Digital Epidemiology: Use of Digital Data Collected for Non-epidemiological Purposes in Epidemiological Studies

Park, Hyeoun-Ae; Jung, Hyesil; On, Jeongah; Park, Seul Ki; Kang, Hannah

We identified four study purpose categories: description (22.9%), exploration (34.9%), explanation (27.5%), and prediction and control (14.7%). Descriptive studies focus on presenting phenomena with incidence, size, and/or other measurable attributes. An example of a descriptive study is a study by Moon et al. [17] , which investigated internet search rates for vitamin D using GT data and reported seasonal variation in public interest in vitamin D. An explorative study investigates the full nature of the phenomenon, the manner in which it manifests, and other related factors through correlation analyses. An example of such a study is Chary et al. [18] , which demonstrated a strong correlation between the geographic variation in social media posts mentioning prescription opioid misuse and government estimates of misuse of prescription opioids. An explanatory study focuses on understanding the underlying causes of a phenomenon or the systematic relationships among variables [13] . An example is Towers et al. [19] , which examined whether news coverage was a significant factor in the temporal patterns of Ebola-related Twitter data using a linear regression model. A predictive and controlled study attempts to predict a phenomenon based on research findings. An example is Ram et al. [20] , which analyzed Twitter, Google [22] . An example of a case-control study is a study by Tourassi et al. [15] , which classified study subjects into a case group (females for whom cancer is the stated or inferred cause of death) or a control group (females for whom there was no mention of cancer) using online obituary announcements in order to examine the association between parity and cancer risk. An example of an intervention study is Edoh's study [16] which conducted an experiment using large range/distance temperature sensors and drones in order to collect infectious diseases-related data from study participants.