Clinical Question: Are conventional anti-psychotic agents safer than atypical agents in older patients?
Study Design: Cohort (retrospective)
Synopsis: A recent public health advisory (http://www.fda.gov/cder/drug/advisory/antipsychotics.htm) warned that there is a twofold increase in the risk of death among older patients with dementia when given atypical antipsychotic agents such as aripiprazole (Abilify), clozapine (Clozaril), olanzapine (Zyprexa), quetiapine (Seroquel), risperidone (Risperdal), and ziprasidone (Geodon). Researchers performed this federally sponsored study to determine whether conventional agents (e.g., haloperidol [Haldol], thiothixene [Navane], loxapine [Loxitane]) carry a similar risk.
They identified patients older than 65 years who had filled a first prescription for any antipsychotic agent between 1994 and 2003. Mortality data came from the Medicare Death Master File, and they used administrative data sets to establish comorbidities, hospitalizations, nursing home residence, and use of other medications. They identified 9,142 patients using a conventional antipsychotic and 13,748 using an atypical agent. Patients taking conventional antipsychotics were more likely to have heart disease or cancer, but they were less likely to have other mood disorders or take other psychotropic medications. In the unadjusted analysis there were more deaths among users of conventional agents (17.9 versus 14.6 percent; P < .001; number needed to harm = 30; 95% confidence interval [CI], 23 to 43). This corresponds to an unadjusted hazard ratio of 1.51 (95% CI, 1.43 to 1.59). After adjusting for comorbidities, the relative risk was somewhat lower but still statistically significant (1.37; 95% CI, 1.27 to 1.49). The risk was greatest in the first 40 days after beginning therapy, in patients without dementia, and in patients taking a higher-than-median dose of the conventional antipsychotic medication.
Bottom Line: It seems reasonable to conclude that conventional and atypical antipsychotic agents are associated with an increased risk of death in older patients. The limitations of this study do not allow for a definite conclusion that older agents are less safe than newer agents. (Level of Evidence: 2b)