H N P D i s c u s s i o N P a P e R 58014 WHO PAYS? Out-of-Pocket Health Spending and Equity Implications in the Middle East and North Africa Heba Elgazzar, Firas Raad, Chokri Arfa, Awad Mataria, Nisreen Salti, Jad Chaaban, Djavad Salehi-Isfahani, Sanaz Fesharaki and Mehdi Majbouri November 2010 WHO PAYS? Out-of-Pocket Health Spending and Equity Implications in the Middle East and North Africa Heba Elgazzar, Firas Raad, Chokri Arfa, Awad Mataria, Nisreen Salti, Jad Chaaban, Djavad Salehi-Isfahani, Sanaz Fesharaki and Mehdi Majbouri November, 2010 Health, Nutrition and Population (HNP) Discussion Paper This series is produced by the Health, Nutrition, and Population Family (HNP) of the World Bank's Human Development Network. The papers in this series aim to provide a vehicle for publishing preliminary and unpolished results on HNP topics to encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account this provisional character. For free copies of papers in this series please contact the individual author(s) whose name appears on the paper. Enquiries about the series and submissions should be made directly to the Editor, Homira Nassery (HNassery@worldbank.org). Submissions should have been previously reviewed and cleared by the sponsoring department, which will bear the cost of publication. No additional reviews will be undertaken after submission. The sponsoring department and author(s) bear full responsibility for the quality of the technical contents and presentation of material in the series. Since the material will be published as presented, authors should submit an electronic copy in a predefined format (available at www.worldbank.org/hnppublications on the Guide for Authors page). Drafts that do not meet minimum presentational standards may be returned to authors for more work before being accepted. For information regarding this and other World Bank publications, please contact the HNP Advisory Services at healthpop@worldbank.org (email), 202-473-2256 (telephone), or 202-522-3234 (fax). © 2010 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved. ii Health, Nutrition and Population (HNP) Discussion Paper Who Pays? Out-of-pocket health spending and equity implications in the Middle East and North Africa Heba Elgazzara, Firas Raada, Chokri Arfab, Awad Matariac, Nisreen Saltid, Jad Chaaband, Djavad Salehi-Isfahanie, Sanaz Fesharakie and Mehdi Majbourif a World Bank, Washington, DC, USA b University of Tunis, Tunis, Tunisia c Birzeit University, Birzeit, West Bank d American University of Beirut, Beirut, Lebanon e Virginia Polytechnic Institute and State University, Blacksburg, Virginia, USA f University of Southern California, Los Angeles, California, USA Paper prepared for and funded by the World Bank Washington, DC, USA, November 2010 Abstract: Ensuring affordable, effective health care and financial protection against the adverse effects of household out-of-pocket (OOP) health expenditures represents an important policy objective in most countries, yet relatively little evidence exists regarding patterns and implications of household health expenditures in the Middle East and North Africa (MENA) region. This paper examines the scope of out-of-pocket expenditures and their implications on living standards and policy reforms in six MENA countries including Yemen, the West Bank and Gaza, Egypt, Iran, Tunisia, and Lebanon. Results show that OOP payments represent a relatively high share of total national health care financing at 49 percent on average in the MENA region as of 2006. Households pay an average of 6 percent of their total household expenditure on health. Most of this OOP is spent on medications, doctor visits and diagnostic services. Lower-income and rural households generally face greater financial risk; yet this is reversed where private health services are utilized and paid for more frequently by higher-income groups. 7 to 13 percent of households face particularly high OOP payments, or catastrophic expenditures equal to at least 10 percent of household spending. Poverty rates tend to increase by up to 20 percent after health care spending is accounted for. Results are discussed in light of ongoing policy efforts to strengthen social protection for health care. Keywords: health financing, out-of-pocket spending on health, equity, health insurance, social protection Disclaimer: The findings, interpretations and conclusions expressed in the paper are entirely those of the authors, and do not represent the views of the World Bank, its Executive Directors, or the countries they represent. Correspondence Details: Heba A. Elgazzar, World Bank, 1818 H Street, Washington, DC, USA; Email: helgazzar@worldbank.org iii iv Table of Contents List of Tables and Figures.................................................................................................. vi Acknowledgements ........................................................................................................... vii I. Introduction: The Relationship between Out-of-Pocket Health Spending and Equity in MENA............................................................................................................................. 1 II. The Reliance on Out-of-Pocket Health Financing in MENA ..................................... 3 III. The Distribution of OOP Health Spending ................................................................. 8 IV. Identifying Catastrophic Health Care Expenditures ................................................. 12 V. The Impoverishing Effect of OOP Spending ............................................................ 17 VI. Policy Implications: Bridging the Gaps .................................................................... 22 VII. Conclusions ........................................................................................................... 29 Annex: Analytic Methods ................................................................................................. 31 References ......................................................................................................................... 35 v List of Tables and Figures Table 1. National health expenditures in MENA countries, 2006 .................................................................. 4 Table 2. Coverage rates by health financing scheme in MENA, 2008 ........................................................... 5 Table 3. Spatial distribution of health insurance coverage over time in Iran, 1996-2006 .............................. 6 Table 4. Average out-of-pocket spending on health care in MENA countries ............................................... 7 Table 5. Distribution of out-of-pocket health care spending (OOP) across socioeconomic status ................ 8 Table 6. Catastrophic expenditures by insurance status in Tunisia, 2005 .....................................................14 Table 7. Incidence of catastrophic expenditures before and after User-Fee Exemption Scheme in Egypt, 2007 .....................................................................................................................................................15 Table 8. Effect of out-of-pocket health care payments (OOP) on poverty rates in MENA countries ...........17 Table 9. Effect of OOP on depth of poverty in MENA countries .................................................................18 Table 10. Spatial distribution of OOP-related impoverishment in Yemen ....................................................19 Table 11. OOP-related impoverishment before and after user-fee exemption policy in Egypt, 2006/7 .......20 Table 12. Summary of policy initiatives taken to alleviate barriers to health care in MENA, 1996- 2006 .....................................................................................................................................................27 Table 13. Data sources .................................................................................................................................31 Table 14. Example of concentration index method. ......................................................................................32 Figure 1. Income inequality in maternal and child health indicators in MENA countries.............................. 1 Figure 2. Forgone health care use due to financial costs in MENA ............................................................... 6 Figure 3. OOP by health service in MENA countries .................................................................................... 9 Figure 4. OOP by health service in Iran over time, 1996-2006 .....................................................................10 Figure 5. OOP by insurance coverage status in Lebanon, 2004/5 .................................................................10 Figure 6. Incidence of catastrophic health expenditures in MENA countries ...............................................12 Figure 7. Spatial distribution of catastrophic expenditures in MENA countries ...........................................13 Figure 8. Spatial distribution of catastrophic expenditures by insurance status in Iran over time, 1996-2006 .............................................................................................................................................................14 Figure 9. Incidence of Catastrophic Spending in West Bank-Gaza, 1998-2007 ...........................................16 Figure 10. Spatial distribution of OOP-related impoverishment over time in Iran, 1996-2006 ...................19 Figure 11. Health Expenditure-Related Impoverishment Over Time in West Bank-Gaza, 1998-2007........20 vi Acknowledgements This report was prepared by Heba Elgazzar (Health Economist, World Bank) and Firas Raad (Senior Health Policy Specialist, World Bank), with contributions by Chokri Arfa (Associate Professor of Economics, University of Tunis), Awad Mataria (Health Economist, Birzeit University), Nisreen Salti (Associate Professor of Economics, American University of Beirut), Jad Chaaban (Associate Professor of Economics, American University of Beirut), Djavad Salehi-Isfahani (Professor of Economics, Virginia Polytechnic Institute and State University), Sanaz Fesharaki (PhD Candidate, Virginia Polytechnic Institute and State University), and Mehdi Majbouri (University of Southern California). The team benefited from comments and suggestions provided by Adam Wagstaff (Research Manager, Development Economics Research Group, World Bank), Bjorn Ekman (Senior Health Economist, World Bank), Ritva Reinikka (Sector Director, MENA Social and Economic Development Department, World Bank), Rekha Menon (Senior Health Economist, ECA, World Bank), Jose-Lopez Calix (Lead Economist, LAC Social and Economic Development Department, World Bank), David Gwatkin (Results for Development, Washington, DC), John Langenbrunner (Lead Health Economist, EAP, World Bank), Moukim Temourov (Senior Human Development Economist, World Bank), Jeffrey Waite (Human Development Coordinator and Lead Education Specialist, World Bank), Akiko Maeda (Sector Manager, MENA Health, Nutrition and Population, World Bank), and Steen Lau Jorgensen (Sector Director, MENA Human Development Department, World Bank). Findings from the paper were presented at the 7th Annual International Health Economics Congress, Beijing, China, during July 2009. The authors are grateful to the World Bank for publishing this report as an HNP Discussion Paper. vii viii I. Introduction: The Relationship between Out-of-Pocket Health Spending and Equity in MENA In many countries, direct household spending