Modelling disease dynamics and management scenarios

Abstract

Mathematical modelling now plays an important role in developing scientific understanding of complex biological processes such as epidemics. Model-based risk assessments make such studies relevant to policy makers and resource managers. However, in providing such advice it is important to ensure that model predictions are robust to alternative plausible assumptions, and also that any predictions arising from such models correctly reflect the uncertainty in current knowledge and any intrinsic variability of the system under study. To see why this is so, contrast a point estimate of the efficacy of a given disease control measure with a prediction which gives the probability associated with varying degrees of success, and crucially, failure. The former gives a false sense of confidence, whilst the latter allows the decision maker to carry out a more complete risk assessment of the proposed strategy. In all cases, model predictions should be interpreted in the light of model structure and assumptions. © Springer 2009.