Series

Health in Southeast Asia 6 Health-nancing reforms in southeast Asia: challenges in achieving universal coverage
Viroj Tangcharoensathien, Walaiporn Patcharanarumol, Por Ir, Syed Mohamed Aljunid, Ali Ghufron Mukti, Kongsap Akkhavong, Eduardo Banzon, Dang Boi Huong, Hasbullah Thabrany, Anne Mills

In this sixth paper of the Series, we review health-nancing reforms in seven countries in southeast Asia that have sought to reduce dependence on out-of-pocket payments, increase pooled health nance, and expand service use as steps towards universal coverage. Laos and Cambodia, both resource-poor countries, have mostly relied on donorsupported health equity funds to reach the poor, and reliable funding and appropriate identication of the eligible poor are two major challenges for nationwide expansion. For Thailand, the Philippines, Indonesia, and Vietnam, social health insurance nanced by payroll tax is commonly used for formal sector employees (excluding Malaysia), with varying outcomes in terms of nancial protection. Alternative payment methods have dierent implications for provider behaviour and nancial protection. Two alternative approaches for nancial protection of the non-poor outside the formal sector have emergedcontributory arrangements and tax-nanced schemeswith dierent abilities to achieve high population coverage rapidly. Fiscal space and mobilisation of payroll contributions are both important in accelerating nancial protection. Expanding coverage of good-quality services and ensuring adequate human resources are also important to achieve universal coverage. As health-nancing reform is complex, institutional capacity to generate evidence and inform policy is essential and should be strengthened.

Lancet 2011; 377: 86373 Published Online January 25, 2011 DOI:10.1016/S01406736(10)61890-9 See Comment page 792 See Comment Lancet 2011; 377: 355, 534, 619, and 700 See Online/Comment DOI:10.1016/S01406736(10)62140-X This is the sixth in a Series of six papers about health in southeast Asia International Health Policy Program, Ministry of Public Health, Nonthaburi, Thailand (V Tangcharoensathien PhD, W Patcharanarumol PhD); Siem Reap Provincial Health Department, Ministry of Health, Siem Reap, Cambodia (P Ir MPH); United Nations University, Kuala Lumpur, Malaysia (Prof S M Aljunid PhD); Gadjah Mada University,

Introduction
The large amount of household out-of-pocket payments for medical bills, resulting in household nancial disruption and impoverishment, was a key motive for the adoption of a World Health Assembly Resolution in 2005 on nancial protection.1 Countries in southeast Asia, home to 87% of the worlds population and that have a fast economic growth and a moderate poverty level of 146%,2 have a high potential to accelerate protection from nancial risks and achieve universal coverage of health care. Universal coverage is dened as securing access by all citizens to appropriate promotive, preventive, curative, and rehabilitative services at an aordable cost.3 Prospects of progress towards this aspiration seem poor,4
Search strategy and selection criteria A common protocol was developed by VT and WP and agreed upon by country authors. The protocol included country health nancing background, analysis of the government eorts in coverage extension to the poor, those in formal employment, and the informal sector. The analysis covered source of revenue, pooling, service coverage, level of nancial risk protection, and the government policy towards universal coverage. In producing each country report, authors reviewed and synthesised the published literatures and other government unpublished documents, such as the Philippine health insurance corporation annual reports, the government statistics year book in Vietnam, International Labour Oce statistics, and Laos Health Financing Strategies 201115. VT and WP then compiled and synthesised the nal references on the reports from the country authors.

particularly for countries whose government scal capacity is low and whose social health insurance for the employed sector is absent or very small, thus restricting the mobilisation of additional resources from payroll contributions. Financing health care in most developing countries greatly relies on out-of-pocket payments,5 with most donors and global health initiatives such as the Global Fund focusing on specic diseases or interventions rather than the broader health system.

Key messages The development of a universal health coverage policy is guided by explicit consideration of how best to cover and nance specic population groups: those in formal employment, the poor and vulnerable, and the informal sector and the rest of the population. Those in formal employment can be given nancial protection through payroll-nanced social health insurance or tax-funded arrangements. The poor and vulnerable are accepted to need highly subsidised arrangements by general budget, and there is good evidence from Laos and Cambodia that demand-side targeted approaches such as health equity funds work better than a simple fee exemptions policy. The informal sector and the rest of the population remain a challenge, with countries such as the Philippines and Vietnam seeking to expand coverage through contributory arrangements, and others such as Thailand using tax funding. In addition to extension of population coverage, eorts should be given to provide adequate nancial risk protection and to design an appropriate mix of provider payment methods that can aect physicians clinical practices towards rational use of medical technologies, eciency, and long-term aordability. Expanding coverage of good-quality services and ensuring adequate human resources are equally important elements of achieving universal health coverage. Comparative analysis such as that presented in this paper is helpful in bringing diverse experiences from the southeast Asia region together to learn lessons and develop a culture of evidence in decision making.

www.thelancet.com Vol 377 March 5, 2011

863

Series

Yogyakarta, Indonesia (Prof A G Mukti PhD); National Institute of Public Health, Ministry of Health, Vientiane, Laos (K Akkhavong MPH); World Bank, Manila, Philippines (E Banzon MD); Vietnam Health Economic Association, Hanoi, Vietnam (D B Huong MSc); University of Indonesia, West Java, Indonesia (Prof H Thabrany PhD); and London School of Hygiene and Tropical Medicine, London, UK (Prof A Mills PhD) Correspondence to: Dr Viroj Tangcharoensathien, International Health Policy Program, Ministry of Public Health, Tiwanon Road, Nonthaburi 11000, Thailand viroj@ihpp.thaigov.net

See Online for webappendix

Countries with a high share of out-of-pocket payments are more likely to have a high proportion of households facing catastrophic health expendituredened as spending more than 40% of household consumption expenditure, excluding food, on health, more than 25% of non-food consumption expenditure of households on health, or more than 10% of total household consumption expenditure on health.6 A 1% increase in the proportion of out-of-pocket payments in total health expenditure is associated with a 22% increase in the proportion of households facing catastrophic health payments. The larger the share of prepayment in health-care nancing, the smaller the proportion of households that face catastrophic health spending.7 However, there is no strong evidence that countries with social health insurance oer better or worse protection than do countries that rely on general taxation.8 Nevertheless, the existence of prepayment does not guarantee nancial protectioninadequate nancial protection has been reported from some prepayment schemes. For example, 15% of individuals enrolled in the insurance scheme of the Self-Employed Womens Association in India faced a nancially catastrophic level of payment even after reimbursement for hospital admission,9 and the Chinese Rural Cooperative Medical System covers only 30% of inpatient expenditure.10 For universal coverage, progress on three general areas is needed: extension of population coverage of health insurance schemes or other forms of prepayment, specication of which types of services should be provided and ensuring their availability and quality, and improving nancial risk protection (webappendix p 1). The breadth is population coverage by insurance schemes, the depth means service coverage such as outpatient, inpatient, and other high-cost services, and the height is the level of nancial protection such as co-payment. The smaller the co-payment by users and

the more comprehensive the service coverage, the higher the protection against nancial risk. We focus discussion on these three areas. The key dilemma in resource-poor settings is the choice between providing a high level of service and nancial protection for a small group of the population versus extending a high level of population coverage but with restricted services and nancial protection. In this paper, we assess approaches to nancing healthcare reform and progress towards universal coverage in seven low-income and middle-income countries in the southeast Asia region. Brunei and Singapore, two highincome countries, were excluded from this analysis, as was Myanmar, for which there is little information on health nancing. On the basis of documentary analysis, we review achievements of the health-nancing reforms of these countries and identify challenges with regards to population coverage, service coverage, and nancial protection to share lessons and to inform the nancingreform eorts of countries outside this region.

