Datafication in Public Health Surveillance : Making Authoritative Accounts

Abstract: The data traces we increasingly leave behind through interactions with information technology are being integrated into public health practice for continuous real-time monitoring and decision-making in a process of datafication. While previous research suggests there are challenges in producing and moving representations, datafication relies on reusing data primarily intended for a different purpose. This raises the question of how data is made to work in new contexts with datafication. Hence, this thesis unpacks how datafication is accomplished and how it becomes an authoritative account in public health practice.The dissertation is based on a qualitative study of two different syndromic surveillance systems based on datafication in the context of public health surveillance, with data generated from citizens’ health information seeking via a website and a phone medical advice line. The study follows the development and use of the syndromic surveillance systems, primarily based on participant observation of the developers who work with the software for analysing and visualizing illness using both types of data. The study is also based on participant observation during weekly meetings of public health practitioners, for whom the two surveillance systems factor into national influenza surveillance.Grounded primarily in Isabelle Stengers’ reading of Alfred North Whitehead, this dissertation contributes to the understanding of how data-driven phenomena play out in practice, in particular on datafication in the domain of public health. It offers a conceptualization of how comparison is accomplished with datafication, by the establishment of a datafication rapport. Drawing on valuation studies, the study shows how value is attributed to datafication in multiple registers. It also highlights the mutability of valuations and the care involved in making datafication work.Datafication rapport is further articulated in relation to public health surveillance with two concepts. Firstly, patchwork data narration conceptualizes how comparability is accomplished in practice with datafication, in the face of challenges such as ambiguities in data provenance, and in how data is accessed and generated. Secondly, asymmetric de/serialization highlights the dynamics in the relationship to the data provider and that datafication is an accomplishment which unfolds over time.