The Impact of Host-Based Early Warning on Disease Outbreaks

Mark Hernandez; Lauren E Milechin; Shakti K Davis; Rich DeLaura; Kajal T Claypool; Albert Swiston

To demonstrate the potential utility of the early warning-enabled quarantine-on-alert model in a more comprehensive QIT policy analysis, we compared outcomes over a range of disease transmission rates, initial exposure conditions, and early warning performance parameters. We varied these parameters in simulations of a mass exposure to an infectious pathogen occurring in a population of 1,000 people. All scenarios were simulated over 50 days. Because we fixed the exposure time for all cases to t = 0, rather than using the quarantine release rate ε, all individuals in quarantine were returned to the susceptible compartment at t = 21 days (the maximum incubation period). We defined policy outcomes via two metrics derived from the policy-dependent SEIR model outputs as shown in Figure 9 . The first, lost duty days, is the percentage of total number of days of work productivity that are lost because of quarantine and isolation in the 50-day simulation; this percentage is proportional to the integral of the curves in Figure 9a . The second metric, cumulative infections, is the percentage of the population that has been infected by the end of the simulation (Figure 9b ). These two metrics characterize a trade space for evaluating QIT policies under different circumstances, as any measure that reduces the number of people in quarantine and isolation may be expected to increase the likelihood of infection in the population. While these metrics are helpful for evaluating the impact of various policy choices, they are by no means comprehensive and particularly do not consider the financial or other costs associated with quarantine. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity.