To estimate individual and household economic impact of cardiovascular disease (CVD) in selected low- and middle-income countries (LMIC).
Methods and Findings
We surveyed 1,657 recently hospitalized CVD patients (66% male; mean age 55.8 years) from Argentina, China, India, and Tanzania to evaluate the microeconomic and functional/productivity impact of CVD hospitalization. Respondents were stratified into three income groups. Median out-of-pocket expenditures for CVD treatment over 15 month follow-up ranged from 354 international dollars (2007 INT$, Tanzania, low-income) to INT$2,917 (India, high-income). Catastrophic health spending (CHS) was present in >50% of respondents in China, India, and Tanzania. Distress financing (DF) and lost income were more common in low-income respondents. After adjustment, lack of health insurance was associated with CHS in Argentina (OR 4.73 [2.56, 8.76], India (OR 3.93 [2.23, 6.90], and Tanzania (OR 3.68 [1.86, 7.26] with a marginal association in China (OR 2.05 [0.82, 5.11]). These economic effects were accompanied by substantial decreases in individual functional health and productivity.
Citation: Huffman MD, Rao KD, Pichon-Riviere A, Zhao D, Harikrishnan S, Ramaiya K, et al. (2011) A Cross-Sectional Study of the Microeconomic Impact of Cardiovascular Disease Hospitalization in Four Low- and Middle-Income Countries. PLoS ONE 6(6): e20821. https://doi.org/10.1371/journal.pone.0020821
Editor: German Malaga, Universidad Peruana Cayetano Heredia, Peru
Received: December 17, 2010; Accepted: May 13, 2011; Published: June 14, 2011
Copyright: © 2011 Huffman et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Global Forum for Health Research, World Bank, and International Union for Health Promotion and Education. MDH was supported by the NIH FICR-F program (R24TW007988) and is supported by an NHLBI training grant in cardiovascular epidemiology (5T32HL069771). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: This research was partially funded by the World Bank via the Global Forum for Health Research. Otherwise, the authors have no other relevant declarations relating to employment, consultancy, patents, products in development or marketed products related to this manuscript. The authors confirm that the inclusion of the World Bank as a partial funder does not not alter their adherence to all the PLoS ONE policies on sharing data and materials.
Cardiovascular disease (CVD) is the leading cause of mortality and among the leading causes of morbidity worldwide . On average, CVD affects patients in low- and middle-income countries (LMIC) ten to fifteen years earlier than high-income country patients, reducing LMIC workforce capacity and potential economic growth . The number of potentially productive years of life lost (PPYLL) due to CVD between ages 35–64 was estimated to be 9.2 million (3,572 per 100,000) in India and 6.7 million (1,595 per 100,000) in China in 2000. These estimates are projected to rise to 17.9 million PPYLL in India (3,707 PPYLL per 100,000) and 10.5 million PPYLL in China (1,863 per 100,000), by 2030 . On a national scale, Tanzania, India and China are projected to lose 2.5 billion, 237 billion, and 558 billion 1998 international dollars in gross domestic product (GDP), respectively, between 2005–2015 due to CVD, diabetes, and stroke . To the best of our knowledge, there are no similar estimates since 2000.
Beyond these macroeconomic projections, the 2000 World Health Report used fairness in financing—defined as the ratio of a household's total health spending to its capacity to pay—as a key indicator of health system performance (4). In this context, Xu and colleagues estimated in 2007 that 150 million people suffer from financial catastrophe (defined as annual health spending ≥40% non-food income) due to out-of pocket spending on health care. More than 90% of these people live in low-income countries . Despite the high prevalence of CVD and out-of-pocket healthcare financing in LMIC, there are limited data evaluating the association between acute CVD events and their economic impact.
We conducted a standardized survey in four LMICs—Argentina, China, India, and Tanzania—in order to quantify the effect of hospitalization for a CVD event on (a) health care expenditures, (b) how people financed health expenditures (c) impoverishment, and (d) changes in functional capacity and productivity. We hypothesized that a CVD event in a family member would produce a higher effect on family financial stress in low income families and families without health insurance.
