In survival analyses, competing risks are encountered where the subjects under study are at risk for more than one mutually exclusive failure event [1]. Competing risks are often analysed using either cause-specific or subdistribution (cumulative incidence) proportional hazards models. Cause-specific hazards model the rate of occurrence of an event, whereas subdistribution hazards model the risk of failure of a specific event. Results of competing risks analyses are being presented more frequently in the medical literature, but the difference in the interpretation of various estimates, compared to standard Cox hazard ratios, is rarely considered.