Country background
Seven countries in southeast Asia with dierent levels of economic development and pace of expansion of healthservice coverage and nancial protection were selected as case studies: two low-income countries with low coverage (Laos and Cambodia), and ve middle-income countries, three of which have more than 50% coverage and clear policies towards universal coverage (Indonesia, the Philippines, and Vietnam), and two of which have achieved universal coverage (Malaysia and Thailand). Table 1 shows the wide variation in economic and poverty indicators among these countries. Fiscal space, the governments ability to collect tax and to spend funds for desired purposes, measured as a share of the gross domestic product, ranges from 82% in Cambodia to 168% in Thailand (by contrast with an average of 225% in high-income countries in 2007).11
Poverty incidence (% below national poverty line) Poverty headcount (%)

Gross national income per GDP yearly growth (%)* capita in 2008 (PPP$)* 2000 Malaysia Thailand Philippines Indonesia Vietnam Laos Cambodia 13 740 5990 3900 3830 2700 2040 1820 89 48 60 49 68 58 88 2005 53 46 50 57 84 71 133 2008 46 26 38 61 61 75 52

Fiscal space: government tax (% of GDP)*

166 (2003) 168 (2007) 143 (2006) 123 (2004) 130 (2007) 101 (2007) 82 (2006)

87 (2004) 210 (2000) 85 (2007) 329 (2006) 202 (2009) 182 (2006) 135 (2008) 320 (2002) 270 (2008) 347 (2004)

NA NA 226 (2006) 294 (2007) 215 (2006) NA 258 (2007)

As data on poverty from national estimates in some countries are scarce and irregularly reported, some countries only have reports of poverty indicators from one year. GDP=gross domestic product. PPP=purchasing power parity. NA=not available. *World Development Indicators database, April 2009.11 Fiscal space of Vietnam was analysed by the country author (DBH) on the basis of data from the General Statistical Oce, Vietnam. Ocial country sources. World Development Indicators database (Aug 31, 2010).11 Percentage of population living at or below PPP US$125 per day.12

Table 1: Economic and poverty backgrounds of seven countries in southeast Asia

864

www.thelancet.com Vol 377 March 5, 2011

Series

The poverty incidence is not only suggestive of the number of people who cannot aord to pay medical costs when they are sick, but also indicates the magnitude of the health budget needed if governments decide to subsidise the poor. This subsidy puts pressure on the scal space, and, in resource-poor countries such as Laos and Cambodia, funding from donors is inevitable to support access for the poor to health services, especially services linked to the Millennium Development Goals.

Health-nancing challenges
Level and prole of health expenditure
Private health expenditure has a dominant role in nancing health care in ve of the seven countries, contributing more than 70% of total spending on health in Laos and Cambodia (table 2),13,14 although the level of catastrophic health expenditure diers between these countries, consisting of 5% of households in Cambodia and 105% in Vietnam.15 Less than 9% of the government budget is allocated to health in ve of these seven countries exceptions being Cambodia (because government funding includes donor support channelled through government) and Thailand (because loans from development banks and bilateral sources were combined with the government budget, which need to be repaid). The high amount of external resources from donors in Cambodia (164% of total health expenditure) and Laos (145%) not only raises questions about long-term sustainability but also about the extent to which donor-funded programmes are in accordance with national priorities.16 Social health insurance nanced by payroll tax ranges from none in Cambodia to 127% of total health expenditure in Vietnam (table 2). Malaysia, an upper middle-income country with a high level of formal sector employment, has yet to establish a social health insurance schemesuch spending is only 04% of total health expenditure. Despite
THE (% GDP) Malaysia Thailand Philippines Indonesia Vietnam Laos Cambodia Low income Lower middle income Upper middle income High income Global 44 37 39 22 71 40 59 53 43 64 112 97 GGHE (% THE)* 444 732 347 545 393 189 290 419 424 552 613 596
13

the well established schemes in Thailand, the Philippines, and Vietnam, their spending was still below the lower middle-income country group average of 158% of total health expenditure, indicative of a smaller population size in the employed sector and a lower benet package than for other lower middle-income countries. Total health expenditure per capita in three of the countries, Indonesia, Laos, and Cambodia, is below the minimum US$4954 per capita17 estimated to be necessary to provide the interventions and health-system platform necessary to meet the Millennium Development Goals.

Population coverage by nancial protection schemes

The best estimates of insurance coverage for the country populations are categorised in webappendix p 2 into four relevant groups for 2009, on the basis of survey or administrative data. Because of the dierent pace of population coverage expansion, the total number of the insured population varies greatly, with low coverage in Laos and Cambodia, medium coverage in Indonesia and Vietnam, and high coverage in Thailand and the Philippines. Malaysia is reported to have 100% coverage because of its tax-funded system (although high out-of-pocket payments suggest eective coverage is less than this level). The high percentage of the uninsured population (webappendix p 2), combined with the high level of outof-pocket payments, put the uninsured people at risk of nancial impoverishment or forfeiting necessary health care, resulting in disability or deaths at home. Social health insurance coverage is low because of the small size of the formal sector.

Coverage and extension of nancial risk protection

The two most commonly used formal nancing approaches are social health insurance for formal sector
External (% of THE) 00 03 13 17 16 145 164 175 10 02 00 02 SHI (% THE) 04 71 77 87 127 23 00 46 158 210 256 246 Out-of-pocket THE (per (% THE) capita US$) 407 192 547 301 548 617 601 483 521 309 140 177 3072 1365 626 418 583 269 368 268 802 4879 44052 8023 THE (per capita PPP int$) 6044 2857 1302 810 1827 839 1081 670 1810 7570 41450 8625

Private health expenditure GGHE (% government (% of THE)* expenditure) 556 268 653 455 607 811 710 581 576 448 387 404 69 131 67 62 87 37 112 87 79 94 172 154

Data from the World Health Statistics, 2010. In accordance with National Health Accounts conventions, external nance is included within government and private shares (which sum to 100%). Private health expenditure includes out-of-pocket payments, private social insurance, and other private insurance. International dollars are used when comparing across countries. US dollars are used when looking specically in one country. THE=total health expenditure. GGHE=general government health expenditure. SHI=social health insurance. PPP=purchasing power parity. int$=international dollar. NA=not available.

