Out-Of-Pocket Spending on Maternal And

Home , Health in Pakistan

Out-of-Pocket Spending on Maternal and Child Health in Asia and the Pacific The Impact of Out-of-Pocket Expenditures on Families and Barriers to Use of Health Services in Pakistan

The Impact of Out-of-Pocket Expenditures on Families and Barriers to Use of Health Services in Pakistan

Evidence from the Pakistan Social and Living Standards Measurement Surveys 2005–2007 COUNTRY BRIEF

1 Summary

• Pakistan is lagging on progress toward Millennium Development Goals 4 and 5. Health services coverage is poor, and overall use of medical services in Pakistan is low compared to other countries in the region. The low rates of utilization are driven by a low sensitivity and recognition of illness symptoms. • Overall healthcare use is not unequal, but poor Pakistanis are certainly sicker, so there is considerable inequity in access to and use of services. • When healthcare is obtained, private providers are the dominant source of treatment, with government healthcare services only accounting for one-fifth of utilization. • The main reasons for not using government facilities, at all income levels, are lack of access and distance to government facilities, followed by lack or inadequate supply of medicines. • Use of government services is pro-poor, although still at a low level, owing to greater use of private services by the nonpoor. • Households spend 4.9% of all of their expenditures on medical care, more than in almost all other Asian countries. This places heavy financial burdens on families. The incidence of catastrophic and impoverishing medical expenditures in Pakistan is much higher than in other countries in the region, and the burden of out-of-pocket spending is greater for the poor than the nonpoor. • To increase access to services, Pakistan needs to increase public financing and risk pooling in an effective manner to reduce the financial burdens faced by households. • The sources of these analyses were the Pakistan Social and Living Standards Measurement Surveys. Compared to other Asian countries, their scope is quite limited, and more comprehensive measures of healthcare use should be adopted in future surveys.

Background significant financial barriers in accessing needed maternal and child healthcare services. Nonetheless, the Government Pakistan’s maternal and child health indicators are among the of Pakistan is committed to reaching the Millennium worst in the Asia-Pacific region and in the world. An estimated Development Goals and to improving the access of its people 260 women die for every 100,000 live births (WHO et al. to adequate healthcare services. There is recognition of 2010). Nearly one in 10 children die before their fifth birthday the negative impact of financial barriers on access to care, (UNICEF 2012), and the Pakistan Demographic and Health reflected in a number of analytical activities supported by the Survey 2007 showed little change in mortality over time Asian Development Bank and other development partners. (National Institute of Population Studies, 2008). More than 65% of women in Pakistan deliver their babies at home, and less than 2 in 5 women deliver with a skilled birth attendant. Data Sources

This poor performance is linked with low levels of government This policy brief presents findings from the Pakistan Social investment in health. According to National Health Accounts and Living Standards Measurement Survey (PSLSMS) 2005– estimates (Sekhar et al. 2009), overall health spending in 06, and the Core Welfare Indicators Questionnaire (CWIQ) Pakistan is low in comparison with regional countries: 2.6% Survey 2006–07 (Federal Bureau of Statistics 2006 and of gross domestic product in fiscal year 2006, compared with 2007). The PSLSMS 2005–06 is a nationally representative 4.1% in India, 3.4% in Bangladesh, and 4.3% in Sri Lanka. household survey of 15,453 households, while the CWIQ Of this, the largest share, 65%, is contributed by private 2006–07 survey sample include 73,953 households. financing, most of which (99%) is household out-of-pocket spending. The CWIQ collects basic data on healthcare use but lacks a detailed household expenditure module. For analysis of Private providers thus play a major role in providing care. In healthcare use inequalities for this brief, the survey population addition to government hospitals and health centers, there in the CWIQ was grouped into equal quintiles of relative is a diverse range of private hospitals, private clinics, and living standards, using a wealth index, estimated using well- other private providers from which Pakistanis obtain medical established methods (Deon and Pritchett 2001). The CWIQ treatment. Government facilities charge user fees for many does not collect information on health expenditures, so medical services, and informal payments to access care are the PSLSMS 2005–06 was also used, which had a detailed widely prevalent in the public sector, so the poor can face household expenditure module, for analysis of out-of-pocket

