If CMS cannot persuade GRAIL of the importance of randomization and using cancer death as the outcome, then it should cancel the study. The magnitude of the problem was evident in a major National ...