Venous ultrasound testing for suspected thrombosis: incidence of significant non-thrombotic findings.

Duplex ultrasound (US) is used to "rule out" deep venous thrombosis (DVT), but can also diagnose other causes of leg pain or swelling in Emergency Department (ED) patients. Recent literature suggests that US imaging is unnecessary among patients with low or moderate clinical probability of DVT with a normal D-dimer. We attempted to determine the incidence of clinically important incidental findings detected using venous US imaging in patients with suspected lower extremity DVT. We conducted a retrospective chart review of all ultrasounds performed by the non-invasive vascular laboratory on ED patients > 18 years old. Results were classified: normal, DVT, or incidental finding. The latter were classified as clinically significant major findings if the diagnosis led to immediate and specific treatment to prevent morbidity, or clinically significant minor findings. A total of 484 US studies were reviewed; 179 were excluded (arterial studies, penetrating trauma, upper extremity US). Findings among 305 studies were: 238 (78%) normal, 28 (9%) DVT, and 39 (12%) incidental findings. Among 39 incidental findings, 10 were clinically significant major findings and 29 clinically significant minor findings. Clinically significant major findings included: pseudoaneurysm, arterial occlusive disease, vascular graft complication, compartment syndrome, and tumor. Among 38 abnormal US studies that required immediate treatment, DVT comprised 74% (95% confidence interval 59%-85%) and important major incidental findings 26% (95% confidence interval 14%-41%). Among ED patients who underwent US to evaluate leg pain and swelling, 26% of positive studies showed clinically important findings other than DVT. Further research is needed to determine if D-dimer plus a clinical probability tool will include or exclude the patients with clinically significant major findings.