Antenatal care (ANC) is one of the key interventions of the Every Newborn action plan to
improve newborn health and prevent stillbirths by 2035. However, little is known about its
relationship with neonatal mortality in sub-Saharan Africa since the 1990s. We use data from 54
Demographic and Health Survey (DHS) from 27 countries to make comparisons of neonatal
mortality by ANC attendance. Each country had two surveys that were categorized as ‘earliest
surveys’ (i.e. conducted since 1990 but before 2010) and ‘latest surveys’ (from 2010 to 2014).
Multi-level logistic regression model and meta-analysis were applied on 1.1 million births that
occurred among women in the 5 years preceding the surveys. Overall neonatal mortality rate (NMR)
was 37.7 (95% CI, 37.4–38.1) deaths per 1000 live births; NMR in the earliest surveys were
46.0 (95% CI, 45.4–46.7) and 33.4 (95% CI, 33.0–33.8) deaths per 1000 live births in the latest
surveys. The overall NMR was also 10% higher than expected NMR (37.7 vs 34.3 deaths per 1000
live births). NMR was 2.2 times higher among births of women with no ANC compared to those who
had at least one ANC visit (42.5 vs 19.6 per 1000 live births). After adjusting for place of delivery,
maternal age at birth, relative household wealth, residence, mother’s education, marital status, birth
order, sex of child, and period of survey, the overall odds ratio (OR) demonstrated that women with
at least one ANC visit were 48% less likely to report neonatal deaths (OR: 0.52; 95% CI: 0.47–0.57)
than women who did not receive ANC. NMR was 27% less likely to occur during the latest surveys
than during the earliest surveys (OR: 0.73; 95% CI: 0.71–0.75). We discuss these results within the
context of calls for continued efforts to deploy interventions aimed at improving the quality of
maternal and newborn care.