2 The Impact of Out-of-Pocket Expenditures on Families and Barriers to Use of Health Services in Pakistan spending. The PSLSMS was conducted in a subsample of the Pakistan Demographic and Health Survey 2006–07 (National CWIQ sample, but is also nationally representative. Institute of Population Studies 2008), show that maternal and child health are much worse in the poor and in rural areas. This The PSLSMS questionnaire is not as extensive as those found in most suggests that one driver of worse health outcomes in the poor Asia-Pacific countries, which limits the analysis that can be done is inadequate use of healthcare by poorer families, owing to a of healthcare inequalities. A more detailed and comprehensive reduced ability to recognize illness when sick. health module was piloted in the PSLSMS 2009, but data from this were not available for analysis. Routinizing the use of this Figure 2: Illness Reporting in Pakistan, by Socioeconomic more detailed health module would permit more systematic and Status, 2006–2007 detailed analysis of healthcare inequalities in Pakistan. 14

Perception of Illness and Treatment Seeking 12

A key driver of whether ill individuals seek healthcare is whether 10 they recognize themselves as being sick. The CWIQ 2006–07 asked whether individuals were ill in the 30 days preceding the 8 survey. In total, 6.3% of all individuals, and 11.6% of children aged less than 5 years, were reported to have been ill. 6

However, the self-reporting of sickness in a survey is an 4 unreliable indicator of the real level or distribution of illness 2 within the population, as it is critically dependent on the ability % reporting sickness any in past 30 days of individuals to recognize illness symptoms as sickness. This is apparent when the levels of reported sickness in the past 30 0 days in similar surveys from other Asian countries are compared Poorest Q2 Q3 Q4 Richest with similar levels of health status to Pakistan. Overall, the levels All Children Quintile of reported sickness in both children and adults in Pakistan Q = quintile are much lower than in most other countries in the region, Source: Authors’ analysis of CWIQ 2007 data set. indicating that Pakistani families are less likely to recognize or respond to signs of illness than in other countries. The situation in Pakistan can be contrasted with patterns in other countries, such as those shown in Figure 1, where rates Rates of reported illness are relatively equal by income level, and of reported sickness are generally higher in the nonpoor than by urban or rural location (Figure 2). Other data, such as the the poor, and in urban populations. This pattern is probably because in most countries, health awareness initially increases Figure 1: Illness Reporting in Pakistan and Other Asian in the better-off and more educated households, and then Countries, Recent Years spreads to the poorer and less-educated households. This 45 trend is not so apparent in Pakistan, but can be seen to a limited extent, since the rates of reported sickness were lowest in the 40 least developed province of Balochistan (4.9% in adults and 35 9.7% in children). The problem of greater underreporting in Balochistan is also mentioned in the Demographic and Health 30 Survey 2006–07, which notes that the mortality estimates for 25 that province may be correspondingly underestimated.

20 Being sick does not automatically lead to seeking medical care; in many countries, the poor who are sick are much 15 less likely to obtain treatment than the rich. In Pakistan, such 10 a steep gradient is not seen, but poor Pakistanis and their children are slightly less likely to be taken for treatment when % reporting sickness any in past 30 days 5 sick than those in higher income groups (95% of sick children 0 in the poorest quintile versus 97% in the richest quintile). Pakistan Lao-PDR Cambodia PNG Bangladesh Timor-Leste Slightly lower rates are also seen in rural areas (94% of all sick

All Children individuals and 96% of sick children) compared with urban areas (96% of all sick individuals and 98% of sick children). Lao PDR = Lao People’s Democratic Republic, PNG = Papua New Guinea Sources: Authors’ analysis of PSLSMS 2006 and CWIQ 2007 data sets, and analyses of The overall rates of treatment seeking when reported sick are Asian Development Bank technical assistance project. actually high compared to other countries, suggesting that

