Evidence for pre‐symptomatic transmission of coronavirus disease 2019 (COVID‐19) in China

Xiang Ren, Yu Li, Xiaokun Yang, Zhili Li, Jinzhao Cui, Aiqin Zhu, Hongting Zhao, Jianxing Yu, Taoran Nie, Minrui Ren, Shuaibing Dong, Ying Cheng, Qiulan Chen, Zhaorui Chang, Junling Sun, Liping Wang, Luzhao Feng, George F. Gao, Zijian Feng, Zhongjie LiInfluenza 2020 · DOI

TL;DR

This study provides crucial early evidence that a significant portion of COVID-19 transmissions occurred before infected individuals showed symptoms. By analyzing early case data outside of Hubei province, the researchers quantified the role of pre-symptomatic spread, highlighting its importance for public health interventions.

What Problem Does This Paper Address?

As COVID-19 began to spread globally in early 2020, a critical unknown was the extent to which infected individuals could transmit the virus before they became symptomatic. This "pre-symptomatic transmission" is a major challenge for containment strategies, as individuals may unknowingly spread the virus while they appear healthy. Previous studies had suggested this phenomenon but often with limited data or without quantifying its prevalence. This paper aimed to address this gap by analyzing early case data from mainland China outside the initial epicenter to estimate the proportion of transmissions that occurred pre-symptomatically.

Methods

The researchers employed a retrospective analysis of epidemiological data from confirmed COVID-19 cases reported to the Chinese Public Health Event Surveillance System. Key methods included:

  • Epidemic Curve Analysis: Constructing epidemic curves based on illness onset dates to estimate growth rates and doubling times.
  • Incubation Period Estimation: Fitting a lognormal distribution to exposure windows and onset dates to estimate the incubation period, accounting for epidemic growth and interval censoring.
  • Serial Interval Estimation: Analyzing clusters of transmission events (primary and secondary cases) to estimate the serial interval (time between onset of illness in successive cases) using a normal distribution, allowing for negative intervals and correcting for epidemic growth.
  • Pre-symptomatic Transmission Inference: Using the estimated incubation period and serial interval distributions, along with detailed contact tracing data for transmission pairs, they employed Monte Carlo simulations to estimate the probability of transmission occurring before symptom onset.
  • Reproductive Number Estimation: Calculating the effective reproductive number (R) using the estimated serial interval.

Datasets

The study utilized data from the Chinese Public Health Event Surveillance System, comprising 449 of the earliest COVID-19 cases detected outside Hubei province. This included detailed demographic information, travel history, exposure history, and clinical data for laboratory-confirmed cases. The analysis focused on 168 confirmed cases for detailed epidemic curve construction and 64 clusters involving 80 transmission events for serial interval and pre-symptomatic transmission analysis.

Key Insights and Ideas

  • Pre-symptomatic transmission is a significant driver: The study provides strong quantitative evidence that a substantial portion of COVID-19 spread occurred before individuals exhibited symptoms. This is a critical insight for understanding the virus's rapid dissemination.
  • Milder secondary cases may reflect true severity: Secondary cases in clusters often presented with milder disease compared to primary or sporadic cases. This suggests that milder infections might be more common than initially apparent and that focusing solely on severe cases might underestimate the true burden of infection.
  • Serial interval is shorter than incubation period: The estimated serial interval was found to be shorter than or comparable to the incubation period, a pattern consistent with pre-symptomatic transmission.

Key Results

  • Epidemic Doubling Time: Estimated at 5.3 days (95% CI: 4.3, 6.7).
  • Median Incubation Period: Estimated at 4.6 days (95% CI: 4.0, 5.2).
  • Serial Interval: Mean of 5.7 days (95% CI: 4.7, 6.8) with a standard deviation of 3.5 days.
  • Effective Reproductive Number (R): Estimated at 1.98 (95% CI: 1.68, 2.35).
  • Pre-symptomatic Transmission: An estimated 32 out of 80 (40%) transmission events were likely or very likely to have occurred prior to symptom onset in the primary case.
  • Severity Comparison: Cluster secondary cases were more likely to be mild (87%) compared to cluster primary and sporadic cases (66%).

Limitations and Future Work

The authors acknowledge several limitations:

  • Bias towards severe infections: The study relied on confirmed cases, which likely overrepresents more severe infections and may miss milder or asymptomatic cases.
  • Ascertainment of symptom onset: Self-reported symptom onset dates can be unreliable, especially in the early stages of an outbreak.
  • Potential overestimation of pre-symptomatic transmission: If primary cases took measures to reduce infectivity after symptom onset, this could artificially shorten the observed serial interval and inflate the estimate of pre-symptomatic transmission.
  • Limited data for non-Wuhan cities: While some cases outside Wuhan were analyzed, the sample size was insufficient to definitively determine sustained transmission in other cities.

Future work could involve prospective studies of close contacts with repeated sampling to better characterize mild infections and viral shedding. The findings strongly support the continued importance of quarantine for exposed individuals and social distancing measures to curb transmission.

Summary updated

Abstract

The majority of transmissions are occurring around illness onset in an infected person, and pre-symptomatic transmission does play a role. Detection of milder infections among the secondary cases may be more reflective of true disease severity.
BibTeX
@inproceedings{ren2020evidence,
  author = {Xiang Ren and Yu Li and Xiaokun Yang and Zhili Li and Jinzhao Cui and Aiqin Zhu and Hongting Zhao and Jianxing Yu and Taoran Nie and Minrui Ren and Shuaibing Dong and Ying Cheng and Qiulan Chen and Zhaorui Chang and Junling Sun and Liping Wang and Luzhao Feng and George F. Gao and Zijian Feng and Zhongjie Li and Zhongjie Li and Zhongjie Li},
  title = {{Evidence for pre‐symptomatic transmission of coronavirus disease 2019 (COVID‐19) in China}},
  booktitle = {Influenza},
  year = {2020},
  doi = {10.1111/irv.12787},
}
Figure