Access Barriers and Facility Delivery Inequalities in Zambia: A Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy.

Laura Sochas, London School of Economics

LMIC progress in reducing intra-country inequalities in maternal healthcare access has lagged behind progress made in other primary healthcare areas. Additional evidence is needed to inform policy-makers on the health system factors shaping disparities in maternal healthcare access, going beyond individual-level determinants. This study draws on established “relational” healthcare access models to define accessibility as the extent to which the health system is fit for purpose given the population’s needs and capacities. It is the first study to apply an innovative method from social epidemiology, Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (MAIHDA), to the study of healthcare access barriers. Findings provide reliable estimates of the predicted probability of facility delivery for 24 barrier combinations, demonstrate very good levels of discriminatory accuracy for the proposed model, and show that geographic, availability, and quality of care barriers are particularly effective in predicting who is likely to access a facility delivery.