UC San Diego

Background: Brazil has the largest crack market in the world. The drug’s low cost, local poverty, and inequities made the country especially vulnerable to the crack epidemic. Structurally vulnerable people who use crack (PWUC) usually consume, buy, and sell drugs in open public spaces, the “open drug scenes” (ODS), locally called “cracolândias” (cracklands). In these settings, PWUC are subject to contextual risks that may sustain or enhance PWUC vulnerabilities, gender inequities, and the acquisition and spread of infections. The Brazilian prevalence of HIV and HCV is significantly higher among PWUC, especially among women. In this dissertation, we used the syndemic framework to explore the syndemic of crack use, HIV, and HCV in Brazil and the role of gender and the open drug scenes.Methods: This dissertation analyzed two levels of data from the parent study, the Brazilian Crack Survey: (i) PWUC survey, containing HIV and HCV tests and socioeconomic characteristics, and (ii) ethnographic field notes with descriptions of the ODS where PWUC were reached. In Chapter 2, we used Generalized Linear Models to explore the potential syndemic interaction between crack use, HIV, and HCV in promoting exacerbated adverse health consequences and the underlying gender differences among the 7,381 PWUC respondents. In Chapter 3, we analyzed the ODS ethnographic field notes through Structural Topic Modeling to unveil their main characteristics (N=3,787). In Chapter 4, we combined both datasets, constructing a latent syndemic risk factor (PWUC survey) and using it to guide a supervised Latent Dirichlet Allocation model to find the ODS features (ethnographic field notes) most associated with the syndemic of crack use, HIV, and HCV. Results: Chapter 2 findings provided epidemiological evidence for the syndemic of crack use, HIV, and HCV in Brazil among PWUC, unveiling gender differences. Women presented a higher cluster of health conditions and a broader spectrum of socioeconomic risk factors associated with and exacerbating the health condition compared to men. Chapter 3 showed that the Brazilian ODS share a common background of vulnerabilities and deprivation. The ODS, however, vary in complexity and dynamics, depending on their location, whether rural, central, or outskirts areas. The police actions also differ, and so does their impact on PWUC behaviors. Chapter 4 findings suggested that “Cracklands” (crowded spaces with PWUC heavily using crack) were associated with the highest syndemic risk. ODS where PWUC share paraphernalia and act on a Cat-and-mouse dynamic with the police, were associated with homelessness and displayed a middle syndemic risk. Other scenes characterized by Drug trade, Lesser deprivation, Low traffic of people, Express drug use, or Hidden use are associated with the Low-middle syndemic risk. Conclusions: These studies emphasize the significance of understanding the environment and tailoring interventions to address the syndemic of crack use, HIV, and HCV at both individual and structural levels. Instead of treating health conditions in isolation and disregarding the intricate contexts in which they manifest, these interventions should recognize the comprehensive range of syndemic vulnerabilities, enhancing the effectiveness of prevention and health care practices. Such an approach is crucial for effectively reducing harm, mitigating stigma, and alleviating the burden on this vulnerable population.