Opparounds

Serial Ketamine Infusions for Treatment-Resistant Bipolar Depression: A Randomized Clinical Trial

Orsini DK, Di Luch S, Tomlinson G, et al. · JAMA Psychiatry · September 2026

This double-blind RCT randomized 63 patients with treatment-resistant bipolar I or II depression to four biweekly IV ketamine infusions (0.5-0.75 mg/kg) or active-placebo midazolam, layered on top of each patient's stable mood stabilizer or antipsychotic regimen. By day 14, the ketamine group showed a 7.3-point greater drop in MADRS scores than the midazolam group.

Notably for a bipolar sample, no cases of mania, hypomania, or psychosis emerged in either arm — the concern that most often keeps clinicians from reaching for ketamine in bipolar depression in the first place.

Read the paper →  doi:10.1001/jamapsychiatry.2026.2658

Benzodiazepine Prescriptions to Older Adults in the United States, 2015 to 2024

Olfson M, Xie F, Bushnell G, et al. · Annals of Internal Medicine · July 2026

This national pharmacoepidemiology study tracked benzodiazepine dispensing to roughly 24.9 million U.S. adults 65 and older from 2015 through 2024. Rates fell steadily from 14.1 to 11.5 fills per 100 persons between 2015 and 2019, then the decline stalled and has since plateaued.

The one segment still climbing is long-term care: dispensing through long-term-care pharmacies nearly doubled, from 0.46 to 0.85 fills per 100 persons, with the sharpest rise coming after 2020.

Read the paper →  doi:10.7326/ANNALS-25-05594

The Citalopram Dose Cap Is Really a Genotype Cap

The FDA's 40 mg/day ceiling on citalopram (20 mg in patients over 60) gets treated as a blanket population rule, but the pharmacokinetics behind it are genotype-specific. In CYP2C19 poor metabolizers, steady-state Cmax rises 68% and AUC rises 107% compared with normal metabolizers at the same dose — the FDA's own labeling caps this group at 20 mg/day for exactly that reason, and recommends avoiding citalopram entirely alongside a concomitant CYP2C19 inhibitor (omeprazole, fluconazole, fluvoxamine) in a poor metabolizer.

The QTc risk tracks the drug level, not the milligram number, which is why two patients on the "same" 40 mg dose can carry meaningfully different torsades risk. If pharmacogenomic testing comes back showing CYP2C19 poor-metabolizer status, that result should change the ceiling you use for that patient, not just sit as a note in the chart. Escitalopram carries the same CYP2C19 dependence and the same logic, just less label attention.

Concordant or Complementary — Your Countertransference Is Telling You Which Character You've Been Cast As

Racker split countertransference into two types, and the distinction changes what your reaction is evidence of. A concordant identification is when you feel what the patient feels — their anxiety becomes your anxiety, their shame becomes your shame — because you've identified with their experiencing self.

A complementary identification is different and often more diagnostically useful: you find yourself feeling like someone else in the patient's internal world — the critical parent, the abandoning partner — because the patient has unconsciously cast you in that role. The same session can produce both, sequentially, and confusing one for the other leads to the wrong intervention: treating a complementary identification as if it were your own empathic attunement mistakes the patient's internal object for your countertransference proper.