6.3 ICU Prophylaxis, Delirium, and Early Mobility (01.L)

The ABPN Content Specifications list ICU prophylaxis as seven leaves: prevention of volutrauma, DVT prevention, prevention of gastroduodenal ulcers, pressure ulcers, CAUTI and VAP, early mobility, and delirium. This section covers all of them. Independent OpenExamPrep teaching here is not an SCCM or ABIM product. Details of ARDS ventilator modes live in the mechanical-ventilation section; the prophylaxis point is that a brain injury does not authorize a 12 mL/kg tidal volume.

Prevention of volutrauma

Volutrauma is alveolar overdistention from excessive tidal volume. The ARDS Network ARMA trial (NEJM 2000) compared 6 mL/kg predicted body weight (PBW) with a plateau-pressure limit of ≤30 cm H2O against 12 mL/kg. Hospital mortality was about 31% versus 40%. That result still governs how you set a volume-control breath in 2026, including in the neuro ICU, unless a brief, documented exception exists (severe uncompensated acidosis, a few minutes of herniation-bridge hyperventilation). PBW comes from height and sex, not from the scale: men 50 + 0.91 × (height in cm − 152.4); women 45.5 + 0.91 × (height in cm − 152.4). A short woman with a 90 kg anasarca weight still gets a ~6 mL/kg PBW breath, not 600 mL because she "looks large."

Neuro-specific caveats do not repeal the 6 mL/kg rule. You may accept a higher PaCO2 if ICP is quiet and you are protecting the lung, or you may briefly lower PaCO2 as a bridge in herniation (avoid PaCO2 <25 mmHg). Those are CO2 and ICP decisions. They are not permission to set 10–12 mL/kg as the day-shift default. Driving pressure and PEEP belong in the pulmonary chapter; the prophylaxis leaf is stop stretching the lung.

DVT prevention and the neuro timing problem

Immobility, hemiparesis, craniotomy, and brain-tissue-factor release make the neuro ICU a venous thromboembolism (VTE) factory. Intermittent pneumatic compression (IPC) goes on at admission when the legs are intact. Graduated compression stockings alone did not prevent VTE in CLOTS-1 and are not a substitute for IPC after ICH (AHA/ASA 2022 ICH guidance treats stockings as not beneficial for this purpose). CLOTS-3 showed IPC reduces DVT after stroke. Pharmacologic prophylaxis is more effective than machines alone once bleeding risk allows it.

After ICH, AHA/ASA 2022 material: start IPC on the day of diagnosis (Class 1). Low-dose unfractionated heparin (UFH) or low-molecular-weight heparin (LMWH) as an adjunct is reasonable around 24–48 hours when follow-up imaging shows a stable hematoma, coagulation is acceptable, and blood pressure is controlled (Class 2b). Neurocritical Care Society (NCS) 2016 VTE guidance similarly suggests prophylactic UFH or LMWH within 48 hours of admission when the hematoma is stable. Do not start enoxaparin in the CT scanner bay while the hematoma is still blooming. Do not wait 10 days in a hemiplegic patient whose 24-hour CT is unchanged.

After TBI, NCS 2016 recommends IPC within 24 hours of presentation (or within 24 hours after craniotomy) and LMWH or UFH within 24–48 hours when hemorrhage is stable, or 24 hours after craniotomy. Brain Trauma Foundation language is less clock-specific and says start pharmacologic prophylaxis when the brain injury is stable and benefit outweighs bleed risk. The examination-friendly synthesis: machines immediately, drug after a stable scan, usually inside 48 hours, sooner after an uncomplicated craniotomy if the surgeon agrees. Polytrauma long-bone fractures raise VTE risk further; delaying heparin "because of the head CT" for a week is how pulmonary emboli happen.

After ischemic stroke, start pharmacologic prophylaxis when the patient is immobile and there is no active hemorrhagic transformation. After intravenous thrombolysis, delay chemoprophylaxis about 24 hours (and usually after a stability scan). IPC can start immediately. After endovascular thrombectomy without lytic, many units start prophylaxis the same day if the infarct is not hemorrhagic; follow local post-lytics timing if alteplase or tenecteplase was given.

