Modelling the Determinants of Typhoid Fever in Benue State, Nigeria Using Binary Logistic Regression

Abstract

Typhoid fever remains a significant public health burden in Benue State, Nigeria, where unsafe water, poor sanitation, inadequate hygiene, and low vaccination coverage continue to fuel transmission. Despite its endemicity, rigorous statistical evidence identifying the key determinants of infection within the state’s unique socio-environmental context remains limited. This study aimed to identify and quantify the independent predictors of typhoid fever in Benue State using binary logistic regression. The study adopted a descriptive and analytical cross-sectional design, utilizing secondary data from hospital records and epidemiological reports. After data cleaning and validation, researchers analyzed a total of 420 complete records obtained from Benue State University Teaching Hospital, Makurdi and related surveillance sources. The dependent variable was typhoid fever status (positive = 1; negative = 0). Independent variables included socio-demographic, environmental, household, behavioral, and clinical factors. Descriptive statistics, chi-square tests, and binary logistic regression were applied at a 5% significance level using standard statistical software. Of the 420 respondents, 168 (40.0%) tested positive for typhoid fever. Binary logistic regression identified ten significant independent predictors of infection. The strongest predictors were lack of typhoid vaccination (OR = 4.31; 95% CI: 2.21–8.41), poor handwashing practice (OR = 3.94; 95% CI: 2.06–7.53), uncovered water storage (OR = 3.49; 95% CI: 1.90–6.42), use of unimproved water sources (OR = 3.06; 95% CI: 1.71–5.47), and prior history of typhoid infection (OR = 2.97; 95% CI: 1.72–5.13). Additional significant predictors included unimproved toilet facilities (OR = 2.66), household size greater than five persons (OR = 2.44), age below 15 years (OR = 2.36); frequent eating outside the home (OR = 2.25); and low educational attainment (OR = 2.10). Sex was not a significant predictor. The model demonstrated good fit (Hosmer–Lemeshow χ² = 6.18, p = 0.63), with Nagelkerke R² = 0.51 and an overall classification accuracy of 79.5% (sensitivity: 76.2%; specificity: 81.7%). Typhoid fever in Benue State is a multifactorial disease driven by behavioral, environmental, household, clinical, and socio-demographic factors. Effective control requires integrated water, sanitation, and hygiene (WASH) interventions, expanded vaccination programs, and targeted health education campaigns focused on high-risk populations.