Pedro Mateu-Gelabert

There is great interest in finding meaningful subgroups of attributed network data. There are many available methods for clustering complete network. Unfortunately, much network data is collected through sampling, and therefore incomplete. Respondent-driven sampling (RDS) is a widely used method for sampling hard-to-reach human populations based on tracing links in the underlying unobserved social network. The resulting data therefore have tree structure representing a sub-sample of the network, along with many nodal attributes. In this paper, we introduce an approach to adjust mixture models for general network clustering for samplings by RDS. We apply our model to data on opioid users in New York City, and detect communities reflecting group characteristics of interest for intervention activities, including drug use patterns, social connections and other community variables

Purpose of Review The social networks of people who inject drugs (PWID) have long been studied to understand disease transmission dynamics and social influences on risky practices. We illustrate how PWID can be active agents promoting HIV, HCV, and overdose prevention. Recent Findings We assessed drug users’ connections and interactions with others at risk for HIV/HCV in three cities: New York City (NYC), USA (n = 539); Pereira, Colombia (n = 50); and St. Petersburg, Russia (n = 49). In all three cities, the majority of participants’ network members were of a similar age as themselves, yet connections across age groups were also present. In NYC, knowing any opioid user(s) older than 29 was associated with testing HCV-positive. In NYC and St. Petersburg, a large proportion of PWID engaged in intravention activities to support safer injection and overdose prevention; in Pereira, PWID injected, had sex, and interacted with other key groups at risk. Summary People who use drugs can be active players in HIV/HCV and overdose risk- reduction; their networks provide them with ample opportunities to disseminate harm reduction knowledge, strategies, and norms to others at risk. Local communities could augment prevention programming by empowering drug users to be allies in the fight against HIV and facilitating their pre-existing health-protective actions.

Background: Heroin production in Colombia has increased dramatically in recent decades, and some studies point to an increase in local heroin use since the mid-1990s. Despite this rapid increase, little is known about the effects of these activities on heroin injection within Colombia. One of the biggest concerns surrounding heroin injection is the potential spread of HIV through drug user networks. Objectives: This article examines injection risk behaviors among heroin injectors in the Colombian cities of Medellín and Pereira to explore the implications for possible increased HIV transmission within this group. Methods: A cross-sectional study used respondent-driving sampling to recruit a sample of 540 people who inject drugs (PWID) over 18 years of age (Medellín: n = 242, Pereira: n = 298). Structured interviews with each participant were conducted using the World Health Organization Drug Injection Study Phase II Survey. An HIV test was also administered. Results: Information regarding the socio-demographics, injection drug use, HIV risk and transmission behaviors, injection risk management, and HIV knowledge and prevalence of participants are reported. The study identified many young, newly initiated injectors who engage in risky injection practices. The study also found that HIV prevalence is fairly low among participants (2.7%). Conclusions/Importance: Findings indicate a potential risk for the spread of HIV among PWID in Colombia given their widespread sharing practices, high rate of new injector initiation, and unsafe syringe cleaning practices. Colombia has a possibly time-limited opportunity to prevent an HIV epidemic by implementing harm reduction interventions among young, newly initiated PWID.