Impact of Timing of ECMO Initiation on Outcomes After Pediatric Heart Surgery: A Multi-Institutional Analysis

Little is known about the relationship of timing of extracorporeal membrane oxygenation (ECMO) initiation on patient outcomes after pediatric heart surgery. We hypothesized that increasing timing of ECMO initiation after heart surgery will be associated with worsening study outcomes. Patients aged ≤18 years receiving ECMO after pediatric cardiac surgery at a Pediatric Health Information System-participating hospital (2004–2013) were included. Outcomes evaluated included in-hospital mortality, composite poor outcome, prolonged length of ECMO, prolonged length of mechanical ventilation, prolonged length of ICU stay, and prolonged length of hospital stay. Multivariable logistic regression models were fitted to study the probability of study outcomes as a function of timing from cardiac surgery to ECMO initiation. A total of 2908 patients from 42 hospitals qualified for inclusion. The median timing of ECMO initiation after cardiac surgery was 0 days (IQR 0–1 day; range 0–294 days). After adjusting for patient and center characteristics, increasing duration of time from surgery to ECMO initiation was not associated with higher mortality or worsening composite poor outcome. However, increasing duration of time from surgery to ECMO initiation was associated with prolonged length of ECMO, prolonged length of ventilation, prolonged length of ICU stay, and prolonged length of hospital stay. Although this relationship was statistically significant, the odds for prolonged length of ECMO, prolonged length of ventilation, prolonged length of ICU stay, and prolonged length of hospital stay increased by only 1–3 % for every 1-day increase in ECMO that may be clinically insignificant. We did not demonstrate any relationship between timing of ECMO initiation and mortality among the patients of varying age groups, and patients undergoing cardiac surgery of varying complexity. We concluded that increasing duration of time from surgery to ECMO initiation is not associated with worsening mortality. Our results suggest that ECMO is initiated at the appropriate time when dictated by clinical situation among patients of all age groups, and among patients undergoing heart operations of varying complexity.