Here is a comprehensive answer based on authoritative sources:
Oxygen Management in Intubated TBI Patients: Should You Keep FiO₂ at 90%?
Short answer: No. You should titrate FiO₂ down to achieve normoxia (SpO₂ 94–98%, PaO₂ ~80–100 mmHg). Sustained hyperoxia is harmful.
The Core Principle: Normoxia is the Target
Multiple major references are clear on this:
"In TBI, normoxia should be maintained."
— Rosen's Emergency Medicine, Secondary Systemic Insults
"Attention to ABCs is critical, with maintenance of normotension and normoxia."
— Mulholland & Greenfield's Surgery, Moderate to Severe TBI
A target SpO₂ of 94–98% with a PaO₂ of ~100 mmHg is the clinically accepted range. Your patient has been sitting at SpO₂ 100% on 90% FiO₂ for >1 hour — that is sustained hyperoxia, and it should be corrected.
Why Hyperoxia is Harmful in TBI
Your clinical concern about cerebral vasoconstriction is correct, and it goes further than that:
1. Reactive oxygen species (ROS) & free radical damage
"Supra-physiologic oxygen concentrations may result in production of reactive oxygen species which have been linked to inflammation, reduced CBF, and necrosis. If there is a potential for hyperoxia improving a patient's condition, the therapeutic window appears to be extremely narrow."
— Plum & Posner's Diagnosis and Treatment of Stupor and Coma, Hyperoxia
2. Microvascular vasoconstriction
"Hyperoxia is associated with vasoconstriction and reduction in microvascular tissue flow, generation of reactive oxygen species, and augmentation of inflammation."
— Miller's Anesthesia, 10e, Hyperoxia
3. Worsened neurological outcomes
"Significant hyperoxia with resultant oxygen toxicity is also associated with worse outcome after TBI, although this relationship is less clearly defined."
— Rosen's Emergency Medicine
Multiple studies have associated brief episodes of arterial hyperoxia with poor outcomes in brain-injured patients. (Plum & Posner)
4. Does NOT reverse secondary metabolic crisis
"Hyperoxia is not effective in completely reversing metabolic crisis. It may lead to persistent secondary injury due to superoxide and free radical generation, mitochondrial dysfunction, and significant exacerbation of neuroinflammation."
— Miller's Anesthesia, 10e, Traumatic Brain Injury (Pediatric ICU section)
What About the "Protective" Effects of Hyperoxia in TBI?
There is some evidence that hyperbaric oxygen therapy (HBOT) — a very controlled, structured intervention — may have neuroprotective mechanisms (improved oxidative metabolism, reduced ICP, reduced apoptosis). However:
- HBOT is a distinct protocol, not simply running high FiO₂ on a ventilator
- It is still under investigation (HOBIT trial)
- This does NOT justify maintaining 90% FiO₂ on a mechanically ventilated patient who is already at SpO₂ 100%
Practical Recommendation
| Parameter | Target in TBI |
|---|
| SpO₂ | 94–98% |
| PaO₂ | ~80–100 mmHg |
| FiO₂ | Titrate down — wean to lowest FiO₂ maintaining SpO₂ ≥94% |
| PaCO₂ | 35–45 mmHg (normocapnia; avoid hyperventilation except for acute herniation) |
Action: Wean FiO₂ stepwise (e.g., 90% → 60% → 40%) with continuous SpO₂ monitoring and periodic ABGs. The goal is the lowest FiO₂ that keeps SpO₂ in the 94–98% range.
One Additional Note on Your Mechanism Concern
You mentioned hyperoxia causing cerebral vasoconstriction → reduced CPP. The mechanism is accurate — hyperoxia-induced vasoconstriction reduces CBF and can worsen cerebral ischemia in vulnerable penumbral zones, especially in a brain already suffering impaired autoregulation after TBI. Maintaining MAP and CPP targets (CPP ≥60–70 mmHg per BTF guidelines) while simultaneously correcting hyperoxia is the right dual approach.
Sources: Rosen's Emergency Medicine; Plum & Posner's Diagnosis and Treatment of Stupor and Coma; Miller's Anesthesia 10e; Mulholland & Greenfield's Surgery 7e