CT venography for deep venous thrombosis: Continuous images versus reformatted discontinuous images using PIOPED II data

OBJECTIVE. This study was designed to determine whether discontinuous CT of the lower extremities for the detection of deep venous thrombosis (DVT) yields results similar to those of complete helical imaging using cases from the Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II). MATERIALS AND METHODS. In PIOPED II, CT venography followed CT angiography (CTA) to detect pulmonary embolus, using 7.5-mm continuous helical imaging from the iliac crest to the tibial plateau. DVT was detected in 105 of 737 patients (14.2%). We randomly chose 54 positive cases and 96 negative cases for our study. The continuous helical images were reformatted as 7.5-mm images and two of every three images were deleted. These images (7.5 mm; skip = 15 mm) were then sent - without identifying information - to the original reviewers. From 1 to 3.5 years had elapsed since the original interpretations. The results of the new interpretations were compared with the original CT venography consensus interpretations of PIOPED II. RESULTS. There was agreement for the presence of DVT in at least one leg (same leg) or for the absence of DVT in both legs in 133 of the 150 study patients (89%). The kappa statistic showed substantial agreement between the consensus interpretations and the test interpretations (kappa = 0.75; 95% CI = 0.64-0.86) per patient. CONCLUSION. There was good - but not perfect - agreement between continuous helical and discontinuous axial imaging for the detection of DVT. Given the vagaries of interobserver and intraobserver variation, there appears to be little difference between the two approaches. Adopting discontinuous imaging and other dose-reduction strategies can reduce pelvic radiation by more than 75%.