Introduction: The aim of this study is to assess concordance between transthoracic echocardiography (TTE) performed by neurologists and cardiologists for the screening of cardioembolic sources in patients with ischemic stroke.
Methods: We examined prospectively 27 consecutive patients with ischemic stroke. A potential source of embolism in all of the patients was sought by means of a TTE study. This study was performed by one neurologist after one-month full-time training on the technique and independently by one cardiologist whose results were considered as the reference standard. Regarding the risk of cardioembolism, findings from TTE studies were considered as normal, or indicative of either moderate or high risk, according to the TOAST criteria. Agreement between the two observers on the diagnosis of cardioembolic source was analyzed with the Cohen's Kappa Test (K). A K > 0,6 was considered as good agreement.
Results: The 27 patients studied had a mean age of 68,7 +/- 10,2 years and 70 % were men. A cardioembolic source was detected in 9 patients by the cardiologist and 8 of these were also identified by the neurologist: in 1 patient a mitral annulus calcification was not detected in the examination performed by the neurologist. In 2 patients with poor echocardiographic window, the neurologist underestimated the left ventricular function. Diagnostic agreement between both examinations was 88,9 % (K: 0,77; p = 0,001).
Conclusion: Compared with cardiologists, trained neurologists show a fairly high degree of accuracy in the TTE screening of patients with stroke. Those cases with poor echocardiographic window or with abnormal findings should be confirmed by the cardiologists.