Friday, October 16, 2026

The AI Psych Brief

🔬 Section 1

Psychiatric Research Article

Psychological Interventions Targeting Auditory Hallucinations in Persons With Psychotic Disorders: A Systematic Review and Meta-Analysis

Fässler L, Koop S, Opper F, Bighelli I, Leucht S, Sabé M, Bajbouj M, Knaevelsrud C, Böge K · JAMA Psychiatry · October 2026 (online June 2026)

This preregistered meta-analysis pooled 23 randomized trials (2,016 adults with schizophrenia-spectrum disorders and current voices) of psychological treatments aimed specifically at auditory hallucinations: CBT, avatar therapy, relating/dialogue therapy, digital programs, and integrative approaches. Overall, the targeted therapies reduced hallucination severity modestly (Hedges g = −0.18, NNT 16) and reduced frequency, distress, delusions, depression, and anxiety, with effects persisting at follow-up. Avatar therapy was the most consistent approach (g = −0.37 for severity, still significant at follow-up), while voice-focused CBT, dialogue, and digital subgroups did not separate from controls on their own. Beliefs about the voices' malevolence did not change, dropout was similar to controls, and GRADE certainty was low.

💡 Why it matters

For a patient whose voices persist despite an adequate antipsychotic trial, a targeted psychological intervention adds a small but durable benefit, and avatar therapy has the strongest current signal. Set expectations about frequency and distress rather than elimination.

Read the paper → doi:10.1001/jamapsychiatry.2026.1443

🩺 Section 2

General Medicine Article

Dexmedetomidine Use and Intensive Care Unit Admissions for Patients With Medetomidine Withdrawal

Boyce TG, Thakrar AP, Durkin BR, Casey ER, Xu L, Lowenstein M · JAMA Internal Medicine (Research Letter) · September 2026

Medetomidine is a veterinary α2-agonist sedative that has spread through the illicit fentanyl supply since 2024. Its withdrawal syndrome causes severe tachycardia, hypertension, vomiting, agitated delirium, and tremor, responds poorly to opioids, and settles with α2 agonists over three to five days. Penn investigators used IV dexmedetomidine as a marker of severe withdrawal and tracked it among hospitalized patients with opioid use disorder at two Philadelphia hospitals from 2020 through September 2025. According to the authors' institution, dexmedetomidine use rose from just under 6% of OUD admissions before medetomidine arrived to about 20% afterward, reaching about one in three in the final quarter studied. ICU management rose in step, because dexmedetomidine infusions require monitoring for bradycardia and hypotension.

💡 Why it matters

On a C-L or addiction consult, delirium with hypertension and tachycardia that does not respond to opioids in a patient using fentanyl should raise suspicion of medetomidine withdrawal; early, escalating α2-agonist treatment and continued MOUD (methadone or buprenorphine) are the core of care, and an unrecognized case can be mistaken for primary agitation or psychosis.

Read the paper → doi:10.1001/jamainternmed.2026.4568

💊 Section 3

Psychiatric Fact

Haloperidol decanoate takes months to reach steady state unless you load it

Haloperidol decanoate's apparent half-life is about three weeks, so on a standard monthly maintenance dose (roughly 10 to 15 times the daily oral dose) it takes three to five months to reach steady state. Stopping oral haloperidol on the day of the first injection leaves the patient underdosed for weeks. That early relapse is often blamed on the depot itself.

The alternative is a loading strategy (Ereshefsky): give about 20 times the stabilized oral daily dose in the first month, split into injections of no more than 100 mg a few days apart, then drop to 10 to 15 times the oral dose monthly. This reaches near-steady levels within the first month and lets you taper oral haloperidol quickly. If you don't load, plan an oral overlap for the first several weeks and expect EPS to show up late, because levels keep rising for months after the dose was set.

🛋️ Section 4

Psychotherapy Teaching Pearl

Opposite action only works when the emotion doesn't fit the facts

Opposite action is often taught as "do the opposite of what you feel," which makes it sound like willpower. In DBT it is narrower and more diagnostic. It is the skill for an emotion that does not fit the facts, or whose intensity or duration doesn't fit them. If the emotion does fit (fear of a genuinely dangerous partner, guilt over a real violation of one's values), the indicated skill is problem-solving, and opposite action would be invalidating. So the clinician first runs a check-the-facts step, then has the patient act opposite to the emotion's action urge all the way: posture, voice, and thoughts included. Half-measures fail. Approaching a feared situation while still mentally avoiding it keeps the emotion alive. The skill is repeated until the emotion's intensity actually falls.

🗒️ Vignette

A 22-year-old in DBT skills group has stopped going to her lab section after a TA corrected her answer in front of others. She feels intense shame and the urge to hide. The resident first checks the facts: does shame fit, meaning would she actually be rejected if others knew? She concedes the TA corrected three other students that week. Shame does not fit, so they plan opposite action all the way: attend the next section, sit in the front row instead of near the door, make eye contact, and ask one question aloud, rather than attending while slumped and silent. The resident predicts that the shame will spike first, then fall. At the next session, she reports it dropped by half after the second class.