Transoesophageal echo-Doppler in patients suspected of a cardiac source of peripheral emboli.

The initial results of a controlled and partly blinded study aimed at evaluating the accuracy of transoesophageal echo-Doppler in detecting cardiac sources of peripheral emboli are reported. A total of 120 consecutive patients suspected of acute embolic events were entered. After completion of all investigations, the patients were classified into three groups: patients who had definitely not suffered an embolic event (controls; n = 56); patients in whom the differentiation between local thrombosis, embolic event originating from a diseased infarct-related artery or embolic event from a cardiac source was not possible (questionable cases; n = 24) and patients in whom a cardiac source of a definite embolic event was highly suspected (cardiac emboli; n = 40). Isolated interatrial septum anomaly and mitral valve prolapse were as frequent in the control group as in the embolism group. Transoesophageal echo-Doppler had a sensitivity of 83% and a specificity of 86% in correctly assigning a patient to the cardiac embolism group or to the control group. The positive and negative predictive values were 80% and 87% respectively. It is concluded that transoesophageal echo-Doppler is highly sensitive but is also specific in demonstrating cardiac sources of peripheral emboli.