexts where competitive pressure is lacking. They also offer greater transparency in the financing of health care, relating provider revenue directly to workload (Street et al., 2007). Finally, they provide greater opportunity for clinician involvement.

4.2.3 Episode–Based (Casemix) Resource Allocation in NSW In 1990, Area Health Services (AHSs) in NSW started using episode –based funding by calculating the shares of each hospital by use of Diagnosis Related Groups (DRGs) data weighted by average cost, despite the difficulties associated with episodic funding. Resources allocated to AHSs through the Resource Allocation Formula were separated into major product types (for example acute hospital care, community health, etc) largely on the basis of historical expenditures. However, resources to particular health care delivery units were allocated in a variety of ways, including output (casemix) data with respect to the funding of hospital inpatient services. A target production level was set for each hospital, and then its share of the AHSs budget was determined by computation of its casemix– weighted share of production (Eagar et al., 1997).

In the absence of a clear state–wide policy in NSW Health, AHSs increasingly began to share ideas and information about casemix funding. This included the establishment of a formal collaborative arrangement in 1994, termed the Casemix Area Network. However, in 1995 a report on the Areas’ health funding strategies by Eagar and Hindle (1995),

103 ______commissioned by NSW Health, came up with recommendations that were endorsed by the Department for the most part.

The first and most important recommendation was to do with all Areas adapting and using a common model of casemix–based funding. Secondly, the tertiary increment payments were to cease, because they simply encouraged inertia in provision of services. Instead, all funding was to be provided to AHSs by the Resource Allocation Formula (RAF), and equally the payments for cross–boundary flows were to be based on bilateral contracts that employed a standard set of casemix classifications (Eagar and Hindle, 1995).

In an Economic Statement for Health published in 1996, the NSW Government expressed its commitment to the recommended changes (NSW Health, 1996) which envisaged a staged implementation beginning in 1997. However, this deadline was not met, in part due to the significant public debate over the proposals to change the location of St Vincent’s hospital. According to some critics, the developments were disrupted by the Carr government’s electoral promise to halve waiting lists and the consequent waste of energy and money on ensuring that the numbers were in fact halved within the promised period, after which they quickly returned to their previous levels (Hindle, 2002).

The Economic Statement for Health re–emphasised the principles of needs–based funding. It re–committed the government to full implementation whereby Areas would receive virtually all of the available funds and begin to act as purchasers across boundaries. As a result, most Areas increased the degree to which casemix data were applied to resource allocation. The central office of NSW Health increased its involvement in providing support to Areas through the undertaking of costing studies, the appraisal of optional ways of measuring and funding, and refinement of accounting and patient care activity reporting systems (NSW Health, 1996).

In 1998, a review of the State’s health care system was conducted by the Independent Pricing and Regulatory Tribunal; it was satisfied with operations (IPART, 1998). However,

104 ______IPART made several important recommendations for change. The first was that there should be a progressive transition toward ‘live’ Resource Distribution Formula (RDF) based budget allocations. In other words, this transition should be associated with the establishment of a system of “dollars–following–patients” whereby an Area purchases the most appropriate services to meet the health needs of its population.

The second recommendation was that the Areas should make greater use of output–based (case–mix) funding models. IPART argued that the use of case–mix–based benchmarking would define which hospitals have the least cost for a procedure and consequently, the clinical pathways could then be replicated in other hospitals to reduce total costs. The third was that the Areas should be encouraged to enter into service agreements (“quasi– contracts”) with each of their major facilities to cover funding, outputs, and minimum standards of quality. The final recommendation was that a transparent capital charge should be established (such as 8% net assets) within the RDF funding allocations. Apparently, this would help reduce the risk of failing to recognise opportunity costs (IPART, 1998).

The IPART review was one of the main reasons the government decided to conduct its own detailed study of the NSW health care system. Therefore, in 1999 a body established by the government and named the NSW Health Council was assigned to undertake the study. The NSW Health Council made wide–ranging recommendations, most of which were accepted by the government and incorporated into the NSW Government’s Action Plan for Health. The conclusion drawn by the NSW Health Council was that the State government health care sector was under increasing financial pressure because of the following: 1) increased care needs and rising input costs; 2) uncertainties in budget allocations restricting care providers’ ability to undertake strategic changes; 3) the need to promote more continuity of care; 4) the need for incentives to encourage innovation; and finally 5) the need for clinicians and consumers to be more involved and informed (NSW Health Council, 2000).

The Government responded by promising to increase the health budget from $6.9 billion to almost $8.1 billion over three years. There was also a definite commitment to the full implementation of needs–based funding to Areas, including an announcement by the

105 ______government that there would be more structure in the way that Areas allocated their funds. In particular, an output–based funding model, known as the ‘episode funding model’, would be established.

In the year 2000/01 there was an overhaul of the way in which Area Health Services (AHSs) were funded and how they re–distributed those funds to the health facilities. According to the Action Plan for Health (2000), one of the requirements was that AHSs introduce and make use of an output–based funding model (episode funding) in order to distribute resources allocated to them through a largely unchanged Resource Distribution Formula (RDF). Each AHS would continue to be funded on a needs basis, while the AHSs would be required to implement a set of output–based resource allocation methods at the local level. General performance requirements were set by NSW. However, each AHS was to be empowered with considerable flexibility to enable it to take into account local factors during the funds allocation (NSW Health Council, 2000).

The introduction of these State–wide policies on output–based funding was associated with increased funding, known as enhancement funding in the policy papers (NSW Health, 1996, NSW Health., 2003a). NSW Health saw this as an opportunity to use episode funding as a way of “...ensuring value in the use of enhancement funding” (NSW Health, 2005). The work of NSW Health was to simply negotiate global activity targets with each AHS, which would often continue to be responsible for most of the decisions regarding the ways in which the targets would be met. In particular, the model would not directly affect the AHSs’ responsibilities for allocating resources between programs. Each AHS would continue to decide how much to spend on, for example, hospitals relative to community health. According to NSW Health (2000), each Area would need to take into account many factors when deciding the allocations between programs, including its strategic plans for changing the shape of service delivery. At the hospital level, the overall aim was to facilitate resource allocation between hospitals, and not to inform the processes of resource allocation between clinical services within a hospital. The latter task was a matter for AHSs and their hospitals to resolve (NSW Health, 2000).

106 ______Even with all the changes, there was no significant increase in reporting requirements since NSW Health believed that it already had adequate access to the information that it needed on patient care activities and finance. Nonetheless, NSW Health recognised that AHSs might need more information but chose to leave the matter for the AHSs to determine. One important factor with respect to the episode funding model was its restriction to only acute inpatients and to large and mid–size hospitals (NSW Health, 2002, NSW Health., 2003a). Other hospitals, predominantly in rural AHSs, were not subject to episode funding. NSW Health defined the scope of episode funding as admissions to hospitals that are within the programs named same day acute inpatient services and overnight inpatient services (Hindle, 2002). Other service types might involve output–based funding at the discretion of the Areas, but NSW Health gave no instructions on how this should be done.

The New South Wales Department of Health made a decision in 2008 to fully implement the Episode Funding Policy as the main resource allocation mechanism as of 2009, despite the casemix–based models presenting technical difficulties and the lack of direct encouragement to promote equity of access to services (as noted earlier); and despite critics of episode–based funding arguing that this type of funding might compromise quality by encouraging hospitals to be overly cost conscious, leading them to ‘cut corners’ or ‘shift costs’ onto other contributors of healthcare providers, such as GPs and specialist medical practitioners (Street et al., 2007). The objective of the current NSW Episode Funding Policy is to derive budgets that are linked to expected outputs, such as episodes of care, to create an explicit relationship between funds allocated and services provided and encourage stronger focus on outputs, outcomes and quality, including encouraging clinicians and managers to identify variations in costs and practices so that these can be managed to improve efficiency and effectiveness (NSW Health, 2008). However, the opponents of this type of mechanism warn that, to guard against such behaviour, activity–based funding has to be supplemented by additional regulatory mechanisms such as activity ceilings, marginal pricing, data auditing, and monitoring of care process, including measurement of patient satisfaction and health outcomes (Street et al., 2007).

107 ______According to NSW Health episode funding policy (2008), the proposed performance measures for monitoring implementation of the funding mechanism are as follows: access (wait list of percentage of elective surgery patients, emergency access, throughput according to weighted acute separation volume); quality (percentage readmissions, inter– hospital transfer, avoidable admissions, standardised hospital mortality ratios); efficiency (percentage target surgery, cost per weighted separation, performance to budget, relative stay in hospital); and finally general (percentage of acute services funded as episode funding and cost–shifting). Clearly equity is not a consideration in this type of mechanism. The next section discusses the population and needs–based resource allocation approaches aimed at promoting equity.

4.3 The NSW Resource Allocation Formula New South Wales (NSW) was the first state in Australia to construct a population and needs–based resource allocation formula modelled on the RAWP approach. Prior to the introduction of the RAF, the NSW Health Department distributed funds to hospitals and other health services largely on an input–based approach where historical cost data were used to estimate levels of resources needed to provide services. This was found to be less equity–oriented as it ignored changes in demographic trends, including the rapid growth of the population in the western suburbs of Sydney and the declining trends in some of the older suburbs where most of the hospitals were located. Due to growing concerns about the input–based approach, NSW Health begun to consider possible funding allocation improvements (NSW Health, 1988). The most important reason was a realisation that needs for health care had been changing as a consequence of demographic trends in some parts of the State of NSW. Some retirement areas such as the Central Coast with rapid population growth were also facing similar problems due to the elderly being high users of hospital services (Hindle, 2002, Eager et al., 2001).

Therefore, around 1980, an informal interest group began exploring issues of access and equity in NSW. As a result, NSW Health decided in principle to change the input–based resource distribution mechanism to a form of needs–based funding and this led to the

108 ______establishment of a formal committee on resource allocation in 1987. By 1989, the first NSW Health Resource Allocation Formula (RAF) was published (NSW Health, 1990, Gilbert and Bennett, 1991, Eager et al., 2001). The needs–based resource allocation formula was based in part on the Resource Allocation Working Party (RAWP) model which was introduced in England in 1976 (Department of Health, 1976). Like the RAWP model, the equity principle underpinning the RAF was equality of access to health services (NSW Health, 1993).

The Resource Allocation Formula (RAF), just like the RAWP9 model, estimated the needs for health care in the population living in defined geographical areas (Eager et al., 2001). The main idea behind the reform was that parts of the State (Areas) would be defined on a predominantly geographical basis, taking into consideration other factors such as actual distribution of health care facilities.

Therefore, the first version of the needs–based funding formula involved the creation of service Areas. They have been adjusted several times to twenty three, then seventeen and currently there are eight Area Health Services (NSW Health, 1990). The changes in numbers and sizes of the Areas was seen as an important shift in the way the NSW health system allocated the health funds, with equity of access to hospital care becoming a key objective since RAF was to distribute funds to hospitals. This led the NSW Health Department to explore the possibility of overhauling the funding allocation system to ensure that equity in access to services was promoted (Gilbert and Bennett, 1991, Eager et al., 2001). The first version of RAF, activated in 1988, “was never intended to be the means of allocating funds, but rather to establish the goalposts for future funding shares”. It was designed to “guide rather than determine resource allocation” (Stoelwinder and Viney, 2000).

9 The resource allocation working party health needs model utilises factors such as standardised premature mortality ratio, morbidity (nervous system, circulatory system, musculoskeletal morbidity), education, income, low birth weight, standardised birth ratio and aged living alone (Department of Health, 2005).

109 ______The RAF took into consideration the total resident population in each Area Health Service and used it as a measure of ‘need’ with adjustments for differences in attributes that predicted per capita variations in need, such as standardised mortality ratio (SMR), age and sex structure, fertility weighting, nursing–home–type workload, usage of private hospitals and net interstate flows. Therefore, the aim of RAF was to estimate funding targets for Area Health Services by matching the projected needs of health services for people over a 10– year planning period. It was seen by the NSW Health as a key strategic policy initiative with a major impact on the operational budgets of the Areas (NSW Health, 1993).

Initially the NSW Resource Allocation Formula (RAF) set resource target shares for a 10 year period to be used as a guide in the setting of the Areas’ recurrent budget forward estimates (Gilbert and Bennett, 1991). The main focus was on how to forecast the needs for hospital services with other elements of the health care system largely ignored. According to Gilbert et al (1991), the formula was meant to redistribute funds over a 10–year period by 1) targeting areas of high population growth for new capital works; 2) relocating services to areas of greater need when such opportunities arose; and 3) targeting under– resourced Areas/Regions in the allocation of growth funds. However, recurrent funding was not to flow in large portions until the Areas had an effective and operational infrastructure in place.

NSW Health, therefore agreed that the model should not be immediately implemented in its entirety due to predicted high levels of investment in infrastructure (Gilbert and Bennett, 1991). The obvious problem was the concentration of teaching hospitals in central Sydney; they could not be closed suddenly or reduced significantly in size and equivalent hospital infrastructure created in growing areas without major losses in efficiency and effectiveness in health services. It was, therefore, agreed by the NSW Health officials that the RAF would be implemented progressively over several years, as opportunities arose to do so without loss of performance of the health care system as a whole. The transitional period was predicted to be ten years (NSW Health, 1993, NSW Health, 1996).

110 ______According to Hindle (2002), some people believed that the Resource Allocation Formula (RAF) was a sufficient resource allocation tool without a need to supplement the model with the casemix –based funding mechanism which, at the time, had generated interest in other parts of Australia. This inconsistent view was not formally acknowledged until 1992 when NSW Health published two policy papers that stated firstly, that casemix data should be used to facilitate improvements in the RAF and secondly, an output–based funding approach should be dominant for the purpose of allocation of budgets to health care facilities within each Area Health Service (NSW Health 1992).

Consequently, the RAF approach was refined each year both in terms of data and the computational formulas, including several important changes made as a result of a policy statement titled the New South Wales Government’s Economic Statement for Health (NSW Health, 1995). The main refinements to the RAF approach following the policy statement included renaming the model to call it the Resource Distribution Formula (RDF); incorporating measures of actual service provision (production) measured largely by Diagnostic–Related Groups (DRG); and splitting the model into two main parts, one relating to tertiary services and the other to primary and secondary. The tertiary, primary and secondary health services’ funding was to be distributed based on needs as in previous versions. However, funding for tertiary services would take partial account of actual service use as defined by DRG data. As noted above, tertiary services are restricted to a subset of the hospitals that are classified as ‘tertiary’. But, in the context of the RAF not all patients in tertiary hospitals are ‘tertiary’, and so a subset of the more complicated casemix (based on DRGs) is used for identification with the aim of estimating the additional costs incurred by the areas in Sydney so as to pay them additional amounts before splitting the funds among the Areas. The areas in Sydney were also the main net importers of patients through cross–boundary flows, and this included some additional costs associated with higher severity patients and teaching functions (NSW Health, 1996).

The RAF has undergone several modifications since it was introduced in 1989/90. The first revision occurred in 1993 (see Figure 4–2, the modified Resource Allocation Formula) following the creation of 23 District Health Services to replace the existing six rural

111 ______regions. Subsequent to the election of the NSW Labour Party into office in 1995, the name of the model was changed from Resource Allocation Formula (RAF) to Resource Distribution Formula (RDF) to emphasise the intention of the new government to specifically shift resources across the state in line with the formula estimates (NSW Health, 1996). Figure 4.2 shows the NSW Resource Allocation Formula (RAF), 1993 Revision.

The renaming of the model came with a change in the funding period with the RDF being used to determine annual funding or budgetary allocations, a departure from the long–term (10 years) funding targets provided by the RAF. The annual redistribution was designed to proceed gradually through the use of the RDF, as occurred within the NHS with the use of the RAWP formula. However, unlike the RAF, which was designed to provide long–term targets, the enhanced RDF was used primarily to determine the annual budgets of Areas. This required the inclusion of as many expenditure areas as possible in the pool of funds distributed by the formula (NSW Health, 1996).

112 ______Table 4–2: The New South Wales Resource Allocation Formula (RAF), 1993 Revision

R = a(SEPi – 0.8 Pr + 0 + H – I + N) + b(SEPi) + cT

Where:

a = Inpatient conversion factor (dollars per DRG cost weight)

b = Non–inpatient conversion factor

c = Teaching conversion factor

R = Projected Resource Requirement of Area/Region

Pi = Population for the ith age/sex weighted for resource consumption

(using DRG cost weights) of Area/Region

S = Standardised Mortality Ratio of Area/Region

Pr = Private Hospital activity (DRG cost weights)

O = Obstetric activity (DRG cost weights) less private hospital births

H = High flow (tertiary) admissions (DRG cost weights)

I = Net interstate activity (DRG cost weights)

N = Nursing Home Type Patient in Acute Hospital activity (DRG cost weight)

T = Tertiary Increment (dollars)

Source: NSW Health, RAF Technical Paper 1993, Revision

4.4 Resource Distribution Formula (RDF) The Resource Distribution Formula (RDF) replaced the Resource Allocation Formula (RAF) in 1995, with the new formula incorporating a number of changes to the method of determining funding allocations to the Area Health Services (AHSs) (NSW Health, 1996). Overall, the RDF was designed to provide funding to AHSs on the principle that the Areas receive a fair and equitable share of resources that reflect the health needs of their populations and enable them to provide their local communities with comparable levels of access to health services (NSW Health, 2005, NSW Health, 2008) to improve the populations’ health outcomes. NSW Health acknowledges that the ‘health gains’ achieved over the past several decade have not been equally shared across the whole population and

113 ______therefore, a ‘health gap’ exists between individuals with the best and worst health in NSW. According to the Department of Health, this has profound implications for the health outcomes of some of the most vulnerable and disadvantaged population groups in the community (NSW Health, 2004). What is not clear from the NSW policy documents is how local communities would be provided with equal access to health services when large hospitals are highly populated in metropolitan AHSs, while the rural and remote AHSs have fewer equivalent health facilities?

According to NSW Health (2005), the main function of the RDF is to guide the allocation of resources from the NSW Department of Health to the geographically defined Area Health Services (AHSs) in order to achieve equity in funding distribution across AHSs and populations. The RDF takes into consideration the size of the local population and its health needs, unavoidable cost differences, the net receipts and expenditures of the AHSs for the financial year, patient flows seeking services outside their AHS, age and sex weights and private sector and state–wide services (NSW Health, 2005). The major difference between the RAF and the RDF is the shift from a “hospital based” resource allocation to a “wider service based” allocation where, unlike the RAF, the RDF includes health programs/services such as population health, primary & community health services, oral health services and teaching & research. Table 4–3 outlines the principles used to guide the development of the RDF.

The Resource Distribution Formula is a mechanism through which these principles are achieved and, as distinct from the RAF which provided budget share allocations towards a 10–year target, the RDF is used to determine annual equitable shares of available resources for Area Health Services to enable them to provide a comprehensive range of health services to the local population (NSW Health, 2005). This necessitates the inclusion of as many expenditure areas as possible, with the exception of special purpose and trust funds, capital funds and special projects that are funded separately (see Figure 4–1). Funding for capital expenditures, grants for special purpose, trust funds, and state or commonwealth grants to specific projects, especially where the projects are unique to a particular Area

114 ______Health Service (AHS), are not included in the pool of funds distributed by the RDF (NSW Health, 2007a, NSW Health, 2005).

Table 4–3: Principles Guiding the Structure of Resource Distribution Formula (RDF)  To guide the allocation of available resources to Area Health Services to enable Areas to provide their local communities with comparable levels of access to health services taking into account: The assessed health needs of the local population. Flows of patients between Area Health Services and to Children’s Hospital, Westmead. The local population’s utilisation of private health services. Additional cost components in providing services to specific populations (such as additional transport and infrastructure costs in rural areas and the cost of interpreter services).  To recognise funding of state–wide and selected specialty services that benefit the entire health system and are provided in limited locations.  To take account of additional costs associated with severity not currently recognised in casemix measures faced by major tertiary referral hospitals and specialist paediatric hospitals, to reflect the higher needs of these patients.  To reflect the need for Area Health Services to provide additional health services to improve the health status of indigenous and homeless population groups who experience significantly lower health status.  To assume that Area Health Services achieve comparable levels of efficiency in the provision of services.  To reflect the strategic directions set for Area Health Services in NSW and the NSW Department of Health. Source: NSW Health, RDF Technical Paper, 2005 Revision

115 ______Figure 4–2: Steps in Calculating Resource Distribution Formula Funding Targets

Source: NSW Health, RDF Technical Paper, 2005 Revision

The funds that are distributed through the use of the RDF are first pooled together as shown in Figure 4–3 for the purposes of calculating the RDF target shares. This is followed by the selection of programs/services that should be funded mainly according to RDF. The funds that are not included in the category of RDF are excluded10 at this stage; in particular, funds that are associated with cross–boundary flows between Areas are removed.

10 Other exclusions include fund flows for eligible veterans and their dependants, since under existing contractual arrangements with the Department of Veterans Affairs, each Area Health Service receives funding according to an agreed price for each eligible Veteran treated. Flows for overseas residents, ineligible patients, and interstate patients are excluded due to other financial arrangements for these patients. Flow of funds to contracted private hospitals (such as Port Macquarie and Hawkesbury) are excluded including flows to the Mental Health Program and other programs other than acute inpatient, emergency, outpatient and rehabilitation and extended care (NSW Health, 2005).

116 ______In principle, each Area Health Service (AHS) allocation would reflect the needs of its resident population as determined by the Health Need Index (see Table 4–4), and payments would be made by the specific Area when any of the residents received care outside of their residential AHS. However, in practice, NSW Health removes some of the funds before allocations are made, and subsequently allocates them among the AHSs that have net inflows of patients from other Areas (Eager et al., 2001). This issue has been debated extensively by NSW Health and the AHSs, but no consensus has yet been reached as to how best to handle the cross–border flows and by whom. According to NSW Health a specific formula is required to guide the cross–border flows (NSW Health, 2005).

Table 4–4: Resource Distribution Formula (RDF), 2005 Revision 2000 Health Need Index = α + β (SMR < 70) + δ (EDOCC) + γ ARIA + ε %ATSI

Where: α = Constant SMR<70 = The Standardised Mortality Ratio for ages less than 70 EDOCC = The Index of Education and Occupation11 ARIA = Measure of Accessibility/Remoteness12 %ATSI = The proportion of the population which is Aboriginal or Torres Strait Islander Final model equation: 2000 HNI = 95.31 + 0.3 (SMR<70) – 0.3 (EDOCC) + 4.0 ARIA + 1.0 %ATSI

Information Source: NSW Health, RDF Technical Paper, 2005

11 The weighting of the formula with the Index of Education and Occupation, an index developed by the Australian Bureau of Statistics, was so as to incorporate factors known to influence or reflect population morbidity due to lack of systems for generating direct measures of population morbidity as observed by NSW Health (2005).

12 The use of the Accessibility/Remoteness Index of Australia (ARIA), based purely on geographic factors, is derived by scoring populated localities on the basis of the road distance from centres of population greater than 5,000 persons to four categories of designated “service centres”. The service categories are based on resident populations (i.e. 1) 250,000 or more; 2) 48,000 to 249,999; 3) 18,000 to 47,999; and 4) 5,000 to 17,999). A score is calculated for each locality and category as a ratio of the road distance to the mean road distance for each category, truncated at 3.0. Therefore, the scores for each of the four categories of service centre are combined to given an ARIA index ranging from 0 in the inner metropolitan Area Health Services (AHSs) to 12 in the most remote AHSs. The index was developed by the National Key Centre for Social Applications of Geographical Information Systems through the ARIA project which was funded by the Federal Department of Health and Aged Care. Following the ARIA project, the Australian Government proposed that the ARIA index be adopted as a national standard for measuring remoteness and since then it has been included in the NSW 2000 Need Index (NSW Health, 2005).

117 ______The Resource Distribution Formula (RDF) is the principle mechanism used by NSW Health for assessing the relative resourcing requirements of each of the eight Area Health Services (AHSs). The RDF is applied to all NSW Health programs other than the Mental Health program. Mental health program funding is earmarked by NSW Health so AHSs cannot utilise the funds for other purposes outside mental health services. It is not clear why it is not included in the RDF pool. According to NSW Health, approximately 81.7 percent of funding derived by the RDF pool of funds is influenced by the assessment of relative needs through the use of the Health Need Index (HNI). The other 18.3 percent is impacted by a range of other factors such as state–wide services and teaching and research (NSW Health, 2005).

The assessment of the relative needs of the AHSs involves taking into consideration the impact of demographic characteristics of Area populations, specifically age and sex composition. Age and sex utilisation weights (adjusted where possible for the cost of services) are calculated based on the best source of data and applied to AHS populations to determine the impact of need. The Health Need Indices are applied to capture the influence of other legitimate factors, such as features of the resident populations (index of education and occupation as shown in Table 4–3) which impact on the need for health services (i.e. Acute Admitted Care, Maternity Services, Rehabilitation, , Emergency Departments, Outpatients and Community Health) (NSW Health, 2005).

According to NSW Health (2005), the construction of the Health Need Index follows a research approach adopted for the English National Health Services (Royston et al., 1992, Carr–Hill et al., 1994), an approach which is also used in a number of other countries as discussed earlier (Rice et al., 2000). In NSW, the HNI seeks to explain variation in hospital utilisation between geographic areas and the population characteristics that can be directly measured, that is age and sex composition. Thus NSW uses the Australian Refined Diagnosis Related Groups (AR–DRGs), the standard casemix classification for grouping admissions to hospital. Each acute care hospital admission is grouped according to an AR– DRG based on the reported diagnoses and procedures. In NSW, a standard approach to assign a relative weight to each admission is adopted (NSW Health, 2002, NSW Health, 2008). This approach is based on reported average costs for each AR–DRG, although other

118 ______adjustments are also applied. The relative weight assigned reflects the relative costliness of the admission. However it does not reflect variations between hospitals in their costs, which may reflect efficiency variations, or other systematic variations in costs (NSW Health, 2001).

However, the use of hospital utilisation as the basis for capitation payments, the amount of health service funds associated with a citizen for a particular time period aiming to effectively place a health care ‘price’ on the head of every citizen, has been criticised, since the expected health care expenditure needs of individuals vary considerably, depending on personal characteristics such as age, morbidity and social circumstances. Firstly, current utilisation might, to some extent, reflect systematic variations in supply, implying that existing inequities might be perpetuated if no adjustment were made for such variations. Secondly, uncritical use of current utilisation as the basis for setting capitation payments might introduce a perverse incentive for Area Health Services to increase current utilisation in order to attract higher capitation payments for their population in the future.

The approaches of using utilisation data, such as episode/casemix–based funding, are intrinsically conservative, in the sense that they assume that, on average, the health system is currently meeting the desirable concept of need, whatever that concept might be (e.g. capacity to benefit, level of sickness, life expectancy and so on). The methods, therefore, fail to reflect ‘legitimate’ health care needs that are not currently met by the system (Culyer, 1995, Smith et al., 2005). ‘Unmet needs’ merely seek to indicate that certain groups of the population systematically fail to receive the health care that policymakers intend and are, therefore, not captured in the utilisation data. Smith et al (2005) argued that, the use of empirical utilisation data as the basis of capitation payments is, therefore, inappropriate as it perpetuates the inequity implied by the existence of unmet needs (Smith et al., 2005).

Accordingly, the modelling for the NSW Health 2000 Health Need Index used step –wise regression to test which variables, out of a broad list of possible factors, best explained variation in hospital use. As illustrated by Table 4–4, four variables were deemed important – Standardised Mortality Ratios (SMRs) for people aged under 70 years; the proportion of

119 ______the population that identified as Indigenous; a measure of rurality (the ARIA index); and a measure of socioeconomic status (the Index of Education and Occupation, developed by the Australian Bureau of Statistics, ABS) (NSW Health, 2005).

The general RDF used the 1998 population projections derived from the census data for planning purposes. The Mental Health Services RDF utilises a blended needs index, with a discounted weight of 0.6 for the proportion of the population never married. This is based on the understanding that mental illness is rarely a direct cause of death and the SMR was an inappropriate measure of the need for mental health services. As Jarman (1983) observed, an unmarried male was six times more likely to be admitted because of mental illness than a married male of the same age (Jarman, 1983).

The index of education and occupation (EDOCC) incorporated into the Relative Need Index is lower, reduced from the level of 0.5 in the previous RAF to 0.4 in the RDF. Additional weighting for the Aboriginal population (the Aboriginality factor) in the primary and community–based service programs is given a higher priority to ensure more funds to areas with groups such as Aboriginals and Torres Straits Islanders (ATSI), people with poor health status (NSW Health, 2005). Other factor taken into consideration was the patient flows to private hospitals. Therefore, patient flows to private hospitals factor was discounted by 70 percent of the total cost weighted separations. For the first time, direct and indirect teaching and research costs were identified separately in the RDF. Direct costs include reported expenditure in special purpose and trust funds, expenditure funded through local revenue, and department–funded expenditure such as overheads. Indirect costs refer to additional costs incurred by teaching and tertiary referral hospitals due to the more severely ill patients treated by these facilities. Both direct and indirect costs were determined by comparing the cost differences between teaching and non–teaching hospitals (NSW Health, 1999, NSW Health, 2005).

Like the original formula, the RDF has undergone several revisions since its inception but the nine key components have remained unchanged. Table 4–5 outlines the nine components and their alignment to the ten programs under which funding is reported in the NSW state budget. Thus, resources in the pool of funds that will be subject to the RDF are

120 ______divided into program expenditure categories based on the expenditure reported by areas across the NSW health programs. For example, the Acute Inpatient Services component of the RDF covers expenditure program 2.2 and 2.3 (Overnight and Same Day Acute Inpatient Services) (see Table 4–5). The share of expenditure of each component is calculated by weighting the population by indicators of health service utilisation and health needs (NSW Health, 1996, Hindle, 2002).

Table 4–5: Components of the Resource Distribution Formula and Program Alignment RDF Components Health Programs Population Health 5.1 Population Health Services Oral Health Services 1.1 Primary & Community Based Services Primary & Community Based Services 1.1 Primary & Community Based Services Outpatients 1.3 Outpatient Services Emergency Department Services 2.1 Emergency Services Acute Inpatient 2.2 Overnight Acute Inpatient Services 2.3 Same Day Acute Inpatient Services Mental Health Services 3.1 Mental Health Services Rehabilitation and Extended Care 4.1 Rehabilitation and Extended Care Services Teaching and Research 6.1 Teaching and Research Source: NSW Health Resource Distribution Formula Technical Paper, 2005

In 2005, the RDF was revised but the share of expenditure of each of the nine components is still calculated by weighting the population by indicators of health service utilisation and health needs. The key indicators used currently include age and sex structure; this is a general needs index developed from an analysis of how the standardised mortality ratio, rural–urban variations and socio–economic status explain the use of health services. A separate index of need was developed from mental health and rehabilitation and extended care. Other key indicators include a factor for utilisation of the private health services that can be substituted for public health services by the local population; a factor for a range of other unavoidable cost components such as additional transport costs in rural areas; and a factor for Aboriginality and homelessness (NSW Health, 2005). According to NSW Health, taking these factors into consideration ensures that each AHS receives a share of the

121 ______available health care budget in proportion to the relative needs for health care of the population of its Area (NSW Health, 2006c). The entities excluded from the scope of the RDF funding model over the years still remain the same. According to NSW Health (2005), the main exclusions are the New Children’s Hospital, Corrections Health, the Ambulance Services, the Blood Transfusion Service, and the Veteran’s13 population.

According to NSW Health, the RDF is continually reviewed on an annual basis by an RDF Advisory Committee (although currently there are no yearly reports published to confirm the updates – the last report to be published was the ‘RDF Technical Paper’ written in 2001 but published in 2005) (NSW Health, 2005). The Committee consists of members with clinical, health administration and academic backgrounds. The meetings occur annually and they are normally chaired by the NSW Health Director of Inter–Government and Funding Strategies Branch and, through wide consultations, they make recommendations for change on the RDF. But what has been the impact of the RDF in resource distribution within the NSW health system?

4.4.1.1 Effects of the Resource Distribution Formula (RDF) Even though the Resource Distribution Formula (RDF) has not been critically evaluated and debated extensively like the RAWP formula, it has nonetheless been hailed as a model that attempts a fair distribution of funds available to the NSW health care system (Gibbs et al., 2002, Gilbert et al., 1992). Prior to the development and adoption of the RAF/RDF, health services in NSW remained in historical settings, reflecting the population distribution of the past and previous year’s expenditure (Eager et al. 2001). Given the changes in population dynamics, there was a need to reform the methods of allocating resources, including closure and modification of some services in order to fund the development of others. Although the government faced opposition from entrenched lobby groups determined to preserve the status quo, they were able to introduce some reforms.

13 According to NSW Health (2005), for the purposes of the RDF, the population, needs, services and funding associated with eligible gold card veterans and their dependants are not considered, as these are dealt with comprehensively by the Department of Veterans’ Affairs. Services provided by the public sector are provided under a general contract NSW Health has entered into with the Department of Veterans’ Affairs.

122 ______The introduction of a Resource Allocation Formula (RAF), currently known as the Resource Distribution Formula (RDF), made it possible to quantify the maldistribution and was used by NSW health authorities to argue for a redistribution of resources in a rational and non–politicised manner.

There are few reports of progress in reducing funding disparities across NSW following the introduction of the RDF. Gibbs et al. (2002) and more recently NSW Health (2005) noted that in 1989/90, approximately 16.4 percent of the health budget needed to be reallocated to achieve equity in funding. However, by 1994/95, this figure was reduced to 9.6 percent and was down to 4.4 percent by 1998/99 and currently its 3%. The reduction is attributed to the use of the RDF. The authors (who were also involved in the development of the funding mechanism) further argued that while all Area Health Services in NSW have received growth in funding, a greater share has been allocated towards historically under–funded population growth Area Health Services such as those in Greater Western Sydney, the Central Coast, and the North Coast of NSW (Gibbs et al., 2002, NSW Health, 2005). However, for almost twenty years since the first Resource Allocation Formula (RAF) was introduced, no external review has been conducted to corroborate the assertion that the RDF has indeed achieved equity in the distribution of resources in NSW Health and hence this is the aim of this thesis.

4.5 Beyond Needs–Based Funding in NSW: Introducing the Research Questions Population and needs–based funding (or risk adjusted capitation approaches as they are often referred to) has been the most common approach to promoting equity and facilitating the shift from historical funding models which have been less concerned with addressing inequities. As discussed earlier, the approach was first developed by the Resource Allocation Working Party (RAWP) in England in 1976 (Mays and Bevan, 1986, Resource Allocation Working Party, 1976) and has since inspired health sector resource allocation reforms in several countries around the world. Among other objectives, the RAWP formula was meant to ensure that the allocation of resources was commensurate with the health needs of the population in order to improve equity by increasing the opportunity of access

123 ______to services through effective allocation of the National Health Service (NHS) budget between geographical areas (Carr–Hill et al., 1994, Smith et al., 2001, Smith et al., 1994). It defined equity in terms of equal opportunity of access to services for patients in equal need regardless of where they live (Shaw and Smith, 2001, Department of Health and Social Security, 1976). Since the RAWP risk adjusted capitation model was published in 1976, several countries, including New South Wales (Australia), various Canadian provinces, New Zealand, Sweden, Italy, Scotland, Northern Ireland and South Africa, have adopted similar a methodology to develop their own needs–based approaches to resource distribution to improve equitable funding among geographical areas and population groups.

Despite their prominence and wider appeal, needs–based mechanisms are usually complex and their design can involve a difficult process, using time and resources. In New South Wales, just like many other countries, needs-based formulae consist of population needs alongside more recently developed market mechanisms of health care financing and resource allocation (Rice and Smith, 2001b). The size of the population in a geographical area often serves as the primary indicator of a need for health services. The population size is usually weighted by a range of other indicators of relative need for health care such as standardised mortality ratios, age and sex composition, the communities’ capacity to benefit from health services, their ability to pay for health care costs, and their level of dependence on public sector health services (Eckstein and Gibberd, 1994, Gilson and McIntyre, 2005, Mooney and Houston, 2004, Sheldon et al., 1994, van Doorslaer et al., 2006, NSW Health, 2005). The effectiveness of a needs–based formula in reducing inequalities may depend on the accuracy of the predictive factors or indicators used in developing it and how they reflect the values of the community (Eyles et al., 1991, Gibson, 2002, Mooney, 2003, Mooney et al., 2002). This may explain why most countries with needs–based models continue to struggle with avoidable inequalities (health differences that are not attributable to biological or natural factors).

Recent debate in the literature about resource allocation suggests that needs–based formulae do not promote equity. This stems largely from the persistent inequities in health among different jurisdictions and population groups, even in countries where needs–based

124 ______models are in use (some countries may have developed a needs–based funding formula but the extent to which it is applied is doubtful). Similarly, needs-based formulas fail to promote equity because they are not based on measurable indicators (such as morbidity and premature mortality) that entirely reflect the actual health needs of the population. According to the NSW Health and Equity Statement – In All Fairness: Increasing equity in health across NSW (2004) policy document, ‘the gaps between the health of the most and least disadvantaged members of the NSW community are persistent and significant, and there is even concern that these gaps may be widening’ (NSW Health, 2004). Mooney and Houston (2004) have more recently argued for an alternative approach to resource allocation that embraces the concept of ‘capacity to benefit’ and the MESH infrastructure (see Chapter 3). MESH stands for Management, Economic, Social and Human infrastructure. According to Mooney and Houston (2004), the capacity to benefit from resource allocation entails good management, requires availability of resources, needs a socially well functioning community, and, ideally, good human resources. Where some or all of the infrastructure components are missing, the resources may be wasted, or at best, used to a lesser effect (Mooney and Houston, 2004). In other cases, resources might be reduced or withheld due to the lack of capacity to utilise the funds (Asante et al., 2006).

The philosophy underpinning the RAWP–type approaches has been criticised, especially by Mooney and Houston (2004) who argued that the RAWP–type resource allocation formulas emphasise the size of the problem in terms of greater need rather than the ‘capacity to benefit’ from the processes of resource allocation. This is because different areas or jurisdictions may have a different capacity to manage resources well; if the MESH infrastructure is not included in the resource allocation formula, jurisdictions with low capacity, especially those in rural and remote parts, might be inadequately resourced (Mooney and Houston, 2004). Indeed, even some disadvantaged population groups may lack the capacity to utilise health services due to a lack of knowledge regarding their condition and the need to access services. As evidence shows, those people and communities with the poorest health often have poorest access to health services and make the least use of health services (Birch et al., 2007, NSW Health, 2006). This can lead to fewer resources being allocated to Areas since most needs–based formulas use service

125 ______utilisation data to derive health needs and allocate resources accordingly. Thus NSW Health acknowledges that if equity in health is to be realised in NSW, the system must recognise that not all Area Health Services or population groups have the same capacity to deal with their health problems; hence it is important to address different Areas and people’s needs in a different way through a targeted approach (NSW Health, 2004).

As a result, the indicators applied in the development of needs–based formulas would be deemed inadequate in terms of estimating the health needs of the jurisdictions and population groups if they do not take into account the capacity to benefit from resources. Similarly, needs–based formulas by themselves are not sufficient to address equity. Gibbs et al. (2002) noted the New South Wales RDF is only one policy for addressing the equity issue, and by itself is an insufficient mechanism. Gibbs et al. (2002) admitted that while the RDF aims to create broad resource capacity for equity to be achieved within the health system, an essential ingredient in delivering on equity objectives is action at the local level within the health system. In short, the obsession with funding formula, or “formula fever” as Trevor Sheldon (1997) puts it, has distracted attention from the important issues of the indicators applied, the capacity of areas and population groups to benefit and most of all how allocated resources are spent at the population level. As Sheldon has suggested, health authorities and general practitioners should focus their attention on whether current spending patterns reinforce socially produced health inequalities and, if so, do something about it from the local level (Sheldon, 1997). Clearly, NSW Health and other countries using needs–based formulas and yet still burdened by avoidable inequities in health, especially in health outcomes, need to take a second look at the mechanisms used in distributing resources and indeed the way resources are used to meet the needs of the worst–off population groups.

Equity in health across NSW Health can be achieved through a commitment to redirect resources to the most disadvantaged population groups in an effort to reduce health inequalities. This will mean a radical departure from the criterion of seeking to offer ‘equal and comparable access’, in effect seeking to secure a redistribution of health, implying that current practices are not securing outcomes in line with policy intentions as noted in the

126 ______NSW Health policy document that says, ‘pursuing equity in health involves all efforts both within and beyond the health system aimed at improving life opportunities for people who are most disadvantaged, so that they have the best chance of achieving and maintaining good health. It implies a need for the redistribution of existing and new resources towards redressing those inequities’ (NSW Health, 2005). This criterion will steer health policy quite determinedly away from the narrow concept of health care equity towards the broader concept of health equity, with its implications for diverse policy areas involving other non– health sectors such as income redistribution, housing, education, environment, and transport as discussed in Chapter 3. The following Chapter details the methodology used in this study, including the limitations encountered.

127 ______Chapter 4: Summary

Chapter 4 has primarily covered the various resource allocation mechanisms aimed at promoting equitable distribution of health resources.

The discussion focused largely on three funding approaches: historical –based funding and population, needs–based funding models and market driven mechanisms such as episode – (activity) based resource allocation.

The RAWP formula and its influence on resource allocation reforms, especially in New South Wales, were reviewed. Resource allocation in New South Wales has been greatly influenced by the RAWP approach. Since 1989/90, NSW Health resource distribution across Area Health Services was achieved using their RAWP–inspired Resource Allocation Formula (RAF), currently known as the Resource Distribution Formula (RDF).

The principle underpinning the allocation of funds from the NSW Department of Health to Area Health Services is that AHSs are able to provide a comparable level of access to health services while taking into account the health needs of their local population. The RDF is used to guide the allocation by taking into consideration the size of the local population and its health needs, unavoidable cost differences, patient flows as well as a number of other factors.

Debate in the literature about resource allocation suggests that the RAWP–type population and needs–based formulas, which emphasise the distribution of resources on the basis of need, do not adequately promote equity on their own. As a result some countries have adopted episode– based (patient casemix), also known as activity–based mechanisms, for funding hospitals.

Episode–based funding seeks to derive budgets that are linked to expected outputs (such as episodes of care within hospitals).

In NSW, episode funding applies to public hospitals within the NSW public health system, including rural and metropolitan Area Health Services and the Children’s Hospital at Westmead.

Resource allocation within AHSs is guided by hospital and community health programs and is services based. The RDF does not guide allocation of resources at the AHSs level.

128 ______

Chapter 5: RESEARCH METHODS

Overview This chapter describes the method of inquiry adopted in this study. The chapter is divided in two sections. The first section presents the aim of the study, the research questions and a detailed explanation of how the study was designed, including a detailed step–by–step approach as to how each of the research questions was addressed. The second section describes the method of inquiry used for the collection and analysis of data and the rationale for the selection of the Area Health Services in New South Wales as the focus of the study. The final section presents the problems encountered in the field and the overall limitations of the methods.

Section 1: Aim, Research Questions and Design

5.1 Study Aim and Research Questions As stated in Chapter 1, the overall aim of this study was to examine whether there has been movement towards equity in resource allocation to Area Health Services under the NSW Health Resource Distribution Formula (RDF) and if there has, whether this has been reflected in equitable resource allocation within Area Health Services. Equity is defined in this study as equality of access to financial resources for health care on the basis of health needs. Health needs is defined in terms of population health status.

The research questions addressed by the study are as follows:

1. To what extent has resource allocation to Area Health Services from the NSW Department of Health been equitable? 2. How are the financial resources received from the State level distributed at the Area Health Service level? 3. Is equity reflected in the allocation of NSW Health resources at the Area Health Service level? 4. What prospect is there for allocation of NSW Health resources both at State and Area Health Service levels to promote equity?

129 ______5.2 Research Design Research design is a fundamental part of any study and refers to the overall research structure. According to Bowling (2002), research design is the logical progression that unites the empirical data to the research questions and, consequently, to its conclusions. It also ensures that the evidence obtained can address the research questions as explicitly as possible (De Vaus, 2001, Johnson and Onwuegbuzie, 2004). The current study was designed to use a mix of qualitative and quantitative methods. Qualitative and quantitative methods are the two main research techniques that continue to be applied in health research (Bowling, 2002). To some researchers, the two terms ‘qualitative’ and ‘quantitative’ refer to methodological approaches but to others the methods describe concepts that are essentially different, involving different views of the world. For example, qualitative research is contextual and driven by the environment of the research, whereas quantitative research is grounded in scientific method and empiricism (Berglund, 2001).

A mixed methods approach is an approach that collects, analyses and integrates quantitative and qualitative data into a single study (Morse, 2005). A quantitative research method involves ‘epistemological differences’ which claim that ‘truths’ exist, and knowledge about the truth can be acquired through experimentation and observation (Phillips and Burbules, 2000, Bowling, 2002). The approach is particularly appropriate if the issue being investigated is known about, relatively simple and unambiguous, and amenable to valid and reliable measurement. On the other hand, qualitative methods are based on social construction which emphasises that the ‘truth’ is not absolute, but rather depends on the meanings put on observations by society (Green and Thorogood, 2004); those meanings are many and diverse and the research should rely as much as possible on the participant’s views of the situation being studied (Neuman, 2000, Lincoln and Guba, 2000).

Within the mixed–methods framework, different research methods are usually employed in order to investigate complex situations fully and to validate the findings (Sidell, 1995, Creswell, 2003). A combination of both quantitative and qualitative data yield a more complete analysis and complement each other, including strengthening the reliability and confirmability of the research through corroboration and mutual assurance (Green, 1994, Lincoln and Guba, 2000). Johnson and Onwuegbuzie (2004) observed that mixed–methods

130 ______research attempts to justify the use of multiple approaches in answering research questions, rather than restricting or constraining researchers’ choices (Johnson and Onwuegbuzie, 2004). Thus, in mixed–methods, the research question is fundamental and ideally research methods should follow the research questions in a way that offers the best chance to obtain useful answers (Johnstone, 2004). The contrasting purposes of the two methods provide scope for using multiple methods in the current study, using one to supplement the other in order to check the accuracy, content, validity and relevance of the data that has been collected. However, although the two approaches are regarded as complementary, each has its own strengths and weaknesses and so the choice between the two has been the subject of intense academic debate (Bowling, 2002, Creswell, 2003).

Previous studies focusing on equity and resource allocation have often used one or other of the research methods to establish the extent of equity in resource allocation. A study based in Ghana, conducted by Asante, Zwi and Ho (2006), was the only study obtained from the literature search to have used a mixed–methods approach to investigate equity in resource allocation. The dearth of such studies worldwide, especially in developed countries, motivated the researcher to explore the research issue using a combination of both methods (qualitative and quantitative). But unlike the Asante et al study, this study was designed to elicit the perspectives of key stakeholders about the processes for resource allocation, the consideration given to equity and the factors that drive or limit the promotion of equity in resource allocation in NSW health system before undertaking the quantitative component of the study.

5.2.1 Research Model: Sequential Exploratory Model In a mixed–method structure, different models exist. During the design of this study, all potential models were considered to establish the most appropriate model and the conclusion was that the sequential mixed (exploratory) model would help achieve the study objectives. The sequential mixed model design is a multi–strand mix of qualitative and quantitative approaches in which the conclusions that are made on the basis of the results of the first strand (i.e. qualitative phase) may lead to formulation of questions, collection and analysis of data for the next strand (i.e. quantitative phase). In this model, qualitative and

131 ______quantitative approaches are used to confirm and cross–validate the findings within a single study (Greene et al., 1998, Johnson and Onwuegbuzie, 2004).

An initial consultation with some key stakeholders during the design of the study set the stage by clearly revealing that the way resource allocation and equity are dealt with lacks transparency. For example, some key stakeholders thought that equity at the state level had been achieved. Others indicated that the equity objective was only applied at the top health care system (State/central) level and that it was not effectively promoted at the lower (Area Health Service/local) levels of the health system. Thus, a decision was made to embrace the sequential exploratory model so as to continue conducting further in–depth qualitative data collection and use the quantitative results to explain the research findings of the qualitative phase. The results of both methods were then merged at the results interpretation stage because of the differences in the tools used to collect and analyse data. The qualitative data analysis involved content analysis undertaken using NVIVO software version 8, while statistical methods consisting of Principal Component Analysis (PCA) and Microsoft Excel were used to analyse the quantitative data. Figure 5–1 below encapsulates the entire study design, showing the approaches adopted to address the research questions and analyse the data.

132 ______Figure 5–1: Approach Taken to Address Research Questions

133 ______Summary of Approaches used to Address the Research Questions This section outlines the approaches used to address each of the research questions as shown in Figure 5–1. As described in the research design, a mix of qualitative and quantitative methods was employed for addressing the research questions. The questions and approaches are specified below:

Research Question 1: To what extent has resource allocation to Area Health Services from NSW Department of health been equitable? The main objective of this question was to establish whether and the degree to which resource allocation to AHSs has been equitable.

Approach: This question was addressed using both qualitative and quantitative methods. First, the opinions of policy makers at the NSW Department of Health and AHS executives on resource allocation and equity were explored using semi–structured interviews. Then, a quantitative analysis of health expenditure data and demographic, socio–economic and health care information from the Australian Bureau of Statistics and NSW Health was undertaken.

The demographic, socioeconomic and health–related data were used to develop two benchmark indices – the General Health Need Index (GHNIdx) was developed using a mixture of socioeconomic, demographic and health related variables. The second index – the Double Variable Need Index (DVIdx) – was developed using only premature mortality and morbidity data. Both indices were used to proxy the need for health care across the eight AHSs.

To determine the extent of equity of the allocation, the need indices were converted to funding shares and used to redistribute the total funding allocated to AHSs to obtain the predicted allocations. These were then compared with the actual allocations to the eight AHSs to assess whether the AHSs with the greatest health needs were allocated resources commensurate with their level of health needs. In addition to the two indices, premature mortality across Area Health Services was also analysed and used to gauge whether

134 ______resource allocation from the NSW Department of Health to AHSs had led to reduction in premature mortality rates.

Research Question 2: How are the financial resources received from the state level distributed at the Area Health Service level? The objective of this research question was to examine and document the process of resource allocation at the AHS level and then analyse the pattern of distribution between hospital and community–based services and programs.

Approach This question was addressed using both qualitative and quantitative methods. First, Area Health Service executives and other key health officials who were in positions of authority were consulted about the resource allocation procedures and processes within the area. This includes the people involved in the allocation decision –making process and the factors taken into consideration when allocating resources. Their views were documented.

After exploring the views of AHS officials on resource allocation, the reported distribution of finance between programs and services over a period of time was examined. Resources received by the AHSs were categorised according to the percentage allocated to hospital– based and community–based health services. Then, the pattern of distribution of AHS funding between these two categories was examined to to determine whether there had been any changes in the traditionally disproportionate emphasis on hospital funding.

Research Question 3: Is equity reflected in the allocation of NSW Health resources at the Area Health Service level? The objective was to establish whether the current system of allocating resources within AHSs promotes equity in terms of needy Local Government Areas (LGAs) within AHSs receiving a fair share of resources.

135 ______Approach This question was mainly addressed using quantitative methods. First, an index to proxy health needs across LGAs in two AHSs using premature mortality and morbidity data only was developed. Since resources within AHSs were not allocated according to geographic regions but by programs/services, it was not possible to compare actual and predicted allocations as was done with the inter–AHSs equity analysis. A pragmatic approach to accessing equity in the allocation was therefore devised and this involved mapping of all the health services available in each LGA and matching that against the level of health needs to determine whether equity was reflected in the system of resource allocation within areas; i.e. whether the LGAs with higher health needs have more services. The issue of cross–border flow of patients from one LGA to another was not able to be taken into account.

Research Question 4: What prospect is there for allocation of NSW Health resources both at the State and Area Health Service levels to promote equity?

The objective was to consider the extent to which alternative ways of allocating financial resources within the NSW health system could better promote equity.

Approach This question was addressed mainly through analysis of documents. The findings of the study were reviewed in the light of the historical, political and other factors that may facilitate or constrain equitable allocation of resources in New South Wales. Key documents about the development of the NSW health system and the relationship between the Commonwealth government and the States in terms of the sharing of responsibilities for health care were reviewed. Recommendations as to how best to promote equity in resource allocation at all levels of the health system were made.

136 ______Section 2: Method of Inquiry This section first describes the method and process of the qualitative component (5.3) of the study and then the quantitative component (5.4).

5.3 Qualitative Method The qualitative phase of the study was designed to explore the knowledge of and opinions regarding resource distribution processes and factors that drive resource allocation of the key players and actors involved in the decision making process. One–to–one interviews were the method used to gather qualitative data. Interviews that are open–ended and in– depth can be conducted individually or in group settings and their end products are quotations about people’s knowledge, beliefs and experiences (Green and Thorogood, 2004). This qualitative research method aims to engage people in their daily (natural) settings in order to collate information on the research topic and is usually less interfering than quantitative investigations. Both the researcher and the participant are ‘intertwined with the natural environment’ of the topic in question. Therefore, interviews with NSW Health (State level authority) and Area Health Service officials and some other key stakeholders were conducted in the participant’s natural settings

Initially six key stakeholders at both levels of the health care system were interviewed. However, the results of the data analysed indicated that the officials had mixed perspectives about the resource allocation processes, and the players and actors involved in the decision –making. For example, some thought that the mechanism (resource distribution formula) used by the NSW Department of Health to promote equity had not achieved its objective of reducing historical funding imbalances among the eight geographically defined Area Health Services. Others thought that the formula was better than the historical–based allocations that occurred before the formula was introduced in early 1990s. According to others, resource distribution is politically driven and, therefore, lacks openness.

Clearly, there was a lack of consensus among the officials regarding the key issues addressed by this study and consequently, further extensive interviews involving 24 stakeholders were conducted to ensure the data were comprehensive and in order to

137 ______enhance data validity and reliability. The decision to conduct further interviews made it difficult (since the researcher needed to travel to different study sites) to simultaneously embark on the quantitative phase of the study. Hence, the qualitative and quantitative components of the study were conducted sequentially and both methods were integrated at the results interpretation stage. The flexibility offered by the sequential mixed–method approach, especially when applied within a single study, was a very important feature of the study. This approach gave the researcher an opportunity to understand how resource allocation is dealt with, factors taken into consideration and the players and actors involved in the decision –making. The next section explains how the study sample was selected.

5.3.1 Sample Selection Australia is divided into seven states and territories (see Chapter 2). New South Wales (NSW) is Australia’s most populous state and was the first state to introduce a population and needs–based resource distribution formula to guide allocation of health funds; this was the main reason NSW was selected as the study site. NSW is divided into eight administrative Area Health Services (AHS) responsible for the planning and delivery of health services to the local population with funding provided from the Commonwealth, NSW State and Local Governments and some other sources (non–government sector). Around 27 percent of the health budget is allocated from the State treasury to the NSW Department of Health which, in turn, distributes the funds to its organisational entities known as AHSs through the use of a Resource Distribution Formula (RDF). The AHSs vary in population size, landmass area and characteristics. According to publications from NSW Health, the use RDF ensures that each AHS receives a fair share of resources based on its population and level of health needs (NSW Health, 2005).

Logistics and time constraints limited inclusion of all the eight Area Health Services (AHSs) in the qualitative part of the study. Consequently, two diverse AHSs were purposefully selected. These were the urban Northern Sydney Central Coast AHS and Greater Western AHS, classified as rural and remote. The two AHSs have distinctive features that make them suitable for comparative analysis within a case study framework and they are also representative of the other six AHSs since they highlight extremes.