on health care can account for the single largest component of household spending after food expenditures. The impact of out-of- pocket payments for health care can become catastrophic over time and can significantly affect the living standards of individuals. Out-of-pocket spending on health care has become a policy concern for three reasons: First, households may be pushed into poverty or into deeper poverty as a result of paying directly for health services. Second, households facing these health expenses may cut back on other essential household spending such as food and clothing. Third, households may, in fact, choose to forgo necessary health care services rather than face the steep financial consequences - creating a vicious cycle of ill health, disability, and poverty. Previous assessments in Egypt1, Jordan2, Iran3, Lebanon4 and other countries suggest that socioeconomic inequities exist in health status and the use of health care services. A comparison of maternal and child health disparities in MENA is presented in Figure 15. However, to date there has been little evidence in the Middle East and North Africa (MENA) on the level and distribution of household out-of-pocket payments for health care, and to what extent household expenditures on health care affect living standards. Figure 1. Income inequality in maternal and child health indicators in MENA countries Prevalence of Severe Severe Acute stunting underweight Malnutrition Infant Under-five Prevalence of Respiratory Total fertility among among among mortality rate mortality rate Diarrhea Infection rate children children women 0 -0.05 Concentration index (-1, 1) -0.1 -0.15 -0.2 -0.25 -0.3 -0.35 -0.4 -0.45 -0.5 Egypt (2000) Jordan (1997) Morocco (2003/4) Yemen (1997) 5 Source: Demographic and Health Surveys, Gwatkin et al, 2008 . Notes: IMR = infant mortality rate; U5MR = under- five mortality rate. *`Concentration index' denotes degree of inequality across socio-economic status; a positive index indicates a pro-rich distribution, where the value of the indicator increases as income increases; a negative value indicates the reverse. 1 The adverse consequences of ill health and inadequate social protection have prompted many countries to examine the consequences on households of paying for health care and to expand and enhance health insurance coverage. In many countries, subsidized and nominal ,,coverage is extended to the general population through access to the subsidized public sector health care system. The problem, however, is that these public sector systems, at times, do not fully command the confidence of individuals, prompting them to opt out of this public coverage; and self-pay directly for needed health care services. This policy issue has encouraged countries to experiment with ways to improve the quality of their subsidized health services and to develop ways to better target the less well-off individuals in society. In recent years, rising global interest in the area of health equity has spawned numerous international research and policy initiatives. These initiatives have largely focused on measuring and explaining inequalities in health status outcomes (infant mortality or maternal mortality), health service use (antenatal care visit), and public subsidies supporting health service utilization. Policy initiatives have sought to alleviate disparities by addressing modifiable factors through organizational, economic and/or regulatory reforms. Another focal area of health equity research has revolved around the progressivity of health care payments and the catastrophic and impoverishing impact of these payments on individual households. A number of global6 as well as regional evaluations in Europe7, Asia8 and Latin America9 have highlighted the variations in household payments for health care and their implications, shedding light on the international focus afforded to the issue of financial protection in health care. This paper presents new evidence on the patterns and effects of out-of-pocket health payments on households in selected MENA countries and assesses options for alleviating financial burdens associated with utilizing health services in the MENA region. The paper examines the scope of OOP expenditures and the implications on living standards in six MENA countries representing various per capita income levels and approximately 55 percent of the overall population of nearly 300 million in MENA. Methods The methods employed are described in detail in the Annex, including a complete list of data sources used for the analysis. Quantitative analyses were conducted using data from representative household surveys from Yemen, West Bank and Gaza, Egypt, Iran, Tunisia, and Lebanon. The paper is structured as follows. Section 1 assesses the extent of OOP payments as a percentage of total national health spending in MENA countries and as a percentage of monthly household spending. Section 2 extends this analysis by presenting the distribution of payments across socioeconomic status, by health care service and by insurance status. Section 3 examines the incidence and intensity of those health payments that are catastrophic. Section 4 evaluates the impoverishing effect of out-of- pocket payments. The final section concludes with key policy implications aimed at strengthening financial protection for health care in the MENA region. 2 II. The Reliance on Out-of-Pocket Health Financing in MENA In light of the epidemiologic and economic transition occurring in most Middle Eastern and North African nations as elsewhere, policy-makers, practitioners, and citizens face a number of policy questions regarding safety nets for health care, such as: · Are health systems in MENA prepared to protect their populations, particularly the most vulnerable, from potentially catastrophic costs of health care? · Which policy options for health care financing offer the greatest promise for expanding coverage whilst maintaining sustainability? · Which other supply- and demand-side policies are expected to alleviate household financial burdens related to health care? This paper evaluates the first question and sheds light on the broader policy implications highlighted in the second two by examining the experience to date in MENA. Despite the existence of public services designed to be almost free at the point of use in most MENA countries, most citizens pay relatively high costs to access health care services. The 2009 World Development Report10 highlights the negative implications of a lack of sound financial protection policies and a lack of access to basic services such as health care on vulnerable groups and the socioeconomic and geographic disparities that result. By having to pay out-of-pocket for services, households may face the choice of delaying or forgoing care in the face of user fees, which can accumulate and pose as a challenge to living standards. Indeed, evidence from Vietnam11 and other countries shows that pharmaceutical expenditures and fees for routine doctor visits add up quickly to impinge on household disposable income. The nature of health systems varies somewhat across the MENA region in terms of health financing, but most share similarities with regards to the provision, organization and regulation of health services. Generally, the public sector accounts for a substantial proportion of the provision of services in a centralized system managed directly by the Ministry of Health, although the recent two decades have witnessed the growth of the private sector in the areas of pharmaceutical and physician consultations. The exception is Lebanon, where most health care is provided through the private sector. Whilst health care professionals are salaried, civil servants in the public sector, a substantial number of physicians also operate private clinics where remuneration is on a fee-for-service basis financed either through direct user fees or, on a limited scale, reimbursement from social health insurance schemes. In some countries, notably Jordan, Egypt and Tunisia, the state has greater institutional capacities to regulate the quality and availability of health services than in others, notably Yemen, Morocco and Lebanon, yet most of the countries do not currently have effective systems of accountability with respect to service delivery performance. The relatively low level accountability has likely contributed to the problems regarding health financing as well. As revealed in Table 1 below, over half of the amount that MENA countries spend on health care comes from out-of-pocket fees in most MENA countries, followed either by general government revenues or social health insurance schemes, depending on the country. OOP health expenditures account for approximately 49 percent of total health 3 financing in the MENA region as a whole, ranging from 4 percent in Saudi Arabia to 59 percent in Egypt in 2006. On balance the GCC countries spend less out-of-pocket on health services than the remaining MENA countries. Saudi Arabia aside, Oman spends 10 percent and the UAE 22 percent out-of-pocket as a percentage of total health expenditures compared to Jordan, Syria and Tunisia, for example, which all spend more than 40 percent out-of-pocket for health care. It is noteworthy, too, that Yemen, one of the poorest countries in MENA, has the second highest percentage (55 percent of total health spending is out-of-pocket) following Egypt. Table 1 also reveals that although the proportion of national health expenditures that is private is similar between MENA (51%) and other regions, out-of-pocket payments make up a larger share of that private spending in MENA (at 90%) than they do in other regions. Table 1. National health expenditures in MENA countries, 2006 Country Health Health Health Health Out-of-pocket expenditure, expenditure, expenditure, expenditure, health expenditure private (% of public (% of public (% of total (% of (% of private GDP) GDP) total health GDP) expenditure on expenditure) health) Algeria 0.8 3.4 81.1 4.2 94.6 Bahrain 1.1 2.5 68.2 3.6 68.0 Djibouti 1.8 5.0 74.1 6.8 98.6 Egypt, Arab Rep. 3.7 2.6 41.4 6.3 94.9 Iran, Islamic Rep. 3.4 3.4 50.7 6.8 94.8 Iraq 0.8 2.7 78.1 3.5 100.0 Israel 3.5 4.5 56.0 8.0 75.3 Jordan 5.5 4.2 43.3 9.7 75.9 Kuwait 0.5 1.7 78.2 2.2 91.6 Lebanon 4.9 3.9 44.3 8.8 76.1 Libya 0.8 1.6 66.3 2.4 100.0 Malta 1.9 6.5 77.0 8.4 90.4 Morocco 3.9 1.4 26.2 5.3 77.3 Oman 0.4 1.9 82.3 2.3 57.7 Qatar 0.9 3.4 78.2 4.3 88.2 Saudi Arabia 0.8 2.5 77.0 3.3 13.4 Syrian Arab Republic 2.0 1.9 47.8 3.9 100.0 Tunisia 2.8 2.3 44.2 5.1 81.7 United Arab Emirates 0.7 1.8 70.4 2.5 69.4 West Bank and Gaza .. .. .. .. .. Yemen, Rep. 2.4 2.1 46.0 4.5 95.2 East Asia & Pacific 2.5 1.8 42.1 4.3 82.1 Europe & Central 1.8 3.6 66.2 5.5 85.6 Asia Euro area 2.3 7.5 76.9 9.8 60.0 Latin America & 3.5 3.4 50.0 7.0 72.2 Caribbean Middle East & North 2.8 2.8 51.3 5.7 90.5 Africa South Asia 2.6 0.9 25.8 3.5 91.4 Sub-Saharan Africa 3.4 2.4 41.6 5.7 46.8 Croatia 1.1 7.1 86.1 8.2 92.2 Turkey 1.3 3.5 72.5 4.8 84.2 Costa Rica 2.4 5.3 68.4 7.7 86.7 Source: World Development Indicators, World Bank, Accessed 5 June 2009. Data reflect 2006 values. 