How much can data on days with heavy drinking decrease the underestimation of true alcohol consumption?

An adjusted quantity-frequency method, with questions on occasions with heavy drinking, was used to estimate the consumption of alcohol during the last 30 days. The purpose was to analyze if it was possible to decrease the underestimation of true alcohol consumption. The questionnaire was mailed to a randomized sample of 1,500 individuals, 20-75 years of age, living in the city of Malmö, Sweden; 930 persons (64.3%) participated. Data on alcohol consumption were validated by comparison to sales of alcohol for the city of Malmö. The estimated per capita consumption of alcohol in the population was equivalent to 77.0% of the registered sale of alcohol in Malmö. By adding days with heavy drinking, the estimated weekly per capita consumption of alcohol among the alcohol consumers increased from 74.5 grams to 77.1 grams (+3.5%; p < .001). Of the alcohol consumers, 15.1% increased their reported consumption. In order to decrease even more the underestimation of the true alcohol consumption, we suggest the use of questions about any alcohol consumption that deviates from the typical consumption of each individual.