Predicting Alcohol Withdrawal in Medical Inpatients

Although alcohol dependence is common in medical inpatients, it often remains undetected. As a result, patients may suffer serious consequences of withdrawal during their hospital stays. Researchers in the United Kingdom assessed whether the AUDIT* and/or blood testing (GGT, AST, ALT, MCV)** could predict risk of withdrawal in 874 medical inpatients aged 16 or older.

• Of the 98 patients who scored >=8 on the AUDIT (a positive test for an alcohol use disorder), 17 experienced clinically significant withdrawal symptoms.
• All patients with withdrawal had an AUDIT score of >=8 (sensitivity 100%).
• All but 1 patient with withdrawal had abnormal blood test results.
• Most patients without withdrawal had a normal AUDIT score (specificity 91%).
• Although a positive AUDIT score plus any 2 abnormal blood tests had a sensitivity of 94% and a specificity of 98%, fewer than half of patients with this combination had withdrawal.

Comments:

Drawing firm conclusions based on a small sample is risky. Nonetheless, given that the AUDIT identifies alcohol dependence, it is not surprising that it can also predict alcohol withdrawal. However, most patients with dependence will not have significant withdrawal symptoms. Adding blood tests improves detection of those at risk of withdrawal but may predict only 50%, at best, of withdrawal cases. Therefore, alcohol screening in the hospital is mainly useful for ruling out risk of withdrawal and identifying patients who might be ready for alcohol-dependence treatment.

Richard Saitz, MD, MPH

Reference:

Dolman JM, Hawkes ND. Combining the AUDIT questionnaire and biochemical markers to assess alcohol use and risk of alcohol withdrawal in medical inpatients. Alcohol Alcohol. 2005;40(6):515–519.