Disclosure: This site earns commissions from affiliate links (Amazon, Etsy, and others) at no extra cost to you.   Full affiliate disclosure →

Updated July 2026 · 10 min read

This article was created with AI assistance.

CIWA & Alcohol Withdrawal Protocol for ICU Nurses 2026

⚕️ Medical Disclaimer: This content is for educational purposes only and is intended for licensed healthcare professionals. It does not constitute medical advice and should not replace clinical judgment, facility protocols, or physician orders. Always verify medications, doses, and procedures with your institution's guidelines.

Part of the ICU Emergencies Hub — browse every related guide in one place.

Alcohol withdrawal is one of the few things on the unit that can go from tremulous and talking to seizing and dying over a single shift. The tools to stop that are almost entirely nurse-driven: a reliable assessment, honest scoring, and medication given on the patient in front of you rather than a clock. Done well, it's quietly one of the highest-impact protocols a nurse runs.

The short version: Alcohol withdrawal is a spectrum from mild tremor to delirium tremens, driven by a brain that adapted to alcohol's GABA-boosting, glutamate-blunting effects and is now hyperexcitable without it. The CIWA-Ar scale quantifies severity across ten items, and symptom-triggered benzodiazepine dosing — medicating to the score rather than on a fixed schedule — treats withdrawal with less total drug and shorter stays. Thiamine before glucose, escalate to phenobarbital or a drip when benzos aren't holding, and watch for the escalation to DTs.

Why the brain does this

Chronic alcohol enhances GABA (the brain's main "off" signal) and suppresses glutamate (the main "on" signal). The brain compensates by turning down GABA receptors and turning up glutamate receptors to keep balance. Remove the alcohol suddenly and that compensation is exposed: a wildly overexcited, underinhibited nervous system. That's the tremor, the anxiety, the autonomic storm, the seizures. Benzodiazepines work because they restore the GABA signal the brain is now missing, which is why they're first-line — and why the treatment is really "replace what's suddenly gone, then taper."

The timeline you're watching against

PhaseOnset after last drinkWhat you see
Minor withdrawal~6–12 hrTremor, anxiety, sweating, nausea, insomnia, mild HTN/tachycardia
Withdrawal seizures~12–48 hrGeneralized tonic-clonic, often single or brief cluster
Alcoholic hallucinosis~12–24 hrHallucinations with a clear sensorium (knows they aren't real)
Delirium tremens~48–96 hrDelirium + severe autonomic instability; a medical emergency

The times overlap and vary, but the shape matters: seizures tend to come early, DTs come later, and a patient who looks manageable at hour 12 can be critically ill at hour 60. That's the reason for serial, honest CIWA scoring rather than a single snapshot.

The CIWA-Ar scale — what it actually measures

The Clinical Institute Withdrawal Assessment for Alcohol, revised (CIWA-Ar), scores ten items, most 0–7 and orientation 0–4, for a maximum of 67. The items are nausea/vomiting, tremor, paroxysmal sweats, anxiety, agitation, tactile disturbances, auditory disturbances, visual disturbances, headache, and orientation/clouding of sensorium. You rate what you observe and what the patient reports, and the total drives dosing. Rough bands most protocols use:

CIWA-Ar scoreSeverityTypical response
< 8–10Minimal / mildOften no medication; continue monitoring
10–15ModerateSymptom-triggered benzodiazepine dose
> 15–20SevereLarger/more frequent dosing, reassess closely, escalate
CIWA has real limits — know them. The scale requires a patient who can communicate. It is not valid in the intubated, sedated, non-verbal, or delirious patient, and it can be confounded by anything else that causes anxiety, tremor, or tachycardia (pain, other withdrawal, sepsis). A patient who can't reliably report symptoms needs a different tool (many ICUs use a protocol like the Minnesota Detoxification Scale/MINDS or a symptom-based ICU pathway) and clinical judgment — not a CIWA number forced onto a patient it wasn't built for.

Symptom-triggered vs fixed-dose

The central idea is symptom-triggered therapy: you give a benzodiazepine dose when the CIWA score crosses a threshold, then reassess and dose again as needed, rather than giving a set dose every few hours regardless. The evidence is consistent — symptom-triggered dosing uses less total benzodiazepine and shortens treatment duration compared with fixed schedules, because the patient gets exactly what their withdrawal requires and no more. Fixed-dose or front-loaded regimens still have a place — the patient at high risk of severe withdrawal or seizures, or a setting where frequent reliable scoring isn't feasible — but for the scorable patient, symptom-triggered is the modern default.

The common agents are lorazepam, diazepam, and chlordiazepoxide. Long-acting agents (diazepam, chlordiazepoxide) self-taper as they clear and give smoother control; lorazepam is preferred when there's significant liver disease because it has no active metabolites and simpler clearance — and it's the go-to when you need IV in the sicker patient. See the lorazepam guide for the propylene-glycol caution on big infusions.

When benzos aren't enough

Some withdrawal becomes benzodiazepine-resistant — escalating doses that don't hold the patient. As withdrawal progresses, the GABA receptors that benzos act on become less responsive, so you need agents that hit the problem differently. Phenobarbital is the workhorse escalation: it enhances GABA and directly opens the chloride channel, so it works where benzos are failing, and its long half-life gives a built-in taper — but it stacks with benzos for respiratory depression, so it demands close monitoring. Ketamine is a benzo-sparing NMDA-blocking adjunct that addresses the glutamate side benzos can't reach. Adjuncts like dexmedetomidine and propofol control the autonomic storm and agitation but don't prevent seizures, so they're add-ons, never the sole therapy. For the escalation agents, see the phenobarbital guide and the ketamine for alcohol withdrawal guide.

Delirium tremens — the emergency

DTs is withdrawal delirium plus severe autonomic instability — profound confusion and agitation, fever, drenching sweats, dangerous tachycardia and hypertension — and it carries real mortality if under-treated. It belongs in the ICU. Treatment is aggressive benzodiazepine (often IV, escalating fast) frequently with phenobarbital, airway protection when sedation demands go high, correction of the electrolyte and volume derangements, and continuous monitoring. The mindset shift from routine withdrawal is that you are now treating a life-threatening condition, and under-dosing out of fear of sedation is the more dangerous error.

The supportive care that gets forgotten

Two things save brains and lives beyond the benzos. First, thiamine before glucose: chronic alcohol use depletes thiamine, and giving a glucose load to a thiamine-deficient brain can precipitate Wernicke encephalopathy. Give thiamine first (and generously) whenever you're about to give dextrose or feed the patient — details in the thiamine guide. Second, replace magnesium, potassium, and phosphate, which are commonly depleted and both lower the seizure threshold and complicate the course; magnesium especially, and you can't fully correct potassium until magnesium is repleted (see the IV magnesium guide). Add a quiet environment, reorientation, fall precautions, and DVT prophylaxis. The pharmacology stops the storm; this supportive spine keeps it from causing lasting harm.

This article is general clinical education for nurses, not a substitute for your facility's protocol, provider orders, or clinical judgment. Always follow your institution's withdrawal pathway and dosing.

Get The ICU Notebook Newsletter

Clinical tools and career insights for ICU nurses. One email per week, no fluff.

Yes, send it free

No spam. Unsubscribe any time.