Extended lower limb venous ultrasound for the diagnosis of proximal and distal vein thrombosis in asymptomatic patients after total hip replacement

Objective. To assess the performance of extended lower limb venous ultrasound (US) for the diagnosis of asymptomatic deep vein thrombosis (DVT) and to estimate a 3-month DVT incidence on repeated US after total hip replacement. Design. Diagnostic performance study and prospective cohort study. Materials and methods. US was compared to phlebography in 70 consecutive patients and interobserver agreement was assessed in the last 48 patients at day 8. US was repeated in these 48 patients at day 13 and day 90. Results. Phlebography demonstrated a DVT in 18170 (26%) patients, with five proximal and 13 distal and US in 23170 (33%) patients, with eight proximal and 15 distal. Sensitivity and specificity of US with 95% Cl were 94% (73-100) and 89% (76-96), respectively. Sensitivity in isolated distal vein thrombosis was 92% (67-99). The Kappa coefficient for agreement between observers was 0.84 (0.66-1.00). Follow-up showed a DVT in 15148 (31%) patients on day 8, in 20148 patients (42%) on day 13. DVT recurred in two patients during follow-up. Conclusions. The incidence of asymptomatic DVT is still significant despite prophylaxis but most DVTs remain distal and occur in the first 2 weeks. Extended US could replace phlebography for systematic screening in clinical trials using surrogate endpoints in view of its high accuracy and reliability.