3 Pakistanis tend to identify themselves as sick only when illness Pakistanis obtain most of their medical care from private is so serious that they cannot avoid seeking treatment. providers (Figure 5), with similar patterns in children (Figure 6). Public providers accounted for only 20.9% of total healthcare Excluding cases where illness was not considered serious use, and 18.2% of healthcare use for children. Most medical enough to require treatment, the two leading reasons why care is obtained from private hospitals and dispensaries sick persons did not seek treatment are that medical treatment (72.4%). Use of traditional or homeopathic providers is not would be too expensive (52% overall and 56% for children), significant (3.3%), but is greater among the poor than the followed by health facilities being too distant (14% overall and nonpoor (6% of all healthcare use in the poorest quintile versus 15% for children). Other reasons related to service quality 3% in the richest quintile), and less in children than adults. or lack of staff or medicines are quite small. The high cost of treatment and distance to facilities are significantly more Private providers are preferred largely because they are closer important for the poor and those living in rural areas, while than public facilities. The PLSMS 2005–06 asked why children quality-related factors are only important for the richest 40% of with diarrhea were not taken to the nearest public provider households (Figure 3).1 Distance is also much more important as well as the reasons for selecting a private provider. Lack in Balochistan (38%) than in other provinces. of a nearby government facility was the reason for not using a government provider in 46% of cases, followed by lack of Figure 3: Barriers to Treatment of Illness in Pakistan, medicines or poor medicine quality in 21% (Figure 7). Similar 2006–2007 reasons were reported for selecting a private provider. 100 14 Figure 4: Use of Medical Treatment in Pakistan, by Age 90 23 3 28 Group, 2006–2007 80 (%) 43 18.0 50 52 51 4 70 1 25 16.0 11 18 60 3 14.0 50 7 5 2 3 12.0 40 6 13 60 6 10.0 30 57 55 43 8.0 20 47 39 Reason for notReason taking treatment 30 10 6.0

0 4.0 Poorest Q2 Q3 Q4 Richest Urban Rural % obtaining medical care in past 30 days% obtaining medical care 2.0 Quintile Sector Other No confidence in service 0.0 Health facility too far Health care too expensive 0 1–4 5–9 10–17 18–44 45–64 65+ Q = quintile Age group (years) Source: Authors’ analysis of CWIQ 2007 data set. All providers Public providers

Source: Authors’ analysis of CWIQ 2007 data set. These results indicate that financial barriers and lack of physical access are major factors behind inadequate and However, in contrast to most developing countries in Asia, the unequal use of healthcare for sick children in Pakistan. poor make more use of public facilities than the nonpoor. However, underlying this is a low responsiveness to illness. The distribution of utilization of public services both overall (Figure 8) and by children (Figure 9) is actually pro-poor, with the poorest 40% of the population accounting for 47% Utilization of Healthcare Providers of all public sector visits. However, the overall distribution of government healthcare expenditures cannot be assessed Overall healthcare utilization varies by age, with a higher from this, since the survey data do not permit analysis of level for infants and young children than for young adults, whether more expensive, higher-level government facilities and then increasing for older adults (Figure 4). In 2006– are disproportionately used by the nonpoor. 2007, infants accounted for 5.3% of all healthcare visits, and children (i.e., less than 5 years of age) 18.6%. Infants accounted for 16.3% and children 18.7% of all visits to Out-of-Pocket Spending on Healthcare public providers (e.g., public hospitals, dispensaries, rural health centers, basic health units, and lady health workers). Out-of-pocket payments not only deter households from seeking care, they can also cause considerable hardship 1 Other reasons given included no confidence in service, no doctor or female staff members available, staff members not helpful, lack of cleanliness, and lack of medicines. 4 The Impact of Out-of-Pocket Expenditures on Families and Barriers to Use of Health Services in Pakistan

