Cost-effectiveness of sertindole versus olanzapine or haloperidol: a comprehensive model

The safety and efficacy of second generation antipsychotics relative to conventional treatment are well documented in<p><p>schizophrenia, although their economic impact has not yet been evaluated. The aim of this study was to evaluate antipsychotics in normal practice using a ten-year cost-effectiveness model based on a six-month Markov cycle tree. The model incorporated three competing drug strategies (sertindole, olanzapine and haloperidol); five care managnement strategies, defined by place of residence (hospital, residential or private home) and intensity of care (intensive or mild); clinical events (extrapyramidial symptoms, sedation, weight gain, sexual dysfunction and relapse); and direct medical costs associated with each. Adverse-event rates were estimated from integrated safety reports; compliance and relapse rates were obtained from meta-analysis of the literature. The model comprised fifteen discrete health states; transition probabilities among these were derived from local patient cohorts. The model computed that the relative risks of relapse on haloperidol or olanzapine compared with sertindole are 1.4 and 1.2 respectively, corresponding to an additional 5.7 and 13.5 months withour relapse over ten years. Furthermore, in most scenarios, sertindole is self-financing because less time is spent in hospital and indeed shows modest net savings.