Introduction: In intensive care units (ICUs), oxygen therapy has shifted toward conservative targets over concerns that liberal oxygenation may increase mortality. However, recent RCTs report conflicting results in ventilated patients. Methods: We conducted a meta-analysis (OXY-BREATHES) of RCTs comparing conservative (SpO2 88–94% or PaO2 < 80 mmHg) versus liberal oxygen targets (SpO2 ≥94% or PaO2 ≥90 mmHg) in mechanically ventilated ICU patients. Databases searched included PubMed, Cochrane CENTRAL, Embase, and ClinicalTrials.gov. Primary outcomes were 30-day, 90-day, and ICU mortality. Secondary outcomes included ventilator- and vasopressor-free days, renal replacement therapy, ICU length of stay, nosocomial pneumonia, and cardiac or cerebral ischemic events. Prespecified subgroup analyses included patients with sepsis/septic shock; exploratory analyses included those with brain injury and post–cardiac arrest. Results: Eight RCTs enrolling 19,658 patients were included. Conservative versus liberal oxygen targets showed no significant differences in 30-day (RR 1.02; 95% CI, 0.97–1.07), 90-day (RR 1.03; 95% CI, 0.93–1.13), or ICU mortality (RR 1.02; 95% CI, 0.97–1.06). Secondary outcomes, including ICU length of stay and adverse events, were similar. Conservative targets yielded more vasopressor-free days (SMD 0.17; p =0.008) in septic patients and a potential survival benefit in post–cardiac arrest patients (RR 0.89; p =0.04). Conclusions: This meta-analysis suggests conservative oxygen targets provide comparable mortality and safety outcomes to liberal targets in mechanically ventilated ICU patients, with possible benefits in select subgroups. (OXY-BREATHES was registered on PROSPERO, under ID: CRD420251082611)
Nguyen et al. (Sun,) studied this question.