Age-adjusted D-dimer, clinical pre-test probability-adjusted D-dimer, and whole leg ultrasound in ruling out suspected proximal and calf deep venous thrombosis.

Abstract

D-dimer (DD) and ultrasonography (US) are part of the diagnostic workup for lower-extremity deep vein thrombosis (DVT). Recent studies have shown that adjusting DD level cut-offs by age or clinical pre-test probability (PTP) decreases the use of US. We compared diagnostic accuracy of PTP-adjusted DD and age-adjusted DD in 3883 patients (F: 61.1%; age: 65.3 ± 16.8 y) referred to our unit for clinically suspected DVT. All patients underwent clinical evaluation, DD, and US. Proximal DVT was detected in 477 (12.4%) patients, and distal DVT was isolated in 342 (8.9%) patients. In the remaining 3064 patients there were 23 venous thromboembolic events (0.75%, 95% CI: 0.50-1.12) during the 3-month follow-up. The specificities of DD, age-adjusted DD, and PTP-adjusted DD in patients without high PTP levels were 47% (95% CI: 45-49), 61% (95% CI: 59-62), and 67% (95% CI: 65-68), respectively. The negative predictive value (NPV) was 96% (95% CI: 95-97) for all diagnostic strategies. When only proximal DVTs were considered, the NPV increased to 99% (95% CI: 98-99). US was avoided in 37% (95% CI: 36-38) of patients with a fixed cut-off DD, 48% (95% CI: 47-50) with age-adjusted DD, and 52% (95% CI: 51-54) with PTP-adjusted DD. The failure rate for all DVTs of DD, age-adjusted DD, and PTP-adjusted DD was 2.0% (95% CI: 1.6-2.5), 2.7% (95% CI: 2.2-3.2), and 2.5% (95% CI: 2.1-3.0), respectively. Compared with the standard DD cut-off, both age-adjusted and PTP-adjusted DD reduced the proportion of patients who required US at the cost of a small increase in failure rate.