A Theory of Information Biases on Healthcare Platforms

Abstract

Online healthcare has emerged as an appealing new channel for healthcare in recent years. A fundamental but unanswered question is whether the classical physician-patient agency problem in the healthcare market can be alleviated, especially with the healthcare platform as a new market player. We address the question by developing a game-theoretic model. Given that physicians want to build the reputation as a provider of accurate diagnostic information (medical reports) in the online healthcare market, we find: the private information that originates from the expertise of the physicians would induce a bias in diagnostic reports, i.e., exaggerating the likelihood of being in an unhealthy condition, and suggesting patients to do more than necessary tests. This information bias would further influence the demand for online healthcare services in two ways. First, the more skeptical the rational patients are towards the potentially biased diagnostic information, the less likely their decision would rely on the diagnostic reports generated by the physicians. Second, the information bias would make certain types of diagnosis come up more often than others. We also find that the private information gives online physicians more incentive to bias their reports if their return of career concern depends on the reputation of being providers of accurate diagnostic information. However, the healthcare platform could reduce the bias by restricting the discretion allowed to physicians. The platform’s profit will be increased if more bias is allowed. We also present a variety of empirical predictions.