Online media and structural influences on new HIV/STI Cases in the US (2006-2020)

University Of Illinois At Urbana-Champaign, Urbana IL

Investigators

Linked publications, trials & patents

Abstract

? DESCRIPTION (provided by applicant): Social and online media are inextricably linked to HIV/STI transmission behaviors and risks by motivating risk behaviors and social contacts in ways not available before the age of Twitter and online hookups. These online social influences are also shaped by and interact with structural biological and social vulnerabilities, including network HIV/STI prevalence as well as the economic and social capital of the population. In this application, we propose an investigation into this complex and difficult problem by modeling the influence of online communications (e.g., Twitter, selected dating/hookup sites linguistically analyzed) and regularly updated, accessible online structural (societal and environmental) data that can influence HIV/STI. The project will provide understanding and prediction of how new HIV and STI diagnoses are affected by temporally and geographically localized online communication and structural factors, using a theoretical model in which linguistic contents interact with social and biological vulnerabilities in an environment (for theory based, top- down analyses), as well as bottom-up, exploratory, data-driven analyses. Cutting-edge cross-sectional and longitudinal analysis of county and zip-code level data are proposed to understand online social communications and structural factors relevant to HIV/STI. Nine models will refine and assess the generalizability of findings with the most readily available but coarse data (i.e., singe year surveillance reported by county or longitudinal surveillance reported yearly by county), and their validity when compared with the more detailed longitudinal models (surveillance reported quarterly in Philadelphia zip codes). The analyses will entail 3 cross-sectional and 6 longitudinal tests predicting new HIV/STI diagnoses in Philadelphia zip codes, US counties, and the zip codes of the most populated US counties. The project will provide understanding and prediction of how new HIV and STI diagnoses are affected by temporally and geographically localized online communication and structural factors, using a theoretical model in which linguistic contents interact with social and biological vulnerabilities in an environment (for theory based, top- down analyses), as well as bottom-up, exploratory, data driven analyses. With a degree of geographic and temporal precision not feasible or cost effective with prior methods, the proposed models should spatially identify future HIV/STI hotspots in real time (at the level of zip code and linked to references to specific venues or venue types), and in relation to references to specific groups (e.g., MSM [Men who have Sex with Men]). This tracking should in turn pinpoint times, spaces and social groups in urgent need of prevention and care mobilization in the immediate future. All tools will then be used to develop HIVRadar, a tool that health departments will be able to use to perform similar tracking and prediction.