Abstract
BACKGROUND: Studies in the literature have found mixed results on the effect of microcredit on health outcomes. Of the five previous experimental studies that included microcredit and a health intervention, three reported no significant changes in health status or behaviors. The purpose of this study was to test for marginal and interactive effects of increased microcredit and provision of basic health services. METHODS: This study had a 4-celled experimental design in 128 villages in rural Bangladesh. For villages in one cell, an additional microcredit worker was assigned. For those in a second cell, a health assistant visited households each month, provided simple medicines and announced a satellite clinic held monthly in each village. For a third cell, both interventions were combined, and villages in a fourth cell served as control. A baseline survey was completed and a follow-up survey was done three years later. Outcome measures were food security, contraceptive use, having a trained birth attendant at last birth, and measles immunization. RESULTS: Comparison of follow-up with baseline levels of the four outcome measures (for 3787 households (96% completeness) and 3687 women (94% completeness)) showed significant improvement in food security in all study arms and a significant increase in trained birth attendant at last birth in the health services villages. Due to confusion within Grameen Bank about which workers would provide the additional microcredit work, that intervention was poorly implemented so in multivariate analyses, the data for that intervention arm were grouped with data from the control arm. Logistic regression with values of the outcomes at follow-up as dependent variable and study arm and women's schooling as covariates showed no significant effects of either separate or grouped study arms. CONCLUSION: Two of the three health behaviors showed no significant changes over time but having a trained birth attendant at last delivery did increase significantly in the health services arm. Therefore, community health education can sometimes be effective in promoting healthy behaviors. TRIAL REGISTRATION: This was a field trial rather than a clinical trial, so trial registration was unnecessary.