The Effect of Socio-Economic Status (SES) on Likelihood of Sickness, Care-Seeking, and Choice of Provider for Children Under-Five in Ten Districts of Uganda

Gita G. Mirchandani, Johns Hopkins University
David Bishai, Johns Hopkins University

Logistic regression was used to test the association between SES and choice of first-level provider (public, private/NGO, and drug shop/traditional provider) for sick children under the age of five. Socio-economic status, as measured by asset-ownership, was not associated with probability of sickness, utilizing care, or the multinomial probability of choice of provider. A conditional logit model was also used to associate both provider characteristics and family SES to the provider that was visited (chosen) versus the second choice provider (not chosen). The conditional logit model showed a significant interaction effect of SES as measured by education of head of household and travel time. Those with any education were 38% less likely (OR=0.62, p=0.09) to choose a provider with an additional hour of travel time as compared to those with no education.

  See paper

Presented in Session 75: Child Health and Socioeconomic Status