American cities that implemented the most aggressive social distancing measures in 1918 succeeded in reducing the number of casualties during the main peak of the influenza pandemic. But in the medium term, these cities were much more sensitive to the next waves and strains of influenza – and they ended up with a higher number of deaths.

That is the central finding of new research by Guillaume Chapelle (CY Cergy Paris Université). His study suggests that by limiting the diffusion of the initial strain of influenza, stronger non-pharmaceutical interventions made populations more sensitive to the subsequent seasonal flu. This has potential implications for today’s struggle with Covid-19: policy-makers should prepare exit strategies from lockdowns, school closures and other restrictive measures with vaccination campaigns that may compensate for lower immunity.

More…

The Covid-19 crisis raises many questions. Past pandemics such as the 1918 influenza offer an opportunity to evaluate the potential effect of non-pharmaceutical interventions (NPIs) such as school closures and social distancing on the spread of diseases and mortality, not only during the pandemic but also afterwards.

The 1918 influenza is key in epidemiological research on the impact of NPIs. Investigating their impact in US cities, the new study highlights two important facts:

• First, in the short run, cities that implemented the strongest NPIs in 1918 managed to ‘flatten the curve’ and thus reduced the number of casualties during the main peak of the pandemic.
• But second, in the medium run, cities that adopted the most aggressive policies in the autumn of 1918 became much more sensitive to the next waves and strains of influenza, and ended with a higher number of deaths. By limiting the diffusion of the initial strain of influenza, NPIs probably made those cities’ populations more sensitive to the subsequent seasonal flu.

The author gathers new data series on yearly and monthly mortality at the city level from archives digitized by the Centre for Control and Disease Prevention. He combines them with data on NPIs for 43 cities previously compiled by epidemiologists. It is then possible to investigate the evolution of various mortality indicators from 1911 to 1924, comparing cities that fought the pandemic more aggressively in 1918 with those that implemented milder responses.

First, the results show that both groups of cities had similar mortality dynamics before the pandemic. Second, aggressive NPIs are clearly associated with a flattening of the epidemic curve in the autumn of 1918 and thus managed to lower the number of deaths.

According to the author’s estimates, a city of 100,000 inhabitants with an average intensity of NPIs avoided about 170 deaths in October 1918. But cities that responded more aggressively to the pandemic saw a divergence in their level of mortality from 1919. If this increase in mortality remains relatively modest (about 100 yearly deaths in the previous example), this phenomenon rapidly offsets the benefits of NPIs, resulting in a higher number of deaths in more aggressive cities.

This medium-term impact might be explained by immunity mechanisms put forward in medical research, in particular by the fact that the first influenza virus that emerged in a region is key to understanding people’s lifelong immunity response to influenza. By reducing exposure to the initial strain of influenza, NPIs increased the sensitivity of people to the disease. This reduced their life expectancy, in particular when they were afflicted by co-morbidities such as cancer, heart diseases and tuberculosis.

To conclude, while the case of the 1918 influenza does not deny the short-term benefits of NPIs that might lower the peak of the pandemic and prevent overcrowding of the healthcare system, it does warn of their potential risk in the medium run. Thus, whenever implementing NPIs, policy-makers should prepare exit strategies as vaccination campaigns that should be able to compensate for the lower immunity when repealing these policies.

‘The Medium-term Impact of Non-Pharmaceutical interventions: Evidence from the 1918 Influenza in US Cities’

Authors:

Guillaume Chapelle (CY Cergy Paris Université)