Abstract
Background: Pre-symptomatic transmission can be a key determinant of the effectiveness of containment and mitigation strategies for infectious diseases, particularly if interventions rely on syndromic case finding. For COVID-19, infections in the absence of apparent symptoms have been reported frequently alongside circumstantial evidence for asymptomatic or pre-symptomatic transmission. We estimated the potential contribution of pre-symptomatic cases to COVID-19 transmission. Methods: Using the probability for symptom onset on a given day inferred from the incubation period, we attributed the serial interval reported from Shenzen, China, into likely pre-symptomatic and symptomatic transmission. We used the serial interval derived for cases isolated more than 6 days after symptom onset as the no active case finding scenario and the unrestricted serial interval as the active case finding scenario. We reported the estimate assuming no correlation between the incubation period and the serial interval alongside a range indicating alternative assumptions of positive and negative correlation. Results: We estimated that 23% (range accounting for correlation: 12 - 28%) of transmissions in Shenzen may have originated from pre-symptomatic infections. Through accelerated case isolation following symptom onset, this percentage increased to 46% (21 - 46%), implying that about 35% of secondary infections among symptomatic cases have been prevented. These results were robust to using reported incubation periods and serial intervals from other settings. Conclusions: Pre-symptomatic transmission may be essential to consider for containment and mitigation strategies for COVID-19.
Highlights
Since its emergence in early December 2019, SARS-CoV-2 has spread rapidly despite multiple layers of interventions designed to prevent a pandemic[1]
Through accelerated case isolation following symptom onset, this percentage increased to 46% (21 – 46%), implying that about 35% of secondary infections among symptomatic cases have been prevented
These results were robust to using reported incubation periods and serial intervals from other settings
Summary
The probability of pre-symptomatic transmission can be expressed as the probability of onward transmission during one’s incubation period. For the no active case finding scenario we use the serial interval reported based on cases that were only isolated at least 6 days after their symptom onset; i.e. 8.0 days (95% empirical confidence interval (eCI): 4.7 – 11.3). Nishiura et al provide estimates of the serial interval from published reports across all affected countries, mostly including attempted rapid case finding and isolation[17]. We use their mean serial interval estimates based on high certainty transmission pairs (4.8 days (95% eCI: 1.1 – 9.6)) in combination with either the incubation time reported by Bi et al.[15] or Li et al.[16]. All analyses were done in R 3.6.319; codes and data are available online[18]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
