Simon Coulton, Veronica Dale, Paolo Deluca, Eilish Gilvarry, Christine Godfrey, Eileen Kaner, Ruth McGovern, Dorothy Newbury-Birch, Robert Patton, Steve Parrott, Katherine Perryman, Thomas Phillips, Jonathan Shepherd, Colin Drummond, Nick Heather

The aim of the study was to explore the relative efficiency and effectiveness of targeted versus universal screening for at-risk alcohol use in a primary care population in the UK. The study was a randomized evaluation of screening approach (targeted versus universal) for consecutive attendees at primary care aged 18 years or more. Targeted screening involved screening any patient attending with one of the targeted presentations, conditions associated with excessive alcohol consumption: mental health, gastrointestinal, hypertension, minor injuries or a new patient registration. In the universal arm of the study all presentations in the recruitment period were included. Universal screening included all patients presenting to allocated practices. A total of 3562 potential participants were approached. The odds ratio of being screen positive was higher for the targeted group versus the universal group. Yet the vast majority of those screening positive in the universal group of the study would have been missed by a targeted approach. A combination of age and gender was a more efficient approach than targeting by clinical condition or context. While screening targeted by age and gender is more efficient than universal screening, targeting by clinical condition or presentation is not. Further universal screening is more effective in identifying the full range of patients who could benefit from brief alcohol interventions, and would therefore have greater public health impact. Current Controlled Trials ISRCTN06145674.