Used in prevalence to determine which covariates should be sampled when
simulating an incident population. This should provide a character vector containing
column names of the original registry data set input to prevalence that
are used by the survival model.
extract_covars(object)The survival object itself
A character vector holding the column names. Can be NULL if no covariates are used i.e. the survival model is built based on population level data rather than individual level data.