Survival Analysis of Patients with Beta-Thalassemia Major in Isfahan Province

Abstract

Beta-thalassemia major is one of the most prevalent forms of anemia that has become increasingly widespread in recent years. If not diagnosed and managed in a timely manner, the disease may lead to patient mortality. The application of appropriate statistical models, such as survival analysis in affected patients, can provide an effective approach for identifying and determining the factors influencing the risk of death among these patients. The present study aimed to identify the best-fitting survival model, using conventional survival analysis methods, to determine the risk factors associated with mortality in patients with beta-thalassemia major. This study utilized data from 190 patients with beta-thalassemia major who attended a treatment center in Isfahan between 2011 and 2021. Using the R statistical software, exponential, Weibull, log-normal, log-logistic, and Gompertz models—with and without frailty terms—as well as the semi-parametric Cox model were fitted to the data. Based on the Akaike Information Criterion (AIC) and a significance level of 0.05, the log-normal model was identified as the best-fitting model, whereas the Gompertz model demonstrated the poorest fit. Accordingly, the parametric log-normal model was selected and recommended for the survival analysis of beta-thalassemia major, and the effects of the relevant variables were subsequently estimated.