DISCREPANCIES BETWEEN PHLEBOGRAPHY AND DU PLEX ULTRASONOGRAPHY FOR THE DIAGNOSIS OF ASYMPTOMATIC VENOUS THROMBOSIS AFTER TOTAL HIP-ARTHROPLASTY - FALSE-NEGATIVE VENOGRAPHY OR FALSE-POSITIVE DUPLEX ULTRASONOGRAPHY

Objectives. Assess the diagnostic performance of phlebography and disc repancies with duplex ultrasonography in screening for asymptomatic de ep vein thrombosis after total hip arthroplasty: Search for arguments which would favor contributing false-negatives or false-positives to o ne of the two exploration methods. Patients and methods. The study inc luded 24 patients who underwent the 2 explorations independently betwe en day 7 and day 14 in a multicentric therapeutic trial of a new hepar in. Discrepancies between the two techniques were recorded. Diagnostic performance of phlebography was calculated from contingency tables. T he phlebograms were then reviewed with knowledge of the duplex ultraso nographic findings. The course of the venous thrombus after treatment was monitored with duplex ultrasonography. Results. Phlebography allow ed the diagnosis of thrombus formation in 9 patients. Ultrasonography provided the diagnosis in 14 cases out of 24. Sensitivity of phlebogra phy compared with duplex ultrasonography was 64 % (95 % CI = 35.1 - 87 .2) (9/14) and specificity was 100 % (95 % CI = 69.1 - 100) (10/10). S everal localizations were not visualized with phlebography: 3 partiall y obstructive thrombi near the femoral junction, and 38 distal thrombi including 31 soleus thrombi (fig. 5). When the phlebograms were revie wed again, defects compatible with a partial thrombus were found for t he 3 proximal localizations and for 2 of the distal localizations (fig . 1, 2, 4) None of the soleus localizations were visualized. Duplex ul trasonographic follow-up concerned 12 of the 14 patients with a thromb us identified by duplex ultrasonography involving 44 localizations. In five case, the operator was different from the operator for the initi al duplex ultrasonography. Follow-up revealed: 1 new localization, 37 narrowings and 6 repermeabilizations. Discussion. The coherence of the duplex ultrasonographic follow-up and the presence of images compatib le with partial thrombus on the revised phlebograms which had been ini tially interpreted as wash out flow constitute a group of arguments su ggesting that the discrepancies observed should be considered as phleb ography false-negatives. This defect in the sensitivity of phlebograph y has been reported by others in the literature both for distal and pr oximal localizations.