The study was approved by a local ethics board for each participating site. Specifically, approval was provided by ethics boards located at the following institutions: Sree Chitra Tirunal Institute for Medical Sciences and Technology, National Institute of Medical Research, Muhimbili National Hospital, Capital Medical University affiliated Beijing Anzhen Hospital & Beijing Institute of Heart, Lung and Blood Vessel Diseases, and Hospital Italiano de Buenos Aires. Written informed consent was obtained from each participant in his or her primary language.
We performed a cross-sectional survey of recently hospitalized CVD patients in Argentina (Buenos Aires, La Plata, Mar del Plata), China (Beijing, Henan Zhoukou City), India (Trivandrum), and Tanzania (Dar es Salaam, Mwanza, Zanzibar). Each country chose three to seven hospitals, including a mix of public and private, urban and rural, hospitals with and without advanced [interventional] treatment facilities to cover a range of hospitalized participants. Hospitalized CVD patients were sampled using a stratified, random-sampling process based on age (<55 years old and ≥55 years old). All decisions related to in-hospital and follow-up medical care were decided by the participant and his/her providing physician. Surveys were developed by experts from the Initiative for Cardiovascular Health Research in Developing Countries to evaluate the individual- and household-level economic impact of hospitalization for a CVD-related event. Questions related to functionality were from adapted from the Short-Form-36.
Standardized surveys (Appendix S1) were translated from English into local languages and culturally adapted by each country site. A pilot study was performed at each site in order to detect implementation difficulties. Surveys were conducted three to fifteen months following hospital discharge. In order to evaluate the economic diversity among the study participants, participants were assigned to one of three income groups based upon the discharge hospital, payment scheme within each hospital, and/or participants' income and/or expenditures. The poorest group represented the poorest 40% (“low”), the middle group represented the middle 40% (“middle”), and the wealthiest group represented the top 20% (“high”) of the national population income distribution.
We included participants aged 25–70 years who had been hospitalized for one or more of the following admission diagnoses: acute coronary syndrome (ACS, including unstable angina or myocardial infarction), stroke, acute heart failure, or peripheral vascular intervention (including amputation). First-ever or repeat CVD events were included. Exclusion criteria were any of the following: active malignancy, end-stage renal disease requiring dialysis, solid-organ or hematopoietic transplant, human immunodeficiency virus infection, or severe mental illness.
Survey data were collected in the outpatient clinic or in the household in 2008–2009 by trained personnel to evaluate: 1) demography, 2) medical history, 3) individual and household economic information, 4) expenditures on CVD treatment associated with antecedent CVD event (including indirect costs), and 5) effect of a CVD event on health and productivity. Estimated costs were based upon the total inpatient and follow-up costs up to 15 months and were confirmed by chart review.
A household's health spending was considered catastrophic if annual out-of-pocket health expenditures were ≥40% of total, non-food household expenditures, since such spending is likely to result in impoverishment , . Distress financing was defined as financial activities, such as borrowing money from relatives/friends, taking loans from banks/other lenders, or selling assets (property, e.g.), that were directly related to the patient's most recent hospitalization.
In order to assess the economic impact of CVD at different income levels, each national group was stratified into income groups. Continuous variables are reported as means ± standard deviation or, if skewed in distribution, as median with interquartile range. Categorical variables are reported as proportions (%). Continuous variables were compared using t tests and one-way ANOVA, and categorical variables were compared using the chi-square test. We used univariate and multivariate logistic regression to assess the determinants of catastrophic health spending and distress financing. The multivariate models were constructed using those variables that were significant (p<0.1) in the univariate models and included a dichotomy variable to reflect the type of event, that is 1 for acute coronary syndrome and 0 for stroke. Income and expenditures across all countries were compared using the purchasing power parity conversion to 2007 international dollars (INT$) (7).
