Placebo-adjusted effect (revision 19)
Old revision·23:00, 6 Oct 2025·INN_Ingrid
| Placebo-adjusted effectTrial methodology | |
|---|---|
| Also called | Between-group difference, treatment difference |
| Computed as | Active-arm change minus placebo-arm change |
| Why it matters | The active-arm change alone is not the drug effect |
| Topic infobox · conventions | |
The placebo-adjusted effect is the difference between the change observed in a treatment arm and the change observed in the comparator arm. It is the quantity a randomised trial is designed to estimate, and it is what can be attributed to the treatment.[1]
The active-arm change alone conflates the treatment effect with everything else operating during the trial: regression to the mean, the natural course of the condition, the lifestyle intervention both arms received, and the effects of being observed.[2]
In obesity trials the distinction is substantial. STEP 1 reported −14.9% on semaglutide and −2.4% on placebo, so the placebo-adjusted effect is about 12.4 percentage points — not 14.9.[3]
What the comparator arm captures
[edit]A placebo arm in a modern obesity trial is not a no-treatment arm. Participants receive the same lifestyle counselling, the same visit schedule and the same measurement, and the difference between the arms is the treatment alone.[2][4]
Several distinct phenomena contribute to the change in a placebo arm. Regression to the mean operates because participants are enrolled at a moment when their measurement was high enough to qualify. The trial's own intervention — diet advice, activity advice, regular weighing — has an effect. Observation changes behaviour. And the natural course of a condition may be towards improvement or deterioration.[2]
The placebo response proper — improvement attributable to expectation — is only one of these, and in objective endpoints such as body weight it is generally small relative to the others.[2]
How figures are misreported
[edit]| Reported as | Actually | Effect on the impression |
|---|---|---|
| "Lost 15% of body weight" | Active-arm change | Overstates the drug effect |
| "12 percentage points more than placebo" | Placebo-adjusted | Correct |
| "Six times the placebo effect" | Ratio of arms | Sensitive to a small denominator |
| "Up to 24%" | Highest dose, best case | Not the trial's headline result |
The third row deserves attention: a ratio between arms is unstable when the placebo change is near zero, and a trial with a placebo change of −0.5% will produce a spectacular ratio from an unremarkable difference.[1]
Cross-trial comparison compounds all of this. Two trials with different lifestyle interventions have different placebo arms, so their active-arm figures are not comparable even when the drug effects are.[3]
Where the adjustment is not enough
[edit]Placebo adjustment removes what the comparator arm experienced. It does not remove differences in population, in duration, in escalation schedule, or in analysis approach — which is why a placebo-adjusted figure from one trial is still not directly comparable with one from another.[1]
Nor does it address blinding failure. An agent producing marked gastrointestinal effects is not perfectly blinded in practice, and participants who infer their allocation may behave differently. This is unquantifiable and is a general limitation of trials of agents with distinctive effects.[2]
For a reader, the practical discipline is to ask three questions of any quoted figure: is it placebo-adjusted, over what duration, and in what population. A figure lacking any of the three is not interpretable.[1]
See also
- Intention-to-treat analysis
- Number needed to treat
- STEP trial programme
- SURMOUNT trial programme
- Randomised controlled trial
References
- ^ a b c d International Council for Harmonisation, E9(R1): Estimands and Sensitivity Analysis in Clinical Trials (2019).
- ^ a b c d e Hróbjartsson A, Gøtzsche PC. "Placebo interventions for all clinical conditions." Cochrane Database of Systematic Reviews (1):CD003974 (2010). PMID 20091554.
- ^ a b Wilding JPH, Batterham RL, Calanna S, et al. "Once-weekly semaglutide in adults with overweight or obesity." New England Journal of Medicine 384(11):989–1002 (2021). PMID 33567185.
- ^ International Council for Harmonisation, E10: Choice of Control Group and Related Issues in Clinical Trials (2000).