Proposed is a prospective dynamic cohort to determine the prevalence and incidence of HIV in recently initiated adolescent and young adult injection drug users or initiates, and adolescent and young adult noninjection drug users or NlUs (heroin/cocaine sniffers, crack smokers) and to identify correlates and risk factors of HIV infection which include newly examined factors in the presence of previously identified HIV risk behaviors. For this proposal, the """"""""new"""""""" variables will include physical/sexual abuse, psychiatric condition, social networks/social support; the established variables include demographics, drug use and sex practices, and circumstances of initiation into injecting Eligible initiates must be between the ages of 15-30 and have initiated injection drug use with the past 5 years. Eligible NIUs must also be between the ages of 15-30 and have used drugs (snort heroin/cocaine or smoke crack) between 1 and 10 years prior to enrollment. The 400 young initiates and 200 young NIUs will be recruited using street outreach methodology which will involve a stationary clinic and mobile clinic for screening, interviewing, testing and follow-up in the Greater Baltimore City area. This dynamic cohort recruited and enrolled over a 5-year study period will undergo baseline, 6, and 12-month interviews along with HIV testing of all eligible participants of whom moderate remunerations will be given. Blood drawn on all participants will be screened for HIV using ELISA and Western Blot assays, and stored in a repository for later testing of other blood-borne pathogens. In addition to the HIV results, other data collected during the interview will include the above mentioned """"""""established"""""""" correlates and """"""""new"""""""" correlates (e.g., physical/sexual abuse, psychiatric condition and social network/social support). The above correlates will be compared and contrasted among initiates and NlUs. A transition rate (from noninjection to injection drug use) by the above covariates will also be estimated. The analysis will include a cross-sectional assessment of baseline data using univariate and bivariate approaches such as 2x2 and 2xK tables confirmed with Mantel-Haenszel techniques. Multivariate analysis will use logistic regression and regression tree techniques. Prospective data on HIV incidence will use person-time techniques for univariate and poisson regression for multivariate analyses Cox regression may also be employed when examining transition rates by selected covariates. We are interested in identifying and clarifying the independent and joint effects of the variables that relate to HIV infection and transition into an injecting career in this population. Our earlier initiate studies have shown the period shortly after initiation into injection drug use is associated with high risk of HIV infection, especially among young IDUs and women. In Baltimore, we are interested in continuing to understand behaviors around the time of initiation that might be modifiable through intervention, as well as the role of physical/sexual abuse, psychiatric condition, and social networks/social support. Additionally, we are complementing our research with the inclusion of NIUs of whom we will compare and contrast with our new injectors and, observe factors associated with the transition into an injecting career.
Showing the most recent 10 out of 24 publications
