Since the data specifies death as a competing risk, the cox analysis must be adjusted. Death will now, as it was before, treated as a censor.
Split HR and CI, print logrank values.
First univariate with covariates coxph models.
