Summary. Background: Pretest clinical probability with the Wells rule and D-dimer have been widely investigated for the diagnosis of symptomatic proximal deep vein thrombosis (DVT) of the lower limbs, but they have not been formally tested for symptomatic isolated distal DVT diagnosis.
Objective: To evaluate the diagnostic accuracy of the Wells rule and D-dimer for isolated distal DVT.
Design, Setting, and Patients: This was a single-center, cross-sectional study including 873 consecutive outpatients with suspected DVT, in whom pretest clinical probability determination, D-dimer determination (STA Liatest; cut-off of < 500 ng mL−1) and complete compression ultrasonography of both lower limbs were performed.
Results: The isolated distal DVT prevalence was 12.4% (90/725). The sensitivity of the Wells rule for isolated distal DVT was 47% (95% confidence interval [CI] 36–57%), the specificity was 74% (95% CI 70–77%), and the negative and positive predictive values were 91% (95% CI 88–93%) and 20% (95% CI 15–26%), respectively. Patients with isolated distal DVT had higher D-dimer levels than patients without DVT (1759 ± 1576 vs. 862 ± 1079 ng mL−1, P = 0.0001). D-dimer was negative in 13 patients with isolated distal DVT. D-dimer sensitivity and specificity for isolated distal DVT were 84% (95% CI 75–91%) and 50% (95% CI 46–54%), respectively, with a negative predictive value of 96% (95% CI 93–98%). In patients with low pretest clinical probability, the D-dimer negative predictive value was 99% (95% CI 95–100%).
Conclusion: In clinically suspected DVT with negative proximal compression ultrasonography, pretest clinical probability with the Wells rule has a low diagnostic accuracy for isolated distal DVT. D-dimer has a better negative predictive value, but alone it does not exclude isolated distal DVT. In patients with low pretest clinical probability, D-dimer had a negative predictive value of > 95% for isolated distal DVT.