138 ______Northern Sydney Central Coast (NSCC) is a wealthy urban AHS with a population density of about 1.1 million people as of 2006, representing 16.4% of the estimated NSW population (close to 7 million) and a landmass of 6,300 square kilometres. Of the NSCCAHS population, 19.1% are aged 75 years or more, an age group that generally needs considerably more health care than the general population. NSCCAHS is an urban/metropolitan area, divided into 13 local government areas (LGAs): Gosford, Hornsby, Hunters Hill, Ku–ring–gai, Lane Cove, Manly, Mosman, North Sydney, Pittwater, Ryde, Warringah, Willoughby and Wyong. Most LGAs in NSCCAHS are well equipped in terms of infrastructure, transport system and health services. The area has the highest number of privately operated hospitals, totalling 23, and 15 day procedure centres (NSW Health, 2006b). Most of the population in this Area has good health outcomes compared to other AHSs with mortality rates significantly lower than for the whole of NSW. Its status as the richest AHS in NSW was determined as measured by the Socio– Economic Indexes for Areas (SEIFA) developed by the Australian Bureau of Statistics. According to the SEIFA index of relative socio–economic disadvantage (IRSD), most Local Government Areas (LGAs) in Northern Sydney Central Coast AHS (NSCCAHS) rank above the 1000 index score. The benchmark used is the Australia–wide scale that is set at 1000. Therefore, any LGAs ranked above 1000 are less disadvantaged, while those ranked below 1000 are more disadvantaged. In NSCCAHS, out of the 13 LGAs, only Wyong was ranked below average, falling within the 4th quintile according to the NSW Socio–Economic Status group, while eight LGAs had the highest SES 1st quintile and two LGAs 2nd quintile. This means Wyong is the most disadvantaged LGA in NSCCAHS (NSW Health, 2008) and it also has the lowest socioeconomic (SES) status ranking (5th quintile) compared to other Local Government Areas in the Northern Sydney Central Coast AHS (see Table 5–1).

139 ______Table 5–1: Index of Relative Socio–economic Disadvantage (IRSD) of Local Government Areas in Northern Sydney Central Coast, 2008

Source: NSW Chief Health Officers Report, 2008

On the other hand, Greater Western AHS (GWAHS) is located within a rural and remote setting sharing borders with South Australia, Victoria and Queensland and covers the far western part of the NSW; it is one of the most socially and economically disadvantaged AHSs. In terms of landmass, it is the largest AHS in the State covering 445,586 square kilometres, representing over 55 percent of the landmass of the whole of state of NSW. With only about 305,440 inhabitants, the area is the most sparsely populated in NSW, suggesting the complexity and costly nature of providing health care services to such a small population yet one that is vastly distributed. The GWAHS population represents 4.6% of the NSW population. Of the total population in GWAHS, 7.9% (24,345) identify as Aboriginal and Torres Strait Islander people. This is significantly higher than the NSW average of 1.9% and is the highest percentage of Aboriginal people in NSW (NSW Health, 2007a). GWAHS is divided into 28 local government areas as shown in Table 5–2, all classified as either regional, rural, remote or very remote.

140 ______Table 5–2: Index of Relative Socioeconomic Disadvantage (IRSD) of Local Government Areas in Greater Western Area Health Service, 2008

Source: NSW Chief Health Officers Report, 2008

In comparison to Northern Sydney Central Coast Area Health Service (NSCCAHS), Greater Western Area Health Service’s (GWAHS) Local Government Areas (LGAs) are more disadvantaged with only three LGAs showing IRSD scores above the Australia–wide average (1000) while most of the other LGAs were ranked below 1000 – being in the

141 ______lowest 4th or 5th quintiles, thus illustrating the high level of socio–economic disadvantage. In addition, when compared to NSCCAHS, which had only one LGA with the lowest socioeconomic status (SES), Greater Western Area Health Services had eight LGAs in the last quintile (5th) when weighed against other LGAs (see Table 5–2).

The sampling units in this study consist of the NSW Department of Health, two AHSs and individuals (key stakeholders) working or accessing health care services (consumers). As mentioned earlier, two contrasting AHSs (urban and rural/remote) in NSW were purposely selected for the qualitative phase. Purposive sampling in qualitative research is a deliberate non–random method of sampling, which aims to sample a setting or group of people with particular characteristics. This is sometimes called judgement sampling, meaning respondents can be selected because they have knowledge that might be valuable to the research process; therefore, the interviewees are explicitly hand–picked in order to generate insightful data to answer a particular question (Bowling, 2002).

According to Green & Thorogood (2004) and Patton (1990), the overall aim of purposive, as opposed to probability sampling, is to include ‘information–rich’ cases to gain an in– depth understanding of what is being studied. To achieve this, a number of other different sampling strategies can be considered such as snowballing, a sampling technique used where no sampling frame exists and cannot be created. For example, there may be no list of people with prior knowledge of the research issue of interest (Bowling, 2002). In such an instance, the researcher asks an initial group of respondents to recruit others they know in the target group who have an understanding of the research interest. The disadvantage of the snowballing method is that it can often include only members of a specific network. However, this study looked at two different levels of the NSW health system, ensuring that those interviewed belonged to different networks. An effort was also made to include several categories of participants, including policy makers, health executives, health managers and other relevant key stakeholders.

In the current study, purposive sampling for the geographically defined areas and the snowballing method for the recruitment of key stakeholders were used. The complex nature

142 ______of the resource allocation and equity issues required that the participants have an idea about the two key issues the study was attempting to address. Therefore, the first group of key stakeholders was mainly executives from the NSW Department of Health, the Northern Sydney Central Coast Area Health Service (NSCCAHS) and the Greater Western Area Health Service (GWAHS) who were then asked to nominate other relevant people who were likely to contribute new data to the study. Using the key concepts that were emerging as important from these interviews, the researcher was also able to identify an informative sample that generated data as proposed by both ‘purposive’ and ‘snowballing’ sampling strategies.

Purposive sampling, unlike ‘convenience sampling’, is based on information–rich cases, while convenience sampling is about selecting cases that are easy to access and inexpensive to the study. Thus, there are a number of weaknesses with the convenience sampling method, including built–in selection biases and non–statistical representativeness (Green and Thorogood, 2004). Therefore, the researcher must be careful not to select cases solely based on one’s personal convenience. The researcher, being new in the country (Australia) at the time of commencing this study, did not have prior experience or knowledge of the NSW health care system and so did not base selection of cases on personal convenience. Green and Thorogood (2004), warn against selecting a particular method or cases based on convenience and cost. The two AHSs: NSCCAHS and GWAHS were selected for their unique contrasting characteristics as noted earlier, rather than for convenience.

5.3.2 Rationale for Selecting NSCCAHS and GWAHS The selection of NSCCAHS and GWAHS for this study was largely based on the study objective of specifically targeting two contrasting areas (urban and rural). However, the final decision in terms of which urban or rural Area Health Service (AHS) to select among the eight AHS was based on the approval to participate in the study by the executive management and ethics committees of the various AHSs initially approached with the study proposal. Contact was first made with Sydney South West Area Health Service (SSWAHS) and Greater Southern Area Health Service (GSAHS); however, they both declined to take part in this study, citing different reasons. Northern Sydney Central Coast and Greater

143 ______Western Area Health Services were both keen to participate in the current study and consequently granted both executive and ethics committee approval. Due to lack of openness and complexities surrounding equity issues in health resource allocation, the process for the ethics application took longer than expected (eight months) from the initial contact with the relevant authorities to the time ethics approval was granted.

The agreement of both NSCCAHS and GWAHS to participate in this study was a good outcome for a number of reasons. First, there are huge disparities in health, including socio–economic conditions, between the two areas (see Tables 5–1 and 5–2). Second, there was a need for the study to contrast both urban and rural sectors in order to improve the representation of all eight Area Health Services (AHSs). Third, the study was designed in such a way as to provide evidence of how two contrasting (urban–richer and rural–poorer) areas deal with the decision –making processes of resource allocation, including the players and actors involved. According to the Australian Bureau of Statistics index of relative socio–economic disadvantage/advantage, Northern Sydney Central Coast Area Health Services (NSCCAHS) is one of the wealthiest areas, while Greater Western Areas Health Services (GWAHS) is one of the poorest areas in NSW.

Similarly, the NSCCAHS population has one of the best health outcomes, while the GWAHS population has the poorest health status in New South Wales. Indeed, Greater Western is, of all the AHSs, the one with the worst health indicators in NSW. For example, in 2006, Greater Western had the highest age–adjusted death rates (male 17% and female 12%) compared to other AHSs and the general NSW rates. By contrast, Northern Sydney had the lowest (male 9% and female 6%) compared to rates in NSW (NSW Health, 2006a). Apart from the differences in deaths from all causes, Greater Western Area Health Service scored lowest in the indexes of relative socioeconomic disadvantage, while NSCCAHS scored highest. This index score is one of the socioeconomic indices for areas (SEIFA) developed by the Australian Bureau of Statistics (NSW Health, 2006a). These extreme differences provided the justification for wanting to explore the views and perspectives of key stakeholders in the two clearly contrasting Area Health Services.

144 ______5.3.3 Data Sources and Recruitment of Participants In this study, the qualitative objective was achieved through primary and some secondary materials obtained from face–to–face interviews with key informants. The interviews were largely open–ended, using interview guides prepared by the researcher with specified broad domains to be covered during the interviews and key issues to be explored. An extensive literature review of policy documents was also undertaken by the researcher to clearly understand how the NSW health system operates, and to identify sources where important data could be gathered for purposes of triangulation. The use of multiple (triangulated) research methods was suggested by Denzin (1989), who argued that triangulation elevates the researcher above the personal biases that stem from single methodologies. Thus, combining methods and data in the same study can partially overcome the deficiencies that flow from one method or same setting data. The use of data triangulation requires the collection of data at different times and places and from different people or groups (Denzin, 1978, Denzin, 1970). Therefore, in this study all relevant sources were explored and related data collected at the same time as the interviews were conducted. The main sources for the secondary qualitative and quantitative data were administrative and policy documents of the NSW Department of Health and AHSs, as well as annual reports.

The process of recruiting potential participants involved self –introduction letters detailing the aims of the study. Once a prospective participant agreed to participate in the study, they were issued with a letter from the University of New South Wales and School of Public Health and Community Medicine approving the research project and provided with a participant’s information statement that further explained the study in detail. Before commencing each interview, permission was once again sought from the participant to ensure they were happy to participate in the interview. A total of 30 interviews were conducted, involving four NSW Health Senior management personnel, six policy makers, six Area Health Directors, twelve Program Managers and six consumers and their representatives. Each interview lasted about an hour. The following are some of the key headings that were used to guide the interviews:

145 ______ Health funding in NSW and understanding of equitable distribution of health resources;  Mechanisms for distributing health funds to administrative Area Health Services;  Resource allocation procedure including players/actors involved in decision making process and factors taken into consideration when funds are distributed;  Barriers to allocating resources more effectively across areas and population groups;  Disadvantaged population groups and health personnel’s understanding of needs– based resource allocation;  Stakeholders’ consultations during resource allocation decision–making;  Area Health Services’ system of financial accountability;  Health services’ and programs’ capacity to meet needs of population;  Role of private hospitals in NSW and level of collaboration;  Sharing of health responsibilities between federal, state and local governments;  Amalgamation of Area Health Services from 17 to current 8;  Level of consultation regarding resource allocation; and  Hospital and community based services funding.

The interviews commenced at the NSCCAHS and NSW Health Department before moving to GWAHS. Some of the interviews conducted at the State level were repeated to ensure the information obtained from the initial interviews was correctly represented, including seeking further explanations for some of the issues raised at the AHSs. The main motive for starting fieldwork at the AHS level was to determine their level of understanding of resource allocation and equity in funding and to give the researcher an opportunity to raise any issues of concern with policy makers at the State level. It was also an opportunity to assess whether financial data were available at those levels and to gain some prior insights into the AHS resource allocation mechanisms. All interviews were conducted between February 2006 and August 2007 and were tape–recorded. The stakeholders were also asked verbally if they could be tape–recorded and reminded that they were free to withdraw their participation any time in the course of the interview on the interview day. All stakeholders consented to being tape–recorded and no concerns about participating in the study were

146 ______raised at any point during the study period. Similarly no requests were made to withdraw from the study or to withdraw the data collected.

5.3.4 Data Collection and Management The primary objective of the qualitative method was to consult with and document views of key stakeholders at both the State (NSW Department of Health) and Area Health Service (AHS) levels about their understanding of the resource distribution procedures, including who were the key players/actors involved in the decision making process and the factors taken into consideration. To accomplish this objective, primary data were gathered from key decision makers and officials within the NSW health care system consisting of the Department of Health, the Northern Sydney Central Coast Area Health Service (NSCCAHS) and the Greater Western Area Health Service (GWAHS). The primary data were gathered through use of digital voice recorded, semi–structured interviews facilitated by interview guides. Each interview was downloaded and properly labelled, including the participant’s name, position, study site, date, time and length of the interview. However, the interview data were given special codes and any information that would potentially identify the participant deleted. All interview materials were downloaded into a laptop folder and a backup folder copied to an external memory for safe storage. A very effective digital voice recorder was used to record all the interviews.

The digital voice recorder is commonly used as a reliable tool for recording interviews and is effective in terms of its capacity to clearly record each interview. However, each interview was reviewed immediately afterwards to ensure that the interview was audible. Transcribing of interviews was carried out over a period of six months. All 30 interviews were transcribed verbatim even if some of the interview conversations were repetitive. Due to the sensitivity of most of the interviews conducted, the transcription of all the interviews was undertaken by the researcher to ensure confidentiality of the information was maintained, including the identity of the participants.

147 ______5.3.5 Data Analysis Qualitative data analysis involves a broad range of approaches with the task of the researcher defined rather differently across that range due to lack of collectively acceptable rules that should be applied in such analysis. According to Green and Thorogood (2004), most approaches to analysing qualitative data attempt to ‘intervene’ at some level, to draw out the ‘meaning’ of the data that is not obvious at a journalistic or narrative reading. In health research, there is a requirement for the researcher to be explicit about how the data were collected, analysed, and how the ‘stories’ or other materials are selected, interpreted and organised. The most common requirement in qualitative methods, is that the researcher must be thoroughly familiar with the field notes, the tapes or digital recordings and their transcriptions and any other data collected in order to analyse and accurately present data (Bowling, 2002). This study used the principles of ‘grounded theory’, a more systematic view of qualitative data analysis developed by Glaser and Strauss (1967) and Strauss (1987). According to Glaser and Strauss;

“…the strength of grounded theory approaches lies in the cyclical process of collecting data, analysing it, developing a provisional coding scheme, using this to suggest further sampling, more analysis, checking out emerging theory,… until a point of saturation is reached, when no new constructs are emerging. At this point, you have a rich, dense theoretical account – but one that is completely grounded in empirical data. It is both inductive and deductive, moving back and forward between emerging theory and data” (pg 67)

A first step in a grounded theory analysis is the emphasis on intense coding of early data. By ‘codes’, Glaser and Strauss mean conceptual labels which identify what general phenomenon is indicated by the extract of talk being analysed and not merely descriptive summaries of the data (Glaser and Strauss, 1967). For example, in this study the coding involved the conceptual labels related to the interview guide prepared by the researcher (i.e. 1.1 Health Funding in NSW Health, 1.2 Understanding of equity and equitable distribution of health resources, 1.3 Mechanisms for distribution of health funds to AHSs etc.). This was followed by ‘open coding’, an intense line–by–line analysis of a transcript to attempt to open up the data. This type of analysis forces the researcher to take a step back and open up all potential avenues of inquiry. The codes were used to generate themes and sub–themes,

148 ______which were then regrouped into equity of resource allocation and factors that influence the process. From this stage, meanings were developed to explain the factors that emerged as influential for resource allocation. The NSW Health policy documents were consulted to make sense of some of the findings. Comparative analysis of the way the resource allocation process was perceived at different levels of the health system was undertaken in order to understand factors behind equitable allocation of resources at the state level and how it reflects equitable allocation at the Area Health Service level.

The qualitative data organisation, preparation and analysis started while data collection was in progress as several studies have shown that making sense of the data collected in order to analyse and present the data is challenging, time–consuming and expensive (Patton, 2002, Marshall and Rossman, 1999, Bowling, 2002). Since time had already been wasted while waiting for Ethics Approval from the NSW Department of Health and the two Area Health Services, an ongoing data analysis, while collecting data in order to inform and improve the research process (Glaser and Strauss, 1967), was necessary. Therefore, most of the data was transcribed, partially analysed, organised and the participants de–identified while the data was being collected. The conceptual labelling of the data involved continuous analysis and interpretation of the transcripts, looking for key themes and patterns and using the following research questions to guide the analysis:

1) To what extent has resource allocation to Area Health Services from NSW Department of Health been equitable? 2) How are the financial resources received from the state level distributed at the Area Health Service level? 3) Is equity reflected in the allocation of NSW Health resources at the Area Health Service level?

The process involved going back to the data to check the emerging ideas, and refining the conceptual labels and theories. Other questions used to guide the analysis included the following:  What does the resource allocation procedure involve?

149 ______ Who are the key players and actors involved in the resource allocation decision – making process?  What factors are taken into consideration at the population level and do they reflect equitable distribution of resources at that level?  How is equity perceived at the State and Area Health Service levels? To what extent does the equity principle influence distribution of funds at both levels of the health system?  What role does politics play in resource distribution decision –making?  To what extent does the population level of health needs influence the way resources are allocated?  How are the Area Health Service officials involved in resource allocation at the state and Area Health Service levels?

The initial qualitative data analysis in this study was mostly undertaken manually during the data transcribing process by highlighting the conceptual labels in different colours. However, all transcripts were thoroughly read and comments were made in the margins during reading regarding the main issues being raised by the data and how to categorise these issues. The next step in the process involved reading the transcripts and undertaking a ‘cut and paste’ process whereby relevant themes in the Microsoft Word document were again highlighted in transcripts and then cut out and pasted onto a Microsoft Excel spreadsheet table, organised in terms of coded themes, as well as with cross–references to the original source to enable the researcher to trace it back to its original context. This proved to be a useful process because, even though the interviews were somewhat categorised by domains, covered as designed in the interview guide and the conceptual labels, new issues emerged and some of the responses were shifted around under different headings from where they were initially given.

The categorisation of the data was according to constructed broad headings in line with the interview domains and emerging themes that emerged from what the respondents were largely talking about. Key stakeholder responses from the State and Area Health Service

150 ______levels were grouped together under various headings. During the development of the codes and categories, consideration was given to the issue of emerging themes; some other parts simply reported common issues mentioned by the key stakeholders since in this area of research not much is known. Green and Thorogood (2004), argue that the researcher should move beyond simply categorising and coding the data to thinking about how the codes relate to each other and asking more complex questions regarding the relationship, links, differences between the themes emerging and context of those particular codes (Green and Thorogood, 2004).

Qualitative data analyses require regular re–visiting of and interaction with the data and the organisation system to help verify the meaningfulness and precision of the categories, including the assignment of data to categories. To cross–check the categorisation, NVIVO 8 software was also used to identify patterns and themes emerging from the State and Area Health Service levels. Themes were not imposed but allowed to emerge from the data. Emerging themes were analysed first for different levels of the health care system, which generated new insights into factors that underlie resource allocation and equity in the NSW health system. Different colours were used to highlight important sentences that were then linked to the appropriate theme codes. NVivo software was used to help manage and retrieve data once the coding scheme and coding of data were developed. The advantage of using NVivo after manual data analysis was to be more thorough and systematic.

Pre–constructed coding schemes were used to identify the respondents. Each code started with abbreviations that indicate the rank of the respondent, for example SHO_DOH (Senior Health Official, Department of Health), SHO_AHS (Senior Health Official, Area Health Service), PM_DOH (Policy Maker, Department of Health), PM_AHS (Program Manager, Area Health Services) including the health system level (NSW Health or Area Health Services), however, due to the sensitivity of the information being shared by respondents, all respondents are referred to as Senior Health Official in this thesis. This study assumed that using the job titles of health personnel and level of the health system when writing the report would put the officials at risk of being identified as having participated in this study.

151 ______It is noteworthy that most of those interviewed were senior personnel since they are the decision–makers and were more conversant with the processes of allocating resources.

5.3.6 Data Interpretation The interpretation of research findings is generally considered to be the core of qualitative analysis. According to Glaser and Strauss (1967), this requires that researchers submerge themselves in the data, read and re–read transcripts, code and organise, cross–examine and actively look for themes, explanations and understandings that can only be produced from getting ‘one’s hands dirty’ with the data. In this study, the grounded theory principle was the main approach adopted to search, identify, code and categorise patterns, topics and themes in the qualitative data. Grounded theory, sometimes called the constant comparative method, is known for its potential ways of developing a deeper analysis of qualitative data. The reason it is known as the ‘constant comparative method’ is its insistence on ‘constant comparison’, and the notion that interpretation of data moves forward through comparing codes, cases, and data sets (Glaser and Strauss, 1967). In the study, the grounded theory principle was used for its strength in guiding the process of data collection, analysis, and development of the provisional coding scheme. Using the coding undertaken in the field as a gauge, further participants were recruited, interviewed and data analysis undertaken to determine whether data collection had reached a point of ‘saturation’ (Green and Thorogood, 2004). This is the stage at which no new data were emerging. At this point, the research data collected was deemed to provide sufficient accounts of how equity and resource allocation are perceived within the New South Wales health system. Through induction and deduction (Gilgun, 1995, Greene et al., 1998) – moving back and forward between emerging theories and data – the researcher was able to interrogate the content of the transcripts by asking questions around the data which helped generate some useful early ideas about the data, and lines of inquiry to follow. The deductive method of investigation helped the researcher to start with general ideas and developed theory. At the same time induction was used to test observations from the initial data collected.

152 ______In quantitative and qualitative research, reliability and validity are important aspects since they determine whether insights resulting from research are truthful from the standpoint of the researcher, the participant, or the reader of the accounts (Bowling, 2002). The main approach adopted to ensure the accuracy of the findings was triangulation of data sources (NSW Department of Health and Area Health Services). The multiple case study design that focused on different administrative Area Health Services (AHSS), the NSW Department of Health levels and the different stakeholders included in the study, facilitated the triangulation of data sources. Information derived from different sources at different times was compared to build a coherent justification for the themes. For instance, interviews and policy documents were compared and interview data were re–examined. Similarly, views of participants regarding resource allocation and equity from different levels of the health care system were compared.

5.4 Quantitative Method Overview Quantitative methods were applied in the analysis of resource allocation at both the State (NSW Department of Health) and Area Health Services (AHSs) levels. As indicated earlier, the extent of equity of resource allocation to Area Health Services was assessed at the State level. In order to do this two main indices were developed – a general health need index (GHNIdx) and a double–variable index (DVIdx). These two indices were used as proxies of health needs and were developed using statistical techniques. At the State level (inter– AHS), both indices were applied but only the DVIdx was applied at the Area Health Service Level (intra–AHS). All the eight AHSs in NSW (Sydney South West, South Eastern Sydney & Illawarra, Sydney West, Northern Sydney Central Coast, Hunter and New England, North Coast, Greater Southern, and Greater Western) were included in the analysis; however, the third research question was addressed by purposefully selecting two AHSs (Northern Sydney Central Coast and Greater Western) as case studies. The next section describes the process of developing the general health need index, including the variables used and where they were obtained.

153 ______General Health Need Index (GHNIdx)

The GHNIdx is the first proxy of health needs used as a benchmark to assess the extent of equity in resource allocation at the State level. It is supposed to give the predicted equity share of resources that should go to each Area Health Service (AHS) in proportion to their level of health needs. The index was derived by weighting the population of each AHS by its level of health needs score constructed using different variables.

Double–Variable Index (DVIdx) is the second alternative proxy of health needs also used as a yardstick to assess the extent of equity in resource allocation from the NSW Department of Health to the eight AHSs (intra–Area Health Services). It is supposed to give the predicted equity share of resources that should go to each AHS in proportion to their level of health needs. The index was constructed using consolidated AHSs morbidity (2001) and premature mortality (2000 – 2002) data. DVIdx was developed based on two measurable indicators (morbidity and premature mortality). A similar index was also used as a benchmark to assess the extent of equity in resource allocation within two AHSs (Northern Sydney Central Coast and Greater Western) but the data for this level of analysis was aggregated according to the Local Government Areas (LGAs) in each AHS.

Socio–Economic Index for Areas – Index For Relative Socioeconomic Disadvantage SEIFA–IRSD was the third index that the study attempted to use as proxy of health needs. SEIFA–Indexes have often been used as a measure of health needs; therefore, this study attempted to contrast the two indices developed with the Australian Bureau of Statistics Socio–Economic Indexes for Areas – Index of Relative Socioeconomic Disadvantage (SEIFA–IRSD)14 but the differences in funding were astronomical. This is because SEIFA indexes are ordinal measures of socio–economic level that can be used to rank areas but cannot be used to measure the size of the difference of socio–economic level between

14 Socio–Economic Indexes for Areas (SEIFA) consists of four indexes developed by the Australian Bureau of Statistics (ABS). Each index summarises a different aspect of the socio–economic conditions of the NSW population using a combination of variables from the Census of Population and Housing i.e. the Index of Relative Socioeconomic Disadvantage (IRSD).

154 ______areas. For example, we cannot infer that an area with an Index of Relative Socio–Economic Disadvantage value of 500 is twice as disadvantaged as an area with an index value of 1,000.

Therefore, although it is possible to compare the distribution of an index across Area Health Services (AHSs) using ‘box plots’ as explained in the SEIFA Technical Paper (Adhikari, 2006), these comparisons cannot be relied upon as the basis for precise allocation of resources between AHSs. As shown in Appendix A, the SEIFA–IRSD index distribution of resources reflects a number of factors which may in some way or another have an influence on health status but are quite outside the control of NSW Health and are not affected by the NSW Health allocation of resources. Therefore, the implementation of the SEIFA indexes derived distribution pattern, as shown by Tables and Figures in Appendix A, would be both politically and administratively impossible due to the enormous gap in funding.

The purpose of attempting to use different indices as proxies of health needs was to estimate which of the indices was more likely to capture population health needs, considering the lack of consensus on which indicators should be used to measure health needs. The following is a brief description of the tools. The process involved in the development and use of the indices is explained in detail later in this chapter.

The reason NSW Health was selected for this study is because it is currently the only state in Australia that has fully implemented a Resource Distribution Formula (RDF) since the early 1990s with the aim of reducing geographical inequities in health funding. Therefore, in order for the current study to address its main objectives, it had to focus on the three levels of the health care system in NSW as mentioned above. The first objective was to establish whether and the extent to which resource allocation from the State level of the NSW health system to the Area Health Services had been equitable. To achieve this objective, the study had to focus on all eight AHSs. The second objective was to examine the pattern of distribution of funding within the AHSs, and the third was to establish whether the current system of allocating resources within the AHSs promotes equity. The

155 ______third objective was pursued using two Area Health Services: Northern Sydney Central Coast and Greater Western as case studies.

5.4.1 Data Sources The quantitative strand of this study involved exploration of a range of published documentary evidence from NSW Health, the Area Health Services (AHSs), the Public Health Information Development Unit (PHIDU) and the Australian Bureau of Statistics. The first set of data collected from these organisations included demographic, socio– economic, and health–related data, including premature mortality and morbidity data. The premature mortality data were compiled by PHIDU from ABS deaths, 2000 to 2002 and the morbidity data were estimated from the 2001 National Health Survey (NHS) conducted by ABS and can be freely accessed online. The assumption made was that the population characteristics had not changed much between 2001 and 2008. The data were used in the development of various indices.

The second set of data was gathered from the NSW Department of Health and AHSs audited financial reports included in their annual reports and accessible online. Funding allocations to the eight AHSs from 2003/04 to 2006/07 were extracted from the reports and entered into Microsoft excel spreadsheets. The four years (2003/04 to 2006/07) selected to assess the extent of equity was in order to coincide with the ‘tail–end’ of the NSW Health ‘average distance from RDF target’ equity graph shown in Chapter 1 (Figure 1–3). This period was chosen because, looking at the NSW equity graph, that is where the gap between the actual resources allocated and the RDF target shares had narrowed. The expectation is that a comparison of actual resources allocated and allocations based on the health need indices developed in this study will show the minimum gap possible (i.e. plus/minus 10%) in order to be judged equitable. Principal component analysis (PCA) (Jolliffe, 2002, Jolliffe, 1986) and Microsoft Excel were the statistical techniques used to develop indices and analyse all the data. Finally, mapping of health facilities in each Local Government Area (LGA) was conducted using the number of health facilities extracted from the NSW Department of Health and Area Health Services annual reports.

156 ______5.4.2 Data Analysis – Principal Component Analysis Principal component analysis (PCA) was the main statistical technique used for the measurement of socio–economic disadvantage and for constructing a general health need index in the current study. PCA is a classical statistical technique that identifies linear combinations of random variables that maximise variance. According to Jolliffe (2002),

“…the central idea of principal component analysis (PCA) is to reduce the dimensionality of a data set consisting of a large number of interrelated variables, while retaining as much as possible of the variation present in the data set. This is achieved by transforming to a new set of variables, the principal components (PCs) which are uncorrelated, and which are ordered so that the first few retain most of variation present in all of the original variables” (Jolliffe, 2002, pg 6).

The first principal component accounts for as much of the variability in the data as possible, and each succeeding component accounts for as much of the remaining variance. The PCA is a multivariate choice method that develops a composite index by defining a number of possibly correlated variables into a smaller number of uncorrelated variables known as principal components. Given a set of explanatory variables, the PCA helps select the most important variable or a limited number of variables from a set of many relevant variables. The PCA has been used in numerous studies, especially those aiming to unravel the relationship between different factors in the population using large data sets, to reveal the internal structure of the data in a way which best explains the variation in the data and the meaningful underlying relationships between all the variables. According to Kirkwood (1999), the principle of this technique lies in the fact that weights are given to each of the variables determined by the variation in the linear composite of the variables (see Table 5– 5).

In the current study, the application of the Principal Component Analysis (PCA) technique enabled the researcher to determine a vector known as the first Principal Component; it is linearly dependent on the variables, having a maximum sum of squared correlation with the variables. After selecting and defining the variables, they were all transformed to improve normality. They were then correlated against each other, using Spearman’s correlation, before being analysed using the PCA. According to Jolliffe (2002), with PCA it is imperative that the variables correlate highly with each other for results to be meaningful.

157 ______Many variables from the initial list of 35 variables (see figure 5.2 below) that did not show evidence of high correlation with other variables were excluded from the PCA. When constructing a health need index, additivity is an important criterion that should be met (McIntyre et al., 2002, Gordon, 1995). Therefore, cross–analysis of the data was undertaken to ensure the variables did not lead to double counting or measuring the same thing. Variables deemed to measure the same thing were excluded from the final analysis. Figure 5.2 shows the 35 variables initially selected for the PCA.

158 ______Table 5–3: Initial Variables Defined Variable Definition

1. AUND14 % of population under 14 years of age 2. AOVER65 % of population aged 65 and over 3. INCUND300 % of population earning under $300 per week 4. INOVR1500 % of population earning $1500 or more per week 5. DISSICBEN % of population on disability & sickness benefit 6. UNEMPBEN % of population receiving unemployment benefit based on active labour force 7. PARBEN % of population receiving parenting benefit 8. ASROBTHFT % of population who reported assaults & robberies & thefts 9. ADPHACT % of population receiving inadequate physical activity 10. FRUVEG % of population consuming fruit & vegetables less than recommended intake per day 11. SMOK % of population from age 16 smoking daily 12. ALCODRNK % of population aged over 16 consuming alcohol at high risk 13. DTHSALL % of deaths rate from all causes per year 14. HOSPCAS % of population hospitalised per year 15. ABOPOP % of Aboriginal population 16. HHOWN % of households owned based on total households 17. ONEPARFAM % of one parent family with dependent children 18. SECEDU16 % of population aged 16 years who have not participated in full/part–time secondary education 19. UNEMP % of active population unemployed in 2003 20. NonESB % of population from non–English speaking backgrounds with poor proficiency in English 21. DWELRGH % of dwellings rented from government housing commission 22. HHWMOTOR % of dwellings without a motor vehicle 23. NOCOMPUSE % of population who do not use computer at home 24. NOPHINS % of population without private health insurance 25. INFFIMM % infants not fully immunised based on total infants population 26. DIABTYP2 % of estimated population suffering from chronic disease–type 2 diabetes 27. MENTHLTH % estimate population suffering from chronic mental & behavioural disorders 28. OBSIT_POP % of estimated population who have type 2 diabetes & are overweight & obese 29. DENCOND % of population suffering from dental conditions – ambulatory care sensitive 30. DGHCWN % of population having difficulties getting health care when needed 31. UGPSERPP Utilisation of GP services per population (%) 32. BEDSPHU Population per one public 33. BEDSLPRH Population per one private hospital bed 34. OUTPATHOP % of population who utilise hospital outpatient services 35. AVDTHSUN75 % proportion of avoidable deaths for persons under the age of 75

Variables that were not showing strong correlations were eliminated and the remaining 16 variables were used as a measure of socio–economic disadvantage and to assess the levels of general health needs within the eight AHSs. As stated earlier, the data were extracted from three main sources: The Australian Bureau of Statistics Census dataset; the Public

159 ______Health Information Development Unit; and the NSW Department of Health and AHSs’ published documents. These data can be grouped into three broad categories: demographic, socio–economic and health–related. Table 5.4 below shows the results of the principal component analysis.

Table 5–4: Unrotated Component Matrix Components/ Variables 1 2 3 DISSICBEN 0.97 0.17 0.12 INCUND300 0.96 0.10 –0.26 UNEMPBEN 0.95 0.11 0.01 DGHCWN 0.93 –0.20 –0.10 OUTPATHOP 0.91 –0.26 –0.09 ONEPARFAM 0.90 0.31 0.17 NOCOMPUSE 0.86 0.10 0.37 PARBEN 0.82 0.36 0.41 DENCOND 0.81 –0.47 –0.08 ABOPOP 0.77 –0.45 0.22 OBSIT_POP 0.69 –0.34 –0.11 NESB –0.65 0.42 0.50 DIABTYP2 0.36 0.87 –0.30 MENTHLTH 0.52 0.80 0.10 AUND15 0.43 –0.30 0.77 AOVER65 0.66 0.07 –0.72 Note: The variables in each component with more than 50% loading together explain a particular dimension of health needs

As shown in Table 5.5, it is a requirement in disadvantaged or health need measurement indices that variables be assigned weightings to differentiate their relative role in the overall index (McIntyre et al., 2002, Crampton and Laugesen, 1995). The PCA produced coefficient score weightings of all the variables and these weightings were used to generate the final general health need index (GHNIdx) for the eight AHSs. The general health need index was constructed through multiplication of the component score coefficient of each variable with the standardised scores (z–scores) for each variable and both coefficient and standardised scores were then summed together. The z–scores are derived by subtracting the mean from the raw score and then dividing the difference by the standard deviation (Kirkwood, 1999). This is represented by the formula shown in Figure 5–2.

160 ______Table 5–5: Component Score Coefficient Matrix Components/ Variables 1 2 3 AUND15 –0.09 –0.04 0.40 AOVER65 0.21 0.03 –0.31 INCUND300 0.12 0.08 –0.10 DISSICBEN 0.02 0.11 0.07 UNEMPBEN 0.05 0.09 0.03 PARBEN –0.10 0.18 0.17 ABOPOP 0.09 –0.09 0.18 ONEPARFAM –0.03 0.16 0.07 NESB –0.24 0.12 0.16 NOCOMPUSE –0.04 0.10 0.18 DIABTYP2 –0.06 0.28 –0.23 MENTHLTH –0.13 0.29 –0.03 OBSIT_POP 0.14 –0.07 0.01 DENCOND 0.17 –0.11 0.05 DGHCWN 0.13 –0.02 0.01 OUTPATHOP 0.14 –0.04 0.02 Note: Variable weights are component score coefficients produced by the PCA. Only coefficients of variables with 50% or more loadings in the first component were used.

Figure 5–2: Health Needs Formula

GHNIdxt = W1 * Vt1 - V1 + W2 * Vt2 – V2 + …+ Wn * Vtn – Vn

S1 S2 Sn

th GHNIdx t = Health Need Index of t Area Health Service

W1 = Weight/component score coefficient of the first variable

V = Raw score of the first variable of tth Area Health Service t1

V 1 = Mean of the first variable

S1 = Standard deviation of the first variable

161 ______5.4.3 The General Health Need Index (GHNIdx) The main assumption underpinning the GHNIdx is that, for resource allocation to be equitable, it should reflect the differences in population size, level of socioeconomic disadvantage and health needs. The GHNIdx was applied as a benchmark against which equity in resource allocation among the eight Area Health Services (AHSs) was assessed. It was calculated by weighting the total population of each AHS by its health need index score. All the AHSs’ GHNIdx scores were first normalised against the score of the least disadvantaged AHS. The normalised score was then used to weight the actual population. The normalisation of scores was undertaken in order to standardise the derived scores by translating those that were negatives to positives and then adding up the GHNIdx to give the expected equity–adjusted allocation that should go to each AHS on the basis of their health needs.

As shown in Figure 5.3, the higher the general health need index (GHNIdx), the greater the funding allocated to an AHS. Finally, the total yearly allocations to the AHSs were re– distributed using the proportions yielded by the GHNIdx to obtain the expected allocations; these were then compared with the actual allocations the AHSs received from the NSW Department of Health. Figure 5–3 also illustrates the constructed GHNIdx where the AHSs are ranked according to their GHNIdx values. The AHSs with the highest positive GHNIdx values were regarded as being highly disadvantaged or having greater health needs and those with low negative GHNIdx were advantaged or had fewer health needs. The next section discusses the rationale behind the development of the Double Variable Index (DVIdx) consisting of premature mortality and morbidity data.

162 ______Figure 5–3: NSW Area Health Services – General Health Need Indices (GHNIdx)

1.50

1.00

0.50

0.00 SSWAHS SESIAHS SWAHS NSCCAHS HNEAHS NCAHS GSAHS GWAHS -0.50

-1.00 General Health Need Index Score Index Need Health General -1.50 Area Health Services

GHNIdx

Note: Negative values imply AHSs with fewer health needs while positive is high levels of health needs

5.4.4 Premature Mortality and Morbidity: The Double–Variable Index (DVIdx) In Australia, illness and sickness are unevenly distributed in the population, with people in lower socioeconomic status groups experiencing more ill health (McClelland and Scotton, 1998, House, 2001, Turrell and Mathers, 2000, Hayes et al., 2002). For example, the health of Aboriginal and Torres Strait Islander people is significantly worse than that of other Australians (Turrell and Mathers, 2001). The Indigenous population has an age– standardised death rate at least twice that of the non–indigenous population, with a male indigenous person aged 35–44 years almost six times more likely to die than a non– indigenous male of the same age (Australian Institute of Health and Welfare, 2001a). Although infant mortality rates per 1000 live births for Aboriginal population groups are declining, they are still three times greater than those for non– (Mathews, 1997, Australian Bureau of Statistics, 2006b). Aboriginal infants’ birth weights are also considerably lower than those for non–Aboriginal infants (Australian Institute of Health and Welfare, 2006, Australian Bureau of Statistics, 2006b).

163 ______Recent studies show that groups from the most disadvantaged socioeconomic backgrounds in Australia have poorer health in terms of higher mortality rates (Turrell and Mathers, 2001), incidence of chronic illness, a greater prevalence of disability and handicap and higher self–reporting of fair and poor health than the rest of the population (Mathers et al., 1999, Healy et al., 2006, Australian Institute of Health and Welfare, 2006). The interdependence of various socioeconomic factors known to affect health status means that the task of attempting to alleviate health inequalities is likely to be particularly complex as improving health status cannot be addressed satisfactorily by simply allocating more resources to health care services.

The focus of national health efforts needs to be first on the improvement of overall health by reducing premature mortality and morbidity rates through targeting disadvantaged population groups with health resources and services and second on improving the broader social, economic and political determinants of health. People who are relatively socially and economically disadvantaged experience worse health than those of higher socioeconomic status for almost every major cause of mortality and morbidity. Therefore, a Double–Variable Index (DVIdx) was developed using premature mortality and morbidity data extracted from the Public Health Information Development Unit, established by the Australian Government Department of Health and Ageing in 1999, to assist in the development of public health data systems and indicators. As explained earlier, the premature mortality data was compiled from the ABS deaths, 2000 to 2002, and the morbidity data from the 2001 ABS National Health Survey. For the first DVIdx, the data was aggregated to an Area Health Service level as shown in Table 5–6.

164 ______Table 5–6: Development of Double Variable Index (DVIdx) for Eight Area Health Services in NSW

Data Source: NSW Health and Public Health Information Development Unit 2007

This basic index of health needs was then used as an alternative proxy of health needs in the AHSs’ population. The DVIdx was developed using the additivity procedure of both variables and then finding an average; this average was used as a benchmark to assess health needs within the eight Area Health Services (AHSs) and their respective Local Government Areas (LGAs). According to the DVIdx index, the AHSs with the greatest incidence of premature mortality and prevalence of chronic morbidity rates attracted higher percentage shares. This is discussed in detail in chapter 6 where the results are also presented.

The Double Variable Index (DVIdx) developed for Local Government Areas (LGAs) used aggregated LGA level data (incidence of premature mortality and prevalence of morbidity for population under 75 years) for two specific Area Health Services: Northern Sydney Central Coast and Greater Western as shown in Tables 5–7 and 5–8. The total chronic disease data used in this study were collated in 2001, while the premature death data was gathered in 2000–2002, being deaths reported as occurring between those periods. The reason behind the variation is the lack of uniformly assembled data consistent in all years. The assumption made in this analysis is that the prevalence of chronic diseases data reported in 2001 has not drastically changed and so does not affect the analysis.

165 ______Table 5–7: Development of Double Variable Index (DVIdx) for Northern Sydney Central Coast AHS Local Population Chronic Premature Chronic Premature Double Government 2007 Diseases Mortality Diseases Mortality Variable Areas 2001 2000–2002 (%) (%) Index (DVIdx) % Gosford 164,512 411,079 1,461 16.4% 19.6% 18.0% Hornsby 146,323 332,719 854 13.3% 11.5% 12.4% Hunter’s Hill 139,396 29,834 85 1.2% 1.1% 1.2% Ku–ring–gai 100,068 229,271 537 9.1% 7.2% 8.2% Lane Cove 57,672 70,121 170 2.8% 2.3% 2.5% Manly 39,350 87,021 223 3.5% 3.0% 3.2% Mosman 32,508 58,698 127 2.3% 1.7% 2.0% North Sydney 14,063 128,161 310 5.1% 4.2% 4.6% Pittwater 108,371 127,994 343 5.1% 4.6% 4.9% Ryde 60,686 223,737 684 8.9% 9.2% 9.0% Warringah 157,911 315,294 871 12.6% 11.7% 12.1% Willoughby 28,214 132,153 305 5.3% 4.1% 4.7% Wyong 65,491 361,235 1,485 14.4% 19.9% 17.2% TOTAL 1,114,565 2,507,317 7,455 100% 100% 100% Data Source: NSW Health and Public Health Information Development Unit 2007

The two indexes constructed and applied in this study were the Double–Variable Index – DVIdx and the General Health Need Index – GHNIdx.. The Resource Distribution Formula (RDF) constructed by the NSW Department of Health was also applied to the actual funds allocated to the eight AHSs from 2003/04 to 2006/07 to determine the extent to which the RDF had indeed been applied as stated in the NSW Health policy documents (NSW Health, 2005) (see Chapter 6). The next section presents the RDF in more detail.

166 ______Table 5–8: Development of Double Variable Index (DVIdx) for Greater Western AHS Local Government Areas Population Chronic Premature Chronic Premature Double 2007 Diseases Mortality Diseases Mortality Variable 2001 2000–2002 (%) (%) Index (DVIdx) % Balranald 2,546 7,379 34 1.1% 1.1% 1.1% Bathurst Regional 37,542 72,482 236 11.0% 7.5% 9.3% Blayney 6,891 15,881 72 2.4% 2.3% 2.4% Bogan 2,994 8,800 64 1.3% 2.0% 1.7% Bourke 3,217 9,300 48 1.4% 1.5% 1.5% Brewarrina 1,998 8,500 38 1.3% 1.2% 1.3% Broken Hill 20,139 56,484 269 8.6% 8.6% 8.6% Cabonne 12,907 41,194 105 6.2% 3.3% 4.8% Central Darling 2,008 8,500 38 1.3% 1.2% 1.3% Cobar 5,112 10,200 56 1.5% 1.8% 1.7% Coonamble 4,342 9,800 53 1.5% 1.7% 1.6% Cowra 13,025 33,400 131 5.1% 4.2% 4.6% Dubbo 39,499 91,148 345 13.8% 11.0% 12.4% Forbes 9,755 25,519 109 3.9% 3.5% 3.7% Gilgandra 4,702 12,745 37 1.9% 1.2% 1.6% Lachlan 6,927 17,600 98 2.7% 3.1% 2.9% Mid–Western Regional 21,983 68,749 297 10.4% 9.5% 9.9% Narromine 6,776 17,238 68 2.6% 2.2% 2.4% Oberon 5,260 11,797 41 1.8% 1.3% 1.5% Orange 37,009 88,413 349 13.4% 11.1% 12.3% Parkes 14,846 37,527 160 5.7% 5.1% 5.4% Walgett 7,199 21,500 109 3.3% 3.5% 3.4% Warren 2,871 9,200 34 1.4% 1.1% 1.2% Warrumbungle Shire 10,208 18,728 88 2.8% 2.8% 2.8% Weddin 3,793 10,295 40 1.6% 1.3% 1.4% Wellington 8,406 22,343 124 3.4% 4.0% 3.7% Wentworth 7,086 18,540 76 2.8% 2.4% 2.6% Unincorporated Far West 756 3,200 16 0.5% 0.5% 0.5% GWAHS Total 304,872 756,462 3,135 100% 100% 100% Data Source: NSW Health and Public Health Information Development Unit 2007

5.4.5 Resource Distribution Formula (RDF) As discussed at length in Chapter 4, the RDF is the main tool used by NSW Health to guide the allocation of funds from the Department of Health to the geographically defined Area Health Services (AHSs). According to NSW Health (2005), funding to AHSs is based on the principle of providing Areas with the resources that would facilitate comparable and comprehensive access to services that meet the health needs of their population. This

167 ______suggests that the focus is more on ‘health care equity’ rather than on ‘health equity’. The Resource Distribution Formula funds allocation to AHSs takes into account a range of factors including the population age/sex; population health needs; private sector substitutability factor; factors for direct and indirect teaching and research. The RDF is used as a planning tool to identify equitable shares of available resources for each AHS and to monitor progress towards the achievement of fairness in health funding. However, the formula does not identify the total amount of funding available for distribution by NSW Department of Health, as this is a matter decided by the NSW Parliament (NSW Health, 2005).

5.5 Mapping of Health Needs and Facilities The Local Government Areas (LGAs) within Northern Sydney Central Coast AHS and Greater Western AHS were mapped to demonstrate for the first time the level of premature mortality consisting of deaths from all causes for people under 75 years of age, prevalence of chronic diseases (DVIdx) and the SEIFA–IRSD index of relative socio–economic disadvantage rankings as measured by the Australian Bureau of Statistics. Regarding the use of SEIFA–IRSD index for mapping relative socioeconomic status of LGAs within an AHS (see Chapter 8, Section 2), the quintile was used showing where each LGA lies, rather than by LGA index value since the LGAs index value is an ordinal value as explained earlier.

The mapping of Area Health Services LGAs illustrates the social geography of health and health needs using routinely collected indicators of population health that are geographically locatable and reflect health needs i.e. premature mortality and chronic diseases.

For decades, premature mortality has been used as a powerful indicator to illustrate disparities in both health status and access to health care. Therefore, mapping all the indicators facilitates an understanding of health needs among small geographical areas, as well as whether the risk of premature death directly increases with socio–economic disadvantage.

168 ______Also included in the mapping are the health services available to each Local Government Area (LGA) within the two Area Health Services: Northern Sydney Central Coast and Greater Western selected as case studies. Health care facilities (private and public hospitals and community health centres), including GP services, were mapped for each LGA level. This level of analysis was conducted to highlight the LGAs level of socioeconomic disadvantage (as defined by the SEIFA–IRSD index), health needs (defined by premature mortality and chronic morbidity), including the share of resources and health service distribution, to determine whether the current system of resource allocation promoted equity. The mapping of health facilities also provided another indicator of population disadvantage in terms of access to services.

Finally, the usefulness of this study lies in part in its potential to alert policy– and decision– makers about the views of health system officials and key stakeholders regarding what they perceive about resource allocation processes, equity and factors that should influence equitable resource allocation in NSW. The mapping of health services and health needs and socio–economic status also shed some light on the importance and effectiveness of simple methodologies that can influence service delivery and distribution of resources to smaller geographical areas.

5.6 Methodological Limitations The first limitation involved the use of the mixed–method approach to study a complex research issue with two independent methods. The use of qualitative and quantitative methods was considered since it has the potential to explore any issue comprehensively. However, it requires greater effort, training, time and good understanding of the two methods to adequately study a single phenomenon within the specified time frame. Although the researcher had undergone extensive training in both methods at the London School of Hygiene and Tropical Medicine in a Masters in Public Health course, the knowledge and skills gained were not exhaustive. The wide–ranging data collection and meticulous analysis of both qualitative and quantitative data were a challenge that the researcher tried to overcome throughout the study by undergoing further training.

169 ______In regard to qualitative data collection, the difficulties encountered ranged from the fact that many respondents did not know how equity was applied within the NSW health care system during resource allocation decision –making due to lack of transparency on how resources are allocated within the system. The majority of those interviewed, mainly senior health officials, could only describe how best health resources should be allocated to achieve equity rather than indicating how they had applied the criterion in their day–to–day resource distribution decision –making. Therefore, most of the data derived from the interviews described the challenges faced by the health professionals and the measures that could be taken to improve the health care system. Due to the range of issues discussed outside the interview guide, the researcher made a decision during the data analysis stage to not restrict emerging themes even though they were outside the issues prepared to guide the interview. All themes relating to the NSW health system derived from the qualitative data have all been discussed in the three chapters presenting the findings (Chapters 6, 7 and 8).

External validity (generalisability) was the next concern of the study. It might be argued that the selection of only two Area Health Services (AHSs) out of the eight in NSW for the qualitative component might have some effects on the validity of the study but, while the argument may be valid, efforts were made to select two different AHSs that were representative of the other remaining six. For example, Northern Sydney Central Coast is an urban AHS that has similar characteristics to other three metropolitan based AHSs. On the other hand, Greater Western, a rural/remote AHS, has similar features to the rest of the rural/remote AHSs.

An important limitation that could not be prevented was the sample selection bias linked to purposive sampling of the two AHSs, since timeliness of ethics approval and accessibility were some of the factors that had to be taken into consideration. However, as explained earlier, both areas were deemed representative of the eight AHSs in terms of generating appropriate ‘information–rich’ cases for in–depth study (Patton 1990: 182).

The validity of quantitative data (premature deaths and chronic morbidity) used in the analysis could not be corroborated for potential under–or over–estimation error due to lack of other data sources. However, the sources, the PHIDU established by the Australian

170 ______Government – Department of Health and Ageing to assist in the development of public health data systems and the Australian Bureau of Statistics (ABS) were deemed reliable and a good source of population health information.

Another limitation encountered in this study was the lack of well –categorised financial data from the NSW Department of Health in terms of the percentage funding that is dedicated to promoting equity. In the NSW Health policy document, it is stated that the main tool (RDF) used to promote equity is only applied to ‘Growth Funding’; however, efforts to find out the actual percentage of growth funding that is allocated through the RDF were fruitless. Therefore, a decision was made to use the total amount of funding allocated to the eight AHSs from the NSW Department of Health labelled ‘initial cash allocations’ in the annual (NSW DoH and AHSs) financial reports, which excludes special grants and capital funds (see Chapter 6). Special grants and capital funding were excluded because they are in most cases a one off payment which does not fall under the category of service provision expenditures as explained by NSW Health (NSW Health, 2007a).

Finally, the financial data obtained from Area Health Services was not categorised according to health facilities or population groups, making it difficult to assess whether there has been movement towards equity at the inter–Area Health Service level. However, this limitation was overcome by mapping out services available and matching them with the identified health needs as defined by the DVIdx to ascertain whether the LGAs with high levels of needs received services commensurate with needs (see Chapter 8).

171 ______Chapter 5: Summary Chapter 5 described the different methods used in this study, including how the study was designed and a detailed approach to addressing research questions captured in a diagram.

A mixed–methods approach consisting of qualitative and quantitative approaches was used in the design of the study to address three questions:

To what extent has resource allocation to Area Health Services from the NSW Department of Health been equitable?

How are the financial resources received from the State level distributed at the Area Health Service level?

Is equity reflected in the allocation of NSW Health resources at the Area Health Service level?

For the qualitative component of this study, the Northern Sydney Central Coast and Greater Western Area Health Services were selected from among the eight Area Health Services in NSW. Area Health Service classifications (urban, rural and remote), and demographic and socio–economic characteristics were used to support the selection of the two AHSs.

The qualitative component to addressing the research questions was tackled by analysing primary and secondary data.

The primary data were collected through one–to–one interviews with 30 participants comprising health policymakers, health executives, health managers and other relevant health officials of the NSW Department of Health and Area Health Services. The participants were purposefully selected. The interviews were conducted from mid 2006 to 2007.

The qualitative data were analysed using content analysis and NVivo 8 software.

The quantitative component to addressing the research questions used two indices constructed in this study. The following is the step–by–step approach undertaken:

 First, health needs across Area Health Services (AHSs) were measured using proxy indicators analysed using principal component analysis (PCA). The output of the PCA was used to develop a general health need index (GHNIdx) for each AHS.

 Second, a double variable index (DVIdx) was developed using premature mortality and morbidity data and was used as proxy for health needs across AHSs.

 The GHNIdx and DVIdx were, in turn, used to derive equity–adjusted shares which were applied as a yardstick against which equity in resource allocation was assessed.

 Actual resources allocated to each Area Health Service from 2003/04 to 2006/07

172 ______were traced and redistributed to determine the proportion of funding differences in the actual allocations and the predicted GHNIdx and DVIdx equity–adjusted shares for each AHS. The predicted equity allocations were compared with the actual allocations to highlight the equity–gap.

 To address research question three, the DVIdx for two AHSs, Northern Sydney Central Coast and Greater Western Local Government Areas (LGAs), were constructed and used as a proxy of health needs to derive equity–adjusted shares for each of the two AHS local government areas. Predicted equity allocations were calculated for each LGA but no comparisons were carried out due to lack of actual funding data since resources within the AHSs are not distributed according to the LGAs.

 Finally, a mapping exercise was conducted showing each LGAs’ socioeconomic status, premature mortality, prevalence of morbidity and health services available to each LGA within the two AHSs, including health resources as determined by the DVIdx. This exercise was conducted to establish whether equity was reflected in the allocation of health services due to lack of funding data.

 Limitations encountered in this study were also discussed

173 ______

Chapter 6: EQUITY IN RESOURCE ALLOCATION AT THE STATE LEVEL: QUALITATIVE AND QUANTITATIVE RESULTS

“…NSW health system needs specific policy to guide equity at each level… the policy should be formulated, implemented and evaluated by a team of experts whose responsibility would be to monitor progress as well as collect data and consult with communities to ensure uniformity of how equity is promoted at the various levels of the system…” (Policy maker, NSW Health)

Overview This chapter presents the qualitative and quantitative results relating to equity in resource allocation at the State level of the NSW health care system. It specifically addresses the first research question: to what extent has resource allocation to Area Health Services from the NSW Department of health been equitable? The chapter is divided into two sections: the first section presents the findings of the qualitative component which explored the perspectives on resource allocation and equity of policy makers, Area Health Service (AHS) executives, program managers and other state health officials involved in the allocation of financial resources to AHSs. The second section presents the quantitative component that examined the extent to which resource allocation at the state level reflects the level of health needs of Area Health Services.

Section 1: Qualitative Findings

6.1 Introduction: Perspectives on Resource Allocation and Equity As indicated earlier, the allocation of financial resources to Area Health Services (AHSs) from the state level has been described as equitable by the NSW Department of Health (see Chapter 4). As part of the overall attempt to assess whether, and the degree to which, resource allocation has indeed been equitable (i.e. to answer the first research question), the perspectives on resource allocation and equity of health policy makers, executives and program managers from two AHSs and the NSW DoH were explored (see Chapter 5 for further details on selection of AHSs). The qualitative phase was done ahead of the quantitative analysis to gain insights into what specifically policy makers and AHS authorities meant by ‘equity’, how much consideration they give to it in practice when making resource allocation decisions, and whether the view that funding allocations to

174 ______AHSs based on the Resource Distribution Formula (RDF) have been equitable is shared by both the State and AHS officials.

