Predictors of bradycardia during the stimulation phase of electroconvulsive therapy

Temporary asystole can be detected at various time points during electroconvulsive treatment. We carefully monitored and documented its incidence during stimulation to evaluate currently known and assumed predictors.

All treatments over a 20-month period in 2 separate institutions were recorded prospectively. Data from 119 patients comprising 720 treatments were suitable for multiple regression analysis.

Electrode placement was the most influential determinant. Treatment series using right unilateral placement (64 patients and 291 traces) produced a mean (SD) asystole duration of 5.64 (2.88) seconds (range, 1.21-11.20 seconds), compared to a duration of 0.80 (0.21) seconds (range, 0.47-1.71 seconds) in bifrontal series (55 patients, 429 traces). Multiple regression analysis showed no independent effect of body weight, age, β-blocker medication, or preexisting heart block. Stimulus dose and succinylcholine dose had no influence in univariate analyses. Cardiac action in bifrontal treatment series was highly predictable using baseline values. During unilateral stimulation, between-subject differences accounted for 79% of the observed variation.

The heart rate during stimulation depends mainly on electrode positioning. A yet unidentified, probably constitutional factor is responsible for the broad range of asystole duration brought about by right unilateral electrode placement. Any assessment using interindividual analysis will therefore be biased as long as that factor remains unknown.