Authors: Rob Hirst, Becky Maxwell, Chris Connolly, Dave McCreary, Andy Neill / Codes: EP1, RP4, SLO10, TP7 / Published: 11/03/2025
Clinical Question
Is too much oxygen bad for trauma patients?
Title
Early Restrictive vs Liberal Oxygen for Trauma Patients. The TRAUMOX2 Randomized Clinical Trial
Author
Tobias Arleth, JAMA Dec 2024
Background
Oxygen is kind of important, there are recommendations to be used for severely injured patients but often with little evidence. Too much oxygen seems definitively bad for premature infants and may be bad for lots of other patient populations. This is thought to be related to nasty things called reactive oxygen species that cause damage to lungs. I won't claim to understand much more beyond that. This is a randomised trial to see if a restrictive oxygen strategy improves outcomes in trauma patients specifically.
Methods
Danish in origin but netherlands and switzerland also enrolled - enrollled those with trauma that triggered a full trauma team activation (not immediately clear what that means) - could be prehospital ranodmisation - excluded potential carbon monixide poisoning which seems prudent - allocation concealment with sealed envelopes - randomised to an 8 r period of either 1) restrictive (sats>94%) or liberal (NRB or Fio2 1.0 if tubed) which could be knocked down to 12L/min or 60% once out of resus - allowed to use normal preoxygenation techniques for intubating. - composite outcome of death and major resp complications at 30 days. This captured pnuemonia and ARDS. These were defined by specific outocme assessors with all the clinical info but blinded to allocation. It seems they went to some effort with this to hide the Fio2 etc... - power calculation was based on similar outcomes from a pilot study but always comes across like a bit of a fudge. Looked for a 5% absolute reduction in the composite from 15% to 10%.
Results
1500 pts (which was their target) - mostly middle aged males with blunt injuries (90%) - 25% tubed at randomisation - 60/40 in hospital/pre hospital randomisation - odds ratio 1.01 and a forest plot showing no difference in the primary outcome. - bhut when you split that up it looks like more oxygen means less death but more resp complciations. Splitting it up is of course not what the trial was powered for but it's interesting to look at. - mortality rate overall ~8% so they weren't especially sick trauma patients. - unsuprisingly there more hypoxic episodes in the retrictive group - they did seem to acheive sepeartion in the amount of oxygen given for the brief study period
Thoughts
This is open label so everyone knew what was happening here (apart from the outcome assessors apparently) - it's a very short intervention to expect it to produce such a big change in outcome.