Treatment-Regimen Estimand
Definition
The estimated treatment effect including the impact of discontinuations and adherence variability — what the drug does at the policy level, regardless of whether participants stayed on it.
Treatment-Regimen Estimand
The treatment-regimen estimand is the more conservative of the two main estimands used in modern Phase 3 trials. It estimates the average treatment effect of the assigned treatment policy as a whole, including the dilution effect of participants who discontinued treatment for any reason. This makes it a closer approximation to the real-world effect of starting a drug, because it accounts for the fact that some patients stop taking it.
The contrasting estimand is the efficacy estimand, which estimates the effect among participants who actually adhere to treatment. The efficacy estimand is typically larger in magnitude because it does not include the dilution from discontinuations.
In retatrutide’s TRIUMPH-1 trial, retatrutide 12 mg produced -25.0% mean weight loss under the treatment-regimen estimand at 80 weeks (versus -28.3% under the efficacy estimand). The 3-percentage-point gap reflects participants who discontinued during the trial. The two estimands are not in conflict; they answer slightly different questions and modern regulatory submissions present both. When comparing across trials and drugs, always compare like estimand with like — mixing efficacy with treatment-regimen numbers creates apparent differences that are an artifact of the estimand choice.