Demography, Baseline Economic Information, Medical History, and Presentation
Mean participant age ranged from 53 years (Tanzania) to 59 years (China), and the majority of participants were male (Table 1). Median number of years of education completed ranged from seven in Tanzania to ten in India. Median overall monthly individual incomes varied considerably across countries and were lowest in India (INT$259/month) and highest in Argentina (INT$975/month). Significant intra-country variability across income groups was present. A similar pattern was seen for monthly household income. Baseline unemployment ranged from 6% to 24% and was most common in the low-income group. Overall, lack of health insurance was common. Hypertension and tobacco use were among the most common CVD risk factors, and ACS and stroke were the most common admission diagnoses. Median length of stay ranged from 5 to 12 days. Median (interquartile range) time to survey completion following hospital discharge ranged from 174 (83, 264) days in Tanzania to 369 (306, 404) days in China (Table 1).
Expenditures, Microeconomic Effects, and Income Effects of CVD Event
When CVD expenditure data were stratified by income groups for each nation surveyed, median 15-month out-of-pocket CVD health spending ranged from INT$374 (Tanzania, low-income group) to INT$2,917 (India, high-income group), and median annual household expenditures ranged from INT$1,701 (India, low-income group) to INT$24,597 (Argentina, high-income group)(Table 2). Out-of-pocket CVD expenditures were significantly higher in high-income strata in India and Tanzania but significantly lower in the high-income group in China. The proportion of 15-month out-of-pocket CVD expenditures to annual household expenditures ranged considerably from 4% (Argentina, all income groups) to 55% (India, middle-income group).
Participants from all countries and all income strata reported a decrease in individual incomes after a CVD event. Participants from low-income strata were more likely to report a decrease in individual income than those in the high-income stratum in Argentina, China, and India, but not in Tanzania (Table 2).
Catastrophic health spending (CHS) occurred in all countries and income strata, ranging from 5% (Argentina, high-income stratum) to 92% (India and Tanzania, low-income strata). Distress financing ranged from 1% (Argentina, high-income stratum) to 64% (India, low-income stratum) (Figure 1). Borrowing from family, friends, and employers was the most common form of distress financing, but the distribution of distress financing type differed across countries (Appendix S2).
Differences across income strata were considered statistically significant (p<0.05) for China (CHS and DF), India (CHS), and Tanzania (CHS and DF).
Private/social health insurance was significantly and inversely associated with CHS in all countries by univariate analysis (Tables 3 and 4) and remained significant after controlling for other significant variables in Argentina (OR 4.73 [2.56, 8.76], India (OR 3.93 [2.23, 6.90], and Tanzania (OR 3.68 [1.86, 7.26] with a trend toward association in China (OR 2.05 [0.82, 5.11]). (Table 3) CHS was also associated with the low-income group in China, India, and Tanzania, an association that remained statistically significant in India (OR 6.59 [2.23, 19.45]), after adjustment for other significant variables. Distress financing was predicted by low-income status in Argentina (OR 3.08 [1.12, 8.43]), China (OR 6.67 [1.69, 26.35]), and Tanzania (OR 3.25 [1.08, 9.75]), whereas income status in India was not associated with distress financing (OR 1.30 [0.68, 2.49]). In India, health insurance was the strongest predictor of avoiding distress financing (OR 11.37 [5.18, 24.95]).
In Argentina, 15-month out-of-pocket CVD expenditures were similar across all income strata with a trend toward increased costs in the high-income group. Private/social health insurance coverage was moderate (52%) across all respondents. Access to ambulatory and in-hospital care is fully granted for all Argentineans, regardless of their insurance status or hospital where they receive care. In this context, CHS was relatively low across all groups but was significantly associated with lack of private or social health insurance (OR 4.72 [2.56, 8.76] Table 3). The proportion of distress financing was similar across all income groups and was significantly associated low-income status (OR 3.08 [1.12, 8.43]). Nevertheless, more than half of all participants from Argentina reported that they had experienced a decrease in individual and household income following their CVD-related hospitalization (Table 2).
Overall out-of-pocket CVD expenditures were higher among participants in the low-income stratum; the ratio of out-of-pocket CVD spending to total household expenditures was also higher. CHS was highest in the low-income stratum (71%) and had a significant negative association with age <55 years (OR 0.47 [0.26, 0.85]) and significant positive association with rural status (OR 2.69 [1.31, 5.53]) (Table 3). A similar distribution was seen for distress financing, which was significantly associated with rural status (OR 5.13 [1.53, 17.13] and low-income group (OR 6.67 [1.69, 26.35]). The subsequent proportion of participants who experienced any decrease in income was highest in the low-income group (53%), though the absolute decrease was highest in the high-income group (Table 2).
