Sixty-five studies (n=100,980) were included: 33 studies were set in EDs and 32 were set in hospital wards. Fifty-six independent samples from 51 studies were included in the main analysis (reasons were given for the exclusion of some eligible studies from the analysis).
For studies set in the ED, screening using BAC tools produced significantly higher prevalence rates of positive screens than self-report: 26.1% (12 studies) versus 15.6% (19 studies); the OR was 1.92 (95% CI: 1.84, 2.00, p<0.001).
For self-report studies, prevalence rates of positive screens were similar for ward and ED studies: 16.5% (29 studies) and 15.6% (19 studies); the OR was 1.07 (95% CI: 1.03, 1.11, p<0.001).
Higher prevalence rates for positive tests were found for males in both ED and ward settings, regardless of the screening tool used. Prevalence rates of positive screens were 18.5% for men (6 studies) versus 8.7% for women (6 studies) when using BAC in the ED setting (OR 2.37, 95% CI: 1.91, 2.94, p<0.001) and 18.2% (9 studies) versus 6.8% (10 studies) when using self-report (OR 3.07, 95% CI: 2.72, 3.46, p<0.001). In the ward, the rates were 24.7% for men (24 studies) versus 7.1% for women (24 studies) when using self-report (OR 4.30, 95% CI: 4.00, 4.62, p<0.001).
The relationship between age and prevalence rates of positive screens was not linear. Findings suggested a trend towards a higher prevalence in ED patients aged 20 to 40 years and in ward patients aged 30 to 50 years.