Can D-dimer testing help emergency department physicians to detect acute aortic dissections?

Abstract

To determine the diagnostic accuracy of D-dimer testing for detection of acute aortic dissection. This study is a retrospective chart review of patients who had been evaluated with suspicion of acute aortic dissection. All patients' D-dimer levels were determined prior to their further work up in the emergency department. The study was conducted in a tertiary care center between February 2006-August 2008. The D-dimer assay used was the immunoturbidimetric assay, with a normal range up to 0.246 µg/ml. Statistical analysis was accomplished using Chi-square test, Student's t-test and a receiver-operating characteristics (ROC) curve analysis. Ninety-nine patients were included in the study, 30 patients were diagnosed as having acute aortic dissection and 69 patients were evaluated in non-acute aortic dissection group. In comparison of the two groups, positive D-dimer results were found to be significantly higher in acute aortic dissection group than in non-acute aortic dissection group (p=0.001). Sensitivity of the D-dimer test in detection of acute aortic dissection was found as 96.6% and the negative predictive value of the test was 97.3%. Specificity and positive predictive value of the D-dimer test were 52.2% and 46.8%, respectively. The area under the ROC curve yielded an acceptable certainty for excluding acute aortic dissection on base of negative results (AUC: 0.764; CI 95%: 0.674-0.855; p=0.001). D-dimer testing is helpful for emergency physicians in detection of patients with suspected acute aortic dissection in the emergency department.