Table 2: Key indicators of health nancing in seven countries in southeast Asia in 2007

www.thelancet.com Vol 377 March 5, 2011

865

Series

employees and general tax nance for the poor and vulnerable because these groups are generally accepted to be the responsibility of the government. Given these approaches, the coverage of the informal sector is a major challenge, described as squeezing the middle at a conference in 2006,18 the middle layer referring to the non-poor or to the not-so-poor informal sector, whereas the top layer consists of formal sector employees and the bottom layer comprises the poor. By social health insurance, we mean a payroll taxnanced scheme for employees in the public or private sectors, in which a specic portion of an employees salary is mandatorily deducted, with the employer also contributing an equal or higher portion. In some countries, the government also contributes. By contrast, tax-nanced schemes draw on general tax revenues and do not need pre-paid individual or household contributions. The population groups are re-categorised in webappendix p 3 to distinguish the economically active (formal and informal sectors) from the poor and the rest of the population, and indicates their size. The rest of the population consists of non-poor children and elderly dependants and other economically inactive groups. The poor include children, elderly dependants, and the poor in the informal sector. Despite the complexity of potential overlapping populations across these four broad groups, these categories are useful to inform policy on how health-nancial protection for each group should be nanced and progress in coverage extension monitored.

Protecting the poor and vulnerable

Cambodia introduced a user-fee policy in 1996 with the aim of improving the capacity of the health-care delivery system, as revenues were used to pay incentives to health workers, supplement the inadequate government budget, and resolve irregularities of budget disbursement.19 However, user fees created a barrier for the poor in the absence of an eective exemption system.19,20 Since the rst pilot trial in 2000, the health equity fund, which is mostly nanced by donors to compensate health facilities for medical expenditures of the poor and to pay some travelling costs, has been gradually expanded, covering about 68% of the poor population, or 23% of Cambodias total population, by 2008.21 Evidence suggests that this fund has improved access of medical services for the poor and potentially provided nancial protection. Although there has not been a methodologically rigorous study assessing the eect of the health equity fund on health-care access and nancial protection, several case studies have indicated a substantial increase in hospital use by poor members of the fund, without a decrease in use by self-paying patients. In most cases, the number of beneciaries of the fund accounted for more than a third of total hospital inpatients.2224 However, the nancial sustainability and government capacity to expand using its own resources have been questioned.25
866

The 1995 user charge policy in Laos26 made provision for exempting the poor from payments, but this approach did not work well as village leaders veried the poor on an ad-hoc basis. Free care for the poor was a mandate with inadequate fundingapart from routine allocations for medicines and sta salary, there was no additional budget line for this purpose.27 Health centres and hospitals were reluctant to subsidise the poor using their own revenue from user fees. A donor-funded health equity fund piloted in 2003 has been expanded after an assessment reported an increased use of services by the poor, and recent government policy dialogues have favoured increasing funding for the poor.28 In response to the 1997 Asian economic crisis, which heavily aected the poor,29 Indonesia introduced a taxnanced targeted scheme for the poor and the near poor, including homeless people and orphans. Finance is from central and district governments, and providers are paid on a case mix-adjusted basis for both outpatient and inpatient services. Nationwide expansion of the scheme reached 334% of the total population by 2008, so almost all the poor and the near poor are covered. From hospital administrative records, use has increased for ambulatory and inpatient care,30 and the gap in the use of services has reduced between the rich and the poor. Because of scal constraints, the per capita government subsidy is only US$6 per year for a package of outpatient and inpatient services (relative to a total health expenditure of $418 per capita), and so might result in a low level of service provision and nancial protection. Out-of-pocket payment also remains high. Since October, 1997, the Philippine Health Insurance Corporation (PhilHealth) has introduced a sponsored programme for poor households that are identied and registered by the local government. The premium for this programme is subsidised by central (mean contribution 80%, range 5090%) and local (mean contribution 20%, range 1050%) governments. Yearly enrolment has depended on local government political will and scal capacity; for example, peaking during election years. Between 1975 and 2002, Thailand operated a scheme designed to reach the poor when universal coverage was introduced. Initially, partial to full exemption was left to the discretion of health workers and, subsequently, a means test (to verify whether an individual or family was eligible for government help) was used to identify the poor; this step was initially applied by health workers and later by a local committee. Despite the community involvement, nepotism resulted in under-coverage of the poor, and the non-poor linked to local politicians commonly beneted.31 A common trend has emerged across the countries that health services for the poor are subsidised by tax through budget allocations to public providers, with additional support in Laos and Cambodia from donors to health equity funds. Historically, means testing to identify the poor has not been very accurate,32,33 and this
www.thelancet.com Vol 377 March 5, 2011

Series

imprecision remains a challenge in the countries that rely on it. Panel 1 compares the targeting experiences in the Philippines, Laos, and Cambodia where lessons can be drawn from the dierent approaches in identifying the poor in these three countries.

Panel 1: Challenges in targeting the poorlessons from Cambodia, Laos, and the Philippines In Cambodia, beneciaries of the health equity fund are identied on the basis of eligibility criteria either at the community level (pre-identication) or at health facilities through questionnaire interviews using proxy means tests (such as durable assets, housing, land ownership, and number of working members, dependants, and disabled members) and estimates of household income, expenditure, and debt. Identication at point of service picks up those missed at the community level. In Laos, poor households eligible for the health equity fund are identied by a village committee, using certain means-testing criteria. In areas not covered by the fund, the village head issues a letter at the request of a patient, certifying the individual as poor on a case-by-case basis. Unlike beneciaries of the fund who get the cost of their free care reimbursed to hospitals, poor individuals in non-fund areas have to negotiate for exemption with providers as there is no budget line to subsidise free care for the poor. In practice, some patients are allowed to delay payment.27 In the Philippines, those who are indigent for a certain period are assessed by local government units, using a family income test, and are enrolled into a programme that has budget subsidies to cover outpatient and inpatient care. The recently elected government in 2010 has now mandated the central Department of Social Welfare and Development to manage this assessment, because income tests are inconsistently applied by local government units. Potential leak of benets to the non-poor is likely in all three countries, although further study is needed, especially in Laos and Cambodia. In these two countries, supporting transport costs for fund beneciaries, in addition to medical costs, has been essential to facilitate access to care by the poor. Lessons Ad-hoc certication in non-fund areas, and limited funding, are major factors in Laos for under-coverage of the poor. The health equity funds in Laos and Cambodia, with clear identication procedures and reliable funding, have improved rates of use of services and tend to provide better nancial protection. Similarly, the sponsored programme of PhilHealth, with clear targeted funding, has improved access and use. In addition to the provision of basic quality health care, support of transport and food for poor patients during their admission to hospital seems to be essential. Although challenging, objective criteria and transparent and participatory engagement by local communities in identifying the poor, as experienced in all three countries, are essential to prevent favouritism and leakage to the non-poor.

