The Shapes of Pandemic Curves

Getting monthly data for even 1957 and 1968 is tough sledding. We finally found in the CDC Vital Statistics Archives monthly data for these pandemics. Monthly data for surrounding years is available for 1968, but only 1957 and 58 for that period. Influenza is a separate line immediately above pneumonia. Our interest was the shapes of the curves, so rather than get into the modern morass of adding flu and pneumonic deaths and looking for “excess”, we report here only deaths ascribed solely to flu.

Corona deaths are the crude totals from Worldometers and there has only been one complete monthly bin so far. CDC will eventually parse these crude deaths into flu and pneumonic, likely reducing the trend. Germany recently reduced its corona deaths nearly 50%. It appears corona will be worse than ’57 and ’68, but nowhere near as bad as 1918.

We show here San Francisco data as a proxy for the entire US based on the following:

We got monthly data by tracing in Autocad according to the monthly grid and apportioning the areas under the curve. We originally were going to use Boston as more representative, but population data proved difficult and SF had a nice round 500k population in 1918.

The graph above is for all causes of deaths. We had previously determined that the shapes of the curves for all causes in ’57 and ’68 closely matched the influenza only curves. On this basis, we apportioned our SF data to match 670k US flu and pneumonia deaths and multiplied it by the ratio of flu only to pneumonia deaths in 1968 (.074) to get flu only deaths in 1918. We then subtracted 1000 to roughly align the start points.

Obviously this is not ideal, but it is at least a rational basis for comparing the shapes.

There is a notable similarity of the shapes, reaching initial peaks in about a month and a half. It is also notable that the fall from initial highs seems much faster than IHME projects.

We have shown above only the second and third peaks of the widely circulated triple whammy from England shown above. US data for the first wave is dismal. A handful of deaths at an army base and nearby towns. The CDC offers this curve as anecdotal without axis data.There is an interesting paper (Olsen et al 2005) arguing for a US beginning in New York as early as 1916.

Sobering image of a ravaged lung from a soldier who died from the 1918 flu from Smithsonian Magazine:

We now feel trying to force everything to actual deaths for influenza only was an error that distorts the true historical perspective, so we make this addendum April 12. Below we have converted everything to deaths per capita and added the IHME model in monthly bins.


From this perspective, it appears very unlikely that Corona will ever make the big time in the US. be about the same as 68-69. Current data suggests that US cases and deaths are at or very near their peaks, somewhat earlier than IHME prediction.

A separate axis is required for 1918. It is a completely different animal. US population was 1/3 current, and deaths were six times higher. Let’s not go there, ever.

Note* We made a spreadsheet error from prior dithering with virus only that incorrectly reduced the IHME model. The graphic and text were modified 4-16-20.

This entry was posted in Pandemic. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.