An epidemiological investigation of off‐label anticonvulsant drug use in the Georgia Medicaid population

Abstract

Purpose The primary objective was to determine the prevalence of off-label anticonvulsant drug use in the Georgia Medicaid population and establish what percentage of this off-label use is evidence-based. The second objective was to investigate differences in the prevalence of off-label use among anticonvulsants marketed before and after 1993. The third objective was to identify patient and physician characteristics associated with off-label use. Method The study design was a retrospective study utilizing pharmacy, inpatient, outpatient and long-term care claims linked with eligibility files for persons with Georgia Medicaid benefits in 1999 through 2000. An anticonvulsant recipient was considered an off-label anticonvulsant user if their anticonvulsant use did not match age or medical diagnoses in the product label. An evidence-based off-label use was defined as off-label anticonvulsant use supported by at least one randomized controlled clinical trial. Logistic regression analysis was used to identify patient and physician characteristics associated with off-label use. Results 34 676 (71.3%) of 48 648 patients on one or more anticonvulsants received an off-label prescription for an anticonvulsant. Gabapentin was the anticonvulsant most widely used off-label (86%). After accounting for labeled and all evidence-based uses for the six most frequently prescribed anticonvulsants, there was a moderate to large percentage of anticonvulsant use not supported by any evidence from controlled trials (range: 19.09–57.07%). The most common comorbidities among patients prescribed the top six anticonvulsants were diabetes mellitus, depression, schizophrenia and pain. Anticonvulsants launched after 1993 had a higher prevalence of off-label use than anticonvulsants marketed before 1993. Off-label drug use varied by age with children and adolescents being the most likely to receive an off-label anticonvulsant. Compared with other practitioners, neurologists were more likely to prescribe anticonvulsants off-label. Conclusions The anticonvulsant off-label use in the Georgia Medicaid population is very high (71%). Only a modest proportion of these off-label uses are supported by evidence from controlled trials. Copyright © 2004 John Wiley & Sons, Ltd.