Question

Explain the concepts underlying the cox regression model.

Answer 1

Concepts underlying Cox Regression

We know that when we need to study or find the relationship between the response variable and several explanatory variables, then we use the multiple linear regression model. But, if the response variable is survival time, then multiple regression model is not good or appropriate for the prediction of response variable survival time. The response variable survival time have special nature of survival data. Survival data consist of censored data and this data is highly skewed in nature. Due to this reason we cannot use the multiple regression model for the prediction of response variable survival time.

For the prediction of survival time, there are so many regression models were suggested by different researchers. Out of these regression models, one of the best regression model for the prediction of survival time is nothing but the Cox regression model which is developed by Cox in 1972. There are also some other regression models such as Weibull regression for the prediction of response variable survival time.

In survival analysis, we study the response variable survival time which is the time between time origin and end point. The end point of this data set is either death or failure or end of subjects participation in the study. In the Cox regression method, we are dealing with the two types of problems. First problem of survival data is often positively skewed nature of the survival data. Second problem of the survival data is that portion of the data are censored.