Pa o2 and Mortality in Neonatal Extracorporeal Membrane Oxygenation: Retrospective Analysis of the Extracorporeal Life Support Organization Registry, 2015-2020.

OBJECTIVES: Extracorporeal life support can lead to rapid reversal of hypoxemia but the benefits and harms of different oxygenation targets in severely ill patients are unclear. Our primary objective was to investigate the association between the Pa o2 after extracorporeal membrane oxygenation (ECMO) initiation and mortality in neonates treated for respiratory failure. DESIGN: Retrospective analysis of the Extracorporeal Life Support Organization (ELSO) Registry data, 2015-2020. PATIENTS: Newborns supported by ECMO for respiratory indication were included. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Pa o2 24 hours after ECMO initiation (H24 Pa o2 ) was reported. The primary outcome was 28-day mortality. We identified 3533 newborns (median age 1 d [interquartile range (IQR), 1-3]; median weight 3.2 kg [IQR, 2.8-3.6]) from 198 ELSO centers, who were placed on ECMO. By 28 days of life, 731 (20.7%) had died. The median H24 Pa o2 was 85 mm Hg (IQR, 60-142). We found that both hypoxia (Pa o2 < 60 mm Hg) and moderate hyperoxia (Pa o2 201-300 mm Hg) were associated with greater adjusted odds ratio (aOR [95% CI]) of 28-day mortality, respectively: aOR 1.44 (95% CI, 1.08-1.93), p = 0.016, and aOR 1.49 (95% CI, 1.01-2.19), p value equals to 0.045. CONCLUSIONS: Early hypoxia or moderate hyperoxia after ECMO initiation are each associated with greater odds of 28-day mortality among neonates requiring ECMO for respiratory failure.

Evolution of Carboxyhemoglobin in Children Supported by Extracorporeal Membrane Oxygenation: An Observational Single-Center Study.

Carboxyhemoglobin (COHb) is potentially a novel marker of hemolysis on extracorporeal membrane oxygenation (ECMO) and may be useful as an indicator for circuit-related complication in adults, but little is known about COHb levels in children. An observational single-center study was performed between January 2018 and December 2021. Fifty-eight children were included and COHb levels were obtained along with routine blood gas analysis before, during, and after ECMO support. From the 6th hour of ECMO support, the COHb level increased relative to the pre-ECMO level, with an adjusted mean difference of 0.44 (95% confidence interval [CI], 0.26-0.62; p < 0.001) and remained higher during ECMO run and within 6 hours after weaning ( p < 0.001). Among the 18 children (31%) who experienced at least one circuit-related complication leading to a circuit change, we observed a significant decrease in COHb levels within 24 hours after the circuit change, compared with the 24 hours before (adjusted mean difference, 0.54%; 95% CI, 0.27-0.80; p < 0.001). The maximal daily COHb level was able to predict circuit-related complications within 24 hours following COHb measurement with an area under the receiver operating characteristic (ROC) curve of 0.85 (95% CI, 0.77-0.92; p < 0.001).