Social and structural determinants of health associated with police violence victimization: A latent class analysis of female sex workers who use drugs in Kazakhstan.

ABSTRACT

INTRODUCTION: Punitive legal environments remain a challenge to HIV prevention efforts in Central Asia, and female sex workers who use drugs are vulnerable to police violence. Little is known about the heterogeneity of police violence against female sex workers who use drugs and factors associated with HIV risk in Central Asia, despite the growing HIV epidemic. METHODS: We recruited a community-based sample of 255 female sex workers who use drugs in Almaty, Kazakhstan between February 2015 and May 2017. We used latent class analysis to differentiate women into distinct classes of police violence victimization, and multinomial logistic regression to identify individual-level health outcomes, HIV risk behaviors, and social and structural factors within the risk environment associated with class membership. RESULTS: A three-class model emerged Low Victimization (51%), Discrimination and Extortion (15%), and Poly-Victimization (34%). Relative to Low Victimization, factors associated with Poly-Victimization included being positive for HIV and/or sexually-transmitted infections (STI) (aOR 1.78 (95% CI 1.01, 3.14)), prior tuberculosis diagnosis (2.73 (1.15, 6.50)), injection drug use (IDU) (2.00 (1.12, 3.58)), greater number of unsafe IDU behaviors (1.21 (1.08, 1.35)), homelessness (1.92 (1.06, 3.48)), greater drug use (1.22 (1.07, 1.39)) and sex work stigma (1.23 (1.06, 1.43)), greater number of sex work clients (2.40 (1.33, 4.31)), working for a boss/pimp (2.74 (1.16, 6.50)), client violence (2.99 (1.65, 5.42)), economic incentives for condomless sex (2.77 (1.42, 5.41)), accessing needle/syringe exchange programs (3.47 (1.42, 8.50)), recent arrest (2.99 (1.36, 6.55)) and detention (2.93 (1.62, 5.30)), and negative police perceptions (8.28 (4.20, 16.3)). Compared to Low Violence, Discrimination and Extortion was associated with lower odds of experiencing intimate partner violence (aOR= 0.26 (0.12, 0.59)), but no other significant associations with the risk environment upon adjusting for socio-demographic characteristics. CONCLUSION: Police violence against female sex workers who use drugs is pervasive in Kazakhstan. Patterns of police violence vary, with greater HIV susceptibility associated with a higher probability of experiencing multiple forms of police violence. Police sensitization workshops that integrate policing and harm reduction, and drug policy reforms that decriminalize drug use may help mitigate the HIV epidemic in Kazakhstan.