High Variability in the Duration of Chest Compression Interruption is Associated With Poor Outcomes in Pediatric Extracorporeal Cardiopulmonary Resuscitation.

ABSTRACT

OBJECTIVES: To determine the association between chest compression interruption (CCI) patterns and outcomes in pediatric patients undergoing extracorporeal cardiopulmonary resuscitation (ECPR). DESIGN: Cardiopulmonary resuscitation (CPR) data were collected using defibrillator-electrode and bedside monitor waveforms from pediatric ECPR cases between 2013 and 2021. Duration and variability of CCI during cannulation for ECPR was determined and compared with survival to discharge using Fishers exact test and logistic regressions with cluster-robust se s for adjusted analyses. SETTING: Quaternary care children's hospital. PATIENTS Pediatric patients undergoing ECPR. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Of 41 ECPR events, median age was 0.7 years (Q1, Q3 0.1, 5.4), 37% (15/41) survived to hospital discharge with 73% (11/15) of survivors having a favorable neurologic outcome. Median duration of CPR from start of ECPR cannulation procedure to initiation of extracorporeal membrane oxygenation (ECMO) flow was 21 minutes (18, 30). Median duration of no-flow times associated with CCI during ECMO cannulation was 11 seconds (5, 28). Following planned adjustment for known confounders, survival to discharge was inversely associated with maximum duration of CCI (odds ratio [OR] 0.91 [0.86-0.95], p = 0.04) as well as the variability in the CCI duration (OR 0.96 [0.93-0.99], p = 0.04). Cases with both above-average CCI duration and higher CCI variability ( sd > 30 s) were associated with lowest survival (12% vs. 54%, p = 0.009). Interaction modeling suggests that lower variability in CCI is associated with improved survival, especially in cases where average CCI durations are higher. CONCLUSIONS: Shorter duration of CCI and lower variability in CCI during cannulation for ECPR were associated with survival following refractory pediatric cardiac arrest.