A Mathematical Analysis of Centipede Game Theory in Emergency Room Interactions

This paper explores physician-nurse interactions in emergency rooms using Centipede Game Theory, a sequential decision-making model. The game illustrates how trust and institutional factors influence cooperation under high-stakes conditions. Participants alternate between cooperating, which increases cumulative rewards, and defecting, which ends the interaction for immediate personal gain. We introduce a trust-sensitive version of the model, featuring a logistic cooperation probability function, PC (t), driven by trust cap (θ), sensitivity (λ), and inflection point (t0). The expected utility function, EUC (t), captures outcomes across ten stages. Using values δ = 1.2, α = 0.6, and θ = 0.8, results show cooperation probability rising from 0.095 at stage 1 to 0.74 at stage 10, and EUC (t) increasing from 0.50 to 3.39. The findings underscore how trust and institutional support like joint training and standardized protocols can shift strategic behavior from early defection to sustained collaboration. This research demonstrates the value of game-theoretic approaches for improving teamwork and patient outcomes in critical healthcare environments.