Vaccination games with peer effects in a heterogeneous hospital worker population

Abstract

We develop a game theoretic model to analyze the Nash equilibrium of vaccine decisions in a hospital population with heterogeneous contacts. We use the model in conjunction with person-to-person contact data within a large university hospital. We simulate, using agent-based models, the probability of infection for various worker types in the data and use these probabilities to identify the Nash equilibrium vaccine choices of hospital workers. The analysis suggests that there may be large differences in vaccination rates among hospital worker groups. We extend the model to include peer effects within the game. The peer effects may create additional equilibria or may further cement existing equilibria depending on parameter values. Further, depending on the magnitude of the peer effects and the costs of infection and vaccination, peer effects may increase or decrease differences in worker group vaccination rates within the hospital.