A large trial found ketamine was not inferior to ECT for treatment-resistant depression. That leaves a practical question for anyone facing the choice: is there anything that suggests one might suit you better?
Researchers went back into that trial's data to look. They examined whether particular characteristics measured before treatment — depression severity, thinking and memory scores, a history of attempted suicide, and whether someone was an inpatient or outpatient — were linked to doing better on one treatment than the other.
The purpose was explicitly to support shared decision-making: helping a person and their clinician choose together, rather than the choice being made purely on availability or habit.
It is worth being clear that this was a secondary analysis, meaning these questions were not the ones the trial was originally designed to answer. Findings from this kind of analysis are a guide for further research and for conversations, not a rule.
Still, the direction is encouraging. The question is shifting from "is there anything left to try" to "which of these suits this particular person". If you are at that point in your own treatment, it is a fair thing to ask your psychiatrist directly.
What the study did
The original ELEKT-D trial compared intravenous ketamine against electroconvulsive therapy (ECT) for treatment-resistant depression (TRD) — depression that hasn't responded to standard treatments — and found ketamine was 'not inferior' to ECT overall. This paper is a secondary analysis of that same trial: instead of asking which treatment is better on average, the researchers (Jha et al.) asked whether certain features measured before treatment began were linked to doing better on one treatment versus the other. Because these questions weren't planned before the original trial started, this is called a secondary, non-prespecified analysis — useful for generating ideas, but not as strong evidence as the trial's main, pre-planned result.
The study included 365 adults (aged 21–75) with nonpsychotic TRD who had been referred for ECT, at five US academic medical centres. They were randomly split into ketamine (195 people, six infusions) or ECT (170 people, nine sessions) over three weeks. Researchers looked at baseline depression severity (measured with two standard scales, the QIDS-SR16 self-report questionnaire and the clinician-rated MADRS), estimated premorbid intelligence (a proxy for cognitive functioning before illness), memory performance, history of suicide attempts, and inpatient versus outpatient status. Statistical models tracked depression scores over time, and a correction called false discovery rate adjustment was applied — this reduces the chance that, with so many comparisons being tested, some 'findings' are just due to chance.
What it found
People who started as outpatients, and those with baseline QIDS-SR16 scores of 20 or below (moderate-to-severe rather than the most severe depression), showed larger symptom reductions with ketamine than ECT. In contrast, people who started as inpatients or had very severe depression (QIDS-SR16 above 20) improved faster with ECT early on, though by the end of the three weeks both groups had similar scores.
Within the ECT group specifically, people with higher premorbid intelligence or a co-occurring PTSD diagnosis showed larger MADRS reductions. Those with impaired memory recall improved faster on ECT during week two, but again ended up similar to others by the final visit. No other differences survived the statistical correction.
Why it matters
These patterns suggest that some clinical features — particularly being an outpatient with moderate-to-severe (rather than very severe) depression — may make ketamine a reasonable first choice to discuss with a psychiatrist, while very severe or inpatient presentations might still favour ECT, at least for speed of early improvement. This kind of finding supports shared decision-making: it gives patients and clinicians something concrete to talk about, rather than choosing based on habit or availability alone.
Its limits
Because this analysis wasn't planned in advance and involved many comparisons, its findings are best treated as hypotheses for future, purpose-designed studies to confirm — not firm treatment rules. The trial was also open-label, meaning patients and clinicians knew which treatment was being given, which can subtly influence self-reported symptoms. If you're weighing treatment options for depression, please talk it through with a psychiatrist or GP who knows your situation.