Restaurants, cafes, and gyms acted as "superspreading" sites for COVID-19 transmission last spring, accounting for the majority of new infections in large U.S. cities, according to a new study.
The findings, published today (Nov. 10) in the journal Nature (opens in new tab), also suggest that reducing maximum occupancy at these venues may be more effective for curbing COVID-19 spread than blanket lockdowns, the authors said.
"Our work highlights that it does not have to be all or nothing," study senior author, Jure Leskovec, a computer scientist at Stanford University, told The New York Times.
The study researchers created a model to mimic the spread of COVID-19 in 10 major U.S. cities: New York, Los Angeles, Chicago, Dallas, Washington, D.C., Houston, Atlanta, Miami, Philadelphia and San Francisco. The model not only took into account standard factors in the spread of an infectious disease (such as how many people are susceptible, exposed, infected and immune to the virus), but also drew on real data that showed how often people came into close contact with others. To do so, they incorporated information on people's behavior using anonymized cellphone data from 98 millions Americans, tracking their movements from their neighborhoods to about 553,000 public locations between March 1 and May 2. They also obtained the square footage of these venues to calculate the number of people per square foot in each of these locations at a given time. .
They found that their model could accurately predict daily COVID-19 case counts in these cities.
The researchers then estimated the number of infections that occurred at each of the public locations, and found that the majority of infections occurred at just a small number of "superspreading" venues. For example, in Chicago, 10% of venus accounted for 85% of the predicted infections, the authors found.
What's more, when the researchers modeled the risk of reopening venues after lockdowns, some venues — particularly restaurants — posed a much higher risk than others in terms of the number of new infections that would occur upon reopening.
"Restaurants were by far the riskiest places, about four times riskier than gyms and coffee shops, followed by hotels," Leskovec said in a news conference, according to the Times. The researchers hypothesized these venus were more risky because they tended to have a high density of people who stayed for long periods.
Some of the less risky venues including car dealerships, gas stations and hardware stores, the study found.
Overall, the researchers found that limiting venue occupancy to 20% of maximum capacity reduced predicted infections by more than 80%, while only reducing overall visits to these venues by 42%.
The study findings also help explain why minority and low-income populations have been hit particularly hard by COVID-19.
During lockdowns, people in lower-income neighborhoods did not reduce their mobility as much as those in more affluent neighborhoods, likely because they had jobs that did not allow them to work from home. In addition, venues such as grocery stores in lower-income areas had higher COVID-19 transmission rates because these venues were smaller and more crowded, and people stayed there longer, compared with similar venues in more affluent areas, the study found.
These findings suggest ways for policymakers to reduce disparities in COVID-19 infection rates, for example, with occupancy caps to reduce crowding, in addition to improved paid leave policies so workers can stay home when sick, the authors said.
"Our results suggest that infection disparities are not the unavoidable consequence of factors that are difficult to address in the short term, like differences in preexisting conditions; on the contrary, short-term policy decisions can substantially affect infection outcomes by altering the overall amount of mobility allowed and the types of [venues] reopened," the authors said.
The authors note their model did not include all public locations, in particular it did not include schools or offices. In addition, because the study used data from the beginning of the pandemic, the findings do not necessarily apply to COVID-19 transmission today. For example, people are more likely to practice social distancing and wear masks at restaurants today compared with March. And health officials have noted that many new outbreaks in the U.S. are being driven by small gatherings in people's homes.
Still, the researchers hope their findings can be used by policymakers to help guide reopenings. They are currently working on a tool to make their model accessible for policymakers and public health officials.
Originally published on Live Science.