Bottom Line
This trial showed that the percentage of mechanically ventilated patient post arrest who received conservative oxygen therapy (SaO2 90-95%) with favorable neurological outcomes was not significantly different compared to the percentage of patients who received liberal oxygen therapy (minimum FiO2 of 0.3, SaO2 > 90%). While avoiding hypoxemia along with hypercapnia remains a hallmark of post arrest care, exact oxygenation goals above SaO2 of 90% remain unclear.
Additional Information
LOGICAL trial - NEJM Aug 2026
Background:
Patients who are status post cardiac arrest remain some of the sickest patients in medicine, requiring optimization of physiologic markers to promote the best neurological outcome. One of the most important parameters is oxygenation. However, exact oxygenation goals have been unclear with some trials showing benefit to using more conservative (lower) oxygenation targets in hopes of preventing neuronal death from relative hyperoxia and reperfusion injury. Thus, the Low Oxygen Intervention for Cardiac Arrest Injury Limitation (LOGICAL) trial was conducted.
Patients: > 18 years old mechanically ventilated with suspected ischemic encephalopathy from 53 ICU's in Australia, New Zealand, and Ireland. Exclusions included previous enrollment or inappropriate to enroll (such as imminent death)
Intervention: Conservative oxygen therapy defined as SaO2 goal of 90-95% with alarms for outside these parameters; FiO2 could be increased if SaO2 at goal but PaO2 < 60 mmHg
Control: Liberal oxygen therapy defined as Sa02 goal > 90% with alarms for below this parameter; FiO2 was at 0.3 minimum for mechanical ventilation
Outcome: Extended Glasgow Outcome Scale (GOS-E) score of 5-8 at 180 days reflecting lower moderate disability to upper good recovery ("favorable neurological outcome"); other outcomes include all cause and specific mortality, duration of mechanical ventilation/ICU stay/hospital stay, and discharge to home
Results:
1840 patients enrolled with 1821 analyzed by intention to treat. There were similar baseline characteristics between groups. Less than 1% of patients in each group did not receive assigned conservative/liberal treatment. 26.3% of patients in the conservative oxygen group failed to have FiO2 weaned and 8.4% of liberal oxygen group had an FiO2 less than 0.3.
A favorable GOS-E (5-8) was observed in 38.2% of the conservative oxygen therapy group vs. 39.7% in the liberal oxygen therapy group (RR of 0.97, p = 0.65). Secondary outcomes showed no statistically significant difference.
Discussion:
Overall, the trial did not find a statistically significant difference in favorable neurological outcome or other outcomes in conservative vs. liberal oxygen therapy. The protocol deviations, noted above, likely lessened between-group difference in oxygen exposure. Additionally, time from ROSC to randomization was longer than previous studies which likely did not have a large effect. The trial adds to the literature that conservative oxygen therapy does not lead to improved neurological outcomes compared to liberal oxygen therapy.
In my practice of patient including post arrest patients, I continue to aim for an SaO2 of mid 90's for most of my patients, avoiding hypoxemia (SaO2 < 90% generally) or the persistent SaO2 of 100.
References
- LOGICAL Investigators and the Australian and New Zealand Intensive Care Society Clinical Trials Group; Hodgson CL et al. Conservative Oxygen for Unresponsive Patients after Cardiac Arrest. N Engl J Med. 2026 Aug 6;395(6):571-581. doi: 10.1056/NEJMoa2513814. Epub 2026 Jun 10. PMID: 42267831.