As part of the Data Then and Now seminar series, Stephen Molldrem from University of California Irvine will be presenting a lecture called “HIV, Meet Data Science: On the Arrival of Big Data Methods, the Open Source Ethos, and Predictive Analytics in U.S. HIV Prevention.” Please join us for Stephen’s lecture on January 29th, 2020 from 4:00-5:00 PM in the Seminar Room of the WRF Data Science Studio.
Following confirmation that antiretroviral treatment for HIV is a highly effective way to prevent transmission, the U.S. HIV strategy has reoriented around a “treatment as prevention” paradigm. The rationales supporting treatment as prevention are widely grasped among stakeholders. However, the infrastructural and epistemological transformations that have facilitated the implementation of programs to support treatment as prevention are poorly understood. This paper describes the arrival of approaches from data science, and from the “open data” and “open source” movements, to HIV prevention in three cases. The first case focuses on how HIV data, formerly a class of “exceptionally” sensitive information with many restrictions on its exchange, were re-articulated during the 2010s. Mandates from CDC for state departments of public health to collect and use routine HIV care data in prevention have made HIV doubly-exceptional: exceptionally sensitive, but also exceptionally important to exchange, share, and use in prevention. Programs in this area often involve partnerships between health departments and correctional agencies, a concern in jurisdictions where HIV nondisclosure is criminalized. The second case considers open source tools for conducting HIV research and prevention. Software packages like HyPhy, PHYLOSCANNER, and HIV-TRACE are maintained by researchers on GitHub. They facilitate new forms of “cluster tracing” to identify people living with HIV in “transmission risk networks.” The third case describes a project in South Carolina which will use “[electronic health records], claims and data from private institutions, housing, prisons, mental health, Medicare, Medicaid, State Health Plan and the department of health and human services” in the hope of developing “machine learning” and “predictive model development” tools for use in prevention. Public health re-uses of HIV data are done without consent. I close by considering bioethical issues, recommending the development of affordances for people living with HIV to assert controls over some re-uses of their data.
The Data Then and Now seminar series explores the social and organizational history of data and data practices in order to better understand the current data-intensive moment through its antecedents and continuities. It features invited speakers from across the country and around the world. For more information, please visit the Data Then and Now web page.