A note about the invariance of the basic reproduction number for stochastically perturbed SIS models

Abstract

In Gray et al. [A stochastic differential equation SIS epidemic model, SIAM J. Appl. Math. 71 (3) (2011) 876–902] a susceptible‐infected‐susceptible (SIS) stochastic differential equation (SDE), obtained via a suitable random perturbation of the disease transmission coefficient in the classic SIS model, has been studied. Such random perturbation enters via an informal manipulation of stochastic differentials and leads to an Itô's‐type SDE. The authors identify a stochastic reproduction number, which differs from the standard one for the presence of those additional parameters that describe the employed random perturbation, and show that, similarly to the deterministic case, the stochastic reproduction number rules the asymptotic behavior of the solution. Aiming to make that random perturbation rigorous, we suggest an alternative approach based on a Wong–Zakai approximation argument thus arriving at a different stochastic model corresponding to the Stratonovich version of the Itô equation analyzed in Gray et al. Rather surprisingly, the asymptotic behavior of this alternative model turns out to be governed by the same reproduction number as the deterministic SIS equation. In other words, the random perturbation does not modify the threshold for extinction and persistence of the disease.

Itô vs Stratonovich stochastic SIR models

“…The previous theorem asserts that in the Stratonovich model (1.5) the conditions for extinction and persistence of the disease coincide with those of the deterministic system (1.1). This result extends to the SIR model an equivalent feature shown recently for the SIS model: see [4], [1], [9]. It also contributes to the fundamental modelling issues discussed for instance in [3] and [13].…”