Wegovy pill trials: OASIS 4 and the limits of the evidence

What OASIS 4 tested

OASIS 4 was a randomised, double-blind, placebo-controlled study of 307 adults, with 205 assigned to oral semaglutide. Participants had obesity, or overweight with a weight-related condition, and received support with diet and activity. The main weight measurement came at 64 weeks.

Randomisation reduces systematic differences between the treatment groups. Blinding limits the effect of expectations. Neither feature makes the participants representative of every patient. Not everyone stayed on treatment, and a trial of this size cannot pick up every rare side effect.

Keep tablet and injection evidence apart

The UK tablet SmPC also discusses trials of injected semaglutide. Those sections add to what is known about the molecule. Keep the way the medicine was given and the people studied in view. A weekly injection study is no head-to-head comparison with the oral tablet, even though both contain semaglutide.

When you check a claim, look for the formulation, dose, number of participants, length of follow-up and how the outcome was defined. Claims about heart events, heart failure or the liver need the right licensed use and evidence. The brand name alone proves nothing.

The question behind an average

An intention-to-treat analysis keeps the benefit of randomisation by counting people in the group they were assigned to. Other analyses estimate what would happen if treatment continued as planned. Missing data, stopping treatment and rescue therapy complicate both. The statistical method decides how these events are handled.

Ask which question each analysis answers and how closely that question matches your decision. Ignore which one gives the biggest number. If you are deciding whether to start, the chance of stopping because of side effects matters to you as much as results among people who kept going.

What one trial cannot settle

A moderate-sized study is poorly placed to measure rare harms. A short study cannot prove safety or lasting results over decades. People with certain conditions were excluded, so uncertainty remains for those groups. These limits leave the study valid and mark where its conclusions stop.

Look for disclosure of manufacturer funding, trial registration, methods and publication status. An independent critique helps interpretation, but independence alone does not make a claim correct. The soundest approach links to the original evidence, describes the uncertainty and avoids ranking medicines by mismatched headline numbers.

An outcome chosen before a trial starts carries more weight than a pattern spotted afterwards. Exploratory subgroup findings raise useful questions, but small groups are prone to chance differences. A careful summary names the primary endpoint and labels exploratory results as exploratory.

Questions for your appointment

When you read a claim, ask whether the figure comes from OASIS 4, a different oral dose or an injection study. Check whether the percentage includes people who stopped treatment, and whether the outcome measured was weight, symptoms or a real clinical event.