Globally, the number of deaths of children under 5 (U5) decreased substantially from 12 million
in 1990 to about 6.9 million in 2011; during the same period under five mortality (U5M)
decreased steadily in Mozambique as a result of the implementation of several interventions.
This study aimed to determine the role of health systems strengthening and HIV in U5 mortality
trends in Mozambique. We performed an exploratory analysis with U5M as the outcome
variable. First, we conducted a univariate analyses by provinces over time periods, as well as
Pearson correlations between each independent predictor and the outcome; second a bivariate
analysis was performed to determine the association between each independent variable and the
outcome variable followed by a multivariate analysis. Model selection was achieved by using
backward selection where in each step were removed the variable with the highest p-value, The II
final significance level selected was 0.05. Overall the U5M in Mozambique dropped
substantially during 2000 to 2010 and for each additional year we predicted a decrease of 7.4 per
1000 live births of the U5M on average across all provinces (95% CI: -9.4 , -5.3). After adjusting
for time trend population per health facility with ß= 2.7 (95% CI 0.19, 5.2), health work force
density with ß= -0.41 (95% CI: -0.81, -0.01) and institutional birth attendance ß= -0.45 (95%
CI:-0.77, -0.14) remained significantly associated to U5M. These results suggest that
improvements on health human resources particularly with maternal and child nurses and
interventions which resulted in improvements of institutional birth attendance were important in
the reductions of the U5M rates in Mozambique during the period of the study. If these results
are confirmed with other studies, investments on health should prioritize innovative interventions
to accelerate human resources trainings, health infrastructure buildings and better access and
quality of services for pregnant women.