The effect of health provider density on health care utilization and inequity in Peru

Background: this study examines the effect of health provider density on the probability of health care utilization for the publicly- and uninsured population in Peru. Evidence on determinants of health care utilization in developing countries is low. The other objective is to examine to what extent health provider density creates inequality in health care utilization, and the underlying factors causing inequity for this part of the population in Peru. This thesis contributes to the knowledge about the effect of health provider density on health care utilization and inequity, in a developing country. It is specifically interesting for policy makers. Methods: the study uses eight binary measures of health care utilization, which are utilization of GP consults, analysis, scans, other tests, ophthalmology, medicines and health care use in general. A logistic regression method is used to obtain the relations. The main explanatory variables are doctor density, nurse density and skilled health provider density (doctors + nurses). To measure inequality concentration curves are obtained showing the distribution of provider density against health care utilization. Health care utilization is standardized by age, gender and health status. Inequality is decomposed to obtain the factors causing inequity in utilization of GP visits and health care utilization in general. Results: skilled health providers are significantly and positively related to utilization of GP consults and ophthalmology. The results are rather contrasting because doctors show a positive and nurses a negative relation with utilization of health care. Doctor density is positively and significantly associated with utilization of GP consults, analysis, other tests and health care overall. Nurse density is negatively and significantly related to utilization of other tests. The three groups of health provider density are not correlated with the other measures of health care utilization. Inequity exists favoring people living in high provider density regions for utilization of GP consults and health care overall. Decomposition shows that doctor density and wealth are the most important contributors of the inequity. Conclusion: the results indicate that especially physician density has a positive relation with health care utilization, especially GP consults. The weakened results when controlling for confounding factors indicate that health provider density is not the only explanatory factor for health care utilization. It is plausible that the quality of provided services plays an important role. Furthermore, the inequity analysis shows that it would be effective to distribute health providers to areas with low provider densities