Figure 5: Overall Healthcare Use in Pakistan, by Figure 7: Reasons for Not Using Public Providers When Socioeconomic Status, 2006–2007 Children in Pakistan Are Ill with Diarrhea, 2005–2006 1 2 1 100 11 Reasons for not using a government 9% 5 4 4 3 6 4 provider (%) 23% 90 5 8 5 No government facility 12% 80 Too far away Medicines ineffective or not available 10% 70 Staff unhelpful 22% 60 Lack of doctors or female staff 65 64 69 70 Other 76 24% 50 Source: Authors’ analysis of PSLSMS 2006 data set. 40 Figure 8: Use of Public Healthcare Services in Pakistan, 30 by Socioeconomic Status, 2006–2007 100 20 4 3 2 2 9 1 15 18

% of individuals seeking care by provider by % of individuals seeking care 90 10 20 19 18 19 15 15 80 10 0 20 Poorest Q2 Q3 Q4 Richest 70 17 Quintile 20 60 Other Traditional or homeopathic providers 20 Pharmacies Private hospitals or dispensaries 50 RHCs, BHUs, LHWs and other public Public hospitals or dispensaries 27 16 40 Q = quintile 23 RHC= Rural Health Centre, BHU = Basic Health Unit, LHW = Lady Health Worker % of total utilization 30 Source: Authors’ analysis of CWIQ 2007 data set. 36 20 32 Figure 6: Children’s Use of Healthcare Facilities in 22 10 Pakistan, by Socioeconomic Status, 2006–2007 100 11 1 0 433 2 2 Public hospitals or RHC or BHUs LHV or LHWs 4 5 4 90 8 8 dispensaries Richest Q4 Q3 80 Q2 Poorest Q = quintile 70 RHC= Rural Health Centre, BHU = Basic Health Unit, LHV = Lady Health Visitor, LHW = Lady Health Worker 60 Sources: Authors’ analysis of CWIQ 2007 data set. 69 74 79 69 75 50 Figure 9: Children’s Use of Public Healthcare Services in Pakistan, by Socioeconomic Status, 2006–2007 40 100 30 12 10 90 24 20 6 10 4 3 80 15 2 3 % of individuals seeking care by provider by % of individuals seeking care 10 17 15 10 14 15 13 11 70 0 20 11 Poorest Q2 Q3 Q4 Richest 60

Quintile 50 30 10

Other Traditional or homeopathic providers 40 24 Pharmacies Private hospitals or dispensaries % of total utilization RHCs, BHUs, LHWs and other public Public hospitals or dispensaries 30 Q = quintile 44 20 35 RHC= Rural Health Centre, BHU = Basic Health Unit, LHW = Lady Health Worker 28 Source: Authors’ analysis of CWIQ 2007 data set. 10 and financial impoverishment, especially among the poor. 0 Public hospitals or RHC or BHUs LHV or LHWs Pakistan’s healthcare system relies predominantly on out-of- dispensaries pocket financing. Richest Q2 Q3 Q4 Poorest The PSLSMS 2005–06 allows further examination of the Q = quintile RHC= Rural Health Centre, BHU = Basic Health Unit, LHV = Lady Health patterns and distribution of out-of-pocket household Visitor, LHW = Lady Health Worker healthcare spending. However, as it does not ask for spending Sources: Authors’ analysis of CWIQ 2007 data set.

