Objective: To document the association between supply-side determinants and AIDS mortality in Mexico between 2008 and 2013.
Materials and methods: We analyzed the SALVAR database (system for antiretroviral management, logistics and surveillance) as well as data collected through a nationally representative survey in health facilities. We used multivariate logit regression models to estimate the association between supply-side characteristics, namely management, training and experience of health care providers, and AIDS mortality, distinguishing early and non-early mortality and controlling for clinical indicators of the patients.
Results: Clinic status of the patients (initial CD4 and viral load) explain 44.4% of the variability of early mortality across clinics and 13.8% of the variability in non-early mortality. Supply-side characteristics increase explanatory power of the models by 16% in the case of early mortality, and 96% in the case of non-early mortality.
Conclusions: Aspects of management and implementation of services contribute significantly to explain AIDS mortality in Mexico. Improving these aspects of the national program, can similarly improve its results.