Validation of the Canadian Clinical Probability Rule and Use of Bedside D-Dimer Testing in the Diagnosis of Acute Venous Thrombosis in an Emergency Room Setting

Abstract

PURPOSE: To validate the Canadian clinical probability tool and utility of bedside rapid D-dimer testing using SimpliRed® among patients with suspected acute deep-vein thrombosis (DVT) in an emergency room setting. METHODS: Consecutive patients with clinically suspected DVT in a leg and no past history of symptomatic DVT were evaluated in the UC Davis emergency room by emergency room staff using the clinical probability tool developed by Wells and colleagues [Lancet,1997; 350:1795], which was provided on a one page form. After completing the form, each patient underwent venous ultrasound (US) imaging, whole blood D-dimer testing using SimpliRed®, and D-dimer measurement using a sensitive ELISA technique. All patients were followed for 3 months. Diagnosis of thromboembolism required objective confirmation of DVT using US or venography, or confirmation of pulmonary embolism using a high-probability lung scan or pulmonary arteriography. RESULTS: Of 102 patients who were evaluated, 17 (17%) were diagnosed as having DVT initially; none of the 85 in whom DVT was excluded developed thromboembolism within 3 months. Ten of 17 (59%, CI 36%–82%) who met criteria for ‘high-probability’, 6 of 44 (13%, CI 5%–21%) with ‘intermediate probability’ and 1 of 41 (2.4%, CI 0.5%–7%) with ‘low probability’ had objectively confirmed DVT. This compares to published values of 49%, 14% and 3%, respectively, from two Canadian emergency rooms. Forty-one of the 102 (40%) patients had an ‘alternate diagnosis as likely or greater than DVT’, which lowered the probability of DVT by one or more levels in 35 (85%) cases. Thirty-eight of the 41 (93%) patients with an ‘alternate diagnosis’ did not have DVT; the 3 who did have DVT were thought to have cellulitis (n = 2) or knee hyperextension (n = 1). The negative predictive value of the SimpliRed rapid D-dimer test was 100%, 100% and 71% (5 of 7) in the low, intermediate and high probability groups. Both ‘false negative’ SimpliRed® tests had normal levels of D-dimer using a sensitive ELISA method, and careful re-review of the US studies revealed findings consistent with subacute or chronic DVT. CONCLUSION: The Canadian probability tool for DVT appears to be a valid instrument in our emergency department. Similarly, we documented the high negative predictive value of bedside D-dimer (SimpliRed®) testing among patients classified as ‘low’ or ‘intermediate’ probability for having DVT, as reported in the literature. Radiographic misinterpretation of subacute or chronic DVT as acute DVT may partially explain the lower negative predictive value of SimpliRed® testing in patients with high clinical probability of having DVT.