Feasibility and safety of conservative versus liberal oxygen targets in the mechanically ventilated patients.

Clinicians consider arterial oxygen saturation (SpO2) of at least 88-90% as an acceptable target for most patients undergoing mechanical ventilation (MV). Although several experimental and clinical studies report that high levels of oxygen (O2) may be injurious, and recent RCTs show that O2 therapy titrated to lower SpO2 targets (88-94%) improves clinical outcomes in some patient groups, major observational studies show that ICU patients often receive much higher levels of oxygenation in the first 24 hours of ICU admission. Indeed, in a recently completed observational study in ICU patients on MV, we observed that high oxygenation levels were usually achieved with the use of liberal amounts of O2 in routine practice. At a teaching hospital, the 95% confidence interval for the mean time-weighted average SpO2 for the first 7 days of ventilation was 96.8-97.4% and the mean FiO2 of 0.40 or more was used to achieve mean SpO2 of >95% on 40% of the ventilated days. There are no data in literature, for adult ICU patients receiving MV, to demonstrate whether or not higher oxygenation targets (SpO2 >95%) lead to different outcomes than the alternative lower targets (SpO2 88-94%). We aim to systematically evaluate whether liberal O2 therapy is beneficial or harmful compared to a more conservative O2 therapy in critically ill patients. To help us determine the optimal size and design of a suitable trial, a pilot study will be first performed to assess feasibility of the two oxygenation strategies. Our hypothesis for this pilot study is that in critically ill patients on invasive MV, an effective separation of SpO2 levels can be achieved between the two (conservative vs. liberal oxygen targets) groups, without any significant problems or safety concerns in implementing the study protocol in different critical care settings. 100 eligible patients, will be randomly allocated to either a liberal oxygenation strategy with target SpO2 of at least 96% or a conservative oxygenation strategy with target SpO2 of 88-92%. The treating team will set PEEP as per usual practice. When FiO2 requirements are >0.80, SpO2 targets will be determined by the treating clinicians. The pragmatic nature of the trial allows flexibility in management decisions and allows the treating clinicians to alter oxygenation targets if necessary according to patient’s current clinical state. Other co-interventions such as sedation, nutrition and weaning from ventilation will follow standard practices. This pilot study will provide crucial preliminary information on the likely effects of two different target levels of oxygenation in ICU patients on invasive MV. If feasibility and safety are demonstrated, we will follow this study with a definitive RCT, the results of which will have the potential to substantially influence clinical practice in relation to O2 therapy in critically ill patients.