Plasma Exchange and Glucocorticoids in Severe ANCA-Associated Vasculitis

To the Editor: Walsh et al. (Feb. 13 issue)¹ designed the PEXIVAS trial on the basis of a median time to event of 6 years,² and the trial sample size was increased after an interim analysis. When the event rate is lower than expected, there will be more censoring than expected, and the assessment of the comparative event rate for a threshold relative hazard can be informative.³ The lower-than-expected rate in this trial was probably due to a period effect from contemporary immunosuppression regimens but could also have resulted from the enrollment of participants who were at lower risk for . . .