Accuracy of D-dimer/fibrinogen ratio to predict pulmonary embolism: a prospective diagnostic study.

D-Dimer and fibrinogen are elevated in many diseases presenting signs and symptoms similar to those seen in patients with pulmonary embolism (PE). We tested the hypothesis that patients with PE have lower fibrinogen and higher d-dimer values than patients in whom the diagnosis is suspected but safely excluded. One hundred and ninety-one consecutive patients with suspected acute PE (85 positive, 106 negative) were investigated with a diagnostic strategy including d-dimer, pretest probability, and helical computed tomography as first-line tests. In 38 of 40 patients with suspected PE and d-dimer <500 microg L(-1), PE was excluded without further testing. During a 3-month follow-up, there was no clinical PE among these 38 and the 68 patients with a negative helical CT. In 151 patients with d-dimer >500 microg L(-1), d-dimer, fibrinogen, and d-dimer/fibrinogen ratio (D/F ratio) were different in PE-positive compared with PE-negative patients [medians (and ranges) for d-dimer: 3793 (780 - 42 195) vs. 992 (621-6957) microg L(-1), fibrinogen: 3.8 (0.4-6.2) vs. 4.7 (2.2-8.4) g L(-1), and D/F ratio: 1.22 (0.15-85.45) 103 vs. 0.25 (0.09-1.03) x 103; P < 0.0001, respectively). The true positive rate was almost twice as high using D/F ratio >1.04 x 103 (49 of 85 patients; 57.6%) compared with d-dimer >7000 micro g L(-1) (25 of 85 patients; 29.4%). Patients with acute PE have lower fibrinogen values than patients with suspected but excluded PE. D/F ratio >103 is highly specific for the presence of acute PE, and causes a doubling of the diagnostic rate compared with d-dimer testing alone.