All Times EDT
Keywords: censoring, hazard ratio, IPCW, inverse-probability-of-censoring weighting, win probability, win proportion, win ratio
The win ratio method has received much attention in methodological research, ad hoc analyses, and designs of prospective studies. As the primary analysis it supported the approval of tafamidis for treatment of cardiomyopathy to reduce cardiovascular mortality and cardiovascular-related hospitalization. However, its dependence on censoring is a potential shortcoming. In this talk, we will present the IPCW-adjusted win ratio to overcome censoring issues (Dong et al., 2020). We consider both independent and dependent censoring, common censoring across endpoints, and right censoring. Our simulation studies show that, as the amount of censoring increases, the unadjusted win proportions may decrease greatly. Consequently, the bias of the unadjusted win ratio estimate may increase substantially, producing either an overestimate or an underestimate. We will demonstrate theoretically and through simulation that the IPCW-adjusted win ratio statistic gives an unbiased estimate of treatment effect.