Diagnosis of deep-vein thrombosis: adherence to guidelines and outcomes in real-world health care.

Current guidelines recommend optimised algorithms for diagnosis of suspected deep-vein thrombosis (DVT). There is little data to determine to what extent real-world health care adheres to guidelines, and which outcome in terms of diagnostic efficiency and safety is achieved. This registry involved patients with clinically suspected DVT of the leg recruited in German ambulatory care between October and December 2005. Registry items were: diagnostic methods applied, diagnostic categories at day 1, and venous thromboembolic events up to 90 days in patients without firmly established DVT. A total of 4,976 patients were recruited in 326 centres. Venous ultrasonography was performed in 4,770 patients (96%), D-dimer assay in 1,773 patients (36%) and venography in 288 patients (6%). At day 1, DVT was confirmed in 1,388 patients (28%), and ruled out in 3,389 patients (68%), and work-up was inconclusive in 199 patients (4%). The rate of venous thromboembolism at 90 days was 0.34% (95% confidence interval [CI]: 0.09 to 0.88) in patients in whom the diagnosis of DVT had been ruled out, and 2.50% (95% CI: 0.69 to 6.28) in patients with inconclusive diagnostic workup. This nationwide evaluation in German ambulatory care revealed that the diagnostic work-up for suspected DVT did not adhere to current guidelines. However, the overall diagnostic safety was excellent, although there is potential for improvement in a well defined minority of patients.