4.4 Mechanical Ventilation

A ventilator in the neuro ICU has two jobs that sometimes fight: protect the lung and protect the brain’s blood flow. This independent OpenExamPrep section covers low-tidal-volume mechanics, how PaCO2 moves cerebral blood flow (CBF), what to target after traumatic brain injury (TBI) versus in a brain-dead donor versus in ARDS, and how liberation and dyssynchrony show up on an ICP trace.

Lung-protective ventilation

The ARDSNet chassis is the default for almost every adult ICU lung, injured or not. Set tidal volume near 6 mL/kg predicted body weight (PBW), not scale weight. PBW for men is 50 + 0.91 × (height in cm − 152.4); for women 45.5 + 0.91 × (height in cm − 152.4). A 180 cm man who weighs 140 kg still has a PBW near 75 kg; 6 mL/kg is about 450 mL, not 840 mL. Keep plateau pressure ≤30 cm H2O. If plateau is 34 cm H2O at 6 mL/kg, you lower volume further toward 4–5 mL/kg and treat the reason (dyssynchrony, pneumothorax, mainstem tube, abdominal compartment). PEEP and FiO2 are titrated to oxygenation, often along ARDSNet tables in true ARDS. Moderate PEEP is compatible with most brain injuries if MAP is supported; extreme PEEP can raise intrathoracic pressure, impede venous return, and lift ICP—watch the ICP and CPP rather than declaring all PEEP illegal.

Oxygenation targets in brain injury avoid both hypoxia and needless hyperoxia. Keep SpO2 generally in the mid-90s rather than 100% on FiO2 1.0 for hours. Hypoxia (PaO2 <60 mm Hg) is not a brain-protective strategy.

Volume-assist control is a common starting mode because it guarantees a minute ventilation while you control PaCO2. Pressure support is a weaning mode, not a herniation mode. Whatever the mode, synchrony matters as much as the written tidal volume: a patient who double-triggers is actually taking 12 mL/kg.

PaCO2, CBF, and the herniation bridge

Carbon dioxide is a potent cerebral vasodilator. Hypercapnia raises CBF and cerebral blood volume and can raise ICP. Hypocapnia constricts arterioles, lowers CBF and blood volume, and can drop ICP—at the cost of ischemia if you overdo it. A working bedside heuristic is that CBF changes on the order of 3–4% per mm Hg PaCO2 in the responsive range, until vessels are already maxed out by injury.

Prophylactic hyperventilation (bagging every intubated TBI patient to PaCO2 25 “because ICP might rise”) is harmful. Brain Trauma Foundation 4th Edition: prolonged prophylactic hyperventilation with PaCO2 ≤25 mm Hg is not recommended. Hyperventilation should be avoided in the first 24 hours after TBI, when CBF is often already critically reduced. If you hyperventilate, older BTF commentary supports watching jugular venous oxygen saturation (SjO2) or brain-tissue oxygen (PbtO2) so you can see the ischemia you just created.

Brief hyperventilation as a herniation bridge is still taught: a few minutes of lower PaCO2 (often toward the low 30s, not below 25 mm Hg) while mannitol or 23.4% saline is infusing and the OR is called. Then return toward normocapnia (about 35–45 mm Hg). Herniation is not treated with a 12-hour night-shift of PaCO2 22. End-tidal CO2 is a trend, not a PaCO2; check an arterial blood gas when the brain is on the line.

Permissive hypercapnia, beloved in ARDS, is a negotiated truce in high ICP. If ICP is 35 mm Hg, you may need a higher minute ventilation and a slightly less “permissive” CO2, while still defending the 6 mL/kg and plateau ceiling as best you can. If ICP is 8 mm Hg in recovering ARDS, let PaCO2 ride in the mid-40s rather than stacking injurious tidal volumes.

Targets: TBI versus brain-dead donor versus ARDS

These three patients can share a bed space and not share a CO2 goal.

After TBI (living patient): prevent hypoxia and hypotension. Ventilate to normocapnia. Treat ICP at the BTF-style threshold near 22 mm Hg with the full bundle (head position, sedation, CSF drainage, osmotherapy, surgery)—not with standing hyperventilation orders. CPP commonly targeted around 60–70 mm Hg. Use PEEP that oxygenates without cratering MAP. Daily awakening is desirable when ICP is quiet and the patient is not in status; it is deferred during active herniation or anesthetic treatment of seizures.

