A small minority of places where people go frequently account for a large majority of coronavirus infections in big cities, according to a new modeling study.
The study, published in the journal Nature on Tuesday, suggests that reducing the maximum occupancy in such places -- including restaurants, gyms, cafes and hotels -- can slow the spread of illness substantially.
"Our model predicts that capping points-of-interest at 20% of maximum occupancy can reduce the infections by more than 80%, but we only lose around 40% of the visits when compared to a fully reopening with usual maximum occupancy," Jure Leskovec, an author of the study and associate professor of computer science at Stanford University, said during a press briefing on Tuesday.
"Our work highlights that it doesn't have to be all or nothing," he said.
The model also found significant racial and socioeconomic inequities in coronavirus infections.
Model spotlights potential 'superspreader' placesThe researchers -- from Stanford University and Northwestern University -- used cell phone location data from SafeGraph to model the potential spread of Covid-19 within 10 of the largest metropolitan areas in the United States: Atlanta, Chicago, Dallas, Houston, Los Angeles, Miami, New York, Philadelphia, San Francisco and Washington DC.
The data, representing the hourly movements of 98 million people, included mobility patterns from March to May.
The researchers examined Covid-19 case counts for each area and took a close look at how often people traveled to certain non-residential locations or "points-of-interest."
Those locations included grocery stores, fitness centers, cafes and snack bars, doctor's offices, religious establishments, hotels and motels and full-service restaurants.
"On average across metro areas, full-service restaurants, gyms, hotels, cafes, religious organizations, and limited-service restaurants produced the largest predicted increases in infections when reopened," the researchers wrote in their study.
The model predicted that "infections are happening very unevenly -- that there are about 10% of points-of-interest that account for over 80% of all infections, and these are places that are smaller, more crowded and people dwell there longer," Leskovec said during Tuesday's briefing.
The model also predicted that people living in neighborhoods with the lowest income, based on Census data, were more likely to have been infected -- driven in part by how places in those areas tended to be smaller in size, leading to crowding and increasing the risk of spread.
"Our model predicts that one visit to a grocery store is twice more dangerous for a lower-income individual compared to a higher-income individual," Leskovec said. "This is because of grocery stores visited by lower-income individuals have on average 60% more people by square foot, and visitors stay there 17% longer."
The study comes with limitations, including that the model is a simulation -- not a real-life experiment -- and the data are based on 10 metropolitan areas and do not capture all places someone could frequent, such as schools, nursing homes and prisons, which also have been associated with Covid-19 outbreaks. More research is needed to determine whether similar findings would emerge among other populations and places.
A deeper understanding of disparitiesThe study's findings "make intuitive sense," Julian Tang, of the University of Leicester in the United Kingdom, said in a written statement distributed by the UK-based Science Media Centre on Tuesday.
"The results reported from this study may not be that surprising -- and they make intuitive sense -- unfortunately highlighting again how this virus exploits socio-economic disparities across populations," Tang, who was not involved in the new study, said in part.
Kevin C. Ma and Marc Lipsitch, both of the Harvard T. H. Chan School of Public Health, co-wrote an editorial that published alongside the new study in the journal Nature on Tuesday.
The maximum occupancy approach in the new study "can be extended to evaluate other types of reopening strategy. For example, time-limited visits to places such as gyms and museums could be modelled by decreasing the average visit length," Ma and Lipsitch wrote.
They added that the new study also "deepens our understanding" of possible factors behind disparities in Covid-19 cases by income level.
"By combining their mobility model with demographic census data, the authors identified two possible main contributors. The first is that lower-income neighbourhoods, which tend to have higher numbers of front-line workers, had less overall reduction in mobility during lockdowns than did higher-income neighbourhoods, a conclusion shared by other studies," Ma and Lipsitch wrote about the new study.
"The second possible contributor is that, across many types of setting, the venues visited by people from lower-income neighbourhoods tend to be more crowded than are the venues visited by those from higher-income areas," they wrote. "This type of observation is possible only because of the high level of detail on venue size and occupancy in the data."
Overall, Ma and Lipsitch wrote, "Further model testing is needed, but given the challenges in gathering and interpreting other relevant data types, these findings could have a valuable role in guiding policy decisions on how to reopen society safely and minimize the harm caused by movement restrictions."