UK-ROX: Conservative vs. Usual Oxygen Therapy in Mechanically Ventilated Adults (2025)

“In mechanically ventilated adults receiving supplemental oxygen in the ICU, minimizing oxygen exposure through conservative oxygen therapy did not significantly reduce all-cause mortality at 90 days.” — The UK-ROX Investigators

1. Publication Details

  • Trial Title: Conservative Oxygen Therapy in Mechanically Ventilated Critically Ill Adult Patients: The UK-ROX Randomized Clinical Trial
  • Citation: Martin DS, et al, for the UK-ROX Investigators. Conservative Oxygen Therapy in Mechanically Ventilated Critically Ill Adult Patients: The UK-ROX Randomized Clinical Trial. JAMA. 2025. doi:10.1001/jama.2025.9663.
  • Published: June 12, 2025, in JAMA
  • Author: Daniel S. Martin, OBE
  • Funding: National Institute for Health and Care Research (NIHR), UK

2. Keywords

  • Critical Care, Mechanical Ventilation, Oxygen Therapy, Conservative Oxygenation, Hyperoxia, Hypoxemia

3. The Clinical Question

  • In mechanically ventilated, critically ill adults receiving supplemental oxygen (Population), does a conservative oxygen therapy strategy targeting an SpO₂ of 88-92% (Intervention) compared to usual oxygen therapy (Comparison) reduce 90-day all-cause mortality (Outcome)?

4. Background and Rationale

  • Existing Knowledge: Oxygen is a ubiquitous drug in the ICU, but the optimal dose is unknown. Both hypoxemia and hyperoxemia (excessive oxygen) are potentially harmful. Previous, smaller trials comparing conservative versus liberal or usual oxygen strategies have yielded conflicting results, leaving significant clinical uncertainty.
  • Knowledge Gap: A very large, pragmatic trial was needed to definitively determine if a conservative oxygen strategy is superior to usual care in a broad population of critically ill, mechanically ventilated patients.
  • Proposed Hypothesis: The investigators hypothesized that a conservative oxygen therapy strategy would be superior to usual oxygen therapy in reducing 90-day all-cause mortality.

5. Study Design and Methods

  • Design: Large-scale, multicenter, pragmatic, registry-embedded, randomized clinical trial.
  • Setting: 97 National Health Service (NHS) ICUs in the United Kingdom.
  • Trial Period: May 2021 to November 2024.
  • Population:
    • Inclusion Criteria: Adults (≥18 years) receiving invasive mechanical ventilation in the ICU following an unplanned admission and receiving supplemental oxygen (FiO₂ >0.21).
    • Exclusion Criteria: Previous randomization in the trial, currently receiving ECMO, or treating clinician considered one of the treatment arms to be either indicated or contraindicated.
  • Intervention: Conservative oxygen therapy: administration of the lowest possible FiO₂ to maintain a peripheral oxygen saturation (SpO₂) of 90% (with an allowed range of 88-92%).
  • Control: Usual oxygen therapy: administration of oxygen at the discretion of the treating clinical team, without a specified upper SpO₂ limit.
  • Management Common to Both Groups: The assigned oxygen strategy was continued until ICU discharge or for a maximum of 90 days. All other aspects of patient care were managed according to usual practice at each site.
  • Outcomes:
    • Primary Outcome: All-cause mortality at 90 days after randomization.
    • Secondary Outcomes: Duration of ICU and hospital stay, days alive and free from organ support at 30 days, and mortality at other timepoints (e.g., ICU discharge, 1 year).

6. Key Results

  • Enrollment and Baseline: 16,500 patients were randomized (8258 to conservative oxygen therapy and 8242 to usual oxygen therapy). The groups were well-balanced at baseline, with a median age of 60 and about a third of patients having sepsis.
  • Trial Status: The trial was completed as planned.
  • Primary Outcome: There was no significant difference in 90-day mortality. 2908 of 8211 patients (35.4%) in the conservative oxygen group died, compared with 2858 of 8183 patients (34.9%) in the usual oxygen group. The adjusted absolute risk difference was 0.7 percentage points (95% CI, -0.7 to 2.0; P=0.28).
  • Secondary Outcomes: There were no significant differences between the groups for any of the secondary outcomes, including duration of ICU stay or days alive and free from organ support.
  • Adverse Events: The incidence of serious adverse events was not reported in the primary abstract, but no safety concerns were highlighted.

