Academic and Community Hospitals differ in stroke subtype classification and cardiac monitoring: the DiVERT multi-center cohort study

Background: Cardiac monitoring strategies to detect occult atrial fibrillation (AF) post-stroke differ among healthcare institutions. This may be related to discrepancies in stroke subtype classification/adjudication, and/or consultation of cardiology specialists at Community Hospitals (CoH) and Academic Centers (AcC). Identifying the degree of heterogeneity may encourage development of guideline-directed monitoring protocols, result in higher AF detection rates and treatments, and fewer strokes.

Methods: The DiVERT (SeconDary Stroke PreVEntion ThRough Pathway ManagemenT) study was designed to characterize post-stroke cardiac monitoring practices in a hospital setting. Care pathways were assessed with in-person stakeholder interviews; patient-level data were reviewed using electronic medical records.

Results: DiVERT identified 2,475 patients with diagnoses of cryptogenic (83.6% vs. 33.1%, p < 0.001), large vessel disease (LVD) (13.3% vs. 37.0%, p < 0.001), or small vessel disease (SVD) (3.1% vs. 29.9%, p < 0.001) stroke, at CoH and AcC, respectively. CoH consulted cardiology significantly less than AcC (12.3% vs. 34.7%, p < 0.001) and ordered significantly fewer short- or long-term cardiac monitors than AcC (6.8% vs. 69.2%, p < 0.001). CoH had shorter length of stay (5.3 vs. 9.4 days, p < 0.001) and patient demographics were significantly different (p < 0.001 for age, ethnicity and race).

Conclusion: Significant heterogeneity in cardiac monitoring post-stroke exists: CoH reported 2.5-times more cryptogenic stroke than AcC yet ordered 10-times fewer short/long-term cardiac monitors to look for AF. Significant differences in patient demographics among institutions may account for this discrepancy. Regardless, efforts to reduce heterogeneity are warranted to improve AF detection and treatment and prevent recurrent stroke.