Article type
Year
Abstract
Objective: To assess the effectiveness of brief interventions to reduce alcohol consumption.
Methods: Meta-analysis of randomised control trials located through MEDLINE & PSYCHLIT searches, manually checking reference lists, and consultation with clinical experts. Setting: Primary care. Outpatient, Inpatient care in 8 countries. Subjects: Patients screened to have hazardous or harmful alcohol consumption. Main outcome measures: Reduction in weekly alcohol consumption, within group changes, relationship between initial consumption and outcome.
Results: The difference in within group changes in consumption between the treatment and control groups was calculated, and a weighted average calculated using meta-analysis to estimate the overall pooled treatment effect. The pooled estimate of the treatment effect was 6.75 UK units of alcohol per week (95% CI, 4.27 to 9.22) which represents a 22.7% reduction in alcohol consumption. A weighted regression of treatment effect against average baseline values suggests that the size of the treatment effect may be greater in patients with higher levels of consumption.
Discussion: Brief intervention including assessment and provision of simple information and advice, may be an effective way to reduce alcohol consumption in nonalcoholic heavy drinkers. The relatively small treatment difference for women is largely a result of the large within group improvement among those receiving assessment only (34.7% for brief intervention, 28.0% for assessment only).
Methods: Meta-analysis of randomised control trials located through MEDLINE & PSYCHLIT searches, manually checking reference lists, and consultation with clinical experts. Setting: Primary care. Outpatient, Inpatient care in 8 countries. Subjects: Patients screened to have hazardous or harmful alcohol consumption. Main outcome measures: Reduction in weekly alcohol consumption, within group changes, relationship between initial consumption and outcome.
Results: The difference in within group changes in consumption between the treatment and control groups was calculated, and a weighted average calculated using meta-analysis to estimate the overall pooled treatment effect. The pooled estimate of the treatment effect was 6.75 UK units of alcohol per week (95% CI, 4.27 to 9.22) which represents a 22.7% reduction in alcohol consumption. A weighted regression of treatment effect against average baseline values suggests that the size of the treatment effect may be greater in patients with higher levels of consumption.
Discussion: Brief intervention including assessment and provision of simple information and advice, may be an effective way to reduce alcohol consumption in nonalcoholic heavy drinkers. The relatively small treatment difference for women is largely a result of the large within group improvement among those receiving assessment only (34.7% for brief intervention, 28.0% for assessment only).