Association of Resting Right‐to‐Left Shunt With Ischemic Stroke Among Patients Referred for Contrast Echocardiographic Screening: A Retrospective Cross‐Sectional Study
Liang Tang, Shutong Zou, Binhan Wu, Meixiang Wang, Bo Zhang, Gecai ChenABSTRACT
Background
Resting right‐to‐left shunt (RLS) may represent a persistent functional shunt phenotype compared with provoked‐only shunting. However, its association with ischemic stroke among patients clinically referred for contrast‐enhanced transthoracic echocardiography (c‐TTE)‐based RLS assessment remains incompletely characterized. Because such patients are referred according to clinical suspicion rather than population‐based screening, findings from this setting require cautious interpretation.
Methods
We conducted a single‐center retrospective cross‐sectional study at Taizhou People's Hospital. Consecutive outpatients and inpatients clinically referred for c‐TTE‐based RLS assessment between July 2024 and June 2025 were screened. Patients with incomplete key clinical, laboratory, or echocardiographic data were excluded, and analyses were performed on a complete‐case basis. Ischemic stroke was identified on the basis of compatible focal neurological symptoms and radiologic evidence of acute infarction. Resting RLS was defined as microbubble passage into the left‐sided cardiac chambers without provocation and was graded semiquantitatively from 0 to 3. Multivariable logistic regression was used to examine the association between resting RLS and ischemic stroke status after adjustment for clinically relevant covariates.
Results
Among 1328 consecutive patients clinically referred for c‐TTE‐based RLS assessment, 717 with complete data were included in the final analytic cohort. Ischemic stroke was present in 161 patients (22.5%), and resting RLS was present in 156 patients (21.8%). Resting RLS was more frequent in patients with stroke than in those without stroke (32.9% vs. 18.5%, p < 0.001). In multivariable analysis, resting RLS was associated with ischemic stroke status (adjusted OR 2.30, 95% CI 1.52–3.50; p < 0.001). When analyzed as an ordinal variable, each one‐grade increase in resting RLS was associated with higher odds of stroke status (adjusted OR 1.50, 95% CI 1.23–1.84; P for trend < 0.001). Exploratory ROC analysis showed an AUC of 0.715 for the model including resting RLS.
Conclusions
In this clinically referred complete‐case cohort undergoing c‐TTE‐based RLS assessment, resting RLS and higher resting RLS grade were associated with ischemic stroke status after adjustment for available conventional vascular risk factors. Given the retrospective referral‐based design, complete‐case selection, and limited availability of medication information, these findings should be interpreted as hypothesis‐generating and require prospective validation in broader, systematically evaluated populations.