On the time shift phenomena in epidemic models

Dobie, Ayse Peker; Demirci, Ali; Bilge, Ayse Humeyra; Ahmetolan, Semra

In the standard Susceptible-Infected-Removed (SIR) and Susceptible-Exposed-Infected-Removed (SEIR) models, the peak of infected individuals coincides with the in ection point of removed individuals. Nevertheless, a survey based on the data of the 2009 H1N1 epidemic in Istanbul, Turkey [19] displayed an unexpected time shift between the hospital referrals and fatalities. With the motivation of investigating the underlying reason, we use multistage SIR and SEIR models to provide an explanation for this time shift. Numerical solutions of these models present strong evidences that the delay is approximately half of the infection period of the epidemic disease. In addition, graphs of the classical SIR and the multistage SIR models; and the classical SEIR and the multistage SEIR models are compared for various epidemic parameters. Depending on the number of stages, it is observed that the delay varies for relatively small stage numbers whereas it does not change for large numbers in multistage systems. One important result that follows immediately from this observation is that this fixed delay for large numbers explains the time shift. Additionally, depending on the stage number and the duration of the epidemic disease, the distance between the points where each infectious stage reaches its maximum is found approximately both graphically and qualitatively for both systems. Variations of the time shift, the maximum point of the sum of all infectious stages, and the in ection point of the removed stage are observed subject to the stage number N and it is shown that these variations stay unchanged for greater values of N.