Targeted intervention for the ultra poor in rural Bangladesh: Does it make any difference in their health-seeking behaviour?

It is now well recognised that regular microcredit intervention is not enough to effectively reach the ultra poor in rural Bangladesh, in fact it actively excludes them for structural reasons. A grants-based integrated intervention was developed (with health inputs to mitigate the income-erosion effect of illness) to examine whether such a targeted intervention could change the health-seeking behaviour of the ultra-poor towards greater use of health services and "formal allopathic" providers during illness, besides improving their poverty status and capacity for health expenditure. The study was carried out in three northern districts of Bangladesh with high density of ultra poor households, using a pre-test/post-test control group design. A pre-intervention baseline (2189 interventions and 2134 controls) survey was undertaken in 2002 followed by an intervention (of 18 months duration) and a post-intervention follow-up survey of the same households in 2004. Structured interviews were conducted to elicit information on health-seeking behaviour of household members. Findings reveal an overall change in health-seeking behaviour in the study population, but the intervention reduced self-care by 7 percentage units and increased formal allopathic care by 9 percentage units. The intervention increased the proportion of non-deficit households by 43 percentage units, as well as the capacity to spend more than Tk. 25 for treatment of illness during the reference period by 11 percentage units. Higher health expenditure and time (pre- to -post-intervention period) was associated with increased use of health care from formal allopathic providers. However, gender differences in health-seeking and health-expenditure disfavouring women were also noted. The programmatic implications of these findings are discussed in the context of improving the ability of health systems to reach the ultra poor.