Intervention Strategies for Epidemics: Does Ignoring Time Delay Lead to Incorrect Predictions?.

Our paper investigates distributions of exposed and infectious time periods in an epidemic model and how applying a disease control strategy affects the model's accuracy. While ordinary differential equations are widely used for their simplicity, they incorporate an exponential distribution for time spent exposed or infectious. This allows for a high probability of unrealistically short exposed and infectious time periods. We propose that caution must be taken when applying intervention methods to basic models in order to avoid inaccurate predictions. Delay differential equations, which use a delta distribution for exposed and infectious periods, can provide better realism but are more difficult to use and analyze. We introduce a multi-infected compartment model to interpolate between an ODE model with exponential distributions and a DDE model with delta distributions in order to investigate the effect these distributions have on the dynamics of the system when an intervention method is also included. Using steady state stability and bifurcation analysis, this paper considers when simpler infectious disease models can be used versus when more realistic time periods must be incorporated. We find that the placement of control measures on subpopulations and the length of the time delay impacts the accuracy of the simpler models.