Électroconvulsivothérapie et niveau de preuve : de la causalité à la relation dose-effet

Reconsidering the seizure threshold? A retrospective study about seizure quality among patients with maintenance ECT

OBJECTIVES

Electroconvulsive therapy (ECT) is a safe and validated technique used to treat various psychiatric conditions. It triggers an artificially-induced seizure. This seizure is defined using several parameters such as the amount of energy, duration, frequency, pulse width and intensity. Efficacy and adverse events depend on the amount of energy delivered. Due to technical control, the amount of energy delivered by our unit's ECT device was limited to 614 mC, 60% of the maximum possible output of the device. We wondered if lowering the dose would lead to better seizure quality among maintenance ECT patients.

METHODS

We assessed seizure quality based on the EEG, using a validated tool created by MacPherson. Two evaluators independently rated the seizures. Pre- and post-control scores were compared using Student's t-test for paired samples.

RESULTS

We analysed data from 15 patients. Mean age was 65 years old. Twelve had depressive disorder, two had schizophrenia and one had schizo-affective disorder. Mean duration of seizure before control was 41.1 s [95% confidence interval (95CI)=26.1, 51.1]. The mean MacPherson's score was 20.3 (95CI=16.2, 24.4). After control, the mean MacPherson's score was 28.2 (23.1, 33.3), showing a significant difference with the pre-control dataset (p=0.032; t=-2.4; df=14). Specifically, peak mid-ictal amplitude increased from 6.9 (95CI=5.1, 8.7) to 10.0 (95CI=7.2, 12.8). Other sub-scores remained unchanged.

CONCLUSION

Lowering the energy delivered led to an overall increase of seizure quality among our sample. This highlights the necessity and utility of retitration during ECT maintenance, possibly leading to better management of our patients.

[What place for placebo in clinical trials conducted on psychiatric patients?]

Placebo effect remains a crucial issue in current clinical trials. Most clinical trials rely on the hypothesis of equivalent placebo response rates in both placebo and specific drug arms ("additive model"). But contrary to this dominant and rarely questioned hypothesis, several aspects may influence placebo response. A few recent meta-analyses and reviews have shown evidence for several clinical and methodological factors, which are able to modulate placebo response. In psychiatry research, placebo response has been mainly explored through antidepressant trials. In early clinical trials, drug-placebo differences were initially significant and robust. However, more recent clinical trials have not yielded similar results, and rather show narrowed antidepressant-placebo differences. Several factors may be involved in this absence of comparability: intrinsic properties of new antidepressants, changes in clinical criteria and classifications, symptomatic remission rather than global remission criteria, industrial and institutional constraints. Moreover, results from antidepressant trials (laboratory conditions) remain hardly fully transposable to clinical routine (ecological conditions).