Modeling the Comparative Impact of Individual Quarantine vs. Active Monitoring of Contacts for the Mitigation of COVID-19

Corey M Peak; Rebecca Kahn; Yonatan H Grad; Lauren M Childs; Ruoran Li; Marc Lipsitch; Caroline O Buckee

We fit the model assuming short (mean of 4.8 days; scenario 1) versus long (mean of 7.5 days; scenario 2) serial interval estimates. Model fitting by SMC for serial interval scenario 1 (mean = 4.8 days) resulted in mean duration of infectiousness 2.4 days [95% CI 1.0, 6.7]; mean time of peak relative infectiousness at 43% of the duration of infectiousness [0%, 97%]; and a mean time of infectiousness onset 0.77 days before symptom onset [1.98 days before, 0.29 days after] (Table 2, Figure S1 ). The longer serial interval in scenario 2 (mean = 7.5 days) resulted in slower disease dynamics: mean duration of infectiousness 4.8 days [95% CI 1.12, 10.5]; mean time of peak relative infectiousness at 38% of the duration of infectiousness [0%, 97%]; and a mean time of infectiousness onset 0.51 days after symptom onset [0.77 days before, 1.50 after] (Table 2, Figure S2 ). Therefore, given the same incubation period distribution with mean = 5.2 days, a serial interval with mean = 4.8 days is best fit by substantial presymptomatic infectiousness (mean = 20.5% [0%, 91.4%]) while a longer serial interval with mean = 7.5 days is best fit by limited presymptomatic infectiousness (mean = 0.065% [0%, 0.88%]).