State-Level Epidemiological Analysis and Exploratory Machine-Learning Classification of Nigeria's 2022-2024 Diphtheria Outbreak

This preprint presents a state-level epidemiological analysis and exploratory machine-learning classification of Nigeria's 2022 - 2024 diphtheria outbreak. The study characterises the geographic distribution of reported disease burden, examines the association between DTP3 vaccination coverage and reported diphtheria cases, and explores whether vaccination and demographic indicators can distinguish states according to relative reported disease burden. State-level secondary surveillance data were compiled from publicly available reports from the Nigeria Centre for Disease Control and Prevention (NCDC), UNICEF, and the International Federation of Red Cross and Red Crescent Societies (IFRC), alongside DTP3 vaccination coverage data from the 2021 Nigeria Multiple Indicator Cluster Survey (MICS) and state population estimates. Descriptive epidemiological, statistical, geospatial, and exploratory machine-learning analyses were conducted using Python. Logistic Regression and Random Forest models were evaluated using classification metrics and five-fold cross-validation. The study identified substantial geographic disparities in reported diphtheria burden and an inverse association between DTP3 vaccination coverage and confirmed case counts. Exploratory machine-learning models demonstrated moderate classification performance, with vaccination coverage emerging as an important predictor. Findings should be interpreted cautiously because of the ecological study design, differences in the timing of vaccination and outbreak data, potential variation in surveillance and reporting, and the exploratory nature of the machine-learning analyses.