Variables significantly associated with antipsychotic use included a diagnosis of psychosis (odds ratio, 7.2), taking anticonvulsants (OR, 2.7), taking lithium (OR, 2.5), having PTSD (OR, 2.3), taking benzodiazepines (OR, 2.0), having a diagnosis of behavioral disorder (OR, 1.93), taking antidepressants (OR, 1.8), and being male (OR, 1.76).
More than three-quarters of patients (82%) received one of the three antipsychotics approved for some indication for patients under age 18 (aripiprazole, quetiapine, and risperidone). However, in 92% of all patients the antipsychotic was still "off label" based on the patient’s age and/or diagnosis.
Use of antipsychotics did not change significantly over the two time periods (47% in the earlier time period vs. 44% in the later time period). There were no significant changes in antipsychotic use for any diagnostic group.
"The prevalence of antipsychotic use varies widely by diagnosis and several other parameters," Dr. Goethe concluded. "Some of these are new findings, and to us they further support the need for clarity and uniformity in reporting prevalence [of antipsychotic use in children and adolescents], both overall and with respect to off-label use.
"We need common definitions, and we need to have clarity about when we’re quoting FDA guidelines and when we are quoting generally accepted guidelines," he said. "The specific antipsychotics that were used were not surprising with regard to the data about second-generation drugs, but were about first-generation drugs."
Also, the high rate of off-label use "was expected, and these findings are consistent with previous reports," he continued. "Calculating off-label prevalence is complicated because definitions vary widely; FDA-approved indications change over time; and some of the reported rates in the literature are misleading in that you can read the rates as reflecting FDA approval, but in fact what they’re reporting is being consistent with or not being consistent with one or more of the published guidelines, so we need to be clear about that."
Dr. Goethe acknowledged certain limitations of the study. It was conducted on a single site, some relevant variables might not have been included, and no posthospital follow-up was conducted. "We are assuming that the discharge medication was one that would be continued, but we don’t know that for a fact," he said. "Further research in this area is needed."
The study was funded by the Burlingame Center Endowment and the Hartford Hospital Research Fund. Dr. Goethe disclosed that he also has received research support from the National Institutes of Health, AstraZeneca, Bristol-Myers Squibb, Eli Lilly, Eisai, Hoffmann-La Roche, Shire, Schering-Plough, Johnson & Johnson, Merck, Neuronetics, Novartis, Otsuka, Pfizer, and Takeda.