What does “evidence level” mean on a treatment page?
An evidence level is a short label for how much confidence a body of research supports, based on its design and quality. In the GRADE system, evidence is rated high, moderate, low or very low, and randomized trials start as high-quality evidence and observational studies as low. A level does not say a treatment is right for you, and different pages can use different systems. Look for what the label rests on: the design, the people studied, the comparison and the timeframe.
Quick answer
An evidence level is a short label for how much confidence a body of research supports, based on its design and quality. In the GRADE system, evidence is rated high, moderate, low or very low, and randomized trials start as high-quality evidence and observational studies as low. A level does not say a treatment is right for you, and different pages can use different systems. Look for what the label rests on: the design, the people studied, the comparison and the timeframe.
What the common labels mean
- GRADE specifies four categories (high, moderate, low, very low) applied to a body of evidence, not to individual studies. Quality means more than risk of bias: it can be lowered by imprecision, inconsistency, indirectness of results and publication bias. Source: GRADE guidelines: 3. Rating the quality of evidence.
- Hierarchy of evidence. Randomized controlled trials are usually placed at the top, controlled observational studies in the middle and uncontrolled studies and opinion at the bottom. The authors of one review note that this ordering has been challenged. Source: Hierarchy of evidence.
- A page’s own labels (for example “emerging” or “established”) may not follow either system. Check how the page defines them.
What a high-level result reports

STEP 1 as an example
As an example, STEP 1 was a randomized, double-blind trial of 1,961 adults with overweight or obesity and no diabetes, who received semaglutide 2.4 mg or placebo for 68 weeks plus lifestyle intervention. Mean body weight changed by -14.9% with semaglutide and -2.4% with placebo, and 86.4% versus 31.5% lost 5% or more. It was funded by Novo Nordisk. The result names the comparison group, the people, the time and the measure. Source: Wilding et al., 2021.
What a level does not tell you
- Whether the result applies to you: the trial population may differ from you in age, health conditions or medicines.
- How large or lasting the effect is for one person.
- Whether a product you can buy is the same as the one tested.
- Whether findings apply to your exact product is a separate question. Check its label and ask your prescriber.
Questions to ask
- What is this label based on?
- Who was studied, for how long, compared with what?
- What are the main limits?
What this does not establish
This page explains the rating terms. It does not rate any treatment. Rating a body of evidence involves judgment, and reviewers may differ.
FAQs
What does evidence level mean? A summary of how confident research findings are, based on design and quality.
Is high-level evidence a recommendation? No. GRADE separates rating the evidence from making recommendations.
Where can I check a product's information? Its label and written information from the dispensing pharmacy.
Sources
Source methods and reading limits
Sources checked against Europe PMC abstracts and product-label text: November 17, 2026. Full text was not read for abstract-only claims. Source dates and reading limits are recorded with the supporting references.