About The Psych Brief
Every week we read the new psychiatry papers, FDA actions, drug shortages, trial results and what clinicians are discussing online, and send you only what could change your practice, with the evidence grade and the catch.
What we read each week
- New psychiatry papers indexed in PubMed, plus online-first papers from major psychiatry and general medical journals.
- Psychiatry preprints on medRxiv, always labeled as not peer reviewed.
- FDA approvals, safety communications and press releases about psychiatric drugs, and FDA shortage data.
- ClinicalTrials.gov: psychiatric trials that posted results, and phase 3 trials that stopped.
- Public clinician discussion on Reddit, Bluesky and X, used to show what colleagues are talking about. Never used as evidence.
How we grade evidence
| A | Clinical practice guideline, or a systematic review/meta-analysis of at least 20 studies |
| B | Meta-analysis of 10–19 studies, or a randomized trial with at least 300 participants and a trial registration |
| C | Meta-analysis of fewer than 10 studies, or a randomized trial with at least 100 participants |
| D | Smaller randomized trial, or an observational study of at least 10,000 people |
| E | Everything else: early signals worth knowing about, not acting on |
Preprints, and syntheses whose authors rate the evidence as low certainty, are graded one step lower. Grades describe the study design, not whether a study is right. We also flag what to watch for: small samples, missing trial registration, industry funding, observational designs.
How it is made
Drafted with AI from primary sources (journals, FDA, ClinicalTrials.gov) and clinician discussion. Every effect size is printed only after it is matched to the source text, and a clinician editor reviews each issue before it is sent. For clinicians; educational, not medical advice. Check the primary source before changing practice.
The "Result" line quotes the abstract's own sentence for the main result, and appears only when that sentence and the number in it were matched against the source. Otherwise we say "see source". Our summaries never state effect sizes of their own.