Decision Making under Compliance Uncertainty

Abstract

In many decision making problems in healthcare, decision makers are different from those who perform them, and compliance with optimal solution is often a concern; optimal policy is not followed and it results in suboptimal outcomes. Based on a conjecture that compliance with optimal decisions is influenced by preferences (or biases) of the decision performer, we formulate a sequential decision making problem that accounts for decision performers’ preferences. A policy solution from this formulation is sub-optimal to the original decision problem, but yields better outcome under less than full compliance environment. Characteristics of the policy solutions are illustrated with examples.