Conservative oxygen targets show no mortality difference in mechanically ventilated critically ill adults

This systematic review and meta-analysis synthesized data from randomized controlled trials comparing conservative oxygen targets (SpO2 88-94% or PaO2 < 80 mm Hg) to liberal oxygen targets (SpO2 ≥ 94% or PaO2 ≥ 90 mm Hg) in mechanically ventilated critically ill adults. The population included 20,447 patients across intensive care unit settings, encompassing those with sepsis, septic shock, post-cardiac arrest, and other critical illnesses. The analysis focused on critically ill adults requiring mechanical ventilation, with the intervention and comparator defined by specific oxygen saturation or partial pressure thresholds.

The primary outcomes were 90-day mortality and ICU length of stay. For 90-day mortality, the risk ratio was 1.01 with a 95% CI of 0.94 to 1.09, indicating no substantial difference between conservative and liberal oxygen targets. For ICU length of stay, the mean difference was -0.17 days with a 95% CI of -0.41 to 0.06, also showing no substantial difference. These results suggest neutral effects for the primary outcomes in the overall population.

Key secondary outcomes included ventilator-free days, vasopressor-free days, renal replacement therapy, nosocomial pneumonia, cardiac ischemic events, and cerebral ischemic events. In septic patients, conservative oxygen targets were associated with more vasopressor-free days, with a mean difference of 2.0 days and a p-value of 0.008. For survival in post-cardiac arrest patients, conservative targets showed a potential benefit with a risk ratio of 0.89 and a p-value of 0.05. Other secondary outcomes were not reported with specific numeric results in the input.

Safety findings indicated that adverse events were comparable between groups. Serious adverse events, discontinuations, and tolerability were not reported. The certainty of evidence was rated moderate for 90-day mortality, ICU length of stay, vasopressor-free days, and ventilator-free days; low for renal replacement therapy and nosocomial pneumonia; and very low for cerebral and cardiac ischemia.

These results compare to prior landmark studies in this therapeutic area, such as the OXYGEN-ICU and HOT-ICU trials, which have explored oxygen targets in critical care. The current meta-analysis reinforces the neutral effect on mortality while highlighting potential subgroup benefits. Methodological limitations include open-label trial designs and imprecision for some outcomes, which may introduce bias and affect the reliability of findings.

Clinical implications suggest that conservative oxygenation is comparable to liberal oxygen targets in mechanically ventilated critically ill patients, with possible advantages in sepsis and post-cardiac arrest scenarios. However, clinicians should not infer causation from association, extrapolate to non-mechanically ventilated patients, or ignore the low or very low certainty for some outcomes. Practice decisions should consider the moderate certainty for primary outcomes and the specific patient contexts.

Unanswered questions remain regarding the optimal oxygen targets for non-mechanically ventilated patients, long-term outcomes beyond 90 days, and the mechanisms underlying the observed benefits in septic and post-cardiac arrest subgroups. Future research should address these gaps to refine clinical guidelines.