A Statistical Analysis of Typhoid Fever in Aliade

Typhoid fever remains a serious public health problem in many underdeveloped regions, with children and young adults (5–19 years) at highest risk. In Benue State, Nigeria, hospital records suggest a rising trend in reported typhoid cases, yet few studies have examined future incidence under current conditions. This research modeled and forecasted daily typhoid fever cases at Saint Vincent Hospital. Aliade (Gwer?East LGA) using an Auto?Regressive Integrated Moving Average (ARIMA) approach, thereby providing policymakers with data-driven projections for resource planning. The research also analyzed 501 daily observations of laboratory?confirmed typhoid cases collected from October 2019 to February 2021. Descriptive statistics (mean = 15.95 cases/day; SD = 8.81) characterized the series, which exhibited non?stationarity (ADF p = 0.301) at level. First and second differencing achieved stationarity (ADF p = 0.01). Autocorrelation (ACF) and partial autocorrelation (PACF) plots guided tentative model selection, and the final ARIMA (2,1,4) model was chosen based on minimum Akaike Information Criterion (AIC = 837.11). Model adequacy was confirmed via residual diagnostics (Ljung–Box test), and forecast accuracy was benchmarked against mean and naïve methods using mean absolute deviation and mean squared error. The ARIMA (2,1,4) model demonstrated strong fit (log?likelihood = –410.56; σ² = 0.2998) and outperformed benchmark forecasts in both MAD and MSE. A 15?year forecast indicates a slight upward trajectory in daily typhoid cases, suggesting continued burden without enhanced interventions. ARIMA modeling provides reliable short? and long?term forecasts of typhoid fever incidence. These projections can inform targeted prevention strategies, resource allocation, and public health planning to mitigate future outbreaks in the region.