METHODS FOR IMPROVING MEASUREMENT OF ALCOHOL CONSUMPTION

点击翻译按钮获取中文摘要

DESCRIPTION: This research is designed to improve the reliability and validity of alcohol consumption and problem measures. It directly contributes to the Center?s charter of gathering and analyzing data in ways that advance methods, and will enhance the Center?s research on the epidemiology of drinking problems in the general population and its service populations. The approach taken here recognizes and addresses the two broad categories of drinking consequences: drinking problems due to hazardous heavy-drinking episodes (as with drunk driving and injuries); and those that result from long-term harmful volume levels (as with liver disease and dependence). Three methodological studies are proposed. The first "Studies of Self-Defined Drink Size", tests and implements innovative methodologies to elicit drink size from telephone survey respondents, hypothesizing that average daily volume as well as rate of heavy drinking will be greater when respondent-defined, beverage-specific drink sizes are considered. Findings have implications for alcohol epidemiology (by potentially accounting in surveys for greater proportion of aggregate consumption based on alcohol sales data; more accurate mapping of the distribution of consumption, by ethnicity, gender, region; etc.), for risk assessment (as with pregnant drinkers, especially those drinking higher alcohol-content beverages in larger-than-standard sizes), and for health services (where, for example, methodologies under study here may be used to help at-risk drinkers become aware of their larger drink sizes). The second study, "Interactive Voice Response (IVR)", examines the value and applicability to alcohol surveys of IVR technology in which respondents interact with a computer (using their telephone key pad) to answer questions about sensitive behaviors such as risky sex, drinking and driving, heavy drinking, etc. Higher prevalence rates for alcohol consumption, dependence, and related risk behaviors are hypothesized for IVR-administered items than when telephone survey interviewers are used. The third study determines the validity of self-report drinking compared to breathanalysis-based Blood Alcohol Levels (BAL) across different contexts and cultures including bars in the U.S. as well as emergency room (ER) settings in the U.S. and five other countries. Analysis of variables which influence biases in self-reported consumption in both ER and bar drinking samples will increase our knowledge about ways that event-specific self-report alcohol measurement might be improved. Each of these three studies interacts with and informs other Center components and addresses its major themes.