Objective: The purpose of this study was to identify children in a state Medicaid population who were newly treated with second-generation antipsychotics from 2001 through 2005, to classify each use of these agents as evidence based or not depending on the child's diagnoses, and to identify factors associated with the likelihood of evidence-based use of the medication.
Methods: A Medicaid claims database was used to retrospectively identify enrollees receiving initial outpatient treatment with a second-generation antipsychotic between 2001 and 2005. To capture all relevant treatments and diagnoses, claims were examined from January 2000 through December 2006. The final sample included 11,700 children under age 18. The primary measure of interest was the proportion for whom use of the antipsychotic was based on evidence. Evidence-based use (categorized as strong, plausible, or weak evidence) was defined as any use of the agent for a diagnosis supported by a clinical trial published before the end of 2005. Trend analysis and logistic regression were used.
Results: The number of children newly treated with second-generation antipsychotics increased from 1,482 in 2001 to 3,110 in 2005. Of the new users of these agents during the study period, 41.3% had no diagnosis for which such treatment was supported by a published study. The medication with the highest level of non-evidence-based use was aripiprazole (77.1%), and risperidone had the lowest (30.6%).
Conclusions: The number of children receiving second-generation antipsychotics doubled in this Medicaid population between 2001 and 2005, and a large proportion of the treatments were not supported by evidence from clinical studies.