Abstract
Complexity, resulting from interactions among many components, is a characterizing property of healthcare systems and related decisions. Such complexity scales up quickly in the face of pandemics, where multiple sources of uncertainty are involved and various contextual factors interacting with policy parameters yield outcome distribution. This paper presents a uni ed framework to assist and inform policy decisions in confronting pandemics. The general framework consists of a model of contagion that makes the policy- relevant variables explicit and exogenous, establishes links between them and the main features of the environment in which the policy is going to be implemented, and treats various sources of uncertainty at different layers of the system. At the macro level, special attention is devoted to the network structure, for which we provide a simple characterization based on two constructive factors. Our results show that by conditioning on these two factors, a large proportion of the stochasticity resulted from the inherent randomness in the network can be captured. Components of the model are synthesized in a broader agent-based model that enables accounting for heterogeneous individual-level attributes that collectively yield the macro-level outcomes. Using several stylized examples and a comprehensive controlled experiment, insights on the overall tendency of the complex system in terms of multidimensional outputs are derived across a range of scenarios and under various types of policy conditions.