The rush from bench to bedside in electroconvulsive therapy

To the Editor

We read with great interest the ‘viewpoint’ of Rosenman (2018), in which he debunks the scientific underpinnings of the stimulus dose titration method in electroconvulsive therapy (ECT). We completely agree and believe this represents the field having rushed from small research datasets to premature implementation in clinical practice.

Our reading of the seminal and influential classic research studies (Sackeim et al., 2000) reveals that the accepted dogma of the lack of efficacy of low-dose stimulation with right unilateral electrode placement (RUL) is based on a total of 43 patients from two studies. We believe that this finding was never adequately, independently replicated before being accepted by the field. How should this be reconciled with other preliminary data (Lapidus et al., 2013) (also, admittedly small and not replicated) showing that a low-dose first RUL ECT session is effective in lowering depression rating scores?

Another idea widely accepted by the field is the extreme range and variability of seizure thresholds (ST). In clinical practice, most patients have a predictable ST in a tight range. The Prolonging Remission in Depressed Elderly (PRIDE) study data showed that with a particular stimulus package at 25 mC, 84% of elderly patients (those typically with the highest STs) had an adequate seizure. ‘Very high seizure thresholds’ are usually the result of concomitant anticonvulsant medication or poor technique in delivering the stimulus; only occasionally, even in busy clinical practices, does a patient have a truly elevated ST.

Trying to balance efficacy and side effects continues to be a crucial goal in ECT practice.

We have a clearer idea of how to dose bilateral ECT than RUL; grossly suprathreshold stimuli likely add more to cognitive adverse effects than to efficacy. This, combined with the above understanding of the predictability of ST for the vast majority of patients, argues against the need for ECT devices with outputs of 1000 mC.

We conclude that some aspects of ECT practice are ahead of the data. Fortunately, ECT is so safe and effective that even this lack of knowledge of exactly the best technique does not seriously limit the ability of ECT to be widely used for benefit to a large number of patients. We remain hopeful that further careful research will answer the remaining technical questions about ECT, the advisability of dose titration among them.