MEASUREMENT AND PREDICTION OF AIDS RISK IN IV DRUG USERS

点击翻译按钮获取中文摘要

Research on behavior related to transmission of HIV involves collecting highly sensitive information about sexual activity and drug use. As yet little is known about how best to gather such information. The proposed study will assess methods for increasing validity of self-reports of high- risk IV drug-related and sexual behaviors. Using a 2x2x2 factorial design, 200 heroin detoxification clients and 200 "street users" of heroin will be randomly assigned to one of the eight interview schedules to test which interview formats (closed-ended vs. open-ended questions, long vs. short questions, questions with standard vs. familiar wording) minimize underreporting of high-risk behavior. These formats were chosen on the basis of literature suggesting that they may have substantial impact on levels of reporting where the content of questions is threatening or where socially desirable responses are expected. We will also assess the extent to which socially desirability, question threat, and sexual self-disclosure affect self-reported high-risk behavior. We will attempt to validate self- reports of needle-sharing with forensic tests of syringes furnished by IV drug-using subjects. We will compare the responses of a subsample of 50 IV drug using couples to questions about high-risk sexual practices, and randomly assign subjects to male and female interviewers to determine the extent of subject by interviewer sex interactions. As a check on self- reported condom use, subjects will be offered condoms at research interviews with records kept of the number of condoms taken and subject's plans for using them. The proposed study additionally will develop and evaluate measures of attitudes and beliefs predicting high-risk behavior. The framework for this effort is the AIDS Risk Reduction Model, which incorporates aspects of the Health Belief Model, self-efficacy theory, interpersonal influences, emotions such as anxiety, decision-making theory, and help-seeking models. The measure will be used to estimate regression models of high-risk behavior to guide preventive efforts with IV drug users.