India had the highest 15-month out-of-pocket CVD expenditures among the comparator countries. Insurance coverage was uncommon in Indian respondents (16%), and overall CHS was common and regressive. CHS was associated with age <55 years (OR 1.66 [1.06, 2.61]), lack of private/social health insurance (OR 3.93 [2.23, 19.45]), and stroke (OR 0.60 [0.37, 0.97]) (Table 4). A similar pattern was seen for distress financing, which was again higher than other comparator countries. Rural status (OR 1.93 [1.27, 2.93]), less than secondary school education (OR 2.27 [1.34, 3.86], absence of private/social health insurance (OR 11.37 [5.18, 24.95]), and stroke (OR 0.32 [0.21, 0.51]) were significantly associated with distress financing. The proportion of participants who experienced any decrease in income was highest in the low-income stratum (43%), but the absolute decrease was also highest in the low-income group (Table 2).
United Republic of Tanzania.
In Tanzania, increased 15-month out-of-pocket CVD expenditures were associated with the high-income respondents, as were annual household costs. Insurance coverage was low throughout Tanzania (14%). As such, CHS was high in all income groups and was associated with rural status (OR 2.00 [1.07, 3.73]) and lack of private/social health insurance (OR 3.68 [1.86, 7.26] Table 4). Distress financing was uncommon, however, and ranged from 4% to 12% and was only associated with low-income group (OR 3.25 [1.08, 9.75]). Nevertheless, two-thirds of respondents reported a decrease in income following their CVD-related hospitalization. The proportion of respondents experiencing any decrease in income was similar across income groups (Table 2).
Functional Capacity and Productivity
The health effects, functional effects, and household effects of CVD-related hospitalization demonstrate the potential indirect economic impact of CVD-related hospitalization (Table 5). Most respondents experienced functional limitations in moderate and vigorous activities with subsequent emotional distress following their hospitalization. Individual decreases in work time and work activities were common coping mechanisms, whereas families responded by a net increase (Argentina, China, India) or net decrease (Tanzania) in work time to compensate for their family member's disability.
Approximately 50% of participants across all income strata reported a decrease in their health with nearly all reporting a decrease in their ability to participate in moderate (86%) or vigorous (88%) activities. The proportion of individuals experiencing emotional problems following their CVD-related hospitalization was also high (61%). Cost was the primary reason for not taking one's medications in 7–13% of respondents.
Most respondents reported a decrease in work time (73%) and reported limiting their work activities (78%) following their CVD-related hospitalization. The low-income stratum was more likely to report feeling limited overall (87%), but this response was common for the entire group (67%). Family members were more likely to increase their work time (or start new work) rather than report a decrease in their work time (or stop work)(Table 5).
The majority of respondents reported a decrease in their self-rated health (62%), while half (51%) of participants were less able to perform moderate physical activities and three-fourths (76%) of participants were less able to perform vigorous physical activities. Approximately one-fourth of the respondents did not take their medications due to cost (Table 5).
Most respondents decreased their work time (81%) and limited their work activities (87%) following their CVD-related hospitalization. The overall proportion of family members who increased their work time (14%) was similar to the proportion that decreased their work time (15%), but the increase in work time by family members was most common in the low-income stratum (Table 5).
Almost two-thirds of respondents reported a decrease in their self-rated health (60%). Less than half of participants reported a worsening of their ability to perform moderate activities (45%), while two-thirds reported a worsening of their ability to perform vigorous activities, both of which trended lower than the other comparator countries. The proportion of individuals reporting emotional problems following their CVD-related hospitalization was also lower than the comparator countries (27%) but highest in the high-income strata (33%). Approximately one in ten respondents across all income strata did not take their medications due to cost (Table 5).
More than three-fourths of respondents decreased their work time, limited their work activities, and felt limited overall. A slightly greater proportion of family members increased their work time rather than decrease their work time following the respondent's CVD-related hospitalization, and this trend was most marked in the low-income stratum (Table 5).
United Republic of Tanzania.
