Interrater reliability of 2 definitions of seizure adequacy in electroconvulsive therapy

Clear evidence of a generalized seizure is regarded as an essential component of a therapeutic electroconvulsive therapy (ECT) session. Recent criteria, in particular, the Royal College of Psychiatry ECT Handbook (2005), suggest that the pattern of electroencephalographic (EEG) seizure is more important than the duration of the seizure when assessing seizure adequacy. We examined the reliability of using EEG criteria compared to simple measures of seizure duration by reviewing 100 sample EEG printouts obtained during ECT sessions in a Scottish hospital ECT unit.

The EEGs were independently assessed by 3 clinicians (a "control" ECT consultant, an experienced ECT therapist, and a less experienced trainee) blind to each other's ratings.

While there was good interrater reliability for measures of seizure duration, kappa scores for EEG criteria were much better for the more than the less experienced rater (κ = 0.68 vs 0.50). Even greater differences were seen for individual components: polyspike activity (κ = 0.80 vs 0.55), spike and wave complexes (κ = 0.80 vs 0.38), and postictal suppression (κ = 0.63 vs 0.35).

The implications of these finding are briefly discussed in the context of developing rational criteria for assessing ECT efficacy that rely upon the consistent and replicable delivery of standardized ECT techniques and consequent training needs of ECT practitioners.