Validation of a clinical prognostic model to identify low-risk patients with pulmonary embolism

Abstract

Objective. To validate the Pulmonary Embolism Severity Index (PESI), a clinical prognostic model which identifies low-risk patients with pulmonary embolism (PE). Design. Validation study using prospectively collected data. Setting. A total of 119 European hospitals. Subjects. A total of 899 patients diagnosed with PE. Intervention. The PESI uses 11 clinical factors to stratify patients with PE into five classes (I-V) of increasing risk of mortality. We calculated the PESI risk class for each patient and the proportion of patients classified as low-risk (classes I and II). The outcomes were overall and PE-specific mortality for low-risk patients at 3 months after presentation. We calculated the sensitivity, specificity and predictive values to predict overall and PE-specific mortality and the discriminatory power using the area under the receiver operating characteristic curve. Results. Overall and PE-specific mortality was 6.5% (58/899) and 2.3% (21/899) respectively. Forty-seven per cent of patients (426/899) were classified as low-risk. Low-risk patients had an overall mortality of only 1.2% (5/426) and a PE-specific mortality of 0.7% (3/426). The sensitivity was 91 [95% confidence interval (CI): 81-97%] and the negative predictive value was 99% (95% CI: 97-100%) for overall mortality. The sensitivity was 86% (95% CI: 64-97%) and the negative predictive value was 99% (95% CI: 98-100%) for PE-specific mortality. The areas under the receiver operating characteristic curve for overall and PE-specific mortality were 0.80 (95% CI: 0.75-0.86) and 0.77 (95% CI: 0.68-0.86) respectively. Conclusions. This validation study confirms that the PESI reliably identifies low-risk patients with PE who are potential candidates for less costly outpatient treatment. © 2007 Blackwell Publishing Ltd.