Transmission interval estimates suggest pre-symptomatic spread of COVID-19

Lauren Tindale; Michelle Coombe; Jessica E Stockdale; Emma Garlock; Wing Yin Venus Lau; Manu Saraswat; Yen-Hsiang Brian Lee; Louxin Zhang; Dongxuan Chen; Jacco Wallinga; Caroline Colijn

is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.03.03.20029983 doi: medRxiv preprint Our serial intervals are notably shorter than our incubation period estimates, suggesting that there is pre-symptomatic transmission, with infection occurring on average 2.89 and 2.55 days before symptom onset of the infector (Tianjin, Singapore). Because the incubation period is different for early-and late-occurring cases in our data, on average transmission for early-occurring cases is 2.79 and 1.2 days before symptom onset (Tianjin, Singapore) and 8.2, 3.3 days before (Tianjin, Singapore) for late-occurring cases. The fact that serial intervals are shorter than incubation periods is robust in our sensitivity analysis (Table S3 ). These estimates are strengthened by the fact that we have estimated both incubation period and serial interval in the same population and by the fact that we obtain the same result in two distinct datasets. In both sets of estimates, samples of the incubation period minus serial interval are negative with probability 0.8 or higher (Tianjin) and 0.7 or higher (Singapore), suggesting that a substantial portion of transmission may occur before symptom onset (see Supplementary Information and Figure S2 ), consistent with the clinical observations reported by Rothe et al. [21] and Bai et al. [22] .