The effectiveness of a brief intervention for intensive care unit patients with hazardous alcohol use: a randomized controlled trial

Abstract

Abstract Background Screening for hazardous alcohol use and performing brief interventions (BIs) are recommended to reduce alcohol-related negative health consequences. We aimed to compare the effectiveness (defined as an at least 10% absolute difference) of BI with usual care in reducing alcohol intake in intensive care unit survivors with history of hazardous alcohol use. Methods We used Alcohol Use Disorder Identification Test-Consumption (AUDIT-C) score to assess history of alcohol use. Patients: Emergency admitted adult ICU patients in three Finnish university hospitals, with an AUDIT-C score > 5 (women), or > 6 (men). We randomized consenting eligible patients to receive a BI or treatment as usual (TAU). Intervention: BI was delivered by the time of ICU discharge or shortly thereafter in the hospital ward. Controls: Control patients received TAU. Outcome: The primary outcome was self-reported alcohol consumption during the preceding week 6 and 12 months after randomization. Secondary outcomes were the change in AUDIT-C scores from baseline to 6 and 12 months, health-related quality of life, and mortality. The trial was terminated early due to slow recruitment during the pandemic. Results We randomized 234 patients to receive BI (N = 117) or TAU (N = 117). At 6 months, the median alcohol intake in the BI and TAU groups were 6.5 g (interquartile range [IQR] 0–141) and 0 g (0–72), respectively (p = 0.544). At 12 months, it was 24 g (0–146) and 0 g (0–96) in the BI and TAU groups, respectively (p = 0.157). Median change in AUDIT-C from baseline to 6 months was − 1 (− 4 to 0) and 2 (− 6 to 0), (p = 0.144) in the BI and TAU groups, and to 12 months − 3 (− 5 to − 1) and − 4 (− 7 to − 1), respectively (p = 0.187). In total, 4% (n = 5) of patients in the BI group and 11% (n = 13) of patients in the TAU group were abstinent at 6 months, and 10% (n = 12) and 15% (n = 17), respectively, at 12 months. No between-groups difference in mortality emerged. Conclusion As underpowered, our study cannot reject or confirm the hypothesis that a single BI early after critical illness is effective in reducing the amount of alcohol consumed compared to TAU. However, a considerable number in both groups reduced their alcohol consumption. Trial registration: ClinicalTrials.gov (NCT03047577).