Findings from this qualitative investigation were meant to complement the quantitative results so that a comprehensive picture of the degree of equity in funding allocation at the NSW State level could be formed. Seven main themes emerged from the qualitative data: 1) equity and how it should be advanced in resource allocation; 2) attempts to balance funding between hospitals and community health; 3) adequacy of the RDF in promoting equitable resource distribution; 4) the political influence of resource allocation; 5) applicability and adequacy of health system funding; 6) amalgamation of AHSs and equity; and 7) consultations on resource allocation. These themes were identified using NVivo qualitative data analysis software.

6.1.1 Equity and how it should be Advanced in Resource Allocation The reduction of inequities in funding to Area Health Services has been a key objective of the NSW DoH in the past decade. According to Department publications, the most important aspect is not just about promoting equity in the distribution of financial resources to the AHSs but also ensuring that resources are allocated within AHSs to promote equity (NSW Health, 2007a, NSW Health, 2008). Policy makers and health executives interviewed expressed divergent views about the meaning of the term “equity” and how it should be promoted within the context of resource allocation.

“Equity is about fairness in resource allocation…meaning health dollars have to be divided to meet the needs of all population mostly those known to be disadvantaged” (Senior Health Official, NSW DoH)

“Equity means that we should provide equal access to essential health services for the entire population so distribution of health dollars should reflect that objective” (Senior Health Official, NSW DoH)

“Equity is about ensuring people have equal access to comparable services in their local communities” (Senior Health Official, NSW DoH)

175 ______

According to Area Health Services health officials: “Equity is about reducing health inequalities and what that means is that those with poor health outcomes have to be prioritised with health dollars to reduce the health inequalities…” (Senior Health Official, AHS)

“Equity means reducing inequalities in access to services by distributing resources to improve access to those experiencing difficulties in the communities” (Senior Health Official, AHS)

“My understanding of equity is to do with tackling health inequalities in health outcomes which means available health funds have to be distributed accordingly to improve health outcomes” (Senior Health Official, AHS)

“I guess what equity means is that in an ideal world we should work with our local population or communities to identify what they see as their health needs and then in the same ideal world again look at what we can do about those health needs and in economic terms usually you have to make some choices because there is not enough funding or resources to address all the needs” (Senior Health Official, AHS)

As explained by a senior health official at the Department of Health (DoH), for resource allocation to be equitable, efforts should be put into funding a ‘universal health system’ which will guarantee a certain basic minimum of services for the general population as follows;

“The best way of allocating health resources in order to be deemed equitable is to fund a universal health system which is the one–size–fits–all to ensure every one is getting access to a minimum level of care and then target the rest of funding to those who are disadvantaged and have greater health needs...”

The policy maker went on; “…effectively what should happen is that we should have a dual target, that we want every one’s health to improve and then we want the gap between those with good health and poor health to diminish over time. The health gradient is such that socially and economically advantaged people generally have a very good health status and those socially disadvantaged very poor health outcomes…” (Senior Health Official, NSW DOH).

176 ______The senior health official, however, conceded that targeting alone will not improve equity unless efforts are made to provide a universal health care system for the entire population and additional resources are invested in targeting disadvantaged groups who usually have much poorer health status compared to the general population; this could be interpreted as emphasising ‘vertical equity’ (the unequal treatment of unequal’s) (see detailed discussion of vertical equity in Chapter 3). The idea of differentially targeting disadvantaged groups with additional resources was shared by other respondents from both the state and AHS levels. The majority of them thought that there was a need to target disadvantaged groups if equity was to be promoted with any seriousness. In terms of how health funds should be allocated in practice, to promote equity there was consensus among respondents [both at the State and AHS levels] that pathways or mechanisms to guide resource allocation so that disadvantaged populations could be effectively targeted were required. However, no such agreement was discernible from the responses from the two levels on what model of resource allocation might achieve that. Several State level respondents thought the current resource allocation mechanism (RDF) used by the Health Department was capable of shifting resources in favour of disadvantaged groups. This was, however, disputed by respondents from the AHSs, who largely expressed doubts about the RDF and its ability to shift resources to promote equity.

A Director at the AHS who disagreed with the view that the current mechanism for resource allocation is capable of distributing resources to areas of greatest need argued that the pathways or mechanisms for resource allocation should be guided by ‘acceptable factors within society’ and that priority setting should ensure that resources are allocated to areas where they are most likely to benefit those in greatest need;

“Equitable distribution of resources should describe a model of resource distribution based upon society acceptable factors and health determinants that will ensure redistribution of resources to where they will do the most good…” (Senior Health Official, AHS).

Another important issue related to the current resource distribution mechanism that was highlighted by respondents was the extent to which the formula is updated with current data to ensure that the distribution of resources is proportionate to the level of health needs.

177 ______Several of them raised concerns about the lack of frequent updating of the RDF with current data in order to capture possible changes in health needs;

“To some extent the RDF takes into consideration some factors like Aboriginality, education and occupation as they are known to impact on population health but I don’t know how often they update the data since population and their needs changes over time and it’s important to keep track of the changes in order to distribute resources according to those needs…” (Senior Health Official, NSW Health)

In terms of equity, there was widespread awareness of its importance as a state–wide health policy objective and the need to consider it during resource allocation. At both the State and AHS levels, most respondents were quick to point out the emphasis placed on equity in policy statements and on the NSW Health Department’s website. However, several respondents added that insufficient funding for health activities is an obstacle to promoting equity;

“There is no doubt that equity should be considered at all levels of health system…it is part of the values and principles that underpin operation of the NSW Health Department both at state and AHS levels. The equity policy statement is there in black and white at the NSW Health website and in key policy documents. This confirms that equity is an issue for NSW Health. However, health dollars we receive are inadequate and so it’s very difficult for us to seriously consider equity when allocating funds” (Senior Health Official, AHS)

The general recognition of the importance of equity as a policy objective and a factor to consider when allocating resources did not match the gap in actual consideration of equity when allocating resources. Most respondents could not describe when they had considered equity objective as a criteria or factor for resource allocation even with the explicit equity statement in the NSW Health policy document [“In All Fairness. Increasing equity in heath across NSW” (NSW Health, 2004a)]. This raises several concerns regarding the interpretation of the equity statement and the capacity of health officials to translate this statement into a resource allocation factor. Furthermore, the issue of insufficient funding and equity suggests that equity in resource allocation cannot be promoted in its own right, irrespective of how much funding is available.

178 ______“Equity is all about achieving health outcomes and not just access to services like imply in our documents (NSW Health), therefore to be equitable we need to put some positive discrimination by weighting the population with greatest health needs higher than the better off …and I think we have done that, to some extent, with the way we give the Aboriginal population that additional weighting in the RDF so they can receive extra resources”(Senior Health Official, NSW DoH)

As this health executive points out, in the NSW Health resource distribution formula (RDF) extra weighting (implying vertical equity as discussed earlier) is applied to Aboriginal population since they are known to have the worst health outcomes. However, although the extra weighting is applied at the central level through the RDF to ensure AHS authorities receive more funding for Aboriginal health programs and services, it is not compulsory for the AHSs to allocate the same amount of money to aboriginal health when they receive their allocations from the State. According to health officials at the Health Department, AHSs are self–governing due to the system of decentralised system15 and therefore the DoH cannot reinforce similar policies implemented at the State level.

“It’s not compulsory for AHSs to allocate same or more money to Aboriginal health. We [NSW Health] are not prescriptive because of decentralisation…the extra weighting for aboriginality is meant to encourage Areas to spend more on Aboriginal health…it’s only an encouragement but areas don’t do it. (Senior Health Official, NSW DoH)

Due to the implications of this, one policy maker proposed that a ‘team of experts’ should be appointed and given the responsibility to formulate equity–focussed policies relevant for each NSW health sector level and to assist the relevant health service personnel in implementation and monitoring of progress each year. The appointed team should also be in charge of consultations with key stakeholders and collection of relevant data to ensure resource distribution matches the needs of the population;

“NSW health system needs specific policy to guide equity at each system level… the policy should be formulated, implemented and evaluated by a team of experts whose responsibility would be to monitor progress as well as collect data and

15 Area Health Services are empowered to redistribute the funds they receive from the state as they deem necessary based on the needs of their populations NSW Health (2008) NSW Episode Funding Policy 2008/09: Inter–Governmental & Funding Strategies Branch, New South Wales Department of Health, Sydney..

179 ______consult communities to ensure uniformity of how equity is promoted at the various levels of the system” (Senior Health Official, NSW Health)

According to the health executive extract below, the lack of community consultation as part of a process of assessing health needs and priority setting is a weakness that might be contributing to inadequate movement towards equity;

“…I think where we [Area Health Services] go wrong with equity is we do not consult communities when identifying health needs and in priority setting to make sure we know the actual health needs and combine that with other indicators of need and then distribute health dollars according to those needs…” (Senior Health Official, NSW Health)

Indeed, as explained by the health official, a mixture of subjective views of population needs and objective indicators needs to be considered when deriving health needs to ensure equitable allocation of resources. It is also necessary that sufficient funding be allocated to deal with these issues through the relevant sectors.

“An ideal system would combine opinions of communities about their health needs with current health data to support the communities views and to capture health needs at all levels of the health system…that’s how we can promote equitable distribution of health funds… however this has to be supported with sufficient funding…communities consultations have to take place including consulting other sectors dealing with social determinants of health and those at the frontline of providing services to the communities…” (Senior Health Official, Area Health Services)

Equally, strong partnerships will need to be formed between the relevant sectors, including development of policies that address the social and economic determinants of health that are known to impact on health outcomes. What will facilitate the successful collaboration, as explained by the health executive, is availability of current data and therefore effective systems for collating data will be necessary.

“…better data collection systems needs to be developed which also requires funding…” (Senior Health Official, NSW Health)

Clearly, having an equity policy document is not enough evidence that equity is being pursued, especially if the resources being allocated are deemed as insufficient to meet

180 ______equity objectives, as well as other urgent priorities. Furthermore, there is a need for clearly defined, equity–focused resource allocation policies relevant to each health system level. The State (DoH) level equity objective would be more concerned with the geographical disparities across AHSs, while at the AHSs the aim would be population groups, especially the worst off as explained by the health official;

“Different levels of health system needs specific equity objectives…state level is dealing with geographical equity so cannot be the same as AHSs level which is population needs…I think it will be necessary to ring–fence the money and set up KPIs [key performance indicators] that there is a very clear expectation that’s how it would be used but if we did that and take it to its logical conclusion that we label every bit of funding for particular use we will not be allowing them to do what they want according to their populations needs. We would like them to maximise the return on every dollar spent… (Senior Health Official, NSW DoH)

6.1.2 Attempts to Balance Funding between Hospitals and Community Health The interviews revealed the challenges being faced by policy makers when it comes to allocation of funding between hospital and community health centres. Historically, NSW Health has been allocating over 70% of its funding to the hospitals. The majority of the policy makers interviewed expressed frustration about the lack of pathways to assist them in striking a good balance between funding hospitals and funding community –based health services. They were of the view that there was a need to reduce the funding distributed to hospitals and instead divert it to the community health centres in order to ease the pressure placed on hospitals by the health demands of the aging population and increases in the number of chronic disease sufferers;

“…putting ‘all’ resources towards hospitals might not necessary meet equity objective but efficiency…the hospitals keeps us travelling down the same track that will be to our disadvantage. The challenges brought about by the aging population and increasing chronic diseases require constant management within the community which can not be improved by increasing investments in hospitals… (Senior Health Official, NSW Health)

According to some respondents, the uneven distribution of resources between hospitals and community –based services might have exacerbated the inequities that exist within the NSW health care system. As explained by most of those interviewed, equity cannot be

181 ______achieved by allocating more resources to hospitals and less to community –based health care, a view also supported by emerging evidence about the best ways to deal with increasing inequities. Globally, there is a consensus that investing more in primary health community –based services is likely to produce better population health outcomes compared to investing more in hospital –based services. Although executives at the AHS (population) level have the flexibility as to how they distribute resources they receive from the State Department of Health, often funding is allocated to hospitals as first priority and whatever is left over goes to health programs.

“…we need to find a way to divert more funding to the community health centres closer to the communities…until we find a way of taking funding away from the hospitals and allocating to community based services, we can’t talk about achieving equity or tackling aging population and chronic diseases challenges…” (Senior Health Official, NSW Health)

At the Area Health Service level, there were similar critical views about the overwhelming emphasis on hospital funding. Program managers complained about the level of funding for non–clinical and non–hospital based services. They indicated that community health services in particular have traditionally not received high priority in resource allocation in NSW partly because outcomes are long term and hard to measure compared with hospital based clinical services, services which are also viewed as revenue generators.

“…health resources are not equitably allocated unfortunately in NSW…community health services have been historically under–funded and because they are not an income earner they don’t have a high priority in the resource allocation decisions…it’s all about the hospitals and response to hospital crisis” (Program Manager, Area Health Services)

While the funding imbalance between hospital and community –based services was acknowledged at both the DoH and AHS levels, there were no clear suggestions as to how it could be resolved. Instead, some degree of ‘blame–passing’ was observed with several respondents from the State level indicating that the AHSs were encouraged to shift some resources to community based programs and services:

“We [NSW Health] encourage managers [at AHS] to develop a more sophisticated understanding of targeting funding and services to community

182 ______based services in order to promote equity but they always argue that the funding allocated for that is not sufficient... (Senior Health Official, DoH)

“Most funding from the Department is allocated to hospital based programs making it very difficult for us [AHS] to take money for example from ICU and allocate it to community…although the community based centres are very important not just because of the quality of life they add to patients by treating them closer to where they live, they can also get in early and prevent illnesses and reduce the demand placed on hospitals…I think extra funding is required mainly for community based programs and services” (Senior Health Official, AHS)

Most of those interviewed did articulate the importance of hospital services but they also emphasised the need for a more comprehensive community –based health services program which would definitely divert the pressure placed on the public hospitals by patients who could have been successfully treated or managed at the community health centres.

Some respondents had some ideas regarding the measures that should be taken to improve the distribution of resources between the hospitals and community health care. However, it was acknowledged that, due to the expensive and less equitable nature of the current system operating in NSW, change would require detailed planning, extra resources, community consultation and political will to shift those resources in order to revamp the weak community health system, while still managing the hospitals and maintaining an effective service.

“…what we [NSW Health] should try and do is ease the pressure on the system as best as we can through doing modest engagement in the sector, distributing resources we have effectively and equitably as well as being more accountable, and at the same time look for other sources of funding where we can find them for example Commonwealth government – although this always turns into a tug of war when we start talking about more resources… and just hope gradually over time that if we are effective in our investment and the population becomes more healthy and slightly less needy of the public hospital services. We [NSW Health] can gradually redirect some of the funding and other resources into community based primary care. But it is a hard road… It will not work except through very careful management and community engagement which is a subject of detailed planning which the system lacks at the moment… we also need less politics…” (Senior Health Official, NSW DoH)

183 ______

As expressed by the respondents, the current system of allocating resources is complex and would benefit from reforms that would ensure hospital funding continues to be allocated while mechanisms are sought to build and strengthen the community –based health services. This means that any resource allocation reforms would have to be clear in order for the hospitals to continue operating while investing in community health/primary health care. The current health budget might not be sufficient to facilitate improvements in both hospitals and community health centres and so alternative sources of funding would be needed.

6.1.3 Applicability and Adequacy of the RDF in Promoting Equitable Resource Distribution Area Health Services (AHSs) officials were largely critical about funding allocation from the State level. The majority of the respondents expressed doubts about whether the RDF was actually used by NSW Health to guide resource allocation as claimed. The lack of openness about how the RDF is used left health officials guessing at the extent to which the formula is used. They speculated that allocation of resources might still be on a ‘historical’ basis, explaining why disadvantaged areas and some programs are not getting adequate funding.

“Resource allocation in NSW is historically based...I don’t think the funding formula [RDF] is used…resource poor AHSs are still poor…health dollars are not spent very wisely on prevention programs…rather they [NSW Health] try to fix problems in hospitals as they occur” (Senior Health Official, AHS)

The issue of historically over–funded AHSs continues to plague the NSW health system since there was consensus among the participants that some urban areas still receive more than their fair share of funding, while the rural and remote AHSs are under–funded, implying that equity has not really been achieved as intended by the RDF. Some respondents felt that the advocates of the RDF are under the impression that the mechanism has indeed achieved one of its objectives of equitable allocation.

184 ______“Now those behind RDF believe that some equity has been achieved but many think the RDF might have worsened things by widening the inequities since areas that were receiving more funding are still receiving more and those under–funded are still receiving less…(Senior Health Official, NSW DoH)

RDF is used to distribute only a very small proportion of the growth funding… the bulk of the health budget is historically distributed to hospitals through episode funding and so I don’t think the RDF will ever be seen as an effective tool when it is barely used...” (Senior Health Official, NSW DoH)

According to other respondents, the RDF is only applied to a small proportion of the “Growth Funding” which was introduced in 2000 by the then Health Minister. The purpose of the growth funding was to try and reduce the inequities that existed between AHSs, although some areas have not had a significant increase in their share of resources. However, according to the informants, the policy of targeting the most disadvantaged AHSs with growth funding was gradually benefiting those worse–off AHSs even though the funds were also allocated to the better–off AHSs, an approach that might continue to widen the funding inequity gap.

“… RDF works as a guide for allocating existing resources. So if area A is regarded as relatively under–funded and as deserving of additional resources then the over–funded area should be allocated less funding...Essentially, this means taking money away from the over–funded and giving it to the under– funded Areas. This is where there is most tension because it’s something that we [NSW Health] have not been able to do since we can’t take money from those historically over–funded and give to those under–funded due to political reasons. So Growth Funding was introduced to help some areas move towards receiving a fair share of resources although even that growth funding is not entirely allocated to those under–funded areas but to all the eight Areas which is all political and this has continued to cause problems for Areas and health system as a whole… (Senior Health Official, NSW DoH)

As this health executive explained, the RDF is meant to guide resources in a way that ensures an Area’s population health needs are commensurate with the level of resources allocated. This would mean that historically under–funded AHSs would receive a fair share of resources that reflect their level of health needs. For example;

“Now if you look [referring to Table 6–], at the difference between those share of resources and the RDF target share which is on a percentage point basis you

185 ______would think that the areas are really quite close to achieving a fair share… like the smallest differences here is 0.6% however when you compare on an amount basis it’s much bigger for all the under–funded areas. So really you can say equity is yet to be achieved by the NSW Health…” (Senior Health Official, NSW DoH)

Table 6–1: NSW Health Actual Share of Resources and RDF Target Shares, 1989– 2008

Source: NSW Health RDF Technical Paper 2005 Revision

Some of those interviewed were of the view that if the RDF was being effectively applied as per the guidelines, then most AHSs would be receiving a fair share of resources based on the RDF target shares. But the question remains, would that RDF target share be commensurate with the AHSs health needs?

“…the RDF has been made a political tool rather than an effective mechanism for distributing resources for example, the current version of the RDF booklet was updated in 2001 when population census data came out but due to external and internal politics the booklet took four years to be published…(Senior Health Official, NSW DoH)

As well as difficulties in updating the funding formula with current population data, respondents related how a change in the Minister for Health or the Department of Health Chief Executive can have profound effects on the applicability of RDF because of what they referred to as fear of ‘political implications’.

“When there is a change in Minister for Health and Chief Executive for NSW Health people inside the department seem to be afraid of political implications that the RDF might have whereas it should not really be perceived as a political tool as such because it is actually a good news story politically since its

186 ______something which can be seen to reduce inequities if used properly to distribute resources to Areas. But somehow it got caught up into all the politics plus internal politics where many people think that it should be abolished but basically there should be a version updated with current population data that comes out every year…” (Senior Health official, NSW DoH)

Others perceived the RDF as a non–existent mechanism since the bulk of the NSW Health funding is allocated mainly to hospitals, using casemix or episode funding policy rather than by using population and needs–based formulas. Furthermore, the tools used to estimate the population health needs are described as being crude, indicating a need for more up–to– date data that can be used in the development of better resource allocation models that truly capture the population health needs.

“I think our health care system [NSW Health] is far from being equitable even with the use of resource distribution formula [RDF] to distribute funding. We have not seen evidence of the use of the formula and wonder if the RDF is really used at all since the funding received is all about episodes of hospital care and not really reflecting our population needs. They [NSW Health] try to define population needs based on very crude parameters (standardised mortality ratios) and crude performance measures. They [NSW Health] should try to measure population needs through different population data collection and then to distribute resources to population where it is perceived as having greatest need…” (Senior Health Official, AHS)

“…what we have been made to understand is that the RDF is all about population needs and equitable distribution of health dollars but for some reason most areas still complain the dollars they receive are not enough to meet the needs of their population…other areas have not been able to pay the bills because the funds are not sufficient. Therefore if the RDF was used to guide funds according to the estimated population health needs then each area should be getting funds to meet those needs as well as achieve equity…this is where I doubt the use of RDF and its ability to estimate the health needs of each area or to address equity because we are far from receiving equitable funds...there needs to be an active formula that is not politicised like RDF that will distribute resources according to population health needs…” (Senior Health Official, AHS)

According to respondents, even after investing large amounts of funding in improving Indigenous population health outcomes, their health still remains comparable to that of developing countries. While better funding mechanisms are needed to capture the

187 ______disadvantaged population health needs and to distribute resources according to those needs, a strong multi–sectoral partnership is also needed.

“…any funding formula needs to take into account the determinants of health as evidence shows…For example, billions of dollars have been poured into Aboriginal population health in NSW if not Australia with little improvement. In fact the health status of Aboriginal population is often compared to that of third world countries. This is because we are focusing health issues alone and not the other underlying determinants of health. So in that sense NSW acknowledges all of this but a pathway and a simple formula is required to effectively capture the health needs of Aboriginal and other disadvantaged population and to allocate health dollars accordingly while collaborating with other sectors to resolve those factors outside health and Department of Health…” (Senior Health Official, AHS)

6.1.4 The Political Influence on Resource Allocation The issue of politics in resource allocation also emerged strongly in the interviews. Many participants from both the State and AHS levels expressed the view that there was a lack of political will to shift resources to where the needs are greatest in order to enhance equity.

“The NSW health budget is not sufficient to meet most of the population health needs and the lack of political willingness to gradually reduce hospital funding and put more funding into primary health care or community health worsen the situation and we [NSW Health] will continue to face crisis. I think we are likely to see more crisis until health funding is de–politicised…” (Senior Health Official, NSW DoH)

According to health officials, the politics of health care funding is exacerbated by the shared responsibility between the Commonwealth and States/Territories and Local governments in Australia and specifically NSW. The funding arrangement was strongly criticised by health officials at all levels. Several respondents indicated that there is significant wastage of resources through duplication of services as a result of the politics between the Commonwealth and the State Governments concerning who has the responsibility to fund different components of health care. This problem was believed to be contributing significantly to service duplication and wastage of resources that could be effectively used to provide services that are effectual. Preferably, they felt, health systems should be funded by one government.

188 ______

“….there is a lot of funding being wasted through the Commonwealth and State Governments funding a little bit here and a little bit there of similar services yet not sufficient to meet the population needs … ideally funding should come from one source and should be more targeted to programs that are known to be effective and to benefit the population…apparently it is estimated that around 2 billion health dollars are wasted due to this historical arrangement…” (Senior Health Official, AHS)

Another issue that emerged from the interviews was the distribution of health resources across and within the AHSs being heavily influenced by vocal, mostly urban populations and by political and other vested interests. Most of the respondents thought that equity or population needs had never been a factor in resource allocation but rather a political issue that was hidden behind the veil of a formula. Using resources to respond to crisis was another common issue cited by respondents.

“…politics have been at the forefront of health funding and its distribution for as long as I can remember… often it’s about the vocal politicians e.g. if they want a new hospital built in their electorate…they make sure they get it even if it means other populations go without services…politicians use health care as a campaigning tool and that is all that matters…the policy makers are left to deal with the consequences which range from either patients are left running mad on the streets or falling and breaking bones due to lack of services or some hospitals are left in poor state not able to provide care due to lack of long term financial sustainability…(Senior Health Official, AHS)

Respondents also described how politics influences resource distribution in NSW, leading to under–utilisation of the RDF, which is technically meant to promote equity by ensuring AHSs are receiving a fair share of resources commensurate with their population health needs.

“… equity is not as important compared to issues motivated by politics as far as the NSW government are concerned…but the what they [government and NSW Health] need to realise if equity is not promoted areas will remain under– resourced and if the areas happen to drift further away from their RDF target then people dealing with the formula are forced to explain why the RDF is not working. This is all because funding has not been allocated according to the RDF principles but according to what sounds good for the Government decision makers…but if they continue responding to crisis only and not making use of the

189 ______RDF fully then we will continue to see more under–funded Areas and increasing inequities…” (Senior Health Official, NSW DoH).

6.1.5 Adequacy of Health System Funding The policy makers at the operational (population) level were critical of and frustrated with the under–funding experienced by some of the AHSs, particularly those in rural and remote areas. A similar concern was expressed by the AHS directors and program managers. They explained that what they receive from NSW Health was insufficient to meet their population health care demands and to facilitate achievement of equity by differentially targeting the worse–off with services. Even though most respondents had different ideas of what equity entailed, including its definition, there was a consensus that the health system needed a better mechanism and political willingness to restructure the way health resources are currently distributed to AHSs and specifically to the most disadvantaged areas, including population groups.

“I guess what equity means is that in an ideal world we should work with our local population or communities to identify what they see as their health needs and then in the same ideal world again look at what we can do about those health needs and in economic terms usually you have to make some choices because there is not enough funding or resources to address all the needs. We should also work with those communities to make those choices but in reality what we are faced with is how can we try and identify the population health needs with the little resources that we have…it’s difficult. Addressing equity issues needs more funding from NSW Health…with more funds it’s possible to address the specific health needs in order to reduce inequities that currently exist in our area…” (Senior Health Official, AHS)

As expressed by this health official, the NSW health system resource allocation processes are politically influenced and have led to the existing inequities. Therefore, the system as a whole requires meticulous planning and political willingness for the changes to effectively take place.

“NSW Health and government as a whole need to act on the health system funding and political system that has resulted in inequities swiftly…” (Senior Health Official, NSW DoH)

190 ______6.1.6 Amalgamation of Area Health Services and Equity The reduction of the Areas from seventeen to eight Area Health Services (AHSs) was another issue that was strongly criticised by many respondents; they saw it as a poor reform that has contributed to increasing existing inequities and health service fragmentation. Some respondents from the AHSs were of the view that the amalgamation of AHSs had increased the cost of delivering health services, including travel costs, and as a result moved health services further away from communities.

“There is no doubt that amalgamation of the AHSs has brought a whole range of challenges. It has made things more complicated by moving health services and management further away from people we serve. There is the issue of extra cost of service delivery with health workers having to travel far and wide to deliver services something which could have been avoided by making areas much smaller” (Senior Health Official, AHS)

“Funding to most programs and services has become more fragmented due to amalgamation Areas …furthermore it is not clear how the inadequate funding received should work to address the historical inequities now areas are even larger …” (Senior Health Official, AHS)

As explained by the respondents, some AHSs have higher needs among their populations and therefore deserve more resources that are commensurate with those needs. According to the respondents, the population needs may have been exacerbated by the amalgamation of areas into larger entities, making it more costly to provide services to the population due to the long distances health professionals have to travel.

“…funding to this AHS [GWAHS] is very small considering the high level of needs…funding allocation does not take into account the actual needs of the population and the difficulties we are going through when delivering services…after amalgamation the area is too large and we require more resources to enable us travel to provide services…travelling takes a lot of resources…also having management teams based for example in Dubbo and travelling to Broken hill cost substantial amounts including time wasted travelling…amalgamation has made things worse…” (Senior Health Official, AHS)

191 ______According to respondents, health services are more fragmented as a result of the amalgamation forcing patients to travel longer distances to access essential services. The Greater Western Area Health Service (GWAHS) was cited as a good example by several respondents: the GWAHS has a landmass of over 400,000 square kilometres and an 0.3 million population. The population is widely dispersed, thus impacting on the amount of resources (financial and manpower) that are required to provide health services. This scenario, many respondents thought, could have been avoided if there had not been any amalgamation.

However, according to NSW Health officials, amalgamation was necessary to strengthen AHSs that were experiencing a lack of comprehensive hospital facilities and a reduced capacity to utilise resources due to manpower shortages.

“Amalgamation of Areas from 17 down to 8 was necessary in NSW health care system to facilitate Areas without large comprehensive hospitals within their boundary merge with another to improve their capacity to provide services. Area experiencing staff shortages were expected to benefit from the merge plus reduce costs of providing health care through consolidating some of the duplicated services but we are yet to conduct an evaluation to see if the reform has achieved its intended purpose” (Senior Health Official, NSW DoH)

6.1.7 Limited Consultations during Resource Allocation Another important issue that emerged from the interviews with regard to resource allocation was inadequate consultations between the State Department of Health and Area Health Services (AHSs), including other key stakeholders. Several frontline health professionals at the AHS level expressed the view that the areas receive inadequate funding because state officials and policy makers dealing with resource allocation rarely consult them. They felt that the lack of consultation had contributed to the financial mismanagement and other problems commonly experienced by AHSs.

“…we are not consulted when resources are being allocated…we are only told what is allocated and the performance targets that we have to meet…there should be consultations with AHSs health officials who are responsible for delivering services with the health dollars allocated. Sometimes we have to cut down the services we provide due to lack of resources either because we can not

192 ______pay a Nurse or to provide the service and when consumers complain it’s us to explain what the problem is while we did not get involved in the resource allocation…another issue is the lack of financial accountability within this Area…” (Senior Health Official, AHS)

“…one would expect NSW Health to consult Areas about the needs of their population and what they perceive as priority in terms of funding before the RDF shares are decided upon but we are never consulted and same goes with most other Areas… to ensure Areas are accountable financially they need to be engaged in decision making so they understand the challenges involved in dividing health budget against many demands rather than push insufficient resources with huge performance targets…” (Senior Health Official, AHS)

Section 1: Summary of Qualitative Findings  This section has provided some insights into the extent to which distribution of resources have been equitable, as perceived by the policy makers, executives and health managers who have the duty to allocate and utilise the available resources. In particular, it has shown that the distribution of health funding to Area Health Services by NSW Health is influenced mainly by a complex set of factors including the following:

 Equity and how it should be advanced in resource allocation,  Attempts to balance funding between hospitals and community health,  Applicability and adequacy of the RDF in promoting equitable resource distribution,  Political influence on resource allocation,  Adequacy of health system funding,  Amalgamation of Area Health Services and equity, and  Limited consultation during resource allocation.

193 ______Section 2: Equity in Resource Allocation at the State Level – the Quantitative Results

6.2 Introduction The results of the quantitative analysis of equity in resource allocation among the Area Health Services (AHSs) in New South Wales (NSW) are presented in this section. As indicated earlier, the level of health needs across the eight AHSs was assessed using two different indices. The first – a general health needs index (GHNIdx) – was developed using multivariate statistics (Principal Component Analysis). The second – a double variable index (DVIdx) – was developed using mortality and morbidity data (see Chapter 5 for further details). Both indices were translated into funding shares and used to redistribute the total funding allocations to the eight AHSs to arrive at the predicted allocations used to assess equity. The assessment of equity was undertaken to compare the predicted allocations, based on the proxy health need indices, with the actual allocations received by the AHSs over a period of four years (2003/04 to 2006/07). The actual allocations are ‘Initial Net Cash Allocations’16 from the NSW DoH to AHSs and are historical figures obtained from financial reports published by the NSW DoH. The decision about the extent of equity in resource allocation was made by analysing the gap between what the AHSs actually received and what they should have received on the basis of health needs. The wider the gap, the more inequitable the allocation was judged to be and vice versa.

In addition to using the two health need indices developed (the GHNIdx and the DVIdx), the actual allocations were also compared with predicted allocations based on the Resource Distribution Formula (RDF) shares published by the NSW Health. As explained in Chapter 5, the purpose of doing this was to ascertain whether indeed the RDF has been used to reduce the inequities in funding distribution to an estimated less than 2 percentage points on average as claimed by NSW Health (see Chapter 1).

16 According to NSW Health (2008), initial net cash allocation reflects the net cost of providing services, excluding capital expenditures.

194 ______6.3 Comparison of Actual Allocations to Area Health Services: 2003/04 to 2006/07 Before analysing the actual and predicted allocations to assess the extent of equity, the actual allocations over the four years were analysed to assess the growth or decline in funding. Table 6–1 shows the amount of funding received by each AHS from NSW Department of Health for the four–year period from 2003/04 to 2006/07.

Table 6–2: Actual Funding Allocations to Area Health Services, 2003/04 to 2006/07

Note: SSW = Sydney South West, SES&I = South Eastern Sydney & Illawarra, SW = Sydney West, NSCC = Northern Sydney Central Coast, H&NE = Hunter & New England, NC = North Coast, GS = Greater Southern and GW = Greater Western. Data Source: NSW Health Annual Reports 2003/04 to 2006/07

As shown in Table 6–2, the percentage growth in funding varied across the eight AHSs with some receiving substantial increments while others received only modest increases in funding in all four years (2003/04 – 2006/07), apart from SES&IAHS that had a decrease in funding between 2003/04 and 2004/05. In 2003/04 to 2004/05, NCAHS was the only AHS to receive over 10% growth in funding, although in the following years (2004/05 – 2005/07), all AHS funding increased by over 11%. In fact the SES&IAHS’s growth in funding was the highest at 22.3% (2004/05 – 2005/06) compared to the previous period when it had a decrease of around –2.2%. In 2005/06 to 2006/07, most AHSs’ growth in funding was reduced to less than 10% apart from NCAHS and GSAHS, which had 10.9% and 10.0% respectively (see Table 6–2).

195 ______6.3.1 Area Health Services Actual Funding per Head of Population: 2003/04 to 2006/07 Figure 6–2 shows the per capita allocation based on the actual resources allocated to each Area Health Services (AHS). GWAHS, a rural and remote classified AHS, had the highest funding allocation per head of population, ranging between $1,200 to $1,700 for the four years 2003/04 to 2006/07, which is a reflection of the population’s health status as illustrated by what NSW Health refers to as ‘potential avoidable deaths’ for persons aged under 75 years amenable to health care in 2006 (see Figure 6–1). As shown in Figure 6–1 the rates per 100,000 population of potentially “preventable” premature mortality in NSW was highest among rural and remote classified AHSs compared to the urban –based AHSs. This did not include SWAHS, which had rates within the same range as the rural and remote AHSs.

Figure 6–1: Area Health Services Potentially Avoidable Premature Deaths from Causes Amenable to Health Care; Persons Aged Under 75 Years in NSW (Rates per 100,000 Population) – 2006

100 90 80 70 60 50 40 30 20 10 0 Number of Avoidable Deaths/100,000 Pop Deaths/100,000 Avoidable of Number

NCAHS GSAHS SSWAHS SWAHS GWAHS SES&IAHS NSCCAHS H&NEAHS Area Health Services

Potentially Premature Deaths Amenable to Health Care (Rates per 100,000 population)

Data Source: NSW Health Chief Health Officers’ Report, 2006

196 ______

Figure 6–2: Area Health Services Funds Allocation per Head of Population: 2003/04 to 2006/07

1,600

1,400

1,200

1,000

800

Amount ($) 600

400

200

0 SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS 2003/04 968 1,062 851 777 970 950 837 1,164 2004/05 1,021 1,038 891 834 1,025 1,049 886 1,234 2005/06 1,178 1,270 1,018 958 1,161 1,192 1,000 1,371 2006/07 1,290 1,383 1,107 1,041 1,276 1,322 1,100 1,500

Data Source: NSW Health Annual Reports (2003–2007)

197 ______Most of the other AHSs allocated less than $1,000 per head of population from 2003/04 to 2004/05, but the amounts increased slightly in the following years (2005/05 to 2006/07). But does the increase in per capita funding reflect the rates of premature deaths? Table 6–3 and Figure 6–3 show that the rates of premature mortality17 for each AHS (persons aged under 75 years) declined significantly from 1987 to 2006, reflecting the increase in health system funding, especially over the period of RDF application, and current understanding of causation and available disease prevention and health care. For example, the gap between the premature death rates in a relatively wealthy AHS (NSCCAHS) and in a less well–off rural/remote AHS (GWAHS) was reduced from 278.9 to 129.8 per 100,000 population (NSCCAHS) and in GWAHS from 417.5 to 193.4 per 100,000 population form 1987 to 2006. The premature death rates in the other four AHSs also declined over the same period, although the gap in the decrease between the AHSs in 1987 and 2006 was not large. The narrowing of the gap in premature rates between the largely metropolitan AHSs: SSWAHS, SWS&IAHS, SWAHS, NSCCAHS and the rural/remote AHSs: H&NEAHS, NCAHS, GSAHS and GWAHS is of particular interest because prior (1987 to 1989) to adoption of the Resource Allocation Formula (now known as the Resource Distribution Formula) funding distribution mechanism, the premature death rates were increasing in some AHSs while others were on a downward trend (see Table 6–3). However, since the RDF was introduced in 1989/90, most AHSs have enjoyed a continuing downward trend, suggesting that the RDF might have had some effect.

The Area Health Services that benefitted most, and hence experienced to a greater extent a larger decrease in premature mortality rates ranging from 211.3 to 224.1, were SSWAHS and GWAHS, while SES&IAHS, SWAHS, NSCCAHS, H&NEAHS, NCAHS and GSAHS had the least (ranging from 149.1 to 197.4) although significant (see Table 6–3 and Figure 6–3).

17 According to NSW Health, potentially avoidable mortality (premature mortality) can be averted by prevention (‘preventable’) or treatment (‘amenable’). Amenable conditions are those for which it is reasonable to expect death to be averted, even after the condition has developed through early detection and effective treatment (such as breast cancer). Preventable conditions are those for which there are effective means of preventing the condition from occurring, for example, where the aetiology is to a considerable extent related to lifestyle factors (such as smoking) (Tobias & Jackson, 2001).

198 ______

Table 6–3: Area Health Services Premature Death Rates per 100,000 Persons Aged Under 75 Years in NSW, 1987–2006

AHSs 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 SSWAHS 362.2 366.1 353.2 326.9 311.1 305.5 285.1 274.6 269.3 258.4 249.7 229.4 236.7 225.8 197.4 195.6 183.7 177.4 164.2 150.9 SES&IAHS 299.4 321.0 317.5 283.8 268.0 256.7 234.9 240.6 232.3 227.6 221.4 204.7 198.0 185.1 174.7 169.0 160.6 159.3 145.7 148.6 SWAHS 351.4 348.0 341.6 309.4 293.0 307.0 277.9 271.7 269.6 249.1 245.7 222.5 228.2 207.7 203.5 200.4 181.3 178.8 165.6 154.0 NSCCAHS 278.9 262.5 257.8 248.1 226.2 231.1 215.8 210.2 205.6 201.7 188.6 182.4 175.9 163.6 157.7 156.8 146.5 141.6 126.4 129.8 H&NEAHS 368.8 344.3 351.3 323.1 294.5 296.5 285.0 274.4 268.5 269.7 253.0 237.0 232.8 234.3 223.5 216.1 194.4 194.3 177.0 171.6 NCAHS 342.9 321.6 322.1 289.0 287.7 279.6 251.8 252.2 244.2 253.3 243.0 228.2 229.5 220.9 213.8 203.2 193.9 180.1 171.5 168.7 GSAHS 350.4 366.0 352.8 319.3 302.1 307.3 280.8 283.7 274.0 262.9 265.5 230.5 250.5 223.4 212.0 220.3 202.8 184.2 181.5 166.0 GWAHS 417.5 384.3 385.4 368.0 344.4 339.0 329.9 328.1 314.7 287.4 273.5 256.5 259.7 258.8 240.3 238.5 229.8 213.0 207.8 193.4

Figure500.0 6–3: Area Health Services Premature Deaths for Persons Aged Under 75 Years in NSW 1987 to 2006 (Rates per 100,000 Populations)

400.0

300.0 ,000 Pop

200.0

100.0

Premature Mortality Rates per 100 0.0

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Years

SSWAHS SES&IAHS SWAHS NSCCAHS H&NEAHS NCAHS GSA HS GW A HS

Data Source: NSW Health Chief Health Officers’ Report, 2006

199 ______

6.3.2 Comparison of NSW Health RDF Target Shares and the Actual Resources Allocated to Area Health Services: 2003/04 to 2006/07 Table 6–4 demonstrates the projected share of funding ‘RDF target shares’ that should have been allocated to the eight AHSs from the NSW Department of Health from 1989/90 to 2007/08. According to the NSW Health policy document ‘the overall inequity within a given year has decreased over time. The weighted average distance from target for the new 8 Area Health Services is around 1.8% as previously under–resourced Area Health Services have been combined with better resourced Areas’ (NSW Health, 2005) (see Chapter 1, Figure 1–3). To determine whether the ‘weighted average distance’ from the RDF target shares is indeed around 1.8% as stated by NSW Health, the actual resources allocated to the eight AHSs from 2003/04 to 2006/07 were analysed and compared (see Tables 6–5 and 6–6).

Table 6–4: NSW Health Resource Distribution Formula Target Shares, 1989–2008

Area Health RDF Target Share 1989–2008 1989/90 – 1994/95 – 1998/99 – Services 1993/94 1997/98 2006/07 2007/08 SSWAHS 18.1% 20.7% 20.7% 20.2% SESIAHS 21.3% 20.0% 19.5% 19.0% SWAHS 14.4% 13.6% 13.9% 15.1% NSCCAHS 14.0% 13.9% 14.4% 13.1% HNEAHS 13.1% 11.7% 11.8% 12.5% NCAHS 6.4% 8.4% 8.1% 7.6% GSAHS 6.7% 6.3% 6.2% 7.3% GWAHS 5.9% 5.4% 5.4% 5.1% Total 100% 100% 100% 100% Source: NSW Health RDF Technical Report 2005 Revision

In Table 6–4, the second to last column (RDF Target Shares 1998/99 – 2006/07) shows the NSW Health RDF estimated shares that each AHS would receive based on the RDF; however, when this is compared to what they actually received from 2003/04 to 2006/07, there are significant differences in funding allocations for most of the AHSs. According to Tables 6–5 and 6–6, five out of eight AHSs were allocated funds that were less than their stated RDF target shares, although the amounts varied considerably. For example,

200 ______NSCCAHS and NCAHS had the largest under–funded amounts while H&NEAHS and SWAHS were the most over–resourced AHSs.

201 ______Table 6–5: Comparison of Area Health Services Actual Funding Allocation and RDF Target Shares, 2003/04 to 2006/07 ($M)

Data Source: NSW Health Annual Reports

Table 6–6: Actual Allocations and RDF–based Target Shares: Average Distance from RDF, 2003/04 to 2006/07 Area Health Services 2003/04 2004/05 2005/06 2006/07 Funding Gap Funding Gap Funding Gap Funding Gap $M (%) $M (%) $M (%) $M (%) SSWAHS –21.7 (–1.7%) –11.5(–0.8%) –16.0 (–1.0%) –14.7 (–0.9%) SESIAHS 11.7 (0.9%) –72.0 (–5.9%) –1.4 (–0.1%) –7.8 (–0.5%) SWAHS 52.1 (5.6%) 56.2 (5.7%) 51.0 (4.6%) 49.5 (4.1%) NSCCAHS –48.2 (–5.6%) –26.3 (–2.8%) –36.6 (–3.4%) –47.4 (–4.1%) H&NEAHS 59.5 (7.4%) 71.9 (8.4%) 62.9 (6.5%) 74.1 (7.0%) NCAHS –63.4 (–14.1%) –39.5 (–7.9%) –55.8 (–9.9%) –52.4 (–8.4%) GSAHS 1.9 (0.5%) 3.2 (0.8%) –7.9 (–1.7%) –5.5 (–1.1%) GWAHS 12.3 (3.5%) 18.0 (4.8%) 3.7 (0.9%) 4.1 (0.9%) Funds for redistribution $M 135.4 149.3 117.6 127.7 As a % of total funds 2.1% 2.3% 1.5% 1.5%

202 ______Table 6–6 shows the difference between what the Resource Distribution Formula (RDF) indicates AHSs should have received as an equitable share of resources compared to what they actually received from NSW Health between 2003/04 and 2006/07. Based on this analysis, NCAHS had the highest under–funded amounts, estimated to be around –$63.4 million in 2003/04, although the amount decreased in 2004/05 but then increased in the following years. The second most under–funded AHS was NSCCAHS with –$48.2 million in 2003/04; however, the funding gap decreased in 2004/05 but increased to –$36.6 million in 2005/06 and to –$47.4 million in the final year (see Table 6–6).

On the other hand, H&NEAHS had the highest over–funded amount ($74.1 million), followed by SWAHS with $49.5 million; the lowest over–funded was GWAHS at around $4.1 million (see Table 6–6). Based on these results, it should be asked whether, or to what extent, was the formula being used to guide the distribution of the health budget to the AHSs as NSW Health asserts? Figure 6–4 shows the percentage point difference of actual funding from the RDF target shares, represented in Tables 6–5 and 6–6 in graphic form. Although the actual funding allocation for the under–resourced AHSs moved closer to their RDF target shares, the funding differences are significant (see Figure 6–4).

Overall, the analysis conducted in this study shows that the amount of funding that was required for redistribution in order to achieve the RDF target shares for the eight Area Health Services (AHSs) was $135.4 million or 2.1% (of total funds) in 2003/04, increasing slightly to $149.3 million (2.3%) in 2004/05, but decreasing significantly to $127.7 million (1.5%) in 2006/07 (see Table 6–6). This suggests that, although there has been movement towards RDF guided equitable resource allocation and significant reductions in premature mortality as noted earlier, inequities in resource distribution still existed between AHSs in NSW in 2006/07. As will be discussed later in this report, one also has to consider what might be regarded as an “inevitable but acceptable” level of inequity, given the many determinants of administrative decisions involved in the actual allocation of resources.

203 ______Figure 6–4: Actual & Resource Distribution Formula (RDF) by Area Health Services: 2003/04 – 2006/07

Actual & RDF-based Allocation by Area Health Services: 2003/04 Actual & RDF-based Allocation by Area Health Services: 2005/06 1,800.0 8.0% 1,400.0 7.4% 5.6% 10.0% 6% 1,600.0 5% 6.0% 1,200.0 4.0% 0.9% 5.0% 1,400.0 1,000.0 -1.7% 3.5% 2.0% 1,200.0 0.0% 0% -5.6% -1% 0.0% 800.0 -0.5% 1,000.0 -3% 1% -5.0% -2.0% 800.0 600.0 -2% -4.0% Amount ($'M) -14.1% Amount ($'M) -10.0% 600.0 400.0 -6.0%

Percentage difference Percentage -10% 400.0 -8.0% difference Percentage 200.0 -15.0% 200.0 -10.0%

0.0 -20.0% 0.0 -12.0% SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS Area Health Services Area Health Services

Actual allocation 2003/04 RDF-Based allocation 2003/04 Pe rce ntage diffe re nce Actual allocation 2005/06 RDF-Based allocation 2005/06 Pe rce ntage diffe re nce

Actual & RDF-based Allocation by Area Health Services: 2004/05 Actual & RDF-based Allocation by Area Health Services: 2006/07

1,600.0 8.4% 10.0% 7.0% 2,000.0 8.0% 5.7% 1,400.0 8.0% 1,800.0 4.1% 6.0% 6.0% 1,600.0 1,200.0 4.0% 4.8% 4.0% 1,400.0 1,000.0 2.0% -0.9% -0.5% -0.8% 2.0% 1,200.0 -2.8% 0.9% 0.0% 800.0 0.8% 0.0% 1,000.0 -4.1% -1.1% -2.0% -2.0% Amount ($'M) 600.0 800.0 Amount ($'M) -5.9% -4.0% -7.9% -4.0% 600.0 400.0 difference Percentage -8.4% Percentage difference Percentage -6.0% 400.0 -6.0% 200.0 -8.0% 200.0 -8.0%

0.0 -10.0% 0.0 -10.0%

SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS Area Health Services Area Health Services

Actual allocation 2004/05 RDF-Based allocation 2004/05 Pe rce ntage diffe re nce Actual allocation 2006/07 RDF-Based allocation 2006/07 Pe rce ntage diffe re nce

204 ______

6.4 Inter–Area Health Service Level Equity in Resource Allocation The results of the use of two indices – the general health need index (GHNIdx) and the double–variable index (DVIdx) – are presented in this section. As discussed in Chapter 5, the GHNIdx and DVIdx were developed in this study as yardsticks to assess equity in resource allocation across the eight Area Health Services (AHSs). The results obtained from both the GHNIdx and DVIdx were used to develop an equity–adjusted share of NSW Health allocated funds. The GHNIdx describes the relative level of health needs within the AHSs, while the DVIdx “index” reflects the ‘burden of disease’ actually borne in the AHSs. The assumption is that the AHSs with a heavy burden of disease will require more resources than those with a lesser burden. To also quantify the strength of each index in deriving shares of funding to meet the AHSs’ level of health needs, the DVIdx was used in this study to corroborate the two multi–variable indices.

Using the two constructed GHN and DV indices, an analysis was undertaken to compare the equity shares and the actual allocations in order to identify funding differences that are referred to as ‘equity gaps’ in this thesis. This analysis was conducted using the funding data allocated to the AHSs from the NSW Department of Health from 2003/04 to 2006/07.

6.4.1 General Health Need Index (GHNIdx) As explained in chapter 5, principal component analysis (PCA) was used to produce the component score coefficients; these coefficients were then used as variable weights in developing the general health need index (GHNIdx). The general health need index for each AHS was derived by summing up each variable’s standardised score (z–score), multiplied by its component score coefficient or weights.

205 ______Table 6–7: AHSs General Health Need Index Principal Component Analysis Area Health Services GHNIdx Sydney South West –1.14 South Eastern Sydney & Illawarra –0.18 Sydney West –0.87 Northern Sydney Central Coast –0.52 Hunter & New England 0.26 North Coast 0.81 Greater Southern 0.54 Greater Western 1.10 Note: The higher the positive GHNIdx (General Health Need Index), the greater the level of “health need” and the higher the negative, the greater the level of “health need”.

Note: Normalisation of the general health need index (GHNIdx) involves computation of the highest negative value to bring it to value 1 (i.e. the highest negative was the Sydney South West value at –1.14) by adding 2.14 to –1.14. 2.14 is the value difference between – 1.14 and 1. To ensure all Area Health Services’ GHNIdx is normalised, the same value (2.14) is added to all other AHSs’ indices to obtain a normalised GHNIdx for all AHSs. The normalised GHNIdx is multiplied by each AHS population to derive a weighted population. To obtain equity shares for each AHS, the weighted population for all AHSs is divided by the total weighted population. The results of the normalisation process are shown in Table 6–7. Table 6–8 shows the eight AHSs’ general health need index (GHNIdx) scores ranked, starting with the AHSs with the lowest GHNIdx score (implying the least health needs) to the AHS with the highest health needs.

Table 6–8: General Health Need Index–Equity Share (GHNIdx–ES)

206 ______Table 6–9: General Health Need Index (GHNIdx) Normalised and Ranked according to Level of Health Needs

Area Health Service GHNIdx Normalised GHNIdx Ranked Sydney South West –1.14 1.00 1 Sydney West –0.87 1.27 2 Northern Sydney Central Coast –0.52 1.62 3 South Eastern Sydney & Illawarra –0.18 1.96 4 Hunter & New England 0.26 2.40 5 Greater Southern 0.54 2.68 6 North Coast 0.81 2.95 7 Greater Western 1.10 3.24 8 Note: Area Health Services are ranked according to their level of health needs. 1 = Area Health Service with the least health needs and 8 = Area Health Service with the highest health needs

Table 6–8 shows Sydney South West as the Area Health Service (AHS) with the least health needs, with a high negative index of –1.14, followed by Sydney West (–0.87), Northern Sydney Central Coast (–0.52) and South Eastern Sydney & Illawarra with –0.18. All four AHSs mentioned above are categorised as urban or metropolitan based, supporting the generally held belief that urban areas have a much healthier population than rural or remote classified Areas, especially in Australia. All the rural and remote based AHSs have a high level of health needs. For example, Greater Western emerged with the highest level of health needs at 1.10, followed by North Coast, Greater Southern and Hunter & New England with 0.81, 0.54 and 0.26 respectively. The results of this index support most indexes of disadvantage constructed by the Australian Bureau of Statistics (ABS), especially the Socio–Economic Indexes for Areas – Index of Relative Socioeconomic Disadvantage (SEIFA–IRSD) (Adhikari, 2006) (for more details about the SEIFA indexes see Appendix A), that has persistently shown a link between rurality/remoteness, socioeconomic disadvantage and burden of disease (Australian Bureau of Statistics, 2006c, Australian Bureau of Statistics, 2006a).

Based on the Area Health Services (AHSs) level of general health need (GHNIdx), as shown in Table 6–8, different AHSs attracted varying funding shares or equity shares (see Table 6–9). The table presents the ranking of AHSs according to their equity shares as estimated by the general health need index (GHNIdx). SES&I ranked first in terms of

207 ______equity shares, attracting 18% of the total health budget. H&NE, a rural classified AHS, was allocated 16%, second highest, and NSCC was third with 14%. Despite differences in population size and classification (urban and rural/remote) SSW, SW and NC received an equal share of around 11% each. Although GS and GW ranked last as the AHSs with the highest health needs, their equity share was around 10% and 8% respectively (see Table 6– 10). This is a reflection of GWAHS’ low population, which is estimated to be around 0.3 million, while the total population for SES&IAHS is four times (1.2 million) that of GW as noted earlier.

Table 6–10: Area Health Services Ranked according to the General Health Need Index (GHNIdx) Equity–Shares Area Health Service GHNIdx Equity–shares Ranked South Eastern Sydney & Illawarra 18% 1 Hunter & New England 16% 2 Northern Sydney Central Coast 14% 3 Sydney South West 11% 4 Sydney West 11% 4 North Coast 11% 4 Greater Southern 10% 7 Greater Western 8% 8

6.4.2 The Double–Variable Index (DVIdx) According to McIntyre, Muirhead, et al 2002, not all complex models yield the best expected outcomes by ensuring equitable distribution of health resources. Often, these statistical models are expensive and time consuming to develop and use due to the technical analysis involved; despite the expense and time involved, they have not managed to secure optimal funding and hence health for the populations. Therefore, there is a need to reassess the value of time and resources spent on constructing intricate models that policy makers often find difficult to implement. As Walt (1994) noted, simple and non–technical policies are easier to introduce and implement than complicated ones.

208 ______As explained in Chapter 5, a double–variable index (DVIdx) consisting of chronic morbidity and premature mortality was developed to investigate whether a simple index involving two key variables could effectively capture the relative level of health needs across and within AHSs. The two variables used to develop this index were the premature mortality (avoidable deaths from all causes for population aged less than 75 years 2000– 2002) and chronic morbidity (total number of chronic diseases within the population 2001) as shown in Table 6–11.

Table 6–11: Development of Double–Variable Index (DVIdx) Equity Shares

Data Source: NSW Health (2006) and Public Health Information Development Unit (2008) established by the Australian Government Department of Health and Ageing in 1999 to assist in the development of public health data systems.

Mortality (premature deaths for population aged under 75 years within one year) and morbidity (total of chronic diseases in a year) data for each AHS were summed to produce an average or mean which was then used as the DVIdx–equity share as shown in Table 6– 12. This approach attaches equal importance to each variable and is seen as appropriate since it reflects the actual population suffering ill–health. Furthermore, most policy documents or even composite indices do not utilise both variables when allocating health resources despite both variables being strong indicators of health needs. The two variables correlate strongly with each other and the data can be routinely collected and updated from official hospital databases, general practitioners (GPs) and other official records. Either variable can be used independently as a measure of some aspect of health need, although

209 ______there are health needs inherent in each of the variables that cannot be adequately captured by the other and this is the reason for applying both indicators.

These two variables have often been linked with populations of low socio–economic status worldwide. They were selected because they reflect the importance of reducing premature mortality and lightening the burden of chronic diseases in an effort to reduce health related inequalities through a more equitable resource distribution. The average proportion of the two variables was used to generate a mean: this was used as a Double–Variable Index (DVIdx) for each of the eight AHSs as shown in Table 6–12. The results from using the DVIdx to assess health needs presents a divergent pattern of health needs within the AHSs compared to that obtained from the general health need index and consequently the pattern of equity shares is different (see Table 6–12). The two equity shares generated from the DVIdx and GHNIdx were used in this study as the yardstick to assess progress towards equity in resource allocation across the eight AHSs in New South Wales. The two indices use two different approaches to assess equity, one incorporating the AHSs’ relative socioeconomic status among other indicators (GHNIdx), and the other the AHSs’ population health status (as defined by premature mortality and morbidity) in an attempt to identify any significant differences between the two approaches and to present a comparison of the results obtained with both approaches.

