Predictors of Catheter-Directed Therapy for Acute Pulmonary Embolism in a Large Multisite Emergency Department Cohort

Pulmonary embolism (PE) is a major cause of mortality in the U.S. While anticoagulation is standard therapy, selected patients with submassive or massive PE may undergo catheter-directed therapy (CDT); rates of occurrence and predictors of CDT receipt in routine emergency care are incompletely described.

We performed a retrospective cohort study using electronic health record (EHR) data from a large, multi-state consortium of hospitals and emergency departments from 2016-2022. Adults with acute PE presenting to the ED, admitted, and initiated on heparin were included. CDT was identified via an EHR algorithm capturing catheter-based pulmonary interventions and/or catheter-directed thrombolytic administration using procedure documentation, orders, and medication administration records (excluding systemic thrombolysis). Multivariable logistic regression evaluated predictors of CDT receipt; discrimination was summarized with the AUC.

Among 28,266 PE patients initiated on heparin, 5.1% received CDT. The model demonstrated good discrimination (AUC 0.75). Clinical presentation features associated with higher odds of CDT included syncope (OR 2.17, 95% CI 1.65–2.82) and higher initial heart rate (OR 1.32 per standard deviation change (SD), 95% CI 1.25–1.40). Oxygenation measures were inversely associated with CDT (minimum oxygen saturation OR 0.74 per SD, 95% CI 0.67–0.81; initial oxygen saturation OR 0.94 per SD, 95% CI 0.89–0.99). Lower minimum systolic and diastolic pressures were inversely associated with CDT (systolic OR 0.56 per SD, 95% CI 0.50–0.62; diastolic OR 0.62 per SD, 95% CI 0.55–0.69), while higher initial systolic pressure was weakly positively associated (OR 1.06 per SD, 95% CI 1.01–1.12), consistent with preferential selection of more hemodynamically stable patients for CDT. Intubation was negatively associated with CDT receipt (OR 0.53, 95% CI 0.33–0.81). Race and payer category showed variation in CDT receipt and was interpreted cautiously as exploratory.

In a large, multi-site ED cohort of adults with acute PE treated with heparin, CDT was uncommon and more likely to occur in patients without profound hemodynamic or respiratory failure. Future work should clarify whether residual variation reflects clinical selection or system-level factors.