RESUMEN
BACKGROUND: There is little literature on the relative efficacy and cognitive safety of right unilateral (RUL), bifrontal (BF), and bitemporal (BT) electroconvulsive therapy (ECT) in schizophrenia. METHODS: We present a randomized, nonblind, naturalist comparison of a fixed course of 8 moderately high-dose RUL (n = 24), threshold BF (n = 27), and threshold BT (n = 31) ECT in patients with schizophrenia. Assessments included the Positive and Negative Syndrome Scale (PANSS), the Wechsler Memory Scale-Revised, and an autobiographical memory interview. A completer analysis was planned and conducted to capture the cognitive outcomes. RESULTS: The sample as a whole improved significantly on all efficacy outcomes and deteriorated significantly on all cognitive outcomes. The primary efficacy outcome, improvement in PANSS total scores, did not differ significantly across groups. The PANSS positive score (but no other subscale score) improved significantly less with RUL relative to BF and BT ECT. For autobiographical memory and for almost all Wechsler Memory Scale subtests, including memory quotient (the primary adverse effect outcome), BT ECT was associated with greater impairment than RUL or BF ECT. Importantly, all statistically significant differences between treatments were clinically small in magnitude. CONCLUSIONS: In patients with schizophrenia who receive a fixed course of 8 ECTs, threshold BT ECT is associated with greater cognitive impairment across a range of measures, and moderately high-dose RUL ECT is associated with poorer efficacy against positive symptoms. Threshold BF ECT exhibits the best efficacy-cum-neurocognitive safety profile. All differences between groups, however, are small and perhaps clinically insignificant.