5 Figure 11: Out-of-Pocket Medical Expenditure per Capital per in relation to individuals in the household, it is not possible Year in Pakistan, by Socioeconomic Status, 2005–2006 to analyze expenditures by gender or by specific age group. 10,000 Moreover, spending is only categorized into two categories of 9,000 medicines/supplies and other costs, so detailed analysis of the components of spending is also not possible. 8,000 7,000 According to the PSLSMS 2005–06, annual out-of-pocket 71 spending on medical care amounted to PRs5,486 ($91) per 6,000 capita, equivalent to 4.9% of total household expenditures. 5,000 This is one of the highest shares of household spending in 4,000 64 the region (Figure 10), and should be expected to result in 58 Pakistan Rupees (%) Rupees Pakistan 3,000 significant financial burdens for many households. Spending 59 on medicines and supplies accounted for 37% of this. 2,000 52 Limitations in the questionnaire prevent estimation of shares 29 1,000 42 36 of spending related to maternal or child healthcare. 48 41 0 Figure 10: Share of Out-of-Pocket Medical Poorest Q2 Q3 Q4 Richest Expenditure in Household Budgets in Regional Quintile Countries, Recent Years Other items Medicine Q = quintile India Note: Figures indicate share of medicines and supplies in total healthcare Cambodia out-of-pocket spending. Source: Authors’ analysis of PSLSMS 2006 data set. Viet Nam Pakistan Bangladesh Figure 12: Out-of-Pocket Medical Expenditure per Capital PRC per Year in Pakistan, by Province and Sector, 2005–2006 Nepal 9,000 Kyrgyz Republic Sri Lanka 8,000 Philippines Indonesia 7,000 Lao PDR Thailand 6,000 Malaysia 68 Maldives 5,000 PNG 4,000 Fiji 60 Timor-Leste 3,000 67 53 65 Pakistan Rupees (%) Rupees Pakistan 0.0 1.0 2.0 3.0 4.0 5.0 6.0 2,000 49 Out-of-pocket spending as a % of total household expenditure 32 47 40 35 1,000 33 51 Lao PDR = Lao People’s Democratic Republic, PNG = Papua New Guinea, PRC = People’s Republic of China 0 Sources: Authors’ analysis of PSLSMS 2006 data set and analyses of Asian Development Punjab Sindh NWFP Balochistan Urban Rural Bank technical assistance project. Province Sector

Other items Medicine There are large disparities in out-of-pocket spending NWFP = Northwest Frontier Province. by income level (Figure 11). Individuals in the richest Source: Authors’ analysis of PSLSMS 2006 data set. quintile spend 4 times more overall than those in the poorest quintile and account for almost four-tenths of all Figure 13: Out-of-Pocket Health Expenditure in Pakistan, out-of-pocket spending. The higher spending among the by Socioeconomic Status, 2005–2006 nonpoor is accounted mostly by much higher spending on doctors and hospital fees, with spending on medicines 10% not skewed. Spending is also higher in Punjab and the Poorest Northwest Frontier Province, but urban–rural disparities 14% Q2 36% are small (Figure 12). Q3 Q4 Most of the out-of-pocket spending is by the richest two Richest 17% quintiles. The poorest quintile of Pakistanis only accounts 23% for 10% of total out-of-pocket medical expenditure, presumably reflecting their lower incomes and ability to Q = quintile pay (Figure 13). Source: Authors’ analysis of PSLSMS 2006 data set.

