The next step evaluated if the performance of each of the covariates was similar across all of the time frames, i.e., the proportionality assumption, by comparing a model with time-varying covariate effects (non-proportional) to a model that constrained the covariate effects to equality across all time points (proportional). None of the variables yielded a significantly better model fit when allowed to vary across time for either outcome. The parameter estimates (“Est” in Tables 2 and 3 which represent the log hazard odds) and corresponding hazard odds ratios (with 95% confidence interval) for the univariate effects of all potential covariates (assumed to be proportional) are listed in Table 2.
