Application of Survival Analysis of TB Patients Using Parametric Model: A Case Study of General Hospital Bayara

Abstract

Aim: We evaluate the performance of parametric models, mixture of generalized gamma frailty model with Gompertz distribution and compare it with Cox proportional hazard model that is commonly used in the analysis of TB patients and also by [1]. Place and the Duration of the Study: The study was carried out in Bauchi State, Nigeria from January, 2017 to January, 2020. Methodology: In this study secondary data was used and gotten from the patients’ treatment card and TB registers from January 2015 to December 2017. The covariates used were, drug, age, marital status, smoking habit, educational level, weight, category, and risk factor. We used AIC and BIC selection tool to select the model with the lowest value and then compare it with Cox hazard model. Data analysis was done in Stata version 14. Results: The result of the analysis shows that mixture of frailty model with Gompertz baseline distribution has the lowest AIC and BIC value when compared to Cox Proportional model therefore shows a better goodness of fit for our dataset. Conclusion: We therefore conclude that mixture of frailty model with Gompertz baseline distribution model can serve as an alternative to Cox Proportional Model.