Two related projects are proposed whose goal it is to evaluate the influence of recruitment methods on studies of populations at risk for HIV. In both projects secondary data analysis will be conducted from multiple study populations in the Seattle area to identify and evaluate differences in sociodemographic characteristics and HIV-related risk and prevention behaviors as a function of recruitment method. One project will focus on injection drug users (IDU), the other on men reporting male-to-male sex (MSM). For the IDU component, four sources of data on Seattle area IDU will be joined together: the second cycle of the National HIV Behavior Surveillance survey of IDU (NHBS-IDU2) (conducted in 2009), NHBS-IDU1 (2005), the Risk Activity Variables, Epidemiology, and Network study (RAVEN) (conducted 1994-1997) and Kiwi (1998-2002). Comparison will also be made with persons diagnosed with HIV and reported to the HIV/AIDS Reporting System (HARS) with injection drug use as a transmission category from 2005 through 2009. The MSM component will join data from the Seattle area NHBS-MSM2 study (2008), data from the Public Health - Seattle &King County STD clinic (2003-2008), two random digit dial surveys ( in 2003 and 2006), and HARS data on MSM from 2003 though 2008. In both the IDU and MSM components, a collection of variables will be synthesized describing sociodemographic characteristics, drug use, sexual behavior and HIV preventive measures that can be evaluated in a comparable manner across the different data sources. These variables will be compared across the different study populations and evaluate the statistical significance of differences. In the IDU component multivariate analyses will be conducted using syringe sharing, a key risk behavior for HIV transmission, as the outcome variable in order to assess whether differences in the frequency of syringe sharing among the different study populations can be accounted for by differing sociodemographic composition and by trends over time. In the MSM component similar multivariate analyses will be conducted on unprotected (i.e. without use of a condom) anal intercourse with a partner of unknown or opposite HIV status and on having an HIV test within the 12 months previous to the survey. The results of these analyses are intended to help guide the choice of recruitment methods for surveying IDU and MSM populations and aid in the interpretation of data collected using these methods.
Surveys of persons at risk for HIV infection are important in providing information to characterize and monitor HIV-related risk and preventive behaviors to inform planning and evaluation of programs to reduce HIV transmission. However, obtaining a representative sample of a population at risk for HIV is problematic, as the different methods generally used for survey recruitment each have well-recognized potential biases. We are proposing two related projects to help evaluate the influence of recruitment methods on studies of populations at risk for HIV. In both projects we will conduct secondary data analysis from multiple study populations in the Seattle area to identify and evaluate differences in sociodemographic characteristics and HIV-related risk behaviors as a function of recruitment method. One project will focus on injection drug users (IDU), the other on men reporting male-to-male sex (MSM). We are in the unique position of already having multiple local sources of data available for such analyses. The results of these projects are intended to help guide the choice of recruitment methods for surveying these populations and aid in the interpretation of data collected using existing methods.
