Association Between Arterial Hyperoxia and Mortality in Pediatric and Adult Patients Undergoing Extracorporeal Membrane Oxygenation: A Systematic Review and Meta-Analysis

BACKGROUND: In patients receiving extracorporeal membrane oxygenation (ECMO) support, the association between arterial hyperoxia and outcomes is unclear. We performed a systematic review and meta-analysis to determine the association between arterial P o2 (Pa o2 ) and mortality in patients with ECMO. METHODS: The meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement and registered in International Prospective Register of Systematic Reviews (PROSPERO; CRD42023467361). We systematically searched PubMed and Embase databases up to September 2023 for randomized trials or observational studies that investigated the association between Pa o2 and mortality in pediatric and adult patients receiving venovenous ECMO (VV-ECMO), venoarterial ECMO (VA-ECMO), and extracorporeal cardiopulmonary resuscitation (ECPR). The predefined outcome was 28-day mortality. We synthesized the data using a random-effects model, calculating odds ratios (OR) and corresponding 95% confidence intervals (CI). RESULTS: Thirteen cohort studies (17,766 participants) were included. All studies used categorical Pa o2 cutoff, with varying thresholds ranging from ≥100 mm Hg to ≥300 mm Hg. When compared with patients with normoxia, elevated Pa o2 levels at all studied thresholds were consistently associated with increased mortality (≥300 mm Hg: OR 1.56, 95% CI, 1.31-1.85, P < .01; ≥200 mm Hg: OR 1.43, 95% CI, 1.10-1.87, P < .01; ≥150 mm Hg: OR 1.51, 95% CI, 1.15-1.98, P < .01; and ≥100 mm Hg: OR 1.44, 95% CI, 1.03-2.02, P = .03). A sensitivity analysis focusing on studies reporting adjusted OR yielded similar results. We observed this association in both adult and pediatric populations. CONCLUSIONS: In critically ill patients on VV- or VA-ECMO, increased Pa o2 values were associated with increased 28-day mortality in ECMO patients. Our results should be interpreted with caution given observational nature of included studies. Further randomized trials are warranted to validate these results.