Heterogeneity in enterotoxigenic Escherichia coli and shigella infections in children under 5 years of age from 11 African countries: a subnational approach quantifying risk, mortality, morbidity, and stunting

BACKGROUND: Diarrhoea, a global cause of child mortality and morbidity, is linked to adverse consequences including childhood stunting and death from other diseases. Few studies explore how diarrhoeal mortality varies subnationally, especially by cause, which is important for targeting investments. Even fewer examine indirect effects of diarrhoeal morbidity on child mortality. We estimated the subnational distribution of mortality, morbidity, and childhood stunting attributable to enterotoxigenic Escherichia coli (ETEC) and shigella infection in children younger than 5 years from 11 eastern and central African countries. These pathogens are leading causes of diarrhoea in young children and have been linked to increased childhood stunting. METHODS: We combined proxy indicators of morbidity and mortality risk from the most recent Demographic and Health Surveys with published relative risks to estimate the potential distribution of diarrhoeal disease risk. To estimate subnational burden, we used country-specific or WHO region-specific morbidity and mortality estimates and distributed them subnationally by three indices that integrate relevant individual characteristics (ie, underweight, probability of receiving oral rehydration treatment of diarrhoea, and receiving vitamin A supplementation) and household characteristics (ie, type of drinking water and sanitation facilities). FINDINGS: Characterising ETEC and shigella subnational estimates of indirect morbidity (infection-attributable stunting) and indirect mortality (stunting-related deaths from other infectious diseases) identified high-risk areas that could be missed by traditional metrics. Burundi and Democratic Republic of the Congo had the highest ETEC-associated and shigella-associated mortality and stunting rates. Mozambique, Democratic Republic of the Congo, and Zimbabwe had the greatest subnational heterogeneity in most ETEC and shigella mortality measures. Inclusion of indirect ETEC and shigella mortality in burden estimates resulted in a 20–30% increase in total ETEC and shigella mortality rates in some subnational areas. INTERPRETATION: Understanding the indirect mortality and morbidity of diarrhoeal pathogens on a subnational level will strengthen disease control strategies and could have important implications for the relative impact and cost-effectiveness of new enteric vaccines. Because our methods rely on publicly available data, they could be employed for national planning. FUNDING: Bill & Melinda Gates Foundation.