EpiBeds: Data informed modelling of the COVID-19 hospital burden in England
Fig 1
Schematic representation of the EpiBeds model.
The construction of the compartmental model was informed by available data. EpiBeds is implemented as a set of ordinary differential equations (ODEs), with one state variable per compartment representing the absolute number of individuals in it. Arrows describe flow between compartments, which occurs at constant rate. Blue compartments indicate infected individuals who are not hospitalised, with a dark and light blue distinction, respectively, for individuals with and without symptoms, while red compartments indicate hospitalised individuals and orange compartments individuals in critical care. The compartments with a red border contain infectious individuals, with a dashed border denoting an infectivity reduced to 25% of that of the other infectious compartments; once hospitalised, it is assumed individuals no longer contribute to the community epidemic. For states in which the waiting times are not exponentially distributed (e.g. Exposed) we use multiple identical compartments enabling us to approximate gamma-distributed waiting times by using Erlang distributions. All variables, rates, and probabilities are described in Tables 2 and 3. The force of infection λ depends on the numbers in the infectious compartments (Section 4.1).