Application of Weibull Model for Survival of Pediatric Patients with Leukemia Cancer: Reference Center in the Central Part of Iran

Although treatment and care for children with leukemia have advanced, the survival rate for children in low- and middle-income countries remains poor. The study aimed to determine the risk factors for remission in children to develop strategies to improve long-term survival. This retrospective cohort performed from 2009 to 2018. Data of 102 children with acute leukemia collected at the Amir Kabir hospital in Arak city. Overall survival and effective risk factors found according to the life table, Cox and parametric model. The overall mortality rate of 1 year, 2 years, 3 years, 4 years and 5 years for patients were 98%, 96%, 74%, 2% and 0%, respectively. Both Weibull [HR=3.79, SE=1.99, 95%CI: 1.40-10.62] and Cox [HR=5.62, SE=3.14, 95%CI: 1.88-16.78] multiple analysis indicated that leukemia type was only identified as an independent prognostic factor in surviving patients. Besides, Weibull model (AIC=9.22) has a better performance than the Cox model (AIC=306.43). Our results suggest that among all potential factors affected only leukemia type as a detecting influence factor, and the AML group have approximately recovered four times better compared to those in the ALL group. This data strongly supported the Weibull model and also it can calculate more precise results.