Emergency department non-invasive cardiac output study (EDNICO): An accuracy study

Abstract

Background: There is little published data investigating non-invasive cardiac output monitoring in the emergency department (ED). We assess here the accuracy of five non-invasive methods in detecting fluid responsiveness in the ED: (1) common carotid artery blood flow, (2) suprasternal aortic Doppler, (3) bioreactance, (4) plethysmography with digital vascular unloading method, and (5) inferior vena cava collapsibility index. Left ventricular outflow tract echocardiography derived velocity time integral is the reference standard. This follows an assessment of feasibility and repeatability of these methods in the same cohort of ED patients. Methods: This is a prospective observational study of non-invasive methods for assessing fluid responsiveness in the ED. Participants were non-ventilated ED adult patients requiring intravenous fluid resuscitation. Sensitivity and specificity of each method in determining the fluid responsiveness status of participants is determined in comparison to the reference standard. Results: Thirty-three patient data sets were included for analysis. The specificity and sensitivity to detect fluid responders was 46.2 and 45% for common carotid artery blood flow (CCABF), 61.5 and 63.2% for suprasternal artery Doppler (SSAD), 46.2 and 50% for bioreactance, 50 and 41.2% for plethysmography vascular unloading technique (PVUT), and 63.6 and 47.4% for inferior vena cava collapsibility index (IVCCI), respectively. Analysis of agreement with Cohen's Kappa-0.08 for CCABF, 0.24 for SSAD,-0.04 for bioreactance,-0.08 for PVUT, and 0.1 for IVCCI. Conclusion: In this study, non-invasive methods were not found to reliably identify fluid responders. Non-invasive methods of identifying fluid responders are likely to play a key role in improving patient outcome in the ED in fluid depleted states such as sepsis. These results have implications for future studies assessing the accuracy of such methods.