Prediction of short-term prognosis in elderly patients with acute pulmonary embolism: validation of the RIETE score.

Essentials The RIETE score was derived to predict 10-day adverse outcomes in acute pulmonary embolism (PE). We externally validated the RIETE score in a prospective cohort of patients with PE. The RIETE score classified fewer patients as low-risk than currently recommended scores. The RIETE score was not superior to other scores in predicting 10-day adverse outcomes.

SUMMARY: Introduction The Registro Informatizado de la Enfermedad TromboEmbolica (RIETE) score was derived to identify patients with pulmonary embolism (PE) at low risk of overall complications. Objective To externally validate the RIETE score and compare its prognostic performance with the Pulmonary Embolism Severity Index (PESI), its simplified version (sPESI) and the Geneva Prognostic Score (GPS). Methods In a prospective multicenter cohort, we studied 687 elderly patients with acute PE. The primary outcome was 10-day overall complications (death, recurrent PE or major bleeding); the secondary outcome was 30-day overall mortality. We compared complications and mortality in low-risk vs. higher-risk patients and the area under the receiver operating characteristic (ROC) curve across scores. Results Overall, 27 patients (3.9%) had complications within 10 days and 22 (3.2%) died within 30 days. The RIETE score classified a smaller proportion of patients as low risk (31%) than the PESI (35%), sPESI (36%) and the GPS (90%). The proportion of low-risk patients based on the RIETE score, PESI, sPESI and GPS who had complications was 1.9%, 1.7%, 1.6% and 2.9%, respectively. The RIETE score had a lower area under the ROC curve (0.60) for predicting complications than the PESI (0.67), sPESI (0.65) and GPS (0.72). The area under the ROC curve for predicting mortality was similar (0.76-0.78) for all scores. Conclusion The RIETE score classified fewer patients as low risk than the other scores. It accurately identified patients at low risk of mortality but was not superior to other scores in predicting 10-day overall complications.

TRIAL REGISTRATION: https://clinicaltrials.gov. Identifier: NCT00973596.