Can someone post the article?
@aygart has constantly pointed out that excess deaths is the best measure. Otherwise, there are arguments to be made that the death toll is higher or lower (depending on what your agenda is)
Most Countries Fail to Capture Extent of Covid-19 Deaths
Many leave the untested off coronavirus tallies; Belgium doesn’t, despite the outcry
By Daniel Michaels
Updated May 28, 2020 6:11 am ET
A growing pool of global death statistics indicates that few countries are accurately capturing fatalities from the new coronavirus—and in some the shortfall is significant.
In the U.S., Russia, the U.K., the Netherlands and many other countries, the number of deaths recorded from all causes has jumped since March and far exceeded the number of deaths those countries report as linked to Covid-19, the disease caused by the coronavirus.
Belgium, which appears to have the world’s highest per-capita Covid-19 death rate, has emerged as an exception. Unlike most countries and many U.S. states, which list only Covid-19 deaths confirmed by tests, Belgium also tallies suspected Covid-19 fatalities.
The approach sparked public criticism in Belgium last month for putting the country in a bad light, but France, the U.K., New York state and other jurisdictions have shifted toward the same methodology as Belgium.
If we had only taken confirmed deaths, we would not have had such accurate data,” said Françoise Renard, a scientist with Belgium’s national health institute, Sciensano.
As countries reopen from lockdowns that have crippled economies, reliable information on the pandemic’s impact is vital to guide health policies. Accurate death data, which indicate the virus’s lethality, “can provide us with critical information on the underlying risk factors,” said a World Health Organization spokesperson.
Lauren Gardner, a Johns Hopkins University associate engineering professor who leads the team that built the school’s widely cited Covid-19 tracker, believes countries including the U.S. and China have broadly undercounted cases and deaths. She cited Belgium as among the rare places where the number of Covid-19-related deaths has closely tracked this year’s surge in excess mortality.
During a pandemic, determining what caused a death can prove difficult—especially so with Covid-19, since many people who have died also suffered from other serious conditions.
The WHO and centers for disease control in the U.S. and Europe have recommended that any death in which Covid-19 was deemed to be a contributing factor—based either on test results or medical judgment—should be listed as a Covid-19 death.
Still, most countries list Covid-19-related deaths mainly for cases where the novel coronavirus was detected with a diagnostic test, either while the person was alive or shortly after death. Since testing remains limited and sporadic world-wide, many Covid-19 deaths don’t get reported as linked to the new coronavirus.
Meanwhile, the total number of deaths has surged in many areas since the pandemic struck, suggesting a link.
“When you get dramatically large spikes, it’s almost certain that most are coronavirus,” said Colin Mathers, a retired coordinator of the WHO’s Mortality and Health Analysis Unit. While death is a lagging indicator of the situation, since people die days or weeks after contracting the virus, “total deaths certainly give you a readout of whether things are getting better or worse,” he said.
Including suspected deaths can have consequences. Belgium by mid-April reported more Covid-19 deaths per capita than any other country with a population above one million people. President Trump, at a White House coronavirus briefing on April 18 in which he said he saw “positive signs that the virus has passed its peak,” displayed a chart with Belgium at the top and the U.S. near the bottom.
Belgian Health Minister Maggie De Block, seeking to blunt public outcry, the same week proposed publishing a second set of data with only confirmed coronavirus deaths, to compare better with other countries. Steven Van Gucht, a virologist at Belgium’s National Crisis Center, countered that since “testing capacity is inadequate, it’s very important to include suspected cases” or deaths would be undercounted.
As more mortality data emerged later last month, the curve of Belgium’s total-deaths chart closely tracked that of its Covid-19 deaths.
“We asked them to wait a little bit,” said Ms. Renard at Sciensano of politicians. With the fuller data, “it was evident it was the right way to calculate.”
Determining what is a Covid-19 death can require judgments by doctors unable to test patients. The WHO, the Centers for Disease Control and Prevention and other agencies offer guidance, but some policy makers have said the uncertainty risks overestimating the death toll from the new coronavirus.
Many medical experts see little risk of that. “We’re almost sure there’s undercounting, not overcounting” in the U.S., said Robert Anderson, chief of the mortality statistics branch at the CDC’s National Center for Health Statistics.
The role of doctors’ judgment isn’t new: only a fraction of U.S. disease deaths prompt an autopsy, yet all list a cause of death. Judgments are critical now as increased understanding of the coronavirus has linked it to a growing list of deadly conditions such as heart attacks and blood clots that weren’t initially seen as connected. Massachusetts is looking back at death certificates from the beginning of the year to see if there are cases that should be considered as probable Covid-19 deaths, despite the lack of test results.
The measure of accuracy, say specialists, is the gap between total recorded deaths from all causes and those linked to Covid-19. The gap should be similar to a country’s expected number of deaths for a given week. All countries with good records expect a certain number of weekly deaths based on trends from recent years. If the gap widens—as has happened in most countries—it means unexplained deaths, which these days are likely Covid-19 deaths.
Dr. Anderson at the CDC cited the week of April 11, when the U.S. recorded roughly 21,000 deaths more than expected and registered 18,000 Covid-19-related deaths. The other 3,000, he believes, were either directly due to Covid-19 or were indirectly related, such as a death of someone who wasn’t infected but couldn’t get care for a heart attack or some other event because hospitals and emergency medical services were overwhelmed.
The gap is evident in Michigan, where death certificates show total deaths have surged above average mortality levels but recorded Covid-19-related deaths have risen more slowly.
Ms. Gardner at Johns Hopkins said tracking probable Covid-19 deaths has helped fill in the gap.
“In those places where they do report those, there is a way smaller unexplained gap,” Ms. Gardner said.
Belgium, with a population similar to Michigan, has a very small unexplained gap. One reason: Since March it has tracked deaths both in nursing homes and hospitals, and always included presumed Covid-19 deaths.
Over half of Belgium’s Covid-19 deaths have been in nursing homes, which were ravaged by the disease. For many weeks, less than 5% of those were confirmed by testing, while most hospital deaths were tested—meaning almost half of the country’s Covid-19 deaths were reported based on medical judgment.
Belgian health officials felt comfortable making judgments about nursing -home deaths because they have long tracked viral outbreaks in the country’s 1,600 elderly homes. France, the U.K. and U.S. states including New York only recorded Covid-19 deaths in nursing homes several weeks after they started seeing cases.
Belgium has reported roughly 80 Covid-19 deaths per 100,000 residents. That compares with 55 in the U.K. and roughly 30 in the U.S., countries where death records suggest Covid-19 links in many more cases.
Ms. Renard said Belgium has been hit hard but probably doesn’t lead the world.
“We are among the countries with the highest mortality,” she said. “But I wouldn’t say we are first.”
—Jon Kamp and Betsy McKay contributed to this article.
Write to Daniel Michaels at dan.michaels@wsj.com
