As with NI and equivalence testing, statements about the relative safety of two treatments in the prospective monitoring context depend on the thresholds that one uses to define “safe.” However, little attention has been given to how these concepts can be applied to prospective drug safety monitoring in the postapproval setting