摘要：Background There is a renewed interest in alcohol's harm to others (AHTO), and survey studies in the general population are often used to estimate the extent of harm, to address the severity and variety of harms, and to identify the victims of such harm. While cross-sectional survey studies are attractive in several respects, they also entail several methodological challenges. Aim We discuss some of these issues, paying particular attention to the problems of causal attribution, transferability, survey data collection and range of harms. Conclusions We offer some suggestions for study design to enhance causal inferences from studies examining alcohol's harm to others.
关键词：Alcohol ; harm to others ; population surveys ; causal attribution ; study design ; data collection
