Modeling and forecasting monthly Hepatitis B virus infections in Benue State using prophet time series model

Abstract

Hepatitis B Virus (HBV) infection remains a major public health challenge in Nigeria, particularly in Benue State, where endemic transmission persists despite ongoing prevention programmes. This retrospective time-series study aimed to model historical trends and forecast monthly HBV infections in Benue State using the Prophet forecasting time series model while evaluating the effects of selected public health interventions. Secondary surveillance data comprising 192 monthly laboratory-confirmed HBV records collected between January 2010 and December 2025 were obtained from the Benue State Epidemiological Unit, Makurdi. Rainy season, vaccination weeks, and public health awareness campaigns were incorporated as intervention variables because they represent environmental conditions and routine control measures expected to influence HBV transmission and case detection. Descriptive statistics, the Anderson-Darling normality test, Prophet decomposition, model diagnostics, and forecast evaluation were performed. The series had a mean monthly value of 61,544 confirmed cases, with significant non-normality (A² = 2.994; p < 0.005). The fitted Prophet model identified a significant upward trend (k = 2.8553, p < 0.001), significant seasonal patterns, increased infections during the rainy season (β = 8.5085, p = 0.0001), and significant reductions during vaccination weeks (β = −3.2244, p = 0.0077) and awareness campaigns (β = −1.1253, p = 0.0060). Model performance was satisfactory (R² = 0.8716; RMSE = 3.6625; MAPE = 6.35%), with diagnostic tests confirming model adequacy. Forecasts for January 2026-December 2027 indicated persistent endemic transmission with seasonal peaks during June-August, accompanied by 95% prediction intervals quantifying forecast uncertainty. The findings demonstrate that the Prophet model provides a reliable framework for HBV surveillance, early warning, and evidence-based public health planning in Benue State.