When Should Doctors and Patients Use Shared Decision-Making Under Bounded Rationality?

Abstract

Recently, clinicians and governments have increased their advocacy for shared-decision making (SDM), a process in which doctors and patients jointly decide amongst appropriate treatment options. Even though both benefits and limitations of SDM have been documented, it is often positioned as a universal recommendation. In contrast, in this paper, we use a stylized analytical model to derive clear guidelines on when and how to employ SDM. Relative to an evidence-based medicine (EBM) approach that decides based on population averages, we first establish that doctors should always engage in SDM if both doctors and patients are perfectly rational. However, we show that EBM can outperform SDM once we account for patients’ and doctors’ bounded rationality (i.e., random errors). We find that when doctors and patients are boundedly rational, administrators should allow doctors to decide whether or not to engage in SDM (versus EBM) on a Case-by-Case (CbC) basis as long as doctors are sophisticated enough to make appropriate adjustments to account for such bounded rationality. If doctors are too overconfident (insufficiently accounting for random errors), it can be best to enforce EBM. If doctors are too underconfident (excessively accounting for random errors), it can be best to enforce SDM. More generally, we provide a set of results that map how patient population and doctor characteristics affect the relative performances of SDM, EBM, or CbC decision-making processes.

Note:

Funding Information: None to declare.

Conflict of Interests: None to declare.

Keywords: shared decision-making, behavioral operations, coproduction in healthcare, bounded rationality, overconfidence and underconfidence