The differences between the DVIdx and GHNIdx equity shares for all the Area Health Services (AHSs) range between 2 and 6 percentage points; for some of the areas, the variation is below 3%. NSCCAHS attained the highest DVIdx predicted equity share of 19.7% , despite historically being categorised as the wealthiest AHS. This high percentage of equity can be explained by this area having the highest level of chronic diseases (25.0%) (see Table 6–11) compared to the other AHSs due to the amalgamation policy18 . Its total premature deaths for three years at 14.3% are also among the highest, with Sydney South West AHSs (SSWAHS) scoring the highest level of premature deaths of 17.2% and 14.9% for total chronic disease, which is reflected in its 16.1% equity–share. SES&I is the other

210 ______AHS experiencing high rates of chronic diseases (15.2%) and premature deaths (17.1%), which explains its 16.2% equity share.

Table 6–12: Area Health Services simple Double–Variable Index (DVIdx) and GHNIdx Equity Shares Area Health DVIdx Equity–Shares GHNIdx Equity– Services shares SSWAHS 16.1% 10.8% SESIAHS 16.2% 18.4% SWAHS 13.3% 11.2% NSCCAHS 19.7% 14.4% HNEAHS 13.8% 16.0% NCAHS 7.2% 11.2% GSAHS 7.5% 10.0% GWAHS 6.3% 7.9% Total 100.0% 100.0%

The two Area Health Services (AHSs) that had closely related total chronic diseases and premature deaths were SWAHS and H&NEAHS, thus attracting equity shares of 13.3% and 13.8% respectively. Their populations differ significantly, but the difference between their level of chronic diseases and premature deaths was minimal. On the other hand, NCAHS, GSAHS and GWAHS differed slightly in their level of chronic diseases; however, both areas had significant differences in premature deaths but the overall equity–share allocation was 7.2%, 7.5% and 6.3% respectively (see Table 6–12 and Figure 6–5).

18 In 2005, NSW Health reduced the number of Area Health Services from 17 to 8 by combining under– resourced AHSs with better resourced Areas. North Sydney was amalgamated with Central Coast AHS with a socially and economically disadvantaged population. This might explain NSCCAHS’ high level of morbidity.

211 ______Figure 6–5: GHNIdx and DVIdx Equity–Shares Compared

25.0%

20.0%

15.0%

10.0% Equity SharesEquity (%)

5.0%

0.0% SSWAHS SESIAHS SWAHS NSCCAHS HNEAHS NCAHS GSAHS GWAHS

Area Health Services

DVIdx Equity-Shares GHNIdx Equity-shares

6.5 Inter–Area Health Services Equity in Resource Allocation Assessment The following section examines whether the actual funding allocated from the State (the NSW Department of Health) level to the AHSs (population) level has been equitable in terms of differentially benefiting the most disadvantaged areas with high levels of health needs. The assessment was conducted using two different health needs indices. First, the general health need index (GHNIdx) was used to assess equity by computing the actual yearly allocation of funds received by the AHSs and then the same amounts were re– distributed using the GHNIdx to derive the predicted equity share of total resources that should go to each AHS, given its population size and level of health needs. These predicted shares derived from the two indices (GHNIdx and DVIdx) constructed for this study were used in the analysis and compared with the actual shares. The actual shares allocated from NSW Health to the eight AHSs were redistributed to determine if there were differences in funding between what was distributed and the two indices’ redistribution for a period of 4 years (2003/04 to 2006/07). Any ‘funding difference’ derived from this analysis will be

212 ______referred to as ‘equity–gap’, which is the difference between the actual and equity adjusted shares. Tables 6–13 and 6–14 present the results of the analysis using the GHNIdx.

213 ______Table 6–13: Comparison of AHS Actual Funds Allocation and General Health Need Index (GHNIdx) Equity Shares, 2003/04 to 2006/07 ($M)

Data Source: NSW Health Annual Reports

Table 6–14: Actual Funds and GHNIdx–based Allocations: Average Distance from GHNIdx, 2003/04 to 2006/07

214 ______Based on Tables 6–13 and 6–14, three main observations can be made:

1. That allocation of resources to two Area Health Services – South Eastern Sydney & IIIawarra (SES&I) and Northern Sydney Central Coast (NSCC) – was fairly equitable with no major differences between the Actual and GHNIdx–based allocation. 2. That resource allocation to Sydney South West (SSW) was inequitable with the Area Health Service receiving nearly twice as many resources as it should have based on the GHNIdx. 3. That allocation of resources to four Area Health Services (AHSs) – Hunter & New England (H&NE); North Coast (NC); Greater Southern (GS); and Greater Western (GW) – all classified as rural and remote were fairly inequitable with all the AHSs receiving less than their GHNIdx–based allocations.

Overall, the resource allocation can be described as fairly inequitable with four out of eight Area Health Services considerably under–funded on the basis of the General Health Need Index (GHNIdx) and two significantly over–funded. Based on the GHNIdx, the average distance from ‘equitable’ resource allocation ranged from 26.5% to 28.6% (see Table 6–14) for the four years analysed.

215 ______Figure 6–6: Actual & GHNIdx–based Allocation by Area Health Service: 2003/04 – 2006/07

Actual & GHNIdx-based Allocation by Area Health Service: 2003/04 Actual & GHNIdx-based Allocation by Area Health Services: 2004/05

1400.0 60.0% 1600.0 60.0% 47% 47% 1200.0 24% 40.0% 1400.0 24% 40.0% 1000.0 6% 1200.0 -6% 20.0% -0.1% 20.0% 1000.0 -3% 800.0 0.0% 0.0% -25% 800.0 -24% 600.0 -20.0% -20.0% -41% 600.0 -39% -49% 400.0 -58% -40.0% -40.0% Amount ($'M)Amount -63% ($'M)Amount 400.0 -61% Percentage difference 200.0 -60.0% Percentage difference 200.0 -60.0% 0.0 -80.0% 0.0 -80.0%

SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS

Area Health Services Area Health Services

Actual allocation 2003/04 GHNIdx-Based allocation 2003/04 Actual allocation 2004/05 GHNIdx-Based allocation 2004/05 Percentage difference Percentage difference

Actual & GHNIdx-based Allocation by Area Health Services: 2005/06 Actual & GHNIdx-based Allocation by Area Health Services: 2006/07

2000.0 47% 1800.0 47% 60.0% 60.0% 23% 1600.0 23% 1800.0 40.0% 5% 40.0% 1400.0 5% 1600.0 1400.0 1200.0 -4% 20.0% -4% 20.0% 1200.0 1000.0 0.0% -26% 0.0% -27% 1000.0 800.0 -20.0% -45% 800.0 -50% -20.0% 600.0 -52% -45% Amount ($'M)Amount Amount ($'M)Amount -40.0% 600.0 400.0 -65% -64% -40.0% 400.0 Percentage difference -60.0% Percentage difference 200.0 200.0 -60.0% 0.0 -80.0% 0.0 -80.0%

SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS Area Health Services Area Health Services

Actual allocation 2005/06 GHNIdx-Based allocation 2005/06 Actual allocation 2006/07 GHNIdx-Based allocation 2006/07 Percentage difference Percentage difference

Note: Allocations within the range of plus or minus maximum 10% of GHNIdx–based shares are judged equitable

216 ______Figure 6–6 shows the percentage funding difference of the actual share of resources allocated to Area Health Services (AHSs) and the GHNIdx predicted shares for the period 2003/04 to 2006/07. As discussed earlier, the GHNIdx constructed for this study used mixed indicators (demographic, socio–economic and health–related). The actual funding distribution when compared with the GHNIdx allocation presents a mixed picture of over– funded and under–funded Area Health Services (AHSs). For example, SSWAHS’ (an urban AHS) funding allocation was almost halved by the general health need index (GHNIdx) compared to what they were receiving. The over–funded difference was around 47% for all the years analysed (2003/04 – 2006/07), suggesting that the AHS funding was over and above the study benchmark of a maximum minus or plus 10%. Therefore, it is fair to conclude that Sydney South West AHS funding was inequitable. SWAHS also received more than its GHNIdx share by 23% points; although it was not at the magnitude of SSW, it still suggests that its AHS funding was fairly inequitable.

On the other hand, four out of the eight AHSs – H&NEAHS, NCAHS, GSAHS and GWAHS – were quite under–resourced with funding shares for the three AHSs almost halved, as shown in Figure 6–6. The two AHSs that scored plus or minus 10% according to this study yardstick, meaning they were classified as equitable, were SES&IAHS and NSCCAHS. SES&IAHS had a mix of being slightly over – and under–funded while NSCC was under–resourced for the four years analysed (2003/04 to 2006/07) (see Figure 6–6).

Overall, Table 6–15 shows the funding gaps (amounts and percentage) for the four years (2003/04 to 2006/07) analysed. According to the figures, the total funds that would have to be redistributed in order for the actual shares to equal the “equity funding shares”, as defined by GHNIdx, was $907.8 million (14.3%) in 2003/04, decreasing to $878.5 million (13.2%) in 2004/05, but then increasing to $1,171.6 million (14.0%) by 2006/07.

217 ______Table 6–15: Funds for Reallocation to Achieve Equity Targets based on GHNIdx, 2003/04 to 2006/07

2003/04 2004/05 2005/06 2006/07 Funds for redistribution $M 907.8 878.5 1,080.2 1,171.6 As a % of total funds 14.3% 13.2% 14.1% 14.0%

Double Variable Index (DVIdx) Tables 6–17 and 6–18 present the results of the analysis using the Double Variable Index (DVIdx) as explained earlier. Three main observations can be made from Tables 6–17 and 6–18: 1. The allocation to four Area Health Services – Hunter & New England (H & NE), North Coast (NC), Greater South (GS) and Greater Western (GW) – were fairly equitable with no major differences between what they were allocated (actual shares) and the DVIdx–based allocation. 2. The allocation to three Area Health Services – Sydney South West (SSW), South Eastern Sydney & IIIawarra (SES&I) and Sydney West (SW) – were inequitable with these AHSs receiving more than their fair share of resources based on their premature mortality and morbidity (DVIdx). 3. The allocation to Northern Sydney Central Coast (NSCC) was inequitable. It received less than its fair share of resources as predicted by DVIdx–based allocations.

Overall, the allocations looks fairly inequitable with three AHSs significantly over –funded on the basis of the Double Variable Index (DVIdx), one considerably under–funded and the other four slightly under–funded. As shown in Tables 6–18 and Figure 6–7, the average distance from equitable resource allocation ranged between 17.4% and 18.8% as determined by the DVIdx.

218 ______Table 6–16: Comparison of AHSs Actual Funds Allocation and Double Variable Index (DVIdx) Equity Shares, 2003/04 to 2006/07

Data Source: NSW Health Annual Reports (2003–2007) and Public Health Information Development Unit (2008) established by the Australian Government Department of Health and Ageing in 1999 to assist in the development of public health data systems.

Table 6–17: Actual Funds and DVIdx Allocation: Average Distance from DVIdx, 2003/04 to 2006/07

219 ______Figure 6–7: Actual & DVIdx–Based Allocation by Area Health Services: 2003/04 – 2006/07

Actual & DVIdx-based Allocation by Area Health Services: 2003/04 Actual & DVIdx-based Allocation by Area Health Services: 2004/05 1,600.0 30.0% 1,400.0 19% 30.0% 22% 21% 1,400.0 20.0% 20.0% 13% 12% 1,200.0 12% 10.0% 1,200.0 10.0% 1,000.0 -10% 1% 0.0% 0.0% 1,000.0 -9% 7% 800.0 -10.0% -7% 800.0 -10.0% 600.0 -13% -20.0% -6% -47% 600.0 -12% -20.0% Amount ($'M)Amount -30.0% ($'M)Amount 400.0 -43% -40.0% 400.0 -30.0% Percentage difference 200.0 Percentage difference -50.0% 200.0 -40.0% 0.0 -60.0% 0.0 -50.0% SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS Area Health Services Area Health Services Actual allocation 2003/04 DVIdx-Based allocation 2003/04 Actual allocation 2004/05 DVIdx-Based allocation 2004/05 Percentage difference Percentage difference

Actual & DVIdx-based Allocation by Area Health Services: 2005/06 Actual & DVIdx-based Allocation by Area Health S ervices: 2006/07

22% 22% 18% 18% 1,800.0 2,000.0 11% 30.0% 10% 30.0% 1,600.0 1,800.0 6% 20.0% 20.0% 1,400.0 1,600.0 10.0% 1,400.0 10.0% 1,200.0 -11% 5% -10% 1,200.0 1,000.0 0.0% 0.0% 1,000.0 -10.0% 800.0 -10% -10% -10.0% -14% Amount ($'M)Amount -15% ($'M) Amount 800.0 600.0 -20.0% -20.0%

Percentage difference 600.0 -44% -45% -30.0% Percentage difference 400.0 400.0 -30.0% 200.0 -40.0% 200.0 -40.0% 0.0 -50.0% 0.0 -50.0%

SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS Area Health Services SSWAHS SESIAHS SWAHSArea NSCCAHSHealth Services H&NEAHS NCAHS GSAHS GWAHS

Actual allocation 2005/06 DVIdx-Based allocation 2005/06 Actual allocation 2006/07 DVIdx-B ased allocation 2006/07 Percentage difference Percentage difference

Note: Allocations within the range of plus or minus maximum 10% of DVIdx–based shares is judged equitable

220 ______Figure 6–7 shows the percentage funding difference for all eight Area Health Services (AHSs) based on the Double Variable Index (DVIdx) from 2003/04 2006/07. Three out of the eight AHSs were within the study yardstick of plus or minus 10%; therefore the funding allocations can be classified as equitable. H&NEAHS and GWAHS were both over– resourced, ranging from 6% to 10% (2003/04 – 2006/07), although H&NEAHS was slightly outside the range (–11%) in 2004/05. Funding for NCAHS was also equitable since its under–funded difference fell within the study range (plus or minus 10%), suggesting that the AHSs’ funding was ‘equitable’ according to the double variable index (DVIdx) developed using premature mortality and prevalence of morbidity. As with the other index (GHNIdx) used in this study, SSWAHS emerged as the Area that had the highest share of actual funds allocated above its indicated DVIdx–equity share.

Finally, Table 6–18 shows the total funding gap (amounts and percentage) for the four years (2003/04 – 2006/07) analysed. Based on the findings, the total funding that would have to be redistributed in order for the actual shares to equal the “equity funding shares”, as defined by DVIdx, was $594.2 million (9.4%) in 2003/04, decreasing to $576.7 million (8.7%) in 2004/05, and then increasing to $721.9 million (9.4%) in 2005/06 and $788.4 million (9.4%) by 2006/07.

Table 6–18: Funds for Reallocation to Achieve Equity Targets based on DVIdx, 2003/04 to 2006/07

2003/04 2004/05 2005/06 2006/07 Funds for redistribution $M 594.2 576.7 721.9 788.4 As a % of total funds 9.4% 8.7% 9.4% 9.4%

221 ______

Chapter 6 Section 2: Summary of Quantitative Findings The quantitative analysis confirms the qualitative findings that inequities still persist in the allocation of resources to Area Health Services from the NSW Department of Health, although there has been some growth in funds allocated to the eight AHSs since the introduction of the RDF in 1989/90, and a significant decrease in premature mortality rates.

Equity in resource allocation from NSW Health to AHSs was difficult to assess due to lack of data indicating the amount of funds that were allocated through the Resource Distribution Formula (RDF), making it difficult to determine whether the decrease in premature mortality was related to the increases in funding.

Therefore, a general health need index (GHNIdx) and a double variable index (DVIdx) were constructed and used to assess levels of health needs among AHSs in NSW Health. Using the indexes as proxy measures of health needs, funds allocated to AHSs from 2003/04 to 2006/07 were distributed to determine whether there had been movement towards equity.

Health needs were greatest in the four rural and remote classified AHSs – H&NEAHS, NCAHS, GSAHS and GWAHS – although NSCCAHS and SES&IAHS, both classified as metropolitan areas, exhibited higher levels of health needs than Sydney South West and Sydney West AHSs. However, due to the large population in the urban –based AHSs – SSWAHS, SES&IAHS, SWAHS and NSCCAHS – these four AHSs attracted greater funding shares.

Resource allocation across the AHSs differed between the urban and rural/remote AHSs. Allocation of resources benefited urban AHSs the most, particularly SSWAHS. Although the historically under–resourced AHSs have also benefited to a large extent, some rural AHSs are still under–funded, taking GHNIdx and DVIdx equity shares into consideration. Resource allocation in NSW Health does not benefit NSCCAHS, an urban AHS, compared to other AHSs. SSW (urban) AHS was the area that received more than an equitable share of funding based on RDF, GHNIdx and DVIdx.

The average distance from achieving equitable resource allocation varied significantly taking the three indices into consideration, suggesting that different need indicators yield different outcomes. The GHNIdx constructed in this study, using a combination of demographic, socioeconomic and health–related indicators, produced the highest funding gap. The DVIdx, constructed using premature mortality and chronic morbidity data, yielded a much less than average distance from equitable resource allocation. This highlights the importance of selecting health need indicators that accurately reflect the population’s actual health needs rather than the population’s socioeconomic disadvantages. The RDF had the least percentage gap in funding from equitable resource allocation and, although NSW Health claimed that the weighted average distance from equity had been reduced to under 2% after the amalgamation of over–resourced AHSs with under– resourced areas, this analysis shows that the average distance has not been achieved.

Overall, growth funds distribution has benefited all the AHSs to a certain degree,

222 ______although disadvantaged Areas continue to receive less than their fair share of funding, which does not reflect their level of health needs.

According to the health needs indices developed for this study, the NSW resource distribution formula (RDF) needs to be revised to include health–related indicators, such as chronic diseases (morbidity) rates, to ensure Area Health Services experiencing a high burden of disease are receiving a fair share of resources commensurate with the geographical differentials in health needs.

Population is a key factor in the distribution of resources to the eight Area Health Services with urban AHSs attracting a ‘larger slice of the pie’ since their population is higher, even though the rural AHSs bear a high burden of premature mortality and chronic morbidity.

223 ______

Chapter 7: RESOURCE DISTRIBUTION AT THE AREA HEALTH SERVICE LEVEL: QUALITATIVE AND QUANTITATIVE FINDINGS

7.1 Overview This chapter presents the qualitative and quantitative findings relating to the second research question: how are the financial resources received from the State level distributed at the Area Health Service level? The main objective was to explore and document the process of resource allocation at the AHS level [which is largely unknown]. The extent to which the allocation within AHSs has been equitable is not part of this chapter; that has been assessed in the next chapter (Chapter 8). The chapter is divided into two sections: the first section presents the findings from the qualitative analysis that explored the perspectives of AHS executives and program managers on the intra–AHS resource allocation process. The second section presents the results of the quantitative component that examined the pattern of distribution of AHS funding through an analysis of expenditure data extracted from annual reports of the areas. The main reason for analysing the pattern of distribution was to determine whether the emphasis of resource allocation in NSW is still on hospital–based services or shifting to community–based programs. Two Area Health Services – North Sydney Central Coast and Greater Western – were the focus of this analysis.

Section 1: Qualitative Findings

7.2 Introduction The State level findings presented in Chapter 6 indicate that there has been some progress towards greater equity in distribution of funding to some Area Health Services while resources to others are inequitable. However, it is not known whether the movement towards equity at the State level has been reflected in equitable resource allocation at the AHS level. Indeed, it remains largely unknown how equity has been pursued in the

224 ______distribution of financial resources at the AHS level and the weight given to equity in the process.

This section presents the opinions of key health officials derived from interviews within two AHSs: Northern Sydney Central Coast (urban) and Greater Western (rural). They were purposely selected for the qualitative component because of their differences in demographics and socio–economic status, including their level of disadvantage. The interviews with health officials were expected to shed light on the process of resource allocation within the AHSs, but from the outset it was clear that resources are distributed according to specific programs and services, most of which have performance targets, making it difficult for health officials to shift resources from one program to another. According to emerging themes, most of those interviewed recounted the difficulties experienced at the AHSs level. The following were the emerging themes from the NVivo analyses of the qualitative data that were seen as influencing equity in resource allocation at the AHS level: 1) inadequate funding and maldistribution of available funds; 2) lack of tools to guide resource distribution at the AHS level; 3) greater emphasis on achieving efficiency which hampers achievement of equity; 4) wastage of resources due to the lack of a coordinated approach to service delivery and collaboration between health providers; 5) lack of resource allocation transparency and accountability, and 6) monitoring and evaluation systems.

7.1.1 Inadequate Funding and Maldistribution of Available Funds Inadequate funding is a major challenge to many health systems. The amount of funding received by Area Health Services (AHSs) in NSW was evident as a major issue affecting resource allocation in general and equity in particular. The majority of the AHS executives and program managers pointed to the inadequacy and unfair allocation of funding between hospital–based and community health services. A significantly large proportion of the health budget is allocated to hospital–based services leaving community health care stifled by lack of funding. According to respondents, the limited amount of funding for the AHSs considerably constrains their efforts to improve equity.

225 ______“The amount of funding we receive from NSW Department of Health is insufficient and even what is received around 80% is allocated to hospital based programs and remainder to community health programs…community health services have been historically under–funded. Also, there is no flexibility to move funds to community based services and as long as there is no movement of resources we cannot claim to be achieving equity which is all about targeting those population groups with greatest health needs with appropriate services… (Director, AHS)

Frontline health professionals dealing with service delivery recounted the constraints they faced in their day–to–day activities as a result of the limited funding. Equally, the consumers interviewed (see Chapter 8, Section 1) were also affected by the insufficient funding of services since they had to travel to hospitals to access even minor services, such as management of diabetes, due to lack of access to services at the community health centres. Apparently, the lack of relevant health services closer to the community resulted from inadequate AHS financial resources and in some places a shortage of general practitioners (GPs). As explained by the policy makers and health executives at the NSW Department of Health (see chapter 6), the history of funding for the NSW health system means that the bulk of funding is allocated to hospitals and a small percentage to community health (see Section 2 for the actual amounts allocated by each AHS to hospitals and community health care). Clearly, the uneven distribution of resources between the hospitals and community health services is a deterrent to any attempts to target resources and services for populations with the greatest health needs.

The program managers believed that some diseases are also unduly favoured when it comes to funding allocation, resulting in diseases such as diabetes reaching epidemic proportions as reflected in the following quote.

“Another problem with the resources allocated is that diseases like diabetes and other chronic diseases are seen as truly unattractive…symptoms progress slowly, the outcomes are however permanent…some medical professionals say diabetes can wait as it’s not a funding priority because there are other priorities like heart attack etc…if you get funding from the NSW Department of Health it would be most likely for cardiac rehabilitation services, ECG machines or Intensive Care Unit. The reason diabetes continues to be an epidemic is because funding is insufficient…” (Program Manager, AHS)

226 ______

Another opinion expressed is that hospital based care was more important than preventative care and health promotion as highlighted in the following quote.

“…Health dollars are mainly allocated to hospitals and not community health centres which are supposed to deal with prevention and management of existing illnesses…GPs are also supposed to provide services including health promotion but some places like rural and remote areas have no GP practices and where the GPs are available they are charging fees to see patients…NSW Health system is all about responding to hospital crisis after the other…basically NSW hospitals are always in the news for all the wrong reasons…” (Diabetes Program Manager, AHS)

Some respondents suggested that the fees charged by GPs who do not bulk–bill might be driving patients to access hospital services, even with minor health problems that are treatable locally.

“…the problem is that some GPs don’t bulk–bill and so patients have to pay and claim some money back from Insurance [Medicare]…So most patients come to the hospitals because of that, causing an overload to already strained workforce and poor quality of services in hospitals and competition for the insufficient health dollars given to areas…” (Director, AHS)

According to the respondents, health professionals at the frontline of service provision feel that there is a need to allocate funding to chronic diseases and aged care, offering services at a community level rather than being hospital based. “…there is an urgent need for funding to help us continue to offer services closer to the communities since the number of chronic diseases and aging population is growing…we need more resources in that direction to facilitate chronic diseases management and education in the communities especially to manage and educate those elderly and vulnerable who are most likely to end up in hospitals for services which could have been provided within the communities either by the GPs or at the community health centres…” (Assistant Director, AHS)

The other concern raised by participants has to do with the 2005 amalgamation reform, reducing the number of Area Health Services (AHSs) from seventeen to eight. The general response from most of the respondents was that this particular reform has contributed to funding and service fragmentation. According to some respondents, merging areas to create

227 ______larger entities might have escalated the historical inequities that existed within the old AHSs. This means that the pockets of population groups with greater health needs that were known to exist previously may be concealed within the larger AHSs, thus defeating the whole purpose of the reform.

“…since the amalgamation funding to most programs has become more fragmented plus services are also further from communities. There is no clear guidance of how after amalgamation the inadequate funding received by Areas should work to address the historical differences...the reform might have disguised inequities that existed… At the moment we are trying to introduce a planning process through clinical networks involving different AHSs, there is also now a clinical network that is charged with addressing equity, charged with making sure that there are consistent services across the entire AHS. But we are yet to see how the team manages to deal with all the issues…” (Executive Director, AHS).

7.1.2 Lack of Tools to Guide Resource Distribution at the AHS Level Area Health Service officials interviewed were concerned about the problem of shifting resources to areas under their respective jurisdictions where health needs are greatest. The respondents recognised the need for resource distribution being health needs based rather than driven by power. As suggested by the following words, the willingness for change is there; what is needed is the right mechanism to facilitate the redistribution of resources.

“…we have identified areas of greatest need and we are working towards finding mechanisms to redistribute our available resources to those areas of greatest need but in “a piece–meal–way”,…there is willingness to change away from the historical domain of retaining resources in the areas of greatest power and moving them to where they will do most good as evidence shows but at the moment we don’t have the right tools…” (Senior Health Official, AHS)

In the absence of a funding distribution tool at the AHS level, funding reallocation was driven by “demands created by hospitals” and not by needs, except when funds were earmarked for a particular program.

“…the problem in our system [NSW health system] is the lack of good tools on how to shift resources according to health needs especially at this level [population level]. As you know the formula [RDF] is used at the Department

228 ______level but at the Area level the Chief Executive has to decide where to put the money not necessary using a formula since we don’t have one…the lack of a formula means that most of the health dollars are allocated to hospitals and the little money left over to programs…the other thing is that even though program managers produce budgets they don’t always get the whole amount they have budgeted for due to the demand created by hospitals…some programs like mental health are Ok since the funding is earmarked and so Chief Executives cannot pinch the funds to allocate to hospitals or other services outside mental health…” (Director, AHS)

The challenge posed by the amalgamation of AHSs into bigger entities, combined with the lack of an effective mechanism to move funds according to population health needs, was reiterated by senior health officials; they emphasised the importance of transparency in the light of amalgamation. According to these participants “they were in need of a formula urgently” as a way of facilitating and “targeting resources to improve access to services” and to “meet specific needs” of “those who have higher needs” such as the Aboriginal population.

“…our area is in need of a formula urgently to help us allocate and target funding to disadvantaged population groups with greatest health needs…after the amalgamation this Area [GWAHS] is very large with dispersed population… some of these population groups have very high health and other needs so we have to find a way to target resources in order to improve access to services…although funding received from NSW Health is not sufficient to meet all the needs we can still make do with what we get for now to reduce the number of chronic diseases…there is also large Aboriginal population groups who have higher needs than the general population and require culturally sensitive services to meet their specific needs…” (Senior Health Official, AHS.

7.1.3 Greater Emphasis on Achieving Efficiency Hampers Achievement of Equity Although concerns about equitable allocation of resources have characterised the NSW Health care system since the 1980s, the conflict between the goals of efficiency and equity has continued to create tension. Indeed, even though equity has been raised to a high level among health policy objectives, the practical notions of equity that should inform policy and the ways in which these should be implemented are far from clear within NSW Health system. Furthermore, there is no consensus on how to deal with policies that may cause a

229 ______conflict between the goals of equity and efficiency – that is, those that may improve equity while decreasing efficiency or improve both (equity and efficiency) equally. According to the respondents, the equity versus efficiency dilemma has been virtually ignored in the political debate, often leading to inconsistent judgments in the development of health and resource allocation policies. This is how the confusion was expressed by policy makers at NSW Health;

“…funding we give to Areas can not be all earmarked in order to target population with greatest needs because, if we tied most of the money for use in a particular way we would actually reduce the overall efficiency of the system. Now, this is a social tension that we policy makers have to deal with when it comes to efficiency and equity…the Resource Distribution Formula (RDF) is all about equity and another way we fund health services is through episode [casemix] funding which is all about efficiency and more historical... equity and efficiency are in tension… (Senior Health Official, NSW DoH)

“Episode funding (case–mix) funding is used in order to ensure that the money which is given out under an RDF is used most efficiently. NSW Department of Health uses a benchmark of service costing. For example, Department of Health costing of services means that if Areas deliver service Y to a client it should cost Areas X dollars but if it cost X + 5 dollars that means that Area [AHS] is inefficient and need to look at their cost practice so that they can bring it down to X amount even if it means the service is affected. This sort of funding will never encourage equity in funding but efficiency and this is where the tension is…the bottom line is that, the Department of Health needs to assess how funding distribution can be done differently to improve equity…” (Senior Health Official, NSW DoH)

At the AHS level, it was clear from the interviews that there was no specific resource allocation policy or mechanism to steer resources to the areas of greatest need away from hospitals. The need to promote equity as well as efficiency was highlighted, although health executives at the State level were not sure how to go about introducing new systems. Politics and financial answerability were expressed as possible challenges that would be difficult to overcome to ensure the systems were implemented without resistance. The inefficiencies inherent in some of the hospitals were raised by the respondents and were seen as hindering progress in equity achievement.

230 ______“NSW Health is concerned about achieving hospital efficiency and equity within all levels of the health care system but politics and lack of accountability when it comes to health dollars will always be an issue...to be honest some hospitals are known to be inefficient so even adjusting for all different factors will not help them achieve efficiency and equity may never be achieved with episode funding based resource distribution policy…We need clear funding distribution policies for each Area taking into consideration the health needs of the populations to guide resource allocation… especially how to determine what percentage amount of dollars should be spent on what programs based on need and identify priority programs that target those disadvantaged with services” (Senior Health Official, NSW DoH).

7.1.4 Wastage of Resources Due to Lack of Coordinated Approach to Service Delivery and Collaboration between Health Providers Interviews with health professionals revealed that there is no coordinated approach to delivering health services in Australia and specifically in NSW; this lack of co –ordination contributes to wastage of limited health resources through duplication. Even within Area Health Services, there is no single concerted approach to tackling the health inequities that exist in accessing services. Indeed, the fragmented nature of service delivery may have exacerbated the chronic disease epidemic through disjointed service provision and lack of comprehensive community based health services, according to the respondents. Equally, the tug–of–war between the Commonwealth and State governments about who funds what services is costing tax payers substantial financial resources, according to the health professionals. The seriousness of this problem was explained as follows:

“…as a result of Commonwealth and State governments duplication of health services, it’s estimated to cost around $2 billion dollars a year…this wastage of resources is really unacceptable when you think about the level of chronic diseases, elderly population and increasing number of obesity in the country…people dying from preventable diseases is five times higher among poor and disadvantaged Australians than among the wealthy people and I think twenty times greater among Indigenous people and yet we claim to be a developed country ranked third among rich countries…is this fair or equitable? No, we need an urgent reform and that reform should involve health being funded by one body – I mean Commonwealth government funded but delivered by the States and an independent commission under Commonwealth government should regulate, monitor and evaluate the services being provided…this hopefully will improve coordination of services and reduce wastage of limited health dollars…” (Senior Health Official, AHS)

231 ______

Health in Australia should preferably be funded by the Commonwealth government and delivered by the States, an arrangement respondents believe will reduce the finite health resources wasted as a result of different authorities providing similar and inadequate services. An estimated $2 billion dollars can be saved through a coordinated approach to service delivery that focuses on providing better preventative care and other health services. According to health officials, NSW Health needs to take a leadership role to ensure duplicated services are merged through formal collaboration. But how can NSW Health take a leadership role when it contributes the least when it comes to health funding?

“…wastage of health dollars through a range of organisations providing same services in this AHS [GWAHS] is great…the organisations have different structures, funding and approaches to service delivery so this is why they are resisting to merge. Merging the services would reduce duplication of services and health dollars...the differences are often difficult to overcome without policy oversight and hinder true collaboration…NSW Health has not developed mechanisms to help Areas work through such difficulties especially after the amalgamation. I generally think NSW Department of Health is not able to manage the health portfolio well and would support a take–over by the Federal Government in order to overhaul the whole health system” (Senior Health Official, AHS)

In NSW, different health providers have the obligation of providing health services to the population. However, according to the health professionals interviewed there is no relationship among the various health service providers. This lack of partnership is the main cause of inadequate management of patients, especially those suffering from chronic diseases, according to those interviewed. Therefore, there is a need for pathways that facilitate collaboration among service providers. According to the respondents, this collaboration would advance management of patients suffering from chronic diseases and reduce the number of patients who end up in hospitals acutely ill from something that could have been prevented.

“…there is little collaboration between the various health providers in NSW…for example, patients suffering from chronic diseases are meant to be treated, managed and monitored by General Practitioners (GP) using the guidelines provided by NHMRC as a standard procedure, however, patients are not

232 ______adequately managed and educated about the chronic diseases and their determinants so most of them end up in hospital with serious complications and acutely ill…this causes overload of hospital with patients who should have been effectively dealt with at the GP or community health centres…but I have to say patients don’t like seeking treatment from GPs since some GPs don’t bulk–bill so patients have to pay out–of–pocket and when we are talking about disadvantaged populations they may not have the money or be willing to pay…the system needs to change to ensure access to essential services outside the hospital…” (Senior Health Official, AHS).

7.1.5 Lack of Resource Allocation Transparency and Accountability Another important issue that emerged from the interviews with regard to resource distribution and decision –making at the Area Health Service level was the lack of adequate information about how decisions are made and what factors are taken into consideration during the allocation. According to most of the respondents both at the Department and Area Health Service level, resource distribution is dealt with in secrecy and managers are only informed about what has been allocated and the performance targets they have to meet with those resources. Health professionals suggested that if they were engaged and consulted during the resource allocation process, this would make a difference in the way they, in turn, would use the funds allocated to ensure the needs of the population are met.

“…resource allocation decision making in NSW Health is never discussed with us, no consultations to identify priority areas since we are the ones providing services to the communities… all we are told is here is the amount of dollars and this is what you have to do with it in terms of performance targets… so no question about that and then there are no financial accountability mechanisms in place to make sure the money is used appropriately…” (Senior Health Professional, AHS)

“…openness about the system of how health dollars are distributed and for what services is important since health professionals, I think would make use the limited funds if they knew the allocation process and the priorities rather than just passing on funding and asking them to achieve so many performance targets with little resources…another important thing is that people are now better informed about making healthy choices and better equipped to make decisions about their healthcare, so it is essential that we use people’s knowledge and experience to improve the services they use by being accountable and open about our dealings...” (Senior Health Official, AHS).

233 ______

Health executives at the AHS level felt that the secrecy involved in resource allocation was the main reason Areas have not promoted equity. At the same time, they acknowledged that the resources allocated are not utilised effectively to improve patient quality of care. The issue of resources spent on teaching and research was raised.

“…the lack of openness on how resource allocation decisions are made leads Areas to be under–resourced and the funds we receive is based on ten programs… equity has not been promoted because of distribution of resources across programs rather than targeting population with health services. The funding to Areas in NSW also poorly utilised leading to poor standards of care... health dollars are also not spent on research…currently no major research is being conducted and there is little influence on teaching in some urban and rural Areas. So this makes it impossible for rural areas to attract and retain the right workforce because of geographical distances from the centres of interest…” (Senior Health Official, AHS)

“To improve how resources are spent NSW health system should have fewer large hospitals like Royal North Shore and large hospitals like Gosford and no hospitals such as Hornsby, Manly, Mona Vale and Ryde – that is what we policy makers and clinicians would say. And on that note also improve community based care through increasing more primary health care centres. But all this needs serious debate involving how resources can be distributed to achieve this…we need accountability and a team we can trust to deliver the services while consulting and engaging the community…” (Senior Health Official, AHS).

7.1.6 Lack of Data Collection, Monitoring and Evaluation Systems According to respondents, data collection needs to be improved to ensure AHSs have accurate information about their populations. As expressed by respondents, the datasets should ideally include the health needs of the population, including who is benefiting from the health resources and services provided. Having a dataset of services outlining which organisation is providing which service would also be helpful in reducing the duplication of services and consequently increase services where they are lacking, according to health professionals. Furthermore, the population data would facilitate effective distribution of available resources, including targeting the most disadvantaged population groups with services. The respondents also explained that adequate data collection systems need to be

234 ______developed to ensure accurate population information is collected so as to identify health needs and to aid in distribution of resources and services according to health needs.

“…the problem as you know is that NSW Health don’t have really good data systems like UK and so we don’t know the actual population needs, services provided or if population groups are benefiting from those health services…at a department level we are actually allocating funding to various data collection systems to be able to capture some data but it is not consistent as it should be…a system is being developed and has not been fully implemented which apparently has its own problems and the Department was looking at a community health and outpatient database. Some areas do collect data but it’s always incomplete and can’t do a comparison or good analysis of population health needs with incomplete data. What we the health system needs one consistent database across the state…” (Senior Manager, NSW Health)

According to respondents, the data collected will not only be useful for guiding resource distribution but also assist in categorising performance indicators pertinent to each Area Health Service (AHS), since all eight AHSs have unique demographics and settings, implying that a service delivery model that might be suitable for urban AHSs will not necessary be suitable for rural and remote AHSs. As noted by the respondents, NSW Health has to actively facilitate the development of the data collection systems;

“…data for each Area is required to guide resource allocation and service provision…the data will also be useful for identifying performance indicators which may be suitable for explaining performance in metropolitan regions, rural and remote areas. Performance indicators for metropolitan areas are not necessarily suitable for rural and remote areas when it comes to explaining changes regarding chronic and complex disease prevention especially…same thing with service delivery – the model for delivering services to urban areas can not be used for rural and remote population… this is what the Department of Health needs to coordinate well rather than assume Areas will get it right…” (Executive Director, AHS)

The interviews also revealed that there are no effective mechanisms for monitoring and evaluating resource allocation and health programs. The lack of evidence as to what really works and what does not, contributes significantly to the poor delivery and duplication of services. Therefore, systems need to be developed for monitoring the data collection and evaluating the use of health resources in terms of health services provided. According to

235 ______decision makers and health professionals, the availability of these essential systems will facilitate distribution of resources to areas of greatest need, rather than the decisions being politically driven.

“…there are no monitoring structures in place to ensure accountability, transparency and each Area Health Service specific performance–related evaluation…this causes duplication of services and wastage of health dollars which could be devoted to targeting those disadvantaged populations with appropriate services…” (Senior Health Official, AHS)

“…comprehensive population data detailing the health needs and locality of the population is needed including monitoring how the areas are performing in terms of data collection not just the hospital based data but also those patients who access GP services and those who don’t access the services... the data will be useful in the distribution of resources and services to where they are most needed according to evidence. It’s the best way to run a health system not politically influenced…” (Senior Health Official, AHS).

236 ______

Chapter 7 Section 1: Summary of Qualitative Findings

This section was expected to highlight the pattern of resource distribution within the Area Health Services (AHSs); instead it provided some insights into factors that influence resource allocation at the AHS level, as perceived by the executives, managers and health professionals. In particular, it has shown that the slight movement towards equity for some AHSs at the State level has not necessarily been reflected in equitable distribution of health resources at the AHSs level for several reasons. The most important reason is the lack of clearly set out and well communicated resource allocation policies that focus on improving the equity of distribution of health care funds at both levels of the health system.

According to the views of the health officials interviewed, while equity is stated as a resource allocation criterion in the NSW Department of Health policies, it is not the case at the Area Health Service level of the health system. In most of AHSs Annual Reports there is little mention of equity. The following are the factors that emerged from the interviews:

Inadequate funding and maldistribution of available funds, Lack of tools to guide resource distribution at the AHS level, Greater emphasis on achieving efficiency hampers achievement of equity, Wastage of resources due to lack of a coordinated approach to service delivery and collaboration between health providers, Lack of resource allocation transparency and accountability; and Lack of data collection, monitoring and evaluation systems.

237 ______Section 2: Quantitative Findings

7.3 Introduction This section presents the quantitative analysis of the pattern of resource distribution within Area Health Services (AHSs) as part of the attempt to document the intra–AHS resource allocation process. As indicated previously, the extent to which the distribution has been equitable is dealt with in the next chapter (Chapter 8). The focus of the analysis here is to show whether or not there have been any changes in the ‘traditional’ emphasis on hospital– based programs in resource allocation. It is important to repeat that resource allocation within AHSs in NSW is based on programs and services rather than on the health needs of geographically defined populations. The analysis was approached in two ways: first, funding for individual programs and services was aggregated into three broad program areas – community–based programs, hospital–based programs and teaching and research programs – and the proportion of funds received by AHSs that goes into each of the three program categories was analysed to show changes in the trend of distribution. Second, changes in per capita allocations for each of the 10 programs in the NSW resource allocation formula were analysed over a three–year period from 2004/05 to 2006/07.

7.4 Distribution of Resources within Area Health Service Level Resources within each of the eight Area Health Services are allocated according to ten health care programs and services as follows:

 Primary & Community health services  Aboriginal health services  Outpatient services  Emergency care services  Overnight acute services  Same day acute services  Mental health services  Rehabilitation and extended care services  Population health services

238 ______ Teaching and research costs

7.1.7 Resource allocation to Area Health Services Programs and Services To establish the amount of funding that is allocated to community health and hospital– based services, the above programs were categorised into three expenditure components: community–based health services, hospital–based health services and teaching and research expenditure. This categorisation of programs is not completely precise since some health programs under the community–based component could be funded and provided within a hospital setting; equally those under the hospital–based component, such as mental health, could possibly be provided in the community. Therefore, the categorisation is only used for analysis in this study to determine the pattern of resource distribution at the Area Health Service level since it is not known how funds are distributed or shifted from one health program to another. The categorisation of the three components is shown in Table 7–1.

Table 7–1: AHSs Health Programs and Services Categorised COMMUNITY–BASED HOSPITAL–BASED SERVICES SERVICES TEACHING & RESEARCH Primary & community health services Outpatient services Teaching and research Aboriginal health services Emergency care services Rehabilitation & extended care Overnight acute services Population health Services Same day acute services Mental health services

The funding allocated to each of the programs as shown in Table 7–1 was then identified and analysed according to the three components (see Table 7–2) to ascertain the percentage funding that is allocated to community health and hospital–based services, including teaching and research. Funding to each of the programs under each component was summed to derive a percentage share of funds allocated to the three components by the eight Area Health Services (AHSs) in New South Wales (NSW). Three funding years were analysed for comparison purposes and the results are presented in Table 7–3 and Figure 7– 1. Table 7–2 shows the total amount of funding spent by the eight AHSs on community and hospital–based services, including teaching and research from 2004/05 to 2006/07

239 ______Table 7–2: Area Health Services Programs/Services Funding, 2005 to 2007

Data Source: Area Health Services Annual Reports, 2004/05 – 2006/07

Based on the results shown in Tables 7–2 and 7–3 and Figure 7–1, hospital based programs and services were more favoured when it came to resource allocation, with all eight AHSs consistently allocating the bulk of their total budget to hospital services for the three years analysed (2004/05 – 2006/07). For most AHSs, the allocation to hospital based services was over 75% of their health budget, apart from Sydney West AHS which allocated 73% and Greater Western 72% in 2004/05. The funding to community –based services was relatively insignificant with seven out of eight AHSs allocating less than 23% of funding to community–based health services from 2004/05 to 2006/07. Greater Western was the only AHS to allocate a slightly larger percentage, ranging from 23% to 27% (see Table 7–3)

Most urban AHSs – Sydney South West, South Eastern Sydney & Illawarra and Northern Sydney Central Coast – allocated much less funding to community health services, totalling between 16% and 17% for each, while Sydney West allocated 21% in all three years. On the other hand, teaching and research received less than 5% funding by most of the AHSs, apart from Sydney South West, South Eastern Sydney & Illawarra and Sydney West, which allocated 5%, 7% and 6% respectively (see Table 7–3).

As shown in Table 7–3 and Figure 7–1, the percentage funding allocated to community health, hospital–based services, and teaching and research across the eight AHSs in 2005– 06 showed minimal differences compared to 2004–05. Funding to the eight AHSs presented a similar pattern in distribution for the year 2005–06, almost identical to the

240 ______previous year (2004–05), with only 1 to 2% growth in hospital funding. For example, Sydney South West allocated a similar percentage of funds to both community health and hospital–based services in both financial years (2004–05 and 2005–06), while Northern Sydney Central Coast AHS allocated the same (79%) to hospital –based services and 1% less to community health–based services (see Table 7–3) in 2005–06. The remainder of the AHSs had a 1% funding decrease for community–based services in the same year as shown in Figure 7–1. Funding for teaching and research was roughly the same with minor increases or decreases for some areas, although for most of the rural and remote areas percentage expenditure was much less than the urban –based AHSs. Similarly, funding distribution to the eight AHSs changed slightly for both community health and hospital– based services in 2006–07 compared to 2005–06. Sydney South West AHSs, for instance, had a funding increase of around 4% for hospital based services from 79% in 2005–06 to 84% in 2006–07, although funding to community based services remained almost identical at 16%. Another area that had a slight growth in funding of 3% was Greater Western AHS from 74% (2005–06) to 77% (2006–07) followed by North Coast with 2%. However, for hospital–based services to benefit generously in both AHSs, community health based services had to suffer a reduction in funding of 2% in the case of North Coast and 3% for Greater Western.

South Eastern Sydney & Illawarra and Greater Southern were the two Area Health Services (AHSs) that experienced a 1% reduction in funding allocation for hospital –based services, while community health services gained by 1% in Greater Southern. South Eastern Sydney & Illawarra AHS community based funding remained the same at 16% for 2006–07, similar to the previous year. On the other hand, Sydney West, Northern Sydney Central Coast and Hunter & New England AHSs maintained the same hospital and community–based funding, although Sydney West community health services had a 2% increase compared to 2005–06. The total expenditure for teaching and research remained between 0.3% and 7% with only a 1% increase or decrease in funding for most AHSs. Sydney West, however, had a sizeable drop from 6% in two successive years to 3% in 2006–07. Three out of four rural and remote based AHSs allocated a negligible percentage of funding to teaching and research compared to the metropolitan areas (see Table 7–3 and Figure 7–1).

241 ______

Table 7–3: Area Health Services Community Health, Hospital Services and Teaching & Research Expenditure, 2004/05 to 2006/07

Figure 7–1: Area Health Services Community Health, Hospital Services and Teaching & Research Expenditure: 2004/05 – 2006/07 90%

80%

70%

60%

50%

40%

30%

20% Programs Expenditure (%) Expenditure Programs

10%

0% 2004/05 2005/06 2006/07 2004/05 2005/06 2006/07 2004/05 2005/06 2006/07

Community Health Services Hospital Health Services Teaching & Research Activity

SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS

242 ______

7.5 Area Health Service Programs and Services per Capita Expenditure This section presents the intended aims of the ten programs and services and their per capita expenditure within each program, comparing the eight Area Health Services (AHSs). The analysis was undertaken to determine the amount of funding allocated to each health program per population per year and, although the aim of this section was to examine how financial resources received by AHSs from the state level were distributed, it was also expected to highlight whether equity had been promoted in terms of benefiting the most disadvantaged population groups. However, a close look at the mechanisms for intra–Area Health Services resource allocation from 2004/05 to 2006/07 revealed a tendency to distribute resources according to where services are located rather than on the basis of differential health needs. As noted in section 7.3, around 80% of the AHSs budget was shared between hospitals in each AHS, regardless of differences in health needs among the population groups. The remaining 20% was distributed to community–based health programs (see Table 7–3).

According to the NSW Department of Health, the allocation of resources based on programs and services is essential since different rules apply to each health service program. In other words, the Resource Distribution Formula (RDF) is actually a set of mini –models built around the programs, each of which applies to a different set of health service needs. Apparently, this involves applying the measures of need to each of the components in the model, in order to split the total budget for the components among the Area Health Services (AHS). NSW Health claims to select the best measures of population and the predictors of per capita variations in need across the AHSs’ populations. It then proposes funding shares for each AHS determined by summing the allocations across the service program components (the aims and objectives of the programs are explained in the next sections) of the RDF, and adding the allocations that originate outside the RDF (that is, from budget elements that are excluded like cross–boundary flows) (NSW Health, 2006c).

Using the amount of funding allocated to each program within the Area Health Services (AHSs), this study calculated the per capita expenditure for all eight AHSs, comparing the

243 ______trend over three years (2004/05 to 2006/07). This type of analysis has not been undertaken before and it will, therefore, add value to what information already exists. It is noteworthy that the AHSs are expected by the NSW Department of Health to report back on the total amount allocated to each of the programs or services and not how the funds were shifted around to benefit various population groups. Table 7–4, shows the per capita expenditure of all the four services that fall under the study categorisation of community–based programs. The aims of the four services – Primary and Community; Aboriginal Health; Rehabilitation & Extended Care; and Population Health – are explained in the next sections.

7.1.8 Per Capita Expenditure of Primary & Community Health Services According to NSW Health, the primary & community health services within each Area Health Service (AHS) play a very important role in improving, maintaining or restoring the health of the population through health promotion, early intervention and assessment, including therapy and treatment services for clients in a home or community setting (NSW Health, 2008). Thus funding to these services would be expected to be fairly large so that they could fulfil their preventative and early intervention roles among others. However, based on the analysis in this study, most AHSs allocated less than $150 per capita annually, with only one AHS exceeding $150 per capita expenditure for the three years assessed.

As shown in Table 7–4, Greater Western AHS (GWAHS), classified as rural and remote based, spent the highest per capita amount on primary and community health services, around $182 in 2004/05, increasing the amount to $189 in 2005/06; however, in 2006/07 this was reduced to $174, which was still relatively high compared to other AHSs. This higher amount can be explained by the higher level of health needs among the population groups in GWAHS, as observed in Chapter 6.

244 ______

Table 7–4: Per Capita Expenditure of Community–based Programmes, 2004/05 to 2006/07

Figure 7–2: Per Capita Expenditure of Primary and Community Health–based Health Services: 2004/05 – 2006/07

450

400

350

300

250

200

150

100 Per Capita Expenditure $ 50

0 2004/05 2005/06 2006/07 2004/05 2005/06 2006/07 2004/05 2005/06 2006/07 2004/05 2005/06 2006/07

Per capita Per capita Per capita Per capita

Primary & Community Aboriginal Health Rehab & Extended Care Population Health

Community-based Health Programs

SSW AHS SESI AHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS

245 ______North Coast, another rural based AHS, allocated a fairly large per capita expenditure of around $136 in 2004/05; this was followed by $145 in 2005/06 but declined to $139 in 2006/07. The other AHS that spend over $130 per person per year on primary and community health services was Sydney West, spending $134 in 2004/05, declining to $132 in 2005/06 and increasing to $140 in 2006/07. Sydney South West followed with a per capita expenditure of $102 in 2004/05, and slightly increasing this to $116 in both 2005/06 and 2006/07. In contrast, Greater Southern AHS (GSAHS) and Hunter & New England AHS’ (H&NEAHS) per capita expenditure for 2004/05 was only $98 and $58 respectively. However, both GSAHS and H&NEAHS raised their per capita expenditure in 2005/06 and 2006/07 to over $100 (see Table 7–4). The AHSs that spent the least per capita for the entire three years analysed were South Eastern Sydney & Illawarra and Northern Sydney Central Coast as illustrated in Figure 7–2.

7.1.9 Per Capita Expenditure of Aboriginal Health Services According to NSW Health (2008), the aim of Aboriginal health services funding is to raise the health status of Aboriginal and Torres Strait Islanders, including promoting a healthy life style among the approximately 150,000 Aboriginal people residing in NSW, where they make up just 2% of the total NSW population. Funding for Aboriginal health services is critical since the Aboriginal people in Australia and specifically NSW continue to experience poor health outcomes and relative socioeconomic disadvantage. The inequities that exist between the Aboriginal population’s health and the non–Aboriginal are great. For example, in NSW Aboriginal people are more likely to die at younger ages and those aged less than 25 years make up around 10% of deaths of Aboriginal people, compared with 2% of deaths among non–Aboriginal people. The infant mortality rate for babies born to Aboriginal mothers is 7.5 per 1,000 births, almost twice the rate for all NSW babies (NSW Health, 2006a).

The leading causes of death for Aboriginal people are the same as for non–Aboriginal people, cardiovascular disease and cancer. However, Aboriginal people are more than twice as likely as non–Aboriginal people to die as a result of diabetes or from injuries. The hospital admission rates for Aboriginal people are 1.7 times the rate of non–Aboriginal

246 ______people. Renal dialysis accounts for the largest number of hospitalisations for Aboriginal people. Compared with the rates for non–Aboriginal people, hospitalisation rates for Aboriginal people in NSW are 140% higher for conditions for which hospitalisation could be avoided by prevention and early management (NSW Health, 2006a). Therefore, considering the high level of need among the Aboriginal population, it is important that AHSs maintain the Aboriginal health program to try and bridge the inequalities and inequities that currently exist when comparing the health outcomes of the Aboriginal and non–Aboriginal populations.

Table 7–4 and Figure 7–2 illustrate the per capita expenditure on Aboriginal health services: in 2004/05, for example, Greater Western was the only AHS that spent around $343 per capita on Aboriginal health services, followed by North Coast spending $276, Greater Southern $265 and South Eastern Sydney & Illawarra $253 in the same year. This was followed by Northern Sydney Central Coast spending $187, followed by Hunter & New England with a per capita expenditure of as little as $66 and Sydney West a minor $56 per capita in 2004/05. In 2005/06 and 2006/07, some areas had a significant increase in per capita expenditure. For example, North Coast had a consistent growth in per capita expenditure in the entire three years while other AHSs reduced their per capita spending. South Eastern Sydney & Illawarra reduced their per capita expenditure from $253 to $182 in 2005/06 and slightly increased this to $192 in 2006/07. Northern Sydney Central Coast is another AHS that reduced its per capita spending substantially from $187 in 2005 to $95 (2005/06) and had a minor increase to $97 in 2006/07. In general, Aboriginal health has been one of the most debated areas in terms of how much per capita should be spent in order to bridge the gap in health outcomes between Aboriginal and non–Aboriginal population groups. As demonstrated by Table 7–4 and Figure 7–2, there is no consensus in terms of how much each AHS should spend on Aboriginal health services or per capita in order to improve the health status of Aboriginal populations.

247 ______7.1.10 Per Capita Expenditure of Rehabilitation & Extended Care Services According to NSW Health (2008), the main objective of rehabilitation and extended care services is to improve the well –being and independent functioning of people with disabilities or chronic conditions, including the frail and the terminally ill within AHSs. As shown in Table 7–4 and Figure 7–2, some AHSs spent more than others on rehabilitation and extended care services based on per capita. For example, Greater Western AHS spent the highest amount, around $263 in both 2004/05 and 2005/06 but reduced it to $217 per capita in 2006/07. This was followed by Greater Southern AHS with a per capita expenditure of $198 in 2004/05, which was increased to $215 in 2006 but decreased to $205 by 2006/07. Third in terms of per capita expenditure on rehabilitation and extended care services was North Coast, followed by South Eastern Sydney & Illawarra AHS. North Coast spent $143 in 2004/05 but reduced the amount to $140 in 2005/06; by 2006/07 this had dropped to $133. The areas that had minimal per capita expenditure in all the years analysed were Sydney South West, followed by Sydney West and Northern Sydney Central Coast AHSs (see Figure 7–2).

