Pulmonary Embolism (PE) Severity Index predicts low risk PE patients.

Clinical Question

Can prognostic models accurately predict short-term mortality in patients with PE?

Bottom Line

For prediction of 30 day mortality, a PE Severity Index (PESI) score of < 3 had a sensitivity and negative predictive value of 100%. (LOE = 2b)

Reference

Jimenez D, Yusen RD, Otero R, et al. Prognostic models for selecting patients with acute pulmonary embolism for initial outpatient therapy. Chest 2007;132:24-30.  [PMID:17625081]

Study Design

Cohort (prospective)

Funding

Unknown/not stated

Setting

Inpatient (any location) with outpatient follow-up

Synopsis

The PE Severity Index (PESI) and Geneva Score are two validated models for predicting outcomes in patients with PE. The Geneva score, based on 6 predictors, classifies patients as either low or high risk, while the PESI uses 11 factors to group patients into 5 risk classes. This study assessed the ability of the Geneva score and PESI to predict 30-day mortality, nonfatal recurrent venous thromboembolism, and nonfatal major bleeding in patients with PE. Performance of models was compared with the original study derivation samples and the ability of the models to identify low-risk patients appropriate for outpatient treatment was assessed. Baseline data was prospectively collected at the time of diagnosis and outcome data was determined by patient interviews and chart reviews. Consecutive ambulatory patients diagnosed with PE in the emergency department of a tertiary care university hospital in Spain were eligible. Patients were hospitalized and treated with intravenous unfractionated heparin or subcutaneous low-molecular-weight heparin until a therapeutic INR was achieved with warfarin. The study enrolled 599 patients with objectively confirmed PE. The Geneva risk score classified significantly more patients as low risk than the PESI (84% vs. 36%, p < 0.0001), and mortality at 30 days was significantly higher among patients identified as low risk by Geneva score compared to PESI (5.6% vs. 0.9%, p < 0.001). Thirty day mortality rates among Geneva and PESI high risk patients were 15.5% and 10.;7%, respectively. While this data is useful in identifying low risk PE patients, a randomized trial evaluating outpatient management of low risk patients is needed to adequately assess the safety of this approach.