Improving the diagnostic process for deep vein thrombosis in orthopaedic outpatients

Prompt diagnosis of proximal lower extremity deep vein thrombosis in outpatients is critical because of the risk of pulmonary embolism. Our purpose was to determine the accuracy of orthopaedists' clinical decisions regarding the diagnosis of proximal deep vein thrombosis in outpatients. A nationally representative random sample of 2300 orthopaedists received a survey of six clinical vignettes. They were asked to estimate the probability of proximal lower extremity deep vein thrombosis using defined criteria and to specify their planned diagnostic tests. A clinical decision rule and evidence-based diagnostic test recommendations from the general literature served as the gold standard for comparison. Six-hundred seventy-six (29%) surgeons completed the survey. The orthopaedists' planned diagnostic tests differed from the gold standard, but these differences varied depending on the probability of deep vein thrombosis. For the moderate and high risk vignettes, the diagnostic test recommendations agreed with the gold standard approximately 70% of the time. With the exception of gender, no differences were found between respondents and nonrespondents. Orthopaedists' approach to the diagnosis of deep vein thrombosis in outpatients potentially could be improved by applying a clinical decision rule and current evidence on diagnostic test usage.