4 Table 2 provides comparative information on coverage levels for seven low to middle income countries. In all countries, nominal coverage levels are relatively high, whether through subsidized government services or through social security schemes. In Libya, Yemen and Egypt, 100 percent of the population is reportedly eligible to use subsidized government services. In Lebanon, Iran and Tunisia, coverage through social security ranges from 52 to 78 percent of the population. Despite these health insurance coverage schemes, OOP health payments in MENA countries are relatively high, indicating a possible shortage of confidence in the quality of the ,,covered services and/or problems related to service availability and access. Table 2. Coverage rates by health financing scheme in MENA, 2008 Est. % of Est. % of population Est. % of population entitled population enrolled in to subsidized/free enrolled in social private health governmental health Country health insurance insurance Est. % uninsured services or insurance Yemen n/a n/a n/a 100.0 Libya n/a n/a n/a 100.0 Lebanon 52.0 15.0 >48 23 Iran 69.3 n/a 30.7 67.0 Egypt 45.0 <1 55.0 100.0 West Bank-Gaza 48.5 10.2* 22.4 18.9** Tunisia 78.0 n/a 14.8 7.2 Sources: Yemen, Ministry of Health, 2007; Libya, Ministry of Health, 2008; Lebanon, Ministry of Finance/Ministry of 12 Health Annual Report; Lebanon, percent receiving Ministry of Health-funded treatment, Ammar, 2003 ; Iran, Household Expenditure Survey, 2006; Egypt, Ministry of Finance/World Bank, Health Policy Note, Public Expenditure Review 2007; Tunisia, Ministry of Health; West Bank-Gaza, Ministry of Health/World Bank Health Policy Note, 2009. *Includes military, private and Israeli insurance schemes. **Covered by United Nations Relief and Works Agency. Attempting to expand health insurance coverage to larger segments of the population is fraught with technical and political challenges, although experience from Iran exemplifies the potential to create schemes that can effectively reach the most vulnerable groups is society. Over the decade between 1996 and 2006, the proportion of households in rural areas that report being covered by health insurance has increased nearly three- fold, from 28.5 percent in 1996 to 76.2 percent in 2006, as shown in Table 3 below. In urban areas, the proportion of households with health insurance has remained unchanged at 61.1 percent during the same period. The sharpest increase in health insurance coverage in rural areas occurred during the period 2001-2006, when the coverage rate increased by 2.3 times (33.3 percent to 76.3 percent of households), as compared to 1996- 2001 when the coverage rate increased by only 1.2 times (from 28.5 percent to 33.3 percent of households). 5 Table 3. Spatial distribution of health insurance coverage over time in Iran, 1996-2006 Proportion of households enrolled in health insurance (%) Rural Urban 1996 28.46 61.12 2001 33.29 60.91 2006 76.24 61.18 Source: Iran, Household Expenditure Survey, 1996-2006. Despite the existence of health insurance schemes that ought to reduce the costs paid by beneficiaries, individuals in MENA countries appear to face economic barriers in seeking health care. According to national household survey data, a substantial proportion of citizens in MENA countries report forgoing necessary health care services because of the expected financial costs. In Yemen, approximately 37 percent of respondents have reported not seeking health care due to financial barriers, as compared to 20 percent in Lebanon and 12 percent in Egypt, as shown below in Figure 2. Figure 2. Forgone health care use due to financial costs in MENA Egypt Yemen Lebanon 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0% Percent of population who forego care due to financial costs (%) Source: Yemen, Household Budget Survey 2005/6; Lebanon, Multi-Country Survey Study, World Health Organization, 2001; Egypt, Multi-Country Survey Study, World Health Organization, 2001. Yet too often citizens in MENA face the choice of immediate versus delayed costs when it comes to health care. Surveys reveal that many choose to carry the financial burden immediately rather than risk greater welfare costs in the future. The fact that citizens of varying socioeconomic levels are willing to pay for health care services does not necessarily mean they can afford these payments. In Morocco, amongst citizens who paid for health services in 2001, approximately 60 percent reported having had to borrow or sell assets to do so, as compared to 35 percent in Tunisia. The difference in the proportion that report having to borrow or sell assets partly reflects differences in the availability and eligibility of health care coverage from country to country, and between different regions or socioeconomic groups in a given country. In MENA, health care expenses often comprise the single largest component of household expenditure after food. In nine MENA countries surveyed as depicted in 6 Table 4 below, households reported spending between 3 percent and 7 percent of their total consumption expenditure on health care-related services. Data by geographic status show that out-of-pocket payments reveal further differences across countries. In Yemen and Tunisia, households in urban and rural areas appear to pay similar proportions of their expenditure on health care on average. In Iran, Lebanon and Egypt, wide gaps exist; healthcare accounts for a greater proportion of household spending in urban areas than in rural areas. Table 4. Average out-of-pocket spending on health care in MENA countries GDP per capita, OOP as % THE Average OOP as % HH Income (%) Country 2006 (current US $) (%) Total Urban Rural Yemen 882 58 2.7 2.8 2.5 Lebanon 6,060 75 6.6 n/a n/a Iran 3,152 44 5.1 4.9 6.1 Egypt 1,489 62 8.9 9.2 8.5 West Bank-Gaza 1,187 40 3.1 3.3 3.2 Tunisia 3,072 56 4.6 4.4 4.8 Sources: GDP data: Current international dollar; World Economic Outlook, April 2009 Database, International Monetary Fund; West Bank-Gaza based on IMF West Bank-Gaza Staff Report February 25, 2009. Consumption and OOP expenditure: Authors' calculations based on household survey data described in the Annex, with the exception of: Egypt figures based on HIECS 2004/5 for total expenditure and HIS 2006 for OOP; urban and rural figures approximated; West Bank-Gaza, Palestinian Consumption and Expenditure Survey, Palestinian Central Bureau of Statistics, 2006; OOP as % THE based on West Bank-Gaza Health Policy Note, World Bank, 2009. Notes: GDP = Gross domestic product, purchasing power parity exchange rate. OOP = out-of-pocket expenditures on health care. THE = total health expenditures at country level. HH = household. Main messages: Nearly half of all health financing in MENA comes from private household spending, suggesting that access to health care is substantively based on ability to pay. 1. Despite the existence of governmental financing and health insurance, out-of-pocket spending has been rising in most countries. 2. Between 10-35 percent of households in countries surveyed report having to forego necessary health care due to costs or up to 60 percent have had to borrow or sell assets to access care. 3. In MENA, health care expenses often comprise the single largest component of household expenditure after food, suggesting that as health care costs rise, better social protection and availability of affordable, effective health services will need to be ensured. 7 III. The Distribution of OOP Health Spending Similarly, socioeconomic differences in out-of-pocket spending appear larger in some countries than in others. Recent data reveal that in most of the countries surveyed, the amount spent by households tends to increase as income levels increase, with the richest quintiles spending nearly double that of the poorest on health care. These trends are revealed below in Table 5. However, in terms of the equity of these expenditures as indicated in the last column using the Kakwani index39, all of the MENA countries exhibit relatively regressive OOP trends. The burden of out-of-pocket payments falls disproportionately higher on lower-income households. In some countries, health care expenditure abroad may not be fully accounted for in national data. In Yemen and Libya, citizens have reported having had to travel abroad for health care services due to the lack of availability or satisfaction with local services. Treatment abroad, for example, in Yemen accounts for approximately one-third of all out-of-pocket spending, although this is reported amongst the top three income categories. Table 5. Distribution of out-of-pocket health care spending (OOP) across socioeconomic status Country Proportion of total household expenditure spent on health Concentration Kakwani care across income quintile (%) Index Index of Progressivity nd nd Poorest 2 Middle 2 Richest Total Poorest Richest Yemen 1.7 1.8 2.2 2.5 3.7 2.7 0.1970 -0.1830 Lebanon 4.8 6.3 6.5 7.5 8.2 6.6 0.0960 -0.2640 Iran 4.3 4.2 4.1 4.2 5.9 5.1 0.0569 -0.3231 Egypt 5.2 3.1 4.4 3.7 1.0 3.4 -0.1888 -0.5088 West Bank-Gaza 3.0 2.9 3.1 2.9 3.5 3.1 0.0162 -0.6238 Tunisia 4.0 4.3 4.4 4.7 5.1 4.5 0.0346 -0.3754 Source: Country survey data. Note: Egypt calculations based on data from pilot program survey in four governorates listed in Annex, used to examine policies at exempting poorer households from user fees. West Bank-Gaza calculations based on Palestinian Consumption and Expenditure Survey, Palestinian Central Bureau of Statistics, 2006. The distribution of out-of-pocket spending reveals that spending patterns vary across countries. In Tunisia and Yemen, the single largest component of household health care expenditures is medications, followed by physician visits and finally inpatient care as indicated below in Figure 3. Expenses classified as ,,other several individual categories such as medical or pharmaceutical paraphernalia, laboratory tests, and dental care. In Iran, the largest component of health spending is inpatient hospital care, followed by physician visits and medications. These patterns reflect differences in supply and demand-side factors in health financing and organization, including differences in user fees for health care services in the private and public sectors, differences in eligibility criteria for various health benefits under public or private insurance schemes, and differences in the availability and distribution of health care providers and services that meet citizens needs and preferences. 8 Figure 3. OOP by health service in MENA countries 100.0% Proportion of total out-of-pocket health 90.0% 80.0% Inpatient stays 70.0% spending (%) Physician visits 60.0% Medications 50.0% Other 40.0% 30.0% 20.0% 10.0% 0.0% Yemen (2005/6) Iran (2006) Tunisia (2005) Source: Authors' calculations using data from national surveys (see Annex). In terms of the growth of OOP health expenditures over time, the case of Iran suggests that the rate of growth depends on the particular component of spending. Since 1996, overall expenditures in Iran have grown at an average rate of 6.6 percent every five years, as shown in Figure 4. This growth has been largely driven by the increase in expenditures for outpatient health care services, which has grown at a rate of 16.0 percent every five years. The rate of growth for overall expenditures also appears to differ spatially; over the period of 1996-2006, expenditures in rural areas grew at a rate of 7.8 percent as compared to 7.3 percent in urban areas. Differences are more apparent in patterns of spending on inpatient services. In rural areas, expenditure on these services grew at a rate of 2.4 percent between 1996 and 2001, and increased sharply to grow at a rate of 9.1 percent between 2001 and 2006. The pattern is reversed in urban areas, where the rate of growth decreased from 12 percent during 1996-2001 to -2.0 percent during 2001-2006. 