By adding behavioral components to an infectious disease model, researchers at Brown University have developed a new modeling approach that captures the peaks and valleys of the new COVID-19 cases seen over the past 16 months.
The approach, published in the journal Scientific reports, could be useful in predicting future trends in the current pandemic, as well as in predicting the course of future ones.
“We know that people’s behavior is important in terms of how an infection spreads,” said Vikas Srivastava, Brown’s assistant professor of engineering and principal investigator of the research. “We wanted to see if we could quantify these behavioral aspects and incorporate them into one model and see if this model is able to capture waves of infection rates that we saw in the United States and elsewhere. “
A commonly used method to model the transmission of infectious diseases is what is known as the SIR model. The approach separates a population into different categories: susceptible, infected, and recovered. The model moves people from one category to another according to two parameters. The transmissibility of the disease, along with the speed with which people come into contact, predict how quickly people go from being susceptible to being infected. The recovery rate moves infected people to recovered ones. (“Recovered” in these models usually means “no longer contagious,” which also includes those who have died from the infection).
The standard SIR model produces a curve with a maximum peak: what infectious disease experts urged people to flatten out through social distancing, masks, and other measures that reduce virus transmission. But over the past 16 months, real case rates in individual states, in the United States as a whole, and in other countries have not produced a single curve. Instead, they produced multiple waves of infection that created a significant challenge for the community of infectious disease models, Srivastava says.
As the pandemic unfolded, Srivastava was teaching a class that included a section on modeling infectious diseases at Brown’s School of Engineering. He and his students were surprised to see the mismatch between model predictions and actual case rates.
“We saw cases going up and down, creating multiple peaks, but the models were not able to capture it,” Srivastava said. “That’s what made us think about using population behavior and response as a way to explain and predict what’s going on.”
Srivastava worked with two Brown undergraduate students, Zachary LaJoie and Thomas Usherwood, to develop the new modeling strategy. They modified a standard SIR model to include the effects of vaccination and also added two behavioral parameters to the model. The first, “level of caution,” estimates people’s tendency toward safe behavior (social distancing, wearing masks, and other safety measures) as reported cases increase. The parameter also captures government actions in response to rising case rates, such as closures and quarantines, which increase safe behavior. A second parameter, the “sense of security,” models people’s confidence in returning to pre-pandemic activities as more people are vaccinated.
The team then used an optimization algorithm to calibrate the values of the new parameters based on the case rates reported in the US. With the parameters optimized, the team found that the model accurately reproduced case rates throughout the pandemic in the United States as a whole and in individual states and cities.
“When we looked at New York City, for example, we saw an increase in our level of caution at the same time that government action went into effect in late March,” LaJoie said. “Then, as the cases went down, we saw the precautionary level go down and there was another increase in cases going on vacation.”
If the model fits the data correctly, it provides information on how the pandemic may develop in the future. For example, the team was able to measure how different vaccine uptake rates could affect case rates. Increasing vaccination rates could lead to a reduction in cases, but it could also reduce precautionary levels and increase the sense of security among unvaccinated people. This could put upward pressure on case rates, even when vaccines reduce them. In fact, the model predicts scenarios in which there are brief periods of rising infection rates as vaccines are established, before they eventually end up declining again.
In the U.S., for example, the model captures a short period of growing infection in mid-April before rates dropped again. The largest rises in places like India look similar to the more extreme post-vaccination climbs predicted by the model. With its current vaccination rates, the model predicts that cases in the US should approach zero by August 2021.
According to researchers, these ideas may be useful in places where vaccination programs are beginning.
“When we developed the model, we focused on the United States, but it would certainly be useful to make predictions in other places like India, Europe, or South America, where case rates are still quite high,” he said. LaJoie.
The modeling approach could also be applied to future outbreaks or pandemics.
“There’s really nothing in this model that limits it to just COVID-19 as a disease,” Usherwood said. “We believe this has applicability to any situation where people’s behavior is important, which is basically any infectious disease.”
Thomas Usherwood et al, A model and predictions for COVID-19 taking into account population behavior and vaccination, Scientific reports (2021). DOI: 10.1038 / s41598-021-91514-7
Citation: The new model explains the effect of behavioral changes to predict cases of COVID-19 (2021, June 11) retrieved June 12, 2021 at https://medicalxpress.com/news/2021-06-accounts -effect-behavior-covid-cases .html
This document is subject to copyright. Apart from any fair treatment for private study or research purposes, no part may be reproduced without written permission. Content is provided for informational purposes only.