Trump Misleads on Reasons for Falling COVID-19 Fatality Rate

For nearly two months, President Donald Trump has touted an 85% decline in the nation’s COVID-19 case fatality rate since April — and has attributed the drop to improvements in treatment. But better treatment is only part of the story.

Experts say part, if not most, of the decline can be explained by expanded testing and a shift toward younger people — rather than higher-risk older folks — catching the coronavirus.

Since the end of July, Trump has wielded the impressive-sounding statistic in press briefings and rallies, playing up the effectiveness of existing therapies and claiming credit for those discoveries.

“Due to the medical advances we’ve already achieved and our increased knowledge in how to treat the virus,” he said in a July 27 briefing in North Carolina, “the mortality rate for patients over the age of 18 is 85% lower than it was in April — think of that: 85%.”

In early August, Trump began mentioning specific drugs, even though several of them have yet to be proven to be effective in reducing COVID-19 mortality — and knowledge about another came from a British clinical trial.

“We’re developing a bounty of therapies such as remdesivir, dexamethasone, antibody treatments — the antibody treatments are really working out well, really well — and many more that have allowed us to reduce mortality by 85% since April,” he said in remarks at a Whirlpool plant in Ohio on Aug. 6.

In rallies and other speeches throughout September, the mortality drop has been one of his go-to lines. “Through pioneering therapies, we reduced the fatality rate 85% since April,” Trump has said, with slight variation, more than 10 times, often mentioning the antiviral drug remdesivir and convalescent plasma.

In an interview with Fox News host Jeanine Pirro, Trump went so far as to suggest that current treatments are so good that they’re “pretty close to a cure.”

Trump, Sept. 12: We also have already come out with therapeutics, if you look at remdesivir and others, the plasma. You take a look at the things that we’ve come out with already and the number, we’re 85% better in terms of fatalities. You look at the kind of numbers, I don’t know if you’ve been seeing this, but we’re in — we’re really rounding the corner. And this is actually without anything further than we already have but we have some very good things out there already in terms of — I would say in terms of — you can’t use the word cure yet but pretty close to a cure. You’re going to be pretty close to a cure.

In fact, few proven therapies exist for COVID-19.

“We are nowhere close [to] a cure,” Dr. Eric Meyerowitz, an infectious disease specialist with the Montefiore Health System in the Bronx, told us in an email, “though we are certainly better off than we were in early March at the beginnings of the major surges in New York City and Boston.”

As we’ll explain, Trump’s statistic is a reference to the case fatality rate, which is an imperfect measure of mortality, as it is highly dependent on diagnostic testing. While he’s right on its precipitous drop, he gives the inaccurate impression that those gains are entirely due to treatment advances.

Increased testing and a greater proportion of infections among younger individuals are significant factors.

“More effective medical management may be playing a role in the falling case fatality rate in the US, but I suspect this improvement has been small relative to the large decrease,” said University of Pennsylvania infectious disease fellow Dr. Aaron Richterman in an email.

Case Fatality Rate

Neither the White House nor the Trump campaign responded to repeated inquiries about the source of 85% statistic, but our colleagues at PolitiFact were told by the White House that the president was referring to the case fatality rate — and were directed to a graph made by Dr. Jay Bhattacharya, a professor of medicine and an economist at Stanford University.

The case fatality rate, or CFR, is the percentage of deaths from the coronavirus among the confirmed cases.

In the graph, which plots the CFR on a weekly basis using a two-week lag to accommodate delayed deaths, the CFR reaches nearly 30% in mid- to late-March, before falling to about 10% on April 1 and steadily declining to less than 2% by late July. 

It’s a bit unclear if Trump’s specific statistic is derived from the graph, although the approximate numbers are consistent with an 85-90% decline between the beginning of April and late July. Bhattacharya told us he shared his graph with some colleagues, but didn’t know how the White House came to be aware of it.

We arrived at a similar result looking through a COVID-19 dataset from the Centers for Disease Control and Prevention. Using the ratio of COVID-19 deaths to probable and suspected COVID-19 cases, we calculated a crude CFR of 7.7% for all ages for the month of April. That dropped 83% to 1.3% in the month of July and further slipped to 0.9% for the month of August, for a total decline of 88% since April. (Note: The cumulative CFR remains much higher, at 3.4%, as it is predominated by the larger number of deaths early in the pandemic; the figures can be found using the visualization tool.)

Regardless of the exact data, Trump is correct that the case fatality rate has fallen substantially since April. Where he goes wrong is in attributing the steep decline to treatment improvements when other factors are also at play.

We’ve noted before the challenges in interpreting the CFR. Because it’s just a raw calculation of the number of observed deaths divided by the number of identified cases, it’s not an accurate estimate of the percentage of people who are infected with the coronavirus die, since it’s hard to know the true number of infections. 

In the beginning of the pandemic, when testing was still extremely limited, the U.S. was not capturing many of the less severe COVID-19 cases, which invariably led to an artificially high CFR. As testing expanded, more and more of those cases were included.

“In the early days, only the sickest cases were getting diagnosed, so unsurprisingly a higher % of these died relative to now, when many other cases are identified,” explained Richterman.

