The comparative value of noninvasive testing for diagnosis and surveillance of deep vein thrombosis

Abstract

This is a prospective analysis of 351 patients in two distinct groups undergoing ascending phlebography, impedance plethysmography (IPG), and/or phleborheography (PRG) within the same 24-hour period. One hundred twenty patients also had a 125I-fibrinogen uptake test (RFUT). The two patient groups consisted of the following: those patients evaluated because of suspicion of deep vein thrombosis (DVT) (diagnostic) and those patients at high risk for postoperative DVT (total joint replacement) who had routine noninvasive testing and ascending phlebography (surveillance). The overall sensitivities for IPG and PRG were significantly better in the diagnosis group (71% [69 of 97 patients] and 78% [82 of 105], respectively) compared with the surveillance group (20% [14 of 71] and 27% [17 of 63], respectively) (p < 0.0001). The sensitivities for IPG and PRG detecting proximal (A/K) thrombi was 83% (68 of 82 patients) and 92% (79 of 86) in the diagnosis group compared with 32% (11 of 34) and 33% (9 of 27) in the surveillance group (p < 0.0001). Although there was no difference in overall incidence of DVT between the diagnosis group (56%, 118 of 212 patients) and the surveillance group (55%, 76 of 139), the results can be explained by the difference in A/K thrombi (84% [99 of 118] and 47% [36 of 76]) (p < 0.001) and occlusive A/K thrombi (84% [58 of 69] and 23% [7 of 31]) (p < 0.0001), respectively. Of the patients with A/K thrombi, 97% (67 of 69) in the diagnosis group had hemodynamically detectable thrombi compared with only 48% (12 of 25) in the surveillance group (p < 0.001). Combining the RFUT with the noninvasive studies for surveillance significantly improved the sensitivity for both A/K and distal thrombi. Patient selection also appears to have a significant influence on the results of the combination of IPG and RFUT when the current surveillance group is compared with similarly performed studies in a previously reported diagnosis group. The location and magnitude of thrombi in any patient population can be skewed depending on indications and timing of testing, thereby significantly affecting the sensitivity of noninvasive tests. IPG and PRG are reliable for evaluating patients with suspected DVT. However, patients with postoperative DVT have a high incidence of nonocclusive thrombi. Because noninvasive hemodynamic tests cannot identify accurately postoperative DVT, they cannot be used to generate epidemiologic data or as end points for studies evaluating efficacy of prophylaxis in patients undergoing total joint replacement, and anatomic studies of the deep venous system continue to be required.