Citation

Ballesteros J, Duffy JC, Querejeta I, Ariño J, González-Pinto A. Alcohol Clin. Exp. Res. 2004; 28(4): 608-618.

Affiliation

Universidad del País Vasco, Facultad de Medicina, Departamento de Neurociencias, Psicología Médica y Psiquiatría, Leioa, Spain. onpbaroj@lg.ehu.es

Copyright

(Copyright © 2004, John Wiley and Sons)

DOI

unavailable

PMID

15100612

Abstract

BACKGROUND: Because recent research in primary care has challenged the findings of previous reviews on the efficacy of brief interventions (BIs) on hazardous drinkers, we conducted a systematic review and meta-analysis to update the evidence of BIs as applied in the primary care setting. METHODS: We obtained source material by searching electronic databases and reference lists and hand-searching journals. We selected randomized trials providing frequency data that allowed assessment of the efficacy of BIs on an intention-to-treat basis. Results were summarized by the odds ratio (OR) of response. When appropriate, risk difference (RD) and its inverse (number needed to treat [NNT] to achieve a positive result) were also computed. Fixed and/or random effect models were fitted according to heterogeneity estimates. RESULTS: Thirteen studies provided data for a dose-effect analysis, 12 for comparison of BIs with reference categories. No clear evidence of a dose-effect relationship was found. BIs outperformed minimal interventions and usual care (random effects model OR = 1.55, 95% confidence interval [CI] = 1.27-1.90; RD = 0.11, 95% CI = 0.06-0.16; NNT = 10, 95% CI = 7-17). Similar results were obtained when two influential studies were removed (fixed effect model OR = 1.57, 95% CI = 1.32-1.87; RD = 0.11, 95% CI = 0.07-0.15; NNT = 9, 95% CI = 7-15). The heterogeneity between individual estimates was accounted for by the type of hazardous drinkers (heavy versus moderate) and by the characteristics of the included individuals (treatment seekers versus nontreatment seekers). The funnel plot did not show evidence of publication bias. CONCLUSION: Our results, although indicating smaller effect sizes than previous meta-analyses, do support the moderate efficacy of BIs. Further research is outlined.

Language: en