All young(18-45 yrs)IS patients underwent both TTE and TEE as standard of care at our comprehensive stroke center from 2011-2018. Stroke classification at discharge was performed according to TOAST criteria by an attending stroke physician unaware of the purposes of the study. All variables were entered through retrospective chart review. To reduce bias, an independent blinded vascular neurologist reviewed the dataset and assigned utility score of 1 to either TTE or TEE or both when test yield contributed to stroke mechanism determination. If TEE confirmed the initial TTE finding, no utility point was assigned to TEE. A point was assigned to both when both TTE and TEE findings had additive utility.

Among 188 study patients (mean age 37±5years, 41% men, 74% African American, median NIHSS 3, IQR 1-8), cardioembolic stroke was diagnosed in 53 (28%). Diagnostic yield for any abnormalities was 32% for TTE, 49% TEE, and 54% both. The utility of TTE was 13% and TEE 18% (p=0.4). However, the combined utility was 28%. There was non-significant correlation between the utility for TTE and TEE (kappa=0.07, p=0.37) and the overall agreement in 75.5% in no utility(i.e. findings insufficient to determine cardioembolic stroke mechanism). TTE is adequate for diagnosis in about a third of the cases, and subsequent performance of TEE further improves diagnosis.

Our study showed that in predominantly African American young IS patients, both TTE and TEE have high diagnostic yield at lower individual utility. However, their combined utility justifies performance of both tests in the young.