Prognostic models for selecting patients with acute pulmonary embolism for initial outpatient therapy

Objective: To assess the performance of two prognostic models in predicting short-term mortality in patients with pulmonary embolism (PE). Subjects and methods: We compared the test characteristics of two prognostic models for predicting 30-day outcomes (mortality, thromboembolic recurrences, and major bleeding) in a cohort of 599 patients with objectively confirmed PE. Patients were stratified into the PE severity index (PESI) risk classes T-V and the Geneva low-risk and high-risk strata. We compared the discriminatory power of both prognostic models. Results: The PESI classified fewer patients as low risk (strata I and II) [36%; 216 of 599 patients; 95% confidence interval (CI), 32 to 40%] compared to the Geneva prediction rule (84%; 502 of 599 patients; 95% CI, 81 to 87%) [p < 0.0001]. Using either prediction rule, the low-risk groups showed statistically relevant 30-day mortality difference (PESI, 0.9%; 95% CI, 0.3 to 2.2; vs Geneva, 5.6%; 95% CI, 3.6 to 7.6) [p < 0.0001], although nonfatal recurrent venous thromboembolisin or major bleeding rates were statistically similar (PESI, 2.8%; 95%.CI, 0.6 to 5.0%; vs Geneva, 4.2%; 95% CI, 2.4 to 5.9%). The area under the receiver operating characteristic curve was higher for the PESI (0.76; 95% CI, 0.69 to 0.83) than for the Geneva score (0.61; 95% CI, 0.51 to 0.71) [p = 0.002]. Conclusions: The PESI quantified the prognosis of patients with PE better than the Geneva score. This study demonstrated that PESI can select patients with very low adverse event rates during the initial days of acute PE therapy and assist in selecting patients for treatment in the outpatient setting.