Application of cure models in survival analysis of thalassemia major disease

Identification of the factors influencing short term and long term survival for patients is one of the important concerns in medical society. With determining the risk factors and also health factors of a disease, valuable information can be obtained. In some cases, such as high rate of censoring, the use of standard survival models for analysis may cause to loss in some of information. The purpose of this study was to apply the cure models in identifying the risk and health factors of patients’ thalassemia major.

The data of this historical cohort study with 296 patients with thalassemia major, were gathered during 1994- 2013 in Zafar Clinic in Tehran. Cure models were used for survival analysis of these patients and determination of its risk factors. Data were analyzed using Stata v.12 and R3.0.2 software.

Mean (±SD) of patients’ age was 29.11 ± 0.47year. Of the patients, 51.7% were women.. The minimum value of Akaike information criterion was for cure non mixture log – normal model with logit link function.

According to the results, the cure non mixture log – normal models is chosen as the best model for identification the risk factors of thalassemia major disease.