C-19: Comparing 'exit strategy' archetypes between countries

Updated: May 12

The COVID-19 pandemic has now reached its fifth month since the initial outbreak in Wuhan, China. Tracking the progression of spread through this period for a range of countries has enabled us to observe the virus' trajectory, due to effects of lock-down and various other control measures. As we contrast how countries have fared, we see a few emerging archetypes and will share them here for discussion

TL;DR, and some introductory notes:

  • On-going mobility reporting has distinctly indicated where populations are beginning to show movement due, in part, to softening of lockdown measures. The trajectory of this mobility can also be extended to forecast future outlook (intuitively this makes sense a there tends to be momentum in movement)

  • In our analysis, we have projected forward the historic 2 week day-on-day mobility trend-lines to aid our forecasting. It also bears noting that as case detection delays range from 5-20 days, and median time from infection to death is roughly ~14 days, mobility data provides a crucial leading indicator to aid a better understanding of death and case detection data trajectories

  • While still some way from the final word on forecasting the pandemic, this modelling approach complements where our initial modelling left off by providing a semi-robust continued trajectory of infections and deaths past this point. The model mechanics kick-in after the initial hard taper from lockdowns, where transmission is assumed to be directly correlated to mobility through a correlative factor

  • This allows an interesting exploration of how different countries will continue from this point onwards - both as a function of their past historic trajectory, and also future mobility increases upon measures to reopen their economies

  • There are a range of extremes across the archetype spectrum: high risk scenarios where transmission has not fallen significantly enough but mobility is rapidly rising, risking a large secondary wave of deaths (e.g. US); to low risk scenarios where transmission has fallen very significantly and mobility can possibly rise to 1.5-2x from current levels without risking epidemic outbreak (e.g. Malaysia)

Analysis notes

  1. For now, our forward projection estimates still use our previously established baseline model mechanics. We will continue to monitor the results from the mobility-based model explained here, and may replace the baseline with it if it proves robust. It is very likely that an approach such as this one will be required moving forward; given that many countries will continue to grapple with moderating controls on and off over the coming months resulting in a continuous shifting of mobilities

  2. While we may assume transmission and mobility are directly correlated for purposes of modelling, this might not always be the case as we had explored in a previous analysis (exact correlations vary from country to country, and at times with ill fitting regression). This relationship is also a function of a number of things; for example, the number of infected (and how close to critical mass for community spread), social distancing practices, personal hygiene, and other evident norms and practices. Regardless, so far, transmission decline has 100% of the time correlated with mobility reduction, and as a corollary, mobility increase should thus result in transmission increase - which is the crux of this approach

  3. The model uses a blended mix of mobility data from both Apple and Google (retail, transit, walking and driving mobilities); with Google/Android assumed as a higher contribution across all countries. As at time of writing, Google data is limited up until 2nd May, beyond which we project forward using historic trajectories

TYPE I(a): Highly eager to reopen economy [risk level: high]

Transmission rates are barely below 1.0, but mobility is already rising distinctly and will almost certainly risk another exponential outbreak and rise in infections and deaths. Massively concerning unless more careful steps are taken

Examples: US, UK

TYPE I(b): Less eager, but waiting for right time to reopen [risk level: moderate]

Taking a slightly delayed/phased tactic to reopening verses Type I(a) countries, by allowing infections and deaths to drop more significantly first; however if not careful could push back into that category as transmission is not significantly <1.0

Examples: Belgium, Spain, Netherlands, Germany, Italy

TYPE II(a): Struggling to taper [risk level: high]

Greatly struggling to reduce transmission to non-epidemic levels despite having reduced mobility very significantly from initial base levels, which is a highly troubling indicator for the future of these countries

Examples: Brazil, India, Bangladesh, Ecuador, Egypt

TYPE II(b): Barely hitting taper point [risk level: moderate]

Have just reached non-epidemic levels mainly through lockdown measures, however proximity to onset of transmission >1.0 means it will be difficult to reopen the economy without taking significant risk. If not careful mobility may rise and tip transmission over the line, resulting in Type II(a) behaviour

Examples: Indonesia, Philippines

TYPE III: Very conservative [risk level: low]

Sizeable headroom gained from lockdown activity resulting in a transmission below 0.95 - likely able to increase mobility significantly by possibly 20-30% without risking epidemic growth

Examples: Malaysia, Thailand

TYPE IV: Delicate balancing act [risk level: low]

Refused harsh lockdowns and have used non-'mobility control' measures to keep infections at bay - however that is a delicate balancing act. Upside is they can now use lockdown controls if necessary to stem any outbreaks from growing too greatly

Examples: Sweden, South Korea, Japan

TYPE V: Finding alternative ways of doing things [risk level: low-med]

Taking a non-standard approach to dealing with virus. Turkey utilises stringent weekend curfews as means to modulate transmission into and out of epidemic levels. It seems to be bearing fruit. Despite being called out for the troubling state of its trajectory in the early phase, the country seems to have found a reasonably good balance between reopening the economy and keeping the spread at bay.

Examples: Turkey


© 2020 Agility Plus 

Pushing Boundaries Together