Behaviours, networks, and sequences: HIV epidemics among people who inject drugs in Pakistan

Loading...
Thumbnail Image
Date
2020-05-07
Authors
Thompson, Laura Hilary
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The estimated overall HIV prevalence among people who inject drugs (PWID) in Pakistan is 21%, with significant geographic heterogeneity. This study explored whether disparate types of evidence about HIV epidemics among PWID in five different cities across Pakistan may be triangulated using a social ecological framework to achieve new insights into what is shaping local epidemics. Evidence that capture HIV risk factors at the level of the 1) characteristics and behaviours of PWID, 2) social organization of PWID, and 3) broader macrosocial context was included. Questionnaire data and dried blood spots were collected from 1,439 PWID in 2014/2015. HIV testing and sequencing of the HIV pol gene was conducted. Networks connecting participants who had named the same injection locations were constructed. Network-level exposures, k-core, and degree centrality were calculated, and dot density maps of injection locations were generated. Phylogenetic data derived from HIV sequences were used to identify molecular clusters. Demographic, behavioural, and network variables were tested for associations with HIV infection. In Hyderabad, use of professional injectors, network position, and network exposures to others who use professional injectors were associated with HIV. In Karachi, injecting at places with a greater proportion of other people who were HIV positive was associated with HIV. In Larkana, injection by a professional injector was protective against HIV, and it is believed that HIV prevention programs may have engaged professional injectors in this city. In the highly dense network context of Peshawar, with scarce HIV prevention services, molecular results revealed a large cluster of 52 individuals who likely became infected just prior to sampling. Although the HIV prevalence among PWID was relatively low in Quetta, use of a professional injector was common and being connected to a network component with a large proportion of other people who received help injecting was associated with HIV infection. The results of this study indicated that generally high levels of risk behaviours and poor access to HIV prevention programs. Local network structures and the practices of professional injectors within networks influence individual risk and shape HIV epidemics locally. The configuration and relevance of city-specific factors were likely highly influenced by the local social economic and security environments.
Description
Keywords
HIV, epidemiology, social network analysis, phylogenetic analysis, social ecological model, Pakistan, people who inject drugs
Citation