COVID-19 superspreading events highlight need for backwards contact tracing

Burnet Institute

07 October, 2020

Image: Burnet Epidemiologist Professor Mike Toole says the COVID-19 cluster which emerged from Chadstone shopping centre is "modest" so far (source: ABC News)

Burnet research has highlighted the role of superspreading events of SARS-CoV-2, the virus that causes COVID-19, emphasising the importance of both forward and backward contact tracing to halt the virus’ spread.

Unlike influenza, diseases caused by coronaviruses have been shown to spread in clusters rather than in a linear fashion, with 10 percent of infected COVID-19 cases infecting more than 80 percent of people.

Burnet Institute epidemiologist, Professor Mike Toole AM told ABC Radio Melbourne the wide dispersion of a cluster can rapidly see the virus spread out of control.

ABC Radio Melbourne 6 Oct 2020

“What is of concern is when a cluster spreads widely and that’s why contact tracing is so important,” Professor Toole said.

The research on COVID-19 superspreading, published by Burnet’s Know-C19 Hub, recommends ‘backward’ contact tracing in order to find the origins of a cluster so that previously unidentified infected individuals can be revealed.

Professor Toole said the recent Chadstone Shopping Centre cluster which has spread to over 30 people across Melbourne and the Mitchell Shire after originating in the city’s south east is a “modest” superspreading event, so far.

“Keep in mind more than 100 people were infected from a cluster linked to a Thai restaurant in Sydney,” he said.

He also drew comparisons to significant superspreading events earlier in 2020 in South Korea and Washington State in the USA which saw hundreds of people infected from one person, often from only a short time in close contact.

“What reassures me is that [Victoria’s] overall numbers are so low that our contact tracers can get onto it.”

Professor Toole said it was crucial that all community resources were mobilised and recruited to halt the spread of a cluster, including church leaders and community leaders who speak different languages.

He said principles reducing people’s movement are also important in limiting the spread of clusters, and drew comparisons to Israel, where people cannot move further than 500m from their home or Bolivia, where one person per household is permitted to go shopping once per week.

“[In Victoria] we don’t have the strictest measures in the world,” he said.

The Know-C19 Hub research analysed studies of COVID-19 from Hong Kong and Singapore which found only 10 percent of people infect a large number of people, while 70 percent of people don’t infect anyone. A combination of high viral load and contact in a poorly ventilated indoor space were found to be catalysts for rapid spreading of the virus.

Read the latest Know-C19 Global Update

Know-C19 Hub COVID-19 Global Analysis October 2020

Overdispersion of the Virus (Superspreading)

  • COVID-19, like other diseases caused by coronaviruses, such as SARS and MERS, tend to cluster rather than spread linearly like influenza. COVID-19 is also subject to large superspreading events where one person may infect dozens or hundreds of others. The metric R0 does not reflect the reality of SARS-CoV-2 transmission.
  • The over-dispersion factor k is a better metric to characterise transmission. The lower the value of k the fewer people are causing most spread of the virus. This value has been estimated as 0.12 for SARS and 0.25 for MERS while estimates for COVID-19 are as low as 0.1 meaning that just 10 percent of infected cases are infecting more than 80 percent of people.
  • Superspreading events have been associated with large indoor gatherings in poorly ventilated spaces and where people may be singing (choir practices and church services) or speaking loudly (nightclubs and bars). It is likely that in this (indoor) environment an individual with a high viral load in their pharynx has introduced the virus into an environment where it spreads rapidly.
  • The concept of over-dispersion has implications for contact tracing. Rather than focusing only on ‘forward’ tracing, namely finding people that have been in contact since the person was infected, there should also be efforts to see who first infected the subject. This may reveal new previously unidentified infected individuals within the cluster.

Contact Details

For more information in relation to this news article, please contact:

Professor Michael Toole AM

Associate Principal Research Fellow


Subscribe to News

Subscribe to receive our latest news: