Abstract
INTRODUCTION: Neonatal hypoxic respiratory failure is commonly assessed with the oxygenation index (OI) to determine severity and guide ECMO initiation. Calculation of the OI requires arterial blood sampling which can be difficult to obtain. A non-invasive alternative, the oxygen saturation index (OSI), has shown promise, but its utility in ECMO determination is not well-described. We aimed to evaluate the correlation between the OI and OSI in neonates requiring ECMO. METHODS: We pursued a retrospective chart review of 64 neonatal ECMO patients at Kentucky Children's Hospital (2012-2022) and analyzed OI and OSI values in the 12 h preceding ECMO initiation. RESULTS: A moderate correlation was observed between the OI and OSI. An OSI >17.41 predicted ECMO initiation, and OI can be estimated with the equation: OI = 1.978(OSI)-6.743. CONCLUSION: These findings suggest OSI may be a useful adjunct to OI for assessing neonatal respiratory failure and could be beneficial when arterial sampling is impractical.