A naturalistic study of risperidone treatment outcome using prognosis-adjusted discharge rates in New York state inpatients

Background: Information concerning the effectiveness of newer atypical antipsychotics is derived largely from controlled clinical trials of relatively short duration. Limited information is available concerning naturalistic outcome of patients selected for clinical treatment with atypical antipsychotics. This study evaluates 1-year discharge rates among all patients treated with risperidone within the New York State inpatient psychiatric hospital system during the calendar years 1994 and 1995 ("period of interest") relative to patients treated with all other antipsychotic medications. Method: Data from the Integrated Research Database at Nathan Kline Institute (Orangeburg, N.Y.) were used. This database maintains complete treatment records for all inpatients within the New York State psychiatric inpatient system along with demographic, diagnostic, admission, and discharge information. Patients were identified at admission or first change in antipsychotic during the period of interest, and 1-year outcome was determined. Results: 2198 risperidone-treated patients were identified versus 3259 treated with other antipsychotics. Length of hospitalization prior to treatment initiation was the primary predictor of discharge rate for both risperidone and control groups. When adjustment was made for between-group difference in prognosis (dischargeability), patients treated with risperidone within 30 days of admission were less likely to be discharged than those treated with all other agents (including clozapine), whereas risperidone was more effective in patients who had been hospitalized for 90 days or more prior to switch from another antipsychotic to risperidone. Conclusion: When database information is utilized to evaluate treatment effectiveness, adjustment must be made for a priori differences in prognosis or dischargeability. With appropriate methodology, database studies may indicate which patient groups are most likely to benefit from newer atypical antipsychotic agents.