Privacy, Data Mining, and Digital Profiling in Online Patient Narratives

Practices of health datafication and inadequate privacy policies are redefining the meaning of online patient narratives. This article compares patient-driven illness narratives and clinic-driven illness narratives to uncover a set of unrecognized assumptions about trust and privacy in health discourses. Specifically, I show how the open sharing of patient stories in social media, blogs, and other public domains collides with privacy regulations and normative assumptions in the US health care system that prevent integration of those stories into electronic health record (EHR) systems. I argue that publicly told stories based on personal experiences of illness are valuable sources of health care information in part because they are subjective, richly detailed, and open ended. Yet, precisely because of their public nature, these patient stories are unprotected sources of data that are barred from integration into health care data ecologies where clinical action takes place. Consequently, an impermeable barrier exists between the officially sanctioned accounts in the clinical record and the contextual richness of patient stories on the social web. The tensions between these two approaches to narrative and data create an opening for exploitative digital profiling practices that can ─ and already do ─ harm patients. Examples are drawn from Hugo Campos and Medtronic, PatientsLikeMe, Apple Health Records, Google Health, Microsoft Health Vault, IBM Watson Health, and OpenNotes.