Median values ought to be banned from all serious discussions of biostatistics, including those occurring at FDA advisory panels. From a mathematical standpoint, medians have no more inherent worthiness than any other arbitrary cross-section of the K-M curves.