Assessment of echocardiographic interpretation of dual-lumen cannula during venovenous extracorporeal membrane oxygenation use for pediatric respiratory failure

Purpose: Echocardiography is considered essential during cannulation placement and manipulations. Literature evaluating transthoracic echocardiography (TTE) usage during pediatric VV-ECMO is scant. The purpose of this study is to describe the use of echocardiography during VV-ECMO at a large, quaternary children's hospital. Methods: A retrospective, single-year cohort study was performed of pediatric patients on VV-ECMO via dual-lumen cannula at our institution from January 2019 through December 2019. For each echocardiogram, final cannula component (re-infusion port (ReP), distal tip, proximal port and distal port) positions were evaluated by one echocardiographer. For TTEs with ReP in the right atrium, two echocardiographers independently evaluated ReP direction using 2-point (Yes/No) and 4-point scales, which were semi-quantitative protocols using color Doppler images to estimate ReP jet direction to the tricuspid valve. Cohen's kappa or weighted kappa was used to measure interrater agreement. Results: During study period, 11 patients (64% male) received VV-ECMO with 49 TTEs and one transesophageal echocardiogram performed. The median patient age was 4.3 years [IQR: 1.1–11.5] and median VV-ECMO run time of 192 h [90–349]. The median time between TTEs on VV-ECMO was 34 h [8.3–65]. Most common position for the ReP was the right atrium (n = 33, 67%), and ReP location was not identified in five TTEs (10%). For ReP flow direction, echocardiographers agreed on 82% of TTEs using 2-point evaluation. There was only moderate agreement between echocardiographers on the 2-point and 4-point assessments (k =.54, kw =.46 respectively). Conclusions: TTE is the predominant cardiac ultrasound modality used during VV-ECMO for pediatric respiratory failure. Subjective evaluation of VV-ECMO ReP jet direction in the right atrium is challenging, regardless of assessment method.