There is a trio of research papers from 2007 that look at how U.S. cities responded to the the Spanish Flu pandemic of 1918-1919. They provide interesting insights into how Seattle and other cities might respond today as the coronavirus sweeps across the country.
There are more than a few similarities between the Spanish Flu pandemic, which struck in multiple waves in 1918 through 1920, and the COVID-19 virus making the rounds now. Both are respiratory viruses and are suspected to be zoonotic in origin. Both started offshore and were imported into the country: Spanish Flu started on the East Coast and worked its way west, while COVID-19 seems to have made its big entrance along the West Coast (and particularly here in Washington) and is now starting to show up on the East Coast. Both viruses have similar patterns of transmission, symptoms and mortality. And it turns out that 100 years ago the public health systems did a pretty good job of record-keeping on deaths and the press did an equally good job of recording government actions to respond to the pandemic.
There are, to be sure, some differences as well. There was no air travel, though there were trains and ships that were well used as long-distance transportation systems, as well as streetcars and other means of getting around cities. Also, governments were more at liberty to impose draconian measures without regard to infringements on civil liberties. Also, 100 years ago families and households were larger, many people lived in crowded boarding houses, children spent fewer years as full-time students, and malnutrition was more widespread. And today there is a broader understanding of germ theory and how diseases are spread, as well as improved basic hygiene. Nevertheless, there are apt comparisons to be made.
The three papers (here, here and here) catalogue the “non-pharmaceutical interventions” (NPIs) taken by local governments to try to stop the spread of the Spanish Flu. Here are some of the findings from the papers, which for the most part independently come to the same conclusions:
- There was large variability across cities in their effectiveness in controlling the spread of the virus — even among cities that used the same measures. Not all of it can be explained.
- Imposing measures early in the outbreak clearly reduced the mortality rate, particularly in terms of delaying and/or reducing the peak mortality rate as measured on a week-to-week basis.
- Certain NPIs were more effective than others: those include closing schools, churches and theaters, and banning public gatherings. Some turned out not to be as effective as hoped, including closing dance halls, bans on public funerals, and isolation/quarantine of those diagnosed with the virus (perhaps because the real danger is when an individual is contagious but is not yet showing symptoms).
- However, certain combinations or “layers” of NPIs (though not all combinations) were found to be even more effective — even combinations that included a measure that wasn’t effective on its own. For example, closing schools and isolation/quarantine together were extremely effective, even more so when other ancillary measures (e.g. face masks, altering work schedules, limited business closures, transportation restrictions) were added on top. But there were many other combinations tried by cities that were not nearly as effective.
- Timing of the introduction of the NPIs was very important, though for a complex mix of reasons. Early introduction had a significant effect, but later weeks, once the virus had established itself at a certain level, had less consistent results — sometimes effective, sometimes not.
- The mortality rate, measured on a week-to-week basis, turned out to be an important threshold for inspiring citizens to self-impose social distancing measures. In other words: once they could see enough people were dying, people got serious about doing whatever it took to stop the virus from spreading further. As tragic as the deaths at the LifeCare facility in Kirkland are, the silver lining is that it has made the threat of COVID-19 all too real for the Seattle area an motivated people to action.
- In some, but not all, cities, there was a second wave of the virus several months after the first wave (and in some there was even a third wave). There are competing hypotheses about why this happened only in a subset of cities. There is evidence that cities that lifted their NPIs too early saw a second wave, perhaps because there was still a substantial population that had not contracted the virus and there wasn’t a big enough group who had contracted it and recovered to create an effective “herd immunity.” There is also a suggestion that it’s possible for a city to impose NPIs that are too strong, quickly damping down the spread of the virus but leaving the city more susceptible to successor waves either from low-level latent cases that remain or re-infection from outside the city once restrictions are lifted. In 1918-1919, cities typically imposed NPIs for 2 to 8 weeks.
- It appears that early intervention with NPIs leads to lower weekly mortality rates, but not necessarily lower total mortality over the course of the pandemic. A more muted response might make the weekly statistics look worse, but might cause the disease to burn itself out more effectively and save lives in the end. Reasons for this could relate to the “second wave” arguments, as well as the suggestion that a higher mortality rate inspires people to take individual action to help prevent further spread. However, that’s a difficult sell — both ethically and as public policy.
We come away from these studies with a very messy picture of the challenge for public officials in deciding how to respond to the COVID-19 outbreak. They need to right-size the response: enough to save lives and slow the spread, but not so strong as to make future waves of resurgence inevitable. They need to respond early and quickly, but still ensure that people see and believe that the threat of the virus is real enough to take personal action. They need to delay the spread of the virus to buy time for the public health response, including treatment facilities, to mobilize. And officials need to decide how long to sustain the interventions, despite the substantial economic and political impacts. That was a challenge that officials faced in 1918 as well, and it looks like China’s leaders will be facing the same issue in the weeks to come: as a result of their draconian measures the COVID-19 outbreak in China seems to have peaked at around 80,000 cases — in a nation of 1.2 billion people. But it’s also severely disrupted their economy, and there will be enormous pressure to lift their NPIs before the virus is eradicated in the rest of the world (and before a vaccine is available in 12-18 months). Even if China completely eradicates the virus inside the country, it will be exposed to the risk of re-introduction from other countries. It appears unlikely that China will avoid a second wave of COVID-19.
Let’s hope Seattle, and the rest of the United States, fare better. There are lessons here for the many cities and states in the U.S. that still have no reported cases: they shouldn’t be waiting for the virus to show up to start preparing their response — or to take action. In 1918 several cities didn’t wait and were well served by being proactive.
In case you’re interested, the CPC maintains a list of research studies on non-pharmaceutical interventions for disease outbreaks.
I hope you found this article valuable. If you did, please take a moment to make a contribution to support my ongoing work. Thanks!