To identify potential predictors of response to sertindole in patients with schizophrenia.

This twelve-month open-label study assessed the safety and efficacy of sertindole doses (4-24mg) in US patients with schizophrenia. Cox's regression analysis was applied to determine the effects of variables on time to sustained response (both CGI-S <=3 and CGI-I <=2, sustained for at least 8 weeks) in 358 patients.

125 patients achieved sustained response. Several factors appear to influence rate of response; amongst these are: treatment with antipsychotic medication before first diagnosis, severity of illness, ECT treatment, drug abuse history and patient weight. Treatment with antipsychotic medication before first diagnosis of schizophrenia increased the response rate, compared with treatment-naive patients. Mildly or moderately ill patients were more likely than more severely ill patients to respond to treatment with sertindole. Patients who never received ECT showed a higher rate of response to sertindole than those who received 1-5 courses of treatment. Patients with at least 6 courses showed a similar response to those who received none. Patients with no history of drug abuse were more likely to respond to sertindole, than patients who had a history of drug abuse. The response to sertindole is influenced by patient weight: for example, a patient weighing 150kg was more likely to respond than a patient weighing 75kg.

A prognostic index could be calculated based on these factors to predict the response of individual patients to treatment with sertindole.