Oxygenation index predicts outcome in children with acute hypoxemic respiratory failure

To define outcome and time dependence of predictors of outcome in pediatric acute hypoxemic respiratory failure, 131 patients (age range, 1 month to 18 years) were prospectively followed. Parametric models were used to describe time-related events, and competing risks analysis was performed for mortality estimates. Multiple logistic analysis was applied to describe time-related predictors of ventilation time and mortality. Overall mortality was 27%. Peak oxygenation index (OI) measured at any time point (p < 0.001, 91% reliability in bootstrapping, after inverse transformation) and Pediatric Risk of Mortality, or PRISM, score within the first 12 hours of mechanical ventilation (p < 0.001, 63% reliability in bootstrapping, after square transformation) were identified as independent predictors of mortality. Peak OI, younger age, and need for renal replacement therapy were significantly associated with a longer time to extubation. Although OI was less reliable as outcome predictor within the first 12 hours of intubation, it still predicted duration of mechanical ventilation. No clear-cut threshold of OI was identified that could accurately predict mortality. Survival was characterized by a peak rate of extubations at approximately 1 week, with a more gradual decline thereafter, whereas death appeared as a constant risk over time, which exceeded chances of survival at approximately 4 weeks. Severity of oxygenation failure at any point in time during acute hypoxemic respiratory failure correlates with duration of mechanical ventilation and mortality. This is best reflected by the OI, which shows a direct correlation to outcome in a time-independent manner.