Semiquantitative Latex Agglutination D-Dimer Assay Mayo Clin Proc, February 2004, Vol 79 164 Mayo Clin Proc. 2004;79:164-168 164 © 2004 Mayo Foundation for Medical Education and Research Original Article Sensitivity and Specificity of the Semiquantitative Latex Agglutination D-Dimer Assay for the Diagnosis of Acute Pulmonary Embolism as Defined by Computed Tomographic Angiography DAVID A. FROEHLING, MD; PETER L. ELKIN, MD; STEPHEN J. SWENSEN, MD; JOHN A. HEIT, MD; V. SHANE PANKRATZ, PHD; AND JAY H. RYU, MD From the Division of Area General Internal Medicine (D.A.F., P.L.E.), Department of Radiology (S.J.S.), Division of Cardiovascular Dis- eases and Internal Medicine (J.A.H.), Division of Biostatistics (V.S.P.), and Division of Pulmonary and Critical Care Medicine and Internal Medicine (J.H.R.), Mayo Clinic College of Medicine, Roches- ter, Minn. This study was supported in part by a grant from Mayo Foundation. Address reprint requests and correspondence to David A. Froehling, MD, Division of Area General Internal Medicine, Mayo Clinic College of Medicine, 200 First St SW, Rochester, MN 55905 (e-mail: froehling.david@mayo.edu). • Objective: To determine the sensitivity and specificity of the semiquantitative latex agglutination plasma fibrin D-dimer assay for the diagnosis of acute pulmonary embo- lism by using computed tomographic (CT) angiography as the diagnostic reference standard. • Patients and Methods: From January 1, 1998, to June 26, 2000, patients who had both semiquantitative latex agglutination plasma fibrin D-dimer testing and CT angi- ography for suspected acute pulmonary embolism were selected for the study. A D-dimer value greater than 250 ng/mL was considered positive for thromboembolic dis- ease. Diagnosis of acute pulmonary embolism was based solely on the interpretation of the CT angiogram. The D- dimer assay results were then compared with the CT angiographic diagnoses. • Results: Of 946 CT studies, 172 (18%) were positive for acute pulmonary embolism. The D-dimer assay was positive for 612 (65%) of the 946 patients. For acute pul- CI = confidence interval; CT = computed tomography; ELISA = enzyme-linked immunosorbent assay monary embolism, the D-dimer assay had a sensitivity of 0.83 (95% confidence interval [CI], 0.76-0.88), a specificity of 0.39 (95% CI, 0.36-0.43), a negative likelihood ratio of 0.44 (95% CI, 0.32-0.62), and a negative predictive value of 0.91 (95% CI, 0.87-0.94). • Conclusions: The semiquantitative latex agglutina- tion plasma fibrin D-dimer assay had moderate sensitivity and low specificity for the diagnosis of acute pulmonary embolism. When used alone, the results of this test were insufficient to exclude this serious and potentially fatal disorder. Approximately two thirds of our patients had positive D-dimer assays and required further evaluation to exclude acute pulmonary embolism. Mayo Clin Proc. 2004;79:164-168 A n estimated 200,000 patients in the United States have acute pulmonary embolism annually, and it is the primary cause of death for about 60,000 of these patients. 1,2 Early diagnosis and prompt initiation of anticoagulation medication markedly reduce this mortality rate. 3 However, diagnosing acute pulmonary embolism is often difficult. A recent 5-year autopsy study from our institution found that the cause of death in 4% of patients was acute pulmonary embolism. This diagnosis was considered antemortem in only half of these patients, and testing for thromboembolic disease was performed in only 22%. 4 Acute pulmonary embolism is probably the most common preventable cause of hospital deaths. 5 Historically, the most common test for the diagnosis of acute pulmonary embolism has been the radionuclide (ven- tilation-perfusion) lung scan. 6,7 However, only 28% of pa- tients in the Prospective Investigation of Pulmonary Embo- lism Diagnosis (PIOPED) study had diagnostic lung scans (normal, near-normal, or high-probability). 8 Most patients require further diagnostic evaluation. Pulmonary angiography is the diagnostic reference stan- dard, but it is invasive and associated with some morbidity and mortality. 9-11 Also, pulmonary angiography has limita- tions, including diagnostic accuracy (particularly for small peripheral emboli), interreader agreement, and incomplete studies. 5 An alternative approach for patients with nondiag- nostic lung scans is noninvasive testing for deep venous thrombosis (by either impedance plethysmography or venous compression ultrasonography), which may be per- formed serially if the initial test result is negative. 12 This second strategy often involves having patients return to the hospital or clinic after dismissal for further testing, but this has drawbacks. The problem of nondiagnostic lung scans has led to the development of 2 new diagnostic tests: the plasma fibrin D-dimer assay and computed tomographic (CT) angiography. 5,13 For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.