Duplex ultrasound in the emergency department for the diagnostic management of clinically suspected deep vein thrombosis.

OBJECTIVES To evaluate the accuracy and safety of an emergency duplex ultrasound (EDUS) evaluation performed by emergency physicians in the emergency department. METHODS Consecutive adult patients suspected of having their first episode of deep vein thrombosis (DVT) presenting to the emergency department were included in the study. All examinations were performed by emergency physicians trained with a 30-hour ultrasound course. Based on EDUS findings, patients were classified into one of three groups: normal, abnormal, and uncertain. Patients with abnormal and uncertain findings were initially treated as having a DVT. Patients with normal EDUS findings were discharged from the emergency department without anticoagulant therapy. A formal duplex ultrasound evaluation was repeated by a radiologist in all patients within 24-48 hours. Patients with normal findings on duplex ultrasound evaluation were followed up for symptomatic venous thromboembolism for up to one month. RESULTS A total of 399 patients were studied. The EDUS findings were normal in 301 (75%), abnormal in 90 (23%), and uncertain in eight (2%). All abnormal test results were confirmed by the formal duplex ultrasound evaluation, and three patients (0.8%) with uncertain findings on EDUS examination were subsequently diagnosed as having a distal DVT (positive predictive value, 95% [95% confidence interval, 92% to 95%]; negative predictive value, 100% [95% confidence interval = 99% to 100%]). No patients with normal findings on EDUS examination died or experienced venous thromboembolism at the one-month follow-up. CONCLUSIONS EDUS examination yielded a high negative predictive value and good positive predictive value, allowing rapid discharge and avoiding improper anticoagulant treatment.