During the month of April, only about 5.3 million tests were performed, versus more than 23 million each in the months of July and August, according to the COVID Tracking Project.

It also matters who is getting sick, and that has also changed over time. Richterman noted that early on, many outbreaks occurred in nursing homes — places with elderly folks with underlying conditions — where COVID-19 mortality is very high.

“Now, the epidemic is being propagated more in younger people, people living in congregate settings, people doing work in close quarters with inability to socially distance, etc.,” he said. “Many of these people have an inherently lower risk of death.”

Indeed, the CDC published a report on Sept. 23 describing the changing age distribution among those infected with the coronavirus, noting that in the early months “COVID-19 incidence was highest among older adults” but that shifted to people ages 20 to 29 over the summer. That younger cohort accounted for more than 20% of cases from June to August. 

Not Yet ‘Close to a Cure’

Improvements in treatment are likely part of the reason for the declining CFR — but not the only one, as Trump suggests.

Bhattacharya, who has said the scientific community has “overreacted” to the pandemic, told us there have been “substantial improvements” in COVID-19 treatment since the start of the pandemic, including “improved protocols for ventilator use for people with severe viral pneumonia, and the availability of drugs like remdesivir and dexamethasone which have definitely improved outcomes.”

But he also said other contributors to the falling CFR were the shift toward lower-risk younger people and the fact that more widespread testing is picking up mild or asymptomatic coronavirus cases.

“I do not know of a study that … systematically decomposes these three contributors to declining CFR,” he said. “All three are important and the fact that the CFR picks up changes in all three makes it a statistic worth tracking.”

In his view, the high CFR numbers reported early on “spread fear and panic,” and he added that “[i]t seems the responsible thing to do to report that the CFR has dropped since those early days.”

Other experts were less certain of the impact of improvements in treatment.

“The mainstay of treatment for severe COVID-19 is excellent supportive care in an intensive care unit, and this has not fundamentally changed over time,” said Richterman. “That said, in areas experiencing very large epidemics (New York City comes to mind), [it’s] plausible that overwhelmed health systems were unable to provide the best possible supportive care early on in the pandemic.”

Tara C. Smith, an epidemiologist at Kent State University’s College of Public Health, also told PolitiFact that she suspects what matters most for the declining CFR are increases in testing and the switch to more infections in younger people — and that drug advances would be “a much smaller piece of the pie.”

Richterman noted that in terms of specific therapies, only two evidence-based treatments are available: remdesivir and the steroid dexamethasone.

“While both might reduce risk of death, only dexamethasone has been proven to do so, and this would account for only a moderate reduction in risk of death relative to the large drop in case fatality rate,” he said.

“We have a long way to go in our treatment,” he added, contradicting the president’s rosy assessment of current COVID-19 therapies.

Although Trump frames the dexamethasone finding as an American development, it was the RECOVERY randomized controlled trial in the U.K. that first identified a mortality benefit of using the drug in select hospitalized COVID-19 patients. The finding was announced on June 16, well after the U.S.’s CFR began to fall.

To be sure, some American doctors used steroids such as dexamethasone prior to the study results, but until that point, whether the medication would hurt or help patients was unclear. And the study revealed that the drug did not help patients who did not need supplemental oxygen, and may have harmed them.

Remdesivir, which is made by the U.S. company Gilead Sciences, has not yet shown that it helps COVID-19 patients survive. The drug received an emergency use authorization from the Food and Drug Administration for those with severe COVID-19 on June 1 based on a study that found the drug shortened the time to recovery among hospitalized patients.

The study’s results hint at a possible survival advantage with the drug: the 14-day mortality was 7.1% among those given remdesivir versus 11.9% for those who received placebo. But Meyerowitz said the finding “did not meet statistical significance, so with the data available there is no clear mortality benefit for remdesivir for hospitalized patients.”

If additional data show that those relative mortality figures hold, he added, then those getting remdesivir would be about 30% less likely to die.

As for convalescent plasma, although some of the results are promising, it too has not yet been proven to reduce COVID-19 mortality, as we’ve written before.

Following the FDA’s EUA for plasma for hospitalized COVID-19 patients on Aug. 23, the National Institutes of Health’s COVID-19 Treatment Guidelines Panel issued a statement explaining that there is “no data from well-controlled, adequately powered randomized clinical trials” that show the therapy is safe and effective for COVID-19.

The statement noted that the FDA’s analysis of the Mayo Clinic’s Expanded Access Program for convalescent plasma, which was at the core of the EUA decision, did not find a difference in seven-day survival among those given high- versus low-concentration plasma. Instead, a survival benefit was only observed in a subset of patients who were not intubated, with 11% of those receiving antibody-rich plasma dying within a week compared with 14% of patients given low-concentration plasma.

Based on the data, the panel neither recommends for nor recommends against the use of plasma.

Editor’s Note: Please consider a donation to FactCheck.org. We do not accept advertising. We rely on grants and individual donations from people like you. Credit card donations may be made through our “Donate” page. If you prefer to give by check, send to: FactCheck.org, Annenberg Public Policy Center, 202 S. 36th St., Philadelphia, PA 19104.