610 - The Guidance document 'Claims Based on Non-Inferiority Trials' states that "In the event the non-inferiority study failed to demonstrate superiority in the primary endpoint, one should not use secondary endpoints to suggest superiority." Does this still apply if the primary/secondary endpoints are unrelated (eg. primary endpoint is an efficacy measure, secondary endpoints are safety measures)?
Jennifer Carroll last edited by
Yes it still applies. Secondary endpoints should not be used from a failed study.