Prediction of in-hospital deterioration in normotensive pulmonary embolism remains elusive: external validation of the calgary acute pulmonary embolism score

Acute pulmonary embolism (PE) is a frequently diagnosed condition. Prediction of in-hospital deterioration is challenging with current risk models. The Calgary Acute Pulmonary Embolism (CAPE) score was recently derived to predict in-hospital adverse PE outcomes but has not yet been externally validated. Retrospective cohort study of normotensive acute pulmonary embolism cases diagnosed in our emergency department between 2017 and 2019. An external validation of the CAPE score was performed in this population for prediction of in-hospital adverse outcomes and a secondary outcome of 30-day all-cause mortality. Performance of the simplified Pulmonary Embolism Severity Index (sPESI) and Bova score was also evaluated. 712 patients met inclusion and exclusion criteria, with 536 patients having a sPESI score of 1 or more. Among this population, the CAPE score had a weak discriminative power to predict in-hospital adverse outcomes, with a calculated c-statistic of 0.57. In this study population, an external validation study found weak discriminative power of the CAPE score to predict in-hospital adverse outcomes among normotensive PE patients. Further efforts are needed to define risk assessment models that can identify normotensive PE patients at risk for in hospital deterioration. Identification of such patients will better guide intensive care utilization and invasive procedural management of PE. Key messagePrognostication of in-hospital adverse outcomes for patients presenting with normotensive acute PE remains challenging.The Calgary Acute Pulmonary Embolism (CAPE) risk score derivation study showed promise for predicting in-hospital adverse PE events. However, this study finds weak performance in an external validation population.Treatment decisions between anticoagulation versus thrombolysis or catheter thrombectomy are challenging in normotensive intermediate risk (sub-massive) PE populations due to unclear prognostics, highlighting the need for improved PE risk models.