>>n case #2, the DSMB might recommend upsizing the enrollment I think that would play havoc with the significance analysis. The case you describe is different from an upsizing following a power analysis. Peter