TL;DR
This study investigated the prevalence of SARS-CoV-2 infection among individuals repatriated from Wuhan, China, during the early stages of the COVID-19 pandemic. By analyzing data from repatriation flights where all passengers were tested upon arrival, the researchers found a low infection prevalence of 0.44%, suggesting that community transmission in Wuhan might have been lower than initially feared at that specific time, despite the city being the epicenter.
What Problem Does This Paper Address?
Early in a novel pathogen outbreak, especially when it originates in a densely populated area like Wuhan, China, accurately assessing the true extent of community transmission is a significant challenge. Local case ascertainment often relies on symptomatic testing, which can underestimate the actual number of infections, particularly if asymptomatic or mildly symptomatic cases are prevalent. Repatriation flights, while not perfectly representative of the general population, offer a unique opportunity to sample individuals directly from the outbreak's epicenter. This paper addresses the need for a more robust estimate of infection prevalence in Wuhan during the critical early period (January-February 2020) by analyzing data from these repatriation efforts, providing a conservative baseline for understanding the virus's spread.
Methods
The researchers identified repatriation flights from Wuhan to 55 countries between January 29 and February 27, 2020. They focused their analysis on 17 specific flights where all passengers (N=2433) were tested for SARS-CoV-2 infection upon arrival, regardless of symptoms. This rigorous testing protocol allowed for a more accurate estimation of infection prevalence compared to relying solely on symptomatic individuals. The infection point prevalence per flight was calculated as the number of positive RT-PCR test results divided by the total number of individuals tested on that flight. To account for variations between flights and provide an overall estimate, a binomial mixed-effects model was used to calculate a pooled infection prevalence with exact 95% binomial confidence intervals.
Datasets
The primary dataset for this study consists of information on 56 repatriation flights that carried a total of 8597 individuals from Wuhan to 55 countries between January 29 and February 27, 2020. Specifically, the analysis focused on 17 flights where all 2433 passengers were tested upon arrival. Data on the number of repatriated individuals per flight, their final destinations, and testing outcomes (total tested, total positive) were collected from international and local news outlets and government press releases. The authors also provide a link to a downloadable data table of identified flights on their public GitHub repository.
Key Insights and Ideas
- Repatriation flights as a proxy for early outbreak assessment: The study highlights the value of analyzing data from repatriation flights as a method to estimate infection prevalence in an outbreak's epicenter when local surveillance is limited.
- Low initial prevalence in Wuhan: The finding of a relatively low pooled infection prevalence (0.44%) among repatriates from Wuhan, even during the reported peak of the epidemic, suggests that the virus might not have been as widespread in the general population as initially feared, or that symptomatic testing was significantly underestimating cases.
- Conservative estimate: The authors acknowledge that the prevalence found in repatriates is likely a conservative estimate. Factors such as travel bans preventing symptomatic individuals from boarding and the diverse socio-economic backgrounds of travelers mean this group might not perfectly reflect the general Wuhan population.
- Comparison with later serological data: The study's findings are contrasted with later serological surveys in Wuhan, which showed higher seroprevalence. This comparison underscores the dynamic nature of outbreaks and the challenges of estimating prevalence at different time points and using different methodologies.
Key Results
The study found that across 17 repatriation flights where all 2433 passengers were tested upon arrival, 13 individuals tested positive for SARS-CoV-2. This resulted in a pooled infection prevalence of 0.44% (95% CI: 0.19%–1.03%). For the five flights departing between January 30 and February 1, which were closest to the reported peak of the epidemic in Wuhan, the pooled infection prevalence was 0.88% (6/685, 95% CI: 0.39%–1.93%). Per-flight prevalence ranged from 0% to 1.9%.
Limitations and Future Work
The authors acknowledge several limitations:
- Not representative of the general population: The repatriated individuals may not accurately represent the general population of Wuhan due to factors like socio-economic status, travel reasons, and potential pre-travel screening.
- Exclusion of post-arrival infections: The study did not consider individuals who tested positive after arrival during their quarantine period, as the focus was on infection prevalence at the time of flight.
- Potential for asymptomatic transmission: While all passengers were tested, the study notes that transmission during the flight itself could not be ruled out and was not explicitly accounted for in the prevalence calculation.
Future work could involve more detailed analysis of transmission dynamics on the flights themselves, and further comparison with serological data from different time points and population groups to build a more comprehensive picture of the outbreak's trajectory. The authors suggest that characterizing infection prevalence from repatriated cohorts is a valuable method for bridging the gap between symptom-based surveillance and seroprevalence surveys, especially during the critical early phases of an epidemic.
Summary updated