Modelling presymptomatic infectiousness in COVID-19

This paper considers SEPIR, an extension of the well-known SEIR continuous simulation compartment model. Both models can be fitted to real data as they include parameters that can be estimated from the data. SEPIR deploys an additional presymptomatic infectious compartment, not modelled in SEIR but known to exist in COVID-19. This stage can also be fitted to data. We focus on how to fit SEPIR to a first wave of COVID. Both SEIR and SEPIR and the existing SEIR models assume a homogeneous mixing population with parameters fixed. Moreover, neither includes dynamically varying control strategies deployed against the virus. If either model is to represent more than just a single wave of the epidemic, then the parameters of the model would have to be time dependent. In view of this, we also show how reproduction numbers can be calculated to investigate the long-term overall outcome of an epidemic.