A complex systems approach for understanding the effect of policy and management interventions on health system performance

Abstract

While many authors in the areas of public health and medicine have encouraged use of complex systems methods for investigating pressing issues such as health system design, few applied examples exist demonstrating potential benefits, limitations, and outcomes of such exercises. In response, we present here the results of an agent-based model, constructed with input from health system managers and designed to reflect the operation of a large-scale road traffic injury insurance, compensation, and rehabilitation system. By analyzing the in silico results of nine separate policy scenarios, we demonstrate the utility of this approach to provide health system managers with insight into the potential effect of interventions on health system performance. Modeling showed that despite the risk of observed performance decrements in the period immediately post-implementation, strategies focused on early intervention and primary prevention were ultimately more effective in improving modeled health system performance than those focused on securing short-term financial viability of the system. Implications for the use of agent-based models in health systems research are discussed.