9 Figure 4. OOP by health service in Iran over time, 1996-2006 Iran 6.00% OOP as % total household 5.00% expenditure (%) 4.00% Inpatient Outpatient 3.00% Medications 2.00% Other 1.00% 0.00% 1996 2001 2006 Year Source: Authors' calculations using data from national surveys (see Annex). These patterns reflect spatial differences in various determinants of health care expenditure, including differences in the type of coverage provided by different social protection schemes. The level of protection offered by insurance or social protection schemes for different health care services and the amount of co-payments levied influence the way in which people utilize health services and the financial costs they face. Indeed, data from Lebanon show that household expenditures on health can vary by insurance coverage as indicated below in Figure 5. While the uninsured spend an average of 7 percent of their household disposable income on health care, households covered by private health insurances schemes spend 5.2 percent on health care. Households covered by public health insurance schemes spend 6 percent out-of- pocket on health care, indicating that there are differences in utilization patterns between households enrolled in different insurance plans. These figures also suggest that the level of financial protection afforded by private health insurance schemes may be more comprehensive than by public health insurance schemes. Figure 5. OOP by insurance coverage status in Lebanon, 2004/5 8.00% 7.00% OOP as % total household 6.00% 5.00% spending (%) 4.00% 3.00% 2.00% 1.00% 0.00% Public Private All insurance Uninsured insurance insurance schemes schemes Health insurance scheme Source: Authors' calculations using data from national surveys (see Annex). 10 Main messages: The level of out-of-pocket spending on health is relatively high in MENA. 1. OOP health expenditures account for 3 to7 percent of household spending in MENA countries, with wealthier households spending more as a proportion of total expenditure than lower-income households. 2. Urban households tend to spend more as a proportion of total expenditures than rural households. 3. Patterns suggest that those with a higher ability to pay for health services may receive better or more frequent health services than poorer counterparts. 11 IV. Identifying Catastrophic Health Care Expenditures The burden on welfare that out-of-pocket payments pose throughout countries in MENA is related to the cycle of ill health and economic hardship. Households that incur excessively high out-of-pocket payments are most vulnerable to being plunged further into this cycle. These households are typically defined as those that pay greater than an internationally-recognized threshold as a proportion of total household expenditure, which may vary between 5 percent and 25 percent of total expenditure, or 15 percent and 40 percent of non-food expenditure13. Figure 6 shows the proportion of households that incur such ,,catastrophic health care expenditures, ranging from 7 to 13 percent of all households in MENA countries. The incidence of catastrophic expenditures is lowest in countries such as Yemen, West Bank-Gaza, Egypt and Iran and highest in Lebanon. Generally, in countries where more than 5 percent of all households incur such catastrophic expenses, financial protection against the costs of health care may be considered insufficient or inadequate for certain services. For example, given that medications account for large proportion of household health spending in Tunisia, insufficient coverage against the costs of medications contributes to high rates of catastrophic payments. Figure 6. Incidence of catastrophic health expenditures in MENA countries 14.00% Proportion of all households incurring 12.00% catastrophic expenses (%) 10.00% 8.00% 6.00% 4.00% 2.00% 0.00% Yemen West Bank and Tunisia Lebanon Iran (2006) Egypt (2007) (2005/6) Gaza (2005) (2005/6) (2004/5) Source: Authors' calculations using data from national surveys (see Annex). Note: Catastrophic spending defined as having incurred at least 10% of total household expenditures. Catastrophic payments also appear to vary depending on geographic location in MENA countries. Evidence from Yemen demonstrates that a greater proportion of urban households as compared to rural households incur catastrophic payments. The reverse is true in Iran, where nearly twice as many rural households as urban households incurred catastrophic payments in 2006, 8.4 percent vs. 4.6 percent, respectively, as shown in Figure 7 below. 12 Figure 7. Spatial distribution of catastrophic expenditures in MENA countries 14.00% Proportion of all households incurring catastrophic expenses (%) 12.00% Total Urban 10.00% Rural 8.00% 6.00% 4.00% 2.00% 0.00% Yemen (2005/6) Iran (2006) Tunisia (2005) Source: Authors' calculations using data from national surveys based on methods developed by 14 Wagstaff et al (see Annex). Note: Catastrophic spending defined as at least 10% of all spending. The extent to which health financing schemes protect households against catastrophic health care payments depends on the nature of the coverage and eligibility criteria, amongst other factors. Examining the case of Iran during 1996-2006, uninsured households have tended to face approximately twice the exposure to catastrophic health care payments; in 2006, approximately 7 percent of insured households reported catastrophic payments as compared to 14 percent of uninsured households as shown in Figure 8 below. Although the expansion of health insurance coverage to rural areas during the late 1990s has been associated with improved access to primary health care services, insured rural households have also reported a substantially higher rate of catastrophic payments over time, reflecting the combination of improved availability of health care services associated with user fees for households that may otherwise forgo necessary health care. Similar patterns exist in Tunisia (Table 6), where the insured tend to report higher rates of catastrophic expenditures than the uninsured, likely attributed to a higher propensity to use more costly services associated with co-payments or those found in the private sector. Attempting to use variations in health insurance coverage to explain variations in the incidence of catastrophic health spending should be done with some degree of caution given the endogenous nature of ,,insurance coverage as an explanatory variable15 16. 13 Figure 8. Spatial distribution of catastrophic expenditures by insurance status in Iran over time, 1996-2006 14 12 Incidence of catastrophic health expenditures 10 UI - Rural 8 UI- Urban (% of households) UI - Country 6 INS - Rural INS -Urban 4 INS - Country 2 0 1996 2001 2006 Year Source: Authors' calculations using data from national surveys (see Annex). Note: Catastrophic spending in Iran defined as at least 25% of total household expenditure. UI: Uninsured population; INS: Insured population. Table 6. Catastrophic expenditures by insurance status in Tunisia, 2005 Type of health insurance coverage Incidence of catastrophic expenditures (%) Social health insurance (SHI) 6.16 Complementary health insurance in addition SHI 0.85 Government-subsidized coverage 3.44 Uninsured 1.62 Total 12.07 Source: Authors' calculations using data from national surveys (see Annex). Note: Catastrophic spending defined as at least 10% of total household expenditure. Depending on how well systems work for exempting the poorest households from having to pay fees at the point-of-use, the effect on financial burdens can vary. During 2006 and 2007, four Egyptian governorates participated in a user-fee abolition system, whereby poorer households would be identified and offered a health card to access primary health care services for free. While this scheme should have resulted in greater access to health care and lower out-of-pocket payments by households, the scheme was unable to demonstrate such an effect. The lack of such an impact has been attributed to weaknesses in the system to adequately ensure that the poorest households received the benefits afforded by the scheme. Table 7 shows that based on catastrophic thresholds of 5 percent or 10 percent of household expenditure, the incidence of catastrophic payments slightly increased following the introduction of the user-fee abolition scheme. Based on a 10 percent 14 threshold, 7 percent of households faced catastrophic payments before the scheme was introduced, as compared to 7.13 percent after its introduction. The higher thresholds reveal that the scheme resulted in a marginal reduction in the incidence of catastrophic payments; using thresholds of 15 percent to 25 percent, the scheme reduced the incidence by 0.3 percent to 0.5 percent. The introduction of the scheme also appears to have marginally reduced the intensity, or overshoot, of catastrophic expenses by approximately 0.5 percent on average. Overall, the introduction of this scheme in Egypt did not substantially reduce the impact of out-of-pocket payments on household living standards as intended. Table 7. Incidence of catastrophic expenditures before and after User-Fee Exemption Scheme in Egypt, 2007 Program Phase Thresholds: Health Care Payments as % of Household Expenditure Baseline 5% 10% 15% 20% 25% Head count (H) (% of households) 12.29 7.00 5.57 4.43 3.86 Overshoot (O) (% of threshold) 2.13 1.67 1.36 1.12 0.90 Mean positive overshoot (MPO) (%) 17.32 23.88 24.37 25.20 23.46 Follow-up Head count (H) (% of households) 12.70 7.13 5.28 4.14 3.14 Overshoot (O) (% of threshold) 1.53 1.07 0.76 0.53 0.36 Mean positive overshoot (MPO) (%) 12.08 15.00 14.37 12.80 11.34 Source: Authors' calculations using data from national surveys (see Annex). Notes: Baseline data were collected before the introduction of the user-fee exemption scheme during 2007. Follow-up data were collected in September 2007, three months after the introduction of the scheme. Conflict-affected environments can also have a big impact on catastrophic spending. These types of environment which exist in the MENA region exacerbate two particular issues. The first issue concerns access to regular care that may mitigate against the propensity to incur sudden, often higher-than-average expenditures. The second issue concerns the weakening of existing financial protection schemes as a result of broader economic and political instability. Using the case of the West Bank and Gaza, longitudinal evidence shows that the incidence of catastrophic expenses had been decreasing over time until 2006, when these expenses increased substantially through 2007, substantially exceeding the incidence of all previous years. These trends are shown below in Figure 9. 15 Figure 9. Incidence of Catastrophic Spending in West Bank-Gaza, 1998-2007 14.00% 12.00% Head Counts Using Total Expenditure 10.00% Percentage of Households Head Count Using Total 8.00% Expenditure Net of Non- Discretionary Expenditure Overshoot Using Total 6.00% Expenditure 4.00% Overshoot Using Total Expenditure Net of Non- Discretionary Expenditure 2.00% 0.00% 1998 2004 2005 2006 2007 Years Source: Authors' calculations using data from national surveys (see Annex). Main messages: Catastrophic health expenditures persist in MENA despite the existence of governmental insurance schemes. 1. The incidence of catastrophic expenditures in MENA is relatively high compared to similar low- and middle-income countries, ranging from 6-14% of all households, despite the existence of directly provided services and health insurance coverage schemes. 