Association between the Pulmonary Embolism Severity Index (PESI) and short-term clinical deterioration.

The Pulmonary Embolism Severity Index (PESI) has been shown to predict 30 and 90 day mortality after PE. However, whether the PESI predicts patients who will be free of clinically adverse outcomes during a typical hospitalisation is not known. Retrospective analysis of Emergency Department patients with PE from May 2006 to April 2008. We compiled demographics, data to calculate the PESI and a composite outcome. Patients were considered to have a negative PESI if they were in category I or II (≤85 points). Patients were considered to have the composite outcome if, within five days of diagnosis, they: A) had recurrent PE; B) developed a new cardiac dysrhythmia; C) required advanced cardiac life support; D) required respiratory support; E) required vasopressors; F) received thrombolysis; G) had major bleeding; H) returned to the ED; I) died. We enrolled 245 patients with PE. Of these, 115 (47%) were male, 204 (83%) were white. The mean age was 57 ± 17 years. The PESI identified 109 (44%) as low risk and 136 (56%) as high risk. Sixty-one (26%) patients had the outcome, of whom nine (14%) were characterised as low risk by the PESI. Test characteristics were: sensitivity 86% (95% confidence interval [CI]: 75%-93%), specificity 55% (95% CI: 47%-62%), NPV 63% (95% CI: 55%-70%), PPV 40% (95% CI: 31%-49%), LR(+) 1.9 (95% CI: 1.57-2.30) and LR(-) 0.26 (95% CI: 0.14-0.48). Of the patients who had an adverse clinical event or required a hospital-based intervention within the first five days after PE diagnosis, 14% were categorised by the PESI as safe for discharge [corrected] .