Oxygen After Cardiac Arrest: The LOGICAL Trial Closes the Book on "Less Is More"

A large multinational RCT finds no functional or survival benefit to conservative oxygen titration in unresponsive post-arrest ICU patients.

For more than a decade, a plausible biological story has driven ICU oxygen practice after cardiac arrest.

Animal data suggested that hyperoxemia during the reperfusion window worsens oxidative brain injury, and several small trials hinted that a conservative oxygen strategy might improve neurologic recovery.

The LOGICAL trial, published this June in the New England Journal of Medicine, was designed to settle the question definitively.

The answer, in short, is that it did not settle in favor of the conservative strategy at all.

What LOGICAL Actually Tested

Investigators enrolled 1,840 unresponsive adults from 53 ICUs across Australia, New Zealand, and Ireland who were undergoing mechanical ventilation after return of spontaneous circulation (ROSC).

Most arrests were out-of-hospital and witnessed, and roughly three in four patients received bystander CPR.

Median time from ROSC to randomization was about seven hours, which is notably longer than in several prior oxygen trials.

Patients were randomized to either a conservative oxygen protocol or a liberal oxygen protocol, with both sharing a default lower SpO2 alarm limit of 90%.

Figure 1. Protocol design of the two randomized oxygen strategies in LOGICAL.

The Primary Result: Convincingly Neutral

At 180 days, a favorable functional outcome occurred in 38.2% of the conservative-oxygen group versus 39.7% of the liberal-oxygen group, a difference that was not statistically significant.

Survival to day 180 was likewise similar between arms, at 48.0% with conservative oxygen and 49.7% with liberal oxygen.

Quality-of-life scores, patient- or proxy-reported health state, and cognitive testing via the Montreal Cognitive Assessment showed no meaningful separation between groups.

Prespecified subgroup analyses, including stratification by time from ROSC to randomization, showed no signal favoring either strategy.

The trial's own investigators described the findings as consistent regardless of when oxygen titration began.

Figure 2. Approximate rates of favorable functional outcome (or comparable neurologic endpoint) across three contemporary oxygen-titration trials in post-arrest patients; values are approximate and trial endpoints/definitions differ, so cross-trial comparison is illustrative rather than statistical.

How LOGICAL Fits With the Rest of the Evidence

LOGICAL is nested within the much larger Mega-ROX program, which will eventually add roughly 2,200 more patients and report in-hospital mortality as its primary endpoint.

The trial's findings align closely with the earlier BOX trial, which also found no difference between restrictive and liberal oxygen targets in comatose out-of-hospital arrest survivors.

They likewise mirror the prehospital EXACT trial, which found no neurologic benefit from a conservative approach and raised concern about titration accuracy in the field.

Together, these three modern, adequately powered trials stand in contrast to an earlier generation of smaller studies, including ICU-ROX, that had suggested a possible mortality benefit with conservative oxygen.

A prior individual patient-level meta-analysis had found a statistically significant mortality reduction with conservative oxygen, but the certainty of that evidence was rated low given bias and imprecision across small trials.

Table 1. Contemporary randomized trials of oxygen strategy after cardiac arrest

Trial	Population	Comparison	Primary Result
LOGICAL (2026)	Unresponsive ICU patients post-arrest, in/out-of-hospital	Conservative vs liberal FiO2/SpO2	No difference in 180-day favorable outcome
BOX	Comatose OHCA survivors	Restrictive vs liberal oxygen target	No difference in mortality/neurologic outcome
EXACT	Prehospital OHCA patients	Conservative vs usual-care oxygen titration	No neurologic benefit; titration challenges noted
ICU-ROX (earlier era)	ICU patients with suspected HIE post-arrest	Conservative vs usual oxygen	Lower day-180 mortality with conservative oxygen (smaller trial)

Practical Implications for the ICU

The most defensible reading of this body of evidence is that hitting a specific low oxygen target does not, by itself, rescue the injured brain after arrest.

An SpO2 floor around 90% to avoid frank hypoxemia remains reasonable, since both LOGICAL arms shared that lower limit.

Chasing an aggressively low ceiling with tight alarms, however, does not appear to add measurable neurologic or survival benefit and adds nursing and respiratory therapy titration burden.

Clinicians should also note that neuroprognostication protocols and their timing did not differ meaningfully between groups, so the neutral result is unlikely to reflect an assessment artifact.

Because LOGICAL enrolled from high-income health systems in Australia, New Zealand, and Ireland, generalizability to resource-limited ICUs remains uncertain pending the broader Mega-ROX HIE dataset.

Table 2. LOGICAL protocol parameters at a glance

Parameter	Conservative Group	Liberal Group
Default lower SpO2 alarm	90%	90%
Upper SpO2 alarm	95%	Not set
Minimum FiO2 during ventilation	0.21 (once SpO2>90%)	0.30
Procedural high-FiO2 exceptions	Permitted	Permitted
180-day favorable functional outcome	38.2%	39.7%
180-day survival	48.0%	49.7%

Case Vignette

A 58-year-old is brought to the ICU after an out-of-hospital arrest with bystander CPR and return of circulation after 22 minutes.

The patient remains unresponsive on mechanical ventilation six hours after ROSC, and the team is deciding on an oxygenation target while awaiting further neuroprognostic testing.

Based on LOGICAL, the team can reasonably set a standard lower SpO2 alarm around 90% without feeling obligated to enforce an aggressively low FiO2 ceiling.

The decision can instead focus on avoiding both hypoxemia and extreme hyperoxemia using routine ICU oxygen management, rather than a rigid conservative protocol.

Bottom Line

In the largest trial of its kind, a conservative oxygen strategy did not improve 180-day survival, functional outcome, quality of life, or cognition compared with a liberal strategy in unresponsive post-arrest ICU patients.

These results are consistent with BOX and EXACT, reinforcing that meticulous low-oxygen titration protocols are unlikely to be the lever that improves outcomes after cardiac arrest.

The larger Mega-ROX HIE program, expected to add roughly 2,200 additional patients with an in-hospital mortality endpoint, is anticipated to be confirmatory rather than to reverse this signal.

Routine, pragmatic oxygen management with a reasonable lower SpO2 threshold remains a sound default pending further data.

This article is intended for physician education and reflects a summary of publicly available trial data; it is not a substitute for independent review of the primary literature or institutional protocols before changing clinical practice.

References

1. Conservative Oxygen for Unresponsive Patients after Cardiac Arrest — New England Journal of Medicine
2. LOGICAL: Conservative Oxygen Therapy Fails to Improve Survival Outcomes After Cardiac Arrest — ACC.org Journal Scan
3. LOGICAL Trial Overview — Australian and New Zealand Intensive Care Society
4. Protocol Summary and Statistical Analysis Plan for the LOGICAL Trial — PMC
5. Higher BP, Oxygen Targets No Help in Comatose OHCA Patients: BOX — TCTMD