Clinical predictors of seizure threshold in electroconvulsive therapy: a prospective study

Jeroen A van Waarde, Lucas J B van Oudheusden, Bastiaan Verwey, Erik J Giltay, Rose C van der Mast

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

At the start and during the course of electroconvulsive therapy (ECT), estimation of the seizure threshold (ST) is useful in weighing the expected effectiveness against the risks of side effects. Therefore, this study explores clinical factors predicting initial ST (IST) and levels of ST during the ECT course. This prospective observational study included patients aged ≥18 years receiving ECT without contraindications for dose titration. At the first and every sixth consecutive ECT session, ST level was measured. Using multivariate linear regression and multilevel models, predictors for IST and change in ST levels were examined. A total of 91 patients (mean age, 59.1 ± 15.0 years; 37 % male; 97 % diagnosis of depression) were included. In multivariable analysis, higher age (β = 0.24; P = 0.03) and bifrontotemporal (BL) electrode placement (β = 0.42; P < 0.001) were independent predictors for higher IST, explaining 49 % of its variation. Also, these two variables independently predicted higher ST levels at different time points during the course. Using multilevel models, absence of a previous ECT course(s) predicted a steeper rise in ST during the course (P = 0.03 for the interaction term time*previous ECT). The age-adjusted dose-titration method is somewhat crude, resulting in some measurement error. Concomitant medication use could have influenced ST levels. Increasing age and BL electrode placement predicted higher (I)ST, which should be taken into account when selecting ECT dosage. Previous ECT course(s) may avoid an increase in ST during the course of ECT.

Original languageEnglish
Pages (from-to)167-75
Number of pages9
JournalEuropean Archives of Psychiatry and Clinical Neuroscience
Volume263
Issue number2
DOIs
Publication statusPublished - Mar 2013

Cite this