Why you should be wary of comparisons of the death tolls from the 1918 flu pandemic and COVID-19

This week, the number of people who have died of COVID-19 in the U.S. surpassed the 675,000 said to have died in the 1918-19 influenza pandemic.

As of Wednesday morning, 675,071 people in the U.S. had died of COVID-19, according to the Centers for Disease Control; news that prompted some to proclaim COVID-19 now the deadliest pandemic in U.S. history.

But there’s a century of time, plus advances in technology, recordkeeping and medical treatment between 1918 and the COVID-19 pandemic that make the claim at least a little dubious.

The terrible W

The influenza pandemic of 1918-19 was caused by an H1N1 strain of flu virus and was first discovered in a military facility in Kansas. It was — and still often is — misleadingly referred to as the “Spanish flu” because cases were more widely reported in Spain — neutral in World War I — as opposed to other countries that censored news of the virus’ spread. The flu swept across the world in waves between 1918 and 1919.

This flu was exceptionally deadly, and compared to other flus — and notably, COVID-19 — commonly killed young people.

“The pandemic in 1918-19 had a mortality graph that epidemiologists call ‘the terrible W’ — it killed a lot of very young people — babies and very young children; it killed a lot of very old people, and then it killed a lot of people who were sort of in their late 20s,”  said Susan Jones, a professor and historian at the University of Minnesota who specializes in the historical ecology of disease. The average age of death was 28.

 

Source: "1918 Influenza: the Mother of All Pandemics," by Armed Forces Institute of Pathology researchers and published in the CDC's Emerging Infectious Diseases.

A sad footnote to those young-adult deaths: most of them were caused by bacterial pneumonia. Today, that would be treated with antibiotics and would be unlikely to lead to death. But antibiotics weren’t discovered until a decade after the flu pandemic.

Comparing death rates

Just how deadly was the 1918-19 flu pandemic?

The U.S. death count of 675,000 accounted for roughly 0.64 percent of the country’s population of 103 million people at the time.

With a death toll of 675,071 so far, COVID-19 has killed about 0.2 percent of the 328.2 million people in the U.S. today.

In Minnesota, the estimated death toll of the 1918-19 pandemic was more than 10,000, which works out to roughly 0.4 percent of the state’s population at the time, compared to 8,025 deaths due to COVID-19 reported to-date, or 0.14 percent of the state’s population.

But the national death toll of 675,000 should be taken with a grain of salt, Jones said.

It’s not just that they didn’t have testing that could confirm whether or not someone who died had died of the flu a century ago, Jones said.  (Actually, though, because many of the deceased were young people who succumbed to bacterial pneumonia caused by the virus — not a common way for the young to die — flu deaths were often pretty evident.)

Another issue? The 675,000 number is based on an extrapolation from incomplete data.

Otis Historical Archives, National Museum of Health and Medicine

Soldiers from Fort Riley, Kansas, ill with Spanish flu at a hospital ward at Camp Funston.

“Mortality statistics collected in the United States in 1918-19 only covered about 75 [percent] to 80 percent of the population. It covered about 30 states and the District of Columbia,” Jones said. Most of the states that did have organized health departments collecting mortality data were in the North and the Midwest. Swaths of the South and West weren’t collecting these statistics.

The person who came up with the 675,000 number took death counts from the states that reported — about 549,000, and added 25 percent to approximate a count for the missing states.

Exclusion of those states in the mortality numbers means specific populations are not well-represented in the official death count.

“Probably over half of African Americans in this country lived in the states that were not part of the mortality registry,” Jones said. Another group probably underrepresented? Indigenous people, including many living in Minnesota.

“We tend to think of mortality statistics today as being absolute numbers that we can really count on and that came from actual data collection. That’s not the case with the 675,000 number, so I guess my point is, we make these things into important numbers — we make these milestones,” Jones said. “Even if we take that number as a pretty good estimate, of course it underestimates what was happening in certain communities.”