Asymptomatic coronavirus infections contribute to over 50% of spread, according to UChicago study
Asymptomatic transmission of COVID-19 contributed significantly to community spread in New York City during the initial phase of the pandemic, according to a new paper from a team of researchers at the University of Chicago. The research, published in the Proceedings of the National Academy of Sciences on February 10, reinforces the continued importance for all people, regardless of symptom status, to follow public health guidance to curb spread of the coronavirus.
In the first mathematical model to incorporate data on daily changes in testing capacity, the research team found that only 14% to 20% of COVID-19 individuals showed symptoms of the disease and that more than 50% of community transmission was from asymptomatic and pre-symptomatic cases.
In emerging diseases, researchers seek to understand outbreaks by estimating epidemiological parameters like the proportion of cases that become symptomatic and which types of infections transmit disease. However, depending on how much and the type of data available, these parameters can be either precisely estimated or have considerable uncertainty.
The investigators initially examined data about the coronavirus outbreak in Wuhan, China, in hopes of bringing a new perspective to modeling of the virus’ spread, but soon found the data lacking.
“Testing data from Wuhan was rather sparse and not having a lot of information can make estimating these epidemiological parameters difficult,” said first author Rahul Subramanian, a PhD student in epidemiology. “It was hard to get good information about COVID-19 spread just from looking at Wuhan, so we found data being reported by New York City and adapted our models.”
The authors fit their model to daily changes in testing capacity reported by New York City, becoming the first peer-reviewed model to explicitly incorporate this data. “We brought together mathematical models and surveillance testing data to add something new to what has already been a very active field of research,” said Mercedes Pascual, PhD, the Louis Block Professor of Ecology and Evolution at the University of Chicago.
“Without taking into account testing capacity, you can’t distinguish between cases that were not reported because they were not symptomatic and cases that were not reported because of a lack of testing capacity,” said Subramanian.
In addition to considering daily testing, the authors also incorporated estimates of herd immunity from an antibody survey by Mount Sinai Hospital, allowing them to estimate the number of symptomatic cases that were underreported and the true number of asymptomatic cases.
“By combining the case data, the serology data, and the testing data, we were actually able to estimate the proportion of cases that were symptomatic, which was quite exciting,” said Subramanian.
The investigators also report that more than half of community transmission is from non-symptomatic cases – either asymptomatic or pre-symptomatic cases. “While we can’t estimate precisely how likely it is that an asymptomatic case will transmit disease, non-symptomatic cases as a whole contribute significantly to community transmission,” said Subramanian. This finding may become important as public health officials are deciding which COVID-19 restrictions to implement, maintain, or roll back. “Whatever measures policymakers implement to control the ongoing outbreak, they also need to include pre-symptomatic and asymptomatic people as well.”
The researchers additionally estimate that the reproductive number or the average number of new infections that will be caused by a currently infected person can be significantly higher than the range of 2-3 that is commonly reported in the literature. Restrictions to curb COVID-19 spread may need to be adapted to account for a higher reproductive number.
Though this research used data from the spring 2020 New York City outbreak, it is likely to be applicable to new coronavirus variants that have begun circulating in the U.S.
“The core findings regarding transmission are not likely to change with new variants,” said Subramanian. “If anything, we are providing a better baseline to compare these new variants to. Researchers can understand more precisely how these changes in new variants affect transmission.”
Further modeling of the COVID-19 pandemic is likely to benefit from continuing to incorporate testing data, but only if researchers can access it. “Making testing data and testing protocols systematically available is key,” said Pascual. “There is valuable information out there that can connect models to data but is not accessible to the researchers doing the modeling.”
The study, “Quantifying Asymptomatic Infection and Transmission of COVID-19 in New York City using Observed Cases, Serology and Testing Capacity,” was supported by the National Science Foundation (1735359).
In the first mathematical model to incorporate data on daily changes in testing capacity, the research team found that only 14% to 20% of COVID-19 individuals showed symptoms of the disease and that more than 50% of community transmission was from asymptomatic and pre-symptomatic cases.
In emerging diseases, researchers seek to understand outbreaks by estimating epidemiological parameters like the proportion of cases that become symptomatic and which types of infections transmit disease. However, depending on how much and the type of data available, these parameters can be either precisely estimated or have considerable uncertainty.
The investigators initially examined data about the coronavirus outbreak in Wuhan, China, in hopes of bringing a new perspective to modeling of the virus’ spread, but soon found the data lacking.
“Testing data from Wuhan was rather sparse and not having a lot of information can make estimating these epidemiological parameters difficult,” said first author Rahul Subramanian, a PhD student in epidemiology. “It was hard to get good information about COVID-19 spread just from looking at Wuhan, so we found data being reported by New York City and adapted our models.”
The authors fit their model to daily changes in testing capacity reported by New York City, becoming the first peer-reviewed model to explicitly incorporate this data. “We brought together mathematical models and surveillance testing data to add something new to what has already been a very active field of research,” said Mercedes Pascual, PhD, the Louis Block Professor of Ecology and Evolution at the University of Chicago.
“Without taking into account testing capacity, you can’t distinguish between cases that were not reported because they were not symptomatic and cases that were not reported because of a lack of testing capacity,” said Subramanian.
In addition to considering daily testing, the authors also incorporated estimates of herd immunity from an antibody survey by Mount Sinai Hospital, allowing them to estimate the number of symptomatic cases that were underreported and the true number of asymptomatic cases.
“By combining the case data, the serology data, and the testing data, we were actually able to estimate the proportion of cases that were symptomatic, which was quite exciting,” said Subramanian.
The investigators also report that more than half of community transmission is from non-symptomatic cases – either asymptomatic or pre-symptomatic cases. “While we can’t estimate precisely how likely it is that an asymptomatic case will transmit disease, non-symptomatic cases as a whole contribute significantly to community transmission,” said Subramanian. This finding may become important as public health officials are deciding which COVID-19 restrictions to implement, maintain, or roll back. “Whatever measures policymakers implement to control the ongoing outbreak, they also need to include pre-symptomatic and asymptomatic people as well.”
The researchers additionally estimate that the reproductive number or the average number of new infections that will be caused by a currently infected person can be significantly higher than the range of 2-3 that is commonly reported in the literature. Restrictions to curb COVID-19 spread may need to be adapted to account for a higher reproductive number.
Though this research used data from the spring 2020 New York City outbreak, it is likely to be applicable to new coronavirus variants that have begun circulating in the U.S.
“The core findings regarding transmission are not likely to change with new variants,” said Subramanian. “If anything, we are providing a better baseline to compare these new variants to. Researchers can understand more precisely how these changes in new variants affect transmission.”
Further modeling of the COVID-19 pandemic is likely to benefit from continuing to incorporate testing data, but only if researchers can access it. “Making testing data and testing protocols systematically available is key,” said Pascual. “There is valuable information out there that can connect models to data but is not accessible to the researchers doing the modeling.”
The study, “Quantifying Asymptomatic Infection and Transmission of COVID-19 in New York City using Observed Cases, Serology and Testing Capacity,” was supported by the National Science Foundation (1735359).