Imperatives for health sector decision-support modelling

Like many other critical infrastructures, the public health systems of most countries have grown to high complexity and broad interdependence. Complex adaptive systems of this kind have been shown to evolve into a precarious state between order and disorder in which remote triggers can lead to cascading failure. The healthcare stresses of ageing populations, emerging infectious diseases and treatment-resistant pathogens continue to grow. However, the economic drive for leaner systems and supply chains, and their increasing interdependencies (power, water, food, fuel etc.), all contribute to the growing fragility of public health. When systems are stressed (during emergencies or disasters), accountability for effective resource use requires that health decision making switch from optimal care of individuals to a resource-conserving mode designed to maximise outcomes for the population. Understanding when and how to operate under such conditions, demands that decision makers have a strong grasp of system capacities, limitations and the beneficial or deleterious effects of parameters they are able to manipulate. These difficulties lie behind a growing effort to model health systems to inform planning and (more aspirationally) support real-time decision making. However, modelling efforts to date have been less than optimally founded and structured. Our purpose in this paper is to reframe conceptual approaches to the strategic modelling of health systems and their dependencies in order to point the way to a more productive path for health sector decision-support modelling.