Brief Interventions for Heavy Alcohol Users

Summary

A Cochrane review [Abstract] 1 included 14 studies with a total of 4041 mainly male participants (16 years or older) admitted to general inpatient hospital care for any reason other than specifically for alcohol treatment. Patients receiving brief interventions (1 to 3 sessions) had a greater reduction in alcohol consumption compared to those in control groups at 6 month (MD in grams per week -69.43, 95% CI -128.14 to -10.72; 4 trials, n=453) and 9 months follow up (MD in grams per week -182.88, 95% CI -360.00 to -5.76; 1 trial, n=479) but this was not maintained at one year (MD in grams per week -33.62, 95% CI -82.27 to 15.03; 4 trials, n=1073). In addition there were significantly fewer deaths in the groups receiving brief interventions than in control groups at 6 months (RR 0.42, 95% CI 0.19 to 0.94; 4 trials, n=1166) and one year follow up (RR 0.60, 95% CI 0.40 to 0.91; 7 trials, n=2396). In one high-quality trial, heavy drinking episodes were decreased at 4, 9 and 12 months (MD -0.56, 95% CI -1.02 to -0.10; MD -0.78, 95% CI -1.32 to -0.24; MD -0.71, 95% CI -1.26 to -0.16 days per week respectively; n=616).

A Cochrane review [Abstract] 2 included 69 studies with a total of 33 642 subjects. Most interventions were delivered in general practice or emergency care settings. Participants receiving brief intervention reduced their alcohol consumption compared to the control group (mean difference: -20 grams/week, 95% CI -28.36 to -11.81 g/week; 34 trials, n=15 197). Sub-group analysis confirmed the benefit of brief intervention both in men and in women. Extended intervention when compared with brief intervention had no greater reduction in alcohol consumption although findings were imprecise (MD 2 g/week, 95% CI -42 to 45; 3 studies, n=552).

Secondary outcome data from a three-arm randomized controlled trial 3 with 6-, 12-, 18- and 24-month follow-ups included 960 patients (18-64 years) with at-risk alcohol use identified through systematic screening on 13 hospital wards. Self-reported number of sick days in the past 6 months was assessed at all time-points. In comparison to assessment only, in-person counselling (OR 2.18, p = 0.047) and computer-generated written feedback (OR = 2.08, p = 0.047) resulted in statistically significant increased odds of reporting no sick days 24 months later.