Injection drug use continues to contribute to new incidence of HIV and remains the primary risk factor for hepatitis C virus in the United States. Accordingly, understanding the social processes associated with injection risk behavior remains an important goal for public health research in effort to inform interventions to reduce the frequency these behaviors. However, previous research has largely focused on who is most likely to engage in injection risk behavior rather than examining when, where, and with whom individuals may be at heighted risk. The current study uses event specific data from the Sexual Acquisition and Transmission of HIV Cooperative Agreement to examine dyadic, network, and situational characteristics associated with injection risk behavior. Data on multiple observations nested within participants (participant n = 784, injection episodes n = 1778) is used to examine both within and between person variation in injection risk behavior via multilevel structural equation modeling. Results are interpreted using Tseng and Seidman's (2007) theory of social settings. Results indicated that injection risk behavior was lower when injecting with new partners. While having an injection partner that is also a sexual partner was associated with greater risk for both males and females, sexual partnership was significantly more positively associated with injection risk for females as compared to males. Furthermore, females were at greater risk when injecting with other females but the gender of their injection partner was not associated with any difference in risk among males. For network characteristics, the number of injectors in the participant's network was not significantly associated with risk behavior. Finally, for situational characteristics no significant relationship with injection risk behavior was found for the location of the injection episode (e.g., if the participant injected at home) but injection risk behavior was higher when more non-injectors were present during the injection episode. These results suggest that differences in social norms or resource availability may create unique risk factors for female injectors as compared to males. Future studies could provide further insight by explicitly measuring mediating social setting variables such as the availability and control of injection resources (e.g., syringes or drugs) as well as setting level norms. Furthermore, the results indicate that intervention and evaluations studies should continue to develop HIV/HCV preventive interventions tailored toward sexual partners and explore the potential for gender specific programming. While the current study provides initial insight into a more complex view of injection risk behavior and associated dyadic, network, and situational variables, significant within-person variability persisted after including all model variables. This suggests that additional dyadic and situational characteristics must be identified to better predict this unexplained variation in injection risk behavior across injection episodes. Accordingly, future work is required to develop a more thorough understanding of social setting mechanisms that may enhance protective behaviors and inhibit risk behavior within injection settings.
Read
