“…Ouwens et al 16 focused instead on the "area-under-the-curve" to compare different models, equivalent to estimating restricted mean survival time, and also presented plots of the estimated hazard function to inform model selection. Klijn et al 17 compared estimates of the conditional survival between models as an alternative means of judging model fit. Nevertheless, there is no established consensus on how to choose the best-fitting models, in spite of there being a large number of studies concerning how to undertake survival analysis for HTA.…”