Brain-dead donor: the neurologic examination is over. The ventilator now serves organs, especially lungs and heart. Stay lung-protective (often 6–8 mL/kg PBW), use enough PEEP (commonly ≥8 cm H2O in lung-donor pathways) to keep alveoli open, avoid fluid drowning the lungs, and keep PaO2/FiO2 in a range transplant teams can use. PaCO2 is usually kept near normal except during the apnea test, which is a diagnostic maneuver (AAN/AAP/CNS/SCCM 2023: no respiration, pH <7.30, PaCO2 ≥60 mm Hg and ≥20 mm Hg above baseline, with extra rules in chronic CO2 retainers)—not a 6-hour hyperventilation recipe. Hormone resuscitation and blood-pressure support belong to the donation chapter; do not “brain-protect” a dead brain with PaCO2 25.

ARDS (including neuro patients with aspiration, neurogenic pulmonary edema, or coincident pneumonia): 6 mL/kg PBW, plateau ≤30 cm H2O, PEEP/FiO2 titration, prone positioning when PaO2/FiO2 is severely reduced and ICP/CPP can be supported in the prone position (prone can raise ICP; it is not automatic in every hemicraniectomy). Neuromuscular blockade is a last-resort synchrony tool and erases the examination—add EEG if you paralyze. You will sometimes accept a PaCO2 in the high 40s if ICP allows; you will sometimes accept a plateau of 30 at 4 mL/kg if the alternative is a pneumothorax. There is no prize for a pretty ETCO2 at the cost of volutrauma.

Liberation, SBT, and dyssynchrony

Liberation is a daily question once the reason for intubation is reversing. Perform a spontaneous awakening trial when you are not in an ICP crisis, open abdomen, or refractory status protocol. Then a spontaneous breathing trial (SBT) for 30–120 minutes on low pressure support or T-piece. A rapid shallow breathing index (frequency/VT in liters) <105 is a classic pass-side number, not a sole law. Neuro-specific fail reasons: weak cough, secretion load, absent airway protection, rising ICP during the trial, or neuromuscular numbers (FVC, negative inspiratory force) still in the intubation range. A cuff-leak test helps when laryngeal edema is likely (prolonged intubation, fluid overload); a failed leak prompts steroids and delay, not a pride extubation.

Dyssynchrony raises ICP. Double-triggering stacks breaths and overdistends. Flow starvation produces a desperate inspiratory effort, a surge in venous pressure, and an ICP spike you can watch in real time. Reverse triggering (entrainment) does the same. Coughing against a closed ventilator circuit is a Valsalva. Fix the match: increase inspiratory flow, adjust trigger sensitivity, treat pain, lighten or deepen sedation to the actual goal, drain the ET tube, consider a mode change. Paralysis is last because you lose the examination. Do not “treat the ICP number” with osmotherapy every time the patient is fighting a 40 L/min flow setting that feels like breathing through a straw.

Failed SBTs in a patient who will not protect for weeks return you to the tracheostomy discussion in 4.3—without promising SETPOINT2-level functional miracles.

Exam traps

Setting tidal volume on actual body weight in obesity. Plateau 35 cm H2O ignored because “the brain needs 8 mL/kg.” Standing orders for PaCO2 25 in every TBI. Hyperventilating through the first post-injury night. Leaving a herniating patient at PaCO2 22 for 8 hours after the OR is already underway. Treating a brain-dead donor like a living high-ICP patient. Refusing all PEEP because of theoretical ICP harm while the PaO2 is 48 mm Hg. Extubating a GCS 5 patient with no cough because RSBI was 80. Ignoring ICP spikes that line up with double-triggering. Using neuromuscular blockade as first-line “ICP control” without EEG.

Worked PBW example: 165 cm woman, PBW ≈ 45.5 + 0.91 × 12.6 ≈ 57 kg; 6 mL/kg ≈ 340 mL. If she weighs 90 kg, 6 mL/kg of actual weight would be 540 mL—already a lung-injurious error before anyone touches CO2.

When an item shows ICP 40, blown pupil, and ETCO2 48, you hyperventilate briefly, give osmotherapy, raise MAP if CPP is low, and call surgery. When the same item shows ICP 16 on hospital day 2 and ETCO2 38, you do not chase 25. The difference is the indication, not your enthusiasm for vasoconstriction.