- Design
- Retrospective independent assessment of prospective registry data, six academic emergency departments
- Population
- 1,731 patients with pulmonary embolism
- Primary outcome
- In-hospital death or clinical deterioration
- Effect
- CPES 3 or higher: sensitivity 72%, specificity 57%, AUC 0.68; advanced intervention AUC 0.78
This retrospective assessment used prospectively collected registry data from six academic emergency departments (2016 to 2020), including 1,731 patients with pulmonary embolism. The Composite Pulmonary Embolism Shock score was compared with the simplified PESI and European Society of Cardiology classes.
The primary outcome of in-hospital death or deterioration occurred in 193 (11.1%). A score of 3 or higher had sensitivity 72% and specificity 57% (AUC 0.68), with likelihood ratios similar to the other two tools. For advanced intervention (7.1%), the score separated better (AUC 0.78; positive LR 2.07, negative LR 0.20). Thirty-day mortality (7.2%) performance was similar across tools. Central thrombus was independently associated with advanced intervention (OR 3.46, 95% CI 1.49 to 8.93).
The score did not clearly beat established tools for the main outcome. Its role, if any, is in anticipating the need for advanced intervention, and that needs prospective testing.
- For death or deterioration, the composite score was no better than sPESI or ESC classes.
- A low score (negative LR 0.20) argued against needing advanced intervention.
- Look for central thrombus; it was linked to later advanced intervention.
- Do not change your risk tool on this study alone.
Why it matters
Choosing which patients with PE need monitoring or intervention is a daily emergency decision with imperfect tools.
Don't overread it
A retrospective analysis from academic centres may not apply to every emergency department.
The statistics, in plain English
An AUC of 0.68 means the score ranks a random patient who deteriorated above one who did not about two thirds of the time, which is modest. The AUC of 0.78 for advanced intervention is better but derives from a small number of events.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for emergency & critical care, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free