Nearly all respondents across all income strata reported a decrease in their self-rated health (96%) with a corresponding worsening in their ability to participate in moderate (84%) and vigorous (96%) activities. The proportion of individuals reporting emotional problems following their CVD-related hospitalization was high overall (70%) and highest in the high-income strata. Approximately one-third of the respondents did not take their medications as prescribed, which was largely due to the highs costs (76%) (Table 5).
Nearly all respondents decreased their work time, limited their work activities, and felt limited overall (99%) following their CVD-related hospitalization. A greater proportion of family members decreased their work time rather than increased their work time to care for the patient (Table 5).
Health Care Spending
We evaluated the individual- and household-level impact of CVD-related hospitalization across four LMIC and found that 15-month out-of-pocket CVD expenditures varied considerably across countries and across income groups within countries (INT$374 [Tanzania, low-income] to INT$2,917 [India, high-income]). By comparison, in-hospital out-of-pocket expenditures for CVD in the United States were estimated to be INT$1,229 in 2006 .
In two of the four countries studied, CVD costs are regressive: poorer respondents pay a higher proportion of income on health care following a CVD hospitalization, as previously demonstrated for other conditions in other countries , . The examples of Argentina and Tanzania in our survey shows that regressive CVD costs are not universal in LMIC  and suggests there is a wide range of country-specific, economic and/or health system development within the LMIC category.
Financing Mechanisms and Impoverishing Effects of CVD
Catastrophic health spending was common in China, India, and Tanzania and most strongly impacted the poorest CVD patients and their families. Our results are markedly higher than mean levels of CHS previously reported among community dwellers in 89 countries (2.3%) , . Global estimates for CHS range from almost zero percent in the United Kingdom, Czech Republic, and Slovakia to >10% percent in Vietnam and Brazil . However, since our study focused on recently hospitalized patients, our figures would be expected to be higher, though perhaps not to this degree. The high proportion of CHS may be underestimated because of individuals who avoid medical treatment due to high, perceived costs, yet might have financial insecurity if they did. On the other hand, distress financing—risky financial activities such as borrowing loans and selling assets—was present in >40% of participants only in India. Participants from Argentina, China, and Tanzania most commonly borrowed money from family, friends, and employers to cover their health care costs, whereas more than half of participants who experienced distress financing in India borrowed from banks/moneylenders or sold assets.
A 2009 survey evaluating the microeconomic impact of stroke across 62 hospitals in China found that 71% of post-stroke patients reported CHS. While the authors used a different definition of CHS (≥30% of annual income), the findings are similar overall to our results (the range of CHS in our sample from China across income groups was 37–71%) . However, in our sample, CHS was not associated with hospitalization due to stroke compared with acute coronary syndromes, which may be due to higher costs of acute coronary syndromes or a lack of power to detect such a difference. This comparison across Argentina, China, India, and Tanzania provides a wider scope of the individual- and household-level economic impact of CVD-related hospitalizations.
Along with high CVD-related costs, many participants reported decreased income, poorer perceived health (including emotional problems), lower functional and productivity capacity, and variable household effects, all of which likely exacerbated their financial instability. These findings, coupled with the relatively low proportion of any form of insurance, likely contributed to the high proportion of CHS. Some argue that CHS overestimates the impoverishing effects of health care costs since families are able to “smooth consumption” by drawing upon savings, assets, credit and loans from friends and relatives . We tried to account for such activities through questions about distress financing, which may offset CHS but contributes to chronic impoverishment .
Lack of private/social health insurance was significantly associated with an increased risk of CHS in 3 of 4 countries studied and a trend toward association in China. Insurance alone, however, does not protect fully against CHS, as evident in our study with the high proportion of CHS in China despite a relatively higher proportion of insurance coverage. This discrepancy may be due, in part, to the proportion of reimbursement by the insurer, which was lowest for both hospital and outpatient charges in the low-income group and rural respondents in China (data not shown). On the other hand, low-income respondents were more likely to experience distress financing in 3 of 4 countries, suggesting that poorer participants may be less likely to have financial reserves to bear the costs of a CVD-related hospitalization. The availability of health services requiring payment, national inequality in health spending, and the capacity of individuals and households to pay (non-subsistence spending) are other key determinants associated with CHS .
