Survival Models By Non-Parametric And Semi-Parametric Methods For Patients Infected With Coronavirus In Al-Kindi Teaching Hospital

ABSTRACT

A study of the non-parametric survival model (Kaplan-Meier) and the semi-parametric Cox Regression model. From the practical side, it was found that the effect of the change of age by (3.483) when the patient's age was transferred from one age group to another on the estimation of the survival function by semi-parametric method using the (Cox Regression) model. From the comparison between the models of survival (nonparametric, semi-parametric) from the mean squares of relative error (RMSE) statistics, it was found that the best model for estimating the survival function is the nonparametric model (Kaplan-Meier). The study came out with several results, the most important of which is that by estimating the survival function by the nonparametric method (Kaplan-Meier), it is possible to obtain the lowest cumulative risk rate for each survival time. This means that the probability of the patient staying in the time period (t) increases and that the risk rate is affected by the change in the patient's age and duration of stay when estimating the survival and cumulative risks by the semi-parametric method (Cox Regerssion).