This time can be different. Learning from the experience of the "Spanish" flu pandemic
Professor of the Online Master in Health Economics and Pharmacoeconomics at UPF Barcelona School of Management
One of the few certainties we have is that the health and economic impact of the pandemic depends not only on epidemiological factors, but also on the behaviour and responses of governments, populations and organizations.
The global spread of the new coronavirus (COVID-19) is infecting not only people but also economic activity and financial markets globally. One of the few certainties we have is that the health and economic impact of the pandemic depends not only on epidemiological factors, but also on the behaviour and responses of governments, populations and organizations. The impact on both population mortality and the mortality of productive and financial organizations and domestic economies will depend on the public and private management of this double health and economic crisis.
Until recently, the study of pandemics, beyond epidemiologists, was only of interest to historians, with little attention, if not almost forgotten, by macroeconomic experts. One of the notable exceptions has been Robert Barro from Harvard University, who has studied the 1918-1920 influenza pandemic (H1N1 virus), so called the "Spanish" flu.
It would be a poor excuse to argue now that pandemics were a thing of the past (as were business cycles before 2008?). It is enough to name the Singapore flu of 1957, the Hong Kong flu of 1968 and the avian influenza A virus (H1N1) of 2009 among the most prominent, not to mention Ebola, the SARS-Cov coronavirus or the MERS-Cov.
The "Spanish" flu of a century ago could give us an idea of the worst scenario of COVID-19 and that is why, despite the differences, some lessons can be drawn. Since the incubation period for COVID-19 (5-6 days) is longer than for influenza, compliance must be sustained for months, with openings and closings that can be repeated.
The "Spanish" flu of a century ago could give us an idea of the worst scenario of COVID-19 and that is why, despite the differences, some lessons can be drawn. This pandemic had three main waves between 1918 -coinciding with the final phase of the First World War- and 1919, with the second one causing the most deaths. The influential recent study from Imperial College on March 16 also suggests that there may be several waves with COVID-19. According to Barro, the mortality rate for 1918-1920 has been estimated at 2% of the world population, with very important variations between countries: India (rate of 5.2%, with 43% of all deaths in the world), United States (0.52%), and Spain (1.36%). Unlike COVID-19, mortality was higher among individuals between 18 and 40 years old and without previous disease, and higher in men in this age group, with devastating effects on household economies. Robert Barro has also estimated that the reduction in GDP in this period due to the pandemic was 6%, with a greater contraction in consumption, with a subsequent rapid recovery but without conclusive evidence on the temporality or permanence of part of the effects.
The need to apply non-pharmacological interventions as soon as possible was already learned in 1918, and this was done by the United States Army in its military camps. However, there were no significant differences in morbidity and mortality between the camps that applied it and those that did not. The reason lies in the degree of compliance with the measures: only a small group applied the isolation measures strictly. In the cholera epidemic in Hamburg in 1892 there were riots against the imposition of quarantines and isolation. To be successful, individuals must follow them and do so over time. In Mexico City, mask use during the 2009-2010 influenza A (H1N1) epidemic peaked at 65% and decreased to 10% after ten days. Since the incubation period for COVID-19 (5-6 days) is longer than for influenza, compliance must be sustained for months, with openings and closings that can be repeated. It is necessary to provide effective measures to mitigate population fatigue and economic mortality.
In 1918, American cities attempted to control the spread of the pandemic with various non-pharmacological interventions, ranging from wearing surgical masks to measures of social distancing with the closing of schools and churches, the prohibition of public gatherings, and the isolation of the sick, which have been studied as a natural experiment. American cities that previously adopted containment measures performed better. Philadelphia had the first cases on the 17th of September, 1918, but allowed a large military parade on September 28th and delayed social distancing measures until October 3rd. St. Louis had the first cases on October 5th and took social distancing measures just two days later. Philadelphia had a spike in the death rate attributable to the pandemic 8 times higher than St. Louis. Did these interventions also reduce the cumulative mortality of the pandemic? Three studies for different American cities published in PNAS and JAMA in 2007 showed that when the adoption of multiple measures, greater intensity of the intervention, was very early, the cumulative mortality was slightly lower (20% on average). It had a moderate impact due to that few cities maintained the measures for more than 6 weeks, although the reduction reached 30-50% of excess mortality in the cities that were most successful.
Economic shock measures are required also to flatten the economic mortality for which we have less evidence than for non-pharmacological interventions.
The COVID-19 pandemic is different from the "Spanish" flu in epidemiological terms and in the context of health systems and economies, so any extrapolation to today should be done with caution. Flattening the epidemiological curve to avoid exceeding the threshold of the capacity of the health system is effective. Instead of having the entire healthcare system collapsed and lacking the appropriate resources for a significant part of those who need intensive cures in a short period of time corresponding to the most acute phase of the disease, if we flatten the curve we will have a longer healthcare collapse but at a lower level with less excess mortality. It is about buying time, waiting for a vaccine, pan-coronavirus if possible, and the production of antivirals and medical supplies on a large scale, and meanwhile increasing the response capacity of the healthcare system with more resources and applying valid technical and clinical diagnostic tests to whom and when necessary. Flattening the epidemiological curve should also help mitigate the economic impact, but economic shock measures are required also to flatten the economic mortality for which we have less evidence than for non-pharmacological interventions.
© Jaume Puig-Junoy