Conservative versus liberal oxygen therapy in mechanically ventilated patients: A systematic review with meta-analysis.

Background: Clinical trials comparing liberal versus conservative oxygen therapy strategies in patients requiring mechanical ventilation have shown contradictory results regarding the best strategy for improving mortality outcomes, length of stay, and ventilator-free days. To summarize the evidence regarding the effects of conservative oxygen therapy compared with liberal oxygen therapy in adult patients admitted to an intensive care unit (ICU). Methods: This systematic review with meta-analysis included controlled and randomized clinical trials obtained from the MEDLINE/PubMed, Embase, the Cochrane Library, Lilacs/bvs, PEDro and ScienceDirect. The effect estimate for mortality was expressed as the relative risk (RR), whereas the other variables were expressed as the mean difference (MD). A meta-analysis of the data was conducted via Review Manager software version 5.3 (Cochrane Collaboration). Results: Nineteen randomized clinical trials involving more than 10,000 patients were included. Liberal oxygen therapy did not significantly differ from conventional oxygen therapy in terms of mortality (RR 1.00; 95% CI: 0.93-1.07, GRADE moderate), length of stay (MD 0.18; 95% CI: - 2.69-3.05, GRADE very low) or ventilator-free days (MD 0.25; 95% CI: -1.78-2.27, GRADE very low). Conclusion: The findings of this review show no significant differences in clinical outcomes between liberal and conventional oxygen therapy in adult ICU patients. As the burden of proof rests on the intervention, the absence of evidence for the superiority of liberal oxygen therapy means its benefit cannot be assumed.