Monitoring of sex-specific mortality up to age five for all countries is challenging because of issues with data availability and quality. Moreover, sex ratios are expected to vary with under-five mortality, which makes it challenging to define “expected levels” and pinpoint to countries with outlying levels or trends, e.g. because of deprivation of girls’ access to health care or proper nutrition. We developed a Bayesian model to estimate sex ratios for infant, child and under-five mortality for all countries. In addition, we estimated the relative difference between national sex ratios and expected sex ratios based on the global relation between mortality and sex ratios. Based on these relative differences, we identified countries with outlying sex ratios and assessed excess female mortality.
This conference paper was presented by Leontine Alkema in the PAA 2014 Conference, Boston, MA, USA.