BACKGROUND AND AIM: Pharmacotherapy for alcohol use disorder (AUD) is recommendable, but under-used, possibly due to deficient knowledge of medications. This study aimed to investigate the real-world effectiveness of approved pharmacological treatments (disulfiram, acamprosate, naltrexone and nalmefene) of AUD.

DESIGN: A nation-wide, register-based cohort study.
SETTING: Sweden.
PARTICIPANTS: All residents aged 16-64 years living in Sweden with registered first-time treatment contact due to AUD from July 2006 to December 2016 (n = 125 556, 62.5% men) were identified from nation-wide registers.
MEASUREMENTS: The main outcome was hospitalization due to AUD. The secondary outcomes were hospitalization due to any cause, alcohol-related somatic causes, as well as work disability (sickness absence or disability pension), and death. Mortality was analysed with between-individual analysis using a traditional multivariate-adjusted Cox hazards regression model. Recurrent outcomes, such as hospitalization-based events and work disability, were analysed with within-individual analyses to eliminate selection bias.

FINDINGS: Naltrexone combined with acamprosate [hazard ratio (HR) = 0.74; 95% confidence interval (CI) = 0.61-0.89], combined with disulfiram (HR = 0.76, 95% CI = 0.60-0.96) and as monotherapy (HR = 0.89, 95% CI = 0.81-0.97) was associated with a significantly lower risk of AUD-hospitalization compared with no use of AUD medication. Similar results were found for risk of hospitalization due to any cause. Benzodiazepine use and acamprosate monotherapy were associated with an increased risk of AUD-hospitalization (HR = 1.18, 95% CI = 1.14-1.22 and HR = 1.10, 95% CI = 1.04-1.17, respectively). No statistically significant effects were found for work disability or mortality.

CONCLUSIONS: Naltrexone as monotherapy and when combined with disulfiram and acamprosate appears to be associated with lower risk of hospitalization due to any and alcohol-related causes, compared with no use of alcohol use disorder (AUD) medication. Acamprosate monotherapy and benzodiazepine use appear to be associated with increased risk of AUD-associated hospitalization.