New approximations, and policy implications, from a delayed dynamic model of a fast pandemic

Chandrika Prakash Vyasarayani; Anindya Chatterjee

With β = 1, p = 0.98 and γ = 0.1, we obtain table 1. It is seen that relatively small reductions in τ achieve significant reductions in the net population affected by the disease. However, if p = 0.92 (i.e., the chance of escaping quarantine is four times greater) and γ remains the same, then instability occurs at τ = 0.22 (about four times smaller), and S(∞) is less sensitive to τ as well (numerical examples omitted for brevity). We conclude that in circumstances where p can be kept high (i.e., if public institutions are strong and detection followed by quarantine is nearly certain), and where the self-recovery rate γ is not extremely small, the system can benefit significantly from even modest reductions in the detection time τ . In such situations, research directed toward earlier detection may yield substantial benefits, and it may even be unnecessary to engage in extreme social distancing. On the other hand, if there is under-reporting and imperfect quarantining, then impractically large reductions in τ may be needed to compensate, and strict social distancing may be more effective.