## Proportionality assumption

The Cox Hazard proportionality model has a very significant feature called
**proportionality assumption**. There are numerous ways through which we can verify that a particular model fulfills the proportionality model. However, in this case, we will examine proportionality by incorporating time-dependent covariates, such as `vcr`

and `temp`

in the `stocx`

command. Here, predictors as well as time are the interactions of time-dependent covariates. In the present case, `interaction`

is used along with `log`

because generally, this function is used more frequently, yet apart from this, one can use any other function. Note that if a time-dependent covariate is important, then it disobeys the proportionality assumption of that predictor. To summarize, we can say that all time-dependent variables support the proportionality assumption hazard:

Stcox age metfordrug j.med j.sests ccs.age j.sests nmohrr vmcr (age metfordrug medicinet ssets) temp(lmlnn(_tnt)) failure _d: censor analysis...