Conservative versus liberal oxygen therapy for acutely ill medical patients: A systematic review and meta-analysis

Background The role of conservative versus liberal oxygen therapy for acutely ill patients remains controversial. Objective To systematically review the available evidence regarding the efficacy and safety of conservative oxygen therapy compared with liberal oxygen therapy for acutely ill patients. Methods A systematic search of Medline, Embase, and the Cochrane Central Register was conducted from their inception until April 5, 2020. Randomized clinical trials evaluating a high-target (liberal) or a low-target (conservative) oxygenation strategy in adults with an acutely ill condition were eligible for inclusion. A meta-analysis using random-effects models was conducted to calculate the risk ratio with corresponding 95% confidence intervals. Heterogeneity and publication bias were evaluated. Results The analyses included 33 randomized clinical trials with a total of 17,780 participants. Compared with conservative oxygen therapy, liberal oxygen therapy was not associated with increased mortality at 30 days (risk ratio 1.09, 95% confidence intervals 0.98–1.22; I2=0%), at 90 days (risk ratio 1.00, 95% confidence intervals 0.88–1.13, I2=37%), or at the longest follow-up (risk ratio 1.04, 95% confidence intervals 0.96–1.12, I2=0%). Good functional outcome was similar between groups. Findings were robust to trial sequential, subgroup, and sensitivity analyses. Conclusions Compared with liberal oxygen therapy, conservative oxygen therapy was not associated with decreased mortality.Tweetable abstract: Compared with liberal oxygen therapy, conservative oxygen therapy was not associated with decreased mortality.