Transmission Dynamics of SARS-CoV-2: Superspreaders and Superspreading Events

The concept of Superspreaders and Superspreading events has recently attracted a lot of attention. In fact it is important to understand that most of SARS-CoV-2 infected persons are in fact not contagious! Thus, transmission is really dependent on a handful of individuals we call Superspreaders who nurture Superspreading events. Let’s distinguish between Superspreaders and Superspreading events. Superspreaders are individuals who infect a high number of persons; why?  This is not clear, and this is a most important issue to clarify in the future.  We know they yield high viral load and that they are generally, but not always, young people.  But, this cannot fully explain their massive contamination impact.  Moreover, they are frequently asymptomatic, thereby significantly increasing risk of dissemination. Superspreading events, which involve at least one Superspreader, are events which favor large scale transmission, such as close contacts in indoor situations. Secondary transmissions from infected people then result in a large number of further infections, and so on. Thus, by some estimations  only 20% of infected individuals cause 80% of infections(1).  Identification of Superspreading events depends upon contact tracing.  Furthermore, DNA sequencing of viral genomes adds a great deal to a clearer understanding of these phenomena, thus, confirming the substantial role of Superspreaders in the pandemic. Let’s look at some well-characterized superspreader events to try to better understand how the majority of SARS-CoV-2 infections occur, enabling us to gain an understanding of what might be done to prevent them.

One of the early recognized superspreading events occurred in mid-March in Skagit County, Washington(2) at a 2.5 hr-long choir practice in which 61 people sang in close proximity. Probable or confirmed infections occurred in 87% of the attendees. Given the low incidence of COVID-19 at the time of this event, it is likely that all the infections originated with a single individual, and that the act of singing vigorously launched many viral laden particles into the air. People were relatively closely positioned. Thus, it is likely that superspreading occurs in an unusually favorable environment. Looking at it another way, the virus got “lucky.”  Another study, unrelated to the choir event, analyzed the sequences of 453 viral genomes collected between February and March in Washington(3).  It is possible to infer the likelihood of how many people have been infected by a single person using viral genomic epidemiology, especially given the relative genetic stability of SARS-CoV-2. A phylogenetic analysis strongly suggested that 84% of the 453 viral genomes derived from a single introduction sometime in early February.  In the choir study, it is clear how virus was transmitted. In the genomic study, it is only clear that a single infected individual somehow infected a great number of people through primary, secondary, and other less direct routes.

Another well-studied superspreader event began with a single infected individual in a meat packing plant is Postville, Iowa. In this case, viral spread could be ascertained by both contact tracing and by genomic epidemiology. Fourteen independent viral introductions were identified in the region, but the only virus to spread widely was the one from the meat packing plant. The virus from this individual passed first to numerous other workers, then to family members, then to the community in Postville (87 cases), and finally to other locations in an area of 185 square miles in Iowa, Wisconsin and Minnesota. These conclusions were supported by genomic sequences from 27 different infected individuals. Again, it appears that prolonged close contact indoors facilitated transmission and suggests that this is a critical feature of superspreading events.

Perhaps, the best characterized superspreading event, originating from an international business meeting of Biogen in Boston(4) in which more than 90 individuals became infected.  This event recently received considerable attention. The large number of infections was indicative of a possible superspreading event. A recent study looked at this event and its consequences in detail, using genomic epidemiology(5). They were able to identify and track the virus in question by a single nucleotide polymorphism (SNP), C2416T. Among 80 separate introductions from four continents into the Boston area early in the pandemic, which they inferred from phylogenetic analyses, the C2416 SNP was unique to one virus. Comparing other viral genomes from various parts of the world with C2416T, the parental origin appears to be in Europe, perhaps France, with an estimated most recent common ancestor existing about two weeks prior to the conference aroundFebruary 26-27. Of all samples collected prior to March 10, the only instances of C2416T were from people who had attended the conference, indicating that a superspreading event had indeed occurred there.

Subsequent samples (744) from infected individuals in Boston and surrounding areas were collected over a period from February to June and genomes were sequenced. Remarkably, 35% of the samples had the C2416 SNP. Since no sample prior the March 10 had this SNP, it suggests that the superspreading event at the Biogen meeting February 26-27 resulted in virus from a single individual infecting more than a third of all infected people in the Boston area. Percentages of the C2416T SNP in regions around Boston ranged from 30-46% in four adjacent counties. In addition, a second SNP, G26233T, appears to have emerged during the event, enabling further tracking. Data from this SNP shows a likely export to other states and countries, with further community spread in Virginia, Michigan and Australia. Some caution is warranted, however, since genomic sampling is not generally done on a randomized basis.

The same report looked at infection clusters at homeless shelters, nursing facilities, and a hospital to gain a better understanding of transmission dynamics. They analyzed 193 viral genomes collected from the Boston Health Care for the Homeless Program and identified 4 clusters of 20 or more highly similar genomes, including two clusters containing the C2416T SNP. They also investigated a superspreading event at a skilled nursing facility, in which 82/97 (84%) residents and 36/97 (37%) of staff were infected. In fact, 75% of viral genomes from different individuals had highly similar genomes, suggesting that they arose from a single recent introduction. This took place even though strict interventional measures were in place. Interestingly, two other clusters of three closely related genomes were detected. This represents independent introductions, but these failed to massively spread. In the case of two clusters of infection at Massachusetts General Hospital, highly similar genomes were not found, suggesting a lack of significant in-hospital spread.

What is our conclusion? First, it is now clear that a majority of transmissions result from superspreading events, facilitated by conducive conditions. These include indoor location, close contact, lengthy contact, indoor activity such as singing or talking, poor air ventilation, and lack of mitigation procedures (i.e., wearing masks and physical distancing). Tracing Superspreading events in Hong Kong also confirmed that the largest cluster (106 cases) was traced to four bars followed by a wedding (22 cases) and attendance at a temple (19 cases) (1). This study suggests that disease control efforts should focus on avoiding gathering events and mitigating their impact.  The rapid tracing and quarantine of confirmed contacts, along with the implementation of physical distancing policies including either closures or reduced capacity measures targeting high-risk social settings such as bars, weddings, religious sites and restaurants, should be efficient to prevent the occurrence of superspreading events. Overall, the issue of Superspreaders and Superspreading events illustrates the impact of molecular epidemiology for deciphering the patterns of COVID-19 dissemination. What we still clearly lack, however, is the understanding of the very early phases of the pandemics in China. This would be very useful for the whole appraisal of transmission dynamics.

1. D. C. Adam et al., Clustering and superspreading potential of SARS-CoV-2 infections in Hong Kong. Nat Med, (2020).
2. L. Hamner et al., High SARS-CoV-2 Attack Rate Following Exposure at a Choir Practice – Skagit County, Washington, March 2020. MMWR Morb Mortal Wkly Rep 69, 606-610 (2020).
3. T. Bedford et al., Cryptic transmission of SARS-CoV-2 in Washington state. Science, (2020).
4. A. Schuchat, C. C.-R. Team, Public Health Response to the Initiation and Spread of Pandemic COVID-19 in the United States, February 24-April 21, 2020. MMWR Morb Mortal Wkly Rep 69, 551-556 (2020).
5. J. Lemieux et al., Phylogenetic analysis of SARS-CoV-2 in the Boston area highlights the role of recurrent importation and superspreading events. medRxiv, (2020).