2. Catastrophic expenditures tend to be more likely in urban in rural areas in Yemen, but the reverse is the case in Iran. 3. Catastrophic expenditures appear to be more prevalent in countries where there is a greater reliance on private health care services, such as in Lebanon, and where out-of-pocket payments account for a greater share of national health expenditures. 16 V. The Impoverishing Effect of OOP Spending To assess the burden of out-of-pocket payments and catastrophic expenditures, the extent to which health care payments impinge on living standards is evaluated. Most poverty measures do not typically take into account health care-related expenditures or do not explicitly show the contribution of health care payments to poverty levels 17. The poverty headcount represents the percent of households that fall below the poverty line, defined in relative or absolute terms. Where data permits, the poverty headcount and depth of poverty as a percentage of the poverty line, or the normalized gap, is estimated before and after accounting for out-of-pocket payments for health care. The poverty lines used have been chosen on the basis of the local context, representing either a relative or absolute poverty line, although other approaches by which to measure poverty exist as described elsewhere18. The relative poverty line is defined in relation to a nations mean income per capita and was used in the case of Egypt. Absolute poverty lines are set according to international poverty lines estimated by the World Bank and were used in the cases of Lebanon, Iran, West Bank-Gaza, Yemen and Tunisia. Amongst the broad effects of out-of-pocket health care payments, the effect on household disposable income is a particularly important consideration. Health care payments in particular may pose as a source of strain on economic resources due to the uncertainty of their timing, duration and magnitude. Households that are entitled to social protection schemes ought to be protected from either falling below the poverty line or being pushed further if already considered poor. The definition of poverty lines may vary across countries and the magnitude of the impact of health care payments will depend on which poverty lines are used. Nonetheless, Table 8 shows that health care payments in MENA tend to impact poverty levels differently across countries. Health care payments tend to increase the number of households that fall below the poverty line in West Bank and Gaza, Lebanon, Iran, Egypt, and Tunisia. The effect tends to be considerable in most of the countries, putting between 8 to 18 percent more households into poverty as a result of these households having to pay for health care themselves. The West Bank and Gaza show a particularly high increase of nearly 80 percent more households falling into poverty. Table 8. Effect of out-of-pocket health care payments (OOP) on poverty rates in MENA countries Country Pre-payment Post-payment Absolute Relative difference poverty poverty difference (percent (percentage headcount (% of headcount (% of difference) difference) households) households) Yemen (2005/6) 20.30 21.90 1.60 7.88 Lebanon (2004/5) 27.50 31.60 4.10 14.91 West Bank and Gaza (2005) 13.70 25.05 11.35 82.85 Iran (2006) 13.64 15.44 1.80 13.20 Egypt (2007) 31.76 36.05 4.29 13.51 Tunisia (2005) 3.69 4.35 0.66 17.77 Source: Authors' calculations using data from national surveys (see Annex). Notes: Data for Egypt shown for baseline levels before introduction of user-fee exemption scheme. 17 The number of households pushed into poverty as a result of having to pay for health care is one dimension of the impact. Another dimension is the depth of the poverty effect, that is, the amount of expenditure by which households fall below the poverty line, adjusted to reflect the percentage of the poverty line or a normalized poverty gap. In the MENA countries surveyed, the intensity of the poverty effect varies between from an absolute increase of less than 1 percent in Tunisia and Yemen to one of nearly 9 percent in the West Bank and Gaza. These data are shown below in Table 9. Table 9. Effect of OOP on depth of poverty in MENA countries Country Pre-payment Post-payment Absolute Relative difference poverty normalized poverty difference (percentage gap normalized gap (% (percent difference) (% of HHEXP) of HHEXP) difference) Yemen (2005/6) 5.85% 6.32% 0.47% 8.03% Lebanon (2004/5) 7.60% 9.00% 1.40% 18.42% West Bank and Gaza 2.20% 10.70% 8.50% 386.36% (2005) Egypt (2007) 12.05% 16.47% 4.42% 36.68% Tunisia (2005) 0.87% 1.01% 0.14% 15.72% Source: Authors' calculations using data from national surveys (see Annex). Notes: Data for Egypt shown for baseline levels before introduction of user-fee exemption scheme. Notes: Data for Iran not available. HHEXP = total household expenditure. The effects of health care payments on poverty in MENA indicate that the greater concern, with the exception of West Bank-Gaza, is not necessarily the depth of poverty, but the total number of households whose poverty status changes before and after accounting for health care payments, as the effect on the poverty headcount is more marked than that on the poverty gap. Therefore, it is the breadth, and to a less extent, the depth, of poverty that is at stake. The more and more people face economic hardship as a result of health care costs, the greater the implications for society-at-large. Consistent with the patterns of poverty within countries, the effect of health care payments varies geographically as well in indicated below in Table 10. Using the case of Yemen, overall poverty rates are generally two-three times as high in rural areas than in urban areas, as tends to be the case throughout MENA countries. At the same time, Yemeni urban households face the effect of health care payments more intensely than their rural counterparts; the relative differences in headcount and the gap are higher amongst urban than rural households. While this may be the case in Yemen given the greater of availability of fee-based health services in urban areas, other countries may show a different pattern depending on the nature and affordability of health services in urban and rural areas. Therefore, spatial differences in the effect of health care payments should be taken into account when designing appropriate health financing schemes within a given country. 18 Table 10. Spatial distribution of OOP-related impoverishment in Yemen Poverty Headcount (%) Poverty Gap (%) Country Pre-payment Post-payment Relative Pre- Post- Relative poverty poverty difference payment payment difference Yemen (2005/6) 20.30 21.90 7.88 5.85 6.32 8.03 Urban 14.90 16.40 10.07 3.99 4.37 9.52 Rural 31.10 33.25 6.91 9.48 10.22 7.81 Source: Authors' calculations using data from national surveys (see Annex). Furthermore, the effect of health care payments on poverty has changed in some countries over time, a result of changes in the demand, supply, and financial protection systems for health care. In the case of Iran, as shown below in Figure 10, the effect of out-of-pocket payments has somewhat widened over time in rural areas, increasing the poverty headcount by 1.6 percent in 1996 to 1.8 percent in 2006. The reverse is found in urban areas, where the net effect has decreased from 0.65 percent in 1996 to 0.47 percent in 2006. Given the observation that rural households have tended to report higher rates of catastrophic expenditures over time, these expenses have not been compensated by overall economic gains. Indeed, international evidence shows that the rate of increase in medical costs is typically higher than overall economic growth. Although poverty rates may be decreasing in MENA over time, the increasing burden of health care costs has not necessarily followed suit; future increases in the demand for health care therefore poses as a growing concern. Figure 10. Spatial distribution of OOP-related impoverishment over time in Iran, 1996-2006 Iran Pre-payment Rural 25 Pre-payment Urban Poverty headcount (% of households) 20 Post-payment Rural Post-payment Urban 15 10 5 0 1996 2001 2006 Year Source: Authors' calculations using data from national surveys (see Annex). 19 The example of the user-fee abolition program in Egypt also suggests that the introduction of financial coverage must be matched by comprehensive systems to ensure that beneficiaries are, indeed, protected from the costs of care sufficiently and adequately. Following the introduction of the scheme, health care payments appear to have increased the poverty headcount to a greater degree than the before the introduction of the scheme. Similarly, the increase in the poverty gap due to health care payments was greater following the introduction of the scheme as shown below in Table 11. Table 11. OOP-related impoverishment before and after user-fee exemption policy in Egypt, 2006/7 Pre-payment Post-payment Absolute difference Relative difference (%) Indicator (%) (%) (%) Poverty head count - Baseline 31.7 36 4.3 14 Poverty head count - Follow-up 38.4 44.5 6 16 Poverty gap ­ Baseline 12 16.5 4.4 36.7 Poverty gap - Follow-up 14.9 24.2 9.3 62.8 Source: Authors' calculations using data from national surveys (see Annex). In the case of West Bank-Gaza, the impact of health care payments on poverty levels has been mixed from 1999-2006 as shown below in Figure 11. The difference in poverty head counts using Palestinian-specific poverty levels increases slightly during the time period while the same difference increases using the World Bank defined poverty level. As for trends related to ,,deep poverty, again, there in an increase in the difference in poverty head counts using the Palestinian-specific poverty level while there is a decrease using the World Bank defined poverty line for deep poverty. Figure 11. Health Expenditure-Related Impoverishment Over Time in West Bank-Gaza, 1998-2007 16.00% 14.00% 12.00% 10.00% 8.00% 6.00% Difference in Poverty Head Counts using the WB Poverty Level Difference in Poverty Head Counts using the WB Deep Poverty Level 4.00% Difference in Poverty Head Counts using Palestinian-Specific Poverty Level Difference in Poverty Head Counts using Palestinian-Specific Deep Poverty Level 2.00% 0.00% 1998 2004 2005 2006 Note: Vertical axis represents poverty headcount over time as percent of households. Source: Authors' calculations using data from national surveys (see Annex). 20 Main messages: Although poverty rates in MENA countries are relatively lower than several comparable countries in Asia and Latin America, the extent that poverty is attributable to healthcare payments is relatively high. 1. Out-of-pocket health expenditures tend to increase the poverty headcount in MENA countries from between 5 percent and 14 percent, with the exception of the West Bank and Gaza where the difference was 80 percent. 2. Policies to mitigate the impoverishing effect of health care payments by exempting vulnerable populations from paying user fees have not been as effective as anticipated, based on the case of Egypt. 3. National poverty reduction strategies should more explicitly take into account the perverse effects of household financial risk, particularly for groups with a higher-than-average need for health care such as infants, mothers, and those with non-communicable chronic conditions. 