Systematic Review of Neonatal and Under-Five Mortality Predictors and Methods in Nigeria (2000–2025)

Neonatal mortality (deaths within the first 28 days of life) and under-five mortality (deaths before the fifth birthday) remain among the most pressing public health challenges in Sub-Saharan Africa, with Nigeria bearing a disproportionate burden globally. According to the United Nations Inter-Agency Group for Child Mortality Estimation (UNICEF et al., 2023), Nigeria recorded an estimated 74 deaths per 1,000 live births for under-five mortality and approximately 35 deaths per 1,000 live births for neonatal mortality in 2022, making it one of the countries with the highest absolute numbers of child deaths globally (Liu et al., 2016; UNICEF et al., 2023). Even though the world is making great strides, the under-five mortality rate has decreased by an average of about 51% in the world in the last twenty years, Nigeria has lagged behind regional and global colleagues in terms of development (WHO, 2023). Nigeria alone accounted for approximately 14% of all global under-five deaths in 2022, representing nearly 800,000 deaths annually (Liu et al., 2016; UNICEF et al., 2023). This chronic burden has significant implications on the capacity of Nigeria to realize Sustainable Development Goal (SDG) 3.2 which aims at reducing the neonatal mortality rate to at least as low as 12 per 1,000 live births and under-five mortality to at least as low as 25 per 1,000 live births by 2030 (WHO, 2023). In the 2018 Nigeria Demographic and Health Survey, chronic rural-urban inequalities, geographical inequity, especially between the geopolitical region of the North West and South West, and socioeconomic gradients in child mortality were recorded and are still poorly explained mechanistically (National Population Commission - NPC & ICF, 2019). The 2024 NDHS, the most recent nationally representative survey, provides a critical updated evidence base from which predictive models and survival analyses can generate actionable insights. Traditionally, the studies of child mortality in Nigeria and Sub-Saharan Africa have been based on classical techniques of survival analysis, such as Kaplan-Meier estimator and Cox Proportional Hazards (CPH) model (Cox, 1972; Kaplan & Meier, 1958). These parametric and semi-parametric methods have played a pivotal role in the determination of covariates including birth weight, maternal education, antenatal care (ANC) use and facility delivery as important predictors of child survival. (Adedini et al., 2021; Kayode et al., 2012). With the advent of machine learning (ML) and the ensemble-based predictive modelling models, such as Random Forests, Gradient Boosting Machines (GBM), Artificial Neural Networks (ANN), and survival forests, however, there are new opportunities to use non-linear relationships, high-dimensional covariate space, and complex interactions among child mortality determinants (Asnake et al., 2026; Brinati et al., 2020; Goldstein et al., 2017). A growing body of literature has begun applying these methods to DHS datasets in Sub-Saharan Africa, though their application within Nigeria remains limited and methodologically fragmented (Ezeh et al., 2014; Yaya et al., 2019). Critically, the methodological approaches, risk factors, and policy implications from studies using Nigerian DHS data or comparable nationally representative surveys has not yet been systematically synthesized in the literature. This gap is particularly acute given the imminent availability of the 2024 NDHS, which incorporates newer data collection modules on malnutrition, vaccination coverage, maternal health-seeking behaviour, and geographic information systems (GIS) coordinates enabling spatial analysis.