After aneurysmal SAH, IPC on arrival. Pharmacologic prophylaxis is commonly started once the aneurysm is secured; NCS language supports UFH. Unsecured aneurysms plus full-dose anticoagulation is a different, uglier problem than prophylactic subcutaneous heparin after clipping or coiling.

Inferior vena cava filters are for acute proximal DVT or PE with a contraindication to therapeutic anticoagulation, as a retrievable bridge — not prophylactic decoration in every ICH patient.

Gastroduodenal ulcer prophylaxis: who actually needs a PPI

Critical illness can produce stress-related mucosal injury. Stress ulcer prophylaxis (SUP) usually means a proton-pump inhibitor (PPI) (or an H2-receptor antagonist). The mistake is a PPI on every neuro-ICU admission order set.

Classic high-risk features, from Cook and subsequent ICU literature, are mechanical ventilation longer than 48 hours and coagulopathy (platelet count under about 50 × 10^9/L, INR over about 1.5, or PTT more than twice control). Shock, high-dose corticosteroids, and a prior significant GI bleed also raise risk. TBI historically appears on high-risk lists. A talking, extubated, eating patient with a small, stable ICH and normal coagulation does not automatically need pantoprazole for a week.

SUP-ICU (Krag and colleagues, NEJM 2018) randomized 3,298 ICU patients at risk for GI bleeding to pantoprazole 40 mg daily versus placebo. Clinically important GI bleeding was less common with pantoprazole (about 2.5% versus 4.2%). 90-day mortality did not differ (about 31% in both arms). So a PPI can prevent some bleeds in at-risk patients; it is not a mortality drug, and it is not free. Costs include Clostridioides difficile, a possible pneumonia signal in older literature, hypomagnesemia, and CYP2C19 interactions (clopidogrel is the cardiology example). Stop the PPI when risk factors resolve and the patient is eating.

Enteral nutrition itself is mucosal protection. Do not treat a PPI as a reason you can ignore feeding.

Pressure ulcers

The occiput, heels, sacrum, and device sites (collar, endotracheal-tube ties, EEG leads, cooling pads, traction) ulcerate in immobile neuro patients. Prevention is scheduled repositioning (often every 2 hours if the spine and ICP allow), a pressure-redistributing mattress, heel offloading, dry skin without maceration, and enough protein that the wound can heal. The Braden scale risk-stratifies; it does not replace turning. Unstable ICP or an unstable cervical spine changes how you turn (log roll, two-person hold), not whether the skin is allowed to die. Document device-related pressure and loosen EEG collodion and cervical collars on a schedule. A stage-3 occipital ulcer after two weeks of "keep the head still for the EVD" is a prophylaxis failure, not fate.

CAUTI and VAP bundles

Catheter-associated urinary tract infection (CAUTI) prevention is mostly not placing a Foley, or removing it at the first moment hourly urine output is no longer required. Aseptic insertion, a closed drainage system, the bag below the bladder, and a daily "does this catheter still have an indication?" question are the bundle. Indications that survive rounds: accurate hourly output in shock or acute kidney injury, urinary retention that cannot be intermittent-catheterized, and selected pelvic or urologic injuries. "The patient is sleepy" is not an indication. Culture the urine when there is a clinical infection story, not because the bag looks cloudy on day 12 of an unnecessary catheter.

Ventilator-associated pneumonia (VAP) prevention is a bundle, not a single mouthwash. Elements that still matter: head-of-bed 30–45 degrees (unless a spine or procedure forbids it), daily sedation interruption and a spontaneous breathing trial when safe for ICP and airway, oral care, subglottic suction endotracheal tubes when available, cuff pressure about 20–30 cm H2O, and minimizing circuit breaks. Chlorhexidine oral care has a more mixed modern evidence story than the original IHI bundle implied; do not treat a brand of rinse as the whole bundle. The same 6 mL/kg lung-protective ventilation that prevents volutrauma is part of not shredding lung and then calling the infection inevitable.

