Incidental detection of pulmonary emboli on routine MDCT of the chest

OBJECTIVE. The objectives of our study were to assess the prevalence of pulmonary embolism incidentally detected on routine MDCT of the chest and to determine whether the use of wide window settings can improve detection of unsuspected pulmonary embolism. MATERIALS AND METHODS. A retrospective review of routine contrast-enhanced MDCT scans of the chest obtained in 589 patients was undertaken. CT angiograms obtained for suspected pulmonary embolism or thoracic aorta disease were not considered. Image evaluation was performed on a dedicated workstation during two separate review sessions using different window settings: standard, with a width of 400 H and level of 40 H; and wide, with a width of 600 H and level of 100-150 H. The quality of vascular enhancement was recorded. RESULTS. Eight examinations were excluded because of poor quality. Unsuspected pulmonary embolism was present in 20 (3.4%) of 581 patients with an inpatient prevalence of 4.0% (19/474) and outpatient prevalence of 0.9% (1/107). Fourteen patients (70.0%) with unsuspected pulmonary embolism had cancer. The proximal extent of pulmonary embolism involved the main pulmonary artery in five patients, a lobar artery in five, and a segmental artery in 10; isolated subsegmental thrombi were never found. The use of wide window settings allowed detection of pulmonary embolism in two more patients than did the standard settings. CONCLUSION. Unsuspected pulmonary embolism can be found in a significant number of patients undergoing a routine MDCT study of the chest, with a higher prevalence among inpatients with malignancy. The use of wide window settings is recommended when interpreting routine CT scans of the chest to improve detection of unsuspected pulmonary embolism.