[ad_1]
Government officials and policymakers have tried to use numbers to capture the impact of COVID-19. Figures like the number of hospitalizations or deaths reflect part of this burden. Each data point only tells part of the story. But no figure describes the true ubiquity of the novel coronavirus by revealing the number of people actually infected at any given time – a figure important to helping scientists understand whether herd immunity can be achieved, even with vaccinations.
Now, two scientists from the University of Washington have developed a statistical framework that integrates key data on COVID-19 – such as the number of cases and deaths from COVID-19 – to model the true prevalence of this disease in United States and individual states. Their approach, published the week of July 26 in the Proceedings of the National Academy of Sciences, projects that in the United States, up to 60% of COVID-19 cases were undetected by March 7, 2021, the latest date for which the dataset they used is available.
This framework could help authorities determine the true burden of disease in their region – both diagnosed and undiagnosed – and direct resources accordingly, the researchers said.
“There are all kinds of different data sources that we can draw on to understand the COVID-19 pandemic – the number of hospitalizations in a state or the number of tests that come back positive. But every data source has its own flaws that would give a biased picture of what’s really going on, ”said lead author Adrian Raftery, professor of sociology and statistics at UW. “What we wanted to do was develop a framework that corrects the flaws of multiple data sources and builds on their strengths to give us an idea of the prevalence of COVID-19 in a region, state or country. in its entirety.”
Data sources can be biased in different ways. For example, a widely cited COVID-19 statistic is the proportion of test results in a region or state that come back positive. But as access to testing and willingness to be tested varies by location, this figure alone cannot provide a clear picture of the prevalence of COVID-19, Raftery said.
Other statistical methods often attempt to correct for bias in a data source to model the actual prevalence of the disease in a region. For their approach, Raftery and lead author Nicholas Irons, a PhD student in statistics at UW, incorporated three factors: the number of confirmed cases of COVID-19, the number of deaths from COVID-19, and the number of tests. COVID-19 administered each. day as reported by the COVID Tracking Project. Additionally, they incorporated random COVID-19 test results from residents of Indiana and Ohio as an “anchor” for their method.
The researchers used their framework to model the prevalence of COVID-19 in the United States and individual states through March 7, 2021. As of that date, according to their framework, about 19.7% of U.S. residents, or about 65 million people, had been infected. This indicates that the United States is unlikely to achieve collective immunity without its ongoing vaccination campaign, Raftery and Irons said. In addition, the researchers found that the United States had an undercoverage factor of 2.3, which means that only about 1 in 2.3 cases of COVID-19 have been confirmed by testing. In other words, about 60% of cases were not counted at all.
This rate of underestimation of COVID-19 also varied widely from state to state and could have multiple causes, according to Irons.
“It may depend on the severity of the pandemic and the amount of testing in this condition,” Irons said. “If you have a condition with a severe pandemic but limited testing, the undercoverage can be very high and you miss the vast majority of infections that occur. Or, you could have a situation where testing is widespread and pandemic is not as severe. There, the undercoverage rate would be lower. “
Additionally, the undercoverage factor fluctuated by state or region as the pandemic progressed due to differences in access to medical care between regions, changes in testing availability, and other factors. , said Raftery.
With the true prevalence of COVID-19, Raftery and Irons calculated other figures useful for states, such as the death rate from infection, which is the percentage of infected people who succumb to COVID-19, as well as the cumulative incidence, which is the percentage of a state’s population that has had COVID-19.
Ideally, regular random testing of individuals would show the level of infection in a state, region or even nationwide, Raftery said. But during the COVID-19 pandemic, only Indiana and Ohio performed random virus tests on residents, essential data sets to help researchers develop their framework. In the absence of widespread random testing, this new method could help authorities assess the true burden of disease in this pandemic and the next.
“We believe this tool can make a difference by giving those responsible a more accurate picture of how many people are infected and what fraction of them are not being addressed by current testing and treatment efforts,” Raftery said.
The research was funded by the National Institutes of Health.
Source link
