RESUMO
AIM: To examine the efficacy and safety of antiepileptic drugs (AED) for the inpatient treatment of patients with moderate to severe alcohol withdrawal syndrome (AWS). METHODS: We searched in databases and gray literature to include randomized controlled clinical trials in adults that compare the use of AED versus placebo or any other medication. Studies that did not specify severity or were performed on an outpatient basis were excluded. The outcomes were improvement of symptoms, delirium tremens, seizures and adverse events. Two researchers independently selected the references, extracted the data and assessed the risk of bias. A qualitative synthesis was made and, when the heterogeneity was mild or moderate, a meta-analysis was performed. The quality of the evidence obtained was evaluated with the Grading of Recommendations Assessment, Development and Evaluation tool. RESULTS: In total, 26 studies with 1709 patients were included. No benefit is described in withdrawal syndrome severity measured by scales or in the development of delirium tremens. The only possible meta-analysis showed that there are no differences in the incidence of seizures (risk ratio [RR] = 1.0; confidence interval (CI) 95% 0.76-1.33), even when compared with placebo (RR = 0.95; CI95% 0.57-1.57). There were also no differences in adverse events, although a higher proportion of some mild cases were described with the use of carbamazepine and valproic acid. CONCLUSIONS: The routine use of AED is not suggested in the treatment of patients with moderate or severe AWS unless indicated for comorbidity (weak recommendation against using moderate quality of evidence).