Modeling diarrhea in children under five in Somaliland: A machine learning analysis using SLDHS 2020 data

Area of Science:

• Public Health
• Epidemiology
• Pediatrics

Background:

• Diarrhea is a major cause of illness and death in children under five, especially in low- and middle-income countries.
• This study examines diarrhea prevalence and its determinants in Somaliland using the 2020 Somaliland Health and Demographic Survey (SLDHS).

Purpose of the Study:

• To investigate the prevalence and identify key determinants of childhood diarrhea in Somaliland.
• To inform targeted public health interventions for diarrhea prevention and control.

Main Methods:

• Cross-sectional study analyzing data from 1,112 women and their children under five.
• Descriptive statistics, binary logistic regression, and supervised machine learning models (Logistic Regression, Probit, Random Forest, Decision Tree, SVM) were used.
• Variables included socioeconomic, demographic, and environmental factors.

Main Results:

• Overall diarrhea prevalence was 7.2% with significant regional variations.
• Nomadic households showed higher incidence (8.62%) compared to rural and urban households.
• Region, household wealth, and sanitation access were significant predictors; maternal education was not.

Conclusions:

• Region-specific public health strategies are vital, focusing on water and sanitation access for nomadic and low-income populations.
• Machine learning models showed high overall accuracy but require refinement for predicting positive diarrhea cases.
• Enhancing infrastructure and sanitation in underserved communities is a priority to reduce childhood diarrhea.