Protecting the formal employment sector

Thailand, the Philippines, Indonesia, Vietnam, and Laos use mandatory social health insurance for the formal sector. This insurance is commonly managed by a non-prot independent body with a clear governing structure, and services are purchased on behalf of members. A percentage of the payroll is deducted from employees and an equal or higher contribution made by employers, and some governments also contribute as in Thailand. A social health insurance scheme can have a major strategic purchasing role in regulating public and private provider behaviour and in achieving goals of eciency, quality, and nancial protection. Dierent provider payment arrangements can have dierent eects on doctors clinical decisions and behaviour on resource use.34 International experience indicates that a fee for service payment stimulates unnecessary diagnosis, prescription, and treatment, resulting in cost escalation; closed-end payment such as capitation and case-based payment have lower costs. The eects of capitation and case-based payment are an increased use of generic medicines and an increased use of proper diagnosis and treatment, resulting in cost containment. The design of PhilHealth does not provide adequate nancial protection for its members. Outpatient services are not covered; inpatient care is reimbursed up to a maximum amount, leading to balance billing, when patients pay additional bills beyond the level of reimbursement. The share of social health insurance in total health expenditure was 11% in 200535 and has declined to 85% in 2007,3638 indicative of increasingly restricted nancial protection to members. PhilHealth found that reimbursement was only slightly more than a third of the total medical bill paid by patients in 2008,39 and aims to improve nancial protection of members. An increased incidence of catastrophic health spending (dened as >25% of non-food consumption expenditure of households) has been observed, from 211% of the total population in 2000 to 221% in 2003 and 297 in 2006.40 Whereas the PhilHealth fee-for-service model ensures free choice of the patient for their provider, the Thai social health insurance scheme introduced in 1991 limits such choice through a capitation contract model. Members register yearly with preferred public or private contractors and, in return, contractors are paid a capitation fee, currently 1900 Baht (US$57) per member, to provide all outpatient and inpatient services. Balance billing is illegal. The scheme covers private employees
www.thelancet.com Vol 377 March 5, 2011

onlytheir dependants fall under the universal coverage scheme, and public employees and dependants fall under a separate, non-contributory scheme nanced by general tax. Panel 2 describes the experiences of resistance to reform in Malaysia. The Thai capitation model ensures containment of cost and transfers nancial risk to providers, whereas fee-forservice transfers nancial risk to PhilHealth members through balance billing. The risk under capitation is inadequate services, so unit costs and rates of use are monitored and members can change contractor yearly if they are not satised. Results from studies have suggested that service use of this model in Thailand is adequate in terms of rate of use (more than two visits per person per year) and good quality of care provided to social health insurance members.4446
867

Series

Vietnam, having experienced the drawbacks of fee-forservice schemes such as excessive diagnosis and treatment and levels of co-payment up to 30% of total bills, introduced a law in 2008 on health insurance that provides for capitation for primary care services and case-based payment to be used for inpatient care. This programme is expected to be fully implemented by 2014. Strategic purchasing, in particular design of benet package and provider payment method, establishes system eciency, and level of out-of-pocket and catastrophic spending. Once a payment system is entrenched, particularly in cases for which private-forprot providers dominate the health-care market, radical reform from fee-for-service to capitation or case-based payment will face united resistance from the medical profession, as experienced in South Korea.47 Introducing the right purchasing strategies at an early stage is a key foundation for the successful performance of social health insurance. Panel 2 indicates some of the complexities of agreeing the introduction and design of social health insurance in Malaysia.

Protecting the informal sector and the rest of the population

The informal sector and the rest of the population make up a large proportion of these countries; for example, 49% in Cambodia, 64% in Indonesia, and 73% in Vietnam. Because of the large numbers, their restricted capacity to pay premiums, and the feasibility of enforcing payment, extension of coverage to this group is especially challenging. These seven countries have faced a key choicebetween a contributory scheme and a general tax-nanced scheme. Both PhilHealth and the Vietnam social insurance scheme use a contributory approach to extend coverage to the informal sector, with premiums collected from groups such as taxi drivers and street vendors. PhilHealth seeks to collect a xed yearly premium of 1200 Peso (US$258) from individual members, but enforcement is not eective despite huge eort and various innovations. Furthermore, the administrative cost of premium collection is high and collection complex because of high mobility and interruption and seasonality of cash income. There is adverse selection because members enrolling individually are mostly chronically ill and have high rates of use. This element of PhilHealth needs subsidies from the payroll tax-nanced component. In Vietnam, tax funding is used to subsidise the premium for the informal sector by 50%. There is a risk that coverage might stall once the easy-to-reach population has been enrolled, and the administrative cost of premium collection will be high in hard-to-reach remote areas. Moreover, information from impact assessment studies suggests that the fund has not reduced average out-of-pocket spending and has had negligible eects on use among the poorest population, although it has substantially increased overall service use and reduced the risk of catastrophic spending.48 In Thailand, despite community-based49 and then publicly subsidised voluntary health insurance,50 30% of the total population remained uninsured in 2001, mostly in the informal sector. In addition to problems of adverse selection and nancial viability,51 enforcement of premium payment in the informal sector is not technically feasible. When an opportunity arose with a political demand to reach universal coverage in a year, as promised in the January 2001 general election campaign, a contributory scheme was ruled out on reasons of speed and because it was politically inadvisable because of its implications for voters supporting the new government. The political context at that time provided no option but to adopt general tax funding for universal coverage, although nancial assessment indicated its feasibility for both short term (ie, 1 year) and medium term (ie, 5 years) at the time.52 The caveat is the question of nancial feasibility in the long term as the population of Thailand age and their demands increase. Thailand has extended tax nancing from the poor to the informal sector and the rest of the population (ie, squeezed bottomup), whereas the Philippines and Vietnam have extended the contributory scheme from the formal to the informal sector (ie, squeezed topdown).
www.thelancet.com Vol 377 March 5, 2011

Panel 2: Complexities of introducing social health insurance in Malaysia In Malaysia, an upper middle-income country, health services are free for all citizens at primary, secondary, and tertiary levels with minimum co-payment, ranging from 100 RM (US$031) for outpatients to 300 RM ($094) per admission day. The country spent $3072 per capita on health in 2007, using supply-side nancing through yearly budget allocations to public-sector providers. Despite this relatively high expenditure, various problems are apparent: high levels of out-of-pocket payment making up 407% of total health expenditure (mostly spent on secondary and tertiary private services); long waiting times for procedures in public hospitals (eg, 23 weeks for orthopaedic surgery41); rising health care costs because of the epidemiological transition in the face of limited public funds; and poorly regulated private fees. Between 1985 and 1996, the Government commissioned ve reviews on health nancing; recommendations were made that the Government should establish a National Health Financing Scheme to pool resources from both public and private sources and to provide universal nancial risk protection based on social health insurance principles. Discussions on health-nancing reform were restarted in 2000. From 2000 to 2006, multi-stakeholder meetings were convened to discuss the National Health Financing Mechanism. However, no decision was made and various barriers can be identied in addition to absence of political will: Loser versus gainer dierences: the proposed introduction of social health insurance necessitates mandatory contributions by the formal sector such as civil servants and private sector employees who have reservations about having to pay on top of personal income tax. The voices of the informal sector and the poor who are potential gainers from the new scheme are not heard. Social solidarity mechanisms seem insucient to overcome opposition. Private interests: there is strong lobbying by private health insurance operators who fear the scheme will dilute their prots. Institutional conict of interest: the proposed National Health Financing Authority, which will administer the national scheme, threatens the Ministry of Health, which might lose all its nancing power to the Authority. Technical barriers: collection of premiums from the informal sector is dicult.
Information from Yon and colleagues.42,43 RM= Ringgit Malaysia.

