Validación externa de puntajes pronósticos para mortalidad hospitalaria y a treinta días en pacientes con tromboembolismo pulmonar en Argentina

Prognostic models have been developed to help make decisions in the treatment of pulmonary embolism (PE). Among them, the Pulmonary Embolism Severity Index (PESI) and simplified PESI (sPESI), however they have not been validated in our setting. The objective was to evaluate PESI and sPESI scores ability to predict in-hospital mortality in patients with PE in Argentina. We analyzed a database of 75 academic centers in Argentina that included consecutive patients with PE from 2016 to 2017. The scores were prospectively calculated, and in-hospital and 30 days mortality were assessed. The validation of the models was assessed through discrimination using the area under the ROC curve (AUC), and calibration with the Hosmer-Lemeshow (HL) test. The cohort included 684 patients. In-hospital mortality was 12% and at 30 days an additional 3.2% mortality was registered. The AUC (95% CI) for in-hospital mortality was 0.75 (0.69-0.81) for PESI and 0.77 (0.71-0.82) for sPESI (p = 0.2 between scores). AUC of 30-day mortality 0.75 (0.68-0.8) and 0.78 (0.74-0.83) for PESI and sPESI (p = 0.2 between scores). Both models presented good calibration. The PESI and sPESI risk scores demonstrated similar performance and good accuracy in predicting hospital and 30-day mortality. Both scores can be established as simple prediction tools for PE patients in Argentina.