What Makes Health Data Privacy Calculus Unique? Separating Probability from Impact

Abstract

Patient health data is heavily regulated and sensitive. Patients will sometimes falsify data to avoid embarrassment resulting in misdiagnoses and even death. Existing research to explain this phenomenon is scarce with little more than attitudes and intents modeled. Similarly, health data disclosure research has only applied existing theories with additional constructs for the healthcare context. We argue that health data has a fundamentally different cost/benefit calculus than the non-health contexts of traditional privacy research. By separating the probability of disclosure risks and benefits from the impact of that disclosure, it is easier to understand and interpret health data disclosure. In a study of 1590 patients disclosing health information electronically, we find that the benefits of disclosure are more difficult to conceptualize than the impact of the risk. We validate this using both a stated and objective (mouse tracking) measure of patient lying.