RESUMO
Preliminary evidence from hospital discharges hints enormous disparities in infant hospital mortality rates. At the same time, public health agencies acknowledge severe deficiencies and variations in the quality of medical services across public hospitals. Despite these concerns, there is limited evidence of the contribution of hospital infrastructure and quality in explaining variations in outcomes among those who have access to medical services provided at public hospitals. This paper provides evidence to address this question. We use probabilistic econometric methods to estimate the impact of material and human resources and hospital quality on the probability that an infant dies controlling for socioeconomic, maternal and reproductive risk factors. As a measure of quality, we calculate for the first time for Mexico patient safety indicators developed by the AHRQ. We find that the probability to die is affected by hospital infrastructure and by quality. In this last regard, having been treated in a hospital with the worse quality incidence doubles the probability to die. This paper also presents evidence on the contribution of other risk factors on perinatal mortality rates. The conclusions of this paper suggest that lower infant mortality rates can be reached by implementing a set of coherent public policy actions including an increase and reorganization of hospital infrastructure, quality improvement, and increasing demand for health by poor families.