New mathematical model predicts what events spread COVID-19

A team of mathematical researchers has created a model for estimating which preventive measures work best to minimize the spread of COVID-19 at different types of events.

The researchers used data from actual single-event, short-duration outbreaks to develop a model for predicting how many new infections can be expected to result from different events if one infectious person were present. 

The quantitative framework the researchers present is useful for evaluating when and how to reopen economic and social activities. It can also be used, they suggest, as a way "to predict outbreak sizes and to simulate outbreaks and interventions under different scenarios for team size, work from home arrangements, and other structures."

In addition, the researchers say, the model can be used to help determine how many people might need to be tested or put in isolation after an outbreak. 

The model also offers guidelines about which precautionary measures are most effective in different situations such as mask wearing, social distancing, and keeping contact groups small ("bubbling").

They found that social distancing is most effective at preventing COVID-19 spread in every type of situation, while wearing masks and limiting interaction were more effective in certain types of events.

The study, by a research team from Simon Fraser University in British Columbia and Imperial College in London, appears in Proceedings of the National Academy of Sciences, Nov. 19. 

The model starts with the concept of "event R," the expected number of new infections caused by the presence of one infectious individual at an event. It considers four parameters for the event: transmission intensity, duration of exposure, proximity of individuals and the degree of mixing.

Transmission rates are different, depending on how many people come in contact with the infectious person and how long they are in contact. The model develops a measure of the risk of transmission per unit time.

From the researchers' examination of the outcome of known events (parties, public transit, restaurant meals, bars), they arrive at a mathematical model that predicts COVID-19 transmission using different types of precautionary interventions. 

They categorized events as being saturating, with high transmission probability, or linear, with low transmission possibility. They found transmission rates that ranged from low (0.02 to 0.05 per hour) for small household groups or a funeral, to high (0.5 to 0.6 per hour) for events involving speaking, singing or eating.

The authors make an interesting observation about mask wearing. It may be difficult to "estimate the effectiveness of masks and other physical barriers to transmission" in saturating settings, because "even an intervention that halves the transmission rate may not have much impact on the number of infections." They note that this might help explain the different results in studies on the benefits of mask wearing. 

However, the researchers note, the evidence is strong that social distancing of 1 meter (3.3 feet) or more has an effect on reducing transmission in all situations.

As for social bubbles, they say that the effectiveness depends on the amount of mixing and saturation. Keeping strictly in a social bubble in crowded spaces can help reduce COVID-19 transmission, the study says.

The researchers note that the event R rate will also depend on "the frequency of the event, the total attendance and the prevalence of the disease in the population." For example the event R rate for a person on a 30-minute bus ride is low, but if the bus is crowded or the person often takes the bus, or if there's a high prevalence of COVID-19 in the area, the event R rate will be higher. 

Decision-makers looking at transportation, classes or other events, will have to find an "acceptable COVID-19 cost-benefit balance," the researchers say.