The edition · Psychiatry
A suicide model that finds attempts better than deaths
Machine learning on 1.9 million Army health checks sorted suicide attempts well but caught under a fifth of suicides in its top tier; alongside, text-message data in therapy, magnetic seizure therapy against ECT, a buprenorphine vial recall, and theta-burst stimulation in late-life depression.
The edition in brief
A JAMA Psychiatry prognostic study used administrative data from 1,932,243 US Army periodic health assessments to predict suicide and suicide attempts over 24 months. The model discriminated attempts well (AUROC 0.81), but suicide only moderately (AUROC 0.72), and the top 5% of predicted risk captured 18.6% of suicides. Predictors of the two outcomes differed. A randomised trial of 101 outpatients found that feeding personalised text-message data into short-term psychotherapy did not improve depression or anxiety but did prompt more clinical actions. The CORRECT-BD pilot (published May 2026) compared magnetic seizure therapy with right unilateral ultrabrief ECT for bipolar depression: remission 20% versus 30%, with less autobiographical memory loss after magnetic seizure therapy (7.1% versus 22.2%). The FDA lists an ongoing Class II recall of buprenorphine 0.3 mg/mL injection vials for crystallisation. A sham-controlled trial in 108 adults aged 60 or over (published March 2026) found bilateral theta-burst stimulation missed its week-6 endpoint but outperformed sham by week 12, with response rates of 52% against 33%.
Suicide prediction from routine health checks: good for attempts, weak for deaths
Use algorithmic suicide risk flags to widen the offer of support, never to judge that a patient is safe.
Reading patients' text messages did not improve therapy outcomes
Reviewing patients' text messages in therapy is not yet shown to improve outcomes; do not adopt it as routine.
Magnetic seizure therapy against ECT in bipolar depression
Magnetic seizure therapy may spare memory compared with ECT, but it is investigational and should not displace ECT outside trials.
FDA recall: buprenorphine 0.3 mg/mL injection vials
Check buprenorphine 0.3 mg/mL injection stock against the FDA recall and do not use vials showing crystallisation.
A risk score is not a safety plan
Assess needs and make a safety plan together; do not let a risk score decide who gets care.
Theta-burst stimulation for late-life depression: slow to start, ahead by week 12
Consider theta-burst stimulation for antidepressant-free older adults with depression, and judge it at 12 weeks rather than 6.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this one is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for psychiatry, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free