Electronic supplementary material
The online version of this article (doi:10.1186/1940-0640-9-2) contains supplementary material, which is available to authorized users.
The authors declare that they have no competing interests.
Author KAB conceived of the study, obtained funding, and oversaw all aspects of the study. Authors ADR, GTL and LC undertook the statistical analyses. Authors KED and AL managed the literature searches and wrote the first draft of the paper. All authors have contributed to and have approved the final manuscript.
The AUDIT-C is an extensively validated screen for unhealthy alcohol use (i.e. drinking above recommended limits or alcohol use disorder), which consists of three questions about alcohol consumption. AUDIT-C scores ≥4 points for men and ≥3 for women are considered positive screens based on US validation studies that compared the AUDIT-C to “gold standard” measures of unhealthy alcohol use from independent, detailed interviews. However, results of screening—positive or negative based on AUDIT-C scores—can be inconsistent with reported drinking on the AUDIT-C questions. For example, individuals can screen positive based on the AUDIT-C score while reporting drinking below US recommended limits on the same AUDIT-C. Alternatively, they can screen negative based on the AUDIT-C score while reporting drinking above US recommended limits. Such inconsistencies could complicate interpretation of screening results, but it is unclear how often they occur in practice.
This study used AUDIT-C data from respondents who reported past-year drinking on one of two national US surveys: a general population survey (N = 26,610) and a Veterans Health Administration (VA) outpatient survey (N = 467,416). Gender-stratified analyses estimated the prevalence of AUDIT-C screen results—positive or negative screens based on the AUDIT-C score—that were inconsistent with reported drinking (above or below US recommended limits) on the same AUDIT-C.
Among men who reported drinking, 13.8% and 21.1% of US general population and VA samples, respectively, had screening results based on AUDIT-C scores (positive or negative) that were inconsistent with reported drinking on the AUDIT-C questions (above or below US recommended limits). Among women who reported drinking, 18.3% and 20.7% of US general population and VA samples, respectively, had screening results that were inconsistent with reported drinking.
This study did not include an independent interview gold standard for unhealthy alcohol use and therefore cannot address how often observed inconsistencies represent false positive or negative screens.
Up to 21% of people who drink alcohol had alcohol screening results based on the AUDIT-C score that were inconsistent with reported drinking on the same AUDIT-C. This needs to be addressed when training clinicians to use the AUDIT-C.