Spatio-temporal clusters of new psychotropic medications among Michigan children insured by Medicaid.

PURPOSE The prescription of psychotropic medications to children and adolescents has increased dramatically over the last decade. However, the development of disparities in prescribing is poorly understood. We examined whether clustering of utilization is a common phenomenon among early adopters of medications described the characteristics of clusters. METHODS We obtained the complete Medicaid Analytic Extract (MAX) files for the State of Michigan between 1 January 2000 and 31 December 2003. We tracked the adoption of: aripiprazole, atomoxetine, escitalopram, methylphenidate OROS, and ziprasidone. We conducted retrospective, space-time analyses, scanning for clusters with high rates of prescribing. Chi(2) analyses were used to compare the attributes of patients living within clusters to patients living in the rest of the state for each medication. Clusters of utilization were identified via the spatial scan statistic. Analysis of variance (ANOVA) was then used to compare the numbers of mental health professionals per capita in geographic areas that did and did not demonstrate clustering of prescriptions for new psychotropic medications. RESULTS All five medications exhibited space-time clustering within the first 90 days following the US Food and Drug Administration (FDA) approval. The Medicaid population surrounding Kalamazoo was more likely to receive a prescription on multiple occasions. Excluding ziprasidone, clusters were not associated with greater geographic access to mental health care professionals. CONCLUSIONS Clustering of new prescriptions for psychotropic medications was a common phenomenon in this population. Surveillance and cluster identification allow the development of disparities to be studied. This information permits interventions to be targeted to locations prospectively.