COMPETING RISKS MODEL BASED ON FINE AND GRAY IN PRESENCE OF INTERVAL CENSORED DATA

Generally, survival analysis is a significant aspect of statistics that helps in anticipating possible outcomes in the various phenomena of study. A competing risk model is widely used in survival analysis since it not only studies the event of interest but also studies the other possible outcomes and this is the main topic of this research. Various models have been developed by statisticians and are widely used in examining competing risks in real-life phenomena where each model seems to have its strength and weaknesses. The Fine and Gray model is a largely employed method in competing risks analysis for its various advantages, such as the accuracy and the ability to consider multiple competing events. The main goal of this thesis is to analyze the effect of covariate on the cumulative incidence function, the Cox proportional hazards model for the subdistribution is used on both right-censored (RC) data and the model for interval-censored (IC) data. We simulate competing risks data, then we use midpoint imputation to handle the simulated interval-censored and right-censored data. In comparison to the Fine & Gray model with interval -censored data, the simulation results show that our model in this study is applicable and performs well. In additional to that both methods were applied to the MERS data set and the results of the two models show that the covariates have no effect on the cumulative incidence function