The semiparametric mixture models for the analysis of survival data of long-term survivors

Objective To introduce the semiparametric mixture models for the analysis of survival data of long-term survivors. Methods Semiparametric mixture models were applied to analyze the survival data of patients with cancer of larynx and cancer of cervix,and the results were compared with Cox’s proportional hazards regression model. Results As for cancer of larynx,only one covariate was statistically significant based on Cox’s proportional hazards regression model,while two covariates had statistically significant effects on the failure times of uncured patients,and one covariate had a statistically significant effect on the probability of being cured from the semiparametric mixture models.As for the CACX(cancer of the cervix),only one covariate was statistically significant based on Cox’s proportional hazards regression model,while two covariates had statistically significant effects on the probability of being cured from the semiparametric mixture models,and one covariate had a statistically marginal significant effect on the failure times of uncured patients. Conclusion Semiparametric mixture models have been proved to be more advantageous than Cox’s proportional hazards regression model for analyzing survival data of long-term survivors.It provides more valuable information from many aspects and has wider application prospects.