Algorithms for the diagnosis of deep-vein thrombosis in patients with low clinical pretest probability

The objective of this current trial was to evaluate the rate of deep-vein thrombosis (DVT) in patients after low clinical risk stratification and to evaluate the value of D-dimer and different imaging techniques in the diagnostic algorithm. A total of 99 consecutive patients were included in this prospective trial. After clinical risk assessment, D-dimer was determined. Final diagnosis was based on the results of duplex sonography, in cases of indeterminate scans on those of ascending venography. Three months after admission, follow-up investigations were performed by a telephone interview to evaluate possible further venous thromboembolism. Final diagnosis was based on the results of colour Doppler ultrasound in 92.9% and on those of venography in 7.1%. DVT was diagnosed in 2%, D-dimer was positive in 48.4%, giving a sensitivity of 100%, a specificity of 52.7% and a negative predictive value of 100%. Follow-up was possible in 89.9% - no further thromboembolic event occurred. In this specific patient group, a negative D-dimer excludes DVT and can therefore reduce the number of imaging procedures by one-half, which, on the contrary, is necessary in patients with positive D-dimer. (C) 2002 Elsevier Science Ltd. All rights reserved.