To estimate the percentage of patients with suspected acute pulmonary embolism in whom a noninvasive diagnosis or exclusion of thromboembolic disease might be safely made on the basis of ventilation-perfusion (VQ) lung scans, single noninvasive tests of the lower extremities, and, in patients with adequate cardiorespiratory reserve, serial noninvasive tests of the lower extremities.
Calculations were made among 662 patients who participated in the collaborative study Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) and who had blood gas values measured while breathing room air and who underwent pulmonary angiography. The diagnostic strategy recommends treatment in all patients with a high-probability VQ scan and no treatment in patients with nearly normal VQ scans. In patients with nondiagnostic VQ scans (intermediate- or low-probability scans), a single noninvasive leg test is recommended. It was assumed that 50% of patients with pulmonary embolism would show deep venous thrombosis with a single noninvasive leg test. If results are abnormal, treatment is indicated. If normal, serial noninvasive leg tests are recommended. Treatment can be withheld if results of serial tests are normal. In patients with poor cardiorespiratory reserve, pulmonary angiography is indicated.
A single noninvasive leg test in patients with nondiagnostic VQ scans would show deep venous thrombosis and, therefore, eliminate the need for pulmonary angiography in 53 (11%) of 468 patients (95% confidence interval [CI], 9% to 15%) who otherwise would require angiography. Serial noninvasive leg tests in patients with adequate cardiorespiratory reserve who had a normal result of a single leg test would either show deep venous thrombosis or exclude it in 222 (47%) of 468 patients (95% CI, 43% to 52%). The need for pulmonary angiography, therefore, would be reduced from 468 (71%) of 662 (95% CI, 67% to 74%) if no noninvasive leg tests were performed to 415 (63%) of 662 (95% CI, 59% to 66%) if only a single noninvasive leg test were performed, and further reduced to 193 (29%) of 662 (95% CI, 26% to 33%) if serial noninvasive leg tests were used where appropriate.
A noninvasive strategy that includes VQ scans, single noninvasive leg tests, and serial noninvasive leg tests would permit a diagnosis of thromboembolic disease or a safe exclusion of thromboembolic disease in 71% of patients with suspected acute pulmonary embolism.(Arch Intern Med. 1995;155:2101-2104)