Alcohol Withdrawal

Rahul Ganatra, MD, MPH

Scott Navarett, MD

You have just admitted a 63-year-old man with a history of coronary artery disease, diabetes, and alcohol use disorder for acute pancreatitis. He was previously drinking a “handle” (1.75 L) of vodka daily, but due to his current abdominal pain, has not consumed any alcohol in the past 72 hours. Review of the chart reveals multiple prior hospitalizations for alcohol withdrawal, including one requiring ICU admission. Since his arrival to the floor, the patient has received multiple doses of diazepam totaling approximately 200 mg in the last 24 hours, but remains tremulous and diaphoretic, and you worry his alcohol withdrawal could worsen significantly.

What features of a patient’s history and current clinical presentation predict development of severe withdrawal?

Higher reported maximum 24-hour alcohol intake and number of prior withdrawal episodes are associated with severe withdrawal syndromes such as delirium tremens (DT) and/or seizures.

In a 1995 case-control study,¹ investigators used a large database of 1648 patients with DSM-III-R alcohol dependence enrolled at six academic medical centers across the United States to identify predictors of severe withdrawal. Among them, 12.8% (160 men, 51 women) reported the primary outcome of having had at least one episode of DT and/or seizures during an episode of withdrawal in the past.

Compared with alcohol-dependent patients who had never experienced an episode of DT and/or seizures, patients who had experienced these outcomes reported a higher number of drinks in any 24-hour period (OR = 1.02 per each additional drink, 95% CI 1.01-1.03; P < .001). Patients who had ever experienced an episode of DT and/or seizures also reported a higher number of antecedent withdrawal episodes (OR = 1.01 per each additional withdrawal episode, 95% CI 1.00-1.02; P < .01). Patients with a history of DT and/or seizures also reported more concomitant abuse of other sedative-hypnotic drugs (P < .05) and more chronic medical problems related to alcohol use (P < .001) compared with patients without a history of DT and/or seizures. Caveats in the interpretation of this study include its case-control design and reliance on self-report of both the exposures and outcome.

You recognize several high-risk features in his presentation, including his multiple prior episodes of withdrawal, high daily alcohol intake, and chronic medical problems. As you prepare to sign out, you inform the covering nocturnist that the patient is at high risk for complications of alcohol withdrawal. Overnight, the patient’s condition worsens, and he requires transfer to the ICU for DT.

The following day, another new patient on your service, a 45-year-old woman with lower extremity cellulitis and a history of alcohol use disorder begins to develop signs and symptoms of alcohol withdrawal syndrome. Based on his experience caring for your previous patient who was sent to the ICU, the nurse asks if starting a scheduled benzodiazepine might be better for this patient than a symptom-triggered approach.

How does a fixed-dose approach compare with a symptom-triggered approach in treating alcohol withdrawal syndrome?

Administration of benzodiazepines via a symptom-triggered approach is the preferred treatment strategy over scheduled benzodiazepines for most patients because it leads to lower cumulative benzodiazepine doses and shorter hospital stays.

In a 1994 double-blind randomized controlled trial,² 101 patients admitted to an inpatient detoxification facility received chlordiazepoxide either on a fixed-dose schedule or as needed, based on a symptom-triggered administration schedule (the Clinical Institute Withdrawal Assessment for Alcohol, revised, or CIWA-Ar). Patients meeting DSM-III-R criteria for alcohol abuse or dependence admitted to a single Veterans Affairs Alcohol Detoxification Unit for management of acute withdrawal were included in the trial. Patients were excluded if they had any history of seizures or if they were already taking medications that could affect the clinical course of withdrawal, including opiates, benzodiazepines, barbiturates, clonidine, or β-blockers. All patients were reassessed by CIWA-Ar at baseline, every 8 hours, and every hour after benzodiazepines were administered. The primary outcomes were the duration of treatment and the total cumulative dose of benzodiazepines administered. Secondary outcomes included severity of withdrawal as measured by the CIWA-Ar score; against medical advice (AMA) discharges; development of hallucinosis, seizures, or DT; degree of general discomfort and alcohol craving; and rates of readmission and compliance with follow-up.

Treatment by a symptom-triggered approach was associated with a shorter duration of treatment (median 9 vs. 68 hours; P < .001) and a lower cumulative benzodiazepine dose (100 vs. 425 mg chlordiazepoxide; P < .001). There were no differences between groups in withdrawal severity, AMA discharges, episodes of seizures, or the development of DT.

Only gold members can continue reading. Log In or Register to continue