Assessing the impact of Pennsylvania’s prior authorization policy intended to reduce antipsychotic prescribing in Medicaid-insured children

Abstract

Introduction: The volume of antipsychotic medications prescribed to children and adolescents has risen sharply since second generation antipsychotics, also referred to as atypical antipsychotics, were introduced in the 1990’s. The concern surrounding the expanded use of antipsychotics was that the medications have significant adverse metabolic side effects and they were often prescribed to treat conditions in young children for which they have not been proven to be safe and effective. While it is not unlawful for health care providers to prescribe medications for uses beyond which they have been approved by the United States Food and Drug Administration, the lack of empirical evidence guiding much of the antipsychotic use in children had professional pediatric medical groups and policy makers concerned for the well-being of children receiving the medications. Several states, including Pennsylvania, enacted prior authorization policies in an attempt to restrict prescribing to children where a medical need has been established. However, the impact of the policies is largely unknown since published data on the topic is sparse. Methods: This retrospective, medical claims-based cohort study, used de-identified administrative Medicaid data from January 2008 to December 2010 to investigate the impact of Pennsylvania’s September 2008 antipsychotic prior authorization policy on antipsychotic prescribing prevalence in children targeted by the policy. Descriptive methods and segmented regression of the interrupted time series were used to assess the effects of the policy on monthly antipsychotic prescribing prevalence. A difference-in-difference analysis compared Pennsylvania’s prescribing to Ohio, a geographically proximate and demographically similar state without a prior authorization policy; and Delaware, a state that enacted a policy 3 years prior to Pennsylvania. The potential for compensatory prescribing was assessed by reporting the prevalence of other psychotropic medications over the study period. Results: An average of 99,074 Pennsylvania Medicaid enrollees ages 0-6 were identified as meeting the study criteria annually from 2008-10. Immediately following the policy intervention, an abrupt, significant reduction in monthly prescriptions of antipsychotics was observed (-51 prescriptions per 100,000; p=0.0052) and sustained over the observation period. The proportion of children filling prescriptions for antipsychotics dropped approximately 46% and the average number of antipsychotic prescriptions filled per month was reduced by 53% in 2010 compared to 2008. In Ohio, a state without such a policy, the proportion of children receiving an antipsychotic increased nearly 10% in 2010 compared to 2008 and the average number of monthly prescriptions increased 30%. Reductions in antipsychotic prescribing in Delaware, a state that had its antipsychotic policy in place since 2005, were comparable to Pennsylvania. There was no evidence that non-antipsychotic psychotropic medications were prescribed in place of the medications restricted by the policy. Conclusions: Pennsylvania’s 2008 prior authorization policy was associated with a significant decrease in annual and monthly antipsychotic prescribing prevalence in Medicaid-insured children targeted by the policy, those ages 0-6 years of age. Reductions in most other psychotropics was also observed, indicating changes in prescribing behavior may have extended beyond antipsychotics. While this analysis suggests the policy may have achieved its primary aim of reducing antipsychotic prescribing, more research is needed to better understand the complex array of factors influencing provider behavior and to explore potential unintended consequences of the policy.