A Proposal for Standardized Evaluation of Epidemiological Models ( A position paper )

Public Health (PH) officials make many decisions during both emergency and non-emergency periods: procurement, vaccine formulation, resource pre-positioning, surveillance intensity, school closure recommendations, etc. These decisions are made in an uncertain environment, and would benefit greatly from accurate estimates of the likelihood of different outcomes. For example, to decide: how much to invest in surveillance of a particular animal virus, it’s important to know whether its probability of jumping to humans over the next X years is 10%, 1% or 0.1%. whether to purchase new respirators, it’s important to know how likely the impending epidemic is to exceed the current capacity. how aggressively to use a given drug, it’s important to know how likely resistance is to emerge within, say, a year, under the different possible drug usage policies. whether to recommend school closure, it’s important to know the likely epidemic peak, timing and attack rate under the different options.