7.1.11 Per Capita Expenditure of Population Health Services In NSW the population health services aim to promote the health of the population and reduce the incidence of preventable diseases and disability by improving access to opportunities for good health. However, despite population health services being crucial in terms of prevention of diseases, none of the Area Health Services (AHSs) allocated more than $60 per capita. Indeed, only two AHSs out of eight had more than $40 per capita expenditure as illustrated in Figure 7–2. Funding for population health services was not consistent across the AHSs. For example, the only two Areas to allocate over $40 per capita were Sydney West and Greater Western followed by North Coast which spent over $30 for two consecutive years while the rest of the AHSs spent under $20 per capita for most of the years, except for South Eastern Sydney & Illawarra which allocated $27 in 2004/05 only. Greater Southern AHS had the least per capita expenditure for two successive years as shown in Figure 7–2.

248 ______7.6 Per Capita Expenditure of Hospital–Based Programs 2004/05 – 2006/07 This section presents the per capita expenditure for the category of hospital–based health services which include Outpatient services, Emergency services, Overnight Acute services, Same Day Acute services and Mental Health services as shown in Table 7–5 and Figure 7– 3.

7.1.12 Outpatient Services per Capita Expenditure 2004/05 to 2006/07 The Outpatient health services are hospital based and the aim of service is to improve, maintain or restore health through diagnosis, therapy and education, including treatment services for those patients who arrive by ambulance in a hospital setting. Based on the importance of this service, most Area Health Services (AHSs) allocated a fairly large per capita sum compared to other health programs. In 2004–05, Sydney South West spent around $189 per capita on Outpatient services; however the amount was reduced to $110 in 2005–06 and then increased slightly to $116 in 2006–07. South Eastern Sydney & Illawarra also spent a large amount: in 2005–06 its per capita expenditure was $181, which was increased to $212 in 2005–06 and to $225 in 2006–07. They ranked first in terms of the largest per capita expenditure in all three years.

Sydney West and Greater Western were the only two AHSs that were consistent in terms of their per capita expenditure for Outpatient services. For example, Sydney West spent between $120 and $130 for the three years and Greater Western’s per capita expenditure was $168 in both 2004–05 and 2005–06; however, in 2006–07 the amount was slightly reduced to $153. The areas that spent under $100 per capita were Greater Southern, North Coast and Northern Sydney Central Coast (NSCC) in all the three years, although NSCC had a modest increase in expenditure from $97 in 2005–06 to $104 2006–07. Hunter & New England was the AHS that showed the greatest increase, from $72 per capita in 2004–

05 to $158 (2005–06) and $172 in 2006–07.

249 ______

Table 7–5: Per Capita Expenditure of Hospital–based Programmes, 2004/05 to 2006/07

Figure 7–3: Per Capita Expenditure of Hospital–based Health Programs: 2004/05 – 2006/07

1000

900 800

700 600

500 400

300 200

100 Per Capita Expenditure $ 0 2004/05 2005/06 2006/07 2004/05 2005/06 2006/07 2004/05 2005/06 2006/07 2004/05 2005/06 2006/07 2004/05 2005/06 2006/07

Per capita Per capita Per capita Per capita Per capita

Outpatient Emergency Overnight acute Same day acute Mental health

Hospital-based Health Programs

SSW AHS SESI AHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS

250 ______

7.1.13 Per Capita Expenditure of Emergency Care Services The emergency care services are hospital based and are designed to reduce the risk of premature death and disability for people suffering injury or acute illness by providing timely emergency diagnostic, treatment and transport services. As shown in Table 7–5 and Figure 7–3, most AHSs, including Sydney South West, South Eastern Sydney & Illawarra, Sydney West and Greater Southern, all had over $100 per capita expenditure for emergency care services in all the three years analysed, while Northern Sydney Central Coast spent only $56 in 2004–05 but increased this to over $100 in the other years. However, Hunter & New England, North Coast and Greater Western AHSs’ per capita expenditure was below $100 for all three years.

7.1.14 Per Capita Expenditure Overnight Acute Services NSW Health explains the objective of the Overnight Acute services as restoring or improving health and managing risks of illness, injury and childbirth through diagnosis and treatment for people intended to be admitted to hospital on an overnight basis. The programs Overnight Acute services attracted the largest per capita expenditure compared to other programs, with most Area Health Services (AHSs) spending over $600. The Greater Western AHS (GWAHS) once again led in per capita expenditure and for this program especially spent around $788 in both 2004/05 and 2005/06; this was substantially increased to $881 in 2006/07. Greater Western is classified as a rural and remote AHS with a population of around 0.3 million who are mainly socially and economically disadvantaged, including having pockets of population with poor health outcomes. The next area to spend high proportions of per capita funding was Greater Southern (a rural area), spending $698 in 2004/05 with an increase to $842 in 2005/06, but declining to $817 in 2006/07. This was followed by Sydney South West AHS, which spent around $694 in 2004/05, slightly increasing to $743 in 2005/06 and $759 in 2006/07.

North Coast is another rural based Area Health Service (AHS) that spent a large per capita sum for the entire three years ($645 in 2004/05; $772 in 2005/06; $858 in 2006/07) with sizeable increases. South Eastern Sydney & Illawarra also had a large per capita

251 ______expenditure of $632 in 2004/05, rising to $701 in 2005/06 and $747 in 2006/07. Northern Sydney Central Coast, on the other hand, spent $645 in 2004/05, $688 in 2005/06 and $729 in 2006/07. The area with the least per capita expenditure was Hunter & New England, spending $299 in 2004/05, the smallest per capita expenditure in all the three years for the eight AHSs. However, this was noticeably raised to $663 in 2005/06 and to $720 in 2006/07. Finally, Sydney West’s per capita expenditure was $582 in 2004/06, increasing to $645 in 2005/06 but declining to $562 in 2006/07. In general, there were relatively minor shifts in per capita expenditure for most of the AHSs when allocations from 2004/05 to 2006/07 were compared (see Figure 7–3).

7.1.15 Per Capita Expenditure Same Day Acute Services The purpose of the Same Day Acute services in all the Area Health Services (AHSs) in NSW is to restore or advance health and manage risks of illness, injury, and childbirth through diagnosis and treatment for people intended to be admitted to hospital and discharged on the same day. As shown in Table 7–5 and Figure 7–3, some AHSs spent more on this service per capita than others. For example, North Coast’s per capita expenditure was the highest for this program, spending $165 in 2004/05; however, this was reduced to $134 in 2005/06 but gradually increased to $159 in 2006/07. Greater Western was the second in terms of per capita expenditure, spending for the entire three years $130 in both 2004/05 and 2005/06 and increasing this to $150 in 2006/07. Greater Western AHS also spent a comparatively large amount, between $100 and $140, in the three years analysed. Most urban classified AHSs spent a much smaller per capita amount compared to the rural and remote areas, with Northern Sydney Central Coast AHS allocating the least amount as shown in Figure 7–3.

7.1.16 Per Capita Expenditure Mental Health Services According to NSW Health, the aim of mental health services in each Area Health Service (AHS) is to improve the health, well being and social functioning of people with disabling mental disorders and to reduce the incidence of suicide, and mental health problems, including mental disorders, in the community. Figure 7–3 shows the levels of per capita

252 ______expenditure among the eight AHSs in NSW. The three –year analysis revealed that GWAHS, as with most other programs and services, had the highest per capita expenditure for mental health in the entire three years ($171 in 2004/05 and 2005/06, increasing the amount to $212 in 2006/07). Hunter & New England followed closely, although it spent only $75 per capita in 2004/05 but this was increased to $148 in 2005/06 and $151 per capita in 2006/07. The other AHSs to exceed over $100 per capita expenditure for the entire three years were NSCCAHS, SSWAHS, NCAHS and SWAHS. Conversely, the AHSs to spend the least per capita on mental health services were SES&IAHS ($78 in 2004/05; $107 in 2005/06; and in 2006/07 a decreased amount of $94) and GSAHS with a per capita expenditure for the three years of $89 in 2004/05, $97 in 2005/06 and a slight increase to $113 in 2006/07 (see Figure 7–3).

7.7 Teaching and Research Total Expenditure Teaching and research funding is, according to NSW Health, allocated to AHSs to help develop the skills and knowledge of the health workforce in order to support patient care and population health. It is also intended to extend knowledge through scientific enquiry and applied research aimed at improving the health and well being of the people of New South Wales. As shown in Table 7–6, the yearly expenditure on teaching and research for most of the urban –based AHSs was greater than for the rural and remote based areas. For example, SES&IAHS spent a substantial amount of over $110 million for the entire three years ($116 in 2004/05, $122 in 2005/06 and $145 million in 2006/07). Other AHSs that spent a fairly large amount per year on teaching and research were SSWAHS ($98 in 2004/05, $103 in 2005/06 and $102 million in 2006/07) and SWAHS ($88 in 2004/05, $90 in 2005/06, decreasing to $41 million in 2006/07). NSCCAHS had the lowest expenditure in this category among the four urban AHSs, spending $39 in 2004/05; however, this was increased to $60 in 2005/06, and to $63 million in 2006/07 (see Table 7–6).

253 ______Table 7–6: Area Health Services Teaching and Research Expenditure, 2005 to 2007 Area Health Teaching & Research Teaching & Research Teaching & Research Services 2004–05 Exp ($M) 2005–06 Exp ($M) 2006–07 Exp ($M) SSWAHS 98 103 102 SES&IAHS 116 122 145 SWAHS 88 90 41 NSCCAHS 39 60 63 H&NEAHS 19 38 42 NCAHS 11 11 10 GSAHS 3 3 3 GWAHS 6 4 3

On the other hand, the rural and remote Area Health Services (AHSs) generally spent under $20 million in the entire three years; the only rural AHS to spend over that amount was H&NEAHS, spending $19 million in 2004/05, raising this to $38 million in 2005/06 and to 42 million in 2006/07. The areas that had minimal expenditure were GSAHS, spending $3 million in all the three years, followed by GWAHS with a teaching and research expenditure of $6 million in 2004/05; this declined to $4 million in 2005/06 and to $3 million in 2006/07 (see Table 7–6 and Figure 7–4). In general, despite the fact that teaching and research funding is vital in terms of helping develop the skills and knowledge of the health workforce to facilitate better support for patient care and population health, it received very little attention in terms of funding across all the AHSs in NSW.

Figure 7–4: Comparison of Area Health Services Teaching and Research Expenditure, 2004/5 – 2006/07

160

140

120 100

80

60 40 Expenditure $M 20 0 SSWAHS SES&IAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS

Area Health Services

Teaching & Research 2004-05 Exp ($M) Teaching & Research 2005-06 Exp ($M) Teaching & Research 2006-07 Exp ($M)

254 ______Chapter 7 Section 2: Summary of Quantitative Findings Funding from the State to the eight Area Health Services (AHSs) is allocated through ten programs and services. According to NSW Health, funding for these programs is calculated using a resource distribution formula (see Chapter 6).

The evidence from the qualitative data showed that the central level (NSW Health), in spite of decentralisation, still controls the resource allocation process within the AHSs.

The quantitative evidence indicates that the bulk of funding within the AHSs, approximately 80%, is consistently allocated to hospital–based health programs and services, while around 20% goes to community–based health services, with no significant changes in the distribution pattern for the three years analysed (2004/05 to 2006/07).

The analysis did not find any evidence of equity or need factors as the driving force of resource allocation within the AHSs. Funding is allocated based on where programs and services (i.e. hospitals and community health centres) are located rather than on the basis of health needs by differentially targeting those identified population groups with resources and services.

This type of allocation poses many problems for health officials at the front line of service provision, as explained in Section 1 of this chapter.

Programs such as Aboriginal Health received an insignificant amount of per capita funding by all AHSs despite the extra weight applied to RDF at the central level to ensure AHSs received more funding for Aboriginal population to help reduce the inequities in health and health outcomes. According to the qualitative data, it is not compulsory for AHSs to apply the same criteria when it comes to Aboriginality at the AHS level.

Population health and teaching and research also received very little funding across the eight AHSs, especially within the rural and remote based AHSs.

255 ______

Chapter 8: EQUITY IN RESOURCE ALLOCATION WITHIN AREA HEALTH SERVICES

Overview This chapter addresses the third research question: Is equity reflected in the allocation of resources at the Area Health Service level? There is a major problem in addressing this question. As will be seen from the analysis of interview material in this chapter, AHS officials are not required to report in “geographic” detail on their allocation of the financial resources they receive from the central authority (the NSW Department of Health). Consequently there is no readily available information as to how the available funds are shared between local government areas. In an effort to gain some insight into of the degree to which equity is reflected in the inter–LGA distribution of resources, a three step procedure was adopted.

The chapter is divided into three steps: the first step presents the findings of the qualitative inquiry which examined the opinions on equity in resource allocation of NSW Health and Area Health Services (AHSs) executives, program managers and consumers to determine whether equity has been reflected in the allocation of resources at the AHS level. The second step presents the results of a quantitative inquiry that assessed the health needs of the Local Government Areas (LGAs) within Northern Sydney Central Coast and Greater Western AHSs to examine what an equitable allocation of resources might look like.

The third step is the report on a mapping exercise conducted to identify the number of health facilities and services (such as hospitals, community health centres and other services) offered in two selected Area Health Services, thus providing a picture of the geographic distribution of health service facilities.

256 ______STEP 1: Qualitative Findings

8.1 Opinions on Equity and Resource Allocation The NSW Department of Health publications state that equity is one of the factors taken into consideration at the state level when allocating health funds to the eight Area Health Services (AHS) using the Resource Distribution Formula (see Chapter 1). However, it is not known if similar equity–focused concerns drive resource distribution at the AHS level. Therefore, this chapter attempts to assess whether the equity–focused resource allocation policies applied at the State level have been applied in the allocation of NSW Health derived resources within two AHSs. The perspectives on funding and equity of NSW Health and AHS executives, program managers and health professionals who have the duty to use these resources to improve population health, and the views of some community representatives and patients who access the services were explored in face–to–face interviews as explained in Chapter 5.

The interview data was analysed using NVivo. Six main themes emerged from the analysis: 1) confusion regarding the equity principles guiding Area Health Service resource allocation; 2) lack of openness in funds distribution; 3) limited capacity to benefit from resource allocation and health services; 4) the need for custom–made models for distributing resources in each AHS; 5) lack of partnership between multiple service providers; and 6) restricted access to health services.

8.1.1 Confusion regarding Equity Principles Guiding AHS Resource Allocation Respondents at all levels of the NSW heath care system had different views about equity in resource allocation, including equity–focused policies aimed at differentially benefiting the populations in greater need. As noted also in Chapters 6 and 7, most of those interviewed expressed a need for a resource allocation strategy, signifying a lack of clear understanding about how resources are currently allocated. According to most of the respondents, equity within Area Health Services had not been achieved, although they were aware of equity as a necessary factor that should guide the resource allocation processes. The respondents identified various factors that impede their effort to promote equity. Further, the lack of

257 ______effective tools to identify population health needs and to distribute health funds according to those needs were some of the issues raised by most of those interviewed at the AHS level.

“Funding distribution within our Area is based on programs and services and not really targeting population with high needs… so in that sense equity has not been achieved. To achieve equity we have to demonstrate that we are targeting resources and services rather based on historically decided upon health programs most of which are provided within hospital setting…” (Senior Health Official, AHS)

“Equity has not been achieved in this AHS because resources are not allocated according to needs of population groups or Local Government Areas [LGAs] and neither do we have defined policies of how we should apply equity…I think this is why it does not differentially benefit the most disadvantaged…resource distribution is purely based on where the hospitals are located and some historical context including how those hospitals work in the community… (Senior Health Official, AHS)

“Equity in distribution of health dollars is yet to be achieved because population who are disadvantaged in terms of not benefiting from resources and health services are still disadvantaged. Funds distribution in most Areas is like a really great iceberg, in that there is huge resources at the bottom of the iceberg and the tip of the iceberg which is where the greatest need is has got less than fair share distribution of resources based upon historical context but not directly based on population needs…in that sense we will continue to disadvantage those most disadvantaged unless we have a method of redistribution across the Areas by taking resources from hospitals to communities which I think will be fairer and more equitable…” (Senior Health Official, AHS)

“Equity has not been achieved in this Area because politics determines where resources are distributed…to achieve equity we need to reduce politics involved in where we put the health dollars” (Senior Health Official, AHS)

According to the participants, there is a willingness to distribute resources to the areas of greatest need; however, the absence of clearly defined equity principles on which to base resource allocation at the population level has left most areas unable to deal with existing inequities. And according to a respondent;

258 ______“…equity in resource allocation means as the phrase goes ‘is to find a way of distributing resources in the right place at the right time and for the right reasons’. In the sense, one has to utilise whatever tools one thinks are the best tools to identify the population health needs and other underlying factors and then allocate health dollars accordingly…that’s what I would call equitable and best practice by international standards, guidelines and evidence based practice something which we [AHS] don’t do” (Senior Health Official, AHS)

The issue of having a service delivery model that is focused primarily on hospitals was also criticised by respondents as consuming most of the health funds and was a disincentive to target resources to populations with greater needs and those who were disadvantaged. The implications of under–funding of AHSs by the NSW Department of Health were also discussed by the respondents. The issue of inadequate funding was articulated by a majority of those interviewed at all levels of the health care system as impinging on the frontline service provider’s efforts at advancing equity and differentially benefiting those in greater need.

“…to make change in a system to do greatest good for most of the disadvantaged population as you can, you need a good funding distribution mechanism and quality improvement model…But instead our system is all about dealing with one hospital crisis after the other because we don’t have a clearly defined plan to allocate health dollars or improving quality of services we provide or even how to apply equity…” (Senior Clinician, AHS)

“…we need to develop resource distribution mechanisms to encourage distribution of growth funds to areas of greatest need to help reduce the inequities…there would also be benefits in the consolidation of resources in few major hospitals and less small hospitals to reduce the inefficiencies that we currently experience…the resources recovered can be put in better use” (Senior Health Official, AHS)

As also noted by respondents, inefficiencies exist in resource allocation at the local level due to the number of hospitals in NSW. According to the participants, some of the inefficient hospitals could be consolidated to create larger and more comprehensive centres that have the capacity to offer quality services. The health professionals thought that the closure of some hospitals deemed inefficient would release funds that could be diverted to community health services and for targeting population groups with greater health needs.

259 ______

“…there are too many hospitals and too many operating theatres taking all the resources yet not providing services enough to meet the needs of the population… we need to merge some hospitals into larger and fewer centres to reduce the inefficiencies in funding and refocus those resources to community health centres and for meeting the needs of the population especially those disadvantaged…” (Senior Health Official, AHS)

While health officials confirmed that inefficiencies in funding exist at the local level, the Resource Distribution Formula (RDF) was also criticised for its inability to distribute resources according to the health needs of AHSs’ populations. There was a general consensus that the RDF needs to be enhanced to ensure that it is an effective tool that captures AHSs’ population health needs.

“The RDF [used by NSW health] has to be improved according to needs of the population in order to allocate this Area enough resources to do be able to target disadvantaged populations…we also need to improve the inefficient system to be to make some gains although they may not last a life time but it will be a starting point into better health of the populations that are currently disadvantaged…” (Senior Health Official, AHS)

8.1.2 Lack of Openness in Funds Distribution The lack of financial accountability systems at the population level was a concern expressed by the respondents. The participants believed that if the Chief Executives entrusted with the financial responsibility are not accountable in terms of how the funds are spent, then the system will remain inefficient and not capable of improving the health of the most disadvantaged population groups. It was evident from the interviews that the system of financial accountability was largely driven by hospitals’ efficiency and there seemed to be very little about equity. Participants felt that, at the population level, there was little concern about the distribution of funds to community health centres; the emphasis was on hospitals. However, health professionals thought this historical structure of focusing a higher percentage of health resources on hospitals ought to change; community health centres needed to be integrated and more resources offered to the most disadvantaged population groups.

260 ______“…each AHS has got the responsibility to it’s population and the Chief Executive has got the responsibility to NSW health and the Government to use resources wisely and if he or she was truly accountable then I would guarantee that there would be a plan, a strategy to diminish the long term commitment to capital by reducing the number of hospital facilities and then try and balance community needs by directing more funding to primary health community centres so as to improve the health of those in greatest need...this will also reduce hospital services demand. But as long as the AHSs remain unaccountable then we are not likely to see much change in these issues especially achieving equity...” (Senior Health Official, AHS)

According to some policy makers, the perceived lack of accountability and inefficient use of resources is also linked to the poor structure of reporting on how the funds received from the state level are distributed to meet the needs of the population. The standardised system of reporting leads largely to lack of transparency on the part of the Chief Executives regarding how much funding is distributed to community health centres. The interviews also revealed that the lack of effective monitoring and evaluation systems in the NSW health system could be affecting those most disadvantaged since its not known how they benefit from resources and services provided.

“There are no measures like monitoring and evaluation to ensure accountability for the resources given to Areas because the system is decentralised…Chief Executives are expected to report in the annual reports by including an audited financial report but this does not tell us who benefited from the health dollars or even how much was allocated to what programs especially those targeting disadvantaged communities” (Senior Health Official, AHS)

According to NSW Health, the AHSs have the flexibility of moving funds within and between health programs. But, based on the method of allocating resources to specific programs with many performance targets, are AHSs able to actually shift funds between programs? When the AHSs report on how resources were spent, they do not indicate nor are they expected to provide a detailed account of the actual movement of funds. The lack of a system to monitor and evaluate the use of resources was also a weakness identified by the respondents at the population level.

“AHS do not include how wider the funding has been allocated to meet their population health needs in their annual reports. The reporting system to NSW Health does not encourage the Areas to report according to how the funds were

261 ______spent and what health programs benefited from the health dollars. For equity to be promoted it’s important to know who does or does not benefit from resources” (Senior Health Official, AHS)

“At the population level there is no funding distribution mechanism to help assess health needs and shift funds according to those needs. After amalgamation of the Areas the concern is that the historical inequities that have always existed within the old Areas are going to be forgotten and the gap will widen while the new Areas struggle to provide services to much bigger population...rural and remote areas like GWAHS has to micro–manage budgets and actions from a base which is 850km from the regional centre and so they are more disadvantaged and their performance indicators should reflect all this and other factors that have an influence…” (Senior Health Official, AHS)

As explained by some participants, the lack of a database to collate population data also means that the health funds are not allocated according to health needs. There was general consensus among those respondents that data collection needs to be improved so as to shed light on the changing needs of the population. The data can also be used to highlight health outcomes and performance in the different Area Health Services (AHSs), since performance indicators suitable for urban areas will not necessarily be appropriate for rural and remote AHSs.

“Due to decentralised nature of the NSW health system, the Department of Health feels that AHSs are in a position to assess their population health needs and allocate funding according to those needs but the AHSs don’t have a mechanism to measure health needs or a database to collect data at a population level and so how can they allocate health dollars according to need if they don’t have data showing the levels of need? Most of the health dollars are allocated to hospitals” (Senior Health Official, AHS)

The other problem that dominates health funding in NSW Health is the failures experienced by hospitals, leading to what is often known as “hospitals crisis”. In many instances when a crisis is reported it leads to substantial resources being shifted to deal with the crisis and to appoint a commission of inquiry to investigate the failures; this might mean that other essential services are under–resourced as a result, according to the respondents. One recent hospital crisis was at the Royal North Shore Hospital, which led to an investigation through a Commission of Inquiry headed by Professor Garling. The commission produced a

262 ______detailed report known as the Garling Report with a 139 recommendations. Most of the factors reported in this thesis as influencing resource allocation and equity in NSW were highlighted in the Garling report (Garling, 2008)

“The hospitals crisis generally experienced by NSW hospitals every other month takes the attention and resources away from community health services and disadvantaged population…funding community health services would be of use in reducing hospital services demand which often leads to some of the malpractices…to reduce the hospital crisis we have to increase funding to community based services to make sure patients with non–complex health problems can be treated or managed in community health centres…” (Senior Health Official, AHS)

8.1.3 Limited Capacity to Benefit from Resource Allocation and Health Services The capacity to benefit from funds and heath services is another factor that emerged from the interview data. The two Area Health Services (urban and rural/remote) clearly had different challenges in terms of their capacity to utilise resources allocated. For example, Greater Western had more hurdles when it came to health resources’ utilisation compared to urban Northern Sydney Central Coast AHS. Other problems related to how population groups benefit from services provided and the utilisation of resources in rural areas due to staff shortages.

“…with preventive care funding we are doing very badly in NSW and Australia because our main focus is more curative. For example, GPs are meant to provide primary health care but they are not funded for preventative medicine. They are funded to assess, prescribe treatment and cure although they contribute to health promotion activities through variety of means…The message about benefits of eating vegetables, the pyramid of eating healthy, exercise, moderate alcohol and not smoking is out there…but diabetes is an epidemic despite all that. We are one of the wealthiest countries and our Area has got some of the wealthiest parts of Sydney and we got chronic diseases like diabetes coming out of our armpits and everywhere, so the message is out there but we are failing. The messages are clear but it’s how we deliver those messages which is the problem…it’s more to do with how we manage our society more than how we manage the health messages…to make a difference all depends on the social society we are living in, there is probably more diabetes in Northern Sydney because of certain behaviours than Dubbo” (Senior Health Official, AHS)

263 ______It was also evident from the interviews that differences existed across AHSs’ population groups in terms of their capacity to utilise services requiring a tailored approach in providing services.

“Due to population being multicultural and differences in levels of capacity to understand and utilise the services, we need to have a more tailored approach towards all health information and health services to ensure each population group has a capacity to benefit from health funds and services we provide” (Senior Health Official, AHS)

“...health services are scarce in some rural and remote part of our AHS because health professionals are not keen to work in rural and remote areas and there are few GPs in these parts like Broken Hill…service providers I think should be regulated through provider numbers by either the state or federal government to improve services in rural and remote parts in order to increase access to services within the communities…inadequate access to services is always caused by lack of health professionals interested to work in rural parts…” (Senior Health Official, AHS)

Respondents emphasised the importance of addressing cultural issues when training health professionals.

“Social and cultural sensitivity is required when training health professionals to make sure they are aware when providing health service…” (Senior Health Official, AHS)

Other health professionals referred to the differences in capacity between urban and rural AHSs to benefit from resources due to lack of appropriate infrastructure and manpower shortages. Those interviewed expressed the need for better incentives to attract qualified health professionals and a plan for how to train and retain the health workforce, especially in rural and remote parts of NSW. According to the participants, ideally the training should also incorporate cultural competencies in order to equip health professionals not just with health–based knowledge but also with the social and cultural customs of the diverse population groups they serve. Different models of resource allocation, service delivery, and health professional training for urban and rural areas need to be formulated and implemented according to the individual needs of each AHS; the ‘one–size–fits–all’ approach will not work as observed by one respondent.

264 ______

“…in NSW different AHSs have different capacity to benefit from health dollars especially rural and remote areas…Our Area [GWAHS] which is rural and remote experience severe difficulties due to environment, infrastructure and staff shortages…health professionals are not willing to work in remote parts but if incentives were good some would take up the jobs…this is something we have been discussing with policy makers” (Senior Health Official, AHS) “GWAHS population is also multicultural with large number of Aboriginal and Torres Strait Islanders…therefore the services have to be sensitive to the cultural differences… meaning we need more funding to train staff to be culturally aware and to provide the necessary services closer to communities since some can not travel to access centralised services. So if you take all these factors on board we as an Area have less capacity than urban areas to benefit and what this means is that the ‘one–size–fits–all’ approach can not be applied in both urban and rural environments…” (Senior Health Official, AHS)

8.1.4 Need for Custom–Made Models for Distributing Resources in each AHS A recurrent theme was the need for a resource allocation model especially tailor–made for each Area Health Service (AHS) in NSW.

“…each AHS has to have a tailor–made model to meet it’s specific population needs because in rationale term Northern Sydney might have a solution to that because it’s in metropolitan area but you can’t take the same model and apply it to Dubbo or Broken Hill in Greater Western, because the hospitals out there are part of the town’s culture and fabric. It’s very difficult to reshape them because they are geographically in remote part of NSW. So AHSs need different models to suit geographical issues in terms of distance and travel are some of the real issues there unlike North Sydney, which is within the metropolitan area. Greater Western AHS struggle very much with many issues especially staff shortages; training and retaining them is a big issue and I think NSW Health should introduce good incentives to help retain staff in the regional and remote areas…” (Director, AHS)

According to officials in Northern Sydney Central Coast AHS, a formula that would improve how resources are allocated within AHSs should look at grouping regions into clusters, consisting of a number of Local Government Areas (LGAs). This would also enhance the planning of health services and how they are provided to different population groups.

265 ______“We [NSCC] plan to build a resource distribution model tailor–made for each of the four sub–areas (clustered LGAs) when there is data at that population level which can be aggregated in order to capture population health needs. Once the model is ready we can provide funding shares according to those clusters but the difficulties is while the demographic data should be very easy to get at that level it’s more of the financial data that is generally difficult for areas to provide because they are not required by NSW Health to provide” (Senior Health Official, AHS)

8.1.5 Lack of Partnership between Service Providers Some interviewees expressed the expectation that, given the limited health resources, service providers would be willing to collaborate with others to maximise the use of resources as well as to improve management of patients. However, as explained by health professionals and patients, most service providers come either under the State or Commonwealth (federal) governments or the private sector. For example, the NSW Department of Health falls under the NSW State government while the general practitioners (GPs) practising in NSW report directly to the Commonwealth government. The lack of partnership and shared patient management information might have contributed to the escalating burden of chronic diseases in NSW.

“Our Area has real difficulties when we try to discuss the need to work in partnership with other organisations providing health services…some organisations have stronger interests and political power to win funds and even influence policy direction and so they never want to engage in any sort of collaboration…and what this means is that various organisations are funded to deliver similar services leading to a culture of duplication and unfair competition for the little resources available…therefore for these organisations the funding submissions are a means to the organisation’s end and do not really relate to the intended purposes of why they received the funding…the State and Federal government have failed to monitor where funds are really going…are they benefiting the population or not?” (Senior Manager, AHS)

The many weaknesses inherent in the NSW health system are partly due to the complex funding arrangements between the Federal and State governments in terms of who funds what type of service and who should provide that service. For example, according to the view expressed below, preventative care is the GP’s responsibility; however, due to a lack of engagement and collaboration between the providers, some respondents doubted the

266 ______effectiveness of the health services provided. Furthermore, it has also been insinuated that the lack of partnership is one of the aspects contributing to the burden of chronic diseases in NSW.

“…preventative care like health promotion should start at the GPs since they are the first–point–of–contact for patients... health promotion has been part of the general practice but their effectiveness is questionable…however the problem is GPs are funded by Federal Government and so State Governments have no say in the services provided by GPs…whether GPs sell the message or whether they see the effects of that message as being understood or in fact being acted upon is something that falls apart a bit and I am not sure that I or NSW Health has seen the right formula for getting GPs more engaged, more responsible and accountable and committed to collaborate with other service providers in order to improve how patients are managed… I guess the lack of collaboration between those providing services in NSW is a factor contributing to high burden of chronic diseases in NSW…” (Director, AHS)

8.1.6 Restricted Access to Health Services According to community representatives and patients interviewed, the multiple service providers might have exacerbated the problem of access to services. For example, GPs are meant to be the primary health care providers and, therefore, the first point of contact for communities; however, the service fee charges and out–of–pocket expenses, especially when some general practitioners (GP) do not bulk–bill, were mentioned as a barrier to accessing health services at the GP ; patients preferred the free hospital services. The inadequate time GPs spend consulting patients also came up as a deterrent to service utilisation since some patients suffering from complex chronic diseases felt that more time was needed to ensure information on how to manage the illness was communicated clearly and effectively, as well as the lifestyle changes that were necessary. Patients explained how they bypass the GPs services preferring to travel long distances to access the hospital services.

“…I am diabetic receiving injections and tablets but sometimes I forget the instructions and I end up with more problems and since the hospital is far I don’t really go their when I am supposed to go because of transport problems…years ago we used to have a good community health centre close by giving treatment for free but now it’s all about GPs who charges fees even with Medicare card. Sometimes I don’t have the money to pay GP…when I feel quite unwell I go to the

267 ______hospital but even at the hospital they don’t keep me long enough because they need the bed for more patient… so they ask me to go home and be seen by GP…the other problem is my wife and son can’t visit me in hospital since it’s far…

I said to the doctor we need community nurses and doctors but he told me there is no money to pay them to come to the community and so I should go to my GP but the problem is the GP has only few minutes to listen to me and explain what I should do very fast…to be honest half of the time I forget…I think the system should be one… not GPs here, hospital there providing a little bit of service which is not very good…we need better treatment…” (Patient, AHS)

According to the respondents, the only advantage with hospitals is that they are free at the point of contact; however, they also have performance indicators that require hospital managers and clinicians to minimise the number of days a patient can occupy a hospital bed in order to reduce costs of providing health care. According to patients this can lead to misdiagnosis and inadequate management of patients.

“…one day my daughter took me to the hospital with severe pains all over my body but when we got to the hospital there were a few people in the emergency room and we spent 4 hours waiting to see a doctor…at some point when my daughter complained how long we had waited and my increasing pains I was given pain killers by a nurse and told to continue waiting because doctors were busy…when I got to see the doctor later it was a young doctor without much experience I think …at around midnight we were told to go home and see my GP in the morning because it was not necessary to stay in hospital overnight…I think it was all to do with hospital beds and lack of staff…anyway we went home but at around 4am same day the pains were getting worse and luckily my daughter had stayed with me and so she called the ambulance but by the time the ambulance arrived I was unconscious and it turned out to be a mild heart attack plus diabetes and I always wonder what could have happened if my daughter was not with me since I live alone…from that time I have no faith in our health system…I don’t trust the services provided in hospitals because it’s either about the hospital bed or lack of doctors and nurses as is often reported in the news…they really don’t care if they send you home in the name of saving hospital money and something horrible happens to you…” (Patient, AHS)

The other issue raised by patients was the unfairness inherent in Medicare in terms of out– of–pocket payments which deter patients from accessing services when they do not have sufficient funds to pay the whole bill up–front and then get the rebate later. Patients avoid

268 ______seeking early treatment or preventative care because of the cost. This raises the issue of equity in a Medicare type system and the delivery of health care by different organisations.

“I am always required to pay out–of–pocket payments on top of Medicare which really defeats the purpose of tax funded health system. For example, if I go to a GP who does not bulk–bill I am expected to pay the full cost and then go find a Medicare shop to get some reimbursement and I have to pay the difference. This means when I am sick and I don’t have money I sometimes don’t go to the GP for treatment. If the sickness gets worse then I go to the hospital and doctors often ask why I did not go to the GP before my sickness was severe to require a hospital emergency” (Patient, AHS)

According to the patients, education on how to effectively manage chronic illness is crucial, including time dedicated to consultations. In addition, having a co–ordinated system where patients can be managed locally would be ideal in enhancing their response to treatment and reducing the rates of complications due to mismanagement. The issue of out–of–pocket costs through accessing services at the General Practitioners (GPs) clinics was raised as a deterrent to service utilisation. As noted by most of the patients interviewed, the lack of ability to pay or willingness to pay the GP charges for those GPs who do not bulk–bill has impacted on the way patients utilise services. This has also contributed to patients travelling long distances to access hospital services, even with non–complex health problems that could easily be provided at the local community health centres if the services were available. Accessing hospital services for minor illnesses has left most hospitals overburdened and unable to cope with the demand for health care, according to those interviewed. An increase in patients accessing health services at the hospitals also means long waiting times in emergency departments before a consultation, not to mention fatigued doctors and nurses due to long shifts.

“…ideally it would be better if there was a coordinated approach to provide continuum of care to patients closer to where they live to reduce the chronic diseases epidemic and complications…GPs and AHSs need to work together although this can only happen if health system was funded by Federal government. At the moment the system is disconnected with everyone providing a bit here and there of health services which are not good enough to meet the needs of population…patients complain that the consultation time is less than 5 minutes and some patients suffering from chronic diseases needs more time with the GP to understand how to manage the treatment and diet…also some GPs charge fees

269 ______and some patients can’t afford to pay so they end up in hospital which are free even with little problems and with the shortage of doctors and nurses hospitals have too many people to see causing long waiting times or poor quality of care…(Community/Consumer Representative, AHS)

The poorly coordinated approach to service delivery and collaboration between health providers was a factor mentioned by most of the respondents as contributing to uneven distribution of health services and poor access to services. The inadequate system of sharing patient’s information/medical histories was also criticised by patients interviewed, indicating that all service providers should ideally have a centralised patient medical history system to reduce the number of mistakes that could be caused by this problem. The issue of improving community health services was also cited as necessary to improve access to services closer to the communities.

“…GPs, specialists, non–governmental organisations, public and private hospitals are all providing services but we still have access problems especially in rural–remote parts…the problem is the providers don’t work together or share patient information to make sure patients are properly looked after…this is the cause of some of the mistakes made by health professionals if they don’t have up– to–date patient history…I always carry copies of my medical records because if anything happens when I am not at home and I am taken to another hospital then they will not know that I am diabetic and have heart problems…so the health system needs to improve…the government has been talking about introducing an electronic database with patient information which I think would be good but there also needs to be community health centres so we don’t have to travel far to access services…” (Patient, AHS)

270 ______

Chapter 8 Section 1: Summary of Qualitative Findings According to those interviewed, inter–Area Health Service equity in resource allocation has not been achieved and several factors emerged as impacting adversely on the achievement of equity in resource allocation at the Area Health Service level.

It was clear from the interviews that there was a lack clearly defined equity–focussed resource allocation policies at the Area Health Service level to guide resource allocation. The following are the key factors that emerged during the interviews as affecting promotion of equity in resource distribution:

Confusion regarding equity principles to guide AHS resource allocation, Lack of openness about distribution of funds, Limited capacity to benefit from resource allocation and health services, Need for custom–made models for distributing resources, Lack of partnership between service providers, and Restricted access to health services.

271 ______STEP 2: Health and Resource Needs Assessment – Quantitative Findings This section presents the results of two case studies analysing health needs and equity in resource allocation by Area Health Service (AHS) officials to Local Government Areas (LGAs). It focuses on two AHSs: Northern Sydney Central Coast (NSCCAHS) and Greater Western (GWAHS), both in NSW. One is classified as urban while the other is rural and remote (see Chapter 5 for AHS selection criteria). Health needs among the LGAs in the two AHSs were assessed and the results were used to develop “LGA equity funding shares”, which may provide an indication of what an “equity based” allocation of financial resources might look like. As elsewhere in this report, “equity” is defined as equality of access to financial resources for health care on the basis of health need. “Health need” is defined in terms of population health status.

8.2 Health Needs among Local Government Areas (LGAs) and DVIdx Shares Health needs among the 13 LGAs within Northern Sydney Central Coast AHS (NSCCAHS) and the 28 LGAs within Greater Western AHS (GWAHS) were assessed using the premature mortality and chronic morbidity health need index (DVIdx).

Premature mortality and chronic morbidity data were transformed into proportions. The proportions of the two variables were obtained and summed to derive an average that was then used as DVIdx “share”. This “share” represents each LGA share of the total health needs of the AHS. This LGA share of health needs is also taken as the index of the equitable share of total AHS resources to be allocated to the LGA (see Chapter 5 for more details). Due to the lack of actual funding distribution data at the LGA level, a comparative analysis between the actual funds distributed and the study predicted equity shares was not possible. Therefore, the analysis could only predict what each LGA should have received based on the two AHSs’ actual shares for 2006/07 if distribution of resources was based on the burden of chronic diseases and premature deaths among the population used in this study as a proxy for heath needs. The next section presents the DVIdx developed for each LGA within the two AHSs.

272 ______8.2.1 Northern Sydney Central Coast Area Health Service – LGAs and DVIdx Shares Northern Sydney Central Coast is an AHS classified as urban with a population of about 1.1 million distributed across a landmass area of around 6,300 square kilometres (population density 177 per sq km). The area is divided into thirteen administratively defined local government areas (LGAs) namely: Gosford, Hornsby, Hunters Hill, Ku–ring– gai, Lane Cove, Manly, Mosman, North Sydney, Pittwater, Ryde, Warringah, Willoughby and Wyong.

Table 8.1 shows the ranking of Local Government Areas (LGAs) in Northern Sydney Central Coast AHS (NSCCAHS) by their level of chronic diseases and avoidable deaths. The table also shows the premature mortality and chronic morbidity cases per thousand of population.

Table 8–1: Chronic Disease and Premature Death Statistics, Northern Sydney Central Coast Area Health Service (NSCCAHS) Local Population Chronic Chronic Premature Premature Government 2007 Diseases Diseases per Deaths Deaths per Areas (2001) 1,000 (2000– 1,000 Population 2002) Population (2000–2002)

Gosford 164,512 411,079 2,499 1,461 8.9 Hornsby 157,911 332,719 2,107 854 5.4 Hunter’s Hill 14,063 29,834 2,121 85 6.0 Ku–ring–gai 108,371 229,271 2,116 537 5.0 Lane Cove 32,508 70,121 2,157 170 5.2 Manly 39,350 87,021 2,211 223 5.7 Mosman 28,214 58,698 2,080 127 4.5 North Sydney 60,686 128,161 2,112 310 5.1 Pittwater 57,672 127,994 2,219 343 5.9 Ryde 100,068 223,737 2,236 684 6.8 Warringah 139,396 315,294 2,262 871 6.2 Willoughby 65,491 132,153 2,018 305 4.7 Wyong 146,323 361,235 2,469 1,485 10.1 TOTAL 1,114,565 2,507,317 Mean=2,250 7,455 Mean=6.7 Note: Premature Deaths = the number of deaths from all causes for people under 75 years of age and Total Chronic Diseases is the prevalence of people suffering from chronic diseases. Data source: Public Health Information Development Unit established by the Australian Government Department of Health and Ageing in 1999.

273 ______The statistics from which DVIdx shares are calculated are presented in Table 8–1. The results of conversion of those statistics into DVIdx based shares are shown in Table 8–2.

Table 8–2: Health Need Index (DVIdx) Funding Shares for Northern Sydney Central Coast AHS Local Population (%) Chronic Premature DVIdx– Government Diseases Deaths 2000– based Areas 2001 (%) 2002 (%) “Shares”

Gosford 14.8% 16.4% 19.6% 18.0% Wyong 13.1% 14.4% 19.9% 17.2% Hornsby 14.2% 13.3% 11.5% 12.4% Warringah 12.5% 12.6% 11.7% 12.1% Ryde 9.0% 8.9% 9.2% 9.0% Ku–ring–gai 9.7% 9.1% 7.2% 8.2% Pittwater 5.2% 5.1% 4.6% 4.9% Willoughby 5.9% 5.3% 4.1% 4.7% North Sydney 5.4% 5.1% 4.2% 4.6% Manly 3.5% 3.5% 3.0% 3.2% Lane Cove 2.9% 2.8% 2.3% 2.5% Mosman 2.5% 2.3% 1.7% 2.0% Hunter’s Hill 1.3% 1.2% 1.1% 1.2% TOTAL 100.0% 100% 100% 100% Note: DVIdx = Premature mortality and morbidity Health Need Index.

The DVIdx based shares reflect some parameters of the “burden of disease” in the respective Local Government Areas (LGAs). It is clear from Table 8–2 that the “burden of disease” and the consequent “DVIdx funding shares” do not mirror exactly the proportionate distribution of LGA populations (see Figure 8–1). This study supports the notion that it is important when developing a need index to base it on measurable indicators such as premature mortality and morbidity.

Figure 8–2 shows the population and ‘burden of disease’ as defined by the premature mortality and chronic morbidity health need index (DVIdx) of each Local Government Area (LGA) in Northern Sydney Central Coast Area Health Service (NSCCAHS).

274 ______Figure 8–1: Population & Index of Health Need (DVIdx) by Local Government Areas: Northern Sydney Central Coast Area Health Service

180 20.0%

160 18.0%

140 16.0% 14.0% 120 12.0% 100 10.0% 80 8.0% 60 Population ('000) Population 6.0% Health Need Index Index Need Health 40 4.0% 20 2.0% 0 0.0% Gosford Wyong Hornsby WarringahRyde Ku-ring-gaiPittwater Willoughby North SydneyManly Lane Cove Mosman Hunter's Hill

Local Government Areas - NSCCAHS

Population Health need index (DVIdx)

Note: DVIdx = Premature mortality and chronic morbidity health need index as shown in Table 8–2.

8.2.2 Greater Western Area Health Service – LGAs and DVIdx Shares Greater Western is an Area Health Service (AHS) classified as rural and remote with a total population of around 304,874 people distributed in a land mass area of approximately 445,587 square kilometres (population density 0.68 per sq km). The Area is divided into 28 (see Table 8–3) administrative Local Government Areas (LGAs) all classified as regional, rural or remote. The Greater Western Area Health Service (GWAHS) was selected for its rurality, vast landmass area, and sparsely distributed population groups with poor health outcomes.

Table 8.3 shows the population, level of chronic diseases, avoidable deaths and DVIdx (premature mortality and morbidity) per population in each LGA. The population ranges from 700 to 40,000 people. However, although the population is small the enormity of the chronic diseases and premature deaths is evident. According to Table 8.3, almost all local government areas (LGAs) had high levels of chronic diseases and premature deaths but the

275 ______LGAs ranked as having the highest levels of mortality and morbidity, between 20,000 and 92,000, were Dubbo, Orange, Bathurst, Broken Hill, Mid–Western Regional, Cabonne, Parkes, Cowra, Forbes, Wellington and Walgett. In the category of medium levels of chronic diseases and premature deaths, eleven LGAs, including Warrumbungle Shire, Wentworth, Lachlan, Narromine, Blayney, Gilgandra, Oberon, Weddin and Cobar, had rates ranging between 10,000 and 20,000. The population was, however, between 3,000 and 8,000, illustrating the high burden of disease among the LGAs. The LGAs with the lowest premature mortality and chronic morbidity rates were Coonamble, Bourke, Warren, Bogan, Brewarrina, Central Darling, Balranald and the Unincorporated Far West. All nine local government areas had premature mortality and morbidity levels averaging between 3,000 and 10,000 even though the population was averaging between 700 and 5,000, demonstrating the magnitude of chronic diseases and premature deaths within the LGAs. For example, Unincorporated Far West has a population of around 756 people but their premature mortality and morbidity rates per 1000 population is 4,254, the second highest in GWAHS and indeed, NSCCAHS by comparison (see Table 8–3).

Table 8.4 shows that in the category of high health needs, there are six local government areas in the Greater Western AHS: Dubbo, Orange, Mid–Western Regional, Bathurst Regional, Broken Hill, and Parkes, Surprisingly, Broken Hill was the only LGA classified as remote among the six and the rest are mainly regional and rural. Most of these LGAs have a population of 40,000 at the highest and 5,000 at the lowest, producing a DVIdx ranging between 5% and 12.5%. Similarly, the medium level of health needs category was dominated by 10 LGAs as follows: Cabonne, Cowra, Forbes, Wellington, Walgett, Lachlan, Warrumbungle Shire, Wentworth, Blayney, and Narromine with health needs (DVIdx) ranging between 2% and 4.8%. The rest of the LGAs had less than a 1.7% DVIdx, signifying their low levels of premature mortality and prevalence of chronic diseases. According to these findings, some LGAs with relatively low populations had higher level health needs; however, due to a larger population in rural LGAs, they tended to attract a higher DVIdx.

276 ______Table 8–3: Chronic Disease and Premature Death Statistics, Greater Western Area Health Services (GWAHS) Local Government Areas Population Chronic Chronic Premature Premature 2007 Diseases Diseases per Deaths Deaths per (2001) 1,000 (2000– 1,000 Population 2002) Population (2000– 2002) Balranald 2,546 7,379 2,898 34 13.4 Bathurst Regional 37,542 72,482 1,931 236 6.3 Blayney 6,891 15,881 2,305 72 10.4 Bogan 2,994 8,800 2,939 64 21.4 Bourke 3,217 9,300 2,891 48 14.9 Brewarrina 1,998 8,500 4,254 38 19.0 Broken Hill 20,139 56,484 2,805 269 13.4 Cabonne 12,907 41,194 3,192 105 8.1 Central Darling 2,008 8,500 4,233 38 18.9 Cobar 5,112 10,200 1,995 56 11.0 Coonamble 4,342 9,800 2,257 53 12.2 Cowra 13,025 33,400 2,564 131 10.1 Dubbo 39,499 91,148 2,308 345 8.7 Forbes 9,755 25,519 2,616 109 11.2 Gilgandra 4,702 12,745 2,711 37 7.9 Lachlan 6,927 17,600 2,541 98 14.1 Mid–Western Regional 21,983 68,749 3,127 297 13.5 Narromine 6,776 17,238 2,544 68 10.0 Oberon 5,260 11,797 2,243 41 7.8 Orange 37,009 88,413 2,389 349 9.4 Parkes 14,846 37,527 2,528 160 10.8 Walgett 7,199 21,500 2,987 109 15.1 Warren 2,871 9,200 3,204 34 11.8 Warrumbungle Shire 10,208 18,728 1,835 88 8.6 Weddin 3,793 10,295 2,714 40 10.5 Wellington 8,406 22,343 2,658 124 14.8 Wentworth 7,086 18,540 2,616 76 10.7 Unincorporated Far West 756 3,200 4,233 16 21.2 Total 304,872 756,462 Mean=2,481 3,135 Mean=10.3 Note: **Unincorp Far West = is the Unincorporated Far West LGA. Premature Deaths = the number of deaths from all causes for people under 75 years of age and Total Chronic Diseases is the prevalence of people suffering from chronic diseases. Data source: Public Health Information Development Unit established by the Australian Government Department of Health and Ageing in 1999.

277 ______Table 8–4: Health Need Index (DVIdx) Funding Shares for Greater Western Area Health Service Local Government Areas Population Chronic Premature DVIdx–based (%) Diseases 2001 Deaths 2000– “Shares” (%) 2002 (%) Dubbo 13.0% 13.8% 11.0% 12.4% Orange 12.1% 13.4% 11.1% 12.3% Mid–Western Regional 7.2% 10.4% 9.5% 9.9% Bathurst Regional 12.3% 11.0% 7.5% 9.3% Broken Hill 6.6% 8.6% 8.6% 8.6% Parkes 4.9% 5.7% 5.1% 5.4% Cabonne 4.2% 6.2% 3.3% 4.8% Cowra 4.3% 5.1% 4.2% 4.6% Forbes 3.2% 3.9% 3.5% 3.7% Wellington 2.8% 3.4% 4.0% 3.7% Walgett 2.4% 3.3% 3.5% 3.4% Lachlan 2.3% 2.7% 3.1% 2.9% Warrumbungle Shire 3.3% 2.8% 2.8% 2.8% Wentworth 2.3% 2.8% 2.4% 2.6% Blayney 2.3% 2.4% 2.3% 2.4% Narromine 2.2% 2.6% 2.2% 2.4% Bogan 1.0% 1.3% 2.0% 1.7% Cobar 1.7% 1.5% 1.8% 1.7% Coonamble 1.4% 1.5% 1.7% 1.6% Gilgandra 1.5% 1.9% 1.2% 1.6% Bourke 1.1% 1.4% 1.5% 1.5% Oberon 1.7% 1.8% 1.3% 1.5% Weddin 1.2% 1.6% 1.3% 1.4% Brewarrina 0.7% 1.3% 1.2% 1.3% Central Darling 0.7% 1.3% 1.2% 1.3% Warren 0.9% 1.4% 1.1% 1.2% Balranald 0.8% 1.1% 1.1% 1.1% Unincorporated Far West 0.2% 0.5% 0.5% 0.5% Total 100% 100% 100% 100% Note: **Unincorp Far West = is the Unincorporated Far West. DVIdx = Premature mortality and chronic morbidity Health Needs Index.

The DVIdx based shares for Local Government Areas (LGAs) in GWAHS reflect some parameters of the “burden of disease”. Based on Table 8–4, it is clear that the “burden of disease” and the consequent “DVIdx funding shares” are not commensurate with the distribution of LGA populations (see Figure 8–2).

278 ______Figure 8–2: Population & Index of Health Need (DVIdx) by Local Government Areas: Greater Western Area Health Service

45 14.0%

40 12.0% 35 10.0% 30

25 8.0%

20 6.0% Population ('000) Population

15 Index Need Health 4.0% 10

2.0% 5

0 0.0% Dubbo Orange Mid-WesternBathurst Regional RegionalBroken Hill Parkes Cabonne Cowra Forbes Wellington Walgett Lachlan Warrumbungle Wentworth ShireBlayney NarromineBogan Cobar CoonambleGilgandra Bourke Oberon Weddin BrewarrinaCentral Warren Darling Balranald **Unincorp Far West

Local Government Areas - GWAHS

Population Health need index (DVIdx)

Note: **Unincorp Far West = Unincorporated Far West local government area. DVIdx = Premature mortality and chronic morbidity health need index as shown in Table 8–4

279 ______

8.2.3 Movement towards Equity in Health Status, 1987–2006 Although the incidence of premature mortality is still high in both Area Health Services, there has been a substantial decrease over a period of twenty years (1987 to 2006) as shown in Figure 8–1. For example, NSCCAHS premature mortality rates per 100,000 of population in 1987 was estimated to be around 278.9 but declined to 129.8 by 2006, while for GWAHS over the same period 417.5 (1987) decreased to 193.4 (2006). With both rates decreasing some 46%, the gap between the NSCCAHS and GWAHS rates was reduced from 138.6 to 63.6 premature deaths per 100,000 of population. Figure 8–3 shows that after the twenty odd years of pursuit of equity through allocation of NSW Health resources, the premature death rate in GWAHS was nearly 50% higher than the rate in NSCCAHS.

Figure 8–3: Northern Sydney Central Coast and Greater Western Area Health Services Premature Mortality Rates per 100,000 Population, 1987 to 2006

500

400

300

200

100 Premature Mortality Rates per 100,000Premature per Pop. Mortality Rates 0 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Years

GWAHS NSCCAHS Data Source: NSW Health Chief Health Officer’s Report, 2008

280 ______STEP 3: MAPPING OF HEALTH NEEDS (DVIdx), HEALTH PROGRAMS AND SERVICES AND EQUITY SHARES

8.3 DVIdx Equity Shares and Health Services Due to lack of funding data at the lower level (LGA) of the health system, this study was not able to analyse the extent to which equity had been promoted in the allocation of resources to differentially benefit the most disadvantaged LGAs and population groups within NSCCAHS and GWAHS. Therefore, an alternative approach to addressing the third research question was adopted. This approach attempted to assess whether equity had been reflected through the provision of health services. Health services available to each LGA within the two AHSs (NSCCAHS and GWAHS) were identified and mapped to ascertain whether health service availability reflected the level of health needs within the LGAs.

As shown in Figure 8–4, significant numbers of the population were experiencing difficulties accessing health care when needing it. In 2007, for example, NSCCAHS (urban AHS) had an estimated 11.3% persons aged 16 years and over with problems related to health care access, while GWAHS (rural AHS) had almost three times that of NSCCAHS at 29.9%, which was even higher than NSW state–wide (17%).

Figure 8–4: Difficulty Accessing Health Care when Needed in NSCCAHS and GWAHS for Persons Aged 16 Years and Over, NSW 2007

35.0% 30.0% 25.0% 20.0% 15.0% 10.0% health care (%) care health 5.0% 0.0% Pop. having difficult accessing accessing havingPop. difficult NSCCAHS GWAHS NSW State-Wide Area Health Services & NSW

Persons Having Difficulty Accessing Health Care (%)

Data Source: NSW Health Chief Health Officer Report, 2008

281 ______The assumption made here is that resources allocated for health are converted into health services; therefore, an alternative to assessing equity where funding data is lacking, especially within smaller geographical areas, is to appraise the health services located in each LGA and AHS. Each LGA within NSCCAHS and GWAHS was characterised according to its level of population, premature mortality and morbidity health needs index (DVIdx), predicted DVIdx equity funding shares, SEIFA–IRSD ranking and finally available health services. These factors were used to compare the level of health needs in the LGA populations and the services available to meet those needs.

Health services within LGAs in both NSCCAHS and GWAHS were collated by putting together the number of hospital beds in public and private hospitals and the number of community health facilities (these were counted as one service), including GP services, per head of population (2001/02).

