National study of antiepileptic use among patients with dementia in long-term care - PROJECT ABSTRACT Nursing homes are an important location of care for patients with dementia, who are at high risk for medication-related harms. Antiepileptic medications such as valproic acid and gabapentin are commonly prescribed off-label to patients with dementia to treat symptoms such as behavioral disturbances and pain despite limited evidence that these medications are effective in treating such conditions. Our work has found that the prescribing of antiepileptics has doubled over the past decade, now prescribed to over 40% of nursing home residents with dementia. The growth in antiepileptic prescribing is largely driven by valproic acid and gabapentin, which are potentially prescribed as substitutes for antipsychotics or opioids, respectively, in light of federal nursing home prescribing initiatives to reduce antipsychotic use and the ongoing opioid epidemic. However, unlike antipsychotic and opioid prescribing, use of antiepileptics is not subject to regulatory oversight and growth has continued largely unchecked. Off-label use of antiepileptics is concerning for older adults given serious associated risks including falls, sedation, and increased mortality. Despite the high rates of antiepileptic prescribing to nursing home residents with dementia and potential harms associated with use, this growth in prescribing has received little attention. Reducing potentially inappropriate antiepileptic use is an opportunity to improve medication safety for nursing home residents with dementia, both through identifying which nursing home residents are at greatest risk as well as the role of clinicians and facilities in these prescribing shifts. However, the design of clinical or policy interventions to reduce potentially inappropriate antiepileptic prescribing cannot occur without understanding why clinicians decide to prescribe these medications. In this explanatory mixed-methods analysis, we will first characterize the array of patient, provider, and nursing home facility characteristics associated with antiepileptic prescribing among all Medicare beneficiaries with dementia residing in nursing homes (Aim 1). Next, in Aim 2, we will use linked Medicare and Minimum Data Set data to examine clinical outcomes associated with antiepileptic medication initiation (e.g., mortality, change in behavioral and pain scores) as compared to alternative medication classes (i.e., antipsychotics and opioids). Finally, in Aim 3 we will interview prescribing clinicians, nurses, and medical directors at high and low antiepileptic prescribing facilities to determine the key drivers, indications, and attitudes towards antiepileptic use. With input from a multidisciplinary Expert Panel including a patient with early dementia and caregiver dyad, we will identify remaining knowledge gaps and potential prescribing intervention targets. Findings will address a major gap in our knowledge of the drivers and safety of off-label antiepileptic use and help identify possible intervention targets to improve nursing home prescribing safety for residents with dementia.