In defence of armchair epidemiology
OK, “Flatten the curve of armchair epidemiology” was very funny; and “Ten Considerations Before You Create Another Chart About COVID-19” makes some important points about avoiding both panic and indifference; as does Slate’s “Stop the epidemic of armchair epidemiology”.
But we armchair epidemiologists, we unsophisticated sirens of social media, Excel crusaders and lackadaisical luminaries of LinkedIn; we too have a role to play.
He who can does. He who can not teaches
(The section headings in this article are all taken from George Bernard Shaw’s Maxims for Revolutionists.)
Professional epidemiologists are busy building models to understand how quickly COVID-19 spreads, what measures will work, how long we will need them; how many hospital beds we will need and how many respirators.
In the meantime, the world’s population is enduring the excruciating economic repercussions of the momentous measures these models tell us are better than the biblical alternative of letting the contagion rage.
With so many of the public at large at the mercy of the mandate of a handful of mathematical models, the public understanding of those models has never been more essential.
Pedagogic epidemiology is a civic duty.
A learned man is an idler who kills time with study. Beware of his false knowledge: it is more dangerous than ignorance
In my article “The Pains of Epidemiology”, I describe developments to the very simple model (SIR) I use in my “Disease Dynamics Distilled”, where I explain some of the terms we hear so much about in the media: exponential growth, flattening the curve, herd immunity, etc.
The purpose of describing these developments is to introduce and begin to explain some of the models that are becoming influential with policy makers and doing the rounds on the internet (not, unfortunately, always in that order).
Some of these are only marginal improvements on SIR, like the SEIR model that is the basis of Gabriel Goh’s Epidemic calculator (left), which has had lot of use, amongst other places with the very influential Tomas Puyeo (“dancing” is all the rage now). Others represent the pinnacle of epidemiological probity, like the WHO’s study on clinical severity, based on data from the Wuhan outbreak.
My motivation for this presentation of more advanced modelling was to help people understand the research that was so influential in the form of their everyday lives at the moment. Then I realized an uncanny resemblance between my gimcrack Excel SEIR model and some of the curves being shown by health policy experts and suddenly the post acquired the additional motivation of spreading the word on the state of the art and why those developments are so important.
Activity is the only road to knowledge.
My appeal to my fellow armchair epidemiologists is to model, but not make models your aim. Build your own models to understand the models of others and then spread that understanding. And don’t let the naysayers get you down.
Shortly after publishing this in the Spring, Julia Gog (Professor of Mathematical Biology at Cambridge) published an article in Nature called “How you can help with COVID-19 modelling”, which is rather good.
Originally published at https://www.stochastic.dk on March 30, 2020.