6 The Impact of Out-of-Pocket Expenditures on Families and Barriers to Use of Health Services in Pakistan

Financial Impacts of Out-of-Pocket Figure 15: Incidence of Catastrophic Out-of-Pocket Expenditures Medical Spending in Regional Countries, Recent Years Viet Nam Although richer households spend far more on medical treatment PRC than poor ones, the burden of out-of-pocket expenditure is India much greater for poor households than rich ones. Healthcare Pakistan spending as a share of household consumption is higher in Nepal poorer households, and as a share of nonfood consumption is Kyrgyz Republic even greater. For example, the poorest quintile devotes 8.3% Philippines of their nonfood expenditures to medical treatment, versus Indonesia 5.9% in the richest quintile (Figure 14). Bangladesh Thailand There are two broad measures that can be used to assess the Sri Lanka financial impact on households of out-of-pocket expenditure. Malaysia One is to measure how many households are pushed Maldives below the poverty line by such spending (i.e., impoverishing Fiji impacts), and the second is how many households have to Timor-Leste devote a large share of their resources for medical treatment 4.02.00.0 6.0 8.0 10.0 12.0 14.0 16.0 expenses (i.e., catastrophic impacts). Previous studies in Asia % of population spending more than 10% of household budget on health have shown that heavy reliance on out-of-pocket spending in Lao PDR = Lao People’s Democratic Republic, PNG = Papua New Guinea, health systems results in high levels of medical impoverishment PRC = People’s Republic of China and catastrophic expenditures (van Doorslaer et al. 2006, Source: Authors’ analysis of PSLSMS 2006 data set, van Doorslaer et al. 2007, and forthcoming estimates by Equitap research network. van Doorslaer et al. 2007). The PSLSMS 2005–06 reveals high levels of both impoverishing and catastrophic impacts from out-of-pocket expenditures on health in Pakistan. Correspondingly, the frequency of catastrophic health expenditures is also high whatever definition is used. In 2005–2006, in Figure 14: Share of Out-of-Pocket Medical Spending any given month, 10.3% of Pakistani families had to allocate in Household Budgets and Nonfood Expenditure by more than 10% of their total household budget, and 1% had to Socioeconomic Status in Pakistan, 2005–2006 allocate more than 40% of their monthly nonfood expenditures to 9 medical treatment costs. These rates of catastrophic expenditures 8 are among the highest in the region (Figure 15).

6 Conclusions

5 The PSLSMS 2005–06 and CWIQ 2006–07 suggest that overall healthcare use in Pakistan is low due to a low 4 responsiveness to illness symptoms. Increasing health 3 awareness and changing health behaviors would be a key part of improving overall health outcomes. Further, the and a % nonfood expenditure 2 overall inequality in healthcare use is not great between 1 rich and poor, but because poor Pakistanis are in worse health, there exists considerable inequity in access and use Out-of-pocket health spending as a % of total HHE 0 of services. Poorest Q2 Q3 Q4 Richest Quintile When healthcare is obtained, private providers are the dominant Health expenditure as a % of total household expenditure source of treatment, with government healthcare services only Health expenditure as a % of household nonfood expenditure accounting for one-fifth of overall use. The main reasons for HHE = household expenditure, Q = quintile Source: Authors’ analysis of PSLSMS 2006 data set. not using government facilities, including for the poor, are lack of access and distance to government facilities, followed Overall, in any given month in 2005–2006, 4.7% of the by lack or inadequate supply of medicines. However, overall population, or 7.5 million Pakistanis, were pushed below use of government services appears to be pro-poor, although the $1 international poverty line2 as a result of their still at a low level. Given this pattern of use, if government household’s medical spending. This is a high level of medical provision is to improve coverage of essential maternal and impoverishment by regional standards, comparable with child health services, priorities might be to expand the facility 3.7% in India, 3.1% in Bangladesh, and 0.2% in Timor-Leste. network in rural areas and to improve the supply of medicines within the government system. 2 The international $1 poverty line is equivalent to a consumption level of 1.08 international (1993 purchase power parity) dollars per day. In 2005, this would have been equal to PRs27 per day. 7 Households spend 4.9% of their budgets on medical care, which is National Institute of Population Studies (NIPS) and Macro International. a higher level than in almost all other Asian countries. The reliance 2008. Pakistan Demographic and Health Survey 2006–07. Islamabad. on out-of-pocket financing imposes heavy burdens on families. The incidence of catastrophic and impoverishing medical expenditures Sekhar, B., I. Bhushan, M. Rani, and I. Anderson. 2009. Incidence and Correlates of ’Catastrophic’ Maternal Health Care Expenditure in India. in Pakistan is high compared to other regional countries, although Health Policy and Planning. 24 (6). pp. 445–456. comparable to India, and the burden of out-of-pocket spending is greater for the poor than the nonpoor. It is likely that use of maternal United Nations Children’s Fund (UNICEF). 2012. The State of the and child health services in Pakistan frequently impoverishes families World’s Children: Children in an Urban World. New York. and acts as a disincentive to greater use of medical services, as it does in India (Sekhar et al. 2009). Consequently, both lack of van Doorslaer, E., et al. 2007. Catastrophic Payments for Health Care in Asia. Health Economics. 16 (11). pp. 1,159–1,184. physical access to nearby healthcare facilities and the financial costs of treatment need to be addressed to improve utilization of van Doorslaer, E., et al. 2006. Effect of Payments for Health Care on Poverty maternal and child health services. Estimates in 11 Countries in Asia: An Analysis of Household Survey Data. Lancet. 368 (9,544). pp. 1,357–1,364.