8.3.1 Northern Sydney Central Coast AHS Health Services Table 8-5 shows the Northern Sydney Central Coast AHS (NSCCAHS) local government areas health needs index (DVIdx) and the number of available health services. Gosford, by virtue of its large population of around 164,512 (see Table 8–5) and being the second most disadvantaged local government area in NSCCAHS, as defined by SEIFA–IRSD (see Chapter 5 for details), has the highest DVIdx of about 18.0%. However, the health services available – three public hospitals with 577 beds, 128 private hospital beds, 7 community health centres and 5.2 GP services per capita (2005) – are not commensurate with the high level of needs in this LGA. Other LGAs with a high DVIdx behind Gosford are Wyong, Hornsby, Warringah, Ryde and Ku–ring–gai LGA. The Wyong LGA has one of the highest levels of disadvantage (1st highest among the 13 LGAs) and has the third largest population in NSCCAHS. It was not, therefore, surprising that it scored the second highest DVIdx equity share (17.2%) although the health services accessible in Wyong – one public hospital with 278 beds, 10 community health centres and 4.9 GP services per capita (2005) – are not proportionate to the high level of needs in this local government area.

282 ______Table 8–5: Population, Health Need Index (DVIdx) and No. of Health Services within NSCCAHS

Hornsby and Warringah LGAs have the second and fourth highest populations, also securing quite large DVIdx equity shares of around 12.4% and 12.1% respectively. The Hornsby local government area is merged with Ku–ring–gai LGA, which is the least disadvantaged in terms of relative socioeconomic disadvantage for the purposes of providing health services. Both LGAs share one public hospital with 283 beds, one public mental health hospital, two specialist hospitals offering 83 beds, and one dental health hospital among others. In terms of population Ku–ring–gai was ranked fifth according to the DVIdx. Ryde LGA, also with a fairly large population, ranked sixth scoring 9% DVIdx equity funding shares (see Table 8–5 and Figure 8–4).

283 ______Figure 8–5: Health Services and Health Needs (DVIdx) within Northern Sydney Central Coast Area Health Service Compared

1400 20.0% 18.0% 1200 16.0%

1000 14.0% 12.0% 800 10.0% 600 8.0% Health Need Index Need Health 6.0%

No. of Health Services Health of No. 400 4.0% 200 2.0% 0 0.0% North SydneyGosford Ku-ring-gaiHornsby Ryde Wyong Manly Pittwater Mosman Warringah Willoughby Lane Cove Hunter's Hill

Local Government Areas (LGAs)

Health Services Health need index (DVIdx)

Note: Health Services = sum of health services within each local government area

On the other hand, four local government areas in Northern Sydney Central Coast Area Health Services (NSCCAHS), Pittwater, Willoughby, North Sydney and Manly, with medium levels of health needs based on the DVIdx, were only retaining around 3 to 5% DVIdx equity shares. This is due to their relatively low populations. Lane Cove, Mosman and Hunter’s Hill retained the lowest DVIdx in NSCCAHS, ranging between 1% and 2.5%. The three also have the lowest populations within the NSCCAHS, less than 35,000 people in each case, with fewer health needs. On the whole, there are considerable variations in the population distributions, with some LGAs with relatively high populations retaining a high DVIdx and vice versa.

In terms of health services, the least disadvantaged LGAs have more than their fair share of health services, considering their low populations and health needs as defined by premature mortality and prevalence of chronic morbidity. For example North Sydney LGA has a population of around 60,686 people with least health needs based on DVIdx in NSCCAHS; however, the health services offered are one public hospital (with 740 beds), one private

284 ______hospital (with 518 beds), 23 community health centres and 4.0 GP services per capita (2001/02) (see Table 8–5 and Figure 8–4). Compare this to the Gosford or Wyong local government areas, which have the highest burden of premature mortality and chronic diseases (DVIdx) and high populations, and yet fewer health services, Clearly, North Sydney has access to health services over and above the needs of its population as shown in Figure 8.4.

Although clearly inequities in health service distribution exist between local government areas in NSCCAHS, especially North Sydney LGA, a more accurate picture of “equity” would have been achieved if the study had looked at the hospital patient utilisation rates by LGA to determine the patient residence (this is discussed further in discussion Chapter 9).

Northern Sydney Central Coast AHS Mapping There are two maps for this AHS. Figure 8.6 shows the location of each Local Government Area (LGA) in Northern Sydney Central Coast Area Health Services (NSCCAHS) with statistics for each LGA: total population, health needs as defined by premature mortality and chronic morbidity (DVIdx), socioeconomic index for areas – index for relative socioeconomic disadvantage (SEIFA–IRSD) (see Chapter 5 for more details on SEIFA– IRSD) and the predicted equity funding shares. The predicted equity funding shares are the amount of funds each LGA would have received if resource allocation was based on the burden of premature mortality and chronic diseases within the population.

Figure 8.7 indicates the health services available to each LGA in NSCCAHS. The health services included are as follows: 1) public hospital beds, 2) private hospital beds, 3) community health centres, 4) child health services, 5) day surgery services, 6) dental services, 7) mental health services, 8) aged care health services, 9) mental hospitals, and 10) equivalent full time GP services per head of population (2001/02). However, Aged Care Home Facilities were excluded from the analysis since they do not provide direct medical services to the population (see Chapter 9 Discussion for further discussion relating to service availability and utilisation).

285 ______Figure 8–6: Population and Health Needs – Northern Sydney Central Coast AHS

286 ______Figure 8–7: Mapped Health Services – Northern Sydney Central Coast AHS

287 ______8.3.2 Greater Western Area Health Services Based on SEIFA–IRSD (see Chapter 5) almost all LGAs in the Greater Western AHS (GWAHS) are highly disadvantaged with scores less than 1000 (ranking 4th and 5th quintiles), with the exception of Bathurst Regional (1004), Cabonne (1007) and Unincorporated Far West (1034), compared to Northern Sydney Central Coast with 8 out of 13 LGAs scoring over 1100. This reflects the high level of health need among most rural and remote based LGAs within GWAHS. In terms of health services, the LGAs in GWAHS were also disadvantaged with most services centralised in the larger regional LGAs such as Dubbo, Orange and Bathurst, with all three having access to public hospitals with over 100 beds each.

The remote LGAs, Broken Hill, Central Darling, Wentworth and Balranald, all had high levels of health needs in comparison to their low populations and were also socially and economically disadvantaged but had less access to health services.

288 ______Table 8–6: Population, Health Need Index (DVIdx) and Number of Health Services within GWAHS

289 ______Figure 8–8: Premature Mortality and Chronic Morbidity Health Needs Index (DVIdx) – Greater Western AHS

400 14.0% 350 12.0% 300 10.0% 250 8.0% 200

150 6.0% No. of Health Services Health of No. Health Need Index Need Health 100 4.0%

50 2.0%

0 0.0% Dubbo Orange Mid-Western Bathurst RegionalBroken Regional ParkesHill Cabonne Cowra Forbes Wellington Walgett Lachlan Warrumbungle WentworthBlayney Shire NarromineBogan Cobar CoonambleGilgandra Bourke Oberon Weddin Brewarrina Central Warren DarlingBalranald **Unincorp Far West

Local Government Areas (LGAs)

Health Services Health need index (DVIdx)

Note: **Unincorp Far West = Unincorporated Far West Local Government Area

290 ______Greater Western Area Health Service Mapping

There are two maps for this AHS. Figure 8.9 shows the location of each Local Government Area (LGA) in Greater Western Area Health Service (NSCCAHS) with statistics for each LGA: total population, health needs as defined by premature mortality and chronic morbidity (DVIdx), socioeconomic index for areas – index for relative socioeconomic disadvantage (SEIFA–IRSD) (see Chapter 5 for more details on SEIFA–IRSD) and the predicted equity funding shares. The predicted equity funding shares is the amount of funds each LGA would have received if resource allocation was based on the burden of premature mortality and chronic diseases within the population.

Figure 8.10 indicates the health services available to each LGA in GWAHS. The health services included are as follows: 1) public hospital beds, 2) private hospital beds, 3) base hospital (multipurpose), 4) district hospital (small in size), 5) mental health hospitals, 6) community health centres, 7) child health services, 8) day surgery services, 9) dental health services, 10) mental health services, 11) aged care health services, 12) Soldiers’ Memorial hospitals, 13) child and adult mental health services 14) youth health services 15) sexual assaults services, and 16) equivalent full time GP services per capita (2001/02). However, the Aged Care Home Facilities were excluded from the analysis since they do not provide direct medical services to the population (see Chapter 9 Discussion for further discussion relating to service availability and utilisation).

291 ______Figure 8–9: Mapped Population and Health Needs – Greater Western AHS

Notes: DVIdx = Premature Mortality and Chronic Morbidity Health Need Index (constructed by this study), SEIFA–IRSD = Socioeconomic Index for Areas–Index of Relative Socioeconomic Disadvantage, developed by the Australian Bureau of Statistics, 2006.

292 ______Figure 8–10: Mapped Health Services – Greater Western AHS

293 ______

Chapter 8 Section 2: Summary of Quantitative Findings Equity in resource allocation from the Area Health Services (AHSs) to population level was difficult to assess due to lack of expenditure data at the AHSs’ levels that detailed how resources have been distributed within or across programs and services, local government areas or showing the communities that benefited from the funds allocated.

A double–variable index consisting of premature mortality and chronic morbidity was developed and used to assess levels of health needs among local government areas (LGAs) within two purposely selected AHSs: (Northern Sydney Central Coast and Greater Western). Health services within each LGA were mapped to establish whether available health services distribution matched the level of health needs.

Based on this study’s analysis, the LGAs in both NSCCAHS and GWAHS that manifested higher levels of health needs had less than a fair share of health services, while those with fewer health needs and less population had more than their fair share of health services.

The study found that the geographical location of hospitals and other health facilities were the key determinant of how resources were allocated within NSCCAHS and GWAHS. Health services, and hence distribution of resources, did not reflect the level of health needs within the LGAs in either of the two AHSs.

294 ______

Chapter 9: DISCUSSION

Overview This chapter discusses the main findings of the study in the light of the four research questions: 1. To what extent has resource allocation to Area Health Services by the NSW Health Department been equitable? 2. How are the financial resources received from the State level distributed at the Area Health Service level? 3. Is equity reflected in the allocation of resources at the AHS level? 4. What prospect is there for allocation of NSW Health resources both at the State and AHS levels to promote equity?

For the discussion to be systematic the four research questions are used as the main headings and the quantitative and qualitative findings are integrated, where appropriate. The discussion draws on the existing literature in support of the arguments raised.

9.1 To What Extent has Resource Allocation to Area Health Services by the NSW Department of Health been Equitable?

9.1.1 Introduction As indicated earlier, several studies have suggested that, through the use of the resource distribution formula (RDF), the NSW Health Department has been able to improve equity by shifting resources to areas where health needs are greatest (Gibbs et al., 2002, Gilbert et al., 1992, NSW Health, 2005). NSW Health claims that over the years progress had been made in reducing disparities in funding across NSW from 13.8% (funds needed to be reallocated to achieve equity) in 1989–90 down to 3% in 2004–05 (see Chapter 1). This certainly reflects a substantial resources shift. In other words, Area Health Services (AHSs) with the relatively greatest ‘health needs’ as assessed by the RDF have received increasing shares of available funds as the years have gone by, implying that there has been a significant movement towards equity in funds allocation so far as RDF assessment of health

295 ______needs is concerned. However, does achievement of the RDF target shares or narrowing the inequity ‘gap’ make the allocation across AHSs equitable? The objective of this research question was to analyse the extent to which resource allocation by the NSW Health to Area Health Services (inter–AHS) level has been equitable. Equity, as stated in other chapters, is defined in this study as equality of access to financial resources for health care on the basis of health needs. Health need is defined in terms of population health status. Thus, AHSs with the greatest health needs were expected to be allocated more funding to meet their higher level of health needs. Two proxy measures of health need were used to assess equity: a general need index (GHNIdx), which combines health and socio–economic variables using multivariate statistics, and a double variable index (DVIdx) based only on premature mortality and chronic morbidity data. In addition to these indices, the NSW Resource Distribution Formula (RDF) was used to assess the extent to which the RDF had achieved its objective of narrowing the inequity gap in resource allocation across the eight AHSs. An attempt was also made to use the Social Economic Index for Areas – Index of Relative Socioeconomic Disadvantage (SEIFA–IRSD) developed by the Australia Bureau of Statistics to further explore how the resource allocation based on an index of socio– economic variables alone might differ from allocations based on other measures of health needs; however, the funding differences (gap) derived by the SEIFA–IRSD index were very large because the SEIFA indexes are ordinal measures of socio–economic level and cannot be used to distribute health resources (see Chapter 5 for further details on these indices and how they were developed).

Although different indices yield different outcomes, overall, a significant degree of inequity was observed in the inter–AHS resource allocation with a number of AHSs receiving considerably more or less than a ‘fair share’ of resources for the four years (2003/04 – 2006/07) examined on the basis of the two indicators (GHNIdx and DVIdx) developed to assess equity. This does not suggest, however, that allocation to all eight AHSs was not equitable for all the years analysed: a few AHSs received a ‘fair share’ of funding for some of the years. The next section discusses the main points from the findings upon which the above conclusion [of inequitable allocation] is based. Evidence from other sources that support these findings is highlighted.

296 ______9.1.2 Resource Distribution Formula and Equity

9.1.2.1 Evidence from the Resource Distribution Formula (RDF) NSW Health has stated that the average distance from the RDF target shares has decreased considerably from 13.8% to 3% between 1989/90 to 2004/05. However, based on the analysis carried out by this study involving the actual shares allocated to the eight Area Health Services (AHSs) from 2003/04 to 2006/07 and the RDF predicted shares (see Chapter 6), 87% of those funds were found to be allocated in accordance with the RDF targets. In 2005, NSW Health reported that around 98% of the available funds were allocated in accordance with RDF targets. This significant adjustment was achieved despite a number of factors affecting the speed and extent of change. These limiting factors, most of them identified in the interview material reported in Chapter 6, included the following: 1) NSW Health policy requiring the existing level of services in each AHS be maintained; 2) only new “growth” money being available for adjustment of inequities; 3) NSW Health not being exempt from political pressure from particular sectional interests; 4) individual directors of services within NSW Health possibly exerting pressure on the funding allocation decision makers in their efforts to meet performance targets set under the Senior Executive service contract system; and 5) the complexity of the structure, activities and financing of the NSW health system inevitably imposing constraints on the rate and scope of achievable change. A NSW Health publication has pointed out that the RDF was regarded as a “guide” to funds allocation rather than a strictly implemented prescription.

Although only 1–2% of available funds were, in terms of “RDF equity”, over the years 2003 to 2007 inappropriately allocated, there were notable differences between AHSs as to the extent to which they were either “overfunded” or “underfunded” within a range of +14% to –4% of their individual RDF targets. In some cases the factors mentioned above as influencing the rate and extent of change were no doubt relevant and other factors may have also influenced the inter–AHSs’ allocation of funding, for example: 1) the existence within an AHS of private hospitals which might have reduced the demand for government funded hospital care; 2) availability and utilisation of other non –NSW Health funded services to fill gaps in the NSW Health funded provision but which entailed costs higher

297 ______than those of government provided services; 3) non –availability of personnel to staff NSW Health services; and 4) trials, innovations, modifications of NSW Health funded services and activities aimed eventually at leading to more effective and efficient service delivery but initially conducted in only one or a few AHSs.

The interviews with senior NSW Heath officials indicated that the reasons some AHSs were over– or under –funded was due to inadequate implementation of the RDF, which has ensured historically under–funded AHSs were receiving sufficient funding. The other reason cited was the difficulties involved in taking funding away from the over–funded AHSs to allocate to those under–resourced. Those interviewed at the AHS level were more concerned that they were not getting sufficient funds to carry out the particular work they had been employed to do, rather than with equity of resource allocation over the whole of the NSW Health services. Although published annual reports from AHSs’ authorities made reference to “equity”, there emerged from the interview material little indication of a shared state wide “equity orientation” among those interviewed. If, in fact, the central NSW Health authority had tried to develop a state–wide “equity culture” among AHSs personnel, this was not apparent from discussions with them.

9.1.2.2 The RDF and Health Needs A NSW Health publication has claimed that the RDF approach to inter–AHS allocation of funds has “worked” or has “been successful”, citing the movement from 87% to 98% of funds being allocated in accordance with RDF targets over eighteen years (1989/90 – 2007/08) period. But was that “success” reflected in reduction of “inequities” in health status? Using crude premature death rates as the indicator of population health status, Table 6-3 and Figure 6–3 in Chapter 6 shows that, over the period of RDF application, the premature death rates dropped in all AHSs populations. The gap between the premature death rates in a relatively affluent Area Health Service (NSCCAHS) and in a less affluent rural/remote AHS (GWAHS) was reduced substantially. For example, in NSCCAHS premature deaths reduced from 278.9 to 129.8 per 100,000 persons in 1987 to 2006 while GWAHS reduction was from 417.5 to 193.4 per 100,000 persons for the same period. The rates in the other six AHSs populations also decreased over the period under review but the

298 ______differences between them at the start and at the end of the period were not large. The narrowing of the gap between the largely urban NSCCAHS and the rural/remote GWAHS population death rates is of particular interest. This is because, in the years 1987 to 1989 prior to the adoption of the RDF approach to distribute funds, although there was a significant downward trend in both urban and rural premature death rates, the gap between the two rates was increasing. How far the adoption of the RDF approach contributed to the continuance of the downward trend and to the narrowing of the gap between urban and rural premature death rates is a matter for further inquiry or speculation. So far as the decrease in the NSCCAHS and GWAHS rates is concerned, Figure 6–2 shows that for the few years (2003 to 2007) of the detailed analysis covered in this study, NSCCAHS was consistently “underfunded” whereas the funding allocated to GWAHS was either equal to or slightly greater than its RDF predicted share.

9.1.2.3 Towards a Greater Equity For the sake of argument, suppose that equality of population premature death rates results equally in resource allocation. At the end of this study review period, per capita funding for GWAHS was $5,269 and for NSCCAHS $3,610, a GW/NSCC AHS ratio of 1.5:1.0. The GW/NSCC AHS premature mortality rates (PMRs) ratio was 1.9:1.2. This may be interpreted as showing inequity in outcome as the result of inappropriate RDF guided funding. For GWAHS–PMR to be brought into line with the NSCCAHS–PMR, perhaps the GWAHS per capita funding should be increased considerably, but it must be remembered that since NSW Health policy is to maintain existing levels of service, this increase cannot be met by decreasing NSCCAHS funding.

It is not clear from this study or from any other reported study just what level of per capita funding for GWAHS would be required to equate the two premature mortality rates. In view of the slow progress of equalisation over the eighteen years of using the RDF as a guide to resource allocation and the fact that GWAHS has for at least several years been receiving its due RDF share of available funds, it seems that further reliance upon RDF funds allocation within the constraints of likely total NSW annual budgets will not greatly change the present situation.

299 ______This brings the discussion to the point raised in Chapter 2 that NSW Health, even with its access to some Commonwealth funding, does not control all the resources and other factors which determine population health. Recognising this constraint, could NSW Health adopt some other approach to inter–AHS allocation of funds, which would further lead to more rapidly and effective reduction of existing inter–AHSs inequities?

9.1.3 Alternative Formula Based Allocation This study looked at three formula based alternatives to the RDF approach. All three – the Socio–Economic Index For Areas – Index of Relative Socioeconomic disadvantage (SEIFA–IRSD) developed by ABS; the general health need Index (GHNIdx); and the double variable index (DVIdx) constructed by this study – produced resource allocation patterns different from those produced by the RDF as one would expect.

9.1.3.1 SEIFA–IRSD Index Based Resource Allocation As explained in Chapter 5, the SEIFA–IRSD is one of the Australian Bureau of Statistics indices constructed using socio–economic indicators. However, the author of the SEIFA– IRSD index Technical Paper 2006 warns potential users of the SEIFA that the indices are ordinal measures of socio–economic level and they can only be used to rank areas but not to measure the size of the difference in terms of socio–economic level between areas. For example, one cannot infer that an area with an IRSD value of 500 is twice as disadvantaged as an area with an index value of 1,000 and, similarly, the difference in relative socioeconomic disadvantage between two areas with values of 900 and 1,000 is not necessarily the same as the difference between two areas with values of 1,000 and 1,100.

Therefore, although it is possible to compare the distribution of an index across AHSs, using box–plots as explained in the Technical Paper (Adhikari, 2006) (pgs 3–4), these comparisons cannot be relied upon as the basis for the precise allocation of resources between AHSs. The attempt to use SEIFA–IRSD to reallocate resources to the eight AHSs produced a huge gap (see Appendix A) in funding that can be explained by the fact that composite indexes such as those produced by ABS reflect a number of factors which may

300 ______in some way or another have an influence on health status, but which are quite outside the control of NSW Health and are not affected by the NSW Health allocation of resources.

In addition to these limitations on the usefulness of SEIFA indexes in resource planning, the precise pattern of resource allocation produced by their misuse (see Appendix A) is so different from the current pattern that the implementation of the SEIFA derived distribution pattern would be both politically and administratively impossible.

Regarding the use of the SEIFA index, when mapping values for the relative socioeconomic status of LGAs within an AHS (as was attempted in Chapter 8), the Technical Paper advises mapping by the quintile in which each LGA lies rather than, as was done in Chapter 8, by LGA index value. LGA index values, being ordinal values, cannot be used for precise allocation of resources to LGAs. Mapping by quintile was not done in Chapter 8.

9.1.3.2 GHNIdx Based Resource Allocation While this index attempts to construct an index reflecting only variables that bear some reasonable degree of recognisable relevance to determination of health status and relevance to factors related to activities under some degree of NSW Health control, there are, as with the SEIFA indexes, problems in the interpretation and utilisation of the index for detailed decision –making on allocation of funds available to NSW Health. In addition, as with the SEIFA –IRSD based funding pattern, the distribution produced is so widely different from the current pattern that its adoption would entail changes that would obviously be politically unacceptable and administratively unworkable.

For example, the quantitative analysis of inter–AHS equity showed a considerable degree of inequity in the allocation of resources to Area Health Services (AHSs) despite few AHSs receiving a fair share of resources for some of the years examined. On the basis of the general health need index (GHNIdx), as demonstrated in Chapter 6, allocations to two AHSs (South East Sydney & Illawarra and Northern Sydney Central Coast) were fairly equitable with no major differences between actual and GHNIdx–based allocations.

301 ______Allocation to four AHSs (Hunter & New England, North Coast, Greater Southern and Greater Western) was fairly inequitable with all four areas receiving less than their GHNIdx–based allocations. Finally, two AHSs (Sydney South West and Sydney West) were over–funded with actual allocations far exceeding what should have been allocated on the basis of health need as measured by the GHNIdx. Sydney South West, in particular, was significantly over–funded, receiving nearly twice as many resources as it should have on the basis of the GHNIdx. In summary, the overall allocation on the basis of the GHNIdx was fairly inequitable with four AHSs reasonably under–funded and two significantly over–funded.

9.1.3.3 DVIdx Based Resource Allocation Of the four approaches to resource allocation considered in this study, a DVIdx based approach is the simplest in concept and the one most obviously and directly related to population health status. One of the statistics from which it is derived, premature deaths by LGA of residence of the deceased, is routinely collected and published annually. Estimates of population by LGA are available with sufficient frequency to permit timely revision of the index. Morbidity data is more problematic. The data used in the DVIdx relates to “chronic diseases cases” according to the tables published by the Public Health Information Development Unit established by the Australian Government Department of Health and Ageing in 1999 to assist in the development of public health data systems and indicators. The morbidity data was collated from a National Health Survey conducted in 2001.

The anomalous statistics for chronic “cases” in NSCCAHS may raise doubts as to whether there was an error in the original data collection or in the subsequent data processing. However, after the amalgamation policy was introduced in NSW Health in 2005, the then Northern Sydney, an affluent AHS, was merged with Central Coast which is socially and economically disadvantaged; this might, therefore, explain the high levels of morbidity in NSCCAHS. But given the outdated nature of the “premature mortality” and “chronic morbidity” statistics used in calculating the DVIdx, one has to view the calculated DVIdx based funding shares with some caution. This does not necessarily mean that an index that reflects premature mortality and morbidity is of no or little practical value in calculating

302 ______funding shares. A major difficulty is finding appropriate and timely morbidity statistics. This difficulty underlies some of the thinking which leads to the conclusion reached by some writers or researchers that a single readily available indicator of health need, such as the number of premature deaths in a population, may, admitting that it is a somewhat “blunt instrument”, be an acceptable indicator of “relative health status”.

Recognising the caution that must be exercised in using the DVIdx as calculated in this study, it is noted that the use of that index to allocate NSW Health funds in 2006/07 would have required a reallocation of 9.0% of the total funds allocated in that year. This is significantly different from the less than 2% reallocation required to align the actual allocation with RDF based equity shares.

Based on the use of the double variable index (DVIdx), the study analysis showed significant inequities in resource allocation to AHSs, despite few AHSs being allocated a fair share of resources for some of the years analysed. Based on the DVIdx as shown in Chapter 6, Section 2, resource allocation to three AHSs (Hunter & New England, North Coast and Greater Western), were fairly equitable with no major differences between the actual and DVIdx–based allocations. Resource allocation to three AHS (Sydney South West, South Eastern Sydney & Illawarra and Sydney West) was inequitable with all the three AHSs receiving more than their DVIdx–based allocations. Finally, two AHSs (Northern Sydney Central Coast and Greater Southern) were also inequitable due to receiving far fewer resources which did not match their levels of health needs as defined by the DVIdx. Overall, resource allocation on the basis of DVIdx was fairly inequitable with two AHSs considerably under–resourced and three over–funded for the four years examined, suggesting that most AHSs received more or less funds than their burden of disease (DVIdx).

In addition to the quantitative evidence, there was evidence from the interview data suggesting that resource allocation to AHSs has been inequitable (Chapter 6, Section 1). Views from the majority of the informants of this study (policy makers, health executives and managers) suggest that NSW Health has not taken steps to improve equity of resource allocation between AHSs despite the attention given to it in policy documents and reports

303 ______claiming that it has moved towards equity as judged by the RDF. These respondents perceived the inability of the NSW Health Department to address the historical imbalance between funds allocation to hospitals and community–based health services as a source of inequity in the health system. Historically, hospitals have been the main providers of health services in NSW and as a result a significant proportion of the health budget (range 70% to 80%) is allocated to hospitals. This in itself may not constitute lack of equity but, as most of the respondents explained, allocation to hospitals was driven by performance indicators that aimed to promote efficiency rather than equity. Consequently, health authorities were concerned more with efficiency than equity when allocating resources to AHSs. Several respondents also indicated that because of the emphasis on efficiency, AHSs with large hospitals received a greater share of resources than their counterparts with relatively small hospitals, predominantly in rural and remote areas (see Chapter 6).

9.1.3.4 Evidence from Other Sources The finding that resource allocation from the State level to Area Health Services (AHSs) has been partly inequitable is supported by evidence from other sources. The recent comments by a NSW State Member of Parliament – Hon. Melinda Pavey – during a health budget session in parliament highlighted the extent of inequities in funding allocation to AHSs. Directing her question to the Deputy Director General of Strategic Development at NSW Health, the Honourable Member noted among other things that AHSs such as North Coast, Greater Southern and Hunter & New England, all classified as rural and remote, were significantly underfunded based on her calculations. According to the MP, the North Coast AHS was under–funded to the tune of $54 million per year, the Greater Southern AHS was under–funded by $27 million per year and the Hunter & New England AHS was under–funded by $9 million per year (NSW Parliament, 2008). Although the Honourable Member’s assertion was disputed by the Deputy Director General of Strategic Development, there was an implicit admission by him that, although inter–AHS equity in resource allocation had not been achieved for all the AHSs, shares of recurrent funding for the three areas cited were “moving closer towards their Resource Distribution Formula (RDF) targets”.

304 ______The basis of the Member’s claims has been analysed and, based on the percentage shares NSW Health (NSW Health, 2005) projected AHSs should be receiving, the analyses support her claims. In the assessment of equity using the RDF, this study found significant discrepancies between the actual allocations received by AHSs and their RDF target shares for the four years examined (2003/04 – 2006/07). This means some AHSs received less than their RDF target shares while others were allocated greater shares than recommended by the RDF. In 2006/07, five out of the eight AHSs were allocated funding that was less than their RDF target shares, while the other four were receiving over and above their target share. For example, North Coast was the most under–resourced AHS by –$52.4 million, followed by Northern Sydney Central Coast AHS (–$47.3 million), while Hunter & New England AHS had the highest over–funding of around $74.1 million, followed by Sydney West AHS ($49.5 million) in 2006/07 (see Chapter 6). A similar pattern of AHSs receiving more or less than their predicted RDF target shares was observed for the remaining years (2003/04 – 2005/06). And although this cannot be taken as definite indicator that movement towards equity in funding has not been achieved across AHSs, it adds to the general perception among health officials in NSW that resource allocation in the health sector has been inequitable, a perception that is supported by the findings of this study. In the light of the evidence summarised above, one would ask why are there still the inequities after almost two decades of the RDF, and has the RDF really achieved its objective? These issues are discussed in the following section.

9.1.4 To What Extent has the Resource Distribution Formula been used? The existence of a funding formula per se does not guarantee equity of resource allocation (NSW Parliament, 2008, Garling, 2008); it depends inter alia on the extent to which the formula is applied and the soundness of the indicators in achieving equity. In trying to understand why, in spite of the RDF, inter–AHS resource allocation in NSW has remained inequitable, the study examined the extent to which the RDF has been used for funding allocation within the health system. Although the Health Department has stated in several policy documents that the RDF has been used to guide the allocation of health funds to Area Health Services and to monitor progress towards achieving geographical equity in health funding since 1989/90 (NSW Health, 1999, NSW Health, 2005, NSW Health, 1990,

305 ______NSW Health, 2008, NSW Health, 1993), evidence from this study and other sources point to the contrary (i.e. indicate that the RDF has not been applied fully).

NSW Health has created the impression that inter–AHS resource allocation is to some extent guided by the RDF (NSW Health, 2005); however as discussed earlier there were discrepancies between the actual resources allocated to the eight Area Health Services from 2003/04 to 2006/07 and the RDF target shares. For example, four out of eight AHSs received funding that was less than their predicted RDF target shares, ranging between –4% to 14% for the four years examined, while those receiving more than their RDF target shares ranged between 0.5% to 9% (see Chapter 6).

Further evidence came from the policy makers, health executives and managers who doubted the extent to which the RDF is actually used by the NSW Department of Health to allocate resources to the eight Area Health Services (AHS) “…resource allocation in NSW is historical based…I don’t think the funding formula (RDF) works…resource poor AHS are still poor…”. This opinion about the way the RDF is applied was shared by many of the respondents. The political motivation to shift resources to AHSs experiencing higher needs was another concern raised “…the issue is how the formula [RDF] is used and the political willingness to shift resources to areas of greatest need…”. The views articulated clearly suggest that the RDF is not effectively used in distribution of resources, while others thought that it is only applied to an insignificant proportion of growth funding with the largest share of funding being allocated to hospitals using the episode funding model “…the issue really is about how the formula is used…it is used to distribute only a very small proportion of the growth funding…the bulk of the health budget is historically distributed to hospitals through episode [casemix] funding and so I don’t think the RDF will ever be seen as an effective tool…” (See Chapter 6, Section 1). This view of a health executive from the NSW Health Department finally paints a clear picture of the RDF. Therefore, as the evidence shows, doubts about the usage of the funding mechanism (RDF) prevail and it is fair to conclude in this study that the RDF is not indeed applied as it is intended. This raises a question as to whether the under 2% gap in reducing funding inequities is accurate as asserted by NSW Health (see Chapter 1).

306 ______Evidence from the comments by Hon. Melinda Pavey during a health budget estimate session in NSW Parliament directed to the NSW Health Deputy Director of Strategic Development, Dr Richard Matthews, as discussed earlier, noted among other things that some Area Health Services like North Coast, Greater Southern and Hunter & New England, all classified as rural and remote, were significantly under–funded based on her calculations; the underfunding ranged from $9 million to $54 million (see Section 9.1.2.3).

Finally, the recent Garling Report produced by a Special Commission of Inquiry appointed by the General Governor to investigate Acute Care Services in NSW Public Hospitals in 2008, highlighted the importance of the equity component inherent in the RDF as important to ensure “fair and equitable access to health services throughout NSW” but also noted that problems occur because the RDF is not being met throughout the regional areas. Indeed, according to the submissions made to the Garling inquiry, North Coast AHS was under– funded by around $70 million per year. The explanation provided by NSW Health to Professor Garling was that’ “…the RDF technical paper has never been strictly applied because historically there have been discrepancies in the allocation of funds to different areas…” and that the slow but sure move to apply the RDF fully was an effort to achieve equity in funding (Garling, 2008). This evidence clearly confirms that the RDF has not been applied fully since its inception (1989/90) as the findings have shown. Noteworthy is that NSW Health has not published details showing the extent to which the formula is applied, including the amount of funding allocated through the mechanism (i.e. is it the entire annual health expenditure?) These comments can be interpreted to mean that the RDF is yet to achieve its main objective of ensuring AHSs receive their target share of resources as predicted by the RDF.

Based on the different types of evidence discussed above, the findings suggest that allocation of funding from the State level to the eight Area Health Services (AHSs) has not been exactly in line with the Resource Distribution Formula (RDF). The views from policy makers and health executives and evidence from the GHNIdx and DVIdx strongly suggest that more needs to be done in order to achieve movement towards equity in funding across AHSs (Chapter 6). Similarly, the comments from the Honourable Member of Parliament,

307 ______Melinda Pavey and the Garling report confirm that equity has not been achieved in the distribution of resources since some AHSs are still receiving less than a fair share of resources, a share that does not match their level of health needs. The above findings indicating lack of equity in resource allocation to AHSs, prompt two questions: Why are there still inequities after almost two decades of the RDF? and is the RDF actually used?

9.1.5 Is the RDF a Good Formula for Achieving Equity in Resource Allocation to AHSs? This section will discuss the resource distribution formula (RDF) in the context of its ease of applicability and whether the indicators are useful in terms of capturing health needs.

9.1.5.1 Level of Comprehension and Ease of Applicability of the RDF An extracts from the Garling Report (Garling, 2008) clearly states that the RDF “is quite complex and, not surprisingly, poorly understood” by policy makers and health executives within the NSW health system. This was evident in the interview data: due to the complex nature of the formula it is barely understood by policy makers and therefore difficult to apply. The majority of those interviewed expressed negative opinions about the RDF. They believed that the formula might have widened the inequalities in funding distribution since the AHSs that were historically under–funded continued to receive less resources, while others doubted the extent to which the formula was used and whether it was an effective tool in estimating the needs of the population as explained by respondents “…I doubt the use of the RDF and its ability to estimate the needs of the population…”. If the formula was much simpler and easy to understand, then policy makers would be able to apply it without difficulty.

Because there is little evidence as to the extent to which the RDF has indeed been implemented, it has left many health officials guessing the level of its application. This evidence suggests that if the development and use of the RDF was communicated openly and was easy to understand, then it would have been applied effectively, thus stopping the speculation and doubts about its use.

308 ______9.1.5.2 Validity of Indicators Used in the RDF In terms of the effectiveness of the RDF in assessing health needs and equity, the major weakness inherent in the formula is the use of the Standardised Mortality Ratios (SMRs) indicator, as well standardising various measures for age and sex. The age, sex and SMRs indicators are used to weight the different RDF program components as discussed in chapter 4. The question of whether mortality is a valid proxy for morbidity and the use of SMRs as the operationalisation of need has generated considerable debate in health research. SMRs have been criticised due to their poor correlation with morbidity variables. Forster (1977) examined the correlation between age and sex standardised mortality rates and morbidity rates from the General Household Survey (GHS) and found that the rank order correlations between mortality and ‘acute sickness’ and between mortality and ‘bed sickness’ were not significant (Forster, 1977). However, there was a statistically significant correlation between mortality and chronic sickness. He therefore concluded that it was doubtful if mortality could be considered a valid indicator of morbidity. Similarly, Snaith (1978) and Palmer (1978) supported Forster’s finding by casting doubt on the appropriateness of using SMRs as a proxy measure for morbidity.

Also, as Wilkins (1987) commented, “…one policy implication…is that in order to provide equivalent services for a given health need, not only age and sex but also some indicators of socioeconomic status must be taken into account. Health planning based on mechanically applying the same age sex specific rates to populations of widely differing socioeconomic characteristics can result in serious under–servicing of the disadvantaged, aggravating rather than redressing inequalities in health status” (Wilkins, 1987). Therefore, since the use of age, sex and SMRs standardisation as the main indicators of health needs has been criticised due to its failure to recognise the impact of poverty, overcrowding and other aspects of social deprivation on health–care needs (Fox, 1978), current needs–based formulas have consistently included various socioeconomic indicators. However, as noted by Eyles et al (1991), inclusion of social deprivation in needs–based resource allocation formulas might affect health–care needs by affecting the ratio of mortality to morbidity. This implies that social deprivation might give rise to higher health –care needs in clinical terms than is indicated by mortality rates alone. The result of escalated health needs is

309 ______increased costs in meeting those health–care needs by, for example, restraining from discharging patients early from hospitals if their social conditions are deemed poor.

Therefore, although strong correlations are observed between social deprivation and mortality, both in terms of individual indicators of deprivation (Forster, 1979, Brennan and Lancashire, 1978, Knox et al., 1980, Townsend et al., 1984, Glover et al., 2004, Carstairs and Morris, 1989) and combined indices (Carstairs, 1981, Carstairs, 1982), to weight needs–based resource allocation for social deprivation may lead to ‘double–counting’ health needs, as noted by the authors of the Black Report (United Kingdom Department of Health and Social Security., 1976). According to Eyles et al (1991), if the main objective of the needs–based resource distribution formulas is to direct resources to areas of greatest need for health care, then the addition of social deprivation weights would only be relevant where these weights represented risks to health over and above those proxied by mortality indices. They conclude that if the prevalence of social deprivation or other socio-economic determinants of health are of concern in terms of contributing to poor health status, then policies other than health–care provision should be considered which would address the problem more directly.

Thus the use of socioeconomic factors (SEIFA Index of Education and Occupation developed by the Australian Bureau of Statistics) by the NSW Health Resource Distribution Formula (RDF) raises two issues. The first is, as discussed earlier, the author of the SEIFA– IRSD index Technical Paper 2006 warning potential users of the SEIFA that the indices are ordinal measures of socio–economic level and that they can only be used to rank areas; they cannot be used to measure the size of the difference in terms of socio–economic level between areas and, therefore, cannot be used for precise allocation of resources. Second is the issue of “double–counting” of health needs.

The SEIFA Index of Education and Occupation (EDUOCC) summarises variables in terms of people in an area who are unemployed, their level of qualification and if employed, the type of jobs they are working in. For example, a high EDUOCC score indicates that an area has a higher proportion of people who are well educated, employed in professional occupations, proportionately few unemployed people and fewer people in low skilled jobs

310 ______(Adhikari, 2006). Nonetheless, the years of schooling received or earned credentials may not capture significant differences in educational quality (Braveman et al., 2005), differential on–the–job training nor the volatility in economic status during adulthood that has been shown to be relevant and to have adverse implications for health (McDonough et al., 1997). But why was SEIFA–EDUCC applied rather than any of the other three SEIFA indexes? NSW Health has not explained the reason behind the choice of the index apart from stating that evidence has shown a strong link between high levels of health needs, less education and particular occupations (NSW Health, 2005).

Braveman et al (2005) concluded that, if a socioeconomic factor was to be included in a need–based formula then multiple socioeconomic measures should be simultaneously included in models rather than just one or two indicators (Braveman et al., 2005) (as proxies of health needs or deprivation) as in the case of the NSW Department of Health RDF. However, the inclusion of multiple indicators usually involves complex methodological difficulties and laborious designs and resources.

The study concludes that the RDF is a complex tool that is poorly understood by policy makers at all levels, leading to inadequate implementation or use. However, even if it was applied fully, it would not have led to equitable allocation of resources between the eight AHSs. The indicators used, for example, the Standardised Mortality Ratios (SMRs) and Index of Education and Occupation as discussed above which do not adequately reflect the AHS population health needs. Therefore, even if the RDF based allocation process were implemented fully, it would not have achieved its objectives of ensuring AHSs receive a fair share of resources commensurate with the differential level of their population health needs. The next question is: can other simpler indicators better achieve equity?

9.1.6 Can Other Alternative Indicators of Health Need be used? Two indicators were developed and used in this study as proxy measures of health needs, in order to assess equity in resource allocation across the eight Area Health Services in NSW. This discussion is centred on the strengths and limitations of these proxy indicators, the General Health Need Index (GHNIdx) and the Double–Variable Index (DVIdx).

311 ______

9.1.6.1 General Health Need Index (GHNIdx) The General Health Need Index (GHNIdx) was developed by combining demographic, socioeconomic and health–related variables to examine whether an index consisting of demographic, socioeconomic and health–related variables could better capture population health needs across the geographical Area Health Services (AHSs) than could two health only variables (DVIdx), without leading to ‘double counting’. There was a high correlation between the variables, with a number of them emerging as important in explaining health needs within the AHSs. However, when the proxy measure was used to measure equity in funding, the results showed that some of the urban AHSs allocations were halved due to the lack of GHNIdx capacity to reflect the actual health needs of the population. In contrast, rural AHSs known to be socially and economically disadvantaged received an allocation that was far greater than their actual health needs.

The findings of the GHNIdx were compared with other indices (RDF, SEIFA–IRSD, GHNIdx and DVIdx) used in this study to identify the strengths and weaknesses of each index. The results showed an apparent over– or under–estimation of health needs which might be attributed to what the authors of the Black Report referred to as ‘double–counting’ health needs for the over–funded AHSs (United Kingdom Department of Health and Social Security., 1976) and under–counting health needs for the AHSs receiving less than a fair share of resources when only socioeconomic indicators were used or when they were combined with health–related indicators.

For example, Sydney South West (SSW), an urban Area Health Service (AHS) with around 1.3 million population, received an 11% funding allocation while South Eastern Sydney & Illawarra (SES&I) also an urban classified AHS with 1.2 million population received 18% of total health funds (see Chapter 6, Section 2). Both AHSs have similar population characteristics and although SES&I might have more pockets of disadvantaged population than SSW, this does not adequately explain the significant difference in funding based on the GHNIdx (see Table 6–9). On the other hand, the study findings showed that the rural and remote AHSs – Hunter & New England, North Coast, Greater Southern and Greater

312 ______Western – were the winners in terms of the funds allocated using the GHNIdx. Therefore, based on these findings, the strength of the GHNIdx index was that it favoured the AHSs classified as rural and remote due to their high levels of socioeconomic disadvantage; actual health needs were over –estimated brought about by the “double–counting” effect as explained earlier. The GHNIdx limitation involved underestimation of the actual health needs of AHSs characterised as urban, leading to excessive under–funding of those AHSs. Although the variables used in the GHNIdx correlated strongly, according to principal component analysis (see Chapter 5), and provided a broader scope than the RDF in terms of measuring health needs, the index was deemed inadequate in identifying the differential health needs (population health status) across the eight AHSs in NSW. For example, the funding gap between actual funding and GHNIdx was enormous, ranging between 5% to 47% for the over–resourced AHSs and –0.1% to –65% for the under–funded AHSs. The next section discusses the double variable index (DVIdx), the second index constructed for this study.

9.1.6.2 Morbidity and Mortality Index (DVIdx) The Double–Variable Index (DVIdx) was constructed using premature mortality and morbidity data (see Chapter 5). Evidence from the DVIdx showed that across the eight AHSs there were significant differences in health needs. Similarly, the findings illustrated the strong relationship that exists between rurality, remoteness and greater health needs based on the premature mortality and morbidity index. For example, health needs were greatest among the four rural/remote based AHSs: Hunter & New England, North Coast, Greater Southern and Greater Western, while in the four urban based AHSs health needs were highest due to their population size (see Table 6–10 in Chapter 6). The findings were not unique, as studies and state–wide surveys have identified the four rural/remote based AHSs as the most disadvantaged and hence that they have higher health needs (NSW Health, 2006a, NSW Health, 2005, Eckstein and Gibberd, 1994). However, as evidenced by the findings of the DVIdx, the level of health needs found in some of the AHSs classified as urban was high. For example, Northern Sydney Central Coast (NSCC) had the highest level of chronic diseases (25.0%) in 2001 and premature deaths (2000–2002) of around (13.9%) when compared to other AHSs, even though it is ranked highest in terms of socioeconomic

313 ______status by SEIFA–Index of Relative Socioeconomic Disadvantage (IRSD) (see Chapter 6, section 2).

This finding is unique and significant since NSCC for the years examined presented a picture of under–funding or allocation of less funding due to its perceived good socioeconomic status, ignoring the high levels of premature mortality and chronic morbidity within the AHS. Although some of the Local Government Areas (LGAs) are highly advantaged in terms of socioeconomic status, pockets of high levels of premature mortality and morbidity exist as shown by the DVIdx. The ability of the DVIdx to estimate the health needs among the population confirms that it is a better mechanism yet a simple tool that uses measurable indicators and can be effectively used to distribute resources. Based on the findings from the DVIdx, the eight AHSs equity shares were much better compared to other indices in explaining the health status of the population and hence distributing a share of resources commensurate with those levels of health needs. Both premature mortality and morbidity indicators were combined since evidence has shown that mortality or SMRs are not a valid proxy of morbidity (Forster, 1977, Birch et al., 1996), and therefore morbidity was used to explain a different dimension of population health status (health needs) that could not be explained by premature mortality solely. Even though the validity of using mortality data has been criticised as “counterintuitive” (Mays and Bevan, 1986) and policy makers are seeking to move away from the notion of the ‘illness–care system and to social determinants of health’ health care system, this study argues that health funding is, in most cases, spent on improving population health status while social policies funds are for dealing with the underlying social determinants of health in order to achieve a “complete physical, mental and socio-economic well-being” of the population.

As shown by the evidence from GHNIdx and the SEIFA indexes applied in this study, the use of socioeconomic indicators in determining the distribution of health resources can produce a different assessment of health needs from those reflected in the premature mortality and morbidity levels. The DVIdx, although derived from a simple aggregation of both premature mortality and morbidity, is deemed in this study to be a more accurate reflection of population health status and hence a fairly good proxy measure of health

314 ______needs. The simplicity of deriving the DVIdx means that policy makers do not necessarily need to involve consultants outside the NSW Department of Health, implying that if the index is constructed internally, it might be more understandable and much easier to implement. Equally important, the availability of mortality data means that the index can be updated as often as possible to capture the changing population health needs rather than wait for Population Census data that is collected every five years in Australia. Although morbidity data might not be readily available, with better data collection information systems, morbidity data and other health–related data, even though difficult to collate on a wide scale can be acquired through detailed planning and implementation of national health data collection systems.

9.2 How are the Financial Resources Received from State Level Distributed at the Area Health Service level?

9.2.1 Introduction While equity is a well –publicised concern at State level resource allocation, there is very little publicly available information as to how Area Health Services (AHSs) decision makers distributes the resources they receive and how much attention they give to equity in the process. The objective of this question was to examine the process and pattern of resource allocation within AHSs using qualitative and quantitative methods of inquiry. The qualitative inquiry elicited views of key health officials at the State (NSW Health) and AHS levels about the processes of intra–AHSs resource distribution. The quantitative approach yielded an analysis of the reported distribution of funds between programs and services (see Chapter 7). The following sections discuss the findings relating to the current processes of allocating resources within AHSs.

9.2.2 What are the Current Processes of Allocating Resources within AHSs? The key finding is that there is no established process for allocating resources within AHSs and that the distribution follows a similar pattern for the years (2004/05 – 2006/07) examined, which is large proportions (70% to 80%) of funds going to hospital services and

315 ______the rest to community based services. The analysis of the processes of resource allocation within AHSs did not yield a clear picture of how resources are shifted from one program/service to the other or of the communities that benefit from the process due to the following factors.

9.2.3 What Role does NSW Department of Health Play in determining the within– AHS Allocation of Funds? The findings of this study confirm what is already known which is that resources at the Area Health Services (AHSs) level are distributed according to programs and services. The distribution (to programs and services) is done by the NSW Department of Health and handed down to the AHSs. Decisions on how much funding to allocate to each program and service is based on the previous year’s expenditures with some growth adjustments. Mainly, the hospitals based programs and services allocation involves episodes of patients care and the complexity of the cases provided by the AHSs; this information is then used to estimate the actual funds that are distributed to the program by the State Department of Health. Therefore, AHSs officials have limited inputs in making resource allocation decisions, apart from provision of budgetary plans based on the previous year’s episodes (casemix funding) of care, including budgets of community based services. According to this study analysis, 70% to 80% of the health budget allocated to AHSs each year is through episodes of care provided within hospital settings. The remaining budget is distributed directly to community programs and services also based on what they spent in the previous year.

Clearly, this type of allocation based mainly on episodes of care highlights what those interviewed called “pursuit of efficiency rather than equity” since the episodes of care funding does not take into account the health needs of those who do not access hospital services. Episode based funding is more concerned with the numbers of patients using the hospital based services. This raises a question about those who might not access treatment from hospitals.

316 ______According to NSW Health, Area Health Services are expected to allocate the resources they receive across these programs based on assessed need and strategic directions (NSW Health, 2008); however, due to the method of allocating funds from the State Department of Health, mostly attached to specific programs and services, AHSs often have difficulties shifting resources from hospital based services to non–hospital services. They also face challenges and lack tools to redistribute funds according to population health needs away from the prescribed programs and services as was explained by health officials (see Chapter 7, section 1). Equally, the Health Department method of allocating the bulk of resources to hospitals raises concerns about how AHSs can achieve equity by differentially benefiting the most disadvantaged population groups. For example, the study findings showed that Northern Sydney Central Coast allocated 80% to hospital based programs in 2004–05. Similarly, Sydney South West, South Eastern Sydney & Illawarra, Sydney West, Hunter & New England, North Coast, Greater Southern and Greater Western allocated between 72 – 79% of their health budget to hospital based programs. The remaining funds were allocated to community based health programs and services according to the study analysis.

9.2.4 What Role does Area Health Services Play in Decision Making? In theory, the NSW health system is decentralised and Area Health Services (AHSs) are supposed to have some level of autonomy to make resource allocation decisions. Decentralisation allows regions to make independent resource allocation decisions in order to meet the health needs of their population. However, this study found that the decentralisation policy in the NSW health system has not created a situation where every AHS can make independent resource allocation decisions due to the way resources are handed down from the state level. Indeed, the allocation of resources by NSW Health based on programs leaves very little room for AHS decision makers to shift resources between programs or services.

Evidence derived from this study shows that, although the health system is decentralised and AHSs, in principle, are empowered to redistribute resources to meet the needs of their populations, they can only do so within specific programs and services, especially those based in hospitals. Similarly, the findings confirm that a disproportionate share (70% to

317 ______80%) of the total funding to AHSs is allocated by NSW Health to hospital–based programs and services leaving only a small proportion for community–based services (See Chapter 7). This type of allocation makes it difficult for executives at the AHSs to shift resources from hospitals to community-based services.

9.2.5 Does the Process of Resource Allocation at the AHS Level appear to reflect a Concern for Intra–AHS Equity? This study did not find any evidence explaining how resources received by Area Health Services (AHSs) are distributed to differentially meet the needs of the different communities or people within the AHSs’ population. Because funding within the AHSs is distributed according to programs and services, it is mainly allocated “geographically” to where the hospitals and other health facilities are located. There was no evidence to show how funds are moved from programs and services to differentially benefit the different population groups within an AHS based on their assessed health needs. Based on the study evidence, equity considerations are not taken into account at either State or AHS level in making decisions about the allocation of NSW Health funding within Areas. This was corroborated by the interview data which repeatedly show that efficiency of service operation is the main goal pursued by NSW Health and the AHSs rather than equity. This is also supported by episode funding policy (which distributes resources according to patients’ episodes of care) used by NSW Health to distribute funding to AHSs. Although episode based funding is not necessarily inequitable, it may be deemed inequitable if there is not equality of access to services.

As shown by the evidence presented in chapter 7, the process of resource allocation at the AHSs’ level has been influenced by the ten programs and services used to distribute resources from the State level. Once the funding is received at the AHSs level, it is maintained within those programs with little movement between and within programs. Indeed, most of the programs and services are provided through the hospitals; therefore the bulk of funding is for hospital–based services. Thus, based on the findings discussed in this section, the study concludes that resources are not distributed to differentially benefit the

318 ______population in greatest need. The next section discusses whether, despite the reported limitations of the allocation processes, equity has been reflected in the allocation of funding within Area Health Services in NSW.

9.3 Is Equity Reflected in the Allocation of Resources at the Area Health Service Level?

9.3.1 Introduction The aim of this question was to examine the extent to which the allocation of resources within the Area Health Services (AHSs) has promoted equity. However, due to the difficulty resulting from lack of information regarding allocation of funding to particular communities (i.e. LGA populations), it was not possible to assess the extent of equity in the allocation of funds. Therefore, the study attempted to assess equity in terms of availability of services. The analysis, as indicated earlier, was undertaken within two AHSs, Northern Sydney Central Coast and Greater Western Local Government Areas (LGAs). The findings of this study clearly demonstrate that there are considerable differences in health status on a geographical basis and the mapping of availability does not seem to reflect the geographical needs (Chapter 8). This section discusses the evidence regarding this question, drawing conclusions as to how far equity is reflected in the allocation and then considers what factors may have contributed to the degree to which equity is reflected.

9.3.2 What Evidence is Available? The evidence presented in the findings chapters includes the following: 1. Opinions expressed in interviews with AHS officials involved in the management of the two AHSs and some other relevant people; 2. Chronic morbidity and premature death data; 3. Mapping of health needs and services/programs; and 4. Evidence from other sources

319 ______9.3.2.1 Opinions from the Interviews Evidence derived from the study findings (chapters 6, 7 and 8) shows that none of the interviewees attempted to give an estimate of the degree of ‘reflection’ among the reasons for their inability to promote equity. Area Health Service officials involved in the intra– AHS allocation of resources cited lack of policies, plans, and tools to guide them in making decisions regarding the adjustment of existing inequities, lack of information as to the “equity situation” in their respective AHSs, ignorance as to how to apply the concept of “equity”, difficulty in understanding an operational concept of equity, and the fact that they did not have to report to the central NSW Health office on “equity” issues. There was a general awareness that inequities in availability and accessibility did exist; however, for the reasons mentioned above and factors such as inability to realign capital investments, recurrent crises in hospital administration (for the most part attributable to insufficient funding from the NSW Health central authority), lack of coordination of or cooperation in the provision of services and activities which required the participation of the AHSs with other service funders and providers, there had been an absence of well directed efforts to address them.

Not specifically mentioned in interviews, but a contributory factor in limiting adjustment of intra–AHS funding allocation patterns, was the structure of the AHCAs. Under the AHCA terms, the Commonwealth and the NSW governments agreed on contributions over a five– year period. However if expenditure exceeded the agreed total, the whole of the excess expenditure was to be met by the NSW government. Despite attempts to curb rising costs of public hospital provision, and in view of the limited taxation powers and other constraints on the ability of the NSW government to raise money, NSW Health was under pressure to minimise health expenditure. The consequence was that AHS managers did not have access to sufficient funds required to bring about the changes that would be required if intra–AHS inequities in service provision were to be adequately addressed. Under the AHCA arrangements, the Commonwealth claimed credit for the provision of a “free” public hospital system, while the State government received the opprobrium generated by the shortcomings of the under–funded public hospital systems.

320 ______9.3.2.2 The Chronic Morbidity and Premature Mortality Data A problem that arose in this study was obtaining data disaggregated to the LGA level. Thus the chronic morbidity data is from a self-reporting survey and is now out of date. The premature mortality data is also out of date. However, this study is concerned with “shares” of disease and mortality between geographically located population groups and it is assumed that the analysis of the available data provided a reasonable indication of the recent situation.

The distribution of mortality and morbidity rates in both Northern Sydney Central Coast and Greater Western AHSs shows very significant differences between LGAs. In both AHSs, the highest LGA premature deaths rate is approximately double that of the lowest; the median LGA rate for GWAHS is significantly higher than that for NSCCAHS. The median LGA rates of chronic diseases are not so different, but the range of rates for GWAHS is considerably wider than that for NSCCAHS (see chapter 8, section 2). Thus it appears that not only are there differences between the AHSs in level of health status (as mentioned in relation to research question 1) but significant differences in health status between the LGAs in each of the two AHSs. Insofar as differences in population health status indicate inequity in health service funding allocation, this analysis indicates that intra–AHS funding has been inequitable in both AHSs to a significant degree.

