Target arterial PO(2) according to the underlying pathology: a mini-review of the available data in mechanically ventilated patients

Demiselle, Julien; Calzia, Enrico; Hartmann, Clair; Messerer, David Alexander Christian; Asfar, Pierre; Radermacher, Peter; Datzmann, Thomas

There is an ongoing discussion whether hyperoxia, i.e. ventilation with high inspiratory O(2) concentrations (F(I)O(2)), and the consecutive hyperoxaemia, i.e. supraphysiological arterial O(2) tensions (PaO(2)), have a place during the acute management of circulatory shock. This concept is based on experimental evidence that hyperoxaemia may contribute to the compensation of the imbalance between O(2) supply and requirements. However, despite still being common practice, its use is limited due to possible oxygen toxicity resulting from the increased formation of reactive oxygen species (ROS) limits, especially under conditions of ischaemia/reperfusion. Several studies have reported that there is a U-shaped relation between PaO(2) and mortality/morbidity in ICU patients. Interestingly, these mostly retrospective studies found that the lowest mortality coincided with PaO(2) ~ 150 mmHg during the first 24 h of ICU stay, i.e. supraphysiological PaO(2) levels. Most of the recent large-scale retrospective analyses studied general ICU populations, but there are major differences according to the underlying pathology studied as well as whether medical or surgical patients are concerned. Therefore, as far as possible from the data reported, we focus on the need of mechanical ventilation as well as the distinction between the absence or presence of circulatory shock. There seems to be no ideal target PaO(2) except for avoiding prolonged exposure (> 24 h) to either hypoxaemia (PaO(2) < 55–60 mmHg) or supraphysiological (PaO(2) > 100 mmHg). Moreover, the need for mechanical ventilation, absence or presence of circulatory shock and/or the aetiology of tissue dysoxia, i.e. whether it is mainly due to impaired macro- and/or microcirculatory O(2) transport and/or disturbed cellular O(2) utilization, may determine whether any degree of hyperoxaemia causes deleterious side effects.