Excluding Pulmonary Embolism with End-tidal Carbon Dioxide: Accuracy, Cost, and Harm Avoidance

Abstract

A non-randomized single center prospective, descriptive, correlational design was used to determine what end-tidal carbon dioxide (EtCO2) level provided the best sensitivity, specificity, and negative predictive value to exclude pulmonary embolism (PE) diagnosis in hemodynamically stable hospitalized adults ( n = 111). The financial impact and harm avoidance of adding EtCO2 to the PE diagnostic process also were examined. PE diagnosis was determined by computed tomography pulmonary angiography (CTPA). PE prevalence was 18.9%. Mean±SD EtCO2 was lower for PE positive than negative participants (28 ± 7.8 to 33 ± 8.1 mmHg respectively 95% CI: 1.22–8.96; P = .01). For PE exclusion, an EtCO2 cutoff ≥42 mmHg yielded 100% sensitivity, 12.2% specificity, and 100% negative predictive value. For every six inpatients assessed with EtCO2, one could be saved from unnecessary CTPA. Eliminating unnecessary CTPA removes the potential harm associated with radiation and intravenous contrast exposure. Additionally, an EtCO2 cutoff ≥42 mmHg could eliminate ~$88,000/year in healthcare waste at this institution.