Children residing in sub-Saharan Africa experience the highest global rates of under-5 mortality, with 74 fatalities per 1,000 live births—14 times greater than the risk for children in North America and Europe. In 2021, sub-Saharan Africa represented over 80% of worldwide under-5 mortality. Tragically, the causes of death, including diarrhea, malaria, and premature birth, are predominantly preventable or treatable.
A multidisciplinary team of artificial intelligence (AI) and global health experts from Washington University in St. Louis aimed to investigate how the utilization of reproductive, maternal, newborn, and child health services might influence these alarming statistics. Previous studies have indicated that such services play a vital role in enhancing child health and averting mortality; the critical inquiries were who is utilizing these services and how they impact health outcomes.
Employing statistical learning to assess a decade’s worth of data from 31 countries in sub-Saharan Africa, the team identified significant socioeconomic factors—such as maternal education and living environment—that strongly correlate with maternal engagement in available health services. The study’s outcomes were published online on Aug. 22 in Nature Communications.
Claire Najjuuko, a PhD student at WashU, examined a dataset comprising over 9,000 births that led to death before reaching 5 years, sourced from the Demographic and Health Survey program. She employed multilevel latent class analysis on 16 health service indicators, distinguishing three specific categories of mothers—low, medium, and high service users—and grouping countries into three classifications based on comprehensive service utilization trends. Additionally, she utilized multinomial regression to uncover connections between socioeconomic factors and service usage patterns across diverse groups.
“Our research demonstrates a robust connection between socioeconomic factors and the use of maternal and child health services,” stated Najjuuko, who is jointly mentored by Chenyang Lu, the Fullgraf Professor in computer science and engineering at the McKelvey School of Engineering and director of the university’s AI for Health Institute, along with Fred M. Ssewamala, previously affiliated with the Brown School at WashU and now a professor of poverty studies at New York University.
Among the 16 variables that Najjuuko considered as indicators for accessing maternal and child health services were prenatal care, facility-based deliveries, postpartum care, breastfeeding, and protective measures like using improved sanitation facilities, having access to clean drinking water, and utilizing clean cooking fuels.
“We observed elevated breastfeeding practices among young mothers with low socioeconomic status, likely because they have the available time, and it’s often the sole option due to lack of alternative nutrition for their infants,” Najjuuko expressed.
Conversely, this same group exhibited lower chances of maternal education, employment, urban residence, or being in a high wealth category.
The medium-utilization category displayed a mixed profile, marked by high prenatal and postpartum care rates but low institutional delivery rates, likely influenced by accessibility challenges like transportation to hospitals for childbirth.
Najjuuko noted that the high-utilization group also showcased the greatest prevalence of protective practices, including improved sanitation facilities, access to clean drinking water, and the use of clean cooking fuels in their households.
“Additionally, this group had a substantial portion of met needs for family planning, appropriate birth spacing, marrying after the age of 18, and utilizing a broad spectrum of care services before, during, and after delivery, encompassing most of the 16 health service utilization indicators analyzed,” Najjuuko stated.
Beyond individual-level classifications, Najjuuko also categorized the data by country. The researchers discovered that service coverage and utilization significantly differed by country due to socioeconomic, behavioral, and cultural variances. More than half of the 31 countries presented relatively high rates of maternal and child health service utilization, likely owing to enhanced accessibility to these services. Conversely, other nations exhibited a more fragmented usage landscape.
“Our results indicate a substantial link between socioeconomic status and the use of maternal and child health services,” Najjuuko remarked. “It is essential to target strategies towards the most disadvantaged socioeconomic groups. Other studies have aimed at improving the socioeconomic conditions of families, as it is crucial for helping individuals escape poverty or gain access to education and knowledge regarding these life-saving services.”
Lu emphasized the dataset’s significance for enhancing public health and policy formulation.
“This multi-country dataset provides critical insights for shaping health policies,” Lu explained. “Many assumptions—such as the necessity of visiting a health facility for childbirth—reveal considerable variations across Africa, as shown by this survey. Policymakers can leverage these data-driven insights to inform their decisions. For instance, as Claire pointed out, some mothers are aware of the need to seek hospital care but face barriers like transportation. These findings offer clear and actionable insights into what types of policies might be effective in various countries, enabling policymakers to utilize resources optimally.”
Najjuuko C, Xu Z, Kizito S, Lu C, Ssewamala FM. Patterns of maternal, newborn, and child health services utilization, and associated socioeconomic disparities in sub-Saharan Africa: Multilevel latent class analysis. Nature Communications. Published online Aug. 22 https://doi.org/10.1038/s41467-025-61350-8.
This research was partially supported by funding from the National Institutes of Health (NIH) Researcher Resilience Training grant (R25MH118935-01), the AI for Health Institute, and the Fullgraf Foundation.
Originally published on the McKelvey Engineering website
The post Data science uncovers patterns in health service use linked to child mortality appeared first on The Source.