Changes in risk behaviour following a network peer education intervention for HIV prevention among male Tajik migrants who inject drugs in Moscow: a cluster-randomized controlled trial.

ABSTRACT

INTRODUCTION: The "Migrants' Approached Self-Learning Intervention in HIV/AIDS for Tajiks" (MASLIHAT) recruits and trains Tajik labour migrants who inject drugs as peer educators (PEs) in delivering HIV prevention information and encouragement to adopt risk-reduction norms and practices within their diaspora social networks while reducing their own HIV risk. METHODS: The MASLIHAT intervention was tested in Moscow in a cluster-randomized controlled trial with 12 recruitment sites assigned to either the MASLIHAT intervention or an equal-time peer-educator training focused on other health conditions (TANSIHAT). From October 2021 to April 2022, 140 male Tajik migrants who inject drugs were recruited as PEs to attend the 5-session MASLIHAT training or the TANSIHAT non-HIV comparison condition. Each participant in both groups recruited two network members (NMs) who inject drugs with the intent to share with them the information and positive strategies for change they had learned (n = 280). All PEs and NMs (n = 420) participated in baseline and follow-up interviews at 3-month intervals for 1 year. All received HIV counselling and testing. Modified mixed effects Poisson regressions tested for group differences in injection practices, sexual risk behaviours and heavy alcohol use over time. RESULTS: At baseline, across both groups, 75% of participants reported receptive syringe sharing (RSS), 42% reported condomless sex and 20% reported binge drinking at least once a month. In contrast to TANSIHAT where HIV risk behaviours remained the same, significant intervention effects that were sustained over the 12 months were observed for receptive syringe and ancillary equipment sharing among both MASLIHAT PEs and NMs (p < 0.0001). Significant declines in the prevalence of sexual risk behaviours were also associated with the MASLIHAT intervention (p < 0.01), but not the comparison condition. Binge alcohol use was not affected in either condition; the MASLIHAT intervention had a transitory effect on drinking frequency that dissipated after 9 months. CONCLUSIONS: The MASLIHAT peer-education intervention proved highly effective in reducing HIV-related injection risk behaviour, and moderately effective in reducing sexual risk behaviour among both PEs and NMs. Network-based peer education is an important tool for HIV prevention among people who inject drugs, especially in environments that are not amenable to community-based harm reduction.