Evaluation of spill-over effects of antiretroviral treatment on sharing injecting equipment in a sociometric network of people who inject drugs.

The evidence on risk compensation due to antiretroviral treatment (ART) use for people, who inject drugs (PWID), is limited. Moreover, interventions, including ART administration, frequently have spill-over effects, i.e., individuals who were not exposed themselves possibly benefit from their connections to those who were exposed. This work aimed to evaluate the causal effects in the presence of spill-over of ART uptake on sharing injecting equipment. Data were obtained from the Transmission Reduction Intervention Project, a network-based intervention among PWID in 2013-2015 in Athens, Greece. An inverse probability weighted estimator (IPWE) was used to assess spill-over effects. The analysis included 218 individuals, with 17% of them being on ART, corresponding to 32% of those living with HIV. The adjusted estimated risk differences (RD) between those on ART and those without ART (direct effects) ranged from 0.105 to 0.155 and were not statistically significant. Comparing participants, who did not receive ART across different levels of ART exposure among their first-degree contacts in the network (spill-over effects), the estimated RDs indicated a protective but not statistically significant spill-over effect of ART. The analyses of PWID-related network data, using a novel estimator, detected no statistically significant effects of ART on sharing injection equipment.