The AI Psych Brief

Sex differences in the efficacy, acceptability, and tolerability of antipsychotics in people with acute schizophrenia: individual participant data pairwise, dose-response, and network meta-analyses

Furukawa Y, Siafis S, Schneider-Thoma J, Nomura N, Fares-Otero NE, Illing S, Storosum BWC, Zantvoord JB, Schoretsanitis G, Schwarz C, Schulz J, Naiser H, Möhrmann KH, Priller J, Davis JM, Sommer IE, Efthimiou O, Salanti G, Leucht S · The Lancet Psychiatry · September 2026

Leucht's Munich group pulled individual participant data from industry trials shared through Vivli and YODA. They identified 33 trials and ran the main analyses on the 18 fixed-dose ones: 7,062 adults with acute schizophrenia (2,168 women) randomised to haloperidol, olanzapine, paliperidone, risperidone, or placebo. Working from individual-level data let them test sex-by-treatment interactions directly instead of inferring them from trial-level averages.

On the primary outcome, PANSS total at about week 6, there was no sex difference in drug effect (difference in SMD 0.04, 95% CI -0.09 to 0.18). Dose-response curves were essentially the same for women and men, both plateauing at 3–5 mg risperidone equivalents. Two differences did hold up. Antipsychotics cut all-cause dropout more in women (ratio of odds ratios 0.74), and they raised prolactin far more in women (difference in SMD 0.83, rated high credibility).

Read the paper →  doi:10.1016/S2215-0366(26)00280-4

Trust and respect declines and recoveries in psychotherapy: Clinical impact of relational feedback

Connolly Gibbons MB, Kodkany N, Liow Y, Crits-Christoph P · Psychotherapy · September 2026

This University of Pennsylvania trial took place in US community mental health clinics. Therapists were randomised to receive either weekly symptom feedback alone or symptom feedback plus their patients' ratings of how much they trusted and felt respected by the therapist. Rather than coding ruptures from tapes, the authors defined a meaningful decline as a session-to-session drop in a patient's trust/respect rating larger than that patient's usual variability, and a recovery as a later return to the pre-decline level. They tracked weekly functioning over 11 weeks; 162 of 233 randomised patients had enough data for this analysis.

In the symptom-only condition, a trust/respect decline that never recovered went with worse functioning for that patient. When therapists could see the trust/respect ratings, that link was attenuated. Whether the decline recovered did not differ between conditions.

Read the paper →  doi:10.1037/pst0000641

When buprenorphine precipitates withdrawal, the fix is usually more buprenorphine

Buprenorphine binds the mu receptor with very high affinity but is only a partial agonist. When it displaces a full agonist, the patient's net opioid effect drops sharply, and that drop is precipitated withdrawal. Fentanyl has made this more common. It is highly lipophilic and builds up in fat with chronic use, so it keeps leaving tissue stores long after the last use. A COWS score that looks adequate doesn't guarantee the receptors are clear.

The instinct after precipitated withdrawal is to stop buprenorphine. That leaves the patient stuck in the trough. Giving additional buprenorphine (often another 8–16 mg) pushes occupancy toward its ceiling, where the partial-agonist effect can carry the patient, and withdrawal usually resolves within hours. Adjuncts like clonidine and antiemetics treat the symptoms but not the occupancy problem.

To avoid it in the first place, use low-dose initiation: start with sub-milligram buprenorphine doses while the full agonist continues, then cross-taper over several days. That way buprenorphine occupancy builds before the full agonist is withdrawn.

Erotic transference is a message to decode, not a feeling to manage away

When a patient declares love or desire for the therapist, trainees tend to fall into one of two errors. One is quietly gratifying it (warmth that turns flirtatious, extra time, a pleased self-disclosure). The other is fleeing it (an abrupt boundary speech, a referral, or suddenly going clinical). Both act on the feeling instead of thinking about it, and both tell the patient it is too dangerous to put into words. Freud's position still holds: the love is real as an experience, and the frame is what makes it workable. The task is to hold the boundary and stay curious at the same time. Ask what the longing is doing right now. It may protect against envy or dependency, test whether the therapist can be seduced (and so whether anyone can be trusted), or replay an early relationship where being special was the only kind of safety. Its timing is often the best clue. Erotic feelings commonly intensify right after something painful came close, or just before a break.