True, safety first. Then, how big an effect are you targeting? Remember statistical precision only grows with the square root of patients: 4x the patients ==> 2x the precision. So, all else equal, a 1600 person trial will have the same confidence in detecting, say, a 20% reduction in death rate, as a 400 person trial has in detecting a 40% drop,

My expectations for leronlimab relative risk reduction in S/C begin at 40%.