Association between administered oxygen, arterial partial oxygen pressure and mortality in mechanically ventilated intensive care unit patients

INTRODUCTION: The aim of this study was to investigate whether in-hospital mortality was associated with the administered fraction of oxygen in inspired air (FiO(2)) and achieved arterial partial pressure of oxygen (PaO(2)). METHODS: This was a retrospective, observational study on data from the first 24 h after admission from 36,307 consecutive patients admitted to 50 Dutch intensive care units (ICUs) and treated with mechanical ventilation. Oxygenation data from all admission days were analysed in a subset of 3,322 patients in 5 ICUs. RESULTS: Mean PaO(2 )and FiO(2 )in the first 24 h after ICU admission were 13.2 kPa (standard deviation (SD) 6.5) and 50% (SD 20%) respectively. Mean PaO(2 )and FiO(2 )from all admission days were 12.4 kPa (SD 5.5) and 53% (SD 18). Focusing on oxygenation in the first 24 h of admission, in-hospital mortality was shown to be linearly related to FiO(2 )value and had a U-shaped relationship with PaO(2 )(both lower and higher PaO(2 )values were associated with a higher mortality), independent of each other and of Simplified Acute Physiology Score (SAPS) II, age, admission type, reduced Glasgow Coma Scale (GCS) score, and individual ICU. Focusing on the entire ICU stay, in-hospital mortality was independently associated with mean FiO(2 )during ICU stay and with the lower two quintiles of mean PaO(2 )value during ICU stay. CONCLUSIONS: Actually achieved PaO(2 )values in ICU patients in The Netherlands are higher than generally recommended in the literature. High FiO(2), and both low PaO(2 )and high PaO(2 )in the first 24 h after admission are independently associated with in-hospital mortality in ICU patients. Future research should study whether this association is causal or merely a reflection of differences in severity of illness insufficiently corrected for in the multivariate analysis.