Conservative or liberal oxygen therapy for mechanically ventilated adults with acute brain pathologies: A post-hoc subgroup analysis.

Abstract

To compare the effect of conservative vs. liberal oxygen therapy in mechanically ventilated adults in the intensive care unit (ICU) with non-hypoxic ischemic encephalopathy (HIE) acute brain pathologies. Post-hoc analysis of data from 217 patients with non-HIE acute brain pathologies included in the ICU Randomized Trial Comparing Two Approaches to OXygen therapy (ICU-ROX). Patients allocated to conservative oxygen spent less time with oxygen saturation ≥ 97% (50.5 [interquartile range (IQR), 18.5-119] vs. 82 h [IQR, 38-164], absolute difference, -31.5 h; 95%CI, -59.6 to -3.4). At 180 days, 38 of 110 conservative oxygen patients (34.5%) and 28 of 104 liberal oxygen patients (26.9%) had died (absolute difference, 7.6 percentage points; 95%CI, -4.7 to 19.9 percentage points; P = 0.23; interaction P = 0.02 for non-HIE acute brain pathologies vs. HIE; interaction P = 0.53 for non-HIE acute brain pathologies vs. non-neurological conditions). In this post-hoc analysis, patients admitted to the ICU with non-HIE acute brain pathologies treated with conservative oxygen therapy did not have significantly lower mortality than those treated with liberal oxygen. A trial with adequate statistical power is needed to determine whether our day 180 mortality point estimate of treatment effect favoring liberal oxygen therapy indicates a true effect.