Catastrophic health expenditure in Vietnam : studies of problems and solutions

Background: In Vietnam, problems of high out-of-pocket payments for health, leading to catastrophic health expenditure and resulting impoverishment for vulnerable groups, has been at focus in the past decades. Since the beginning of the 1990’s, the Vietnamese government has launched a series of social health insurance reforms to increase prepayment in the health sector and thereby better protect the population from the financial consequences of health problems. Objective: The objective of this thesis is to contribute to the discussion in Vietnam on how large the problems of catastrophic health expenditure are in the population as a whole and in a special subgroup; the elderly households, and to assess important aspects on health insurance as a means to reduce the problems. Methods: Catastrophic health expenditure has been estimated, using an established and common method, from two different data sources; the Epidemiological Field Laboratory for Health Systems Research (FilaBavi) in the Bavi district, and Vietnam Household Living Standards Survey (VHLSS) 2010. Results from two cross-sectional analyses and a panel study have been compared, to gain information on whether the estimates of catastrophic health spending may be overestimated when using cross-sectional data. Then, the size of the problem for one group, the elderly households; hypothesized to be particularly vulnerable in this context, has been estimated. The question of to what extent a health insurance reform; the Health Care Funds for the Poor (HCFP), has offered protection for the insured against health spending is being assessed in another study over the period 2001 – 2007, using propensity score matching. The value that households attach to health insurance has also been explored through a willingness to pay (WTP) study. Results: Comparing results from two cross-sectional studies with a panel study over a full year in which the respondents were interviewed once every month, the estimates of catastrophic spending vary largely. The monthly estimates in the panels study are half as large as the cross-sectional estimates; the latter also having a recall period of one month. Among the elderly households, catastrophic health spending and impoverishment are found to be problems three times as large as for the whole population. However, household health care expenditure as a percentage of total household expenditure was affected by the HCFP, and significantly reduced for the insured. In the study of household WTP for health insurance, it was iiifound that households attach a low value to this insurance form; WTP being only half of household health expenditure. Conclusions: Cross-sectional studies of catastrophic spending with a monthly recall period are likely to be affected by recall bias leading to overestimations through respondents including expenditure in the period preceding the recall period. However, such problems should not deter researchers form studying this phenomenon. If using the same method, estimates of catastrophic spending and impoverishment can be compared between different groups – as for the elderly households – and over time; e.g. studying the protective capacity of health insurance. It should be used more, not less. The VHLSS rounds offer the Vietnamese a possibility to regularly study this. The HCFP were found to be partly protective but important problems remain to be solved, e.g. the fact that people are reluctant to use their health insurance because of e.g. quality problems and possible discrimination of the insured. The findings of a low WTP for health insurance may be another reflection of this