21 VI. Policy Implications: Bridging the Gaps Who Pays? Main findings Who pays? Households in MENA continue to foot the health care bill, despite efforts to expand financial protection in most countries, particularly those households in the lowest two wealth quintiles. Their experiences demonstrate that as a share of monthly household expenditures, OOP payments on health in the six countries account for 6 percent of total annual household expenditure on average, with Egyptian and Palestinian households spending the most on health care. The bulk of OOP payments in these countries go towards purchasing drugs, private outpatient services, and laboratory and diagnostic services. In many cases, lower-income households pay a larger share of their consumption expenditure on health care than higher-income households, but the reverse was also observed in some countries. Between 7 to 13 percent of households from five countries incur catastrophic payments, defined as 10 percent or greater of annual household expenditure. As to the impoverishing impact of health payments, the number of households who are considered below the poverty line when accounting for OOP health payments is 23 percent higher on average than when not accounting for these expenses. Households that report having health insurance coverage tend to incur lower OOP payments than households that do not have health insurance coverage, but this not necessarily the case. Geographic setting also appears to play a role, with households in rural locations often paying a greater share of expenditures on health care than urban counterparts. Patterns of out-of-pocket payments may be influenced by nature of health financing arrangements and the access and quality of health care received at public and private facilities. Given that data is generally lacking in the MENA region on utilization patterns as well as more information pertaining to spending patterns over time, more systematic efforts should be taken to both measure and address the consequences associated with out-of-pocket payments for health care. Overview of policy responses These patterns of out-of-pocket spending on health care in MENA suggest that policy responses ought to focus on both demand- and supply-side issues related to why households fall into hardship when faced with illness. Their experiences imply that financial protection may be ill-devised at the point-of-use, but that other demand- and supply-side factors may also impact a households ability to address their health care needs, such as the quality of care available at low cost and the distribution of necessary and effective services. As shown in Table 12, a number of MENA countries have managed to reduce financial barriers with seeking care to some extent as a result of three main types of policy interventions: (a) on the demand side, effective risk-pooling for households; (a) on the demand- and supply-side, appropriate incentives to use and provide cost-effective health services, such as primary care; and (c) at a central, system- level, the use of comprehensive information systems to monitor and ensure the most vulnerable receive adequate care through beneficiary registries and systematic assessment of service delivery performance indicators. These are discussed in turn. 22 Demand-side policy responses On the demand side, the persistent reliance on households to foot a significant proportion of health care costs appears to have negatively impacted living standards in MENA, particularly in countries in which risk-pooling measures have been inadequate and where basic health care benefits, such as preventative care, have not been guaranteed by the state. This pattern appears to be the case regardless of the main revenue-generating system, whether through social health insurance, general revenues and taxes, or a mix of financing agents. Malnutrition and increasing rates of chronic health conditions amongst poorer groups in peripheral areas and urban slums pose as increasing barriers to future social and economic attainment in lagging areas. The nature of the disparity will often dictate the type of intervention, whether integrative, targeted, or a combination of both. Strategies that aim to improve financial protections schemes and the quality of health care provisioning at a population level may alleviate spatial and socioeconomic inequity in health outcomes to a greater extent than selective, targeted schemes such as conditional cash transfers19. Although there is evidence that conditional cash transfer programs in Honduras, Mexico, Nicaragua and Brazil have been associated with increased utilization of preventative, maternal and child health services, these development and success of these programs often depend on the wider economic, institutional and social contexts in which they operate20 21 22 23. To date, there is a lack of information on the effects of such a scheme in MENA, but may prove to be an option in the future. In terms of risk-pooling and financial protection schemes, experience in MENA has shown that broad-based strategies that reduce the financial and administrative burden on the part of beneficiaries often yield the greatest gains in terms of improved health care access and outcomes. This effect appears to have been biggest in the case of Iran and Tunisia, where the rates of insurance coverage have increased substantially to cover over 85 percent of their populations and where the utilization of preventative health services has risen. The 2009 World Development Report highlighted that although economic growth is seldom distributed evenly within or between nations, the most successful nations also tend to instituted policies that evenly distribute basic living standards 24. The level of financial protection afforded to facilitate access to health services is an integral component of such policies. These policies range from geographic targeting to broad, integrative programs aimed at reducing the level of social exclusion in health that may exist. In MENA, particularly vulnerable populations are often lower-income, marginalized communities such as the urban slums in Cairo, refugee locales in Lebanon, Jordan, Syria and the West Bank and Gaza, mountainous villages in Yemen and Morocco, and desert communities throughout the Sahara and the Arabian Peninsula. Yet Iran has pursued a number of measures during the past two decades that have managed to narrow such socioeconomic and geographic inequities. Policies such as the expansion of government- subsidized health insurance, the strengthening of primary health care through local 23 staffing, and more regular nation-wide monitoring have contributed to expanding the use of primary health care services to reach over 90 percent of the population and health insurance coverage to reach 80 percent25 26. The Tunisian experience shows that, whilst the establishment of health insurance schemes for the informal sector can help to increase enrolment rates, these policies should be linked to adequate service provision for those facilities contracted with public insurance schemes. Social health insurance schemes, operated through the National Health Insurance Fund (Caisse Nationale d'Assurance Maladie, or CNAM), cover approximately 66 percent of the population as of 2008. Since 1958, state-subsidized health insurance has operated in parallel to pre-existing social health insurance schemes for the formal sector. These schemes, previously known as Régimes d'Assistance Médicale Gratuite (AMG 1 and AMG 2, respectively)27, have been consolidated as of 2009 and have gradually succeeded in increasing the proportion of enrolled the majority of informal workers and low-income individuals, currently covering approximately 33 percent of the Tunisian population, although 10 percent of the population continue to be uninsured28. These gains have been matched by exponential increases in out-of-pocket expenditures, partially due to the shift away from public providers to private providers, and partially due to cost-sharing tariffs that exist even under state-subsidized, "free" schemes. By the same token, targeting mechanisms may supplement broader programs, providing they are well-designed and cost-effective. A recent assessment of the Moroccan experience has demonstrated that a combination of highly targeted geographic and means-tested poverty alleviation schemes may be the best way forward29. Likewise, programmatic schemes such as the National Human Development Initiative, or L'initiative nationale pour le developpement humaine (INDH), have specific components each targeting certain geographic zones, including poorer rural communes and urban slums30. In 2008, the Moroccan government also initiated a health card program called Le Regime d'Assistance Medicale (RAMED), in order to expand coverage of social health insurance, improve financial protection and improve utilization of primary health services for the poorest groups in a pilot region. Whether the program succeeds or whether other, supply-side measures and incentives are needed to address inadequate public primary care provision in remote, poorer areas remains to be seen. More specifically, the success of targeting policies that seek to reduce the burden of health care costs for the most vulnerable will depend on how likely these policies can be implemented and operated at a relatively low economic and opportunity cost. In Egypt, the recent user fee exemption policy demonstrated modest decreases in the number of individuals facing catastrophic health care expenditures; but overall, the impoverishing effect of health care payments increased after the introducing of the policy, both in terms of the absolute number of head counts and the poverty gap. Contrary to the intended aim of the program, more than half of the interviewed households identified as poor continued to pay out-of-pocket fees at least once after the implementation of the policy. Moreover, more than half of those eligible to benefit from the scheme indicated they were not aware that such a policy existed, indicating the need to ensure awareness and understanding on 24 the part of citizens regarding benefits and entitlements. Although such initiatives are intended to improve welfare, they need to be part of a broader package of reforms that include addressing operating budgets and provider incentives in order to offset lost revenues attributed with exempting citizens from footing the bill. Better implementation of social protection schemes tends to be a result of commitment and information capacity to manage these systems towards their intended aims. In the West Bank and Gaza, the unification of information systems from the Governmental Health Insurance bureaus resulted in a more accessible and transparent beneficiary registry, improving the identification of beneficiaries and benefits alike. Together with quality improvement standards and guideline development, these improved information technology systems were associated with an increased procurement, delivery and availability of pharmaceuticals in lagging areas; an increase of 30 percent in the utilization rates of rehabilitated facilities; and a nearly seven-fold improvement in diabetes management as indicated by the percent of diabetics reporting poor metabolic indicators over nearly five years31. Supply-side policy responses Parallel to effective demand-side measures such as effective risk-pooling schemes, strengthening supply-side capacity and incentives are likely to alleviate inequity in financing. Indeed, the experience of China in addressing access to health services and financial payments suggests that supply-side reforms such as provider incentives and remuneration systems tend to have a greater effect on reducing the incidence of catastrophic spending, than did demand-side reforms, such as expanding health insurance alone32. Notably, where countries have prioritized investing in adequate provision and quality of basic health services such as primary and preventative care, have health outcomes increased