*223. Variation in Medication Management of Schizophrenia in Veterans Affairs Hospitals
TJ Hudson, Center for Mental Outcomes and Research (CeMHOR); RR owen, Center for Mental Outcomes and Research (CeMHOR); W Feng, Center for Mental Outcomes and Research (CeMHOR); MA Austen, Center for Mental Outcomes and Research (CeMHOR)
Objectives: This study will compare baseline data of medication management(MM) for schizophrenia with data from a multi-site study comparing the effectiveness of a conceptually based, enhanced intervention to that of a basic educational strategy with regard to improving guideline concordant MM.
Methods: Baseline data were extracted from VistA for13 VAMCs prior to initiation of the project. MM analysis was based on the first antipsychotic prescription following hospital discharge for schizophrenia. Demographic characteristics of subjects receiving novel and guideline concordant doses of antipsychotics were compared using bivariate analyses. Characteristics associated with increased likelihood of receiving guideline concordant or novel antipsychotics were evaluated with logistic regression models. Seven VAMCs participated in the 6-month study of MM in 391 subjects with exacerbation of schizophrenia. In 11/00, extraction of MM data from medical records and VistA will be complete. Analgous techniques will be used to evaluate MM during the study period and will be compared to baseline data.
Results: At baseline, 97% of the 599 subjects who received an antipsychotic prescription were male. Mean age was 48.6 years. About 16% were married, 45% were African American, 48% were Caucasian and 6% were Hispanic. Overall 47% received a novel antipsychotic. Caucasians were significantly more likely than non-Caucasians to receive novel antipsychotics. (chi-square = 12, df=1, p=0.0006) Although there was significant variation among facilities, no significant differences in marital status, age or gender were found between recipients of novel and conventional antipsychotics. After controlling for facility in the logistic regression model, Caucasians remained more likely than non-Caucasians to receive novel antipsychotics. (chi-square =14.20, df=1,p=0.0002) Antipsychotic dose was guideline concordant in 76% of the subjects and was more likely in subjects receiving novel antipsychotics. After controlling for facility in a logistic regression model, receipt of a novel antipsychotic and age were the strongest predictors of guideline concordant dosing (p<.001 and .017 respectively). In the multi-site study, subjects were 95% male with a mean age of 46. Fifteen percent were married; 56% were African American, 27% were Caucasian and 6% were Hispanic. Race was missing for 43 subjects. Once data collection is finished, an analysis of medication management analogous to that of baseline will be completed.
Conclusions: The baseline data from VistA indicate that novel antipsychotics were prescribed to approximately 50% of subjects; were more likely to be prescribed to Caucasians and to be prescribed in guideline concordant doses. Because these data are 3 years old, the percent of subjects receiving novel antipsychotics is likely to increase when data from the multi-site study are analyzed. The effect of age and race are difficult to predict.
Impact: Non-concordant medication management practices can place veterans with schizophrenia at risk for adverse outcomes. This is particularly concerning since use of novel antipsychotic medications could decrease adverse medication effect. This study suggests that some patients may not have access to high quality medication management, including the newer antipsychotics. It is important to identify and evaluate this problem to optimize the outcomes of all veterans with schizophrenia.