Reliability and reproducibility of the venous excess ultrasound (VExUS) score, a multi-site prospective study: validating a novel ultrasound technique for comprehensive assessment of venous congestion

Abstract

Though the novel venous excess ultrasound (VExUS) score is increasingly used as a noninvasive means of venous congestion measurement, the inter-rater reliability (IRR), inter-user reproducibility (IUR), and utility of concurrent ECG have not been evaluated. We conducted a multicenter study of the IRR, IUR, and utility of ECG for VExUS interpretation between four attending physicians of diverse specialties, reporting the Kappa statistic (KS) and Intraclass Correlation Coefficient (ICC) for IRR and IUR for scans with and without ECG. Eighty-four paired VExUS exams from 42 patients, 60 of which had a concurrent ECG tracing, were interpreted. They showed substantial IRR, with a KS of 0.71 and ICC of 0.83 for the overall VExUS grade (p < 0.001), and IUR, with a KS 0.63 and ICC of 0.8. There was greater agreement among images with an ECG tracing. These results suggest that ECG-augmented VExUS may be a reliable and reproducible measure interpretable by clinicians with diverse backgrounds.

Fig. 1

Inter-rater reliability and concordance of VExUS exam components.Concordance statistics for the cohort. Concordance as measured by intraclass correlation coefficient and Cohen’s kappa statistic was greater for the overall VExUS grade than each of its individual components. It was also greater among images with a concurrent electrocardiogram tracing than for images without one