Extubation Failure in Neonates Following Congenital Cardiac Surgery: Multicenter Retrospective Cohort, 2017-2020

Objectives: Extubation failure (EF) in neonates recovering from congenital cardiac surgery is associated with morbidity and mortality. Adding continuous physiologic monitoring data and risk analytics algorithms to clinical factors has the potential to assist clinicians in identifying those neonates at high risk for EF. We aimed to evaluate the association of two physiologic risk analytics algorithms evaluating the probability of inadequate delivery of oxygen index (IDo2) and inadequate ventilation of carbon dioxide index (IVco2) with EF in neonates receiving mechanical ventilation (MV) after cardiac surgery. A secondary aim was to evaluate the clinical factors associated with EF. Design: Multicenter retrospective cohort study. Setting: Eight international pediatric cardiac ICUs. Patients: Neonates (age < 1 mo at the time of surgery) receiving MV for longer than 48 hours following cardiac surgery between January 1, 2017, and December 31, 2020. Interventions: None. Measurements and Main Results: Data from 736 neonates were analyzed with 102 (13.9%) having EF (defined as reintubation within 48 hr of extubation). In multivariable analysis (odds ratio [OR] and 95% CI), preoperative respiratory support (OR, 1.72 [95% CI, 1.11-2.67]) was associated with greater odds of EF. In all, 611 neonates had pre-extubation IDo2 data and 478 neonates had both pre-extubation IDo2 and IVco2 data. In multivariable analysis of patients with both pre-extubation IDo2 and IVco2 data, single ventricle anatomy (OR, 2.50 [95% CI, 1.27-4.92]) and high IDo2 (≥ 25) or high IVco2 (≥ 50) in the 2 hours preceding extubation (OR, 1.77 [95% CI, 1.01-3.12]) were associated with greater odds of EF. Conclusions: In this 2017-2020 cohort, EF is high in post-cardiac surgery neonates receiving at least 48 hours of MV. The IDo2 and IVco2 algorithms may be useful in assessing risk of EF in such neonates.