Wells Clinical Probability and D-dimer Testing: Risk Stratification for Acute Lower Extremity Deep Vein Thrombosis

Abstract

Current vascular guidelines recommend duplex ultrasound (DUS) to rule out acute lower extremity deep vein thrombosis (DVT). The combined use of D-dimer levels with Wells clinical probability scores has been proposed as an alternative method to exclude DVT and minimize the use of DUS in low-risk patients. We hypothesized that high-risk patients who would benefit from further evaluation for DVT with DUS can be reliably identified using a combination of the plasma D-dimer level and Wells score. A prospective observational study was performed. Outpatients presenting to the emergency department with a chief complaint concerning for acute lower extremity DVT underwent risk stratification into three groups on the basis of pretest probability using the Wells score and plasma D-dimer level. All patients were evaluated with whole leg ultrasound to the confirm diagnosis. Of 3087 patients, 74.2% were classified as low risk, 18.6% as moderate risk, and 7.2% as high risk. The prevalence of DVT was 7.3% on completion of DUS. The negative predictive value for a negative D-dimer level and negative Wells score was 99.8%. A positive D-dimer level with a positive Wells score had a sensitivity and specificity of 97.4% and 91.9% for DVT, respectively. The plasma D-dimer levels correlated with the DVT level (η2 = 0.22) and were highest in those with proximal DVT. The combined use of the Wells clinical probability score and plasma D-dimer level is a safe and effective method of risk stratifying emergency room patients with lower extremity symptoms concerning for acute DVT. Patients with a low probability Wells score and negative plasma D-dimer level are unlikely to benefit from DUS, and patients with a positive D-dimer level and Wells score would benefit from whole leg DUS to rule out high-risk thrombi.