Olanzapine for chemotherapy-induced nausea and vomiting control

[...]cutoffs for non-inferiority margins are not defined by regulatory authorities and are instead subject to clinical decision making.4 Most clinicians would like to retain at least half of the benefit seen with the intervention in the pivotal trial. In the seminal olanzapine study, in the overall period, the proportion of patients with no nausea was 37% (an absolute difference of 15% vs placebo), the proportion with no clinically significant nausea was 67% (absolute difference 18%), and the proportion with complete response was 64% (absolute difference 23%).3 In another study using 5 mg olanzapine, the proportion of patients with complete control was 76% (absolute difference vs placebo 15%) and the proportion with complete response was 78% (absolute difference 14%).5 In Bajpai and colleagues’ study, the primary endpoint was complete control; however, for the sample size calculation, the rate of nausea control (37%) was considered, not the rate of complete response or the rate of no clinically significant nausea, which is more similar to complete control, both by definition and numerically. [...]we urge clinicians to read non-inferiority studies with caution.