Combining clinical risk with D-dimer testing to rule out deep vein thrombosis

We sought to determine whether the combination of low-intermediate clinical risk of acute lower extremity deep vein thrombosis (DVT) and negative ELISA D-dimer assay can eliminate the need for duplex ultrasonography. Three hundred thirty-six patients prospectively underwent clinical risk stratification (low, intermediate, and high), D-dimer testing, and duplex ultrasonography. Thirteen of 145 intermediate-risk patients had acute DVT; 11 (85%) had a positive D-dimer. Two of 118 low-risk patients had acute DVT; both had a positive D-dimer. Intermediate-high risk stratification alone had sensitivity of 93.9% (95% CI: 80.3-98.3%) and a NPV of 98.3% (95% CI: 94.0-99.5%) for acute DVT. For all patients, a positive D-dimer alone had a sensitivity of 93.9% (95% Cl: 80.3-98.3%) and a NPV of 98.6% (95% CI: 95.1-99.6%). The combination of D-dimer and intermediate-high risk classification had a sensitivity of 100% (95% CI: 89.4 - 100%) and a NPV of 100% (95% CI: 98.9-100%). In suspected acute lower extremity DVT, the combination of intermediate-high clinical risk and positive D-dimer has a high sensitivity and NPV, possibly eliminating the need for duplex ultrasound in this group of patients. (C) 2004 Elsevier Inc.