868

Series

Thailand 100 80 64% 64% 60 Population Insurance coverage

100% 98%

Philippines

100% 76%

45% 40 20 0 Indonesia 100 80 64% 60 40 20 0 Laos 100 80 60 40 20 0 Formal sector 14% 4% 2% The poor 2% Informal sector and ROP 27% 17% 8% 0% Total Formal sector The poor 59% 48% 35% 23% 1% Informal sector and ROP Total 24% 100% Cambodia 16% 15% 20% 20% 13% 13% 13% 14% 14% 48% 28% 55% 100% Vietnam 100% 73% 27% 25% 9% 9% 35% 22% 33% 18% 23%

100%

Figure 1: Insurance coverage for three population groups in 2009 Malaysia is not included because it has 100% coverage. In the Philippines, the formal sector covered by PhilHealth (35%) includes public and private employees and their spouse and dependants, whereas the target population (22%) from the International Labour Oce statistics covers only the public and private sector employees. Data from webappendix pp 23. ROP=rest of population.

Figure 1 summarises the achievements in insurance coverage extension by 2009 for three population groups (including the informal sector and rest of population groups together) in six countries. Laos faces challenges in coverage extension to all groups, whereas Vietnam has fully covered the formal sector and the poor but has a major challenge covering the informal sector and the rest of the population through a contributory scheme. Cambodia has made good progress in using health equity funds to cover the poor, although this achievement needs to be sustainedintroducing social health insurance for the formal sector and devising arrangements to cover the large informal sector is a huge challenge both for scal capacity and programme management. The Philippines face two major challenges, to extend coverage to the poor by encouraging increased local government nancial commitments, and to enrol the hard-to-reach informal sector into the individual contributory scheme. Huge challenges in Indonesia are also coverage extension to the informal sector and the rest of the population with a clear policy on sources of nancing, while sustaining coverage of the poor and near-poor in a fully decentralised system.
www.thelancet.com Vol 377 March 5, 2011

By contrast, Thailand and Malaysia have reached a coverage for the whole population. This region of southeast Asia still has a huge gap of insurance coverage, which is a daunting challenge in the next wave of reform eorts.

Discussion and recommendations

Table 3 summarises achievements in the three areas of population coverage, service coverage, and nancial protection. Population coverage has been established by willingness and capacity to subsidise the poor, enforce formal sector enrolment into social health insurance, and protect the rest of the population through prepayment, whether through tax or contributions. Service coverage is indicative of previous and current investments in the health-service infrastructure, and decisions on benet packages for the various schemes. The level of nancial protection is established by willingness and scal capacity to purchase a large or small benet package, and by co-payment policy. The 76% estimate of insurance coverage for the Philippines is from PhilHealth; however, a recent household survey from the National Statistics Oce53 estimates
869

Series

Population coverage Health service coverage by nancial protection schemes by nancial protection schemes Malaysia 100% Primary care services focus on maternal and child health; curative services are free for all. Services are rationed by waiting time and number of family physicians in health centres; patients opt to pay for private services; survey reports 62% of ambulatory care was provided by private clinics. Comprehensive benet package, free at point of service for all three public insurance schemes. Benet package covers admission only except for the sponsored programme, which also covers outpatient services; high level of co-payment for all PhilHealth components: average reimbursement is 54% of the total medical bill, the balance being paid out-of-pocket. Although the policy intention is to provide comprehensive services, the low per capita government subsidy for the poor of US$6 per year for a package of outpatient and inpatient services might result in inadequate service provision, high levels of self-payment, and low levels of nancial protection. Benet package is comprehensive but has a substantial level of co-payment: 520% of medical bills. In principle, there is a comprehensive coverage for social health insurance and government employee schemes, but low level of funding results in a small service package. The poor covered by the health equity fund are entitled to a comprehensive package, including transport cost and food allowance, but the scope and quality of care provided at government health facilities are restricted.

Financial protection for the total population (measured by out-ofpocket costs as % of THE, 2007) 407%

Thailand Philippines

98% 76%

192% 547%

Indonesia

48%

301%

Vietnam Laos Cambodia

548% 77% 24%

548% 617% 601%

Information is from synthesis of the authors research. THE=total health expenditure.

Table 3: Summary population, service coverage, and nancial protection in seven countries in southeast Asia in 2009

100 % Insurance coverage 80 60 40 20 0 0 20

Malaysia (166%) Philippines (143%)

Thailand (168%)

Vietnam (130%) Cambodia (82%)

Indonesia (123%)

Laos (101%)

40 GGHE as % THE

60

80

100

Figure 2: Fiscal space in the context of insurance coverage and general government expenditure The size of the spheres indicate the size of the scal space as measured by tax revenues as percentage of gross domestic product. GGHE=general government health expenditure. THE=total health expenditure.

a national coverage of 38%, suggesting the need to improve PhilHealths electronic membership database. All three insurance schemes in Thailand (covering the formal private sector, civil servants, and the rest of the population) provide a comprehensive benet package with almost no co-payment. Out-of-pocket payments have decreased from 33% of total health expenditure in 2001 before universal coverage, to 177% in 2008,54 and the reduced incidence and intensity of catastrophic payment has especially beneted the poor population.55,56 With universal coverage, Thailand implemented a purchaser-provider split and mandated that people choose a local primary care unit at which to register, with their costs covered through capitation and case-based payment. There is evidence that health-care providers are improving their responsiveness to patients.57 Malaysia has retained the traditional Ministry of Health power of nancing and provision. The perceived absence of responsiveness of public providers has led to
870

the high level of out-of-pocket payments for private sector care, which is a major source of public concern.5860 Figure 2 depicts the association between insurance coverage and general government health expenditure as a percentage of total health expenditure, and the size of each sphere indicates the scal space for each country. Three country groups are apparent: a tax eort of more than 15% of the gross domestic product (Malaysia and Thailand), 10%15% (the Philippines, Indonesia, Vietnam, and Laos), and less than 10% (Cambodia). Longterm scal capacity to sustain the universal coverage in the Thailand scheme is a major policy challenge, especially given its large benet package. For Malaysia, public sector responsiveness needs to improve and a much greater proportion of funding needs to be channelled through prepayment arrangements.61,62 As an upper middle-income country, Malaysia has a high potential either to increase the general government health expenditure from 444% of total health spending (table 2) or to introduce payroll tax-nanced social health insurance, given the large proportion of the formal employed sector. In reducing out-of-pocket spending by households, it is dicult for the national government in a decentralised system, such as that in the Philippines and Indonesia, to mobilise political will and improve nancial commitment to the poor and vulnerable. The US$6 per year for a package of outpatient and inpatient services for poor individuals in Indonesia can cover only a very limited set of services, resulting in high levels of out-of-pocket expenditure, and the contributory premium of US$258 for the informal sector in the Philippines also provides only a small package and thus co-exists with high levels of out-ofpocket payment. General tax could be used to nance individual members in PhilHealth, although this approach is a major political decision as it departs from the current law. The government needs to broaden the tax base and
www.thelancet.com Vol 377 March 5, 2011

