BAYESIAN CONTINUOUS-TIME SURVIVAL ANALYSIS FOR RECOVERY OF TUBERCULOSIS PATIENTS IN BULUK

Abstract

Tuberculosis is an infectious bacterial disease caused by the bacilli Mycobacterium tuberculosis. The study modeled the prognostic factors for TB patients in Buluk using Kaplan-Meier, Bayesian Continuous-time Survival models, log-logistic and logistic regression models. Age, treatment time, and smear results were found to be associated with treatment outcomes. Pulmonary positive TB was identified to be the most prevalent disease category. The study unveiled that the median recovery time for patients was 171 days with average recovery of 170 and 174 days for males and females respectively. Also, the percentage recoveries for males and females were found to be 79.32% and 87.32% respectively. In addition, recovery among children, adults and the aged were 100%, 81.28% and 80.28 % respectively. The patients who reported for treatment for the first time had lower recovery rates compared to relapsed and other forms of non-new patients. There was a high HIV testing rate of 97.73% with an alarming TB/HIV co-infection rate of 10.39%. TB/HIV Co-infection was found to be associated with TB related mortality with a fatality rate of 18.75%. Finally, it was established that a unit increase in age is associated with 1.2% decrease in the odds of recovery among TB patients whilst a unit increase in treatment time is associated with 3.9% increase in the odds of recovery.