- Design
- Randomised, double-blind, midazolam-controlled, 3 sites
- Population
- 68 adults with treatment-resistant bipolar I or II depression on a mood stabiliser or antipsychotic
- Primary outcome
- MADRS change from baseline to day 14
- Effect
- Difference -7.3 points (95% CI -12.0 to -2.5); d 0.7; no mania or psychosis
Ket-BD was a double-blind trial at three Ontario sites. 68 outpatients aged 21-65 with bipolar I or II disorder, a current moderate to severe depressive episode (MADRS 21 or more) and at least two failed evidence-based treatments were randomised to four 40-minute infusions over two weeks of ketamine (0.5-0.75 mg/kg) or the active placebo midazolam, on top of a mood stabiliser or antipsychotic.
At day 14, MADRS was 7.3 points lower with ketamine (95% CI -12.0 to -2.5; Cohen d 0.7). No mania, hypomania, psychosis or suicide attempt occurred in either group; one case of subthreshold mixed features appeared in each.
Ketamine's evidence has come almost entirely from unipolar depression, and fear of switching has kept it away from bipolar patients. This trial gives a controlled estimate of benefit and reassuring early safety data in exactly that group.
It is small, and it measured two weeks. It supports ketamine as an option in specialist services for bipolar depression that has failed standard treatment, with a mood stabiliser or antipsychotic in place, not as a routine step.
- Confirm at least two adequate failed trials of evidence-based bipolar depression treatments first.
- Keep a mood stabiliser or antipsychotic in place, as every participant had.
- Screen for mixed features and rate hypomanic symptoms before and after each infusion.
- Plan what happens after the course; the trial did not test maintenance or relapse.
Why it matters
It challenges the assumption that ketamine is too risky for switching to use in bipolar depression.
Don't overread it
With 63 people analysed over 14 days, the trial cannot rule out uncommon switches or say how long the benefit lasts.
The statistics, in plain English
A 7.3-point MADRS difference is clinically meaningful, and a Cohen d of 0.7 is a moderate-to-large effect. The confidence interval is wide, from 2.5 to 12 points, which is typical of a small trial. Midazolam was used so that both groups felt sedated, but 47% guessed their allocation after the first infusion, so blinding was imperfect.
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