If there’s one thing about the number of deaths from COVID-19 that researchers seem to agree with. That is, the official number might be too low.
But the extent of undercounting is a source of controversy.
That might help explain why. When the influential Seattle-based Institute for Health Measurement and Evaluation launched a new model this month. which indicates the number of deaths from COVID-19 The actual global figure is more than twice the figure from the World Health Organization. Other experts mix
Some researchers interviewed on the matter said the model looked robust. But, as many have criticized, saying the team overlooked the uncertainty contained in estimates like these and did not provide enough detail about how their statistical sausages were made.
The IHME model suggests that the extent of underreporting varies from region to region around the world. But there was an increase of nearly 7.3 million as of May 16, higher than the official total of about 3.4 million.
The IHME team also predicted the future: by September 1, according to the model, 9.1 million people will die.
in california The death toll was around 120,515 or so on May 16, doubling what the model calls the state’s reported number of 62,596.
Dr. Mark Kali, Minister of Health and Human Services of the State expressing doubts about doubling numbers
“We definitely believe that. It will be too few. because the data needs to be considered more critically,” Ghaly said, “but the idea that it is 50% real – I think a lot of us don’t believe it is correct, but [we] It takes more time to look at it.”
Dr Rochelle Walenski, director of the US Centers for Disease Control and Prevention, said the agency would review the IHME’s revised model and decide whether to count any additional deaths. Only with covid?
“We will consider this matter carefully,” she said at a briefing.
Several research groups have estimated the number of deaths expected in the absence of a pandemic. and compare those estimates with the actual number of reported deaths to create a so-called statistics. “Excess death”
Not all of the excess deaths during the pandemic were caused by SARS-CoV-2 infections, the World Health Organization said, but in some ways, the numbers were a more reliable marker of the number of deaths from the pandemic. This is partly because many locations lack the infrastructure and resources to accurately track COVID-19 deaths.
There’s a good reason why the official count might not reflect reality. the researchers said Among them: Cases may sometimes go undetected and in places with low coronavirus testing rates, deaths of older adults at the start of the pandemic may be due to other reasons. In some places, officials may not. Revealing accurate death numbers for political reasons
that’s a problem Because the correct number of deaths is critical to understanding the outbreak and predicting how it will spread. and helps policymakers weigh the exchanges between public health, economic and other priorities. more precisely as they tried to counter
James Scott, a statistician and data scientist at the University of Texas at Austin, said: “It is vital that we have a clear view of the true burden of this disease.
Dr. Timothy Brewer, an infectious disease expert and epidemiologist at UCLA, points out that the IHME model takes data from a few countries and predicts what will happen in other parts of the world. He said, “it may or may not be a reasonable assumption.”
“I don’t think we can assume that what is happening in the United States or California is the same as what is happening in Gabon, Ghana or anywhere else,” Brewer said. “I think that’s the biggest challenge I have with that.”
Nicholas Jewell, a biostatistician and epidemiologist at the London School of Hygiene and Tropical Medicine, has the same concern. Taking data from places with reliable death records and applying it to places where records are less reliable is a tricky business, he said.
“It’s an analysis that deserves a full review,” he said — one “where statisticians can assess the methods used in full detail and replicate the findings if necessary.”
But overall, Brewer expressed confidence that the model was well thought out.
“I think the way they do it is complicated. They do a great job of extracting existing data,” he said.
Ruth Etzioni, a population health scientist at the Fred Hutchinson Cancer Research Center in Seattle, said that counting only COVID-19 deaths directly underestimated the actual number of deaths. Due to the pandemic, medical care was delayed and caused other consequences. to many people She said the IHME model’s estimates seemed reasonable.