Series

diversify the sources of government non-tax revenue. However, improving the current low contribution to people in the informal sector is consistent with the policy direction of PhilHealth. A clear message emerges from the analysis of Vietnam; the government needs to increase scal space for health subsequent to consistent favourable economic performance to full its commitment towards universal coverage by 2014. With a contributory scheme for the informal sector, government subsidies might increase enrolment but those people hard to reach will not be covered, and at some point there will need to be consideration of a taxnanced scheme that pays the premiums of the poor and enrols them in the Vietnam social security scheme. This scheme would demand strong political leadership supported by scal capacity. Because scal space constraints limit coverage extension to the poor in Laos and Cambodia, resources from donors are inevitable. There are opportunities to harmonise and reorient funding from global health initiatives to strengthen health systems, in compliance with the Paris Declaration on aid eectiveness,63 in particular to strengthen primary health care. The Declaration calls for synergies of donor programmes in line with national priority and furnishes an opportunity to improve primary health care that is accessible to the rural poor population. Better access to quality care is one aspect of universal health coverage, and good-quality and accessible primary care services can contribute to improved service use by the poor.46 Improvement of the eectiveness of means testing is possible through active engagement by community members in identifying the poor, using approaches such as quantitative assessment of consumption levels and qualitative assessment to rank households by wealth.64 The experience in Cambodia exemplies the advantage of demand-side nancing by a health equity fund in improving the accountability of providers to the poor. Removing user charges without additional funding to subsidise health care for the poor might be harmful.65 Newly established social health insurance schemes should learn from the experiences of dierent provider payments regarding strengths and weaknesses of various payment models. PhilHealth not only provides limited nancial protection to its members, but also loses its potential monopsonistic purchasing power to steer healthcare providers to improve eciency. As the largest or only purchaser of medical services in the country, PhilHealth has an opportunity to exert its purchasing power to achieve eciency, such as introducing capitation and case-base payment system. The PhilHealth 2008 annual report stated that: PhilHealth must move away from fee for service towards provider payment schemes where it can easily leverage its purchasing power of more than 185 billion Pesos of health care purchases in 2008.66 Social health insurance in Laos, although mandatory, does not cover the full eligible population and eorts should be made to expand coverage. Cambodia has yet to
www.thelancet.com Vol 377 March 5, 2011

establish social health insurance to encompass the rapidly increasing formally employed sector. Although expanding the community-based health insurance can lead to adverse selection, this approach can be a temporary means for coverage extension to the informal sector, as indicated in Thailand. In general, coverage extension to the informal sector and the population outside formal schemes is divided, with contributory schemes leading one way and tax nancing another. The choice depends on political and health-system contexts. Well functioning contributory arrangements need an eective government and administrative capacities. When scal space is more favourable, the case in Thailand indicates that taxnanced arrangements are feasible. Although decisions on extending coverage to the various population groups can be made on pragmatic grounds, governments need to move towards harmonisation of benet packages, levels, and methods of provider payment across these schemes as members move from one scheme to another. Dierences between schemes within a country is also a major source of inequity. In a decentralised context, particularly in Indonesia and the Philippines, evidence is needed on the proper balance between national and local government nancing and roles in coverage extension. Financing reform is complex and necessitates contextspecic evidence; national institutional capacity to generate evidence and eective translation into policy decisions are vital.67,68 Regular assessment of cost drivers, long-term nancial projections, and capacity to generate and act on evidence about cost-eective interventions are needed. However, there is scope for countries to learn from each other. Partnership and collaborative work among co-authors in southeast Asia are strong foundations for further regional collaboration in the eorts towards better nancial risk protection and universal coverage in this region. As we have experienced and discussed, there are great opportunities to share experiences among countries in this region in the movement towards universal coverage for improved health care. Moreover, the challenges the governments face, including how to improve the responsiveness of public services, expand social health insurance, and identify and protect the poor, and whether coverage of the informal sector is better implemented through contributory arrangements or tax nance, are ones faced across the developing world. This paper is a timely contribution to the current global debates on how to provide nancial risk protection to the poor and vulnerable, how to extend coverage to the formal and informal sectors, and how to reach universal coverage, drawing on experiences and lessons from seven countries in southeast Asia with dierent paces of development. We have discussed the strengths and weaknesses of dierent designs of strategic purchasing and debated nancing sources for the informal sector. The experiences of each country indicate the diversity of country choices, related to political decisions, historical precedence, and social value.
871

Series

However, our focus on nancing reforms should not be interpreted to imply that additional supply factors, notably quality of care and human resources, are not also important elements of achieving universal coverage. In conclusion, governments hold responsibility to protect their citizens from catastrophic health expenditure and impoverishment, or welfare loss from inability to use health services when needed. Key messages emerge for resource-poor settingsrst, the extension of functioning and aordable primary health-care services is an initial priority for governments as geographical access to services is still a major problem. This extension needs to be matched with nancial risk protection for the poor, including eective identication of the poor, user fee exemption, and adequate levels of subsidy. Second, even though the formal sector might be small, social health insurance can make an important contribution to insurance coveragealthough general tax funding might be preferable in the long run.69 Finally, when the poor are adequately protected by tax-funded schemes, and in cases in which scal capacity allows, introducing partial subsidy for the informal sector can be an appropriate choice. These practical steps of reform should maintain a longterm objective of harmonising all prepayment or health insurance schemes with a universal and equal coverage.
Contributors VT was the lead author and was responsible for setting the conceptual framework of the paper, undertaking the literature search, verifying data, data interpretation, writing the manuscript, and ensuring full participation and contributions by country authors. WP helped to set the conceptual framework of the paper, and provided country data (Thailand), compiled data and data analysis for seven counties, produced tables and gures, and gave comments on the content of the draft. PI (Cambodia), SMA (Malaysia), AGM (Indonesia), KA (Laos), EB (Philippines), and DBH (Vietnam) provided and veried country data. HT (Indonesia) provided country data. PI, SMA, and AGM gave comments on the framework and on the general content of the draft. PI, SMA, KA, and EB helped to rewrite the text relevant to the country they provided data on. DBH gave comments relevant to Vietnam. AM was the scientic guarantor, responsible for setting the conceptual framework of the paper with VT, giving comments on the draft, redrafting the whole paper until reaching the nal version. Conicts of interest We declare that we have no conicts of interest. Acknowledgments This paper is part of a Series funded by the China Medical Board, Rockefeller Foundation, and Atlantic Philanthropies. We thank the China Medical Board and the Regional Steering Committee in convening various workshops. References 1 WHO. WHA58.33: Sustainable health nancing, universal coverage and social health insurance. http://apps.who.int/gb/ebwha/pdf_les/ WHA58/WHA58_33-en.pdf (accessed Feb 14, 2010). 2 Chongsuvivatwong V, Phua KH, Yap MT et al. Health and health-care systems in southeast Asia: diversity and transitions. Lancet 2011; published online Jan 25. DOI:10.1016/S01406736(10)61507-3. 3 Carrin G, Mathauer I, Xu K, Evans D. Universal coverage of health services: tailoring its implementation. Bull World Health Organ 2008; 86: 85763. 4 Garrett L, Chowdhury M, Pablos Mendez A. All for universal health coverage. Lancet 2009; 374: 129499. 5 ODonnell O, van Doorslaer E, Rannan-Eliya RP, et al. Who pays for health care in Asia? J Health Econ 2008; 27: 46075.