The reduction of the gap in premature mortality rates between the two AHSs (see Figure 8–3) is interesting as indicating the possible beneficial impact of RDF guided resource allocation between AHSs. But it explains nothing about changes over time in the equity of resource allocation within each of the two AHSs.

9.3.2.3 The Mapping of Health Needs and Services/Programs Further evidence from the mapping exercise of each local government area (LGAs) in the two Area Health Services (AHSs) showed that the distribution of health services did not match the level of health needs in the most disadvantaged LGAs. Across both AHSs, there was a strong association between rurality, remoteness and high health needs; however there was no relationship between the distribution of health services and the high level of health

321 ______needs. In the Northern Sydney Central Coast AHS (NSCCAHS) and Greater Western Area Health Service (GWAHS), the study systematically found a pattern of high health needs among rural and remote LGAs compared to urban ones but the distribution of health services favoured urban based LGAs and were even much higher around the big cities, confirming that equity was not reflected in the provision of health services to match the high levels of health needs (see Chapter 8). This association between rurality, remoteness, high health needs and less health services in this study support the findings of several other studies that have shown a similar link (Cloke et al., 1994, NSW Health, 2006a, Ocana– Riola et al., 2008). Although this finding is not unique, especially in NSW where a significant number of the population reside in rural and remote parts, it has significant implications for policy making in terms of addressing the difficulties associated in changing the location of existing health facilities to where need is greatest. Other important factors that need to be taken into consideration relate to how health professionals can be made to travel to local service delivery points to ensure the under–staffed facilities are running effectively. This suggests that cost of travelling should be factored into the resource distribution equation in order to address the transport difficulties faced by health officials. As the study found some rural and remote LGAs have limited health services, one aspect that might be taken into account is introducing patient transport systems that would guarantee patients are delivered to the nearest functional health facilities to access services. It is worth noting that health care providers such as the GPs, medical specialists and some allied health professionals are not employees of NSW Health since they are funded separately by the Commonwealth government.

The mapping for NSCCAHS and GWAHS of relative levels of LGA health needs as indicated by the DVIdx, LGA SEIFA–IRSD and the DVIdx based “equity funding” provides a graphical aid which may be of some use to AHS officials in developing plans for future allocation of health resources.

Similarly, the mapping of the location of service/program points is potentially an aid for resource allocation planners. This service/program mapping is particularly relevant to this research thesis because it draws attention to the role of other funders/providers of health service activity in addition to NSW Health. It graphically makes the point that in allocating

322 ______resources, NSW Health and AHSs’ decision makers have to take into account the presence of other funders/providers. In pursuit of equity in the allocation of resources, NSW Health cannot simply rely on a formula such as the RDF (or DVIdx) to allocate resources either at the AHS or LGA level because NSW Health is not the only allocator of resources for a given AHS or LGA.

NSW Health has quite rightly stated that the RDF is used as a guide rather than a prescription for resource allocation. The data relating to the distribution of private hospital beds and the differential supply of general practitioner services in the two case study AHSs point to the need to explore the potential benefit, in terms of equitable outcomes, of having a single authority “controlling” the allocation of all relevant sources of funding and service provision.

For purposes of planning the intra–AHS allocation of resources, one has to consider whether the LGA is the appropriate geographic/demographic unit. Some AHS officials referred to instances of two or more LGAs being “clustered” to form what was thought to be a more appropriate geographic/demographic unit for planning purposes. The impacts of these administrative decisions were not examined in this study but are of obvious interest.

Some officials pointed to a need for “tailoring” resource allocation both between and within AHSs. In view of the multiplicity of funders and providers within the NSW health system, this approach may be worth exploring. The exploration would have to include assessment of the capacity of officials at central (NSW Health) and the AHS level to undertake the information gathering, processing, interpretation and utilisation of relevant data, and how far the resultant tailor–made plans could in fact be implemented. Undoubtedly their formulation and implementation would require the cooperation of the other funders/providers involved in service delivery in the relevant geographic area or sub–area.

9.3.2.4 Evidence from Other Sources As discussed in section 9.1, evidence from the State level policy makers and health executives indicated that equity has not been reflected in the allocation of health resources

323 ______since the RDF is not used appropriately to distribute resources to AHSs. What this means is that if equity–focused policies are not promoted at the central (State) level, it would be difficult to expect the AHSs to apply equity principles, since resources allocated in the first place might not be commensurate with the level of health needs. Findings from the NSW Member of Parliament Hon Melinda Pavey confirmed that equitable allocation of resources has not been promoted, citing it as the reason AHSs are still under–funded. The admission by the Deputy Director General of NSW Health that the only mechanism (RDF) that is intended to promote equity was not strictly used is clear indication that equity is not currently reflected in allocation of resources. Finally, the use of the Episode Funding Policy to allocate the bulk of health budget to hospitals illustrates that efficiency rather than equity was being promoted.

Findings from this study confirm that equity within AHSs is not taken seriously and that very little would be achieved in terms of improving the health of the most disadvantaged by the current systems of allocating resources. The study found that intra–AHS resource allocation in the NSW health system has not been equitable in terms of differentially benefiting the populations with greatest health needs due to the programs and services system of allocating resources. The programs and services are mainly determined by the location of hospitals and other health facilities rather than the health needs of the population. Similarly, the findings showed no indication that most disadvantaged LGAs differentially benefited from resource allocation. Instead, it found that within the two AHSs, resource allocation largely favoured LGAs that were less disadvantaged and in urban areas or major cities rather than their rural and remote counterparts, which had higher levels of health needs (premature mortality and morbidity).

Finally, there was no evidence indicating that equity objectives are key drivers of resource allocation in the two AHSs in the NSW health system. The different levels of health needs found in this study emphasise the fact that health needs differ across AHSs and LGAs in NSW, and that if resources were to be allocated more equitably, these differences in need ought to be well–understood and taken into account in the distribution process. Resource allocation decision–making did not achieve the desired goal of greater equity in distribution

324 ______of funds, partly because AHSs have largely ignored the issue of differential needs for health care. To a large extent, the limited attention to differential health needs reflects the lack of a meaningful equity–driven policy and mechanism for allocating resources among different regions and population groups. The next section discusses the key factors that were identified as limiting the degree to which equity is reflected in the allocation of resources at the state (inter–AHS) and AHS (intra–AHS) levels.

9.4 What Prospect is there for Allocation of NSW Health Resources both at the State and Area Health Service Levels to Promote Equity? “…NSW health system needs specific policy to guide equity at each level… the policy should be formulated, implemented and evaluated by a team of experts whose responsibility would be to monitor progress as well as collect data and consult communities to ensure uniformity of how equity is promoted at the various levels of the system…” (Policy maker, NSW Health 2007)

Equity may be a key policy objective within the NSW health system but the economic and political realities of the State (NSW) may affect the extent to which resources are allocated to achieve it. This question was intended to identify and analyse the prospects available for health sector resources to be allocated to promote equity at all levels of the NSW health system. It draws broadly on the findings of the various components of the study and discusses them in the light of the economic, political, social and historical contexts within which decisions around health system resource allocation are made. The discussion is structured around the following points:

1. Difficulty in defining and assessing equity and health needs;

2. Historical and political factors in the allocation of health care resources;

3. Limited role of NSW Health as (a) funder and (b) provider of services;

4. Reallocation of resources based on programs/services and actual costs of implementation;

5. Inadequate supply of finance, personnel and facilities to provide an adequate level of services;

6. Lack of relevant and timely data for development of tools;

325 ______7. Multi–sector collaboration for protection and promotion of health status; and

8. Expert opinion and advice, consultation and collaboration in the resource allocation process.

Difficulty in Defining and Assessing Equity and Health Needs Among the relevant decision makers within the NSW health system, there is much confusion and uncertainty about what is meant by equity including the lack of a single agreed definition and measure of equity in NSW. “Equity” is a concept interpreted differently by different people (see Chapters 3 and 6). The way equity is construed has significant implications for how it is operationalised. In many countries’ health systems, as observed by Mooney and McIntyre (2007), equity is stated as an important goal but the equity goals the country seeks to achieve are rarely established clearly or fully. They note that equity in terms of a policy objective “has been paid little more than a lip–service” (McIntyre and Mooney, 2007). In NSW Health, for example, the lack of agreement as to the exact meaning of equity or the equity policy to be pursued among NSW health officials impedes further progress towards equity as defined in the “The NSW Health and Equity Statement – In All Fairness”. The equity principle emphasised in the policy document is “equality of access to comparable health services” across Area Health Services (AHSs), which could be interpreted as ‘horizontal and vertical equity’ principles, because in the same policy document there is a statement about the need to address inequities related to geographical AHSs and health outcomes by redistributing resources to the most disadvantaged AHSs and population groups (NSW Health, 2004c). This type of redistribution of resources could be based on the premise of achieving vertical equity (see Chapter 4).

The emphasis on diverse equity principles renews the debate that has been ongoing about which equity principle is most favoured by health policy makers. Those who have been engaged in the equity debate argue that equity in access is the most popular equity objective among most health care systems (Culyer et al., 1992b, Mooney et al., 1991, Goddard and Smith, 2001) as is the case of NSW Health. The focus on equity in access, as McIntyre and Mooney (2007) noted, has meant that what health economists have been measuring is ‘use’,

326 ______diverting attention from the actual term ‘access’. However, there is a counter–argument in that health policy documents of many countries include references not only to ‘equity of access’ but also to ‘distribution according to need’ (i.e. NSW Health) and ‘equality of health’ (Culyer et al., 1992b, Culyer et al., 1992a). This focus on different equity objectives may be due partly to the difficulty in defining concepts such as equal access, equal need and equal health, which are mainly used to specify how equity is being pursued in many instances. The confusion brought about by the lack of consensus, even among health economists, as to how equity should be defined has largely left policy makers without a point of reference against which to judge the consistency of their policies in terms of the equity concepts.

In the context of NSW Health, the key challenge with regard to the different equity goals is the degree to which these goals have been communicated within the different levels of the health care system, including the mechanisms developed to support their operationalisation as shown by the study findings. There needs to be well–defined equity principles that guide allocation of health resources at the different levels of the health care system in NSW, bearing in mind that “one–size–will–not–fit all”.

Historical and Political Factors affecting Allocation of Health Care Resources Historically, the allocation of health sector resources has been clouded by politics, particularly in many countries where the process of resource allocation usually lacks transparency and openness. The findings in this study revealed direct political influence on the amount of resources allocated to various Area Health Services in NSW. According to health officials, the Resource Distribution Formula (RDF) implementation is influenced by the political climate so that each time there is a new government or Minister for Health, this adversely influences the resource distribution process, contributing to maintaining the existing historical inequitable distributions. As explained by health officials, politicians often demand hospitals or community health centres to be built in their electorate with no regard for other areas (see Chapter 6, section 1). Clearly, politicians and bureaucrats do not seek to optimise public funds efficiency but rather to maximise their own choices of staying in power for as long as they can. They consciously choose to provide services and benefits to an array of areas they believe will help them to retain office.

327 ______

In some instances, politicians are pressured either by communities or interest groups to provide their Area Health Services (AHSs) with health facilities, leaving other AHSs without powerful politicians continue with whatever services they have. This can lead a politician to interfere with the resource allocation process to ensure his/her community’s demands are met. Politics is an entrenched ‘culture and fabric’ of the NSW health system’s financing and resource allocation and according to evidence any extra resources introduced into the system are also subject to politics. For example, the growth funding introduced by the government in 2000 to help narrow the inequity gap by allocating most of the funding to historically underfunded AHSs is not shared accordingly; instead it is distributed to even urban based over–funded AHSs (See, Chapter 6, Section 1). Therefore, despite the strong policy commitment to addressing inequities in theory, government interference in resource allocation and continuous investment in large hospitals means equity–focused policies will always remain ‘lip–service’ with long term implications for the NSW health care system. The large hospitals not only adversely affect the level of funding available for other areas but also the recurrent costs in favour of areas where the hospitals are located. Resources are shifted towards the urban-based areas which also have large numbers of privately owned hospitals and away from the most disadvantaged and vulnerable residing in rural and remote areas. There needs to be less capital investment and political interference and more transparency through better reporting and monitoring, to show how resources are shifted to differentially benefit the most disadvantaged population groups in NSW.

Limited Role of NSW Department of Health as (a) Funder and (b) Provider of Services The NSW Department of Health is the main provider of public health sector services, mainly through hospitals, and therefore its capacity to distribute funding and provide services is crucial. However, based on the evidence from this study, in NSW, health care funding and resource distribution is a process that is highly vulnerable to political interference and complex bureaucracy. Although the published RDF target shares are meant to reduce these risks, the lack of transparency in how the RDF is used, including the actual amounts allocated through the RDF, leaves it open to risks. As findings have shown, the

328 ______RDF has been in existence for over a decade but the formula has been politicised, according to health officials, rather than used as recommended by the guidelines.

Peacock and Segal (2000) observed that an explicit and transparent resource allocation mechanism provides a constraint on the use of subjective political or bureaucratic power, and identifies key value judgements that underpin the resource distribution process. Furthermore, if made transparent and publicly available, resource allocation mechanisms can generate democratic debate over their appropriateness, which can eventually add to their viability. As this study found, the RDF and specifically the process of distributing financial resources in the NSW health system is anything but transparent. Indeed, at the NSW Department of Health, it was extremely difficult to establish the actual amount of resources allocated through the RDF. Policy makers, also at the Inter–governmental & Funding Strategies Branch, who are responsible for the construction and implementation of the RDF within NSW Health, could not provide any concrete information on the actual RDF figures that were allocated to the Area Health Services (AHSs) since they were also not informed of the process by the executives at the top level within the Department (see Chapter 6).

NSW Health as a funder, and the NSW Government in general, have contributed to the lack of transparency surrounding the resource allocation within the system, including how the RDF is used to promote equity by redistributing resources equitably. As in many other health care systems, the resource allocation mechanism (RDF) in NSW health system was supposed to overcome inequities by distributing funds towards the most disadvantaged AHSs. But as NSW Health explicitly stated, the RDF has not been fully implemented (discussed in Section 9.1). Lack of transparency as to how resources are distributed or how funding priorities are decided upon in NSW Health provides a good opportunity for politics to influence the allocation process, often to the detriment of disadvantaged AHSs and population groups with less political power. As noted by policy makers and executives, NSW Health has often used the resources to “deal with a hospital crisis” within the health care system.

329 ______For a health care system to have the capacity to lead in service provision there needs to be some level of leadership, extensive collaboration and consultations. However, the nature of the collaboration process for the purpose of providing services across the whole of NSW health system is ineffective. This study found limited opportunities for effective collaboration and consultations between the State and Area Health Services and other health providers (like GPs and specialists) within the state and communities. In NSW, there is a need for increased health funding to enable health service providers to target disadvantaged populations with services and to develop mechanisms that would facilitate effective collaboration and consultation with all sectors involved in the provision of health services.

Reallocation of Resources based on Programs/Services and Actual Costs of Implementation Funding to hospital and community health based programs and services plays a crucial role in the NSW health systems just as in many other countries where there is a chronic imbalance in the way health funds are allocated to hospitals and community health centres (WHO, 2008). The proportion of funding allocated to either hospital– or community–based services has significant implications for easing or aggravating the problems faced by a health system. While most of those interviewed discussed the challenges being faced by policy makers when making decisions about where to allocate the health budget, the historical nature of the NSW health care system has prompted allocation of 70% to 80% of its budget to hospitals and the rest to community health (See Chapter 6). The allocation of a large part of the funding controlled by NSW Health to public hospitals is hardly surprising because a large part of the health care outside of public hospitals is not under the control of NSW Health. For example, General Practitioners (GPs) and allied health professionals provide most of the primary health care. However, the imbalance in funding within the NSW health system, as this study found, might be contributing to inequities across geographical areas since Area Health Services in urban areas tend to have more and bigger hospitals than their counterparts in rural and remote AHSs. Equally, evidence from the study has demonstrated that the intra–AHS re–distribution of resources across and within

330 ______programs/services has been restricted, meaning that resources are not shifted to differentially benefit the areas and populations with the greatest health needs.

The main pathways by which funding to hospitals might compromise equity in the health system is through the use of the episode or casemix funding model. The episode funding method distributes resources to hospitals based on episodes of patient care, implying that those populations who do not utilise hospital services are not included in the episodes of care calculations for the purposes of resource distribution. In the last two decades, NSW has made significant progress in increasing funding in the health sector with the larger proportion allocated to hospital based services rather than being used to target the hard to reach disadvantaged populations with health services.

The community health centres, on the other hand, have not benefited much from State funding since primary health care is deemed to be the responsibility of Medicare (funded directly by the Federal Government) and thus provided by General Practitioners (GPs) and specialists. The division of responsibility and hence different methods of funding has led to wastage of resources through a tug–of–war regarding who should fund and provide what services. The conflict between who should provide what services and where have negative implications for patients, especially in rural and remote parts where there might be shortage of hospitals or GPs. Another effect of the division of funding and service provision on the NSW Health public hospitals is the demand brought about by ambulant patients who can not afford or are unwilling to pay GPs charges (for those GPs who do not bulk–bill).

Inadequate Supply of Finance, Personnel and Facilities to Provide an Adequate Level of Services Without sufficient resources AHSs cannot afford to differentially target the most needy population groups with timely services. For AHSs to also have the capacity to address historical inequities, they require extra resources because even though it is possible to differentially target certain groups with the available resources, the AHSs risk taking away resources from people who may also be insufficiently provided. For example, the Commonwealth government, through programs under its direct control, still spends around

331 ______74 cents on Indigenous Australians while it spends $1.00 on the non–Indigenous (Ring and Brown, 2002). And in NSW Health, per capita expenditure on Indigenous health programs across the AHSs ranged between $97 and $404 per capita in 2006/07 (see Chapter 7, Section 2). Considering the high level of health needs experienced by Indigenous people, advocates of the well–being of the Indigenous population argue that spending on indigenous health programs should be double or triple the general expenditure on the non– indigenous population. For example, for every one dollar spent on non–indigenous people, double the amount should be spent on indigenous population so as to narrow the gap in health outcomes (Mooney, 2006), as well as to reduce the inequities in resource allocation and access to health services for the most disadvantaged population groups.

Therefore, unless additional funds are made available, there will probably be limited further improvement in population health status and hence movement towards greater equity. The issue of inadequate supply of finances needs to be addressed at the different levels, State (Department of Health) and AHS levels, to ensure the extra resources be prioritised to differentially meet the health needs of the needy population commensurate with the assessed health needs to yield meaningful improvements. While the amount of funding is not the main factor that facilitates the promotion of equity within a health care system, inadequate funding can lead to a lack of equity–focused policies being communicated effectively based on the assumption that the resources received are not sufficient to promote equity. Insufficient funding can also lead to a dearth of services that does not match the level of population health needs.

In the NSW health system, the capacity of Area Health Services (AHSs) to utilise resources received efficiently is one of the important factors behind resource allocation, particularly at the State level. While evidence in this study has shown that policy makers and health executives see the need to shift resources to areas of greatest need to address inequities, the interviews revealed that resources could not be allocated to areas where there were no facilities or capacity to utilise the resources efficiently.

332 ______Frontline service providers recounted the constraints they faced in their day–to–day efforts to provide quality services as a result of limited funding. Also, findings from the study demonstrate how patients have to travel long distances to access even minor services like management of diabetes, demonstrating that funding allocation is based on availability of health facilities and health professionals to provide services. The evidence in this study confirms that the system of resource allocation in the NSW health system is about where the hospitals are located.

A key determinant of the AHSs capacity to make efficient use of the resources provided to them is health workforce availability. This was the biggest concern of policy makers and health executives in Greater Western Area Health Service (GWAHS). As in many rural and remote areas, GWAHS suffers from a shortage of health professionals due to its rurality and remoteness, leaving AHS significantly understaffed with wide variations in capacity to provide services to highly disadvantaged populations. In 2008 for example, GWAHS had access to 0.6 doctors per 1000 population compared to NSCCAHS with 1.0 doctor per 1000 population (see Chapter 2). In general, the health workforce is congregated in the urban and major cities, especially in the Sydney South West and other urban AHSs where big teaching and research hospitals are based (see Chapter 2).

Most doctors like to work in urban areas where they may have a social lifestyle, teaching and research; therefore, the rural and remote Area Health Services like Greater Western will need extra incentives to attract and retain the right workforce and more resources to improve the health outcomes of the most disadvantaged population.

Lack of Relevant and Timely Data for Development of Tools In NSW, as in many countries equity in resource allocation has been largely promoted at the central (state/national) levels, ignoring the local (Area Health Service) level which has the bigger responsibility of ensuring that available resources are redistributed to differentially benefit those in greatest need. The lack of a tailor–made equity plan for resource distribution at the inter–AHS and intra–AHS levels based on assessed local needs, coupled with a substantial proportion of the health budget going to hospitals, partly

333 ______explains the limited equity in allocation of health funds between and within AHSs despite the decentralisation policy.

As explained by those interviewed, the allocation of most of the available resources to hospital based services did not facilitate redistribution of resources to the most disadvantaged local government areas or population groups. However, health executives at the AHS level expressed the willingness to move funds to the areas of greatest need but they lacked an appropriate tool to guide the process. The study found that the process of allocating resources at the AHSs was anything but transparent. The lack of an AHS tailor– made resource allocation mechanism left many guessing how resources are shifted between and within programs. Therefore, health executives and managers were of the opinion that a tailor–made formula for each AHS is necessary to facilitate redistribution of resources to areas of greatest need where they will do most good. Based on health officials’ views, the formula for rural and remote AHSs has to take different aspects of the area into consideration, implying that if there is a “model” to guide AHSs officials it must be a model that can accommodate the particular problems, costs and resources relevant to that particular AHS. In other words it is a “planning guide” rather than a prescriptive formula. The development of such a resource distribution model will require extensive consultations to ensure the mechanism meets the needs of the specific AHSs and their population. The formula also needs to be simplified by using measurable indicators that reflect the actual health needs of the population (i.e. premature mortality and morbidity) to assist policy makers in achieving a greater degree of equity in resource allocation.

The study findings revealed that there is a shortage of relevant data at AHS level to guide the development of a simpler funding mechanism. Therefore, appropriate data collection systems at the population level are paramount for the successful development of a formula because while this is lacking a meaningful formula cannot be effectively constructed. These findings clearly illustrate the magnitude of the problem and the urgent need for a tool to enable AHSs to shift available funds to areas and populations with the greatest health needs.

334 ______Intra–NSW Health and Multi–sector Collaboration for Protection and Promotion of Health Status In NSW, the study found that collaboration between the different health care providers and different levels of the NSW Health specifically was limited. The limited collaboration was partly due to the diverse health system funders (Commonwealth and States/Territory/Local Governments and the non–governmental sector), multiple service providers and NSW Health being decentralised, thus creating little room for effective collaboration between the State and Area Health Services (AHSs) and within AHSs in terms of how they deal with resource allocation. The poor collaboration and consultation was largely the result of the secrecy behind the way resources were allocated to AHSs and the ineffective use of the Resource Distribution Formula “…we are never consulted when resources are being allocated…” (See, Section 6.2.7). NSW Health policy makers need to promote effective collaboration across and within AHSs, including finding pathways for how to encourage collaboration between the health system, General Practitioners (GPs) and other health care providers within NSW. The lack of an effective partnership between these health sectors has contributed to a poor continuum of care for patients due to an ineffectual integrated model of health care. Having multiple health service providers not collaborating has also led to duplication of health services in NSW, leading to wastage of the already inadequate health budget.

Consultation and Collaboration in the Resource Allocation Process In NSW the health system is supposed to be decentralised, allowing Area Health Services authorities to make independent resource allocation decisions. However, as found in this study, the NSW health system is in fact centralised to the extent that AHSs have limited room to make resource allocation decisions because most resource allocation decisions are made at the central level (NSW Department of Health) through programs and services. The NSW Department of Health decides the proportion of funds that should be allocated to individual AHSs based on programs and services and, although AHSs are supposed to decide the amount of funds that should go to Local Government Areas (LGAs) or communities based on their set priorities in order to meet the differential needs of the population, this has not been the case. This study found that the NSW health system is more

335 ______centralised rather than decentralised and has created a situation where NSW Health independently handles resource allocation issues without consulting or collaborating with AHSs in resource allocation decision making, as explained by health officials. The lack of consultations and collaboration with the relevant health managers meant they were not aware of the processes of distributing resources or the factors taken into consideration. At the State level, the senior managers did not know how resources were allocated at the Department and AHS levels and vice versa (see Chapter 6 & 7). The lack of opportunity to share innovative approaches for resource allocation and deal collaboratively with solving difficult issues associated with developing effective equity–focused mechanisms was apparent from the State and AHS levels of the health system. As observed by Peacock and Segal (2000), the design of a resource allocation mechanism entails a range of technical, bureaucratic and political issues that may be better dealt with in cooperation with others, rather than NSW Health dealing with it, without consulting AHS health officials in the allocation process.

NSW Health needs to create a situation where problems within the entire health system can be dealt with collaboratively by learning from any success stories and from one another rather than being tackled in isolation. Area Health Services facing more challenges, especially those in rural and remote regions, may learn from urban AHSs if they have the chance to collaborate on resource allocation matters with NSW Health taking a leading role. Policy makers within the NSW health system need to find a way of promoting collaboration within the different levels of the system while increasing the autonomy of AHSs in designing their own tailored area–specific resource allocation mechanisms, as was explained by health officials. To a large extent, the lack of consultation and effective collaboration in inter–and intra–AHS resource allocation has restricted the sharing of innovative approaches and policies that would have enhanced equity throughout the NSW Health care system.

336 ______9.5 Study Limitations This section discusses a number of limitations inherent in this study. First, the financial data analysed to assess equity across AHSs was only for a period of four years (2004 to 2007). While some might argue that this period is not sufficient to draw conclusions about the extent to which equity has been achieved, four years were deemed adequate to make judgements about equitable allocation of resources. The extent of equity in resource allocation was assessed by analysing the NSW Health audited financial data to establish the amount of funding allocated to eight AHSs in NSW. While the financial data was available for all the years analysed, there was no categorisation of the funds allocated through the use of RDF. Therefore, an assumption was made that the ‘initial’ (as stated in the NSW Health Annual Report) allocation of resources to the eight AHSs, with an exception of special programs grants, was allocated through the use of RDF.

Another problem in relation to data was that, at the AHS level, financial data showing the movement of funds between and within programs was lacking, including how resources have been shifted to benefit the most disadvantaged population groups. Although the system applied by NSW Health of allocating resources to AHSs according to programs and services has been discredited by health officials for its inability to target the populations in greatest need, it would have been useful to this study if the actual movement of resources could have been reported by AHSs. The effects of the gaps in financial data were deemed minimal to the overall study results. This is because firstly, the main emphasis of the first question of the study was on equity in resource allocation to the eight AHSs (inter–AHSs) through the use of RDF and therefore the use of the total amount was sufficient to achieve that objective. Second, the main concern of the second and third questions of the study was to examine the process of distributing resources at the AHSs to establish whether equity is reflected in the process. The main limitation with the analysis of the third research question was that the study could not ascertain whether resources had been differentially distributed within the specific programs and services to benefit the most disadvantaged population groups because AHSs do not report the allocation of resources according to population groups or local government areas. Therefore, question three was addressed by examining the health services available to determine whether they are consonant with the level of health needs at the LGA level.

337 ______

A further limitation was the exclusion of internally generated revenue from the State (inter– AHSs) level analysis. The study recognises the importance of internally generated revenue in the analysis of equity in resource allocation and exclusion of the component automatically overlooks an important proportion of the health budget. The study acknowledges that some AHSs have a better capacity to generate funds internally; therefore excluding the funds from the analysis might disadvantage some AHSs. The objective of the study was not only to establish the extent of equity in resource allocation at the state level but to also assess how resources received at the AHS are distributed to promote equity. Therefore, resources used in the analysis at the AHSs level (Chapters 7 & 8) took into consideration the revenues generated internally.

The other limitation was with the Australian Bureau of Statistics (ABS) socioeconomic data used for developing the general health needs index (GHNIdx) and used to assess equity in resource allocation at the State level. The Census datasets used in this study were largely 2001 instead of 2006. This is because the 2006 datasets were only published in 2008, by which time data analysis had been completed. The study assumption was that the population’s socioeconomic status had not changed significantly enough to affect study analysis. Another limitation concerned the morbidity data used to construct the Double Variable Index (DVIdx). Unlike mortality data, which was available for three years (2000– 2002), the morbidity data was only available for one year (2001), limiting a comparative analysis to establish if there has been change in the burden of chronic diseases over time.

In terms of methodology–related limitations, the study raises some conceptual issues. Vertical equity (defined as differential allocation of resources to benefit the Area Health Services, Local Government Areas and population groups with the highest health needs) is the main equity principle underpinning the analysis as noted in Chapter 3. However, NSW Health distributes resources to achieve different equity objectives. While resources are allocated to achieve ‘equal access to comparable services’ for the eight Area Health Services (horizontal equity), they are also allocated to address existing inequities by targeting the most disadvantaged Area Health Services with resources, a vertical equity principle. In effect, NSW Health promotes both horizontal and vertical equity objectives in

338 ______allocation of resources. Despite this study intention to focus on vertical equity, it could not overlook the goal of horizontal equity (equal access for equal health needs). This study recognises the need to ensure that the health of a well population is sustained by providing universal services (horizontal equity) to facilitate continuity of care while differentially targeting (vertical equity) those in greatest need with resources. However, this study’s weakness of operationalising the vertical equity principle was to do with inadequate attention be given to informed community values and preferences (Wiseman and Jan, 2000, Mooney, 2006) in the development of health needs indices. In Australia generally, and NSW specifically, special attention to community values and preferences is essential due to considerable differences in cultural interpretations of health. This would ensure that health is conceptually understood from different societal perspectives.

339 ______Summary of Key Points in the Discussion Chapter  In the NSW health system, there was general recognition that equity is an important factor that should be taken into consideration when allocating resources. Most of the policy documents regarding the resource distribution formula acknowledge that the reduction of inequities is a key objective of developing and applying the RAF and RDF formulae. However, evidence from the study’s analysis indicates that inequities still persist in NSW Health even with formula driven health system funding. Indeed, the health officials interviewed doubted that the use of the formula as a basis for allocating funds would result in equitable resource distribution to the eight Area Health Services (AHSs).

 The study found that, although the RDF has been in use for over a decade in NSW Health it has not achieved its intended objective, suggesting that the formula could be a crude instrument that might be exacerbating inequities within the health care system or at least slowing down progress towards attainment of equity goals.

 Based in the overall study findings, the needs-based formulas fail because they are not based on measureable indicators that fully reflect the actual health needs of the population. Therefore, NSW Health needs to simplify the RDF using measurable indicators that reflect the actual needs of the population (i.e. premature mortality and morbidity) to assist policy makers in achieving a greater degree of equity in resource allocation.

 Area Health Services lack relevant expertise in identifying population health needs and the ability to plan and set priorities in the resource distribution process based on health needs.

 Based on the overall study findings, the NSW Department of Health and Area Health Service health executives could also benefit from training relating to population health needs assessment and priority setting against available resources.

 Discussed a number of historical, systemic, operational, financial and administrative factors that are likely to make further substantial and reasonably rapid progress towards equity in resource distribution and health status unlikely. It is suggested that persistence with the RDF–guided allocation of resources is not likely to produce much by way of desired change, and a movement towards more detailed planning, based on a more precise assessment of health status and the inputs required to meet differential health needs, may be a preferred option.

340 ______

Chapter 10: CONCLUSION AND SUGGESTIONS FOR FURTHER RESEARCH

Overview The NSW Health Department has made significant efforts to reduce inequities in allocation of funds between the Area Health Services. However, the findings of this study demonstrate that after more than a decade of RDF–guided resource allocation, inequities in health funding still persist for a variety of reasons as discussed in Chapter 9. The prospect for future allocation of resources in the health system to further promote equity has been examined by identifying the key factors limiting the equitable allocation of resources. This concluding chapter examines the policy implications of the findings and offers recommendations as to how equity in resource allocation in NSW can be enhanced. It also reflects on the general contributions the study has made to the current stock of knowledge and debate on resource allocation and equity, and offers suggestions for further research.

10.1 Policy Implications of the Study Four key policy implications of the study are examined: (a) how equity in resource allocation is assessed; (b) the development of appropriate machinery to guide resource allocation; (c) resolving the historical and political factors affecting the equitable allocation of health funds; and finally, (d) encouraging multi–sectoral collaboration for delivering and promoting health.

10.1.1 Implication for Examining how Equity with regards to Resource Allocation is Assessed The findings of this study have significant implications for the way equity of resource allocation has been assessed within the NSW health system. First, equity of resource allocation has been pursued largely at the State level of the health system, focusing narrowly on inter–Area Health Service equity through the use of the RDF. While it is essential that allocations from the State level to AHSs are made as equitable as possible, it is at the AHS level where resources are translated into programs and services to meet the

341 ______health needs of the population. The equity with which this is done (i.e. resources are allocated at the AHS level) will ultimately determine the extent to which the health system will meet its overall equity objective, which is to eliminate [avoidable] differences in health among different population groups (NSW Health, 2004b). If the allocation of resources at the state level is “inappropriate” there is little that the AHSs can do to correct the imbalance. It is therefore important that the Health Department develop and implement, in collaboration with the AHSs, concrete measures to promote equitable resource allocation within AHSs. The starting point for this could be assisting AHSs to develop appropriate mechanisms to guide intra–AHS resource allocation, including tools for assessing health needs, setting priorities and service targets. There were clear indications from the qualitative data of this study that AHS executives, although desiring to pursue equity of resource allocation, feel constrained in terms of not having any tools to guide them in doing so, and were unclear as to how far they were empowered to make changes to the pattern of resource utilisation within their respective AHSs.

Another issue with regards to how equity of resource allocation is assessed relates to the indicators used in the RDF. The findings of this study, as discussed in Chapter 9, suggest that indicators in the RDF do not adequately reflect the level of health needs of AHS populations. This is based on the fact that the health status of every population is largely reflected in the rate at which people die (mortality) or the rate at which people get sick (morbidity) (Forster, 1977). Logically, there should be a close degree of correlation between indicators of health needs and mortality/morbidity indicators. However, there was limited correlation between the RDF, a composite formula based on four indicators of “population needs”, and the need index developed in this study using the mortality/morbidity variables (see Chapter 9 for a full discussion of this). It is suggested that the RDF is an inappropriate tool for the allocation of resources between AHSs and within AHSs.

342 ______10.1.2 Implications for Developing Appropriate Tools to Guide Resource Allocation This study clearly indicates that inter– and intra–Area Health Services inequities in resource allocation still exist, two decades after the introduction of the RDF/RDF. This finding has major policy implications, requiring the adoption of well–designed and transparent equity–focused machinery for inter – and intra–AHSs resource allocation. As evidence from the health officials has shown, they are not confident that a uniform resource allocation formula that will ensure equity at the State and AHS levels can be developed. It takes more than having a state–wide formula for resource allocation to promote equitable allocation of resources in a health system and, given each AHSs’ specific differences (i.e. geographical and in terms of health needs), a single formula will not work. Indeed, as explained by health officials, AHSs differ greatly in health status and health needs and therefore a formula which would ensure equity across AHSs does not mean it can achieve an equitable distribution of resources within AHSs. Thus focusing on one level at the expense of others constrains the promotion of equity. There is a need for attention to equity in resource allocation at all levels of the health system and policy makers need to seek sound and practical mechanisms for allocating resources if equity is to be widely promoted within the NSW health system.

This could begin with a consideration of current resource allocation policies of the eight AHSs to assess the types of mechanisms that can be developed to improve equity, with overall guidance from NSW Health regarding priority setting and taking into account cross–border flows and internally generated revenues to ensure equitable resource allocation. Within the NSW health system, some AHSs have better administrative and operational arrangements and greater capacity to raise revenue locally than others. Therefore, under an equity–focused policy, such AHSs should be allocated less resources from the State (central) level. Equally, any new formulated equity–focused policies for the State and AHS levels should be communicated well within the entire health system and be transparent and open to scrutiny. Relevant health officials at the NSW Department of Health and AHS levels should be involved in the development process to ensure they are methodologically practical with ease in implementation and that they reflect the actual population health needs. The development of a funding mechanism at the AHS level should

343 ______involve extensive consultations with the health officials at the frontline of service provision since they have superior knowledge about the health needs of their populations and are, therefore, better positioned to allocate resources more equitably to areas where needs are greatest.

10.1.3 Resolving the Historical and Political Factors Affecting the Equitable Allocation of Health Funds The findings of this study have implications for resource allocation and promotion of equity. In NSW Health as indicated earlier, politics influence resources allocated to the Area Health Services (AHSs). Despite the strong policy commitment to addressing historical inequities in theory, the government interference in resource allocation and continuous investment in hospitals in urban areas especially has long term implications for the NSW health system. There is a need to review policy regarding the appropriate distribution of hospitals and hospital based services. Ideally this review should be conducted by independent reviewers.

10.1.4 Encouraging Multi–Sectoral Collaboration for Delivering and Promoting Health For a complex health care system to have the capacity to provide comprehensive services effectively, efficiently and equitably, there needs to be a high level of leadership, extensive collaboration and consultations. However, the nature of collaboration for the purpose of providing services across the whole of the NSW health system is largely ineffective. This study found limited opportunities for effective collaboration and consultations between the State, the AHSs and other health providers (such as GPs and specialists) in the state and communities.

Effective collaboration is crucial not only within the entire NSW health system but also between other sectors responsible for the social determinants of health like social services, housing, education and environmental services. This collaboration would benefit the populations, especially those disadvantaged, by providing a platform for sharing innovative

344 ______approaches to resource allocation across the various sectors for the purposes of improving the health and socioeconomic conditions of the population groups. The social determinants of health can only be effectively addressed by multi–sector policies (e.g. those relating to education, social services, housing, environmental, agriculture) which require strong collaboration and partnership between all relevant sectors whose activities have considerable implications for health and health outcomes. NSW Health can provide the best avenue for such inter–sectoral partnership with support from the Federal government.

10.2 Contributions to Knowledge In most countries equity is a key health policy goal. It is, however, a difficult concept to define and assess. This study makes a number of conceptual and methodological contributions to the equity and resource allocation debate as shown below.

10.2.1 Contributions to the Method of Assessing Health Needs The study has made several methodological contributions to the equity and resource allocation debate. First is the empirical data from the NSW health system to show the extent of equity in resource allocation. The equity debate, in the context of resource allocation, has been largely theoretical, focusing more on need–based funding formulas and the equity principles that should drive resource allocation. Equally, a considerable amount of the health care literature has emphasised the role of the needs–based formulas in reducing funding inequities (see Chapter 4). Although a few empirical studies, mainly from England, have looked at these theoretical arguments, none has been undertaken in NSW or Australia. By analysing resource allocation qualitatively and quantitatively at the State and Area Health Service levels of the NSW health system, this study has contributed valuable insights into how the issue of equity in resource allocation is handled, in practice, at different health system levels. This will contribute to filling the current gap in knowledge regarding how resources are distributed at the inter– and intra–AHSs to promote equity by policy makers and health executives.

345 ______The qualitative component analysed perspectives of policy makers and health executives involved in the resource allocation process in the NSW health system to reveal the factors taken into consideration when resources are being allocated. This important qualitative information has shed some light on why use of the resource distribution formula (RDF) and decisions on resource allocation at the state and AHS level in the NSW health system have not resulted in the movement of more funds to the most disadvantaged AHSs and to population groups with the greatest health needs, as proponents of the RDF policy have argued. The qualitative evidence provided by this study and the empirical quantitative evidence may contribute significantly to informing and supporting health policy reforms that aim to promote equity in resource allocation in the NSW health system and other countries.

A second contribution of this study in terms of methodology is the use of a mixed–method approach to investigate the extent of equity in resource allocation and to explore the factors that influence the equitable allocation of resources. The literature review found no detailed report of the use of qualitative inquiry to inform quantitative analysis of health resource allocation in either developed or developing countries. The review of the literature found no study that goes beyond the quantitative measure of equity in resource allocation by using a qualitative method to first explore the factors that underpin the allocation process in such a comprehensive manner. This distinctive use of mixed–methods to study the issue of equitable resource allocation at different health system levels within NSW Health has provided a more complete picture of the equity and resource allocation phenomenon. For that reason, this study has gone beyond conventional methodological boundaries to promote the use of a mixed–methods approach and, in this case, used a quantitative approach to inform qualitative findings in addressing equity in resource allocation.

A further contribution made by this study in terms of methodology is the use of mortality and morbidity data to develop a proxy measure for health needs. In NSW, as in many other countries, analyses of health needs measurements have involved socioeconomic data derived from Census data and standardised mortality ratios (SMRs). Currently, NSW Health and most other needs–based formulas depend largely on SMRs and socioeconomic

346 ______indices estimated from the Australian Bureau of Statistics for the assessment of health needs levels among AHSs and to guide resource allocation. This study has demonstrated that it is possible to measure health needs using two key variables (DVIdx) which relate to the actual health status of the population in the NSW health system, using some routinely available datasets. Although the DVIdx might appear to be a crude measure of health needs in the simplest terms, it may encourage analyses which focus on the real population health status based on premature mortality and chronic morbidity. Resuscitating the old debate about the use of mortality and morbidity as indicators of health needs for the purposes of allocating health resources may also stimulate debate on alternative and simpler methods of allocating health sector resources that policy makers will find easier to implement than more complex composite “needs indicators”.

10.2.2 Contributions to Current Debate on Resource Allocation and Equity As discussed in Chapter 3, this study supports the use of the principle of treating unequal individuals unequally (vertical equity) and the ‘capacity to benefit’ as important concepts for guiding resource allocation. While there is no consensus on the principle policy makers should use to promote equity, the emphasis is more on horizontal equity ideology to guide resource allocation in most health systems. Recently, there has been a resurgence of literature in support of more emphasis on vertical equity and capacity to benefit goals to drive resource allocation. This study draws on this literature and the different perspectives on social justice to support the contention that society has a moral duty to improve the lives of its most disadvantaged and vulnerable members. This study has re–emphasised the view that it is through discrimination against the well–off in resource allocation that will make the least–advantaged better off or benefit from resource allocation. However, with a limited resource pool (as in NSW) alternative resources need to be sought to facilitate the differential funds allocation to better the worst–off.

The conceptualisation of equity, as undertaken in this study, supports the choice of a vertical equity principle which emphasises unequal but equitable treatment of the unequal, rather than promoting horizontal equity which emphasises equal treatment for equal need. Similarly, the study supports the view that for equity to be achieved, capacity to benefit and

347 ______MESH infrastructure must be also taken into consideration in the processes for allocating health resources.

10.3 Suggestions for Future Research This study suggests two directions for future research. First, there is the need for further research to examine the use of needs–based and population–based formulas’ effectiveness in terms of assessing health needs geographically or within a population and their impacts in promoting equitable resource allocation. A number of health care systems claim to use needs or population–based funding formulas and within those countries inequities still persist between geographical areas and population groups. Therefore, research is necessary to establish how the funding mechanisms are implemented and their capacity to promote equity within a health care system.

Due to time and resource constraints, this study focused on two Area Health Services (Northern Sydney Central Coast and Greater Western) of the eight AHSs for the qualitative component of the study and the AHS level quantitative analysis. Therefore, further research is necessary to broaden the analysis in this study to the remaining six AHSs so that a more complete state–wide picture can be formed of how the issue of equity in resource allocation is handled at the State and AHS levels. A state–wide analysis, in addition to the findings of this study, will provide more robust evidence for developing state–wide equity–focused policies for inter and intra–AHS resource allocation.

Research aimed at analysing how resources have been allocated within different AHSs and population groups, as defined by their differential health status, will be useful in supporting intra–AHS resource allocation and the case for capacity to benefit from resources and vertical equity. Such analysis could not be undertaken in this study because the relevant financial data showing the movements of funds between and within health programs and services at the AHS level were not available.

348 ______Despite several limitations, this study has demonstrated that having a needs–based formula does not guarantee equity in distribution of funds and that there is a complex range of factors that influence the equitable allocation of resources within the NSW health care systems. Therefore, for equity in resource allocation to be widely promoted within all levels of the health system, any mechanism developed for distributing resources has to go beyond simply writing “needs–based formula” and has to pay attention to the all factors involved in the entire process of allocating resources.

10.3.1 Further Work is required in Developing Better Health Information Systems for Area Health Services NSW Health’s health management information system remains inadequate, making it very difficult to monitor and improve the delivery of interventions in a timely and effective way. Availability of appropriate and timely information will facilitate development of useful resource allocation mechanisms especially for AHSs. Therefore, there is a need to develop integrated health information systems that support the delivery of high quality health care as well as enhance the way health resources are allocated and targeted to priority programs and services. Public health informatics requires the cooperation of and data contribution from both health care and non–health care sources to obtain a complete picture of a population’s health and risk status, including how resources are shifted to differentially benefit the most disadvantaged and vulnerable populations. To support resource allocation decision making for both health care and social determinants of health, data must flow not only to policy makers at the NSW Department of Health and AHSs levels but also rapidly to other non–health sectors dealing with the social determinants of health policies. This will facilitate a multi–sectoral collaboration in tackling with population health and the underlying determinants of health.

Other areas that could benefit from further research include the following:

development of better Health Information Systems for Area Health Services;

establishment of whether the NSW health system is more equitable than other states, and of the reasons for any inter–state differences;

349 ______examination of whether the Episode Funding Policy is any different from the RDF in terms of resource distribution;

the best way how to define and assess population health needs, including the indicators to use; and

how the split and partially shared responsibility for funding health services, involving contributions from the Commonwealth, State/Territory and Local governments and a range of non–government sources, affects equity in health care access and deliver in NSW.

Questions Addressed by this Study This thesis set out to and has answered the following research questions: 1) To what extent has resource allocation to Area Health Services by the NSW Health Department been equitable? 2) How are the financial resources received from the state level distributed at the Area Health Service level? 3) Is equity reflected in the allocation of resources at the AHS level? and 4) What prospect is there for allocation of NSW Health resources both at the state and AHS levels to promote equity?

As a result of this study the debate on the allocation of health resources can develop based on better information and informed decision –making.

350 ______Appendix A Socio–Economic Indexes for Areas (SEIFA)

The Socio–Economic Indexes for Areas (SEIFA) is a product constructed especially for those interested in the assessment of the welfare of Australian communities. The Australian Bureau of Statistics (ABS) has developed four indexes to allow ranking of regions or areas, providing a method for determining the level of social and economic well–being in each region. The index that this study attempted to use was Index of Relative Socio–economic Disadvantage (IRSD) constructed from Census variables related to disadvantage i.e. low income, low educational attainment, unemployment, and dwellings without motor vehicles. Therefore, high scores on the IRSD occur when the area ranks high in the measure of relative advantage. Conversely, low scores on the index occur when the area has high measure of relative disadvantage. For example, an area that has a Relative Socio–Economic Disadvantage Index value of 1200 is less disadvantaged than an area with an index value of 900 (Adhikari, 2006).

The geographical areas in NSW are ranked differently by each index according to a ‘score’ that is created for the area based on characteristics of people, families and dwellings within that area. The reference score for the whole of Australia is set to 1,000; lower scores indicate lower socioeconomic status. For example, Greater Western on each of the SEIFA– IRSD rankings has the lowest score (around 957 score) (NSW Health, 2006a).

Table 10–1 is a comparison of SEIFA–IRSD and the two indices (DVIdx and GHNIdx) developed for this study. According to SEIFA–IRSD shares, Sydney South West, Hunter & New England and Greater Western surprisingly attracted almost the same funding shares. SSWAHS and H&NEAHS have much larger population than GWAHS with only 0.3 million population; however, due to its rurality and remoteness, including low levels of socioeconomic status, the area received large funding shares. North Coast was the AHS that attracted the highest funding shares of around 15.1% even though its population is low compared to some AHSs.

351 ______Table 10–1: Comparison of DVIdx, GHNIdx and SEIFA–IRSD Funding Shares

Note: **SEIFA–IRSD is Index of Relative Socioeconomic Disadvantage (IRSD) constructed by the Australian Bureau of Statistics using 2001 population Census data

Based on Tables 10–2 and 10–3 three main observations can be made: 1. That allocation to Hunter & New England (H&NE) Area Health Service was fairly equitable with no major differences between the actual and SEIFA–IRSD–based resource allocation. 2. Allocations to three Area Health Services – Sydney South West, South Eastern Sydney & IIIawarra and Northern Sydney Central Coast – were inequitable with the Areas receiving nearly twice as much resource allocation as they should on the basis of the SEIFA–IRSD–based allocation. 3. That allocations to three Area Health Services – North Coast, Greater Southern & Greater Western – were inequitable with all the Areas receiving less than half of their SEIFA–IRSD–based allocations.

Overall, the allocations can be classified as inequitable with three AHSs significantly under–resourced on the basis of the SEIFA–IRSD and three significantly over–funded. The average distance from equitable resource allocation for the eight Area Health Services ranged between 44.3% and 49.7% for the four years (2003/04 – 2006/07) analysed, based on SEIFA–IRSD (see Table 10–4).

352 ______

Table 10–2: Comparison of Area Health Services Actual Funds Allocation and SEIFA–IRSD Shares, 2003/04 – 2006/07

Data Source: NSW Health Annual Report (2003–2007) and Australian Bureau of Statistics (2001). Note: **SEIFA–IRSD is Index of Relative Socioeconomic Disadvantage (IRSD) constructed by Australian Bureau of Statistics using 2001 population Census data

Table 10–3: Actual Funds and SEIFA–IRSD–based Allocations: Average Distance from SEIFA–IRSD, 2003/04 – 2006/07

353 ______

Table 10–4: Actual and SEIFA–IRSD–based Allocation by Area Health Services, 2003/04 – 2006/07

Table 10–4 shows the percentage funding difference between the actual and the predicted SEIFA–IRSD–based share of funds for 2003/04 to 2006/07. Two Area Health Services (AHSs), South Eastern Sydney & Illawarra and Northern Sydney Central Coast, were substantially over– funded, receiving more than 45% of their predicted SEIFA–IRSD based shares. The biggest losers were North Coast, Greater Southern and Greater Western Area Health Services which received over 100% below their SEIFA–IRSD shares. The percentage funding difference is highlighted in Figure 10–1.

354 ______

Figure 10–1: Actual and SEIFA–IRSD–based Allocation by Area Health Services: 2003/04 – 2006/07

Actual & SEIFA-IRSD-based Allocation by Area Health Services: 2003/04 Actual & SEIFA-IRSD-based Allocation by Area Health Services: 2004/05

1,600.0 100.0% 1,600.0 100.0% 45% 51% 45% 51% 1,400.0 29% 1,400.0 29% 17% 50.0% 17% 50.0% 1,200.0 -10% 1,200.0 -10% 0.0% 1,000.0 1,000.0 0.0%

800.0 -50.0% 800.0 -50.0% -101% -100% -101% -100% 600.0 Amount ($'M)Amount

Amount ($'M)Amount 600.0 -100.0% -100.0% -154% 400.0 Percentage difference 400.0 -154% Percentage difference -150.0% -150.0% 200.0 200.0

0.0 -200.0% 0.0 -200.0%

SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS Area Health Services Area Health Services

Actual allocation 2003/04 SEIFA-IRSD-Based allocation 2003/04 Actual allocation 2004/05 SEIFA-IRSD-Based allocation 2004/05 Percentage difference Percentage difference

Actual & SEIFA-IRSD-based Allocation by Area Health Services: 2005/06 Actual & SEIFA-IRSD-based Allocation by Area Health Services: 2006/07 1,800.0 100.0% 48% 51% 2,000.0 48% 50% 100.0% 1,600.0 29% 29% 1,800.0 16% 50.0% 15% 1,400.0 1,600.0 50.0% -13% -12% 1,200.0 0.0% 1,400.0 0.0% 1,000.0 1,200.0 -50.0% 1,000.0 -50.0% 800.0 -105% -105% -102% -103%

Amount ($'M)Amount 800.0 Amount ($'M)Amount 600.0 -100.0% 600.0 -100.0% Percentage difference 400.0 -164% Percentage difference -164% 400.0 -150.0% -150.0% 200.0 200.0 0.0 -200.0% 0.0 -200.0%

SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS SSWAHS SESIAHS SWAHS NSCCAHS H&NEAHS NCAHS GSAHS GWAHS Area Health Services Area Health Services

Actual allocation 2005/06 SEIFA-IRSD-Based allocation 2005/06 Actual allocation 2006/07 SEIFA-IRSD-Based allocation 2006/07 Percentage difference Percentage difference

Note: Allocations within the range of plus or minus maximum 10% of SEIFA–IRSD–based shares are judged equitable

355 ______BIBLIOGRAPHY

Acheson, R. (1978) The definition and identification of need for health care. J Epidemiol Community Health, 32, 10–15.

Adhikari, P. (2006) Socio–economic indexes for areas: Introduction, use and Future directors.. ABS Catalogue no. 1351.0.55.015. Canberra, Australian Bureau of Statistics.

Anderson, E. S. (1999) ‘What Is the Point of Equality?’ Ethics, 109, 287–337.

Anderson, R. (1978) Health status indices and access to medical care. American Journal of Public Health, 68, 458–463.

Aristotle (1996) The politics, trans. B. Jowett, rev. trans. J. Barnes, ed. S. Everson, Cambridge: Cambridge University Press.

Asada, Y. & Kephart, G. (2007) Equity in health services use and intensity of use in Canada. BMC Health Services Research, 7, 41.

Asante, A. D., Zwi, A. B. & Ho, M. T. (2006) Equity in Resource Allocation for Health: A Comparative Study of the Ashanti and Northern Regions of Ghana. Health Policy, 78, 135– 148.

Asthana, S., Gibson, A., Moon, G., Dicker, J. & Brigham, P. (2004) The pursuit of equity in NHS resource allocation: should morbidity replace utilisation as the basis for setting health care capitations? Social Science & Medicine, 58:, 539–551.

Australian Bureau of Statistics (2004) Population Estimates by Age & Sex NSW 2001. Catalogue No.3235.0.55.001. Canberra, Australian Bureau of Statistics.

Australian Bureau of Statistics (2006a) Deaths 2006, Australia. Cat. No. 3302.0. Australian Bureau of Statistics, Canberra.

Australian Bureau of Statistics (2006b) Population Distribution, Aboriginal and Torres Strait Islander Australians. Cat. No. 4705.0. Canberra. Australian Bureau of Statistics.

Australian Bureau of Statistics (2006c) Socio–Economic Indexes for Areas (SEIFA) – Technical Paper, 2006: 2039.0.55.001, Australian Bureau of Statistics.

Australian Bureau of Statistics (2007a) New South Wales in Focus 2007, Belconnen, ACT, Australian Bureau of Statistics.

Australian Bureau of Statistics (2007b) Regional Population Growth Australia, 2005–06. Canberra, Australian Bureau of Statistics.

Australian Bureau of Statistics (2008a) Health Care Delivery and Financing: 1301.0 Year Book Australia. Australian Bureau of Statistics, Canberra.

Australian Bureau of Statistics (2008b) NSW State and Regional Indicators. Australian Bureau of Statistics, Canberra.

356 ______Australian Bureau of Statistics (2009) NSW State and Regional Indicators, 1338.1. Australian Bureau of Statistics, Canberra.

Australian Institute of Health and Welfare (2001a) Australian health trends. AIHW cat. No. PHE 24, Australian Institute of Health and Welfare, Canberra.

Australian Institute of Health and Welfare (2006) Australia’s health 2006. AIHW cat. no. AUS 73. Australia lnstitute of Health and Welfare, Canberra.

Australian Institute of Health and Welfare (2007a) Health expenditure Australia 2005–06. Health and Welfare Expenditure. Series no.30. Cat. no. HWE 37: Australian Institute of Health and Welfare, Canberra.

Australian Institute of Health and Welfare (2007b) Rural, regional and remote health – A study on mortality (2nd edition). Australia Institute of Health and Welfare, Canberra.

Australian Institute of Health and Welfare (2008) Health expenditure Australia 2006–07. Health and Welfare Expenditure Series no. 31. Cat. no. HWE 37. Australian Institute of Health and Welfare, Canberra.

Bardsley, M., Coles, J. & Jenkins, L. (Eds.) (1989) DRGs and health care. The management of casemix. King’s Fund Publishing Office.

