applied,” but “did some good … when and where they were rigidly carried out” (Soper, undated draft report).

Even if isolation only slowed the virus, it had some value. One of the more interesting epidemiologic findings in 1918 was that the later in the second wave someone got sick, the less likely he or she was to die, and the more mild the illness was likely to be.

This was true in terms of how late in the second wave the virus struck a given area, and, more curiously, it was also true within an area. That is, cities struck later tended to suffer less, and individuals in a given city struck later also tended to suffer less. Thus west coast American cities, hit later, had lower death rates than east coast cities, and Australia, which was not hit by the second wave until 1919, had the lowest death rate of any developed country.

Again, more curiously, someone who got sick 4 days into an outbreak in one place was more likely to develop a viral pneumonia that progressed to ARDS than someone who got sick 4 weeks into the outbreak in the same place. They were also more likely to develop a secondary bacterial pneumonia, and to die from it.

The best data on this comes from the U.S. Army. Of the Army’s 20 largest cantonments, in the first five affected, roughly 20 percent of all soldiers with influenza developed pneumonia. Of those, 37.3 percent died (Soper, 1918; undated draft report).

In the last five camps affected—on average 3 weeks later—only 7.1 percent of influenza victims developed pneumonia. Only 17.8 percent of the soldiers who developed pneumonia died (Soper, 1918).

Inside each camp the same trend held true. Soldiers struck down early died at much higher rates than soldiers in the same camp struck down late.

Similarly, the first cities struck—Boston, Baltimore, Pittsburgh, Philadelphia, Louisville, New York, New Orleans, and smaller cities hit at the same time—all suffered grievously. But in those same places, the people struck by influenza later in the epidemic were not becoming as ill, and were not dying at the same rate, as those struck in the first 2 to 3 weeks.

Cities struck later in the epidemic also usually had lower mortality rates. One of the most careful epidemiologic studies of the epidemic was conducted in Connecticut. The investigator noted that “one factor that appeared to affect the mortality rate was proximity in time to the original outbreak at New London, the point at which the disease was first introduced into Connecticut…. The virus was most virulent or most readily communicable when it first reached the state, and thereafter became generally attenuated” (Thompson and Thompson, 1934a: 215).

The same pattern held true throughout the country and the world. It was not a rigid predictor. The virus was never completely consistent. But places hit later tended to suffer less.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement