Introduction and Objectives Prognostic accuracy and clinical utility of the Pulmonary Embolism Severity Index (PESI) & simplified PESI (sPESI) at predicting very early mortality (<7 days) has not been previously investigated. We use this time frame to suggest outpatient imaging could be used to confirm a suspected pulmonary embolism (PE).

Methods A retrospective chart review of adult patients (≥18 years), who presented to two emergency departments, January 2013 – December 2015, with symptomatic pulmonary embolism (PE) confirmed on computed tomography pulmonary angiogram was carried out. Demographic and clinical parameters were recorded. PESI and sPESI scores were calculated and grouped into classes. Patients were followed up to 90 days. The primary outcome was death at 1, 3, 7, 30 and 90 days.

Results Two hundred and eighty eight patients were eligible for inclusion. Mean age was 63 years (SD 18) and 51% were male. Twenty-two patients died during the follow-up period. PE was attributed to the cause of death in 59%. There was a significant correlation between PESI risk classes (I to V) and death at 3, 7, 30 and 90 days (p<0.01) but not day 1. No deaths occurred in the low risk sPESI class and no deaths occurred in PESI Class III. The discriminatory ability for each study end-point expressed as the area under the ROC curve was high (AUROC ≥0.8, p<0.001) for both PESI and sPESI.

Conclusion This findings of this study supports the use of PESI scoring systems in predicting early mortality. Study data suggests that sPESI has non-inferior predictive properties compared to the PESI, and therefore may prove of higher utility in day-to-day clinical practice. These tools may be reliably used to consider outpatient management of patients with PE, which includes imaging up to 72 hours after A&E attendance.