Assessing parameter identifiability in compartmental dynamic models using a computational approach: application to infectious disease transmission models.

摘要: Mathematical modeling is now frequently used in outbreak investigations to understand underlying mechanisms of infectious disease dynamics, assess patterns epidemiological data, and forecast the trajectory epidemics. However, successful application mathematical models guide public health interventions lies ability reliably estimate model parameters their corresponding uncertainty. Here, we present illustrate a simple computational method for assessing parameter identifiability compartmental epidemic models. We describe parametric bootstrap approach generate simulated data from dynamical systems quantify uncertainty identifiability. calculate confidence intervals mean squared error estimated distributions To demonstrate this approach, begin with low-complexity SEIR work through examples increasingly more complex that correspond applications pandemic influenza, Ebola, Zika. Overall, issues are likely arise (based on number equations/states parameters). As being jointly increases, surrounding tends increase, average, as well. found that, most cases, R0 often robust affecting individual model. Despite large higher other parameters, can still be precision accuracy. Because policies influenced by results studies, it important conduct analyses prior fitting available report estimates quantified The described helpful these regards enhances essential toolkit conducting model-based inferences using dynamic