Evaluation of a Symptom-triggered Protocol for Alcohol Withdrawal for Use in the Emergency Department, General Medical Wards, and Intensive Care Unit.

BACKGROUND: Alcohol withdrawal is common in hospitalized patients and symptom-triggered guidelines have been shown to reduce treatment duration, length of stay, and need for mechanical ventilation.

OBJECTIVES: To assess the feasibility of incorporating symptom-triggered alcohol withdrawal guidelines early in the hospital course and to evaluate outcomes of patients before and after implementation of the guidelines.

METHODS: This was a retrospective pre-post study of adult patients admitted from the emergency department to an urban, academic, tertiary care center. Subjects in the preguideline (PRE) group were given benzodiazepines in a nonprotocolized manner at the discretion of the treating physician, whereas subjects in the postguideline (POST) group were treated according to the alcohol withdrawal guidelines with treatment beginning in the emergency department.

RESULTS: The PRE group involved 113 admissions for severe alcohol withdrawal and the POST group involved 103 admissions for severe alcohol withdrawal. The median benzodiazepine dose per day, in milligrams of chlordiazepoxide, was higher in the POST group (100 mg in the PRE group vs. 141 mg in the POST group; P<0.02). A higher percentage of patients in the POST group were admitted to the intensive care unit (4.4% in the PRE group vs. 12.6% in the POST group; P=0.05); however, more patients in the PRE group than in the POST group received continuous intravenous sedation and mechanical ventilation, although the difference was not statistically significant (P=0.37 for both variables). There was no difference between the 2 groups in length of stay in the intensive care unit or hospital or discharge disposition.

CONCLUSIONS: Incorporating symptom-triggered guidelines for alcohol withdrawal early in the hospital course at a large medical center is feasible. This approach may result in increased benzodiazepine use, but it seems that it is safe and does not result in adverse outcomes.