Incidence and distribution of lower extremity deep venous thrombosis at indirect computed tomography venography in patients suspected of pulmonary embolism

Indirect computed tomography (CT) venography reportedly provides high accuracy for detection of venous thrombosis in patients suspected of pulmonary embolism (PE). Nevertheless, the extent of the scanning range for lower limb and abdominal veins remains to be determined. It was the objective of this study to investigate the distribution of venous thrombosis in order to identify the most appropriate extent of scanning range when using CT venography. We reviewed 1,408 combined CT pulmonary angiographies (CTPA) and indirect CT venographies of the lower limbs, performed in patients suspected of PE. Percentage of venous thromboembolism (VTE), which includes PE and/or venous thrombosis was calculated. Location and the upper end of clots were recorded in 37 venous segments per patient from calf to diaphragm. PE, venous thrombosis and VTE, were found respectively in 272 (19.3%), 259 (18.4%) and 329 (23.4%) patients. Addition of CT venography to CTPA increased depiction of VTE in 17.3%. The upper end of venous thrombosis was located below the knee in 48%, between knee and inguinal ligament in 36% of the patients, and above the inguinal ligament in 15%. Ninety-six patients had thrombosis in a single vein, of which none occurred above the iliac crests in a patient without PE at CTPA. In conclusion, when added to CTPA, optimal scanning of CT venography should extent from calves to the iliac crests in patients suspected of VTE.