Lessons on social distancing from the 1918 flu pandemic: St. Louis shut down public gatherings quickly. Philadelphia held a parade and saw death rates spike

Lipsitch is an epidemiology professor and in 2007, he was one of the authors of a study that looked at public health interventions during the 1918 influenza pandemic. At the time he and his co-authors were writing the paper, there was concern about how the world would handle the next pandemic. Because a vaccine and effective antiviral medications were likely not going to be “widely available,” they wanted to measure the effectiveness of “nonpharmaceutical interventions” — bans on public gatherings, school closures, and other acts of social distancing. Lipsitch and his colleagues looked at American cities and how they reacted as Spanish influenza hit their communities in the autumn of 1918. They found that cities that acted quickly to shut down schools, churches, and ban social gatherings had peak death rates about 50 per cent lower than cities that didn’t and had “less-steep epidemic curves.”

Early intervention, “before it feels bad,” spares the health system from intense stress “like Philly vs. St. Louis,” Lipsitch wrote on Twitter.

Philly vs. St. Louis has become a shorthand in epidemiology circles these days. You might have seen the chart comparing the excess death rates during late 1918. St. Louis closed down the social life of its city two days after the first case of the virus was reported. By the time Philadelphia took similar steps, 16 days had passed since the city’s first reported case. It was too late. Philadelphia had an exponential spike in excess death in mid-October 1918, while St. Louis showed a much lower curve.

Authorities in Philadelphia didn’t make much fuss at first. “Many sufferers not severely attacked and doctors see no cause for fear,” read one early headline after the city’s first reported cases on Sept. 17, 1918. Spanish influenza, the papers said, had made hundreds of sailors sick at the city’s Navy yard and authorities were isolating the sick and urging people not to participate in “promiscuous sneezing.” There were no rules about avoiding large crowds.

In late September, as stories about entire families wiped out by the virus appeared in the news, the city held a patriotic parade to raise money for war loans. It was “the mightiest and the most beautiful, the most solemn and the most compelling of the pageants Philadelphia yet has had,” the newspaper gushed, with no hint of alarm. It was expected that 10,000 would show up, “but the city did it honour by bringing twenty times that number into the streets.” By the next week, Oct. 3, Philadelphia announced theatre and schools closures and other bans on public gatherings. But the virus had already grown unwieldy.

Hospitals overflowed and morgues couldn’t handle the bodies: “Philadelphia had the highest death rate of any major American city during the pandemic,” The Mutter Museum of the College of Physicians of Philadelphia notes on its online exhibit about the pandemic. “More than 12,000 people died in six weeks; over 20,000 died in six months.”

St. Louis, Missouri was about 1,400 km west, and had the advantage of warning. St. Louis officials watched the epidemic spread with “alarming rapidity” on the eastern seaboard. Two days after the first reported cases in St. Louis in early October, the city’s restrictions were front-page news. Spanish influenza — “A malignant infectious and contagious disease” was already here. With a disease that was “quick in its ravages,” public health officials figured it would become an epidemic, so they enacted rules to stop “the toll which this disease has already begun to exact on our city.” The city’s health commissioner and mayor ordered all theatres, schools, moving pictures shows, billiard halls, churches, Sunday schools, cabarets, lodges, societies, public funerals, open-air meetings, dance halls and conventions closed until further notice. The mayor knew there would be economic loss and inconvenience, but the situation was “so alarming” he had to act.

The 2007 study confirmed that St. Louis’s early intervention had its desired effect. In Philadelphia, the peak weekly excess death rate was 257 deaths per 100,000 people. In St. Louis, when restrictions were in place, the peak excess death rate was 31 deaths per 100,000 people.

Most cities relaxed their restrictions around the end of October and into November. The virus had not gone away, and a second wave of the epidemic occurred. From early September to the end of December, the cumulative excess death rate was 719 per 100,000 people in Philadelphia and 347 per 100,000 people in St. Louis. The analysis in the study suggested that aggressively enacting interventions “resulted in flatter epidemic curves and a trend toward better overall outcomes in the fall of 1918.”

Dr. Jeff Kwong, an epidemiologist and professor at the University of Toronto, was struck by that second wave that hit cities after the restrictions were lifted in 1918. He says there is no question that social distancing measures can and should be enacted — a good body of evidence, and recent examples in other countries like China, show that they’ve been able to dampen the COVID-19 outbreak. He just wonders about “kicking the can down the road.”

“We can choose to have our pain early or we can choose to have it later, but the pain is coming,” Kwong said. “I think the only question is, how much of the overall mortality can we reduce through these social distancing measures?”

The study found that overall, cities with early and aggressive restrictions “showed a trend toward lower cumulative excess mortality,” through late 1918, but the difference was smaller and less statistically significant than the data for peak death rates.

The study’s authors underscored “the need for prompt action by public health authorities” and “the need for very rapid interventions to stem the spread of the disease.” The day after Lipsitch made his point about Boston not following Philadelphia’s disastrous parade example, the Boston parade was cancelled. Toronto’s parade was also cancelled this week, joining a growing list of shuttered events and institutions as North American cities take precautions to limit the spread of COVID-19 and flatten the curve of the epidemic.

Kwong says avoiding sharp spikes in illness will help the health-care system. Simply put, if a hypothetical hospital only has 10 ventilators, it is better to have cases spread out over time, as opposed to what is happening in Italy right now. While many people are working on vaccines and antiviral medications, Kwong believes a vaccine for COVID-19 could be a reality in a year or two, if we’re lucky. In the meantime, he doesn’t know what the recommendations should be when it comes to closures and restrictions in the long term. “The problem is weighing what is practical,” he says. “All I do know, that I can say with a fair bit of certainty, is that we’re not talking about weeks here, not even months. This could be going on for a few years, so we just have to prepare ourselves for this new reality.”

As the New York Times noted in a recent article, while the public concern around the virus’ spread is similar to 1918, medically, we are worlds apart. Researchers in China have already sequenced the genome of COVID-19, but in 1918, the electron microscope did not exist. As the Times reported, “It was impossible to test people with mild symptoms so they could self-quarantine. And it was nearly impossible to do contact tracing because the flu seemed to infect — and panic — entire cities and communities all at once.”

The 1918 flu pandemic was a “totally different virus,” Kwong says, and it targeted a different demographic. “In 1918, there were a lot of young and healthy adults getting very sick and dying from it, versus now, it is very old people who are getting the most sick, and people with chronic conditions.”