Application of Weibull Survival Regression for Predicting Hospital Length of Stay Among Hepatitis B patients in Maiduguri, Nigeria

Abstract

Hepatitis B virus (HBV) infection continues to pose major challenges for global health systems, particularly in resource-limited areas where hospital congestion and prolonged hospital stays strain available services. This study aims to model and identify the predictors of hospital length of stay (LOS) among HBV patients using a Weibull survival regression approach. Retrospective clinical and demographic data from 60 HBV patients managed in Maiduguri, Nigeria, were analyzed. Key variables included age, gender, comorbidity status, and antiviral treatment, and the Weibull model was specified to estimate time-to-discharge while accounting for the skewed distribution of LOS. The results show that antiviral therapy significantly reduces LOS, whereas older age, comorbid conditions, and male gender were associated with increased LOS. The model demonstrated strong statistical adequacy (log-likelihood = –148.264), supported by residual diagnostics and goodness-of-fit plots. Only 5% of patients remained hospitalized beyond 14.8 days, with an average LOS of 10.4 days. In conclusion, the Weibull regression model effectively captured LOS dynamics and highlighted key risk factors influencing hospitalization duration. These findings provide evidence useful not only for hospital management but also for public health planning, optimized clinical decision-making, and resource allocation strategies for high-risk HBV patients.