Diagnosing Pulmonary Embolism with Four–Detector Row Helical CT: Prospective Evaluation of 216 Outpatients and Inpatients

PURPOSE: To prospectively evaluate multi-detector row helical computed tomography (CT) for the diagnosis of pulmonary embolism (PE), with focus on the proportion of diagnostic studies and frequency of subsegmental and chronic PE. MATERIALS AND METHODS: Institutional review board approval and patient consent were not required. A total of 220 consecutive CT angiography studies, 124 (56%) of which involved inpatients, were assessed. Thoracic CT angiography was performed in 216 patients; there were 101 male (age range, 25-93 years; median, 66 years) and 115 female (age range, 15-98 years; median, 67 years) patients. Contiguous 1.25-mm sections were acquired through the entire thorax after injection of 140 mL of contrast material at a rate of 4 mL/sec. CT venography was combined with thoracic CT angiography in 178 patients over 40 years of age. CT studies were interpreted first in the emergency setting and subsequently by two experienced chest radiologists. Untreated patients with normal results were contacted by telephone after 3 months. Proportions were compared with the chi(2) test, and agreement was assessed by calculating the kappa statistic (for thoracic CT angiography). RESULTS: Concordance between the two reading sessions was good (kappa = 0.88; 95% confidence interval: 0.77, 0.98). The proportion of nondiagnostic thoracic CT angiography studies was 9% (20 of 220). PE was found in 54 (24.5%) of 220 cases; eight (15%) of 54 patients had only subsegmental PE, which was associated with a calf vein thrombosis in two patients, and six patients (11%) had chronic PE. CT venography demonstrated venous thrombosis in 15% (26 of 178) of the patients thus studied, as well as in 45% (21 of 47) of patients with positive results at thoracic CT angiography and 4% (five of 131) of patients with negative results at thoracic CT angiography. The 3-month rate of thromboembolic events after negative results was 1.8% (two of 111) (95% confidence interval: 0.2%, 6.4%). CONCLUSION: Multi-detector row CT enables diagnosis in 91% of cases and identification of isolated subsegmental or chronic PE in a relatively high proportion of patients.