On Wednesday, May 7, the Federal Trade Commission (FTC) held the third of its Spring Seminars on emerging consumer privacy issues. This session focused on consumer-generated health information (CHI). CHI is data generated by consumers’ use of the Internet and mobile apps that relates to an individual’s health. The “H” in CHI defies easy definition but likely includes, at minimum, data generated from internet or mobile app activity related to seeking information about specific conditions, disease/ medical condition management tools, support and shared experiences through online communities or tools for tracking diet, exercise or other lifestyle data.
In the United States, many consumers (mistakenly) believe that all of their health-related information is protected, at the federal level, by the Health Information Portability and Accountability Act (HIPAA). HIPAA does offer broad privacy protections to health-related information, but only to identifiable health information received by or on behalf of a “covered entity” or a third party working for a covered entity. Covered entities are, essentially, health plans and health care providers who engage in reimbursement transactions with health plans (referred to as “Protected Health Information” or “PHI”). When HIPAA was enacted in 1996, PHI was the primary type of health information, but CHI, which is generally not also PHI, has changed that. As FTC Commissioner Julie Brill noted her in her opening remarks, CHI is “health data stored outside the HIPAA silo.”
Without the limitations imposed by HIPAA, online service providers and mobile apps generally (except where state law requires differently) can treat CHI like other digital non-health data that they collect from consumers. As a result, the FTC expressed concerned that CHI may be aggregated, shared and linked in ways that consumers did not foresee and may not understand.
The panelists at the FTC discussed the difficulty in defining CHI, and whether and how it is different from other kinds of data collected from consumers. One panelist noted that whether a consumer considers his or her CHI sensitive is highly individualized. For example, are the heart rate and exercise data collected by mobile fitness apps sensitive? Would the answer to this question change if these data points were linked with other data points that began to suggest other health or wellness indicators, just as weight? Would the answer change if that linked data was used to predict socioeconomic status that is often linked to certain health, wellness and lifestyle indicators or used to inform risk rating or direct to consumer targeted advertising?
Panelists also discussed the larger and more general question of how to define privacy in a digital economy and how to balance privacy with the recognized benefits of data aggregation and data sharing. These questions are compounded by the difficulty of describing data as being anonymized or de-identified – foundational principles in most privacy frameworks – because the quality of being “identifiable” in the digital economy may depend on the proximity of a piece of data to other pieces of data.
Though the “how” and “what” of additional [...]