Delirium and Withdrawal in a Ventilated Patient With Alcohol Use Disorder
A 58-year-old man is day 4 in the ICU after an open reduction and internal fixation of a femur fracture sustained in a fall. He has a history of heavy alcohol use (approximately 12 drinks per day, last drink the day of admission) and smokes. He is intubated for respiratory failure from pneumonia. He is receiving midazolam 4 mg/h and fentanyl 50 mcg/h for sedation. Overnight he becomes agitated, diaphoretic, and tremulous when sedation is lightened, with a heart rate of 128, blood pressure 168/96, temperature 38.3°C, and a respiratory rate of 30. The Richmond Agitation-Sedation Scale fluctuates between −1 and +2. He is pulling at lines. Sodium is 132, platelets 88,000, and the wound is clean.
1. Construct a differential for his agitation and autonomic instability, and explain how you would distinguish among the leading possibilities.
Read the worked answer
Differential.
- Alcohol withdrawal syndrome, including progression to delirium tremens - time course (day 4, last drink at admission), autonomic hyperactivity (HR 128, BP 168/96, fever, diaphoresis, tremor), and fluctuating agitation fit well. DT typically emerges 48–96 hours after the last drink.
- Delirium from critical illness - multifactorial: infection, medications, sleep disruption, pain, immobility. The fluctuating RASS and inattention are consistent, but the marked autonomic instability favors withdrawal as a major driver.
- Sepsis - pneumonia is present; fever and tachycardia could reflect worsening infection. Wound is clean but the chest source is active.
- Medication effects - midazolam can cause paradoxical agitation and accumulates with prolonged infusion; fentanyl can contribute.
- Hypoglycemia, electrolyte disturbance, hepatic encephalopathy - sodium 132 is mild; check glucose, ammonia, and liver function.
- Pain and urinary retention - common, easily missed.
Distinguishing them.
- Time course and pattern: withdrawal has a characteristic escalation over days with autonomic signs tracking agitation; sepsis tends to show fever with other signs of infection (purulent secretions, worsening oxygenation, rising white count).
- Response to treatment: if benzodiazepine administration produces calm and reduces autonomic signs, withdrawal is likely the dominant driver.
- Workup: blood gas, lactate, glucose, electrolytes, ammonia, liver panel, chest imaging, sputum and blood cultures, urinalysis, and review of all medications. Examine the wound and lines.
- Delirium screening: use a validated tool (CAM-ICU) once sedation is light enough; document fluctuation and inattention.
The two leading possibilities - withdrawal and sepsis - are not mutually exclusive and may coexist.
2. The team wants to increase midazolam to control agitation. Evaluate this plan and propose an alternative sedation strategy.
Read the worked answer
Evaluation. Increasing midazolam is the wrong move for several reasons:
- Benzodiazepine tolerance and accumulation: in a heavy alcohol user, GABA-A receptor sensitivity is altered; escalating midazolam may produce disinhibition and paradoxical agitation rather than calm, while its active metabolites accumulate, especially with any hepatic dysfunction, prolonging delirium.
- It masks rather than treats: if sepsis is contributing, deeper sedation delays recognition.
- It confounds assessment: deeper sedation prevents delirium screening and prolongs ventilation.
- Withdrawal specifically requires benzodiazepines, but the right approach is symptom-triggered dosing with a defined scale, not an open-ended infusion increase.
Alternative strategy.
- Treat withdrawal with a protocolized approach: symptom-triggered benzodiazepines using a validated scale (e.g., CIWA-Ar or a withdrawal-specific ICU tool), with a scheduled taper for patients at high risk of DT. Consider phenobarbital in refractory cases per institutional protocol, with attention to respiratory depression.
- Adjuncts: thiamine before glucose, folate, magnesium replacement, and correction of electrolytes. Avoid relying on antipsychotics alone, which lower the seizure threshold and do not treat autonomic instability.
- Optimize analgesia: multimodal analgesia; avoid escalating fentanyl as a sedative.
- Change the sedative backbone: if ongoing sedation is needed for ventilation, use a non-benzodiazepine agent (e.g., propofol) with daily interruption, while recognizing that propofol does not treat withdrawal.
- Non-pharmacologic delirium prevention: sleep hygiene, daytime mobilization, orientation, glasses and hearing aids, family presence, and minimizing tethers.
- Treat the pneumonia and reassess for other delirium precipitants.
The judgment is to separate the two problems - withdrawal needs benzodiazepines titrated to a scale, while the rest of the sedation plan should minimize benzodiazepine exposure.
3. His agitation escalates despite protocolized benzodiazepines, and he requires increasing oxygen. What is your next step, and what safety considerations guide it?
Read the worked answer
Next step. Escalating oxygen requirement with worsening agitation should prompt reassessment for a new or worsening process - particularly aspiration, progression of pneumonia, or pulmonary edema - before attributing everything to withdrawal. Obtain a chest radiograph and blood gas, review ventilator mechanics, and check for a missed source (urinary, line, wound, abdomen).
If withdrawal remains refractory after adequate protocolized benzodiazepines, options include:
- Phenobarbital as an adjunct, with careful monitoring for respiratory depression - particularly relevant here given his worsening respiratory status.
- Dexmedetomidine as an adjunct for autonomic hyperactivity, recognizing it does not prevent seizures and can cause bradycardia and hypotension.
- Escalation of airway support if agitation threatens oxygenation - reintubation or increased support may be needed.
Safety considerations.
- Sedatives that suppress respiratory drive are hazardous in a patient with worsening oxygenation; the priority is to secure the airway and identify the cause before deepening sedation.
- Antipsychotics should not be used as monotherapy for withdrawal because they lower the seizure threshold and do not treat autonomic instability.
- Reassess for aspiration risk and consider whether enteral feeding should be held.
- Involve the surgical team if the femur wound or hardware is a possible source.
- Document the reasoning and communicate with nursing and pharmacy, since this is a coordination problem as much as a pharmacologic one.