Review of Ferguson et al “Impact of non-pharmaceutical interventions...” Chen Shen† , Nassim Nicholas Taleb∗ , Yaneer Bar-Yam† † New England Complex Systems Institute, ∗ School of Engineering, New York University Second version, March 21, 2020. Corresponding author: yaneer@necsi.edu Neil Ferguson and an Imperial College team perform de- as performed in China in Hubei province, would have been tailed simulations of outbreak response [1]. This is an im- possible instead of national lockdowns. portant work because they model social/government response, A few other issues are of importance: They ignore the not just contagion. They show suppression (lockdown so that possibility of superspreader events in gatherings by not in- R0 < 1) is essential because mitigation (R0 > 1, “flattening cluding the fat tail distribution of contagion in their model. the curve”) necessarily results in massive overload of hospitals They don’t provide details in this paper, but prior works use and many dead. This is an important conclusion that should Gamma distributions that are exponentially decaying and don’t inform policy makers. represent fat tails, i.e. subexponential class. This leads them to However, they make structural mistakes in analyzing out- deny the importance of banning them, which has been shown break response. They ignore standard Contact Tracing [2] to be incorrect, including in South Korea [7]. Cutting the fat allowing isolation of infected prior to symptoms. They also ig- tail of the infection distribution is critical to reducing R0 [8]. nore door-to-door monitoring to identify cases with symptoms The model they use, while agent based, uses assumptions [3]. Their conclusions that there will be resurgent outbreaks are inspired by standard SIR differential equation methods used wrong. After a few weeks of lockdown almost all infectious in epidemiology. For example, it does not include spatial people are identified and their contacts are isolated prior to inhomogeneity, and is therefore not well suited for incor- symptoms and cannot infect others [4]. The outbreak can be porating real world conditions at fine or large scale. These stopped completely with no resurgence as in China, where new include (1) significant interactive local dynamics and travel cases were down to one yesterday, after excluding imported restrictions missing from aggregate quantities or averages international travelers that are quarantined. across geographic locations, (2) non-normal distributions of Their assumptions are equivalent to ergodicity, as they the number of infections per person (superspreader events) as consider new infections to be a function of infected fraction well as the infection period, and (3) dynamic or stochastic and immunity, and not influenced by where in the outbreak values of parameters that arise from variations in sampling of trajectory they are, distinguishing going up from going down. distributions as well as the impact of changing social response efforts. Despite including details of the contagion and response They also don’t specify whether achieving less than one options, their model is several degrees of abstraction away case (extinction of the virus) is possible in their model. The from what is warranted by the situation. actual minimal number for resurgence is larger than 1 because While the efforts to model social response are important, (1) a significant percentage of those in close contact with leaving out critical aspects of the response yields incorrect confirmed cases are not infected, indeed only 5% of close answers. Focusing on details but using incorrect assumptions contacts of infected individuals traced in China subsequently makes for bad policy advice. Where lives are at stake, it is tested positive [2], and (2) small outbreaks can be stopped essential for science to adhere to higher standards. by contact tracing, which is enhanced by the availability of testing [5]. The availability of testing is also not included R EFERENCES in their analysis. These interventions imply the exponential [1] Ferguson et al., Impact of non-pharmaceutical interventions (NPIs) to re- growth they report after relaxing restrictions would require a duce COVID-19 mortality and healthcare demand https://www.imperial. significant number of initial cases. ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/ Imperial-College-COVID19-NPI-modelling-16-03-2020.pdf Since lockdowns result in exponentially decreasing num- [2] Report of the WHO-China Joint Mission on Coronavirus Disease 2019 bers of cases, a comparatively short amount of time can be (COVID-19) https://www.who.int/docs/default-source/coronaviruse/ who-china-joint-mission-on-covid-19-final-report.pdf sufficient to achieve pathogen extinction, after which relaxing [3] China Goes Door to Door in Wuhan, Seek- restrictions can be done without resurgence. Since the expo- ing Infections https://www.courthousenews.com/ nential decay is highly sensitive to the interventions made by china-goes-door-to-door-in-wuhan-seeking-infections/ [4] Tracking down Wuhan’s new confirmed COVID-19 cases https://www. both government and social action, simulating their effects is shine.cn/news/nation/2003023260/ less helpful than the advice to “go all out” and refine the effort [5] South Korea pioneers coronavirus drive-through over time with improved tracing, testing, and other protocols. testing station https://www.cnn.com/2020/03/02/asia/ coronavirus-drive-through-south-korea-hnk-intl/index.html Finally, the use of geographic boundaries and travel re- [6] Daniel Cooney, Vincent Wong, and Yaneer Bar-Yam, Beyond strictions allows for effective and comparatively low cost contact tracing: Community-based early detection for Ebola re- imposition and relaxation of interventions. Such a multiscale sponse, PLoS Currents Outbreaks (May 19, 2016). https://necsi.edu/ beyond-contact-tracing approach accelerates response efforts, reduces social impacts, [7] Chen Shen and Yaneer Bar-Yam, First thoughts on allows for relaxing restrictions in areas earlier that are less superspreader events, NECSI (February 28, 2020). https: affected, enables unifected areas to assist in response in the //necsi.edu/first-thoughts-on-superspreader-events [8] Joseph Norman, Yaneer Bar-Yam, and Nassim Nicholas Taleb, Systemic ares that are infected, and is a much more practical and risk of pandemic via novel pathogens—Coronavirus: A note, New Eng- effective way to stop otherwise devastating outbreaks [6]. If land Complex Systems Institute (January 26, 2020). https://necsi.edu/ actions had been taken earlier, successful local lockdowns, systemic-risk-of-pandemic-via-novel-pathogens-coronavirus-a-note