Back to the real world: connecting models with data

Mathematical models for infectious disease are often used to improve our understanding of infection biology or to evaluate the potential efficacy of intervention programs. Here, we develop a mathematical model that aims to describe infection dynamics of Mycobacterium avium subspecies paratuberculosis (MAP). The model was developed using current knowledge of infection biology and also includes some components of MAP infection dynamics that are currently still hypothetical. The objective was to show methods for parameter estimation of state transition models and to connect simulation models with detailed real life data. Thereby making model predictions and results of simulations more reflective and predictive of real world situations. Longitudinal field data from a large observational study are used to estimate parameter values. It is shown that precise data, including molecular diagnostics on the obtained MAP strains, results in more precise and realistic parameter estimates. It is argued that modeling of infection disease dynamics is of great value to understand the patho-biology, epidemiology and control of infectious diseases. The quality of conclusions drawn from model studies depend on two key issues; first, the quality of biology that has gone in the process of developing the model structure; second the quality of the data that go into the estimation of the parameters and the quality and quantity of the data that go into model validation. The more real world data that are used in the model building process, the more likely that modeling studies will provide novel, innovative and valid results.