On Thursday, Missouri’s state health department reported 30 new COVID-19 deaths, the highest daily toll so far during the pandemic.
The total number of deaths reached 661, more than double the total of 329 at the end of April. Other indicators, meanwhile, suggest the coronavirus may have a greater impact than state numbers indicate.
The COVID-19 data released by the Missouri Department of Health and Senior Services includes numbers of people who’ve died of the disease as reported by doctors, medical examiners, hospitals and local health agencies.
But health data experts, including Virginia Tech professor of statistics Ron Fricker, say such tallies likely miss deaths that are never attributed to COVID-19.
“There’s a suspicion that, in fact, people are getting the disease, not going to the hospital, but then passing away — maybe because of comorbidities — but they are less likely to be identified as having COVID and therefore may be undercounted,” Fricker says.
Many COVID-19 deaths that take place in people’s homes or in nursing homes may be misdiagnosed as being the result of other ailments, such as the flu.
As the weeks go on and more data becomes available, however, different theories are emerging to suggest deaths may be much higher than the numbers indicated by official reports.
Health data experts say a more complete indicator of COVID-19's toll may be found in excess deaths reported by the federal Centers for Disease Control and Prevention.
Generally, 2020 has seen fewer deaths than recent years, which is likely because large numbers of people across the country have been staying at home.
However, the CDC has reported surges of deaths in many states this spring that surpass averages from previous years and are thought to be the result of COVID-19 fatalities.
This theory, Fricker says, has been bolstered by research in Massachusetts, a state with strong public health reporting, where many excess deaths were reported early in the pandemic.
“They saw exactly that phenomenon,” Fricker says. “Total number of deaths were up, and there was some sense that the cause was attributable to COVID-19.”
Missouri had 603 excess deaths between the week ending on April 11 and the week ending May 2, according to CDC estimates, which are weighted to account for reporting delays and provisional data.
However, the state health department reported a total of 353 COVID-19 deaths as of May 2, suggesting that the state may have missed 250 deaths from COVID-19 in its tally, undercounting the full number by more than 40%
Fricker acknowledges that the excess death measurement is an inexact tool because the deaths could be the result of numerous causes.
Doctors think that many people may be avoiding health care and even emergency services due to fears of contracting COVID-19, and this may be leading to deaths that might not occur in other years.
Though excess deaths are, at best, a rough tool for determining COVID-19 deaths, Fricker says they can offer a glimpse of the extent of the spread of the virus, which is especially important given the low rates of testing.
Studies on COVID-19 outbreaks in several areas have estimated that the disease generally causes a mortality rate of roughly 1% among people who have been infected (although some experts think this may also be an underestimate).
Applying that 1% figure to Missouri would suggest 35,300 cases in the state, based on Missouri’s official total of 353 deaths as of May 2. The state's official estimated case tally, however, is 8,154.
If the larger number of COVID-19 deaths estimated from excess death data is correct, then Missouri’s total number of cases as of May 2 would be 60,300 — more than seven times the reported number.
And there may be more excess deaths in Missouri. The CDC has not estimated excess deaths for the state since May 2. This may be due to lags in death reporting, which are common.
Even with the jump in reported deaths on Thursday, Fricker explains that it may still be weeks before there’s adequate data to understand the current impact of the disease.
“We will learn things about the virus from two or three or four weeks ago, but not about what’s going on today,” Fricker says. “So all we can really say is, yes, some historical evidence say that the virus is in a location, or it’s growing or not growing, but that doesn’t tell you much about today, which is what we would most care about.”