To the Editor: In a study of the EEG changes associated with antipsychotic drug treatment, Franca Centorrino, M.D., et al. (1) reported that the "EEG abnormality risk varied widely… [and was] particularly high with clozapine and olanzapine, moderate with risperidone and typical neuroleptics, and low with quetiapine" (p. 109). The authors’ use of "abnormality" to describe changes in the EEG records should be criticized. The term reflects a practice in clinical neurology that is not applicable to describing the effects of psychoactive drugs (2).

In psychiatric practice, adequate serum clozapine levels are deemed necessary for its beneficial effects. These are not characterized as "abnormal" unless they exceed safety levels. EEG changes also vary with serum levels and are as necessary as adequate serum levels for clinical efficacy (3, 4). "Abnormality" suggests a deleterious effect, and from reading the report’s summary sentence as quoted, the reader would conclude that quetiapine has a favorable balance of toxicity over clozapine and risperidone. The reverse is probably true—that the failure of quetiapine to elicit EEG changes probably reflects a lesser clinical effect.

Stevens (5) presented a cogent argument for evaluating the EEG slow waves of clozapine as evidence of its therapeutic activity. She found the centrencephalic effects of clozapine to be the basis for its clinical efficacy and advised that these changes be heralded rather than decried. A parallel relationship of EEG effects is found in studies of ECT, in which electrophysiologic slowing reflects its therapeutic effects (6). Failure to elicit interseizure EEG slowing during an ECT course is associated with poor clinical outcome. The development of EEG slow-wave activity is a necessary part of the ECT process, and when these changes are absent, so too are the clinical benefits.

The term "abnormality" comes from a neurologic literature that uses visual impressionistic methods to assess EEG records. But psychoactive drugs induce subtle changes that are not appreciated in visual analysis. They require quantitative digital computer processing for true estimates of effect (2). Describing the EEG changes associated with psychoactive drugs as "abnormal" or "normal" is misleading. Disregarding the excellent technical methods of analysis for EEG that match the advances in other brain imaging methods degrades the science.