7. Medical Statistics

  • Analysis Principle: The primary analysis was by intention-to-treat.
  • Statistical Tests Used: The primary outcome was analyzed using a mixed-effects logistic regression model.
  • Primary Outcome Analysis: Comparison of 90-day mortality rates.
  • Key Statistic(s) Reported: Adjusted absolute risk difference: 0.7%; 95% CI, -0.7 to 2.0; P-value = 0.28.
  • Interpretation of Key Statistic(s):
    • Confidence Interval (CI):
      • Calculation: 95% CI, -0.7% to 2.0%
      • Clinical Meaning: The confidence interval for the absolute risk difference crosses zero, indicating no statistically significant difference between the two groups. The range is compatible with both a small benefit (0.7% absolute risk reduction) and a small harm (2.0% absolute risk increase) for the conservative oxygen strategy.
    • P-value:
      • Calculation: P=0.28
      • Clinical Meaning: The p-value of 0.28 is well above the 0.05 threshold, indicating that the small observed difference in mortality is not statistically significant and is likely due to chance.
  • Clinical Impact Measures:
    • Absolute Risk Reduction (ARR):
      • Calculation: ARR = 34.9% – 35.4% = -0.5% (an absolute risk increase).
      • Clinical Meaning: For every 200 patients treated with conservative oxygen therapy, there was approximately 1 additional death, though this difference was not statistically significant.
    • Number Needed to Treat (NNT): Not applicable as the result was not statistically significant and trended towards harm.
  • Subgroup Analyses: No significant differences in the primary outcome were found in any of the prespecified diagnostic subgroups (e.g., sepsis, COVID-19, acute brain injury).

8. Strengths of the Study

  • Study Design and Conduct: This was an exceptionally large, pragmatic, randomized trial, providing a very high level of evidence. Embedding the trial within existing national audit registries (ICNARC) made it highly efficient and cost-effective.
  • Generalizability: The inclusion of a very large and diverse cohort of critically ill patients from nearly 100 ICUs across the UK makes the findings highly generalizable.
  • Statistical Power: The trial was powered to detect a small but clinically important difference in mortality (2.5%), and the null finding is therefore very robust.

9. Limitations and Weaknesses

  • Separation Between Groups: While statistically significant, the separation in oxygenation between the two groups was modest. The mean of the median SpO₂ was 93.3% in the conservative group versus 95.1% in the usual care group. This small difference may have been insufficient to affect clinical outcomes.
  • Internal Validity (Bias): The open-label (unblinded) nature of the intervention is a potential source of bias, though unavoidable for this type of trial. The primary outcome of mortality is objective and less prone to bias.
  • Other: A large number of screened patients were excluded based on clinician discretion, which could introduce selection bias, although the randomized nature of the trial mitigates this concern for the enrolled cohort.

10. Conclusion of the Authors In a large cohort of mechanically ventilated, critically ill adults, a strategy of conservative oxygen therapy targeting an SpO₂ of 88-92% did not significantly reduce 90-day all-cause mortality compared with usual oxygen therapy.

11. To Summarize

  • Impact on Current Practice: This trial provides strong evidence that targeting a lower-normal SpO₂ range (88-92%) is not superior to usual care and does not improve survival. It suggests that there is no benefit to be gained from a widespread policy of aggressive oxygen restriction in this population.
  • Specific Recommendations:
    • Actionable Intervention: It is safe to allow SpO₂ levels to be in the lower-normal range (e.g., 90-92%) for mechanically ventilated ICU patients, but there is no evidence to suggest this should be actively targeted to improve survival.
  • What This Trial Does NOT Mean: This trial does not mean that hyperoxia is harmless. It simply shows that this specific conservative strategy was not beneficial. It also does not apply to patients with specific conditions where higher or lower oxygen targets are indicated (e.g., paraquat poisoning, certain congenital heart diseases).

12. Context and Related Studies

  • Building on Previous Evidence: This is the largest trial of its kind, providing a much more definitive answer than previous, smaller studies (like the ICU-ROX trial) which had suggested potential benefits or harms in certain subgroups. UK-ROX did not find evidence of such subgroup effects. It provides a crucial piece of the puzzle ahead of the even larger, ongoing MEGA-ROX trial.

13. Unresolved Questions & Future Directions

  • Unresolved Questions: Whether a more personalized approach to oxygen therapy, guided by markers of tissue oxygenation rather than just SpO₂, could improve outcomes remains an open question.
  • Future Directions: Future research will likely focus on individualizing oxygen targets based on patient-specific pathophysiology and exploring the results of the 40,000-patient MEGA-ROX trial.

14. External Links

15. Framework for Critical Appraisal

  • Clinical Question: A fundamental and highly important question in critical care medicine.
  • Methods: The pragmatic, registry-embedded randomized design is a major strength, allowing for a large, generalizable, and efficient trial.
  • Results: The results are clear and provide a null finding for the primary outcome with a high degree of certainty due to the large sample size.
  • Conclusions and Applicability: The conclusion that this conservative oxygen strategy does not improve survival is strongly supported. The findings are highly applicable to the vast majority of adult patients receiving mechanical ventilation in the ICU.

16. Disclaimer and Contact This summary is provided by the Academic Committee of ESBICM (ACE) to facilitate the understanding of this study; readers are advised to refer to the original trial document for a deeper understanding. If you find any information incorrect, or missing, or it needs an update or have a request for a specific critical care trial summary, kindly write to us at academics[at]esbicm.org.

Scroll to Top