Peer review is how most scientific journals decide what to publish. Editors send a submitted paper to independent researchers in the field, who judge its methods, analysis and conclusions and ask for changes. A paper that passes is described as peer reviewed.
It is a filter, not a guarantee. A 2024 BMJ meta-analysis of psilocybin for depression passed peer review, then was corrected after errors were found in how effect sizes were calculated and in data counted more than once (BMJ correction, 2025). Microdosing research has had a published author correction too (Rootman et al., 2022, author correction). Corrections are a normal part of the record, and readers should check for them.
Work that skips peer review, such as a preprint, company press release or book, can still be informative, but needs more caution. For example, the phase 3 results reported for COMP360 come from the company’s own filings (Form 10-Q, 5 August 2026). This site’s journal labels which sources are peer reviewed and which are not.
Sources
- BMJ correction to Metaxa and Clarke, 2025
- Author correction to Rootman et al., 2022
- COMPASS Pathways Form 10-Q, 5 August 2026
Last checked .