Ketamine vs midazolam for suicidal thoughts: pooled results from 10 randomized trials

The question

In adults in an acute suicidal crisis, does a single treatment with ketamine reduce suicidal thoughts and depression more than midazolam, a sedative often used as an active comparison drug?

What they did

Researchers searched PubMed, Embase and Cochrane Central up to July 2025 for randomized controlled trials comparing ketamine with midazolam in adults with acute suicidality. Ten trials with 649 participants in total were combined. The main outcomes were suicidal thinking, measured by the suicidal ideation item of the Montgomery-Asberg Depression Rating Scale (MADRS-SI) and the Beck Scale for Suicide Ideation (BSS). Depression severity (MADRS total score) and side effects were also pooled, using a statistical method that allows for differences between trials.

What they found

Ketamine was associated with lower suicidal ideation scores than midazolam: a mean difference of -1.23 points on the MADRS suicidal ideation item (95% confidence interval -2.14 to -0.32) and -4.30 points on the Beck Scale for Suicide Ideation (95% confidence interval -8.01 to -0.59). Depression severity was also lower with ketamine, by -6.23 points on the MADRS total score (95% confidence interval -10.37 to -2.08). Side effects including nausea, emotional disturbance, derealization (a feeling of unreality) and dizziness were much more frequent in the ketamine groups.

The fine print

The pooled sample of 649 people across ten trials is still modest, and the confidence intervals are wide, meaning the true size of the benefit is uncertain. The abstract reports only short-term outcomes and does not say how long effects lasted, whether suicide attempts were reduced, or how many people had each side effect. Midazolam causes noticeable sedation but not the distinctive effects of ketamine, so blinding may have been imperfect. This summary is based on the abstract alone.

What it means

Pooled trial evidence suggests ketamine reduces suicidal thinking and depression scores more than midazolam in the short term, at the cost of more temporary side effects; how long any benefit lasts remains unclear.

Evidence grade: Moderate. A meta-analysis of randomized trials, but the pooled sample is modest, confidence intervals are wide, and follow-up is short-term.

This summary is based on the study’s published abstract.