Introduction: Many strokes arise from a distant embolic source such as the heart or large arteries. It has recently been proposed that a minimum evaluation to identify such sources should include imaging of the intracranial and extracranial arteries, an echocardiogram, and continuous heart-rhythm monitoring for ≥24 hours. We sought to evaluate the utilization rates of these tests in a nationally representative cohort of U.S. patients with ischemic stroke.
Methods: We analyzed inpatient and outpatient claims data from 2009-2014 in a 5% sample of Medicare beneficiaries ≥66 years of age. Using validated ICD-9-CM codes, we identified all patients with inpatient admission for ischemic stroke. Our outcomes were the following recommended diagnostic evaluations: intracranial arterial imaging (CTA, MRA, or TCD), extracranial imaging (CTA, MRA, or Doppler ultrasound), echocardiography (transthoracic or transesophageal), and heart-rhythm monitoring (Holter or loop recorder). We used CPT codes to identify whether these tests were performed within 3 months prior to and 6 months after the stroke.
Results: We identified 23,475 patients with ischemic stroke. Extracranial vascular imaging was performed in 20,600 patients (87.8%; 95% confidence interval [CI], 87.3-88.2%), intracranial vascular imaging in 11,726 patients (50.0%; 95% CI, 49.3-50.6%), echocardiography in 19,278 patients (82.1%; 95% CI, 81.6-82.6%), and heart-rhythm monitoring in 2,063 patients (8.8%; 95% CI, 8.4-9.2%). Only 49 (0.2%) underwent the complete recommended diagnostic battery for the detection of embolic sources. These rates were not substantially different even after excluding patients with known atrial fibrillation, the most common source of embolic stroke.
Limitations: We could not capture inpatient cardiac telemetry that was not formally interpreted by a physician.
Conclusions: Among Medicare beneficiaries with acute ischemic stroke, the majority did not receive a full diagnostic evaluation for potential embolic sources of stroke. Promoting the widespread adoption of a more systematic interrogation of potential etiologies of stroke may allow for the implementation of more personalized secondary stroke prevention strategies.