How should prevalence of alcohol use disorders be assessed globally?

Alcohol is a major risk factor for global burden of disease, and alcohol use disorders make up a considerable portion of this burden. Up to now, prevalence of alcohol use disorders has been estimated based on general population surveys with the Composite International Diagnostic Interview (CIDI) as the gold standard for assessment. However, three major problems have been identified with the current conceptualization of alcohol use disorders and its measurement via CIDI: cultural specificity of key criteria measured such as loss of control; lack of convergence of diagnoses identified by CIDI with clinically relevant diagnoses in primary health care; and impact of stigma on measurement. As a solution, it is proposed to measure alcohol use disorders via heavy drinking over time, with thresholds taken from the European Medicines Agency (60 and more grams on average per day of pure alcohol for men, and 40+ grams for women). Current data on level of drinking (per capita consumption) assessed via taxation and other means allow for a measure of less bias. If these thresholds are considered too low and there is more emphasis on need for specialized treatment, then thresholds for very heavy drinking can be taken as alternatively (100+, and 60+ grams per day pure alcohol for men and women, respectively).