Incidence of excess oxygen use in critically ill children and its impact on in-hospital mortality.

Abstract

Objective: To evaluate excess oxygen incidence and the prognostic value of Cumulative Excess Oxygen Exposure (CEOE), by using non-invasive SpO2, for in-hospital mortality in a Pakistani Pediatric intensive care unit (PICU). Study Design: Prospective observational cohort study. Setting: Department of Pediatrics and Child Health, Liaquat National Hospital, Karachi, Pakistan. Period: May 2024 to April 2025. Methods: Using non-probability consecutive sampling, this study included all children aged 1 month to 15 years admitted in PICU and requiring invasive or non-invasive respiratory support within the first 24 hours of the admission. Children with cyanotic heart diseases and who had a stay of <24hrs were excluded. CEOE was calculated as the mean hourly FiO2 exceeding 0.21, for the hours in which corresponding SpO2 was ≥95%, over the first 24 hours of respiratory support. Primary outcome was CEOE, secondary was in-hospital mortality. Results: Among 197 children for analysis, 98.5% experienced excess oxygen exposure of varying severity. In-hospital mortality was 10.7%. Median CEOE was 29.00 (IQR 19.08-40.87). CEOE of >39.46 showed 71.45 sensitivity and 77.8% specificity. CEOE significantly predicted mortality (AUC 0.754; 95% CI 0.687-0/812; p<0.0001). Higher CEOE was significantly associated with invasive mechanical ventilation (MV), prolonged PICU stay, mortality and Multi-Organ Dysfunction Syndrome (MODS) (all p<0.01). Shorter PICU stay (adjusted OR 0.202; p=0.017) and High-flow nasal cannula (HFNC) (adjusted OR 0.123; p<0.001) use were independently protective against high CEOE. Conclusion: Nearly all critically ill children experienced excess oxygen exposure. CEOE, a potentially modifiable risk factor, predicts in-hospital mortality. This highlights the significance of oxygen titration, practicable in resource-limited settings like ours by using non-invasive SpO2.