Evidence Brief — Estimand Methodology
The Oral GLP-1 Estimand Gap
Headline weight loss can move by 0.7–3.0 percentage points depending on how discontinuations and intercurrent events are handled.
Estimand delta = |efficacy / trial-product WL| − |treatment-policy / treatment-regimen WL|. A descriptive within-trial difference; no interpretation beyond that is asserted.
01 — The gap, visualized
From treatment policy to efficacy
How much efficacy disappears when an oral GLP-1 moves from an efficacy / trial-product estimand to a treatment-policy / treatment-regimen estimand?
Model-based estimates as reported by each sponsor. Sponsor-specific estimand terminology is preserved from source documents and is not normalized across programs.
02 — Methodology
Estimand pairs by trial
Exact source terminology preserved; terms are not normalized across sponsors.
03 — Reading the gap
What the data show
- Oral semaglutide shows the largest active-arm estimand delta: 3.0 pp.
- Orforglipron's active-arm gap is smaller at 1.2 pp, but placebo adjustment expands the methodological gap to 2.4 pp.
- Aleniglipron Phase 2b shows only 0.7 pp, but shorter duration and Phase 2 design limit direct inference.