References World Health Organization (WHO) et al. 2010. Trends in Maternal Mortality: 1990 to 2008. Geneva. Deon, F. and L. Pritchett. 2001. Estimating Wealth Effects without Expenditure Data—or Tears: An Application to Educational Enrollments in States of India. Demography. 38 (1). pp. 115–132. Suggested Citation Federal Bureau of Statistics. 2006. Pakistan Social and Living Standards Measurement Survey 2005–2006. Islamabad. Chandrasiri, J., C. Anuranga, R. Wickramasinghe, and R.P. Rannan- Eliya. 2012. The Impact of Out-of-Pocket Expenditures On Families and Federal Bureau of Statistics. 2007. Core Welfare Indicators Question- Barriers to Use of Health Services In Pakistan: Evidence from the Pakistan naire (CWIQ) Survey, Pakistan Social and Living Standards Measure- Social and Living Standards Measurement Surveys 2005–07 - RETA– ment Survey 2006–2007. Islamabad. 6515 Country Brief. Manila: Asian Development Bank.

ADB RETA 6515 Country Brief Series Poor maternal, neonatal, and child health adversely affects women, families, and economies across the Asia and Pacific region. This burden of illness must be reduced if the Millennium Development Goals (particularly 4 [reduce child mortality] and 5 [improve maternal health]) are to be achieved and improvements made in the health and economic well-being of households and nations. Progress in this regard will require an increased supply of effective healthcare services, as well as demand for such services. This series of country briefs provides evidence from national household surveys on the financial burdens imposed on the poor by private expenditures on public and private healthcare services. Countries can use this information in building awareness within health systems and policy bodies of financial constraints on healthcare, and in designing demand-side interventions to increase the use of maternal, neonatal, and child health services. Summaries of the analysis of household data from Bangladesh, Cambodia, the Lao People’s Democratic Republic, Pakistan, Papua New Guinea, and Timor-Leste, and a summary overview, are included in the series. This country brief was prepared by the Institute for Health Policy in Sri Lanka under an Asian Development Bank (ADB) technical assistance project, Impact of Maternal and Child Health Private Expenditure on Poverty and Inequity (TA–6515 REG). The Institute for Health Policy and authors gratefully acknowledge the funding made possible by ADB that was financed principally by the Government of Australia. Australia is taking a leading role in global and regional action to address maternal and child health. A key part of this is to strengthen the evidence for increased financial support and the most effective investments that governments and donors can make to meet Millennium Development Goals 4 and 5. Australia supported this technical assistance project as a part of this commitment.

About the Asian Development Bank ADB’s vision is an Asia and Pacific region free of poverty. Its mission is to help its developing member countries reduce poverty and improve the quality of life of their people. Despite the region’s many successes, it remains home to two-thirds of the world’s poor: 1.7 billion people who live on less than $2 a day, with 828 million struggling on less than $1.25 a day. ADB is committed to reducing poverty through inclusive economic growth, environmentally sustainable growth, and regional integration. Based in Manila, ADB is owned by 67 members, including 48 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance. 8 © Asian Development Bank. Publication Stock No. ARM135435-3 December 2012 Printed on recycled paper