Early mobility

Early mobility means sitting, standing, and marching in place as soon as shock, ICP, and the spine allow — including in many intubated patients. The ABCDEF bundle (Assess and treat pain, Both SAT and SBT, Choice of sedation, Delirium, Early mobility, Family) is how ICUs operationalize this. The 2018 PADIS guideline (Pain, Agitation/sedation, Delirium, Immobility, Sleep) and the 2025 SCCM PADIS focused update both push enhanced mobilization over "usual" bed rest. Hold mobility for active herniation, an unstable fracture that has not been cleared, femoral intra-aortic balloon or ECMO cannulas that the cannulating service has not cleared, and profound undifferentiated shock. Do not hold mobility because the patient "might pull the EVD" if you can supervise the walk. Immobility causes DVT, pressure ulcers, delirium, and ventilator days — the other leaves on this list.

Delirium: PADIS, not a standing haldol order

Delirium is an acute disturbance of attention and awareness. Hyperactive, hypoactive, and mixed forms all count. Screen with CAM-ICU or the Intensive Care Delirium Screening Checklist (ICDSC). Hypoactive delirium in a quiet TBI patient is still delirium.

PADIS 2018 (Devlin and colleagues, Critical Care Medicine) is the framework the examination still expects you to speak:

• Treat pain first (behavioral pain scale or CPOT in the ventilated patient). Untreated pain is agitation; more midazolam is not analgesia.
• Use the lightest sedation that meets the goal. Avoid benzodiazepines when you can; they are deliriogenic. Exceptions include alcohol withdrawal, benzodiazepine-responsive seizures, and some deep ICP or status pathways.
• Nonpharmacologic care is first-line prevention and treatment: reorientation, visible clocks, glasses, hearing aids, night-day lighting, earplugs as tolerated, family presence, and early mobility.
• Do not routinely give antipsychotics to prevent delirium or to "clear" hypoactive delirium. MIND-USA (Girard and colleagues, NEJM 2018) found that haloperidol and ziprasidone did not improve delirium-free days versus placebo. AID-ICU (Olsen and colleagues, NEJM 2022) found intravenous haloperidol did not improve days alive and out of hospital. The 2025 PADIS focused update could not issue a recommendation for or against antipsychotics for ICU delirium treatment. A small dose of an antipsychotic for dangerous agitation, after pain and the ventilator are addressed, is a safety maneuver — not evidence-based disease modification.
• Dexmedetomidine is the drug PADIS 2018 suggested for delirium with agitation that precludes weaning or extubation. The 2025 update additionally suggests dexmedetomidine over propofol when light sedation and delirium reduction are the priorities (watch bradycardia). Dexmedetomidine is not burst-suppression therapy and not an antipsychotic substitute for a screaming, untreated bladder spasm.

Find the cause: hypoxia, hypercarbia, hyponatremia, infection, urinary retention, restraint, withdrawal, new focal ischemia, nonconvulsive seizures. Treating "ICU psychosis" with scheduled quetiapine while an EVD clogs is not PADIS care.

Prophylaxis by neuro diagnosis

Worked bedside scenarios

A 5-foot-tall woman with ICH is ventilated at 550 mL because "she has a stiff brain." Calculate PBW from height. 550 mL is volutrauma unless her PBW is near 90 kg, which it is not.

A man with a 20 mL putaminal ICH has a stable 24-hour CT, normal coagulation, and a hemiplegic leg. IPC should already be on. Starting prophylactic LMWH now is consistent with AHA 2022 / NCS-style timing. Waiting until day 10 is not conservative — it is neglected VTE prevention.

An extubated, eating SAH patient is still on pantoprazole from the admission order set with no coagulopathy and no prior GI bleed. Stop it.

A ventilated patient screens CAM-ICU positive and is quietly inattentive. Scheduled intravenous haloperidol is not supported by MIND-USA or AID-ICU as a way to shorten delirium. Fix sleep, glasses, pain, urea, and the Foley.

Exam traps

12 mL/kg ventilation "for ICP." Chemoprophylaxis through an expanding ICH. IPC omitted because "we will start Lovenox later." PPI for every admission. Foley for convenience. Calling VAP prevention a chlorhexidine brand. Restraining a patient instead of walking them. Treating hypoactive delirium with an antipsychotic as if MIND-USA had been positive. Using dexmedetomidine to burst-suppress. Claiming this guide is approved by ABIM or SCCM — it is independent teaching covering listed prophylaxis topics.