ICU-ROX: Conservative vs Usual O2 Therapy During Mechanical Ventilation in the ICU

Background: The IOTA trial, was a systematic review and meta-analysis of 25 RCTs enrolling 16,000 patients with sepsis, critical illness, stroke, trauma, MI, cardiac arrest, and/or emergency surgery. In this review it was found that liberal use of O2 resulted in a higher in-hospital and 30d mortality with NNH of 143 and 125 respectively compared to conservative O2 therapy.  Since supplemental oxygen is commonly used in the critically ill, it is important to establish parameters for oxygen supplementation, especially in patients undergoing mechanical ventilation where there is no good data regarding strategies for oxygen administration.

What They Did:

  • Intensive Care Unit Randomized Trial Comparing Two Approaches to Oxygen Therapy (ICU-ROX)
  • Investigator-initiated, parallel group, randomized clinical trial
  • 1000 adult patients anticipated to require mechanical ventilation beyond the day after recruitment in the ICU
  • Randomized to:
  • Conservative oxygen therapy: Upper limit of oxygen saturation measured by pulse oximetry (SpO2) was 97% and the FiO2 was decreased to 0.21 if the SpO2 was above the acceptable lower limit
  • Usual oxygen therapy: No specific measures limiting the FiO2 or the SpO2
  • In both groups lower limit for oxygen saturation measured by pulse oximetry was 90%


  • Primary: Number of ventilator-free days from randomization until day 28 (total number of calendar days or portions of calendar days of unassisted breathing during the 1st 28d after randomization)
  • Key Secondary:
    • Death from any cause at 90 and 180d
    • Duration of survival
    • Proportion of patients in paid employment at baseline who were unemployed at day 180
    • Cognitive function and health-related quality of life at day 180
    • Cause-specific mortality


  • ≥18 years of age
  • Expected to receive mechanical ventilation in the ICU beyond the day after recruitment
  • Received <2hrs of invasive mechanical ventilation or noninvasive ventilation in the ICU


  • Hyperoxia clinically indicated for reasons including but not limited to carbon monoxide poisoning or a requirement for hyperbaric oxygen therapy
  • Avoidance of hyperoxia is clinically indicated for reasons including but not limited to COPD, paraquat poisoning, previous exposure to bleomycin, or chronic hypercapnic respiratory failure
  • Pregnancy
  • Death is deemed inevitable
  • Life expectancy <90d
  • Drug overdose (including alcohol intoxication)
  • Long term dependence on invasive ventilation
  • Confirmed or suspected diagnosis of Guillain-Barre syndrome, cervical cord injury above C5, muscular dystrophy, motor neuron disease
  • Enrollment not considered in patient’s best interest
  • Enrolled in any other trial of targeted oxygen therapy
  • Previously enrolled in the ICU-ROX study


  • Enrolled 1000 patients in 21 ICUs in Australia and New Zealand
    • Consent withdrawn in 35 patients
    • Conservative Oxygen Therapy: N = 484
    • Usual Oxygen Therapy: N = 481
  • Median Number of ventilator-free days:
    • Conservative Oxygen Therapy: 21.3d (Range 0 to 26.3)
    • Usual Oxygen Therapy: 22.1d (Range 0 to 26.2)
    • 95% CI 02.1 to 1.5; p = 0.80
    • No Statistical Difference
  • Median Time in the ICU with an FiO2 of 0.21:
    • Conservative Oxygen Therapy: 29hrs (Range 5 to 78)
    • Usual Oxygen Therapy: 1h4 (Range 0 to 17)
    • 95% CI 22 to 34
  • Time with SpO2 Exceeding 96%
    • Conservative Oxygen Therapy: 27hrs (Range 11 to 63.5)
    • Usual Oxygen Therapy: 49hrs (Range 22 to 112)
    • 95% CI 14 to 30
  • 90d Mortality:
    • Conservative Oxygen Therapy: 34.7%
    • Usual Oxygen Therapy: 32.5%
    • uOR 1.10 (95% CI 0.84 to 1.44)
  • 180d Mortality:
    • Conservative Oxygen Therapy: 35.7%
    • Usual Oxygen Therapy: 34.5%
    • uOR 1.05 (95% CI 0.81 to 1.37)
  • Adverse Events: No difference between groups (2 pts in conservative oxygen therapy group vs 1 pt in usual oxygen therapy group)


  • Asks a clinically important question
  • Statistical analysis plan was reported before the completion of enrollment
  • Used an intention-to-treat analysis which is a more realistic characterization of results to the real world
  • Clear separation in oxygen exposure between the two groups


  • There were 1510 patients who were eligible for enrollment but missed the enrollment window, making this more of a convenience sample as opposed to a consecutive sample
  • Hawthorne effect – non-blinded and everyone knew a study was going on. Clinicians may have worked harder and titrated oxygen more often than they normally would have if a study wasn’t going on
  • Clinicians and research staff were aware of trial-group assignments (non-blinded) which may have reduced the impact of the intervention
  • Some quality of life and cognition data was missing, and these data may not be missing at random as patients with better (or worse) outcomes might have been harder to contact or less likely to complete interviews
  • Liberal oxygen strategy in this trial may have been more conservative than prior liberal oxygen use trials, which may dilute potential benefit of conservative oxygen therapy


  • In the subgroup of patients with suspected hypoxic-ischemic encephalopathy there was a difference in median ventilator-free days, but not in any of the other prespecified subgroups:
    • Conservative Oxygen Therapy: 21.1 (Range 0 to 26.1)
    • Usual Oxygen Therapy: None (Range 0 to 26)
    • This finding is hypothesis generating
  • Mortality difference of ≈1%. In a larger study with this as the primary outcome, this may have been statistically significant.  This number is in line with the IOTA trial and, a 1% absolute mortality reduction would be important for a free intervention.  In the absence of harm, minimal FiO2 would be a strategy that makes more clinical sense

Author Conclusion: “In adults undergoing mechanical ventilation in the ICU, the use of conservative oxygen therapy, as compared with usual oxygen therapy, did not significantly affect the number of ventilator-free days.”

Clinical Take Home Point: This trial showed no difference in conservative vs usual oxygen therapy in adult patients undergoing mechanical ventilation. However, the usual care oxygen strategy used in this trial may have been more conservative compared to prior studies.  The weight of evidence supports continuing to wean FiO2 as quickly and safely as possible to minimize patient harms. Given the absence of harm we should clearly embrace lower FiO2 since it reduces cost, and there is lots of evidence saying we should avoid it. 


  1. The ICU-ROX Investigators et al. Conservative Oxygen Therapy During Mechanical Ventilation in the ICU. NEJM 2019. PMID: 31613432

For More Thoughts on This Topic Checkout:

Post Peer Reviewed By: Anand Swaminathan, MD (Twitter: @EMSwami)

Cite this article as: Salim Rezaie, "ICU-ROX: Conservative vs Usual O2 Therapy During Mechanical Ventilation in the ICU", REBEL EM blog, November 21, 2019. Available at:

Like this article?

Share on Facebook
Share on Twitter
Share on Linkdin
Share via Email

Want to support rebelem?