A chart of weight-loss percentages can look like a league table of oral medicines. It usually combines separate experiments, however, with different participants and protocols. Before ranking the bars, check whether the products were ever randomized against one another. When they were not, differences between percentages may reflect study design as well as treatment. A careful comparison preserves that uncertainty.
OASIS studied particular semaglutide regimens
OASIS 1 tested oral semaglutide 50 mg against placebo over 68 weeks in adults with overweight or obesity without type 2 diabetes. That research dose should not be confused with the marketed 25 mg tablet formulation. OASIS 4 separately studied oral semaglutide 25 mg in adults with overweight or obesity. Identify which study an advertisement means before attaching its result to a product.
Even within a development program, the dose, duration and analysis need to stay attached to the outcome. An estimate concerning all randomized participants can differ from an estimate assuming continued treatment. Ask which analysis generated the headline number. If a chart omits that information, it is not yet detailed enough to support a confident comparative conclusion.
PIONEER asked a diabetes question
PIONEER 4 directly compared oral semaglutide with injected liraglutide and placebo in people with type 2 diabetes. Its primary endpoint concerned glycated hemoglobin at week 26; body weight was also assessed. This is a genuine randomized comparison, but its injectable comparator was liraglutide, not weekly injectable semaglutide. “Pill versus injection” would be an incomplete description.
ATTAIN is another program, not the other arm of OASIS
ATTAIN-1 tested orforglipron against placebo in adults with obesity or overweight. It did not randomize participants against oral semaglutide. A side-by-side presentation of ATTAIN and OASIS can help explain the landscape, but it cannot be described as a head-to-head trial. Research dose labels must also be distinguished from the current marketed product instructions.
Use a comparison sheet with five checks
| Check | Why it belongs beside the percentage |
|---|---|
| People | Diabetes status and other entry criteria define the population. |
| Product | Ingredient, formulation, route and study dose identify the intervention. |
| Time | A result at one visit cannot be silently treated as another duration. |
| Comparator | Placebo and active-drug comparisons answer different questions. |
| Analysis | Discontinuation, missing data and assumptions affect interpretation. |
Add a sixth column for tolerability and discontinuation, using the same care with definitions. A single effectiveness number does not describe the whole experience of a treatment. It also does not answer whether a particular person can take it safely, follow its administration conditions or sustain access over time.
Turn the comparison into a clinical question
Bring the completed sheet to the consultation and ask which trial most closely addresses the clinical goal. Ask which important differences prevent a direct comparison. The answer may remain uncertain, and that is useful information. Treatment selection can then consider eligibility, warnings, other medicines, routine and access alongside the research, instead of pretending the largest percentage settles every question.
Keep the source and analysis with any number you save or share. That makes the comparison reproducible and protects against a common error: a research result gradually losing its population, formulation and time horizon as it moves from a paper to a marketing graphic.
Sources and reading scope
- OASIS 1: Oral semaglutide 50 mg in adults with overweight or obesity
- OASIS 4: Oral Semaglutide at a Dose of 25 mg in Adults with Overweight or Obesity
- PIONEER 4: Oral semaglutide versus subcutaneous liraglutide and placebo
- ATTAIN-1: Orforglipron, an Oral Small-Molecule GLP-1 Receptor Agonist for Obesity Treatment
Research draft prepared November 7, 2026. This is educational synthesis, not medical review or an individualized treatment recommendation. Check the linked original records and current labels before making a decision.