Ventilation-perfusion scanning and pulmonary angiography: correlation in clinical high-probability pulmonary embolism.

During a 3-year period, 173 clinically selected patients underwent pulmonary angiography to confirm or exclude acute pulmonary embolism. All patients had undergone ventilation-perfusion (V/Q) scanning (167 patients) or perfusion scanning alone (six) before angiography. Angiography was done because the results of the V/Q scanning did not satisfy the clinician's needs for certainty. The results of the V/Q and studies were compared to determine the relative accuracy of V/Q scanning in this clinical setting. Pulmonary embolism was found in seven (15%) of 47 patients with low-probability scans, 11 (32%) of 34 patients with intermediate-probability scans, 22 (39%) of 57 patients with indeterminate scans, and 23 (66%) of 35 patients with high-probability scans. In this clinically selected population, low-probability scans were more accurate in excluding pulmonary embolism than were high-probability scans in establishing that diagnosis.