Auditing the AUDIT: A systematic review of cut-off scores for the Alcohol Use Disorders Identification Test (AUDIT) in low- and middle-income countries.

INTRODUCTION: The World Health Organization's (WHO) Alcohol Use Disorders Identification Test (AUDIT) is used extensively across the world, with cut-off scores recommended by the WHO. We reviewed the use and validity of AUDIT cut-off scores in low- and middle-income countries as cultural contexts are expected to influence the detection of alcohol use disorders. MATERIALS AND METHODS: The systematic review was guided by an a priori defined protocol consistent with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement. We searched Cochrane library, Medline, EMBASE, PsycINFO, CINAHL, Indmed, LILACS, and AJOL databases using appropriate search terms. We conducted a narrative synthesis of the data. RESULTS: We identified 54 distinct studies that used AUDIT cut-off scores which were not in alignment with those recommended by the WHO. India (n = 10), Nigeria (n = 9), and Brazil (n = 9) produced most of these included studies. Most of the studies (n = 42) did not conduct psychometric evaluations of AUDIT cut-off scores. Of the twelve studies which did report psychometric results, a wide range of cut-off scores performed well. In these studies the cut-off scores to detect hazardous drinking ranged from >3 to >5, for harmful drinking from >5 to >16, and for dependent drinking from >7 to >24. DISCUSSION: AUDIT is being widely used in LMICs and non-recommended cut-off scores are considered to be appropriate in these countries. It is important to systematically evaluate the psychometric properties of those cut-off scores to ensure the internal validity of the studies in which they are used.