Dr. Ramkumar.T Balan, Hassan Swedy Lunku, Ismail Juma
Abstract: The aim of this study was to evaluate a semi-parametric survival model on time of survival of breast cancer patients and a case study is considered for describing the model. The study is on survival times of breast cancer patients treated at ORCI which may help to reduce breast cancer outcomes. Data of socio-demographic characteristics of patients, their reproductivity, stages of disease, treatments undergone and follow up of the treatment were considered to develop the model. Cox proportional hazard regression model is adopted considering the covariates incurring in the system. The Wald statistics evaluates whether the coefficients of given variable is statistically significant and found that stage of disease was significant while other covariates -history of multiple sexual partners, smoking cigarette, alcohol use, cancer grade fails to be significant.Also Kaplan Meier Estimate is derived and median survival time is estimated. The results of this paper showed that non parametric and semi parametric survival models were better to predict survival time of breast cancer patients.
Keywords: non parametrics, semi parametric, survival model, breast cancer