Characterising communities impacted by the 2015 Indiana HIV outbreak: A big data analysis of social media messages associated with HIV and substance abuse

Abstract

Introduction and aims: Infoveillance approaches (i.e. surveillance methods using online content) that leverage big data can provide new insights about infectious disease outbreaks and substance use disorder topics. We assessed social media messages about HIV, opioid use and injection drug use in order to understand how unstructured data can prepare public health practitioners for response to future outbreaks.

Design and methods: We conducted an retrospective analysis of Twitter messages during the 2015 HIV Indiana outbreak using machine learning, statistical and geospatial analysis to examine the transition between opioid prescription drug abuse to heroin injection use and finally HIV transmission risk, and to test possible associations with disease burden and demographic variables in Indiana and Marion County. Tweets from October 2014 to June 2015 were compared to disease burden at the county level for Indiana, and classification of census blocks by presence of relevant messages was done at the census block level for Marion County. Marion County was used as it exhibited the highest total count of Tweets.

Results: 257 messages about substance abuse and HIV were significantly related to HIV rates (P < 0.001) and opioid-related hospitalisations (P = 0.037). Using 157 characteristics from the American Community Survey, a linear classifier was computed with an appreciable correlation (r = 0.49) to risk-related social media messages from Marion County.

Discussion and conclusions: Communities appear to communicate online in response to disease burden. Classification produced an accurate equation to model census block risk based on census data, allowing for high-dimensional estimation of risk for blocks with sparse populations.

Keywords: acquired immunodeficiency syndrome; ecological epidemiology; geospatial analysis; human immunodeficiency virus; social media.

Figure 1.

Observed Tweets, 2015 non-fatal opioid-related hospitalisations and new 2015 HIV cases per 100 000 population, for all Indiana counties, denoted in a choropleth gradient. Values of zero were imputed for data points redacted due to low sample sizes, and these values are illustrated in grey.

Figure 2.

Classification of 632 census blocks in Marion County, Indiana, for social media engagement with HIV/opioid/intravenous drug use themes during (and immediately pre/post) the 2015 HIV outbreak, computed using 157 demographic characteristics from the American Community Survey, with redder shades indicating higher propensity for social media engagement and points indicating observed Twitter posts on these topics.