Ketamine vs esketamine (Spravato): what the trials actually compared

The two names get used interchangeably and they should not be. Esketamine is one half of the ketamine molecule, sold as a nasal spray called Spravato and FDA-approved for treatment-resistant depression. Racemic ketamine is the original mixture of both halves, given by IV, and it is approved as an anaesthetic — so using it for depression is off-label.

That produces a genuinely odd situation, and it is the reason this page exists: the approved drug has the thinner trial record, and the off-label one has the larger independent trial. Here is the evidence, study by study.

This is information to bring to the clinician managing your treatment, not medical advice, and not a reason to start or stop anything. Both drugs are used in treatment-resistant depression, where the stakes run in both directions.

The trial that made Spravato a medicine

TRANSFORM-2 is the short-term pivotal trial. In 227 adults with treatment-resistant depression, esketamine plus a newly started oral antidepressant beat placebo spray plus antidepressant by 4.0 MADRS points at day 28 — statistically significant, and a modest absolute difference. Dissociation, nausea, vertigo and dizziness were markedly more common on esketamine [1].

Study summary: TRANSFORM-2 — moderate evidence, 63/100

Two things sit in the caveat on that record, and both belong in any honest description of Spravato:

• It was the single positive short-term pivotal trial. Two sibling trials, TRANSFORM-1 and TRANSFORM-3, missed their primary endpoints — something the approval package itself acknowledged.
• Sponsor employees authored the paper.

That is not an accusation and it is not a reason to dismiss the drug. It is the context in which a 4.0-point difference on one of three trials became a marketed medicine, and patients are rarely told that part.

The maintenance trial, and what it accidentally shows

SUSTAIN-1 asked a different question: once someone is stable on esketamine, what happens if they stop? Among 297 patients already stabilised, continuing esketamine roughly halved relapse risk compared with switching to placebo spray — 27% versus 45% relapse among stable remitters, a number needed to treat of about 6 [2].

Study summary: SUSTAIN-1 — moderate evidence, 59/100

Read that as two findings, because it is two findings. It is evidence of maintenance benefit — and it is evidence that stopping esketamine carries a real relapse risk, which is a thing worth knowing before starting a treatment that is generally not framed as long-term.

The recorded caveat is important: randomised-withdrawal designs can overstate maintenance benefit, because abrupt discontinuation effects get counted as "relapse". The trial was sponsor-funded and sponsor-authored.

That funding pattern is not unusual, and it is not evenly spread across the literature. When we audited funding across our whole library, industry money concentrated in the individual trials — 14 of 48 readable ones — and was absent from the meta-analyses of randomised trials, 0 of 45 readable. SUSTAIN-1 sits squarely in the first group. The full funding audit is here, with all 235 records listed.

The strongest trial in this whole area is about the off-label drug

ELEKT-D randomised 403 patients with nonpsychotic treatment-resistant depression to IV ketamine or ECT over three weeks. Ketamine was noninferior to ECT — 55.4% response versus 41.2% — with less memory impairment [3]. It is the largest independent trial of ketamine for depression to date, and it carries the highest trust score on this page.

Study summary: ELEKT-D — gold standard, 87/100

Two limits are recorded against it and neither is hidden: the trial was open-label by necessity — you cannot mask ketamine or ECT, so subjective outcomes may drift toward whichever treatment a patient preferred — and it excluded psychotic depression, which is exactly where ECT performs best. So this is not "ketamine beats ECT". It is "for this population, over three weeks, ketamine held its own against the treatment ECT is famous for."

So which one is better? There is no head-to-head trial

This is the question everyone actually arrives with, and the honest answer is that nobody has run the study.

The nearest thing is a meta-analysis of 24 randomised trials in 1,877 participants. Both forms outperformed placebo for unipolar and bipolar depression, and indirect comparisons suggested larger response and remission effects for IV racemic ketamine than for intranasal esketamine [4].

Study summary: Comparative efficacy of racemic ketamine and esketamine — strong evidence, 80/100

The authors call that hypothesis-generating, not head-to-head, and our record says the same in its caveat: the difference rests on comparing across trials with different designs and populations, not on randomising people between the two drugs. One coauthor is an inventor on ketamine-related patents. An erratum was issued in 2021 — we checked, and it is an erratum, not a retraction.

If someone tells you IV ketamine is proven better than Spravato, this is the study they are reaching for, and it does not say that.

Where the whole field started: seven people

It is worth seeing the size of the original observation. In 2000, seven patients in a double-blind crossover received a single IV ketamine infusion, and depression scores fell by a mean of 14 Hamilton points within 72 hours, versus no change on saline [5].

Study summary: Berman 2000 — moderate evidence, 55/100

That is the first controlled demonstration of rapid NMDA-pathway antidepressant effects, and it is in our library for historical significance rather than as actionable evidence. Seven people is far too small for any clinical conclusion, and the record says so. A quarter of a century of research grew out of it, which is remarkable — and it is also a good reminder of how small the founding observation of a large field can be.

The regulatory picture, separately

Which of these is approved for what, and by whom, is its own question with its own moving parts — including how "treatment-resistant depression" is defined and what the REMS requirements around Spravato administration are. We keep that on a separate page: ketamine vs esketamine and where the FDA actually stands.

One structural note that applies here as it does everywhere in psychiatry: of the five studies on this page, TRANSFORM-2 measured at day 28, ELEKT-D ran three weeks, and Berman measured within 72 hours. SUSTAIN-1 and the meta-analysis state no duration at all in their records. These are drugs for a chronic condition, tested over weeks. The trial-duration audit has the full count.

Questions worth taking to your clinician

• Is the plan esketamine (nasal, on-label, clinic-administered) or IV ketamine (off-label)? The evidence differs, and so does the insurance situation.
• What happens if it works — how long would I stay on it, and what does stopping look like? (SUSTAIN-1 suggests stopping carries a relapse risk.)
• What is the plan for dissociation, nausea and dizziness, which were markedly more common in the pivotal trial?
• Has ECT been considered, and if it has been ruled out, why? (ELEKT-D compared them directly.)
• Given TRANSFORM-1 and TRANSFORM-3 missed their primary endpoints, what is the realistic expectation here?

Why this page shows its working

Every study summary linked above is free to read on the web with no account, and each carries a trust score calculated from recorded facts about the paper — study design, funding, journal, sample size, preregistration — recomputed on every read rather than stored. Where a study has a specific weakness, a written caveat sits next to the score. That is why the two Spravato trials show 63 and 59 while the independent comparison shows 87.

The scoring rubric is public. Browse the whole evidence library if you want to check our work.

If you are in crisis right now, please do not wait: call or text 988 (Suicide & Crisis Lifeline) or text HOME to 741741 (Crisis Text Line). Both are free, confidential and staffed 24/7.