Assessment of deep venous thrombosis using routine pelvic CT

The purpose of this study was to assess routine pelvic CT both for accuracy in diagnosis of deep venous thrombosis and for frequency of detection of clinically unsuspected pelvic thrombi. We reviewed the CT records of patients who had undergone pelvic CT during a 6-month period and cross-referenced these records to reports on lower extremity venous sonography to identify patients who had undergone this test within 1 week of pelvic CT. We compared the frequency and location of venous thrombosis revealed through these two tests. To identify unsuspected deep venous thrombosis detected on pelvic CT, we reviewed the charts of patients for whom a clot was reported to determine if it had been clinically suspected before CT. Among the 52 patients who underwent both CT and lower extremity venous sonography, the findings were in agreement for 49 (94%). The techniques disagreed for three patients; two patients had deep venous thrombosis detected on sonography but not on CT and the other had a clot detected on CT but not on sonography. The prevalence of unsuspected deep venous thrombosis detected by CT was 1.1%, and 73% of these patients underwent anticoagulation therapy without further confirmatory tests. Unsuspected deep venous thrombosis is commonly seen on routine pelvic CT and should be carefully sought during such examinations. CT appears to be relatively accurate in the detection of deep venous thrombosis. Thus, CT venography combined with pulmonary CT angiography may significantly increase the percentage of patients who are appropriately treated for thromboembolic disease because both deep venous thrombosis and pulmonary emboli can be identified by this combined test.