Solutions to Avoid Catastrophic Health Spending and Distress Financing
The primary means to avoid distress financing typically includes prepayment either through tax financing, social health insurance programs, or private health insurance programs, the latter two which have been shown to reduce CHS incidence in Mexico . On the other hand, social health insurance programs have been shown to increase per capita health spending by 3–4% without improved outcomes or even at the cost of 5–6% potential life years lost in one report , . Other studies have demonstrated an increase in CHS with a decrease in the depth of insurance, including evaluations of China's urban and rural insurance schemes, which promote more complex, expensive care that is not wholly covered , . Another mechanism to decrease CHS and distress financing is to reduce the supply side of health care through restrictions in spending opportunities (treatment protocols, essential drug lists, restriction of unnecessary interventions, e.g.), which has been shown to be more successful at reducing CHS in China than expansion of insurance coverage alone .
Our study has inherent limitations. First, our survey sample was hospital-based, and did not sample patients who avoided seeking care for a CVD event nor those who did not survive a CVD event during the follow-up period: this may bias the results by describing only the most severe events, or by missing patients who avoided seeking medical care or those who did not survive a CVD event. Second, participants were asked to report sensitive income and expenditure information up to 15 months after hospitalization, which may be susceptible to reporting bias. Third, our survey captured participants from selected hospitals from each country, which should not be considered generalizable to other hospitals. We did, however, attempt to capture a large number of respondents to effectively evaluate the economic effects of CVD-related hospitalizations in these selected LMIC, though country sample sizes per country were not proportionate to overall country population sizes. Fourth, our sample size may have been too small to detect differences in the distribution and determinants of CHS and DF. Fifth, many patients could not afford costly treatments like medications, percutaneous coronary intervention or coronary artery bypass graft surgery nor diagnostics and follow-up care and subsequently did not undergo them, which may underestimate expenditures in an evidence-based treatment setting. Sixth, we did not exhaust all the potentially relevant microeconomic consequences of CVD hospitalization, such as savings or labor supply.
Patients in Argentina, China, India, and Tanzania bear a significant burden of out-of-pocket payments, as defined by CHS and DF, following CVD hospitalization, though substantial variations exist across and within countries. Lack of insurance appears to be a major, remediable source of the financial burden of CVD in these countries. As CVD prevalence increases in LMIC, the household economic impact of CVD may worsen without the development of alternative health spending models that enhance patients' capacity to pay or without more active policies to prevent or at least postpone the onset of CVD in LMIC.
Dorairaj Prabhakaran, Stephen Leeder, K. Srinath Reddy, Krishna D. Rao, Andres Pichon-Riviere, Dong Zhao, S. Harikrishnan, Kaushik Ramaiya, Thomas A. Gaziano, Lorenzo Rocco, Marc Suhrcke, Andrew E. Moran.
V.S. Ajay, Krishna D. Rao, Poornima Prabhakaran, Shifalika Goenka, Mark D. Huffman, K. Srinath Reddy, Dorairaj Prabhakaran.
English language example of survey.
Patterns of distress financing across Argentina, China, India, and Tanzania following CVD-related hospitalization. (Gray = sold land or other assets; black = borrowed money from bank or moneylenders; red = borrowed money from friends, family, and employer).
