Response

We read with interest the recent editorial by Nigam et al, ‘Advancing ketamine in the treatment hierarchy for refractory depression’. ^{Reference Nigam, King and Forcen1} While we welcome discussion of emerging antidepressants such as ketamine, the authors’ strong statements that ‘providers often consider ketamine as an alternative to electroconvulsive therapy (ECT)’ and that ketamine ‘may be at least as effective as ECT’ are premature and predicated on a single citation to a non-inferiority trial with significant methodological caveats. ^{Reference Anand, Mathew, Sanacora, Murrough, Goes and Altinay2} This selective focus ignores the broader context of the available high-level evidence, including another large randomised controlled trial (RCT) that arrived at a contrary conclusion ^{Reference Ekstrand, Fattah, Persson, Cheng, Nordanskog and Åkeson3} and several meta-analyses demonstrating superiority of ECT over ketamine. ^{Reference Rhee, Shim, Forester, Nierenberg, McIntyre and Papakostas4–Reference Petrucci, Fernandes, Reis, da Silva, Recla and de Mendonça6}

First, the editorial’s section on the comparative efficacy of ketamine and ECT relies solely on the Anand et al ^{Reference Anand, Mathew, Sanacora, Murrough, Goes and Altinay2} ELEKT-D trial, a study with an atypical patient population treated with a weak form of ECT. The ELEKT-D trial exclusively recruited patients with non-psychotic, unipolar depression with a history of two failed lifetime antidepressant trials. This is a non-standard definition of treatment resistance that deviates from the commonly accepted threshold of inadequate response to two or more antidepressant trials in the current episode. In routine clinical practice in Western countries, psychotic and bipolar depression are key indications for ECT, and its superior efficacy in these populations is established by meta-analytic evidence. ^{Reference van Diermen, van den Ameele, Kamperman, Sabbe, Vermeulen and Schrijvers7,Reference Bahji, Hawken, Sepehry, Cabrera and Vazquez8} Furthermore, the ELEKT-D patient sample was relatively young (mean age 46), a demographic in which ECT is known to be less effective, as older age is associated with superior ECT response. ^{Reference van Diermen, van den Ameele, Kamperman, Sabbe, Vermeulen and Schrijvers7} Moreover, the ECT modality used in the ELEKT-D trial was ultrabrief pulse (0.3 ms) right unilateral ECT, with switching to 0.5 ms pulse width bilateral ECT permitted in patients deemed to be non-responding. A 2015 meta-analysis by Tor et al ^{Reference Tor, Bautovich, Wang, Martin, Harvey and Loo9} has shown that, while this form of ECT offered a superior cognitive side-effect profile, it was significantly less effective than standard brief-pulse (1.0–1.5 ms) right unilateral ECT with a remission rate of only 33.8%. Choosing a less potent form of ECT as the comparator in a non-inferiority trial inherently lowers the bar for ketamine. Therefore, the finding of non-inferiority in this narrow context cannot be interpreted as non-inferiority to optimally dosed ECT.

Second, Nigam et al ^{Reference Nigam, King and Forcen1} do not mention the only other large RCT in this field, the KetECT trial by Ekstrand et al. ^{Reference Ekstrand, Fattah, Persson, Cheng, Nordanskog and Åkeson3} This study is particularly relevant as it sampled a more prototypical ECT population of in-patients with a higher mean age, and including those with psychotic depression. Its findings were starkly different from those of the ELEKT-D trial: the KetECT trial found ECT to be significantly more effective than ketamine, with remission rates of 63% in the ECT group compared with 46% in the ketamine group (P = 0.026). The divergence between the ELEKT-D ^{Reference Anand, Mathew, Sanacora, Murrough, Goes and Altinay2} and KetECT ^{Reference Ekstrand, Fattah, Persson, Cheng, Nordanskog and Åkeson3} trials underscores how patient selection criteria and ECT parameters can dramatically affect clinical outcomes, and it highlights the danger of drawing broad conclusions from a single idiosyncratic study of an atypical patient population.

Third, and most importantly, the highest level of evidence available, systematic reviews and meta-analyses, shows that ECT is superior to ketamine in reducing depressive symptoms with a medium effect size. A 2022 meta-analysis of six studies (five RCTs and one non-randomised study) by Rhee et al ^{Reference Rhee, Shim, Forester, Nierenberg, McIntyre and Papakostas4} reported a standardised mean difference (SMD) of –0.69 (95% CI (–0.89, –0.48)) in favour of ECT, with a subgroup analysis of the five randomised trials showing a similar effect size (SMD = –0.69, 95% CI (–0.91, –0.47)) favouring ECT. A 2025 update of their meta-analysis by Rhee et al, ^{Reference Rhee, Shim, Davoudian, Espinoza and McIntyre5} including the subsequently published ELEKT-D trial, reported an SMD of –0.59 (95% CI (–1.01, –0.17)) in favour of ECT over ketamine from seven studies, six of which were RCTs. A 2024 meta-analysis by Petrucci et al, ^{Reference Petrucci, Fernandes, Reis, da Silva, Recla and de Mendonça6} which included all six RCTs, reached a similar conclusion, finding that ECT was superior to ketamine (SMD = 0.56, 95% CI (0.05, 1.06)) in the overall population, as well as in the subgroup of participants treated as in-patients (SMD = 0.71, 95% CI (0.38, 1.04)).

In conclusion, while ketamine may be a valuable addition to the therapeutic toolkit for treatment-resistant depression, the current evidence base does not support the claim that it is an alternative to ECT with comparable efficacy. Such an assertion is built upon a selective reading of the literature, highlighting a single high-profile trial with significant methodological limitations while ignoring countervailing evidence. Clinical decision-making must be guided by the totality of the evidence, preferably highest-level evidence from methodologically sound meta-analyses.