Objective: To evaluate the utility of a modified calculation of the alveolar dead space fraction (Vd/Vt), combined with plasma D-dimers, to aid in the exclusion of acute pulmonary embolism (PE).
Methods: A prospective comparison of screening modalities was performed in a metropolitan teaching ED. Ambulatory patients evaluated for PE underwent simultaneous end-tidal CO2 and arterial blood gas determinations, as well as venous latex-agglutination D-dimer quantification. The modified Bohr equation was used to calculate Vd/Vt as an index of alveolar dead space. Acute PE was diagnosed or excluded using appropriate combinations of clinical suspicion, ventilation-perfusion lung scanning, lower-extremity venous Doppler ultrasonography, pulmonary angiography, and comprehensive follow-up.
Results: Of 170 subjects studied, PE was confirmed (PE+) in 26 (15%) and excluded (PE-) in 144 (85%). In the PE+ group, Vd/Vt was 0.31 +/- 0.13 (mean +/- SD), and in the PE- group, Vd/Vt was 0.06 +/- 0.10 (p < 0.05, t-test). Regarding false-negative rates, Vd/Vt was normal (i.e., < 0.2) in 3/26 PE+ patients and D-dimer concentrations were normal (< 0.5 microgram/L) in 4/26 patients in the PE+ group. The combination of a normal Vd/Vt and D-dimer concentration was 100% sensitive (95% CI = 88-100%) in excluding PE. False-positive testing (either test positive) occurred in 49/144 subjects (specificity 65%, 95% CI = 52-73%). The age-adjusted alveolar-arterial O2 gradient was 33 +/- 38 torr in the PE+ group vs 13 +/- 37 torr in the PE- group (p = 0.11).
Conclusions: In ambulatory patients, the finding of Vd/Vt < 0.2 and D-dimers < 0.5 microgram/L lowers the probability of acute PE.