Unveiling Patterns in Health Service Usage and Child Mortality Through Data Science

Recent research utilizing advanced data science techniques has shed light on the intricate relationship between health service utilization and child mortality rates across sub-Saharan Africa. Children in this region face the highest under-5 mortality rate globally, with 74 deaths per 1,000 live births—significantly higher than in North America and Europe. In 2021, over 80% of worldwide under-5 deaths occurred in sub-Saharan Africa, often caused by preventable or treatable conditions such as diarrhea, malaria, and premature birth.

A multidisciplinary team from Washington University in St. Louis employed artificial intelligence and statistical learning methods to analyze a decade of health data from 31 countries in the region. The goal was to understand how reproductive, maternal, newborn, and child health services impact child mortality. The team examined socioeconomic variables—including maternal education and residence—linked to the likelihood of mothers accessing available healthcare services.

Analyzing data from more than 9,000 cases of child death before age five, the researchers identified three distinct groups of mothers based on their healthcare usage: low, medium, and high utilization. They also classified countries by overall service usage patterns. Multilevel latent class analysis helped recognize these groups, while multinomial regression revealed how factors like maternal education, employment, urban residency, and household wealth influence service use.

Findings indicated a strong correlation between socioeconomic status and access to health services. For instance, mothers with lower socioeconomic standing displayed higher breastfeeding rates, possibly due to having fewer nutritional alternatives but faced lower chances of receiving education and employment opportunities. The medium-utilization group showed high prenatal and postpartum care but lower institutional delivery rates, primarily due to accessibility issues like transportation.

Conversely, the high-utilization group demonstrated comprehensive engagement with health services, including family planning, prenatal and postnatal care, and protective health practices such as improved sanitation and clean cooking fuels. This group also had favorable health behaviors like appropriate birth spacing, marrying after age 18, and frequent use of health services across the continuum of care.

Moreover, the study revealed variation across countries driven by cultural, behavioral, and socioeconomic factors. Many countries with better access had higher overall service use, though disparities still persisted. The research emphasizes that addressing socioeconomic inequalities is key to improving maternal and child health outcomes.

Public health policymakers can leverage these insights to tailor strategies for disadvantaged populations, focusing on increasing access and education. The dataset’s cross-country perspective offers invaluable guidance for designing effective policies to reduce child mortality and improve health service uptake.

The full findings were published in uture Communications,urther details can be accessed through the DOI: 10.1038/s41467-025-61350-8, with support from Washington University in St. Louis.