and become more equitable, such as Iran. A substantial proportion of out-of-pocket payments in many MENA countries are associated with the use of outpatient and pharmaceutical services. In the case of costly services, the extent and nature of health care benefits included in the ,,basic benefits package covered through financing schemes will dictate the extent to which ability-to-pay plays a role. In the case of preventative and primary care, both demand- and supply-side incentives to utilize or promote such services are traditionally perceived to be low without direct governmental intervention. Disadvantaged communities that lack such services face a vicious cycle of insufficient preventative care, elevated risk of ill health, and impaired welfare due to either forgoing necessary treatment or facing catastrophic health care costs. As such, evidence from MENA has demonstrated that improved primary care provision has helped to ameliorate disparities in health status. Beginning in the 1970s, Iran has prioritized public primary health care at a national level, ensuring that all communities are strengthened through training of local health workers (behvarz) to provide family planning and reproductive health care33. As a result, service coverage in rural areas has increased from nearly 40 to 90 percent over nearly 15 years (1985-2001). In Egypt, improvement in the basic benefits package covering essential maternal and child health services has increased vaccination rates, maternal nutrition and child health indicators in 25 poorer, rural governorates34. The Social Priorities Program Basic Health Project in Morocco, which targeted facility, equipment and medication supply rehabilitation and the establishment of lodging facilities for health personnel in the 14 most disadvantaged and least urbanized provinces, was associated with nearly 100 percent higher rate of medically-assisted deliveries over approximately five years (18.5 to 47 percent)35. In contrast to Iran, the distribution of goods and services such as medications is grossly inequitable in Yemen, in part due to the lack of strategic planning and availability of local health care personnel. Several social safety net programs operate in Yemen, designed to provide in-kind and cash transfers to the poorest households; but similar to many social programs, these schemes do not operate nation-wide, with tribal politics often influencing the extent of central governmental involvement in social assistance. The three main public programs include the Social Welfare Fund, the Social Fund for Development, and the Public Works Program. There has been some evidence that some of these programs reach a greater share of the poorest income groups than do other programs, yet the effect of these transfers on health and health care patterns have not been well documented36. In similar low-income countries where labor markets are fragile and household disposable income is low, the states role becomes paramount in ensuring the availability of functional primary health care facilities and goods. Overall, social protection mechanisms should nonetheless achieve three basic functions to alleviate inequity in financing and access. These include: (a) accurate information on beneficiary welfare through national registry and information systems; (b) effective incentive systems for health care personnel to provide and citizens to utilize cost- effective health services, such as preventative and primary care; and (c) well-designed risk-pooling schemes to ensure coverage for the most vulnerable populations. 26 Table 12. Summary of policy initiatives taken to alleviate barriers to health care in MENA, 1996-2006 Policy Issues Policy Country Intervention Outcomes* approach Example a) Lack of Demand- Iran25 Expansion of free/subsidized Increased coverage rates effective side health insurance in both urban and rural financial Tunisia27 Increased coverage rate protection, nationally particularly 33 for vulnerable Supply- Iran Supply-side: Comprehensive Increased utilization groups side primary health care services rates of reproductive included in basic benefits health services b) High out-of- package covered by insurance; pocket increased training and staffing expenditures of local health workers (behvarz) Morocco35 Supply-side: Free lodging in Increased rates of c) High rates of rural areas for providers; medically-assisted improved availability of deliveries in rural and forgone care subsidized supplies and underserved regions due to pharmaceuticals financial and Egypt34 Supply-side: Increased Increased immunization geographic coverage of maternal and and prenatal care barriers child services within basic utilization rates in benefits package poorer, rural governorates System- West Bank System-level: Increased Increased availability of level and Gaza31 enrolment and availability of pharmaceuticals; services to beneficiaries increased utilization through linkage between rates of health services governmental health insurance scheme databases Iran25 System-level: Better Increased utilization identification of health needs rates of PHC amongst through monitoring and rural regions evaluation of vital health statistics in all provinces Sources: Governmental data and World Bank staff reports (see References). *Note: Outcomes may be attributed to policy interventions described and exogenous factors. 27 28 VII. Conclusions The experience of MENA to date highlights that out-of-pocket spending for health care has increased over time and the consequences can entail economic hardship and reduced access to health services. A concerted policy approach is needed to respond to these social concerns, one that balances synergistic demand- and supply-side policies. These policy responses will take on different manifestations depending on the broader institutional and infrastructure context. Further examination is also needed to evaluate the broader effects of health financing strategies on use of different health services and long term implications for health and economic growth. In low-income countries as well as sub-national provinces that face resource constraints at a facility and household level, shifting investment to ensuring basic, primary care benefits are sufficiently resourced remains a clear priority. In middle-income countries in which a mix of social health insurance and governmental subsidies operate such as Morocco, Tunisia and Egypt, ensuring that targeting mechanisms are well functioning and sufficient outreach and awareness programs exist for eligible beneficiaries are recommended. In middle to upper middle-income countries such as in the Eastern Mediterranean and Iran in which social health insurance schemes widely operate, refined ,,micro-tuning in existing systems ought to address institutional-level issues such as weaknesses in reaching the most vulnerable populations; and administrative complexity in realizing benefits for beneficiaries and operating agencies alike. Leveraging the synergy between demand- and supply-side factors associated with effective social protection can help reduce the economic burden associated with seeking health care in the Middle East and North Africa. 29 30 Annex: Analytic Methods Data Sources Data were available for six MENA countries from nationally representative surveys conducted in each country. Whilst these account for a fraction of the fifteen non-Gulf Cooperation Countries in MENA, they represent the spectrum of economic levels and health financing systems. The economic levels range from low income, as represented by Yemen, to upper middle income, as represented by Lebanon. Egypt and West Bank-Gaza represent lower-middle income countries, and Lebanon, Iran and Tunisia represent middle-income countries. Likewise, governmental, general revenue-funded health systems are found in Iran, Egypt, Yemen and West Bank-Gaza; social health insurance schemes are found largely in Tunisia as well as in Lebanon, Iran, Egypt and West Bank- Gaza. Lastly, private health insurance schemes are most prominent in Lebanon relative to the other six nations. An overview of the survey data and poverty lines used for the analyses are presented in Table 13. Table 13. Data sources Country Year Survey Sample size Survey Type Recall Poverty Line (No. of period households) Yemen 2005/6 Household Budget 13,136 National 1 month World Bank Survey poverty line of $1.08/day Lebanon 2004/5 Multi-Purpose Survey 13,000 National 1 month National poverty of Households line of $4.4/day West Bank 1998, Palestine 1,440 National 1 month World Bank and Gaza 2004, Consumption and poverty line of 2005, Expenditure Survey $2.15/day 2006, (PECS) 2007 Iran 1996, Household 22,000 ­ National 1 month World Bank 2001, Expenditure and 31,000 poverty line of 2006 Income Survey $2.15/day Egypt 2006/7 User-Fee Exemption 2,954 Four 1 month National poverty Program, Baseline and governorates line of LE Follow-up Survey; (Alexandria, 118.59/day included for purposes Manoufieh, of impact evaluation Qana, Suez); example semi-nationally representative Tunisia 2005 National Health 6,538 National 1 month World Bank Survey (Lenquête poverty line of nationale de santé) $2.15/day Source: Yemen: Central Statistical Organization; Lebanon: Ministry of Social Affairs and Central Administration for Statistics; West Bank and Gaza: Palestinian Central Bureau of Statistics; Iran: Statistical Center of Iran; Egypt: Ministry of Health; Tunisia: Ministry of Public Health. 31 Analyzing the distribution of out-of-pocket payments Out-of-pocket payments are defined as payments made directly by the household at the point-of-use, excluding health insurance premiums and payments that are reimbursed by third parties. The amount spent on medications, outpatient health services and inpatient hospital stays are extracted and compared across the cases. Outpatient health services are defined as visits to public or private clinics where patients do not stay overnight. Inpatient hospital stays are defined as visits to hospitals where patients stay at least one night. The methods that are employed to quantify the degree of equity in health care include descriptive and regression techniques using national household survey data. Descriptive methods include comparison of means and the ,,concentration index technique37. The concentration index is a measure of how equally a health variable is distributed across a population ranked by income level38 39. As a single numeric, the concentration index allows degrees of equity to be easily captured and compared, in order to determine the extent and nature of policy reform that is necessary. The concentration index represents twice the difference between the line of equality and the concentration curve for the health care variable. The index is represented on a scale ranging from -1 to +1. Values equal to zero indicate an income-neutral distribution; values less than zero indicate a pro- poor distribution; and values greater than zero indicate a pro-rich distribution. This is shown in the example below (Table 14). Table 14. Example of concentration index method. Shown below is the percentage of individuals using doctor services, by quintile, using Australian National Health Survey for 2001. `Conc/HI Index' = concentration index. 