Iribarren S, Galante J, Chirón L, Albuquerqu C, Curotto C, Curotto O, Manus LU, Ruth C, De La Torre A, Alessio E, Romoli M, D'abramo A; Hospital Interzonal Agudos y Crónicos San Juan de Dios (La Plata); de la Vega SG, Villanueva VS, Hospital Interzonal General de Agudos (Mar del Plata), Alejandro GE, Ignacio BR, Cristina GE, Hospital General de Agudos “Dr. C. Argerich” (Buenos Aires), Tesolin P, Zurrú MC, Rubinstein F, Alvarez G, Colica D, Max N, Ortega FM, Hospital Italiano (Buenos Aires), Manzano ML, Saconi A, Causerano J, Clínica Opsa – Obra Social for Sanity Personal (Buenos Aires), Kuhn G, Cosentino A, Eugenio O, Paula B, Natalia F, Mara DO, Fernandex AC, Ramirez M, Sanatorio Franchin (Buenos Aires), Shejtman A, Antoniotti C, Bonaverdi R, Amoedo PM, Arozarena I, Loza P, Silvia O, Leonardo C, Hospital Universitario de a Universidad Abierta Interamericana (Buenos Aires); Liu J, Wang, M, Qin Lm Sun J (Beijing), Ren J, Feng T, Cao F Anzhen Hospital (Beijing), Zhang Q, Zhang X, Yang C, Sun C Zhoukou Hospital (Zhoukou); Sukumaran S, Sarma PS, RamanKutty V, Remya S, Rajasekharan, Sree Chitra Tirunal Institute for Medical Sciences & Technology (Trivandrum); Bahuleyan CG, Shyalaja PN, Ananthapuri Hospitals (Trivandrum); Suresh K, Mathew R Trivandrum Medical College (Trivandrum), RajaLekshmi K, SUT Hospital (Trivandrum), Anil Kumar TV, Remash K, SK Hospital (Trivandrum), Pradeep P, Medical Mission Hospital (Trivandrum); Muhimbili National Hospital (Dar es Salaam), Mnazi Moja Hospital (Zanzibar).
Conceived and designed the experiments: VSA KDR SG KSR SL DP. Performed the experiments: AP-R SH DZ KR SG. Analyzed the data: MDH KDR DP VSA AP-R SH DZ KR SG JIC JEC YL JL SN KRT MD JvE TAG SL AEM AM KSR MS. Contributed reagents/materials/analysis tools: MDH KDR AP-R DZ SH KR VSA SG JIC JEC SN YL JL KRT MD JvE AM AEM TAG MS KSR SL DP. Wrote the paper: VSA MDH.
- 1. World Health Organization (2004) Global burden of disease: 2004 update. World Health Organization2004Global burden of disease: 2004 update.Geneva, Switzerland. Geneva, Switzerland.
- 2. Leeder S, Raymond S, Greenberg H, Liu H, Esson K (2004) A Race against time: The challenge of cardiovascular disease in developing countries. S. LeederS. RaymondH. GreenbergH. LiuK. Esson2004A Race against time: The challenge of cardiovascular disease in developing countries.Trustees of Columbia University, New York, NY. Trustees of Columbia University, New York, NY.
- 3. World Health Organization (2005) Preventing chronic diseases: A vital investment. World Health Organization2005Preventing chronic diseases: A vital investment.Geneva, Switzerland. Geneva, Switzerland.
- 4. Xu K, Evans DB, Carrin G, Aguilar-Rivera AM, Musgrove P, et al. (2007) Protecting households from catastrophic health spending. Health Aff (Millwood) 26: 972–983.K. XuDB EvansG. CarrinAM Aguilar-RiveraP. Musgrove2007Protecting households from catastrophic health spending.Health Aff (Millwood)26972983
- 5. World Health Organization (2001) World Health Report 2000—health systems: improving performance. World Health Organization2001World Health Report 2000—health systems: improving performance.Geneva, Switzerland. Geneva, Switzerland.
- 6. Kawabata K, Xu K, Carrin G (2002) Preventing impoverishment through protection against catastrophic health expenditure. Bull World Health Organ 80: 612.K. KawabataK. XuG. Carrin2002Preventing impoverishment through protection against catastrophic health expenditure.Bull World Health Organ80612
- 7. (2006) Medical expenditure panel survey. 2006Medical expenditure panel survey.Agency for Healthcare Research and Quality. Available at: http://www.meps.ahrq.gov/mepsweb/data_stats/quick_tables_results.jsp?component=1&subcomponent=0&tableSeries=2&year=-1&SearchMethod=1&Action=Search. Accessed 17 December 2009. Agency for Healthcare Research and Quality. Available at: http://www.meps.ahrq.gov/mepsweb/data_stats/quick_tables_results.jsp?component=1&subcomponent=0&tableSeries=2&year=-1&SearchMethod=1&Action=Search. Accessed 17 December 2009.
