Background: We compared investigative methods to study the diagnostic capability of ultrasonography. Subjects were 75 patients with proximal femur fractures who underwent surgery from March 2002 to December 2003.
Methods: Biochemical assays were carried out on days 3 and 7, and D-dimer levels were investigated as a time series. Ascending lower limb venography and ultrasonography were carried out nearly simultaneously in all cases on day 7.
Results: The incidence of deep vein thrombosis (DVT)-positive findings on venography was 61.3% (46/75). Venography results demonstrated a diagnostic sensitivity of 95.5% and specificity of 91.4% using a D-dimer cutoff of 1microg/ml on postoperative day 7. By ultrasonography, diagnostic sensitivity for DVT was 78.3%, and specificity was 96.5%. Among all proximal area cases, there were 18 limbs with DVT seen on venography among which were 17 limbs in cases demonstrating DVT at the same site on ultrasonography, a correct diagnosis rate of 94.4%.
Conclusions: For the diagnosis of a thrombus more proximal than the popliteal area, a location at high risk for occurrence of pulmonary embolism, ultrasonography provided results nearly on par with those obtained by venography. Thus, we believe that ultrasonography allowed a more noninvasive, accurate diagnosis and more rapid treatment in cases where D-dimer values were 1 microg/ml or more on postoperative day 7.