Comparative Safety of Seizure Prophylaxis within the Medicare Program

PI Name(s): Moura, Lidia Maria Veras Rocha de Project Title: Comparative Safety of Seizure Prophylaxis within the Medicare Program 1R01AG073410-01 Each year thousands of Medicare beneficiaries with Alzheimer’s disease and/or Alzheimer’s-related dementias (AD/ADRD) receive anticonvulsant drugs for seizure prophylaxis. These drugs are life-saving given the severity of seizure sequalae, but also life-threatening because of serious adverse events. Patients with AD/ADRD are particularly vulnerable given their high susceptibility to both stroke and adverse events. The decision to start and stop anticonvulsants also might change when the patient have AD/ADRD, in part because anticonvulsant therapy management becomes more complex when patients have difficulty with therapy having a narrow therapeutic range or difficulty recognizing early complication symptoms. There are no trial data to guide decisions about prophylaxis after an ischemic stroke among patients with or without AD/ADRD. The sparse and conflicted existing literature has been limited by the difficulty assessing relevant measures, e.g., stroke severity or AD/ADRD status, in claims. Medicare claims offer potential, but lack well-validated definitions for stroke severity, AD/ADRD status, seizures, or adverse effects. We will address these issues using a novel dataset with individual-level linked longitudinal information from registries, electronic health records, and Medicare claims, and will develop prediction models for key baseline and outcome variables. Then, using a random 20% sample of national traditional, fee-for-service Medicare claims over 18 years (2006-23), we will emulate clinical trials assessing two critical decisions in the setting of first mild to moderate ischemic stroke among patients with or without AD/ADRD: a) initiating outpatient prophylaxis versus not; and b) stopping prophylaxis at six months versus continuing. We will apply multiple analytical tools to address confounding and other challenges. We will stratify our population based on AD/ADRD status because the patterns of strokes, post-stroke seizures, and treatment complications differ among patients with compared to those without AD/ADRD, as could the clinical management patterns. We have three aims: 1) To validate claims-based prediction models for our study variables; 2) To emulate a clinical trial comparing outpatient prophylaxis initiation versus not; and 3) To emulate a clinical trial comparing stopping outpatient prophylaxis at six months versus continuing. This comparative safety study could inform future policy and clinical care and improve stroke and AD/ADRD care, e.g., through adjustments in Medicare quality incentives and clinical guidelines. Moreover, these data could help inform patients, families, clinicians, and policy makers about how we can improve care for patients who could require seizure prophylaxis.