Predictive Analytics Offers Insight into COVID-19 Spread, Disparities

November 12, 2020 - A predictive analytics model examined human mobility patterns in major cities across the US and revealed that non-residential locations – such as restaurants, gyms, and cafes – account for high rates of COVID-19 spread, according to a study published in Nature.

For more coronavirus updates, visit our resource page, updated twice daily by Xtelligent Healthcare Media.

The study also revealed that because minority or low-income people leave home more often for their jobs and tend to shop at smaller, more crowded establishments, these population groups are contracting the virus at higher rates, leading to exacerbated health disparities.

Researchers developed the model by collecting data for ten US cities, including New York, Los Angeles, Chicago, Dallas, Atlanta, Miami, Philadelphia, and others.

The team received data showing which of 533,000 public locations people visited each day, for how long, and the square footage of each establishment so that the group could determine the hourly occupancy density.

READ MORE: Predictive Analytics Tool Helps Fill Gaps in COVID-19 Data

The team analyzed data from March 8 to May 9 in two distinct phases. In phase one, they fed their model mobility data and developed their system to calculate the transmission of the virus under a range of different circumstances in ten cities.

While it’s impossible to know in real life when and where an infectious and susceptible person come into contact to create a new infection, the model can estimate the likelihood of infectious events at different places and times.

The tool was also able to solve for the unknown variables because researchers fed it information on how many COVID-19 infections were reported to health officials in each city each day.

The team refined the model until it could determine the transmission rate of the virus in each city. The rate varied from city to city, depending on factors ranging from how often people ventured out of the house to which types of locations they visited.

After obtaining transmission rates for the ten cities, the group moved to phase two, which involved testing the model by asking it to multiply the rate for each city against their database of mobility patterns to predict new COVID-19 infections.

READ MORE: 3 Ways Healthcare is Using Predictive Analytics to Combat COVID-19

The results showed that the model’s predictions matched closely with actual reports from health officials, demonstrating the model’s reliability and accuracy.

“We built a computer model to analyze how people of different demographic backgrounds, and from different neighborhoods, visit different types of places that are more or less crowded,” said Jure Leskovec, the Stanford computer scientist who led the effort, which involved researchers from Northwestern University.

“Based on all of this, we could predict the likelihood of new infections occurring at any given place or time.”

Researchers also combined their model with demographic data available from a database of 57,000 census block groups – 600 to 3,000 person neighborhoods – to show how minority and low-income people leave home more often because their jobs require them to, and tend to shop at smaller, more crowded stores than people with higher incomes who can work from home and use home delivery to avoid shopping.

The study showed that it’s about twice as risky for non-white populations to buy groceries compared to whites.

READ MORE: Predictive Analytics Model Forecasts Future Flu Outbreaks

These findings shine a new light on the significant health disparities experienced by low-income and minority populations throughout the pandemic, researchers noted.

“In the past, these disparities have been assumed to be driven by preexisting conditions and unequal access to healthcare, whereas our model suggests that mobility patterns also help drive these disproportionate risks,” said study co-author David Grusky, a professor of sociology at Stanford’s School of Humanities and Sciences.

The results suggest that reopening business with lower occupancy gaps tend to benefit disadvantaged populations the most.

“Because the places that employ minority and low-income people are often smaller and more crowded, occupancy caps on reopened stores can lower the risks they face. We have a responsibility to build reopening plans that eliminate — or at least reduce — the disparities that current practices are creating,” Grusky said.

The model’s specificity could help officials reduce the spread of COVID-19 by showing the tradeoffs between new infections and lost sales if establishments open at 20 percent or 50 percent capacity.

“By merging mobility, demographic and epidemiological datasets, we were able to use our model to analyze the effectiveness and equity of different reopening policies,” said Serina Chang, a Stanford PhD student and member of the study team.

The study offers strong evidence that the stay-at-home orders enacted this past spring effectively slowed the rate of new infections. The predictive analytics model could help inform new or similar policies going forward.

“In principle, anyone can use this model to understand the consequences of different stay-at-home and business closure policy decisions,” said Leskovec.