Literature Review: Covid-19 Series. If Manikins Could Fly (Eikenberry, et al, 2020)

Article Under Review

Steffen E. Eikenberry, Marina Mancuso, Enahoro Iboi, Tin Phan, Keenan Eikenberry, Yang Kuang, Eric Kostelich, Abba B. Gumel

 To mask or not to mask: Modeling the potential for face mask use by the general public to curtail the COVID-19 pandemic ( Infectious Disease Modelling 5 (2020) pp. 293-308)

This is the second review in a series of articles regarding the efficacy of community-wide mask-wearing.

The Eikenberry article is very well written, tremendously useful, and very well thought-out.  The work by the Eikenberry team is near the pinnacle of academic achievement and their hard work is commendable. 

Contrary to the title, the article is not really about masks; the article is about models. The models thus developed by the Eikenberry team is not just a marvelous mathematical tool, it's a thing of beauty.

The excellence of the model notwithstanding, it rather reminds of a wonderfully complex and elegant model I saw in the 1980s that clearly demonstrated why bumblebees can't fly.  Regarding that particular model, one certainly couldn't argue against the model from a mathematical or aerodynamic perspective, and as long as one ignored the bumblebees flitting about outside, the model was a standalone example of mathematical excellence.

At the heart of the Eikenberry model, the authors explore possible nonlinearities in mask coverage and effectiveness and the interaction of these two parameters during the initial stages of an outbreak of an infectious illness at various values of R0, and mask coverages, and capturing efficiencies (with both inward and outward considerations).

Like the model that demonstrates the impossibility of bumblebee flight, there are no real problems with the model - it is a work of art (and mathematical prowess). 

The model is very "clean."  If we took a million identical manikins and placed them in a single, large, environmentally controlled room, and placed an identical mask on each, and observed how an infectious agent may move through the population, this model would probably predict the outcome very nicely.

Manikins don't move. Manikins don't wear bandanas, crochet masks, masks made from old T-shirts, or the lovely knitted mask made by their Grandmother manikins.  Manikins don't sport beards or wear dentures and never hang their dirty mask on the rear view mirror of their automobiles. Manikins don't wear their masks under their nose, or lift their mask to facilitate a sneeze.  Manikins don't fidget and constantly touch their mask, and then touch surfaces such as credit cards, doorknobs, or money.  Manikins are well behaved and never flush toilets and they don't worry about aerosolizing droplets after going poo.  Manikins don't remove their masks to kiss their kids or wives, and seldom remove their masks to cook dinner or drink water.  In fact, manikins don't ever remove their masks.  Manikins don't wear clothes and never worry about secondary surface contacts.  Manikins never touch their nose.  Therefore, unlike humans, manikins are ideal subjects when trying to model the effects of mask-wearing on the progression of an infectious particle through a population.  Except for this small problem, the model does a fine job - even the bumblebees would approve.

The authors, who conclude that their models suggests that that community mask wearing may possibly "bend the curve," are quiet cognizant of the above limitation and forthrightly address that very fact:

 Our theoretical results still must be interpreted with caution, owing to a combination of potentially high rates of noncompliance with mask use in the community, uncertainty with respect to the intrinsic effectiveness of (especially homemade) masks at blocking respiratory droplets and/or aerosols, and even surprising amounts of uncertainty regarding the basic mechanisms for respiratory infection transmission.

The authors pitted their model against observations from New York City and the State of Washington, and they were quite pleased with the comparisons. The authors don't mention Japan, Nigeria, Sweden, or the multitude of other real life societies that have been affected by the pandemic in a manner that is not quite like New York City and for which the model has unknown predictive value. 

 The ultimate conclusion of the authors can be found in the very first sentence of the article:

 Face mask use by the general public for limiting the spread of the COVID-19 pandemic is controversial, though increasingly recommended, and the potential of this intervention is not well understood.

One simply cannot find fault with the model or the excellent article which was written by modelers for modelers with a keen insight into very good models.

 To quote the authors:

Shakespeare could not have parsed it better and you'll find no disagreement from me.