10 11 12 13

14

15 16

17

18

19 20

21

22

23

24

25

26

27

28

Xu K, Evans D, Kawabata K, Zeramdini R, Klavus J, Murray CJL. Household catastrophic health expenditure: a multicountry analysis. Lancet 2003; 362: 11117. Kawabata K, Xu K, Carrin G. Preventing impoverishment through protection against catastrophic health expenditure. Bull World Health Organ 2002; 80: 612. Xu K, Evans D, Carrin G, Aguilar A, Musgrove P, Evans T. Protecting households from catastrophic health spending. Health Aairs 2007; 26: 97283. Ranson MK. Reduction of catastrophic health care expenditures by a community-based health insurance scheme in Gujarat, India: current experiences and challenges. Bull World Health Organ 2002; 80: 61321. Yip W, Hsiao WC. Market Watch: the Chinese health system at a crossroads. Health Aairs 2008; 27: 46068. World Bank. World Development Indicators. http://data.worldbank. org/indicator (accessed Nov 18, 2010). World Bank. Poverty indicators. http://data.worldbank.org/indicator/ SI.POV.DDAY (accessed Nov 15, 2010). WHO. World Health Statistics 2010. Geneva: World Health Organization, 2010. http://www.who.int/whosis/whostat/ EN_WHS10_Full.pdf (accessed Nov 15, 2010). WHO. World Health Statistics 2008. Geneva: World Health Organization, 2008. http://www.who.int/whosis/whostat/ EN_WHS08_Full.pdf (accessed Nov 15, 2010). Van Doorslaer E, ODonnell O, Rannan-Eliya RP, et al. Catastrophic payment for health care in Asia. Health Economics 2007; 16: 115984. Lane C. Scaling up for better health in Cambodia. A case study for World Health Organization in follow-up to the High-Level Forum on the Health Millennium Development Goals. Geneva: World Health Organization, 2007. http://www.who.int/hdp/publications/ subhcambodia.pdf (accessed Nov 15, 2010). Taskforce on Innovative International Financing for Health Systems. Constraints to scaling up and costs. Working Group 1 Report. Geneva: World Health Organization, 2009. http://www. internationalhealthpartnership.net/pdf/IHP%20Update%2013/ Taskforce/Johansbourg/Working%20Group%201%20Report%20%20 Final.pdf (accessed Nov 15, 2010). PhilHealth. Conference on extending social health insurance to informal economy workers, 1820 October 2006, synthesis report January 2007. Manila, Philippines. PhilHealth, GTZ, ILO, WHO, and World Bank, 2007. http://www2.gtz.de/dokumente/bib/07-0986.pdf (accessed Nov 15, 2010). Jacobs B, Price N. The impact of the introduction of user fees at a district hospital in Cambodia. Health Policy Plan 2004; 19: 31021. Akashi H, Yamada T, Huot E, Kanal K, Sugimoto T. User fees at a public hospital in Cambodia: eects on hospital performance and provider attitudes. Soc Sci Med 2004; 58: 55364. Bureau of Health Economics and Financing. Annual Health Financing Report 2008. Phnom Penh, Cambodia: Ministry of Health, 2009. Hardeman W, Van Damme W, Van Pelt M, Por I, Kimvan H, Meessen B. Access to health care for all? User fees plus a Health Equity Fund in Sotnikum, Cambodia. Health Policy Plan 2004; 19: 2232. Jacobs B, Price N. Improving access for the poorest to public sector health services: insights from Kirivong Operational Health District in Cambodia. Health Policy Plan 2006; 21: 2739. Noirhomme M, Meessen B, Griths F, et al. Improving access to hospital care for the poor: comparative analysis of four health equity funds in Cambodia. Health Policy Plan 2007; 22: 24662. Ir P, Bigdeli M, Meessen B, Van Damme W. Translating knowledge into policy and action to promote health equity: the health equity fund policy process in Cambodia 20002008. Health Policy 2010; 96: 20009. WHO. Essential Medicine Monitor 2010. http://apps.who.int/ medicinedocs/documents/s16784e/s16784e.pdf (accessed Nov 14, 2010). Patcharanarumol W, Mills A, Tangcharoensathien V. Dealing with the cost of illness: the experience of four villages in Lao PDR. J Int Dev 2009; 21: 21230. WHO. Health-nancing strategy for the Asia Pacic Region (20102015). Manila, WHO Western Pacic Region 2009. http://www.wpro.who.int/internet/resources.ashx/HCF/ HCF+strategy+2010-2015.pdf (accessed Nov, 14, 2010).