- 8. World Health Organization (2003) Coping with health-care costs: implications for the measurement of catastrophic expenditures and poverty. World Health Organization2003Coping with health-care costs: implications for the measurement of catastrophic expenditures and poverty.Health system performance assessment: World Health Report 2003. Geneva, Switzerland. Health system performance assessment: World Health Report 2003. Geneva, Switzerland.
- 9. Waters HR, Anderson GF, Mays J (2004) Measuring financial protection in health in the United States. Health Policy 69: 339–349.HR WatersGF AndersonJ. Mays2004Measuring financial protection in health in the United States.Health Policy69339349
- 10. Makinen M, Waters H, Rauch M, Almagambetova N, Bitran R, et al. (2000) Inequalities in health care use and expenditures: empirical data from eight developing countries and countries in transition. Bull WHO 78(1): 55–65.M. MakinenH. WatersM. RauchN. AlmagambetovaR. Bitran2000Inequalities in health care use and expenditures: empirical data from eight developing countries and countries in transition.Bull WHO7815565
- 11. Xu K, Evans DB, Kawabata K, Zeramdini R, Klavus J, et al. (2003) Household catastrophic health expenditure: a multicountry analysis. Lancet 362: 111–117.K. XuDB EvansK. KawabataR. ZeramdiniJ. Klavus2003Household catastrophic health expenditure: a multicountry analysis.Lancet362111117
- 12. Flores G, Krishnakumar J, O'Donnell O, van Doorslaer E (2008) Coping with health-care costs: implications for the measurement of catastrophic expenditures and poverty. Health Econ 17: 1393–1412.G. FloresJ. KrishnakumarO. O'DonnellE. van Doorslaer2008Coping with health-care costs: implications for the measurement of catastrophic expenditures and poverty.Health Econ1713931412
- 13. Knaul FM, Arreola-Ornelas H, Mendez-Carniado O, Bryson-Cahn C, Barofsky J, et al. (2006) Evidence is good for your health system: policy reform to remedy catastrophic and impoverishing health spending in Mexico. Lancet 368: 1828–1841.FM KnaulH. Arreola-OrnelasO. Mendez-CarniadoC. Bryson-CahnJ. Barofsky2006Evidence is good for your health system: policy reform to remedy catastrophic and impoverishing health spending in Mexico.Lancet36818281841
- 14. Wagstaff A (2009) Social health insurance vs. tax-financed health systems–evidence from the OECD. A. Wagstaff2009Social health insurance vs. tax-financed health systems–evidence from the OECD.Washington, DC. World Bank: Policy research working group # WPS 4821. Washington, DC. World Bank: Policy research working group # WPS 4821.
- 15. Wagstaff A, Moreno-Serra R (2009) Europe and central Asia's great post-communist social health insurance experiment: Aggregate impacts on health sector outcomes. J Health Econ 28: 322–340.A. WagstaffR. Moreno-Serra2009Europe and central Asia's great post-communist social health insurance experiment: Aggregate impacts on health sector outcomes.J Health Econ28322340
- 16. Wagstaff A, Lindelow M (2008) Can insurance increase financial risk? The curious case of health insurance in China. J Health Econ 27(4): 990–1005.A. WagstaffM. Lindelow2008Can insurance increase financial risk? The curious case of health insurance in China.J Health Econ2749901005
- 17. Wagstaff A, Lindelow M, Gao J, Ling X, Juncheng Q (2007) Extending health insurance to the rural population: an impact evaluation of China's new cooperative medical scheme. A. WagstaffM. LindelowJ. GaoX. LingQ. Juncheng2007Extending health insurance to the rural population: an impact evaluation of China's new cooperative medical scheme.Washington, DC. World Bank: Impact evaluation series #12. Policy research working paper #WPS 4150. Washington, DC. World Bank: Impact evaluation series #12. Policy research working paper #WPS 4150.
- 18. Wagstaff A, Yu S (2007) Do health sector reforms have their intended impacts? The World Bank's Health VIII project in Gansu province, China. J Health Econ 26(3): 505–35.A. WagstaffS. Yu2007Do health sector reforms have their intended impacts? The World Bank's Health VIII project in Gansu province, China.J Health Econ26350535