“I think it’s possible” that all deaths are doubled on the official count, Etzioni said. But being that it was much higher than reported For me, it’s something that cannot be disputed.”
for their model First released May 6, researchers at IHME estimate excessive mortality rates for various locations. Based on weekly or monthly data up to May 2nd (although it’s been updated since then) and projected to Sept. 1, they split the data into six categories:
death from direct SARS-CoV-2 infection
Deaths caused by medical treatment delayed due to pandemic
Mortality from mental health disorders such as depression, higher alcohol consumption and increased use of opium
Avoiding deaths as stay-at-home orders reduce injuries from traffic accidents and more.
Avoid death as masks and social distancing reduce the spread of other viruses. that could be life threatening including influenza and measles
Mortality from chronic diseases such as cardiovascular disease has decreased. Because people who will succumb to these conditions die of COVID-19 instead.
Researchers assessed the excess mortality at each location with statistics weekly or monthly. They removed deaths due to causes unrelated to COVID-19 and accounted for avoided deaths from the pandemic.
This information is used to create a model that applies to all areas. including data without mortality data. The results were mixed.
According to estimates as of May 10, the United States had 913,081 deaths, researchers estimated nearly 60 percent more than the 578,985 deaths collected from official reports.
India’s death toll is 737,608, nearly three times the reported death toll of 248,307. Mexico’s estimated death toll of 623,571 is nearly three times the official number of 219,925.
Although the number of deaths from COVID-19 In the Russian Federation it is slightly lower. But the discrepancy is much higher, with the IHME estimated to have 607,589 deaths from COVID-19, compared to the official number of 111,909.
The IHME model estimates 175,488 deaths for Egypt (more than 12 times the official number of 13,962) and 84,453 deaths in Kazakhstan (more than 12 times the official figure of 13,962). official at 5,810 to 14 times)
The results were met with great skepticism by many researchers outside the IHME.
Scott from the University of Texas It was one of a handful of people who expressed deep concern about the fact that the estimated death tolls are very accurate, ignoring the mathematical uncertainty about those numbers.
A model used by the IHME team to estimate deaths from COVID-19. based on several assumptions And each assumption creates a slight uncertainty in the review process, he said. It doesn’t take long to add to that uncertainty.
That’s why such numbers are often accompanied by error bars that show the degree of uncertainty. “Giving or Receiving” in Estimates That is not the case with the IHME model, although future mortality projections have error bars. But estimates of the mortality that have occurred have not been found.
“You want error bars,” Scott said.
“The error bar is what turns the back-envelope calculations into something that anyone can judge and contribute as a scientific endeavor,” he added.
Eili Klein, an epidemiologist at Johns Hopkins University, also expressed concern that the calculations were so clear.
“There must be some uncertainty in their models that they should report,” Klein said. “They’re saying this is 900,000 deaths, plus or minus two or three, or 900,000 plus or minus 500,000? I don’t know how uncertain the model is. As a result, I cannot judge the accuracy of the estimates.”
Ali Mokdad, a public health researcher at IHME and one of its senior faculty members, is leading the COVID-19 modeling effort, saying historical mortality calculations do not require large error bars. This is because estimates are based on actual numbers — the total number of deaths in a given location.
Amelia Apfel, IHME’s media relations officer, added that for now The team did not report any uncertainty about past mortality estimates. “To simplify the modeling process,” though, is something they may explore in the coming weeks.
Many independent researchers think that the IHME numbers for the United States and elsewhere are It sounds like it’s in a ballpark. Still, some say the bigger issue isn’t whether the numbers are correct, but where they come from.
In a way, Brewer said, it could be a matter of trust in a particular group and their work.
“In general, this group is a very famous group. and has worked reasonably before,” he said, according to descriptions and methods detailed online. “Looks like they’re doing their best to get it right. That doesn’t mean they got it right. But it looks like they’re trying to get it right. So I sorted give them trust or discrepancy as is.”
Mokdad said the criticism was part of the research process and he welcomed opinions from others.
“That’s normal in science,” he said. “We do new and creative things, and yes, we fix as we keep going. if there is an error I’m not worried about that.”
This story originally appeared in the Los Angeles Times.