Mathematical model of COVID-19 with imperfect vaccine and virus mutation

TL;DR

This paper develops a detailed mathematical model to analyze COVID-19 dynamics considering imperfect vaccines and virus mutations, providing insights into stability, reproduction number, and bifurcation phenomena.

Contribution

The study introduces a comprehensive five-compartment model for COVID-19 with vaccine imperfections and mutations, including stability analysis and bifurcation analysis.

Findings

Identified conditions for disease-free and endemic equilibria.

Calculated the basic reproduction number for the model.

Demonstrated the existence of backward bifurcation in the system.

Abstract

This study examines the effect of a partially protective vaccine on COVID-19 infection with the original and mutant virus with the help of a deterministic mathematical model developed. The model we developed consists of five compartments and thirteen parameters. The model consists of S (susceptible), V (vaccinated), I_1 (infected with original virus), I_2 (infected with mutant virus) and R (recover) subcompartments. With the established model, imperfect vaccines and mutant virus for the COVID-19 pandemic in Turkey were examined. We examined the effect of both artificial active immunity (vaccinated) and natural active immunity (passing disease) in the model. Since it is known that the recovery and death rates of the original virus and the mutant virus are different in COVID-19, we considered it in the study. We performed local stability and global stability analysis by calculating the disease-free equilibrium point and endemic equilibrium point of the model. We also obtained the basic reproduction number with the help of the next generation matrix method. We estimated three model parameters with parameter estimation and identified model-sensitive parameters with local sensitivity analysis. Next, we showed the existence of a backward bifurcation using the Castillo-Chavez and Song Bifurcation Theorem. Additionally, to illustrate the underlying fundamental mechanisms of the model dynamics and to support the analytical findings, we presented three distinct simulations.