Modeling and Simulations for Infectious Diseases

摘要

Modeling and simulations of infectious diseases are useful tools to estimate epidemic size and determine effective intervention methods. Models are idealized, simplified representations of reality. Therefore, a simple model should be the first design, with complex models always able to be added later to better represent reality. The following recommendations are made to ensure accurate simulations: (1) minimal feasible complexity, (2) mathematically well-defined, models should be constructed with well characterized stochastic processes rather than 'ad-hoc'behavioral models, (3) if predictions are being made, unknown parameters need to be robustly estimated using the model from epidemiological data, (4) where data is unavailable, a detailed sensitivity analysis needs to be undertaken, and (5) even if models are only being used to explore dynamics, sensitivity analysis is important to understand mechanisms. For the effective control of infectious diseases with models and simulations, we need effective information systems (laboratory and clinical data) for communicable disease control as a permanent infrastructure, closer collaboration between academic and service providers on new developments in communicable disease control, and enhanced monitoring of veterinary infections and 'minor' zoonoses, particularly in resources poor settings.