Modeling patients' illness perception and equilibrium analysis of their doctor shopping behavior

Abstract

When a patient's illness perception is inconsistent with a doctor's diagnosis, she may seek opinions from multiple doctors without referrals, a behavior called doctor shopping. In this study, we model and derive patients’ optimal doctor shopping decisions. After each visit, patients update their beliefs about their health status following the Bayes’ rule. We show that the patients’ doctor shopping decisions are critically affected by the diagnosis accuracy, the relative value of identifying a severely ill patient, and the cost per visit. We examine how the patients’ doctor shopping behavior affects social welfare from two aspects, namely, an objective one that accesses whether doctor shopping improves the judgment accuracy regarding the patient's health status, and a subjective one concerning whether doctor shopping relieves patients’ anxiety. We find that allowing patients to conduct doctor shopping exacerbates the system congestion (congestion effect), but it can help those patients who have decided to join obtain a higher reward (reward effect). There exists a diagnosis accuracy threshold above which allowing doctor shopping incurs a welfare loss and below which allowing doctor shopping improves welfare. Moreover, this diagnosis accuracy threshold increases as patients become more pessimistic or hold more diverse initial illness perceptions. The objective welfare maximization prefers a higher doctor shopping rate than the subjective welfare maximization does only when the value of identifying a severely ill patient is high enough, which may help explain why doctor shopping is encouraged for the critical illness such as cancer.