OBJECTIVE: To evaluate whether pulmonary embolism (PE), as detected by perfusion lung scan, could be predicted by the ultrasonic (US) characteristics of the thrombi in patients with deep venous thrombosis (DVT) of the lower limbs. PATIENTS: Ninety-three consecutive patients with DVT and no symptoms of lung involvement (52 men, 41 women; mean age, 67 +/- 17 years). MEASUREMENTS AND RESULTS: The degree of thrombotic involvement of the lower limbs was assessed using a US score system ranging from 1 (indicating a subsegmental, nonocclusive thrombus) to 16 (massive, occlusive). According to the echographic and color-Doppler features, the thrombi were classified in terms of echoreflectivity, adhesiveness to the vein wall, and organization. The diagnosis of PE (PIOPED criteria) was highly probable in 46% of the patients, intermediate in 15%, low in 8%, and very low/normal in 31%. No correlations were found between the lung scan findings on one side and the venous scoring system or the US features of the thrombi on the other side. CONCLUSIONS: While confirming that the prevalence of PE in patients with DVT is elevated, we failed to define a subgroup of patients at higher risk. Our data imply that lung scan should be used extensively for the detection of silent PE and that anticoagulation should not be graded on US findings.
OBJECTIVE: To evaluate whether pulmonary embolism (PE), as detected by perfusion lung scan, could be predicted by the ultrasonic (US) characteristics of the thrombi in patients with deep venous thrombosis (DVT) of the lower limbs. PATIENTS: Ninety-three consecutive patients with DVT and no symptoms of lung involvement (52 men, 41 women; mean age, 67 +/- 17 years). MEASUREMENTS AND RESULTS: The degree of thrombotic involvement of the lower limbs was assessed using a US score system ranging from 1 (indicating a subsegmental, nonocclusive thrombus) to 16 (massive, occlusive). According to the echographic and color-Doppler features, the thrombi were classified in terms of echoreflectivity, adhesiveness to the vein wall, and organization. The diagnosis of PE (PIOPED criteria) was highly probable in 46% of the patients, intermediate in 15%, low in 8%, and very low/normal in 31%. No correlations were found between the lung scan findings on one side and the venous scoring system or the US features of the thrombi on the other side. CONCLUSIONS: While confirming that the prevalence of PE in patients with DVT is elevated, we failed to define a subgroup of patients at higher risk. Our data imply that lung scan should be used extensively for the detection of silent PE and that anticoagulation should not be graded on US findings.