Duplex Ultrasonography for the Detection of Deep Vein Thrombi After Total Hip or Knee Arthroplasty

Abstract

The usefulness of real-time duplex ultrasonography (DU) as a screening test for deep vein thrombosis (DVT) in high-risk patients remains uncertain. To determine the sensitivity and specificity of DU for the detection of DVT, the authors prospectively studied 178 consecutive patients after total hip (n=113) or total knee (n = 66) arthroplasty. The deep veins from the inguinal ligament to the ankle were examined first by continuous wave and then by pulsed Doppler signals as needed with real-time gray-scale ultrasound imaging using the criteria of vein noncompressibility to define DVT. Ascending contrast venography was performed within twelve hours after DU studies. Venograms and DU were interpreted independently. DU was attempted on 177 lower extremities (2 patients refused) but was judged adequate for interpretation for only 145 (82%). Venography could not be performed for 28 lower extremities and was technically inadequate for 8 studies. The primary analysis included 119 examinations for which adequate DU and ascending venograms were interpreted. DU was positive in 17 of 27 lower extremities with DVT (23 calf, 4 proximal) diagnosed by venography (sensitivity = .63; 95% confidence interval [CI] = .42 to .81), and DU was negative in 85 of 92 lower extremities with normal venograms (specificity = .92; 95% CI = .85 to .97). A secondary analysis of 81 prospectively collected anatomically complete DU studies demonstrated a sensitivity of .80 (95% CI = .56 to .94) and a specificity of .90 (95% CI = .80 to .96). The authors conclude that real-time DU is not sensitive for the detection of venographically demonstrable calf DVT in the operated-on lower extremity following total hip or knee arthroplasty. Real-time DU is specific for the diagnosis of DVT following total hip or knee arthroplasty.