Evaluating the Reliability of Data-Driven Compartmental Model (SIR or VH) on Dengue Fever for the Ongoing Outbreak in America.

The compartmental epidemiological model is commonly used to study dengue dynamics; some of these models precisely consider the mosquito population, and others indirectly capture its role in disease transmission term. In this article, we have performed a comparative analysis between a simple SIR model and a vector-host interaction model (VH model) by fitting the dengue fever data of the ongoing outbreak in America. Parameter estimations for both models have been performed using the Nelder-Mead simplex algorithm, considering the normalized root mean square error (NRMSE) as an optimization function. The significance and reliability of the estimated parameters towards the models' predictions have been analysed through uncertainty and sensitivity analyses. Uncertainty analysis utilizing the Latin hypercube sampling (LHS) method has been performed to evaluate how various parameters within the models influence the basic reproduction number ([Formula: see text]). Sensitivity analysis for the basic reproduction number ([Formula: see text]), has been carried out by the partial rank correlation coefficients (PRCC) method. Additionally, we have computed the parameter regions ensuring the persistence of equilibrium points of both models. This study offers profound insights into model selection, parameter estimation, and forecasting future data trends for the ongoing dengue outbreak in America. However, this article focuses on exploring two key scientific questions: (1) Which type of compartmental model (SIR or VH) is more suitable to capture the trend of data on dengue fever for the ongoing outbreak in America? (2) Is America likely to face a prolonged dengue outbreak in the near future?