MetaTOC stay on top of your field, easily

Space–Time Cluster Analysis to Detect Innovative Clinical Practices: A Case Study of Aripiprazole in the Department of Veterans Affairs

, , , , , , , , ,

Health Services Research

Published online on

Abstract

Objective To identify space–time clusters of changes in prescribing aripiprazole for bipolar disorder among providers in the VA. Data Sources VA administrative data from 2002 to 2010 were used to identify prescriptions of aripiprazole for bipolar disorder. Prescriber characteristics were obtained using the Personnel and Accounting Integrated Database. Study Design We conducted a retrospective space–time cluster analysis using the space–time permutation statistic. Data Extraction Methods All VA service users with a diagnosis of bipolar disorder were included in the patient population. Individuals with any schizophrenia spectrum diagnoses were excluded. We also identified all clinicians who wrote a prescription for any bipolar disorder medication. Principal Findings The study population included 32,630 prescribers. Of these, 8,643 wrote qualifying prescriptions. We identified three clusters of aripiprazole prescribing centered in Massachusetts, Ohio, and the Pacific Northwest. Clusters were associated with prescribing by VA‐employed (vs. contracted) prescribers. Nurses with prescribing privileges were more likely to make a prescription for aripiprazole in cluster locations compared with psychiatrists. Primary care physicians were less likely. Conclusions Early prescribing of aripiprazole for bipolar disorder clustered geographically and was associated with prescriber subgroups. These methods support prospective surveillance of practice changes and identification of associated health system characteristics.