We acknowledge that it is possible that the proportion of very late stent thrombosis using 1 minus the Kaplan-Meier curve might have biased our estimate of the number of stent thromboses, because of the competing risk for death. We have performed a new analysis using the methods suggested by Dr. Schutt ( Figure 1 ) . We have validated this using the manual approach. When the competing risk for death is considered, it does not appear to drastically modify our results. The very small difference between these curves could be explained by the low rate of events. Previous studies in this field have not taken competing risks into consideration (see references 7 and 10 in our original report). To allow readers to compare our results with those of other research groups, we used the same methods as they did. Potentially, a small difference could be explained by the fact that for many patients (with stent thrombosis), stent thrombosis was the cause of death.