Barry, B. (1989) Theories of Justice, Berkeley: University of California Press.

Bayer, R., Caplan, A. L. & Daniels, N. (1983) IN SEARCH OF EQUITY: Health Needs and the Health Care System, New York, Plenum Press.

Beauchamp, T. L. & Childress, J. F. (1989) Principles of Biomedical Ethics, New York, Oxford University Press.

Bedard, K., Dorland, J., Gregory, A. W. & Roberts, J. (2000) Needs–based health care funding: implications for resource distribution in Ontario. Canadian Journal of Economics/Revue canadienne d’economique, 33, 981–1008.

Bedard, K., Dorland, J., Gregory, A.W., Rosenberg, M. (1999) Standardised mortality ratios in capitation models: empirical issues from Canadian data. Canadian Public Policy, 25, 47– 64.

Belli, P. C. (2004) The impact of resource allocation and purchasing reforms on equity. Washington, DC, The International Bank for Reconstruction and Development/ World Bank.

Bennett, A. E. & Holland, W. W. (1977) Rational planning or muddling through. The Lancet, 1, 464–466.

Bentley, T., Effros, R., Palar, K. & Keeler, E. (2008) Waste in the US health care system: A conceptual framework. RAND Corporation, Milbank Quarterly, 86, 629–659.

Berglund, C. A. (2001) Health Research, Oxford University Press.

357 ______Besley, T., Gouveia, M. & Dreze, J. (1994) “Alternative Systems of Health Care Provision”. Economic Policy, 9, 199–258.

Birch, S. & Abelson, J. (1993) Is reasonable access what we want? Implications of, and challenges to, current Canadian policy on equity in health care. International Journal of Health Services, 23, 629 – 653.

Birch, S. & Chambers, S. (1993) To each according to need: a community–based approach to allocating health care resources. Canadian Medical Association Journal, 149, 607–612.

Birch, S., Eyles, J. & Newbold, B. K. (1996) Proxies for healthcare need among populations: validation of alternatives – a study in Quebec. Journal of Epidemiology and Community Health, 50, 564–569.

Birch, S., Eyles, J., Newbold, K.B (1993) Equitable access to health care: methodological extensions to the analysis of physician utilisation in Canada. Health Economics, 2, 87–101.

Birch, S., Haas, M., Savage, E. & Van Gool, K. (2007) Targeting services to reduce social inequalities in utilisation: an analysis of breast cancer screening in New South Wales. Australia and New Zealand Health Policy, 4, 12.

Black, M. & Mooney, G. (2002) Equity in Health Care from a Communitarian Standpoint. Health Care Analysis, 10.

Blane, D., Barley, M. & Davey–Smith, G. (1997) Disease aetiology and materialist explanations of socioeconomic mortality differentials. European Journal of Public Health, 7, 385–391.

Bone, R. (2003) Agreeing to Fair Process: The Problem with Contractarian Theories of Procedural Fairness, 83 Boston University Law Review 485

Bourne, D., Pick, W., Taylor, S., Mcintyre, D., Klopper, J (1990) A Methodology for Resource Allocation in Health Care for South Africa: Part 3: A South African Health Resource Allocation Formula. South African Medical Journal, 77, 456–59.

Bowling, A. (2002) Research methods in health: investigating health and health services, Open University Press, England.

Bradshaw, J. S. (1972) A taxonomy of social need. In: McLachlan, G. (Ed.). Problems and Progress in Medical Care: Essays on Current Research. Oxford, Oxford University Press.

Braveman, P. (2003) Monitoring equity in health and healthcare: a conceptual framework. J Health Popul Nutr, 21, 181–92.

Braveman, P. & Gruskin, S. (2003) Defining equity in health. British Medical Journal, 57, 254– 258.

Braveman, P. A., Cubbin, C., Egerter, S., Chideya, S., Marchi, K. S., Metzler, M. & Posner, S. (2005) Socioeconomic Status in Health Research: One Size Does Not Fit All. Journal of the American Medical Association 294, 2879–2888.

Brennan, M. & Lancashire, R. (1978) Association of childhood mortality with housing status and unemployment. J Epidemiol Community Health, 32, 28–33.

358 ______Buchanan, A. (1981) ‘Justice: a Phillosopher Review’, in E.E. Shelp (ed.), Justice and Health care, Dordrecht: Reidel. pp. 3–21.

Buchanan, A. (1997) Health care delivery and resource allocation. In R.M. Veatch (Ed.), Medical Ethics (pp.321–361). Boston, MA: Jones and Barlett.

Buchanan, A. E. (1984) The right to a decent minimum of health care. Philosophy & Public Affairs, 13, 55–78.

Buckmaster, L. & Pratt, A. (2005) Not on my account! Cost–shifting in the Australian health system, Research Note No.6, 2005–06, Parliamentary Library, 2 September 2005. Available at: http://www.aph.gov.au/library/Pubs/rn/2005–06/06rn06.htm.

Butler, J. (2009) Reforming Hospital Financing: A Medicare Hospital Benefits Scheme?, Presentation to a public seminar organised by the Menzies Centre for Health Policy, at http://www.menzieshealthpolicy.edu.au/MCHP_V3/site/events.php#sust, accessed on 11 February 2009.

Buxton, M. & Klein, R. (1978) Allocating Health Resources. Royal Commission on the NHS. Research Paper 3. London: HMSO.

Carr–Hill, R. (1994) Efficiency and equity implications of the health care reforms. Social Science & Medicine, 39, 1189 – 201.

Carr–Hill, R. & Dixon, P. (2006) Review of the resource allocation adjustment for healthcare needs due to morbidity and life circumstances and other factors. NHS Scotland Resource Allocation Committee, Technical Report, England.

Carr–Hill, R. & Sheldon, T. (1992) “Rationality and the use of formulae in the allocation of resources to health care”. Journal of Public Health Medicine, 14, 117–26.

Carr–Hill, R. A., Jamison, J. Q., O’reilly, D., Stevenson, M. R., Reid, J. & Merriman, B. (2002) Risk adjustment for hospital use using social security data: cross sectional small area analysis. British Medical Journal, 324, 390–392.

Carr–Hill, R. A., Sheldon, T. A., Smith, P., Martin, S., Peacock, S. & Hardman, G. (1994) Allocating resources to health authorities: development of method for small area analysis of use of inpatient services. British Medical Journal, 309, 1046–1049.

Carstairs, V. (1981) Multiple deprivation and health state. Community Medicine, 3, 4 – 13.

Carstairs, V. (1982) Health and social deprivation. In Recent Advances in Community Medicine (Edited by Smith A.). Churchill Livingstone, Edinburgh.

Carstairs, V. & Morris, R. (1989) Deprivation, mortality and resource allocation. Journal of Public Health 11, 364–372.

Castro–Leal, F., Dayton, L., Demery, L. & Mehra, K. (2000) Public spending on health care in Africa: do the poor benefit? Bulletin of the World Health Organisation, 78.

359 ______Centre for Epidemiology and Research (2007) 2006 Report on adult population health in New South Wales. Sydney: NSW Department of Health. Available at www.health.nsw.gov.au/publichealth/survey/hsurvey.html.

Cloke, P., Milbourne, P. & Thomas, C. (1994) Lifestyles in rural England. London, Rural Development Commission.

Cohen, J., Neumann, P. & Weinstein, M. (2009) Does preventative care save money? Health economics and the presidential candidates. New England Journal of Medicine, 358, 661– 663.

Crampton, P. & Laugesen, M. (1995) The use of indices of need in resource allocation formulae for primary health care. Discussion Paper No. 3. Wellington, Health Services Research Centre.

Creswell, J. W. (2003) Research design: qualitative, quantitative and mixed methods approaches (2nd edition), London, Sage: Thousand Oaks, California, USA.

Crowder, G. (2003) Pluralism, Relativism and Liberalism in Isaiah Berlin. Refereed paper presented to the Australasian Political Studies Association Conference, University of Tasmania, Hobart 29 September – 1 October 2003. http://www.utas.edu.au/government/APSA/GCrowder.doc.pdf.

Csdh (2008) in a generation: Health equity through action on the social determinants of health. Final Report of the Commission on Social Determinants of Health. Geneva, World Health Organisation.

Culyer, A. J. (1976a) Need and the National Health Service, London, Martin Robertson.

Culyer, A. J. (1976b) Need and the National Health Service, Oxford, Martin Robertson.

Culyer, A. J. (1991) Need, Equality and Social Justice, University of York, Centre for Health Economics.

Culyer, A. J. (1995) Need – the Idea Won’t Do – but We Still Need It. Social Science & Medicine, 40, 727–730.

Culyer, A. J. (2001) Equity – some theory and its policy implications. Journal of Medical Ethics, 27, 275–283.

Culyer, A. J. (2006) The bogus conflict between efficiency and vertical equity. Health Economics, 15, 1155–1158.

Culyer, A. J., Lavers, R. J. & Williams, A. (1971) ‘Social indicators: health’. Social Trends, 2, 31– 42.

Culyer, A. J., Van Doorslaer, E. & Wagstaff, A. (1992a) Comment: Utilisation as a Measure of Equity. Journal of Health Economics, 11, 93–98.

Culyer, A. J., Vandoorslaer, E. & Wagstaff, A. (1992b) Access, Utilisation and Equity – a Further Comment. Journal of Health Economics, 11, 207–210.

360 ______Culyer, A. J. & Wagstaff, A. (1993) Equity and Equality in Health and Health–Care. Journal of Health Economics, 12, 431–457.

Daniels, N. (1985a) Fair Equality of Opportunity and Decent Minimums – a Reply to Buchanan. Philosophy & Public Affairs, 14, 106–110.

Daniels, N. (1985b) Just Health Care, Cambridge University Press, Cambridge, 140–142.

Daniels, N. (1999) Decisions about access to health care and accountability for reasonableness. Journal of Urban Health–Bulletin of the New York Academy of Medicine, 76, 176–191.

Daniels, N. (2000) Accountability for reasonableness. BMJ, 321, 1300–1.

Daniels, N., Kennedy, B. P. & Kawachi, I. (1999) Why justice is good for our health: The social determinants of health inequalities. Daedalus, 128, 215–251.

Daniels, N. & Sabin, J. (1997) Limits to health care: Fair procedures, democratic deliberation, and the legitimacy problem for insurers. Philosophy & Public Affairs, 26, 303–350.

Daniels, N. & Sabin, J. (1998) The ethics of accountability in managed care reform. Health Affairs, 17, 50–64.

Daniels, N. & Sabin J.E (2002) Settings limits fairly : can we learn to share medical resources?, Oxford, Oxford University Press.

De Vaus, D. A. (2001) Research design in social research, London, SAGE Publications Ltd.

Denier, Y. (2007) Efficiency, Justice and Care: Philosophical Reflections on Scarcity in Health Care, International Library of Ethics, Law, and the New Medicine, Springer, New York

Denzin, N. K. (1970) The Research Act in Sociology. London: Butterworth.

Denzin, N. K. (1978) Sociological Methods: A Sourcebook, 2nd edn. New York, NY: McGraw– Hill.

Department of Health (1976) Sharing Resources for Health in England: Report of Resource Allocation Working Party (RAWP). London, HMSO

Department of Health (2005a) Exposition Books: 2006–07 and 2007–08 Primary Care Trust inital revenue resource limits. http://www.dh.gov.uk/en/policyandguidance/organisationpolicy/financeandplanning/Alloca tions/DH_4104471 Department of Health, UK.

Department of Health (2005b) Resource Allocation: Weighted Capitation Formula, 5th Edition, United Kingdom.

Department of Health (2009) Tackling Health Inequalities: 10 Years On – A review of developments in tackling health ineqalities in England over the last 10 Years. Department of Health, England.

Department of Health and Social Security (1976) Sharing Resources for Health in England: Report of the Resource Allocation Working Party. London: HMSO.

361 ______Diderichsen, F. (2004) Resource Allocation for Health Equity: Issues and Methods. HNP Discussion Paper, The International Bank for Reconstruction and Development/ The World Bank, Washington, DC.

Diderichsen, F., Varde, E., Whitehead, M. (1997) Resource allocation to health authorities: the quest for an equitable formula in Britain and Sweden. British Medical Journal, 315, 875– 878.

Doggett, J. (2009) OUT OF POCKET: Rethinking health copayments. Centre for Policy Development, Occassional Papers – Number 9.

Donabedian, A. (1973) Aspects of medical care administration, Cambridge: MA, Harvard University Press.

Draper, G., Turrell, G. & Oldenburg, B. (2004) Health inequalities in Australia: mortality. Health Inequalities Monitoring Series No.1. AIHW Cat. No. PHE 55. Queensland University of Technology and Canberra, Australian Institute of Health and Welfare.

Drummond, M. (1991) Output measurement for resource–allocation decisions in health care. In: McGuire A., Fenn P, Mayhew K, eds. Providing health care. The economics of alternative systems of finance and delivery. Oxford: Oxford University Press.

Drummond, M. F., Iglesias, C. P. & Cooper, N. J. (2008) Systematic reviews and economic evaluations conducted for the National Institute for Health and Clinical Excellence in the United Kingdom: A game of two halves? International Journal of Technology Assessment in Health Care, 24, 146–150.

Drummond, M. F., Sculpher, M. J., Torrance, G. W., O’brien, B. J. & Stoddart, G. L. (2005) Methods for the Economic Evaluation of Health Care Programmes, Oxford University Press

Drummond, M. F., Sculpher, M. J., Torrance, G. W., O’brien, B. J. & Stoddart, G. L. (2006) Methods for the Economic Evaluation of Health Care Programmes, Oxford University Press

Duckett, S. (2000) ‘The evolution of the purchaser role for acute in–patient services in Australia’, in A. L. Bloom (ed.). Health Reform in Australia and New Zealand, Oxford University Press, Melbourne, pp. 147–160.

Duckett, S. (2008a) ‘The Australian health care system: reform, repair or replace?’ Australian Health Review, 32, 322–329.

Duckett, S. (2008b) Design of price incentives for adjunct policy goals in formula funding for hospitals and health services. BMC Health Services Review, 8.

Duckett, S., Daniels, S., Kamp, M., Stockwell, A., , G. & Ward, M. (2008) Pay for Performance in Australia: Queensland’s new Clinical Practice Improvement Payment. Journal of Health Services Research and Policy, 13, 174–177.

Duckett, S. J. (1998) Casemix funding for acute hospital inpatient services in Australia. Medical Journal of Australia, 169, S17 – S21.

362 ______Duckett, S. J. (2004) The Australian Health Care System. South Melbourne, VIctoria. Oxford University Press

Eagar, K., Gordon, R., Hodkinson, A., Green, J., Eagar, L., Erven, J., Cromwell D., Leotta, T. & Rigley, A. (1997) The Australian National Sub–Acute and Non–Acute Patient Classification (AN–SNAP): Report of the National Sub–Acute and Non–Acute Casemix Classification Study,. Wollongong: Centre for Health Service Development, University of Wollongong.

Eagar, K. & Hindle, D. (1994) Casemix in Australia: an Overview. Commonwealth Dept of Health and Human Services, Canberra. ISBN 0 642 200327.

Eagar, K. & Hindle, D. (1995) Funding the NSW Health System: options and opportunities. Centre for Health Service Development, University of Wollongong and NSW Department of Health, Sydney.

Eager, K., Garrett, P. & Lin, V. (2001) Health planning: Australia perspectives. Allen & Unwin.

Eaves, D. (1998) An examination of the concept of equity and the implications for health policy if equity is re–asserted as one of the key government objectives for the National Health Service. Jounal of Nursing Management, 6, 125.

Eckstein, G. & Gibberd, R. (1994) A Relative Health Need Index for New South Wales Areas and Districts, Newcastle, Health Services Research Group: University of Newcastle.

Elshaug, A., Moss, J., Littlejohns, P., Karnon, J., Merlin, T. & Hiller, J. (2009) Identifying existing health care services that do not provide value for money. Medical Journal of Australia, 190, 269–273.

Ensor, T. & Witter, S. (2001) Health economics in low income countries: adapting to the reality of the unofficial economy. Health Policy, 57, 1–13.

Evans, D., Tandon, A., Murray, C. & Lauer, J. (2001) Comparative efficiency of national health systems: cross national econometric analysis. BMJ, 323, 307 – 310.

Evans, R. G. (1979) Health Care Economics, Canadian College of Health Service Executives, Ottawa.

Eyles, J., Birch, S., Chambers, S., Hurley, J. & Hutchison, B. (1991) A Needs–based Methodology for Allocating Health Care Resources in Ontario Canada: Development and an Application. Social Science and Medicine, 33, 489–500.

Foley, M. (2000) The changing public–private balance. In: Bloom AL, ed. Health reform in Australia and New Zealand. Melbourne, Oxford University Press: 99–114.

Forster, D. P. (1977) Mortality, Morbidity, and Resource Allocation. The Lancet, 309, 997–998.

Forster, D. P. (1979) The relationship between health needs, socio–environmental indices, general practitioner resources and utilisation. J. Chronic Dis., 32, 333.

Fox, P. (1978) Managing health resources. In By Guess or by What? Information without Design in the NHS (Edicated by McLachlan G.). Nuffield Provincial Hospitals Trust, London.

363 ______Gakidou, E. E., Murray, C. J. L. & Frenk, J. (2000) Defining and measuring health inequality: an approach based on the distribution of health expectancy. Bulletin of World Health Organisation, 78, 42–54.

Garling, P. (2008) Final Report of the Special Commission of Inquiry: Acute Care Services in NSW Public Hospitals. Sydney, Special Commission of Inquiry.

Gibbs, A., Sondalini, R. & Pearse, J. (2002) The NSW health resource distribution formula and health inequalities. NSW Public Health Bulletin, 13, 42–44.

Gibson, A., Asthana, S., Brigham, P., Moon, G. & Dicker, J. (2002) Geographies of need and the new NHS: methodologies issues in the definition and measurement of the health needs of local populations Health and place, 8, 47–60.

Gilbert, R. & Bennett, C. (1991) Move to fairer resource allocation: NSW Health Department. NSW Public Health Bulletin, 2, 30–31.

Gilbert, R., Gibberd, B. & Stewart, J. (1992) The New South Wales Resource Allocation Formula: A Method for Equitable Health Funding. Australian Health Review, 15, 6–21.

Gilgun, J. (1995) We share something special: The moral discourse of incest perpetrators. Journal of Marriage and Family, 57, 265–281.

Gillon, R. (1986) Philosophical Medical Ethics, John Wiley & Sons.

Gillon, R. (1994a) The four principles revisited. In Gillon R, editor. Principles of Health Care Ethics. Chichester: Wiley, p319–334.

Gillon, R. (1994b) Preface: medical ethics and the four principles. In: Gillion R, editor. Principles of Health Care Ethics. Chichester: Wiley, pxxi–xxxi.

Gilson, L. (1994) Equity in the Finance and Delivery of Health–Care – an International Perspective – Vandoorslaer,E, Wagstaff,a, Rutten,F. Health Policy and Planning, 9, 347–349.

Gilson, L. & Mcintyre, D. (2005) Removing user fees for primary care in Africa: the need for careful action. BMJ, 331, 762–765.

Glaser, B. & Strauss, A. (1967) The discovery of grounded theory: strategies for qualitative research. New York: Aldine de Gruyter.

Glover, J., Hetzel, D. & Tennant, S. (2004) The socioeconomic gradient and chronic illness and associated risk factors in Australia. Australian and New Zealand Health Policy, 1.

Goddard, M. & Smith, P. (2001) Equity of access to health care services: : Theory and evidence from the UK. Social Science & Medicine, 53, 1149–1162.

Gold, M. R., Siegel, J. E., Russell, L. B. & Weinstein, M. C. (1996) Cost–effectiveness in health and medicine. New York: Oxford University Press.

Gordon, D. (1995) Census based deprivation indices: their weighting and validation. Journal of Epidemiology and Community Health, 49, S39–S44.

364 ______Grant, C. & Lapsley, H. M. (1993) The Australian Health Care System, 1992, in Health Services Administration, No. 75, University of New South Wales, Sydney.

Green, A., Ali, B., Naeem, A. & Ross, D. (2000) Resource allocation and budgetary mechanisms for decentralised health systems: experiences from Balochistan, Pakistan. Bulletin of the World Health Organisation, 78.

Green, J. (1994) Qualitative paradigm evaluation. In: N. Denzin and Y. Lincoln (eds), Handbook of qualitative research, Sage: Thousands Oaks, California, USA, pp. 530–544.

Green, J. & Thorogood, N. (2004) Qualitative Methods for Health Research, London, SAGE Publications Ltd.

Greene, J. C., Caracelli, V. J. & Graham, W. F. (1998) Towards a conceptual framework for mixed– method evaluation designs. Educational Evaluation and Policy Analysis, 11, 255–274.

Greenhalgh, T. & Peacock, R. (2005) Effectiveness and efficiency of search methods in systematic reviews of complex evidence: audit of primary sources. British Medical Journal 331, 1064 – 1065.

Grumbach, K., Vranizan, K. & Bindman, A. B. (1997) Physician supply and access to care in urban communities. Health Affairs, 16, pp. 71–86.

Gutmann, A. & Thompson, D. (1996) Democracy and Disagreement. Cambridge, Massachusetts: The Beknap Press of the Harvard University Press.

Guyatt, G., Baumann, M., Pauker, S., Halperin, J., Maurer, J., Owens, D. K., Tosteson, A. N. A., Carlin, B., Gutterman, D., Prins, M., Lewis, S. Z. & Schunemann, H. (2006) Addressing Resource Allocation Issues in Recommendations From Clinical Practice Guideline Panels. CHEST Journal of American College of Chest Physicians, 129, 182–187.

Gwatkin, D. R. (2003) How well do health programmes reach the poor? Lancet, 361, 540–541.

Gwatkin, D. R. (2004) Assessing inequalities in maternal mortality. Lancet, 363, 5–5.

Gwatkin, D. R. (2005) How much would poor people gain from faster progress towards the Millennium Development Goals for health? Lancet, 365, 813–817.

Gwatkin, D. R., Bhuiya, A. & Victora, C. G. (2004) Making health systems more equitable. Lancet, 364, 1273–1280.

Haddix, A. C., Teutsch, S. M. & Corso, P. S. (2002) Prevention Effectiveness: A Guide to Decision Analysis and Economic Evaluation, Oxford University Press.

Hailey, D. (1997) Australian economic evaluation and government decisions on pharmaceuticals, compared to assessment of other health technologies. Soc Sci Med, 45, 563 – 581.

Hanson, K., Ranson, K., Oliviera–Cruz, V. & Mills, A. (2001) Constraints to Scaling up health interventions: A conceptual framework and empirical analysis. CMH Working Paper Series WG5:14. World Health Organisation (WHO).

Hay, B. (1997) Procedural Justice––Ex Ante vs. Ex Post, 44 UCLA Law Review 1803.

365 ______Hayes, L. J., Quine, S., Taylor, R. & Berry, G. (2002) Socio–economic mortality differentials in Sydney over a quarter of a century, 1970–94,. Australia and New Zealand Health Health, 26, 311–17.

Health Canada (2008) “Certain Circumstances”: Issues in Equity and Responsiveness in Access to Health Care in Canada: A Collection of papers and reports prepared for Health Canada.

Healy, J., Sharman, E. & Lokuge, B. (2006) Australia: Health system review. Health Systems in Transition. 8, 1–158.

Hindle, D. (2002) Health care funding in New South Wales: from health care needs to hospital outputs. Australian Health Review, 25, 71.

Holland, W. W. (1986) The RAWP review: pious hopes. Resource Allocation Working Party. Lancet, 8, 1087–90.

House, J. S. (2001) Understanding social factors and inequalities in health: 20th century progress and 21st century prospects, Health and Social Behavious, .

House of Commons Health Committee (2009) Health Inequalities: Third Report of Session 2008– 09, Volume 1, Report, together with formal minutes ordered by the House of Commons, UK.

House of Representatives Standing Committee on Health and Ageing (2006) The Blame Game: report on the inquiry into health funding, Canberra.

Hsiao, W. & Liu, Y. (2001) Health Care Financing: Assessing Its Relationship to Health Equity. Challenging Inequities in Health from ethics to action.

Hsiao, W. C. (2007) Why Is A Systemic View Of Health Financing Necessary? Health Affairs, 26, 950–961.

Hutubessy, R., Baltussen, R., Tan Torres–Edejer, T. & Evans, D. (2003a) WHO–CHOICE: Choosing interventions that are cost–effective. In: Health system performance assessment: debate, new methods and empiricism.

Hutubessy, R., Chisholm, D., Edejer, T. & Who–Choice (2003b) Generalised cost–effectiveness analysis for national–level priority–setting in the health sector. Cost Effectiveness and Resource Allocation, 1, 8.

IPART (1998) A review of NSW Health: Independent Pricing and Regulatory Tribunal, Sydney.

Jarman, B. (1983) Identification of underprivileged areas. British Medical Journal, 286.

Johannesson, M. (1996) Theory and Methods of Economic Evaluation in Health Care, Kluwer Academic Publisher, Amsterdam.

Johnson, R. B. & Onwuegbuzie, A. J. (2004) Mixed Methods Research: A Research Paradigm Whose Time Has Come. Educational Researcher, 33, 14–26.

Johnstone, P. (2004) Mixed methods, mixed methodology health service research in practice. Qualitative Health Research, 14, 259–271.

366 ______Jolliffe, I. T. (1986) Principal component analysis. Springer Verlag, New York.

Jolliffe, I. T. (2002) Principal component analysis (2nd ed.). Springer–Verlag, New York Inc.

Kaplan, G. A., Pamuk, E. R., Lynch, J. W., Cohen, R. D. & Balfour, J. L. (1996) Inequality in income and mortality in the United States: analysis of mortality and potential pathways. 312, 999–1003.

Kapur, D. & Webb, R. (2000) Governance–related Conditionalities of the International Financial Institutions, G–24 Discussion Paper Series No. 6. New York and Geneva: United Nations.

Kimberly, J. R., Pouvourville, G. D. & D’aunno, T. (2008) The Globalisation of Managerial Innovation in Health Care. Cambridge University Press

King’s Fund (2005) Payment by Results. Available at: http//www.kingsfund.org.uk/resources/briefings/payment_by.html

Kirigia, J. (1998) Cost–utility analysis of schistosomiasis intervention strategies in Kenya. Environment and Development Economics, 3, 319 – 346.

Kirkwood, B., Kirkwood, B.R., Sterne, J. (1999) Essentials of medical statistics, Blackwell Scientific Publication Oxford.

Kluge, E.–H. W. & Tomasson, K. (2002) Health Care Resource Allocation: Complicating Ethical Factors at the Macro–allocation Level. Health Care Analysis, 10, 209–220.

Knox, E., Marshall, T., Kane, S., Greene, A. & Mallett, R. (1980) Social and health care determinants of area variations in prenatal mortality. Community Med., 2.

Koivasulo, M. (2003) “European health policies – moving towards markets in health?”. Eurohealth, 9, pp.1–4.

Korda, R. J., Butler, J. R. G., Clements, M. S. & Kunitz, S. J. (2007) Differential impacts of health care in Australia: trend analysis of socioeconomic inequalities in avoidable mortality. International Journal of Epidemiology, 36, 157–165.

Kumaranayake, L. (1997) ‘The role of regulation: influencing private sector activity within health sector reforms’ Journal of International Development, 9, 641–9.

Lake, S. (2000) Modeling ‘Need’ in Health Sector Resource Allocation Formulae: Application in a Low–income Country. Health Systems Financing in Low–Income African and Asian Countries. CERDI, Clermont Ferrand.

Lauer, J. A., Rohrich, K., Wirth, H., Charette, C., Gribble, S. & Murray, C. J. L. (2003) Making choices in health : WHO guide to cost–effectiveness analysis PopMod: a longitudinal population model with two interacting disease states. Geneva, World Health Organisation:, 1.

Le Grand, J. (1978) The distribution of public expenditure: the case of health care. Economica, 45, 125 – 142.

367 ______Le Grand, J. (1982) The strategy of equality: redistribution and the social services, London; Boston, G. Allen & Unwin.

Le Grand, J. (1987) Equity, health and health care. Social Justice Research, 1.

Le Grand, J. (1991) The distribution of health care revisited: a commentary on Wagstaff, van Doorslaer and Paci, and O’Donnell and Propper.[comment]. Journal of Health Economics, 10, 239–45.

Leeder, S. R. (2003) Achieving Equity in the Australian Healthcare System. The Medical Journal of Australia, 179, 475 – 478.

Lincoln, Y. S. & Guba, E. G. (2000) Paradigmatic controversies, contradictions and emerging confluences. In. IN DENZIN, N. K. & LINCOLN(EDS), Y. S. (Eds.) Handbook of qualitative research (2nd edition, pp. 163–188). Thousand Oaks, CA: Sage.

Litaker, D. & Love, T. E. (2005) Health care resource allocation and individuals’ health care needs: examining the degree of fit. Health Policy, 73, pp. 183–193.

Lowe, A. (2000) “Accounting in health care: some evidence on the impact of casemix systems”. British Accounting Review, 32, pp. 189–211.

Macintyre, S., Mckay, L., Der, G. & Hiscock, R. (2003) Socio–economic position and health: what you observe depends on how you measure it.

Macklin, R. (1987) Mortal Choices. Houghton Miflin Company, Boston, 149–164.

Marchand, S., Wikler, D., Landesmanm, B. (1998) Class, health and justice. The Midbank Quarterly, 76, 449–467.

Marini, G. & Street, A. (2006) A transaction costs analysis of changing contractual relations in the English NHS. Health Policy, 72, pp. 42–68.

Marmot, M. (2001) Inequalities in health. New England Journal of Medicine, 345, 134–136.

Marshall, C. & Rossman, G. B. (1999) Designing qualitative research (3rd ed.), Thousand Oaks, CA: Sage.

Mathers, C. (1994) Health differentials among adult Australians aged 25 – 64 years. Australian Institute of Health and Social Welfare. Health Monitoring Series No. 1.

Mathers, C., Vos, T. & Stevenson, C. (1999) The burden of disease and injury in Australia, AIHW Cat. No. PHE 17. Canberra, Australian Institute of Health and Welfare.

Mathews, J. D. (1997) Historical, social and bilogical understanding is needed to improve Aboriginal health. Recent Adances in Mibrobiology, 5, 257–84.

Maynard, A. & Ludbrook, A. (1980a) Applying resource allocation formulae to constituent parts of the UK. Lancet, 85–87.

Mays, N. (1987) Measuring morbidity for resource allocation. British Medical Journal (Clin Res Ed), 295, 703–706.

368 ______Mays, N. (1995) Geographical resource allocation in the English National Health Service, 1971– 1994: the tension between normative and empirical approaches. International Journal of Epidemiology, 24, 96–102.

Mays, N. & Bevan, G. (1986) Resource allocation in the health service: A review of methods of the resource allocation working party (RAWP). Occasional Paper on Social Administration No. 81. Bedford Square Press.

Mcclelland, A. & Scotton, R. (1998) Poverty and health, in R. Fincher and J. Nieuwenhuysen (eds) Australian Poverty: Then and Now., Melbourne, VIC: Melbourne University Press.

Mcdonough, P., Duncan, D. J., Williams, D. R. & House, J. (1997) Income dynamics and adult mortality in the US, 1972–1989. American Journal of Public Health, 18, 341 – 378.

Mcintyre, D. & Gilson, L. (2000) Redressing dis–advantage: Promoting vertical equity within South Africa. Health Care Analysis, 8, 235–258.

Mcintyre, D. & Gilson, L. (2002) Putting equity in health back onto the social policy agenda: experience from South Africa. Social Science & Medicine, 54, 1637–1656.

Mcintyre, D. & Mooney, G. (2007) The Economics of Health Equity, Cambridge University Press.

Mcintyre, D., Muirhead, D. & Gilson, L. (2002) Geographic patterns of deprivation in South Africa: informing health equity analyses and public resource allocation strategies. Health Policy and Planning, 17, 30–39.

Mcintyre, D., Thiede, M., Dahlgren, G. & Whitehead, M. (2006) What are the economic consequences for households of illness and of paying for health care in low– and middle– income country contexts? Social Science & Medicine, 62, 858–865.

Mckneally, M. F., Dickens, B. M., Meslin, E. M. & Singer, P. A. (1997) Bioethics for Clinicians: Resource Allocation. Canadian Medical Association Journal, 157, 163–167.

Mcmillan, J. (2002) Allocation of resources. Surgery, Oxford, 20, pp. 117–120.

Mcpake, B., Kumaranayake, L. & Normand, C. (2002) Health Economics: An International Perspective, Routledge, London and New York.

Midwinter, A. (2002) Territorial resource allocation in the UK: a rejoinder on needs assessment. Regional Studies, 35, pp. 563–577.

Mitton, C. & Donaldson, C. (2003) Resource Allocation in Health Care: Health Economics and Beyond. Health Care Analysis, 11, 245 – 257.

Mooney, G. (1983) Equity in health care: confronting the confusion. Effective Health Care, 1, 179– 85.

Mooney, G. (1986) Economics, Medicine and Health Care, Wheatsheaf, Brighton.

Mooney, G. (1987) What does equity in health mean. World Health Statistic Quarterly, 40, 296 – 303.

369 ______Mooney, G. (1989) QALYs: are they enough? A health economist’s perspective. J Med Ethics, 15, 148–152.

Mooney, G. (1992) Just health care: only medicine? Economics, Medicine and Health Care. 2nd ed. Hertfordshire, Harvester Wheatsheaf.

Mooney, G. (1994a) Equity: Key issues in health economics. New York; London,, Harvester Wheatsheaf.

Mooney, G. (1994b) Priority setting in health care. Key Issues in Health Economics. New York; London, Harvester Wheatsheaf.

Mooney, G. (1995) The Politics and Economics of The Rural Health System: Equity and Efficiency. 3rd National Rural Health Conference. Mt Beauty, 3–5 February 1995

Mooney, G. (1998a) “Communitarian claims” as an ethical basis for allocating health care resources. Social Science and Medicine, 47, 1171–1180.

Mooney, G. (1998b) How much should we be spending on health services for Aboriginal and Torres Strait Islander people? Medical Journal of Australia, 169, 508–509.

Mooney, G. (2000a) Judging goodness must come before judging quality––but what is the good of health care? International Journal for Quality in Health Care, 12, 389–94.

Mooney, G. (2000b) Vertical equity in health care resource allocation. Health Care Analysis, 8, 203–15.

Mooney, G. (2003) Inequity in Australian health care: how do we progress from here? Australian & New Zealand Journal of Public Health, 27, 267–70.

Mooney, G. (2006) Communitarian claims as an ethical basis for allocating health care resources. Social Science & Medicine, 47, pp. 1171–1180.

Mooney, G., Hall, J., Donaldson, C. & Gerard, K. (1991) Utilisation as a Measure of Equity – Weighing Heat? Journal of Health Economics, 10, 475–480.

Mooney, G. & Houston, S. (2004) An Alternative Approach to Resource Allocation Weighted Capacity to Benefit Plus MESH Infrastructure. Applied Health Economics and Health Policy, 3, 29–33.

Mooney, G. & Jan, S. (1997) Vertical equity: weighting outcomes? or establishing procedures? Health Policy, 39, 79–87.

Mooney, G., Jan, S., Seymour, J. & Wiseman, V. (2002) Staking a claim for claims: a case study of resource allocation in Australian Aboriginal health care. Social Science & Medicine, 54, 1657–1667.

Mooney, G., Russell, E. M. & Weir, R. D. (1986) Choices for health care: a practical introduction to the economics of health care provision. London: Macmillan.

Mooney, G. & Scotton, R. (1998) Economics and Australian Health Policy, Allen & Unwin, Sydney, Australia.

370 ______Morse, J. (2005) Evolving trends in qualitative research: advances in mixed–method design. Qualitative Health Research, 15, 583–585.

Murray, C. & Frenk, J. (2000) A framework for assessing the performance of health systems. Bulletin World Health Organisation, 78, 717 – 731.

Murray, C. J. L., Gakidou, E. E. & Frenk, J. (1999) Health inequalities and social group difference: What should we measure? Bulletin of the World Health Organisation, 77, 537–543.

National Health and Hospital Reform Commission (2008) ‘Beyond Blame Game: Accountability and performance benchmarks for the next Australian Health Care Agreements’ at http://www.nhhrc.org.au/internet/nhhrc/publishing.nsf/Content/BeyondTheBlameGame.pdf , accessed on 20 March 2009.

National Health and Hospital Reform Commission (2009) The Australian Health Care System: The Potential for Efficiency Gains, A review of the literature: Background paper prepared for the National Health and Hospitals Reform Commission, Australian Government.

National Health Strategy (1992) Enough to make you sick: how income and environment affect health: Australia National Health Strategy, Canberra.

Neuman, W. L. (2000) Social research methods: quantitative and qualitative approaches (4th edition), Boston, Allyn and Bacon.

Newbold, K. B., Eyles, J. & Birch, S. (1995) Equity in health care: Methodological contributions to the analysis of hospital utilisation within Canada.

Normand, C. (1991) Economics, health and the economics of health. British Medical Journal, 303, 1572–1577.

Nozick, R. (1974) Anarchy, state, and utopia. New York, Basic Books,149–82.

NSW Health (1990) A Resource Allocation Formula for the NSW Health System:NSW Department of Health. Sydney.

NSW Health (1996) NSW area strategic planning guidelines for budget allocation, achievement of benchmark cost and cross–boundary purchasing. Sydney, New South Wales Department of Health.

NSW Health (1999) Resource Distribution Formula Technical Paper 1998/99 Revision, Structural and Funding Policy Branch, Policy Division, NSW Department of Health, Sydney.

NSW Health (2000) Episode Funding Guidelines 2000–2001, State Health Publication No. (SEP) 000134, NSW Department of Health, Sydney.

NSW Health (2004a) Annual report 2003–04. Sydney, NSW Department of Health.

NSW Health (2004b) NSW Funding Guidelines 2004/05: Inter–Governmental & Funding Strategies Branch Sydney.

NSW Health (2004c) NSW Health and Equity Statement: In All Fairness: increasing equity in health across NSW. NSW Department of Health 2004. Sydney.

371 ______NSW Health (2006a) The health of the people of New South Wales: Report of the Chief Health Officer: NSW Department of Health, Sydney: http://www.health.nsw.gov.au/publichealth/chorep/atsi/atsi_avopremdth_full.asp, accessed 2008.

NSW Health (2006b) Northern Sydney Central Coast Area Health Services – 2005–2006 Annual Report. Sydney.

NSW Health (2006c) NSW Funding Guidelines: 2006/2007. Sydney, NSW Health Inter– Government & Funding Strategies Branch 2006.

NSW Health (2007a) Annual Report 2006/07: New South Wales Department of Health, NSW Health, Sydney.

NSW Health (2007b) A New Direction for NSW – State Health Plan Towards 2010. NSW Department of Health, North Sydney.

NSW Health (2008) The health of the people of New South Wales: Report of the Chief Health Officer. Sydney: NSW Department of Health. http://www.health.nsw.gov.au/pubs/2008/chorep08_summary.html accessed June 2009.

NSW Health (2009) New South Wales Government Department of Health Organisation Structure. http://www.health.nsw.gov.au/aboutus/structure.asp Accessed January 2009. IN HEALTH, D. O. (Ed.). Sydney.

NSW Health (1988) The Case for a Redistribution of Health Resources. Sydney, NSW Department of Health.

NSW Health (1990) A resource Allocation Formula for the NSW Health System: NSW Department of Health, Sydney.

NSW Health (1993) A resource allocation formula for the NSW health system: 1993 revision. Sydney, New South Wales Department of Health: Service Development and Planning Branch.

NSW Health (1996) Implementation of the economic statement for health. Sydney, New South Wales Department of Health: Structural & Funding Policy Branch.

NSW Health (1999) Resource Distribution Formula: Technical Paper 1998/99. Sydney, New South Wales Department of Health.

NSW Health (2001) Report of the Greater Metropolitan Services Implementation Group, State Health Publication No. (HPA) 010074, NSW Department of Health. Sydney.

NSW Health (2002) NSW Funding Guidelines for Emergency Department Services, 2002/2003, North Sydney: NSW Department of Health.

NSW Health (2004) Health and Equity in New South Wales Department of Health. NSW Public Health Bulletin, 15.

NSW Health (2005) Resource Distribution Formula. Technical Paper: 2005 Revision, New South Wales Department of Health, Sydney.

372 ______NSW Health (2006) The health of the people of New South Wales: Report of the Chief Health Officer: NSW Department of Health, Sydney: http://www.health.nsw.gov.au/publichealth/chorep/atsi/atsi_avopremdth_full.asp, accessed 2008.

NSW Health (2008) NSW Episode Funding Policy 2008/09: Inter–Governmental & Funding Strategies Branch, New South Wales Department of Health, Sydney.

NSW Health Council (2000) Report of the NSW Health Council – A Better Health System for NSW, Sydney: NSW Government.

NSW Health. (2003a) NSW Funding Guidelines 2003/04, NSW Health Department. Sydney.

NSW Parliament (2008) Health Resource Distribution Formula – Extract from Hansard and Papers Thursday 4 December 2008: comments by Hon. Melinda Pavey directed to Deputy Director General of Strategic Development, Dr Richard Matthews. Parliament Proceedings Transcripts Page 12634.

O’donnell, O. & Propper, C. (1991) Equity and the distribution of U.K. national health service resources. Jounal of health economics, 10, 247 – 249.

Ocana–Riola, R., Saurina, C., Fernandez–Ajuria, A., Lertxundi, A., Sanchez–Cantalejo, C., Saez, M., Ruiz–Ramos, M., Barcelo, M. A., March, J. C., Martinez, J. M., Daponte, A. & Benach, J. (2008) Area deprivation and mortality in the provincial capital cities of Andalusia and Catalonia (Spain). J Epidemiol Community Health, 62, 147–152.

Oliver, A. (2005) “The English National Health Service: 1979–2005”. Health Economics, 14, pp. S75–S99.

Oliver, A., Healey, A., Le Grand, J. (2002) Addressing health inequalities. Lancet, 360, 565–567.

Olsen, E. O. & Rodgers, D. L. (1991) The welfare economics of equal access. Journal of Public Economics, 45, 91–106.

Pain, C., Frankovitch, F., Cook, G. (1996) Setting targets for CABG survey in North Western Region. Journal of Public Health Medicine, 18, 449–456.

Palmer, G. R. (1991) ‘The Use if DRGs in the Management and Planning of Hospital Services’, Australian Economic Review, 1st Quarter, pp. 62–70.

Palmer, S. & Raftery, J. (1999) Economics notes: Opportunity cost. BMJ, 318, 1551–1552.

Paton, C. (1985) The policy of resource allocation and its ramifications: A review. Nuffield Provincial Hospitals Trust, London.

Patton, M. Q. (2002) Qualitative research and evaluation methods (3 edition), Sage Publications.

Peacock, S. & Segal, L. (2000) Capitation funding in Australia: Imperatives and impediments. Health Care Management Science, 3, 77–88.

Pearson, M. (2002) Allocation public resources for health: developing pro–poor approaches. London, Department for International Development (DFID).

373 ______Petrou, S. & Wolstenholme, J. (2000) A review of alternative approaches to healthcare resource allocation. Pharmacoeconomics, 18, pp. 33–43.

Phillips, D. C. & Burbules, N. C. (2000) Postpositivism and educational research, Lanham, MD, Rowman and Littlefield.

Phillips, D. L. (1979) Equality, Justice and Rectification: An Exploration in Normative Sociology, London, Academic Press.

Pollock, A. M., Price, D., Viebrock, E., Miller, E. & Watt, G. (2007) The market in primary care. British Medical Journal, 335, 475–477.

Private Health Insurance Administration Council (2007) Operations of the Registered Health Benefits Organisations Annual Report 2006–07: Private Health Insurance Administration Council, Australian Government, Sydney.

Productivity Commission (2006) Potential Benefits of the National Reform Agenda, Report to the Council of Australian Governments, Canberra.

Puffer, F. (1986) Access to primary care: A comparison of US and the UK. Journal of Social Policy, 15, 293 – 313.

Ranade, W. (Ed.) (1998) Markets and health care: a comparative analysis. Essex: Longman.

Rawls, J. (1972) A Theory of Justice, Oxford, Oxford University Press.

Rawls, J. (1985) Justice as fairness: political not metaphysical. Philosophy and Public Affairs, 14, 223 – 251.

Rawls, J. (1999) A Theory of Justice, revised edition, Oxford: Oxford University Press 1999 Chapter II, Section 14.

Resource Allocation Working Party (1976) Sharing Resources for Health: Resource Allocation Working Party. HMSO, London.

Rice, N., Dixon, P., Lloyd, D. C. & Roberts, D. (2000) “Derivation of a needs based capitation formula for allocating prescribing budgets to health authorities and primary care groups in England: regression analysis”. British Medical Journal, 320, 284 – 288.

Rice, N. & Smith, P. (2001a) Capitation and Risk Adjustment in Health Care Financing: an International Progress Report The Milbank Quarterly, 79, 81 – 113.

Rice, N. & Smith, P. (2001b) Ethics and geographical equity in health care. Journal of Medical Ethics, 27, pp. 256–261.

Ring, I. T. & Brown, N. (2002) Indigenous health: chronically inadequate responses to damning statistics. Medical Journal of Australia, 177, 629–631.

Roemer, J. E. (1996) Theories of distributive justice, Harvard University Press.

374 ______Royston, G. H. D., Hurst, J. W., Lister, E. G. & Stewart, P. A. (1992) Modelling the use of health services by populations of small areas to inform the allocation of central resources to larger regions. Socio–Economic Planning Sciences, 26, 169–80.

Russell, L. B. (1992) Opportunity costs in modern medicine. Health Affairs, 11.

Saltman, R. & Figueras, J. (1997) European : analysis of current strategies. Copenhagen; WHO Regional office for Europe.

Samuelson, P. A. & Nordhaus, W. D. (2000) Economics, seventeenth edition, New York: McGraw– Hill.

Scotton, R. B. & Owens, H. J. (1990) Case payment in Australian hospitals: issues and options: Public Sector Management Institute, Monash University, Melbourne.

Segal, L. (2008) A new Model of Primary Care: A paper for the National Health and Hospitals Reform Commission, www.nhhrc.org.au.

Segal, L. & Bolton, T. (2009) Issues facing the future health care workforce: the importance of demand modelling. Australia and New Zealand Health Policy, 6, 12.

Segal, L., Dalziel, K. & Mortimer, D. (2007) Review of Australian Economic Evaluation in Health: Time to look at the bigger picture – the role of the funding environment.

Segal, L. & Mortimer, D. (2006) A population–based model for priority setting across the care continuum and across modalities. Cost Effectiveness and Resource Allocation, 4, 6.

Segal, L. & Richardson, J. (1995) ‘Efficiency in resource allocation’, in A.H. Harris (ed.) Economics and Health: 1994: Proceedings of the Sixteenth Australian Conference of Health Economists, Australian Studies in Health Service Administration No. 79, University of New South Wales, Sydney.

Segall, M. (2003) District health systems in a neoliberal world: a review of five key policy areas. The International Journal of Health Planning and Management, 18, S5–S26.

Sen, A. (1995) Rationality and Social Choice. American Economic Review, 85, 1–24.

Sen, A. (1998) Social choice and justice. Trimestre Economico, 65, 479–504.

Sen, A. (2002) Why health equity? Health Economics, 11, 659–666.

Shannon, M. (1976) A measure of the fulfillment of health needs. International Journal Aging Human Development, 7, 253–261.

Shaw, R. & Smith, P. (2001) Allocating health care resources to reduce health inequalities In: Appelby J, Harrison A (eds) Health care UK. London, King’s Fund.

Sheldon, T. A. (1997) Formula fever: allocating resources in the NHS. British Medical Journal, 315, 964.

375 ______Sheldon, T. A., Smith, P., Borowitz, M., Martin, S. & Hill, R. C. (1994) Attempt at deriving a formula for setting general practitioner fundholding budgets. British Medical Journal, 309, 1059–1064.

Sheldon, T. A. & Smith, P. C. (2000) Equity in the allocation of health care resources. Health Economics, 9, 571–574.

Sidell, M. (1995) Health in Old Age Buckingham, Open University Press.

Smith, C. P., Ginnelly, L. & Sculpher, M. (2005) Health Policy and Economics: Opportunities and Challenges, New York, Open University Press.

Smith, P., Rice, N. & Carr–Hill, R. (2001) Capitation funding in the public sector. Journal of the Royal Statistical Society, Series A, 164, 217–57.

Smith, P., Sheldon, T. A., Carr–Hill, R. A., Martin, S., Peacock, S. & Hardman, G. (1994) Allocating resources to health authorities: results and policy implications of small area analysis of use of inpatient services. British Medical Journal, 309, 1050–1054.

Sou2000:91 (2001) Health on Equal terms. Report of the national public health commission of Sweden. Stockholm. Scand J Public Health.

Stoelwinder, J. & Viney, R. (2000) ‘A tale of two states: New South Wales and Victoria’, in A.L. Bloom (ed.) Health Reform in Australia and New Zealand, Oxford University Press, Melbourne, pp. 201–222.

Street, A., Vitikainen, K., Bjorvatn, A. & Hvenegaard, A. (2007) Introducing activity–based financing: a review of experience in Australia, Denmark, Norway and Sweden. Centre for Health Economics, University of York, CHE Research Paper 30.

Sydney Morning Herald (2008) Aboriginal malnourishment a national disgrace. Sydney Morning Herald. Sydney.

Szende, A. & Magyorosy, Z. (2004) “Health care provider payment mechanisms in the new EU members of Central Europe and the Baltic states: Current reforms, incentives, and challenges”. European Journal of Health Economics, 5, pp. 259–262.

Talbot–Smith, A. & Pollock, A. M. (2006) The New NHS—A Guide, Routledge, London.

Townsend, P. & Davidson, N. (1988) Black Report, The (1982/1988) in P. Townsend and N. Davidson (eds) Inequalities in Health. London: Penguin.

Townsend, P., Simpson, D. & Tibbs, N. (1984) Inequalities in Health in the City of Bristol. Department of Social Administration, University of Bristol.

Turrell, G. & Mathers, C. (2000) Socioeconomic status and . Medical Journal of Australia, 172, 434–8.

Turrell, G. & Mathers, C. (2001) Socioeconomic inequalities in all–cause and specific–cause mortality in Australia: 1985–1987 and 1995–1997. Int J Epidemiol., 30, 231 – 239.

376 ______United Kingdom Department of Health and Social Security. (1976) Sharing Resources for Health in England. HMSO, London.

United Nations Development Programme (2007) World population prospects: the 2006 revision population database, United Nations Population Division (http://esa.un.org/unpp/, last accessed 10 June 2008).

Van De Ven, W. P. M. M. (1996) “Market–oriented health care reforms: trends and future options”. Social Science and Medicine, 43, pp. 655–666.

Van Doorslaer, E., Masseria, C. & Koolman, X. (2006) Inequalities in access to medical care by income in developed countries. Canadian Medical Association Journal 174, 177–183.

Van Doorslaer, E., Wagstaff, A. & Rutten, F. (1993) Equity in the Finance and Delivery of Health Care: An International Perspective. Oxford: Oxford University Press.

Van Doorslaer, E., Wagstaff, A., Van Der Burg, H & Al., E. (1998) Equity in the delivery of health care: Further international comparisons. Erasmus University, Rotterdam.

Varian, H. R. (1975) Distributive justice, welfare economics and the theory of fairness. Philosophy & Public Affairs, 4, 223–247.

Viney, R., Keith, A., Williams, P. & Laffey, J. (1991) ‘Case payment: a New South Wales perspective’. Australian Health Review, 14, pp. 235–244.

Wagstaff, A. (1991) QALYs and the equity–efficiency trade–off. Journal of Health Economics, 10, 21–41.

Wagstaff, A., Paci, P. & Van Doorslaer, E. (1991a) On the measurement of inequalities in health. Social Science and Medicine, 33, 545–57.

Wagstaff, A. & Van Doorslaer, E. (1993) Equity in the finance and delivery of health care: concepts and definitions., Oxford University Press, Oxford.

Wagstaff, A. & Van Doorslaer, E. (1998) Equity in health care finance and delivery. IN CULYER, A. J. & NEWHOUSE, J. P. (Eds.) North Holland Handbook of Health Economics, (eds.).

Wagstaff, A., Van Doorslaer, E. & Paci, P. (1991) On the measurement of horizontal inequity in the delivery of health care. Journal of Health Economics, 10, 169 – 205.

Wagstaff, A., Van Doorslaer, E., Paci, P (1989) Equity in the Finance and Delivery of Health Care: Some Tentative Cross–country Comparisons. Oxford Review of Economic Policy, 5, 89– 112.

Wagstaff, A., Van Doorslaer, E., Paci, P (1991b) On the measurement of horizontal inequity in the delivery of health care. Journal of Health Economics, 10, 169–205.

Weinstein, M. & Stason, W. (1977) Foundations of cost–effectiveness analysis for health and medical practices. New England Journal of Medicine, 296, 716–21.

Westphal, J. E. (1996) Justice, Indianapolis/Cambridge, Hackett Publishing Company Inc.

377 ______Whitehead, M. (1992) The Concepts and Principles of Equity and Health. International Journal of Health Services, 22, 429–445.

Whitehead, M. (1994) Who Cares About Equity in the NHS. British Medical Journal, 308, 1284– 1287.

Whitehead, M., Dahlgren, G. & Evans, T. (2001) Equity and health sector reforms: can low–income countries escape the medical poverty trap? The Lancet, 358, 833–836.

Whitehead, M., Judge, K., Hunter, D. J., Maxwell, R., Scheuer, M. A. (1993) Tackling Inequalities in Health – the Australian Experience. British Medical Journal, 306, 783–787.

Who Commission on Macroeconomics Health (2001) Macroeconomics and health: investing in health for economic development. Report of the Commission on Macroeconomics and Health: Executive Summary. Geneva, World Health Organisation.

Wilkins, R. (1987) Health inequalities in Canada: some policy implications. Second meeting of the Canada Heart–Health Network, Montreal.

Williams, A. (1974) Need’ as a demand concept (with special reference to health). In. IN CULYER (ED.), A. (Ed.) Economic Policies and Social Goals. London, Martin Robertson.

Williams, A. (1988) Priority setting in public and private health care: A guide through the ideological jungle. Journal of Health Economics, 7, 173–183.

Williams, A. (2005) Thinking about equity in health care. Journal of Nursing Management, 13, 397–402.

Williams, A. H. (Ed.) (1993) Equity in health care: The role of ideology. In Van Doorslaer, E., Wagstaff, A. and Rutten, F. eds.,, Oxford, Oxford University Press.

Wilson, I. B., Rogers, W. H. & Chang, H. (2005) Cost–related Skipping of Medications and Other Treatments Among Medicare Beneficiaries Between 1998 and 2000: Results of a National Study. Journal of General Internal Medicine., 20, 715–720.

Wilson, J. Q. (1993) The Moral Sense, New York, Free Press.

Wilson, R. & Van Der Weyden, M. (2005) The safety of Australian healthcare: 10 years after QAHCS. Medical Journal of Australia, 182: 260–261.

Wiseman, V. & Jan, S. (2000) Resource Allocation within Australian Indigenous Communities: A Program for Implementing Vertical Equity. Health Care Analysis, 8, 217–233.

Wiseman, V., Mooney, G., Berry, G. & Tang, K. C. (2003) Involving the general public in priority setting: experiences from Australia. Social Science & Medicine., 56, 1001–12.

Wolfson, M. & Rowe, G. (2001) On measuring inequalities in health. Bulletin of the World Health Organisation, 79, 553 – 60.

World Bank (2006) World Development Report 2006: Equity and Development. Washington DC, World Bank.

378 ______Wright, J., Williams, R. & Wilkinson, J. R. (1998) Development and importance of health needs assessment. British Medical Journal, 316, 25 Apr., 1310–1313.

Zere, E., Mandlhate, C., Mbeeli, T., Shangula, K., Mutirua, K. & Kapenambili, W. (2007) Equity in health care in Namibia: developing a needs–based resource allocation formula using principal components analysis. Int J Equity Health, 6.

Zere, E. & Mcintyre, D. (2003) Equity in self–reported adult illness and use of health service in South Africa: Inter–temporal, comparison. Journal of Health Population and Nutrition, 21, 205–215.