Survival analysis in tuberculosis patients of exponential accelerated failure time model

The survival analysis can be used in time to event data, this one of the tools to estimate the probability of survival in patients. To estimate the patient's survivorship based on time-independent variable this called nonparametric model. Acceleration failure model is considered the parametric model, it takes covariate and multiple effects of survivorship, is measured through a log-linear model taking the logarithm of survival time and the outcomes of the dependent variable. The model assumes that follow survival function is known as assuming the effect of a covariate is to accelerate or decelerate the life of patients by important constant. Hence the AFT model is alternative to proportional hazard models because in this model analysis the effect of a covariate to multiply the hazard by the constant. This study discusses the survival time of patients by acceleration failure time model in TB patients, like the variable is age, Regimen, sex, and weight are considered. We have been check the model fit from the failure distribution, whether it is fitting for the model in that distribution based on deviation method (-2LL) using partial log-likelihood functions.