The Clinical Features of Acute Pulmonary Embolism in Ambulatory Patients

OBJECTIVE: To identify clinical findings and predisposing conditions associated with acute pulmonary embolism (PE) in ambulatory patients being evaluated for PE. METHODS: A prospective observational study was conducted. A standardized multicomponent data collection form was administered to ambulatory subjects being evaluated for PE. The diagnosis of PE was confirmed or excluded using a combination of scintillation lung scanning, lower-extremity venous Doppler ultrasonography, and selective use of pulmonary angiography. RESULTS: Data collection was completed for 170 subjects, with 26 (15%) cases of PE. Subjects with PE were significantly older (56 vs 41 years, 99% CI for difference of 15 years [6 to 25 years]), were more likely to report unexplained dyspnea (92% vs 69%, 99% CI for difference of 23% [7% to 40%]), and waited longer after symptom onset to seek medical evaluation (73 vs 36 hours, 99% CI for difference of 37 hours [11 to 63 hours]). No difference was found for multiple variables commonly associated with PE. Assignment to risk categories was of limited diagnostic utility. For example, low-risk assignment yielded 85% sensitivity, 20% specificity; high-risk assignment: 31% sensitivity, 85% specificity, with diagnostic accuracy below 80% in both categories. CONCLUSIONS: Among outpatients selected for evaluation for PE, further risk stratification demonstrated poor diagnostic utility. Clinical features alone cannot be used to differentiate presence or absence of PE in at-risk ambulatory patients.