Objectives To determine the diagnostic yield of echocardiography and its utility in changing medical management; and to derive a risk score to guide its use in patients with in-hospital stroke or transient ischaemic attack (TIA).
Methods We carried out a retrospective chart review from January 2009 to June 2010 of patients with acute ischaemic stroke or TIA who had undergone transthoracic echocardiography (TTE) or transoesophageal echocardiography (TOE). Clinical and imaging findings at baseline were noted and ‘potential clinically relevant findings’ identified on TTE and TOE. A multivariable logistic regression was used to identify predictors of potential clinically relevant findings on TTE or TOE and derive a risk score.
Results Of 370 patients, 307 (83.0%) had TTE and 63 (17.0%) had additional TOE. Potential clinically relevant findings on echocardiography were noted in 28 (7.6%) patients. Change in medical management was noted in 19/307 (6.2%) patients on TTE and in 7/63 (11.1%) patients on TOE. Male sex (OR 3.05, 95% CI 1.19 to 7.84; p=0.021), abnormal admission ECG (OR 4.39, 95% CI 1.79 to 10.79; p=0.001), and embolic pattern imaging at baseline (OR 2.38, 95% CI 1.05 to 5.40; p=0.038) were independent predictors of findings on TTE or TOE. A risk score including these three variables had modest discrimination (c-statistic 0.69, 95% CI 0.59 to 0.80).
Conclusions Echocardiography detected potential clinically relevant findings in a minority of patients (7.6%), but these findings changed medical management 90.5% of the time. A risk score using sex, ECG abnormality, and embolic pattern imaging at baseline could help predict which patients are more likely to have these echo findings.