Ketamine's antidepressant effect is established in unipolar depression, and the reason it has not moved into bipolar depression is fear of precipitating mania. Ket-BD, a double-blind trial at three Ontario sites, randomised 68 adult outpatients with bipolar I or II disorder in a moderate-to-severe depressive episode, MADRS of 21 or more, who had failed at least two evidence-based pharmacotherapies, to four flexibly dosed 40-minute infusions of ketamine 0.5-0.75 mg/kg or midazolam over two weeks, adjunctive to a stable mood stabiliser or antipsychotic.
At day 14, adjusting for sex, bipolar type and baseline severity, MADRS was 7.3 points lower with ketamine (95% CI -12.0 to -2.5, P = .003, Cohen's d 0.7). No case of mania, hypomania, psychosis or suicide attempt occurred in either group, with a single case of subthreshold mixed features in each.
The choice of midazolam as comparator matters: it produces sedation and dissociation-adjacent effects, making it a far more demanding control than saline. Even so, 47% correctly guessed their allocation after the first infusion, so some expectation effect remains. With 68 patients this cannot exclude an uncommon manic switch, and every participant was on a mood stabiliser or antipsychotic — which is exactly how it should be used. Within those bounds this supports offering adjunctive ketamine to a patient with genuinely treatment-resistant bipolar depression rather than withholding it on theoretical grounds.
- MADRS 7.3 points lower at day 14 (95% CI -12.0 to -2.5, Cohen's d 0.7)
- No mania, hypomania, psychosis or suicide attempt in either arm
- All participants remained on a mood stabiliser or antipsychotic
- Midazolam comparator, but 47% still guessed allocation correctly
The statistics, in plain English
A Cohen's d of 0.7 is a moderate-to-large effect, and the confidence interval, -12.0 to -2.5 points, stays clear of zero. Its width reflects the small sample: the true benefit could be anywhere from just-noticeable to very large. The safety finding is reassuring but cannot be conclusive — with 34 patients per arm, an event occurring in 1 in 50 patients would very likely have been missed entirely.
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 finding 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