In a new guidance released last week, the FDA supplements the International Council on Harmonization’s (ICH) E9 guideline on statistical analysis with clarification of how clinical trials should define and analyze a drug’s treatment effect.

The treatment effect, also known as an estimand, observed in a trial should dictate the trial design to be used, the guidance says, adding that one trial design may be suitable for one estimand but not others. Certain estimands may warrant or benefit from alternative trial designs, including run-in, enrichment, randomized withdrawal or titration designs.

Trial designs should take into consideration that there may be multiple estimands being studied, the FDA says, each of which would need statistical testing and information.

Read the addendum here: https://bit.ly/33FCUpA.