Research report Prediction of antidepressant response in both 2.25 × threshold RUL and fixed high dose RUL ECT James N. Kimball ⁎ , Peter B. Rosenquist, Aaron Dunn, Vaughn McCall Department of Psychiatry and Behavioral Medicine, Wake Forest University School of Medicine, Winston Salem, NC 27157, United States Received 7 February 2008; received in revised form 18 March 2008; accepted 24 March 2008 Available online 9 June 2008 Abstract Some forms of electroconvulsive therapy (ECT) can result in generalized seizures that lack efficacy, therefore physiological markers of treatment adequacy would be helpful. EEG measures of seizure quality, such as EEG regularity and post-ictal suppression, have largely supplanted seizure duration as a marker for seizure adequacy, yet no predictive algorithm has gained wide clinical acceptance. Electrographic seizure durations of less than 25 s still prompt re-stimulation in many settings. We re-examined the utility of EEG seizure duration and other measures of EEG seizure as predictors of antidepressant response to right unilateral (RUL) ECT. Methods: Seventy-two adult patients with major depression were randomized to either titrated RUL ECT at 2.25 times initial seizure threshold or RUL ECT at a fixed dose of 403 mC. Intent-to-treat responder status (defined by 60% reduction in HRSD scores and final score of 12 or less after the last RUL ECT session) was identified as the dependent variable in a nominal logistic regression model including EEG seizure quality candidate variables, controlled for age and gender. Results: A model including EEG seizure duration, EEG regularity, post-ictal suppression, age and gender and randomization status was significantly predictive of intent-to-treat responder status at treatment 2 (R2 = .21 p b .003; N = 66) and treatment 4 (R2 = .27 p b .0004; N = 67). The model remained significant at these time points even when randomization status (titrated moderately suprathreshold vs. high fixed dosage) was removed (Treatment 2: R2 = .18 p b .007; Treatment 4: R2 = .23 p b .0007). Conclusion: EEG markers of seizure adequacy, including EEG seizure duration, are modestly predictive of antidepressant response for both titrated moderately suprathreshold and high fixed dosage RUL ECT. © 2008 Elsevier B.V. All rights reserved. Keywords: ECT; Electroconvulsive therapy; Depression; Major depression; Seizure duration; ROC curves; Antidepressant response; Seizure threshold 1. Introduction Electroconvulsive therapy (ECT) is the most effec- tive treatment for major depressive episode (MDE) (American Psychiatric Association, 2001). Even so, in a clinical trial setting the remission rate, as defined by a Hamilton Rating Scale for Depression (HRSD) of b 7, is usually no better than 80%. Of greater concern, in community settings, the antidepressant remission rate for ECT is well under 50% (Prudic et al., 2004). Given these remission rates, predictors of response that would be available early in the course of ECT would be of value. For example, if a measurement taken at the second or third treatment indicated that the patient was unlikely to respond to the strategy chosen for ECT, then a more aggressive treatment approach could be taken, such as increasing stimulus intensity. Journal of Affective Disorders 112 (2009) 85 – 91 www.elsevier.com/locate/jad ⁎ Corresponding author. Tel.: +1 336 716 9702. E-mail address: jkimball@wfubmc.edu (J.N. Kimball). 0165-0327/$ - see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jad.2008.03.030