Low-dose ketamine does not improve the speed of recovery from depression in electroconvulsive therapy: a randomized controlled trial

Electroconvulsive therapy (ECT) is an established, highly effective treatment for major depressive disorder, with reported remission rates up to 87%, but its precise mechanism remains uncertain. Routine practice uses a short-acting general anaesthetic such as propofol to ensure patient comfort and safety; however, propofol is a potent antiepileptic that can shorten seizure duration, a factor that might reduce ECT efficacy. Ketamine, a dissociative anaesthetic with N-methyl-D-aspartate (NMDA) receptor antagonist properties, has shown rapid antidepressant effects at low, sub‑anaesthetic doses in outpatient settings and can prolong seizure duration when used as an induction agent, but higher doses are associated with haemodynamic and psychotomimetic effects. Woolsey and colleagues designed this trial to test whether adding low-dose intravenous ketamine to a propofol-based anaesthetic regimen for ECT would accelerate clinical improvement in depression compared with placebo. The primary hypothesis was that ketamine would reduce the number of ECT treatments required to achieve a 50% reduction in depressive symptoms as measured by the Montgomery–Åsberg Depression Rating Scale (MADRS); secondary hypotheses addressed more modest symptom reductions, global severity, physiological parameters, seizure duration, and safety outcomes.

This was a double-blind, randomised, placebo-controlled trial conducted at the University of Alberta Hospital. The protocol included an interim analysis and an adaptive dosing plan: the first 14 participants were randomised to receive ketamine 0.2 mg/kg or saline; if no signal emerged at interim analysis the ketamine dose was to be increased to 0.5 mg/kg and additional participants recruited. In practice, after the interim analysis showed no difference, a further 31 patients were randomised with ketamine at 0.5 mg/kg or saline. A total of 48 patients consented, and 45 completed the study. Inclusion criteria targeted adults aged 18–70 referred for ECT with a DSM-5 diagnosis of major depressive disorder and baseline MADRS ≥ 24; participants were required to be in a first or new depressive episode of no more than 3 months' duration and judged by a psychiatrist to require ECT. The protocol listed multiple exclusions, including recent ECT, significant medical instability, substance dependence (unless in full remission), pregnancy or lactation, and inability to provide informed consent. Allocation concealment used sequentially numbered opaque sealed envelopes prepared by the study coordinator and handled by the research pharmacy. Patients, treating psychiatrists, anaesthesiologists, the study coordinator, and the statistician were blinded to allocation. The pharmacy prepared syringes of either ketamine (0.2 mg/kg for participants 1–14, 0.5 mg/kg thereafter) or an equivalent volume of normal saline; the anaesthesiologist administered the study syringe immediately followed by propofol and succinylcholine. Anaesthetic dosing and adjunct medications were titrated by the attending anaesthesiologist to achieve adequate anaesthesia for each ECT treatment. Procedures followed an institutional course of up to 12 ECT treatments given three times weekly (Monday, Wednesday, Friday) using a MECTA SpECTrum 5000Q device with titrated stimulus dosing per standard unilateral ultrabrief and bilateral protocols. Primary outcome was the number of ECT sessions required to reach a 50% reduction in MADRS. Secondary psychiatric outcomes were number of sessions to a 25% MADRS reduction and change in Clinical Global Impression–Severity (CGI‑S). Non‑psychiatric outcomes included heart rate, blood pressure, oxygen saturation, respiratory rate, temperature, seizure duration (visually and by EEG), anaesthetic doses, and adverse events collected at every session. Data were recorded in REDCap. Sample size calculations aimed for 14 subjects per group to achieve 80% power based on prior ketamine infusion literature and assumptions about response rates to single ECT sessions. Analyses were intention-to-treat. Psychiatric endpoints (MADRS, CGI‑S) were compared using independent samples t-tests with Levene's test for variance; non‑psychiatric outcomes used independent samples t-tests. Results are reported as mean ± standard deviation (SD) and p < 0.05 was considered significant. The interim analysis assessed only psychiatric endpoints.

