The dead space to tidal volume ratio in the diagnosis of pulmonary embolism

In order to assess the value of the measurement of the physiologic dead space (VD) to tidal volume (VT) ratio in pulmonary embolism (PE), a prospective study was performed in hospital inpatients suspected to have PE (n = 110; mean age +/- SD, 52.2 +/- 15.5 yr). In 16 of 29 patients in whom the diagnosis of PE was excluded on the basis of a normal radioisotope perfusion scan and/or normal pulmonary angiogram, VD/VT was less than 40%; in the other 13 patients, a VD/VT greater than 40% was associated with an abnormal spirogram. In all patients in whom PE was angiographically diagnosed (n = 16), VD/VT was greater than 40%. In the remaining 65 patients, a high probability of PE was rarely (6%) associated with a normal VD/VT, whereas in patients with a low probability of PE, 71% had normal VD/VT values. These data indicate that a VD/VT value of less than 40% makes the diagnosis of PE extremely unlikely, whereas VD/VT value greater than 40% in the presence of a normal spirogram is highly suggestive of PE. The diagnostic sensitivity of a VD/VT greater than 0.4 with a normal spirogram as a positive test of PE, and a VD/VT less than 0.4 excluding the diagnosis of PE was 100%, whereas the specificity was 94%; applying Bayesian analysis, the probability of a correct diagnosis of PE using these criteria in a similar population would be 90.5%, and of excluding PE, 96.7%. Thus, as a diagnostic test in PE, VD/VT measurement is comparable, in terms of sensitivity and specificity, to radioisotope lung scanning.