40 Source: van Doorslaer et al, 2008 . Note: a positive (negative) index indicates a pro-rich (pro-poor) distribution. Index in bold typeface indicates statistically significant difference from zero at 95% confidence level. To measure the degree of progressivity of out-of-pocket payments, the distribution of these payments along income (expenditure) quintile is then compared to the overall degree of income (expenditure) inequality. Based on methods described in ODonnell et al, the Kakwani index is applied as one of the most widely accepted measures of expenditure progressivity. Specifically, the Kakwani index is used to test the progressivity of out-of-pocket payments in this paper, as an assessment of other sources of financing such as taxes and government subsidies was outside the scope of this paper. The Kakwani index (K.I.) is definied as the difference between the concentration index (C.I.) and the Gini coefficient (G.I.) of the ability-to-pay variable, defined as household expenditure in this analysis, where K.I. = C.I. ­ G.I. The value of the Kakwani index ranges from -2 to 1. A negative number indicates regressivity, where the share of out-of- 32 pocket payments is disproportionately higher amongst lower-income groups. A positive number indicates progressivity, where the share of out-of-pocket payments is disproportionately higher amongst higher-income groups. The proportion of households that incur an exceptionally large amount of out-of-pocket health expenditures is evaluated, referred to as the incidence of ,,catastrophic expenditures. The incidence of catastrophic expenditures is a measure of the extent to which health care payments impinge on living standards. The threshold for catastrophic health expenditures is typically set at a value between 5% and 25% of total household consumption expenditure; or 25% to 40% of non-food household expenditure39. For purposes of this analysis, the threshold is set at 10% of total household expenditure for most of the case studies, with the exception of Iran, where the threshold is set at 25%. Assessing the poverty impact of out-of-pocket payments To assess the burden of out-of-pocket payments and catastrophic expenditures, the extent to which health care payments impinge on living standards is evaluated across the case studies. Most poverty measures do not typically take into account health care-related expenditures or do not explicitly show the contribution of health care payments to poverty levels41. The poverty headcount represents the percent of households that fall below the poverty line, defined in relative or absolute terms. Where data permits, the poverty headcount and depth of poverty as a percentage of the poverty line, or the normalized gap, is estimated before and after accounting for out-of-pocket payments for health care. The poverty lines used have been chosen on the basis of the local context, representing either a relative or absolute poverty line, although other approaches by which to measure poverty exist as described elsewhere42. The relative poverty line is defined in relation to a nations mean income per capita and was used in the case of Egypt. Absolute poverty lines are set according to international poverty lines estimated by the World Bank and were used in the cases of Lebanon, Iran, West Bank-Gaza, Yemen and Tunisia. 33 34 References 1 Rannan-Eliya, R.P., Blanco-Vidal, C., Nandakumar, A.K. (1998). The distribution of health care resources in Egypt: implications for equity: an analysis using a national health accounts framework. Data for Decision Making Project, Department of Population and International Health. Boston MA: Harvard School of Public Health. 2 Ekman B (2007). The impact of health insurance on outpatient utilization and expenditure: evidence from one middle-income country using national household survey data. Health Res Policy Syst 5(6). Published online 30 May 2000. 3 Hosseinpour AR, Mohammed K, Majdzadeh R, Naghavi M, Abolhassani F, Sousa A, Speybroeck N, Jamshidi HR and Vega J (2005). Socioeconomic inequality in infant mortality in Iran and across its provinces. Bulletin of the World Health Organization 83(11): 837-844. 4 Elgazzar H (2009). Income and the use of health care: an empirical study of Egypt and Lebanon. Health Economics, Policy and Law 4(4): 445-478. 5 Gwatkin DR, Rutstein S, Johnson K, Suliman E, Wagstaff A, and Amouzou A (2007). Socio economic differences in health, nutrition, and population within developing countries: an overview. Washington, DC: World Bank. 6 Commission on Social Determinants of Health (2008). Closing the gap 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 Organization. 7 van Doorslaer E, Wagstaff A, van der Burg H et al (1999). The redistributive effect of health care finance in twelve OECD countries. J Health Econ 18(3):291-313. 8 ODonnell et al (2005). ,,Explaining the incidence of catastrophic expenditures on health care: comparative evidence from Asia, Equitap Project Working Paper #5. Rotterdam: Erasmus University. 9 Baeza CC and Packard TG (2006). Beyond survival: protecting households from health shocks in Latin America. Washington DC: World Bank, and Berkeley: Stanford University Press. 10 World Bank (2009). World Development Report 2009: Reshaping Economic Geography. Washington DC: World Bank. 11 Lieberman SS and Wagstaff A (2009). Health financing and delivery in Vietnam: looking forward. Washington DC: World Bank. 12 Ammar W (2003). Health system and reform in Lebanon. Cairo: Regional Office for the Eastern Mediterranean, World Health Organization; and Beirut: Entreprise Universitaire dEtudes et de Publications. 13 Wagstaff A and van Doorslaer E (2003). Catastrophe and impoverishment in paying for health care: with applications to Vietnam 1993-1998. Health Economics 12(11): 921-934. 14 Wagstaff A, van Doorslaer E, et al (1999). Equity in the finance of health care: some further international comparisons. Journal of Health Economics 18(3): 263-290. 35 15 Wagstaff A and Lindelow M (2008). Can insurance increase financial risk? The curious case of health insurance in China. Journal of Health Economics 27(4): 990-1005. 16 Wagstaff A (2009). ,,Measuring financial protection in health, in Smith PC, Mossialos E and Leatherman S (eds)., Performance measurement for health systems improvement: experiences, challenges and prospects. Cambridge: Cambridge University Press. 17 van Doorslaer E, O'Donnell O, Rannan-Eliya RP et al (2006). Effect of payments for health care on poverty estimates in 11 countries in Asia: an analysis of household survey data. Lancet 368(9544):1357-64. 18 Coudouel A Hentschel JS and Wodon QT (2002). ,,Poverty measurement and analysis, in Klugman J (ed.), A sourcebook for poverty reduction strategies, Washington DC: The World Bank. 19 Rassekh B and Segaren N (2009). ,,Review of community empowerment strategies for health, in Peters DH, El-Saharty , Siadat B et al (eds.), Improving health service delivery in developing countries. Washington, DC: World Bank. 20 Gertler P (2000). Final report: the impact of PROGRESA on health. Washington, DC: International Food Policy Research Institute. 21 Morris SS, Flores R, Olinto P and Medina JM (2004). Monetary incentives in primary health care and effects on use and coverage of preventative health care interventions in rural Honduras: a cluster randomized trial. Lancet 364:2030-2037. 22 de la Briere B and Lindert K (2005). Reforming Brazils Cadastro Unico to improve the targeting of the Bolsa Familia Program. Social Protection Discussion Paper Series, report no. 527. Washington DC, World Bank. 23 Lagarde M, Haines A and Palmer N (2007). Conditional cash transfers for improving uptake of health interventions in low- and middle-income countries: a systematic review. Journal of the American Medical Association. 298(16): 1900-1910. 24 World Bank (2009). World Development Report 2009: Reshaping economic geography. Washington, DC: World Bank. 25 World Bank (2008). Islamic Republic of Iran: Health Sector Review, Volume I, Main Report. Report no. 39970-IR. Washington, DC: World Bank. 26 World Bank (2008). Islamic Republic of Iran: Spatial patterns of poverty and economic activity. Report no. 46603. Washington, DC: World Bank. 27 World Bank and Ministry of Public Health, Tunisia (2006). Tunisia Health Sector Study. Report No. 41018. Washington, DC: World Bank. 28 Ministry of Public Health, Tunisia (2009). Health insurance coverage statistics, 2009, Tunisia. 29 Douidich M, Ezzari A and Lanjouw P (2008). Simulating the impact of geographic targeting on poverty alleviation in Morocco: what are the gains from disaggregation? Policy Research Working Paper, no. 4724. Washington, DC: World Bank. 36 30 World Bank (2008). Ciblage: comment advancer pour reduire la pauvrete au Maroc. Washington, DC: World Bank. 31 World Bank (2005). Implementation completion report: West Bank and Gaza Health System Development Project. Report no. 33572. Washington, DC: World Bank. 32 Wagstaff A and Yu S (2007). Do health sector reforms have their intended impacts? The World Banks Health VIII Project in Gansu Province, China. Journal of Health Economics 26(3): 505-535. 33 Sadrizadeh B (2004). Primary health care experience in Iran. Medical Journal of Iranian Red Crescent 7(1):79-90. 34 World Bank (forthcoming). Impact evaluation of the Egyptian health sector reform project. Washington, DC: World Bank. 35 World Bank (2004). Implementation Completion Report, Morocco: Social Priorities Program Basic Health Project. Report No: 28390-MOR. Washington, DC: World Bank. 36 World Bank (2007). Yemen Country Social Analysis. Report No. 34008-YE. Washington, DC: World Bank. 37 Wagstaff A and van Doorslaer E (1993). ,,Equity in the finance and delivery of health care: concepts and definitions, in van Doorslaer E, Wagstaff A and Rutten F (eds.). Equity in the finance and delivery of health care: an international perspective. Oxford: Oxford University Press. Pp. 8-19. 38 Wagstaff A, Paci P and van Doorslaer E (1991). On the measurement of inequalities in health. Social Science & Medicine 33(5):545-557. 39 ODonnell O, van Doorslaer E, Wagstaff A and Lindelow M (2008). Analyzing health equity using household survey data: a guide to techniques and their implementation. Washington, DC: World Bank. 40 van Doorslaer E, Clarke P, Savage E, and Hall J. (2008). Horizontal inequities in Australias mixed public/private health care system. Health Policy 86: 97­108. 41 van Doorslaer E, O'Donnell O, Rannan-Eliya RP, Somanathan A, Adhikari SR, Garg CC, Harbianto D, Herrin AN, Huq MN, Ibragimova S, Karan A, Ng CW, Pande BR, Racelis R, Tao S, Tin K, Tisayaticom K, Trisnantoro L, Vasavid C, and Zhao Y (2006). Effect of payments for health care on poverty estimates in 11 countries in Asia: an analysis of household survey data. Lancet 368(9544):1357-64. 42 Coudouel A, Hentschel JS and Wodon QT (2002). ,,Poverty measurement and analysis, in Klugman J (ed.), A sourcebook for poverty reduction strategies, Washington DC: The World Bank. 37 About this series... This series is produced by the Health, Nutrition, and Population Family (HNP) of the World Bank's Human Development Network. The papers in this series aim to provide a vehicle for publishing preliminary and unpolished results on HNP topics to encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account this provisional character. For free copies of papers in this series please contact the individual authors whose name appears on the paper. Enquiries about the series and submissions should be made directly to the Editor Homira Nassery (hnassery@worldbank.org) or HNP Advisory Service (healthpop@worldbank.org, tel 202 473-2256, fax 202 522-3234). For more information, see also www.worldbank.org/ hnppublications. THE WORLD BANK 1818 H Street, NW Washington, DC USA 20433 Telephone: 202 473 1000 Facsimile: 202 477 6391 Internet: www.worldbank.org E-mail: feedback@worldbank.org