Forty-eight patients consented between October 2015 and May 2019; one was excluded after enrollment and two withdrew from ECT, leaving 45 completers included in the analysis. Seventy-three per cent were women, mean age was 42.9 years, and mean baseline MADRS was 37.1. The initial 14 patients were randomised 7:7 to ketamine 0.2 mg/kg or saline; after an interim analysis showed no difference, the remaining 31 patients were randomised to ketamine 0.5 mg/kg (n = 16) or saline (n = 15). No significant differences emerged on the primary or secondary psychiatric endpoints. In the post‑interim (final) analysis the ketamine group required a mean of 8.25 ± 2.72 ECT sessions to achieve a 50% reduction in MADRS compared with 7.73 ± 2.89 sessions in the placebo group (p = 0.61). For a 25% MADRS reduction the ketamine group required 4.25 ± 1.52 sessions versus 5.47 ± 2.95 in the placebo group (p = 0.34). Final MADRS scores at the end of the ECT course were 13.11 ± 4.83 for ketamine and 10.83 ± 6.86 for placebo; final CGI‑S scores were 1.67 ± 1.23 and 1.58 ± 1.17, respectively. Interim (pre‑interim) subgroup data also showed no significant differences (for example, final MADRS 11.17 ± 9.04 vs 10.71 ± 7.93 in ketamine vs placebo, respectively). Among non‑psychiatric outcomes, maximum heart rate was higher in the ketamine group (98.64 ± 8.06 bpm) than placebo (90.32 ± 12.77 bpm), reaching statistical significance (p = 0.037). Mean arterial pressure did not differ significantly (112.64 ± 11.75 vs 110.70 ± 7.98 mmHg; p = 0.60), and when normalized to baseline there were no significant differences in maximum heart rate or maximum mean arterial pressure. Propofol dosing showed no difference in the interim 0.2 mg/kg analysis (1.21 ± 0.47 vs 1.25 ± 0.24 mg/kg; p = 0.85), whereas the 0.5 mg/kg analysis revealed a trend toward lower propofol requirements in the ketamine group (1.10 ± 0.31 vs 1.36 ± 0.39 mg/kg; p = 0.053). Seizure duration did not differ meaningfully (35.74 ± 11.82 sec for ketamine vs 35.09 ± 6.97 sec for placebo; p = 0.86). Maximum post‑ECT respiratory rates were similar (23.71 ± 1.90 vs 23.20 ± 2.22; p = 0.50). No unexpected adverse events were reported.

Woolsey and colleagues interpret these findings to indicate that adding low‑dose ketamine to a propofol‑based anaesthetic regimen for ECT does not accelerate clinical recovery from depression. Although there was a trend toward reduced propofol requirements with ketamine at 0.5 mg/kg, this did not translate into shorter seizure duration or improved MADRS or CGI‑S outcomes. Ketamine was well tolerated at the doses used and did not produce unexpected haemodynamic instability. The authors situate their null findings within a heterogeneous literature: some trials and meta‑analyses report early benefit from ketamine adjuncts to ECT, whereas others find no symptomatic advantage and some report increased risk of side effects such as delirium or psychotomimetic phenomena. They highlight methodological distinctions of their study — starting at a very low ketamine dose (0.2 mg/kg) with a predefined interim analysis and dose escalation to 0.5 mg/kg, comprehensive blinding that included treating anaesthesiologists, allowance for clinical titration of propofol, and a longer index ECT course (up to 12 treatments) than many prior trials — which strengthen external validity and safety monitoring. Several explanations for the absence of additive benefit are proposed by the investigators. First, there may be a ceiling effect of ECT such that any short‑lived early improvement from ketamine is overtaken by subsequent ECT sessions. Second, ECT and ketamine may share overlapping antidepressant mechanisms — for example, similar effects on regional cerebral blood flow, cerebral metabolism, hypertensive responses, and upregulation of brain‑derived neurotrophic factor — leaving little scope for additive action. Third, the therapeutic effects of ketamine observed in outpatient settings may in part relate to the conscious dissociative experience, an effect muted when ketamine is given as part of general anaesthesia. The authors acknowledge limitations: the sample size was small, which constrains power for some comparisons and precludes detailed time‑point analyses of non‑psychiatric outcomes; no formal post‑treatment cognitive testing was performed so subtle cognitive effects cannot be excluded; and time spent in the post‑anaesthesia care unit was not systematically recorded. They also note a potential risk that blinded anaesthesiologists might have given relatively higher propofol doses during early treatments, although the observed trend toward lower propofol use in the ketamine group suggests appropriate clinical titration occurred. On balance, the investigators conclude that low‑dose ketamine should not be used solely to improve depression outcomes during ECT, but that it appears safe and might have utility in particular clinical situations that warrant further study.