872

www.thelancet.com Vol 377 March 5, 2011

Series

29

30

31

32

33 34

35

36

37 38

39

40

41

42

43 44

45

46

47 48

49

Waters H, Saadah F, Pradhan M. The impact of the 199798 East Asian economic crisis on health and health care in Indonesia. Health Policy Plan 2003; 18: 17281. Rokx C, Scheiber G, Harimurti P, Tandon A, Somanathan A. Health nancing in Indonesia: a reform road map. Washington DC, USA: World Bank, 2009. http://siteresources.worldbank.org/intindonesia/ Resources/Publication/280016-1235115695188/58471791243406029796/Full.Report.en.pdf (accessed Nov 15, 2010). Nitayarumphong S, Pannarunothai S. Achieving universal coverage of health care through health insurance: the Thai situation. In: Nitayarumphong S, Mills A, eds. Achieving universal coverage of health care. Bangkok: Oce of Health Care Reform, Thailand Ministry of Public Health, 1998. Gilson L, Russell S, Ruyajin O, Boonchote T, Pasandhanathorn V. Exempting the poor: a review and evaluation of the low income card scheme in Thailand. PHP Publication No 30. London: London School of Hygiene and Tropical Medicine, 1998. Pannarumothai S, Mills A. The poor pay more: health-related inequality in Thailand. Soc Sci Med 1997; 44: 178190. Langenbrunner JC, Cashin C, O Dougherty S. Designing and implementing health care provider payment systems how-to manuals (eds). Washington DC, USA: The International Bank for Reconstruction and Development/The World Bank, 2009. http://siteresources.worldbank.org/healthnutritionandpopulation/ resources/peer-reviewed-publications/providerpaymenthowto.pdf (accessed Nov 15, 2010). National Statistical Coordinating Board. Philippine National Health Accounts, 2005. http://www.nscb.gov.ph/stats/pnha/2005/ sources.asp (accessed Nov 15, 2010). National Statistical Coordinating Board. Philippine National Health Accounts, 2007. http://www.nscb.gov.ph/stats/pnha/2007/ 2007pnhatables.asp (accessed Nov 15, 2010). National Statistical Coordinating Board. 2008 Philippine Statistical Yearbook. Manila, Philippines: NSCB, 2008. WHO. National Health Accounts, Estimates for the Philippines, 2009. http://www.who.int/nha/country/phl/en/ (accessed Jan 26, 2009). Department of Health. Health sector reform agenda monograph no 9. Bridging to future reforms. Manila, Philippines: Department of Health, 2010. World Bank. Philippines: fostering more inclusive growth. Report No 49482-PH. Washington DC, USA: World Bank, 2010. http://www.worldbank.org.ph/wbsite/external/countries/ eastasiapacicext/philippinesextn/0,,contentMDK:22673984~menu PK:3968175~pagePK:64027988~piPK:64027986~theSitePK:332982,00. html (accessed Nov 15, 2010). Sharifa Ezzat WP, Azimatun NA, Jasmin NM, Aljunid S. Waiting times for elective orthopaedic surgeries in a teaching hospital and their inuencing factors. Medicine Health 2009; 4: 5360. Yon R. Health care nancing in Malaysia: future trends. In: Aljunid SM, Mohsein NAA, eds. Health economics issues in Malaysia. University of Malaya Press: Kuala Lumpur, 2002: 87112. Yon R, Hamidy MA, Lin CY. Evaluation of the seventh Malaysia plan: a new approach. Asia Pac J Public Health 2001; 13: 5458. Tangcharoensathien V, Supachutikul A, Lertiendumrong J. The social security scheme in Thailand: what lessons can be drawn? Soc Sci Med 1999; 48: 91323. Mills A, Bennett S, Siriwanarangsun P, Tangcharoensathien V. The response of providers to capitation payment: a case-study from Thailand. Health Policy 2000; 51: 16380. Prakongsai P, Limwattananon S, Tangcharoensathien V. The equity impact of the universal coverage policy: lessons from Thailand. In: Chernichovsky D, Hanson K, eds. Innovations in health system nance in developing and transitional economies. London: The Emerald Group Publishing Limited, 2009: 5781. Kwon S. Payment system reform for health care providers in Korea. Health Policy Plan 2003; 18: 8492. Wagsta A. Health insurance for the poor: initial impacts of Vietnams health care fund for the poor. Impact evaluation series No 11 policy; research working paper No WPS 4134. The World Bank, 2007. Tangcharoensathien V. Community nancing: the urban health card in Chiangmai, Thailand. PhD thesis, University of London, London School of Hygiene and Tropical Medicine, 1990.

50

51

52

53

54

55

56

57

58 59

60 61

62

63

64

65

66

67

68

69

Tangcharoensathien V, Teokul W, Chanwongpaisarn L. Challenges of implementing universal health care in Thailand. In: Kwon HJ, ed. Transforming the developmental welfare state in east Asia. United Nations Research Institute for Social Development. Palgrave Macmillan, 2005. Pannarunothai S, Srithamrongsawat S, Kongpan M, Thumvanna P. Financing reforms for the Thai health card scheme. Health Policy Plan 2000; 15: 30311. Tangcharoensathien V, Prakongsai P, Limwattananon S, Patcharanarumol W, Jongudomsuk P. From targeting to universality: lessons from the health system in Thailand. In: Townsend P, ed. Building decent societies: rethinking the role of social security in development. Houndmills, Basingstoke, Hampshire: Palgrave Macmillan, 2009: 31022. National Statistics Oce. The 2008 National Demographic and Health Survey results. Manila, Philippines: NSO, 2010. http:// philippines.usaid.gov/sites/default/les/resources/key_documents/ NDHS_2008.pdf (accessed Nov 15, 2010). Thai working group on National Health Account. The report on 19942008 National Health Expenditure. Nonthaburi, International Health Policy Program, Ministry of Public Health, 2010. http://www. ihppthaigov.net/nha/thai_nha_1994-2008.xls (accessed Feb 20, 2010). Somkotra T, Lagrada L. Payments for health care and its eect on catastrophe and impoverishment: experience from the transition to universal coverage in Thailand. Soc Sci Med 2008; 67: 202735. Limwattananon S, Tangcharoensathien V, Prakongsai P. Catastrophic and poverty impacts of health payments: results from national household surveys in Thailand. Bull World Health Organ 2007; 85: 60006. International Health Policy Program. Eectiveness of public health insurance schemes on nancial risk protection: the assessments of purchasers capacities, contractors responses and impact on patients. Research report, Consortium on Research into Equitable Health Systems. Nonthaburi, Thailand: Ministry of Public Health, 2010. Yu CP, Whynes DK, Sach TH. Equity in healthcare nancing: the case of Malaysia. Int J Equity Health 2008; 7: 114. Yu CP, Whynes DK, Sach TH. Assessing progressivity of out-of-pocket payment: with illustration to Malaysia. Int J Health Plann Manage 2006; 21: 193210. Chee HL. Ownership, control, and contention: challenge for the future of healthcare in Malaysia. Soc Sci Med 2008; 66: 214556. Nor Hayati I, Azimatun NA, Rozita H, Sharifa Ezat WP, Rizal AM. In-patients satisfaction in the medical and surgical wards a comparison between accredited and non accredited hospital in the state of Selangor. J Community Health 2010; 16: 6068. Sharifa Ezat WP, Jamsiah M, Aniza I, Suryati AA. Client satisfaction and relationship with ISO certication status in Negeri Sembilans health clinics. J Community Health 2008; 13: 1121. Organisation for Economic Co-operation and Development. The Paris Declaration and Accra Agenda for Action. http://www.oecd.org/docu ment/18/0,3343,en_2649_3236398_35401554_1_1_1_1,00.html (accessed Nov 14, 2010). Patcharanarumol W. Health care nancing for the poor in Lao PDR. PhD thesis, University of London, London School of Hygiene and Tropical Medicine, 2008. James CD, Hanson K, McPake B, et al. To retain or remove user fees?: reections on the current debate in low- and middle-income countries. Appl Health Econ Health Policy 2006; 5: 13753. Philippines Health Insurance Corporation. PhilHealth Board resolution 1,113 April 30, 2008 and 2008 PhilHealth Annual Report. http://www.philhealth.gov.ph/about_us/annual_report/ar2008.pdf (accessed Nov 15, 2010). Tangcharoensathien V, Wibulpolprasert S, Nitayarampong S. Knowledge-based changes to health systems: the Thai experience in policy development. Bull World Health Organ 2004; 82: 75056. Green A. Reforming the health sector in Thailand: the role of the policy actors on the policy stage. Int J Health Plann Manage 2000; 15: 3959. Wagsta A. Social health insurance reexamined. World Bank policy research working paper 4111, January 2007. http://www-wds. worldbank.org/servlet/WDSContentServer/WDSP/IB/2007/01/09/ 000016406_20070109161148/Rendered/PDF/wps4111.pdf (accessed Nov 15, 2010).

www.thelancet.com Vol 377 March 5, 2011

873

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.