Oxygenation Index as a Predictor of Outcome in Children with Acute Hypoxemic Respiratory Failure

Abstract

Wehave several concerns related tomethodology and applicability of the findings in the recent article by Trachsel and colleagues (1), which sought to identify determinants of prognosis in pediatric acute hypoxemic respiratory failure (AHRF). First, the ventilation strategy used by the authors is unclear. The reader is referred to the online supplement, but no further details have been provided. Were lung protective strategies incorporating permissive hypercapnia or acceptance of lower oxygen saturation employed? Were compliance characteristics used to determine adequate lung recruitment? How was “aggressive” use of positive end-expiratory pressure (PEEP) defined? Judging by Trachsel and coworkers’ Figures E1 and E2, the mean peak PEEP in the patients who died appears to be around 10 cm H2O and mean airway pressure (MAP) around 18 cm H2O over the first 48 hours. In many centers, these values would constitute early settings during aggressive management of AHRF, and involve progression to using higher PEEP, or the employment of high-frequency oscillatory ventilation (HFOV) with a significantly higher MAP. The authors attribute the observed mortality of 27% to the presence of underlying medical conditions or multiorgan dysfunction. Could objective composite scores such as the Pediatric Logistic Organ Dysfunction score (PELOD) (2) have helped discriminate the impact of organ dysfunction in this heterogeneous population? In addition, the authors do not present information about actual Pediatric Risk ofMortality (PRISM) scores, even though they conclude that it is an independent predictor of mortality. Could the authors provide additional specifics about the ventilator parameters and escalation of ventilator strategy employed for the 35 patients who died, and especially in the 16 children described as dying of “refractory oxygenation failure,” to help in interpreting the significance of these findings. A last PaO2 of less than 50 mm Hg alone does not indicate refractoriness and could be found if a blood gas is obtained in most dying intensive care unit patients. Were HFOV or extracorporeal membrane oxygenation (ECMO) considered in their management? With additional data and further clarification from the authors, it may become clearer whether the mortality risk assessment determined in this study can be applied to patients at other centers using aggressive conventional or oscillatory ventilation strategies.