By now, everyone has heard about the “Imperial College report on COVID-19”, which apparently was the reason for the 180-degree shift in the trump administration’s response and call for preparedness to the COVID-19 pandemic. A few days ago, trump was claiming that this is a minor bump and the virus will miraculously go away soon; on Monday, something seemed to have put the fear of God in the White House; a national emergency was declared and trump was talking about social distancing and washing hands. In the UK, the crazy idea of letting everyone get infected to achieve “herd immunity” has been replaced by proclamations of mitigation and suppression.
The change in strategy has followed the publication on Monday of a UK government-commissioned report from Imperial College London’s COVID-19 Response Team, which models different strategies and outcomes, for the UK and the US. An early copy of the report was shared with the White House over the weekend.
Here, we present a summary of the report.
Top Level Summary
The authors of the study consider 3 scenarios (and sub-scenarios) and use simulations to predict fatalities in each scenario.
# |
Scenario |
Deaths |
Notes |
1 |
No mitigation steps |
4 million |
80% of Americans will get the disease
Deaths will peak in June
The study mentions 2.2 million deaths with the assumption that sufficient ICU beds are available for the critically ill, which will not be true
|
2 |
Mitigation steps for 3 months |
2 million |
All symptomatic cases and their families in isolation/quarantine. All Americans over 70 social distancing.
Flattens the curve a bit but not sufficiently
|
3 |
Suppression and mitigation steps for 5 months |
1000’s |
#2 + social distancing for the entire population, all public gatherings/most workplaces shut down, schools and universities closed.
This step will have to be repeated periodically for over 18 months until everyone gets vaccinated
|
Note that in all cases, the elderly and those with pre-existing conditions will experience a higher fatality rate.
The authors claim that scenario #3 is the only viable option at this time (assuming there are no vaccines or drugs that can significantly slow down the virus).
Paradoxically, while #3 is implemented, people and politicians will feel that the steps are overkill since the death rate will be so low.
Scenario 1
Scenario 1 is defined as the case where no mitigation steps are taken (which of course is academic since some mitigation steps have already been taken).
In this scenario,
- Infections spread exponentially with a certain R0 values above 2.0.
- 80% of the population gets infected.
- 2.2 million people will die in the U.S. over the next 3-6 months, assuming there are sufficient number of ICU beds.
- About 4 million people will die in the U.S., since the number of ICU beds will be a small fraction of that required (the survival rate for those in ICU with ventilators is 50%)
- 8-15% of all Americans over 70 would die.
- Jeremy Young points out that if we extrapolate these numbers to the rest of the world, this gives us ~90 million deaths globally from COVID-19, in 3-6 months.
- Note that most survivors will have immunity against future infections
Unmitigated epidemic scenarios for GB and the US. Projected deaths per day per 100,000 population in GB and US
Scenario 2
In this scenario, some mitigation steps are taken for a period of 3 months, that help to “flatten the curve”.
In this scenario, the first 3 interventions in the table below are implemented. The last 2 are used in scenario 3.
The study uses simulation to quantify this flattening for various mitigation steps and combinations thereof. The graph below shows the number of ICU beds needed for the various mitigation steps.
Mitigation strategy scenarios for GB showing critical care (ICU) bed requirements.
Note that mitigation steps reduce R0 but not to below 1.
The study estimates that if the first 3 mitigation steps are implemented —
- The death count will be 50% of that in scenario 1, i.e., about 2 million people will die in the U.S.; the number of ICU beds demand will still vastly exceed the supply
- Deaths peak in June/July
Scenario 3
In this scenario, various mitigation and additional suppression steps are taken for a period of 5 months, that help to seriously “flatten the curve”. The study uses simulation to quantify this flattening for various mitigation and suppression steps and combinations thereof. The graph below shows the number of ICU beds needed for the various cases
Suppression strategy scenarios for USA showing ICU bed requirements. (B) shows the same data as in panel (A) but zoomed in on the lower levels of the graph.
With all mitigation and suppression steps in place, ICU demand (orange line) stays below supply during the 5 month period. Total deaths are in the 1000s.
Note that suppression+mitigation steps reduce R0 to below 1.
But after the 5 month periods, when restrictions are lifted (we cannot go on like this forever), the lines start to rise.
So, the authors advise, that scenario #3 needs to be replayed periodically every few months for about 18 months until everyone gets vaccinated. The graph below shows one such 18 month scenario (for GB not USA), where certain mitigation steps are implemented all the time but some of the suppression steps are lifted when ICU cases per week fall below a threshold (50) and are brought back when they exceed another higher threshold (100).
Illustration of adaptive triggering of suppression strategies in GB, for R0=2.2, a policy of all four interventions considered, an “on” trigger of 100 ICU cases in a week and an “off” trigger of 50 ICU cases. The policy is in force approximate 2/3 of the time. Only social distancing and school/university closure are triggered; other policies remain in force throughout.
Notes
The simulation uses inputs such as ICU occupancy rate of infected people, death rates by age groups measured so far and estimates of the spread parameter R0 and how it changes as mitigation and suppression steps are deployed.
The study does not take into account any new or repurposed treatments and drugs that might reduce fatalities, although the measured mortality rates account for the effect of some of the drugs that have been tried out.
The study does not assume any reduction in infection rates with summer temperatures.
The study focused on UK and USA only. Forecasts for developing countries is a question mark.
The study only hints at steps countries must take to beef up healthcare facilities and equipment.
Some Sobering Thoughts
Jeremy Young has some sobering thoughts on life in the next 18 months of Coronavirus -
The authors of the study end with these somber words —
We therefore conclude that epidemic suppression is the only viable strategy at the current time. The social and economic effects of the measures which are needed to achieve this policy goal will be profound. Many countries have adopted such measures already, but even those countries at an earlier stage of their epidemic (such as the UK) will need to do so imminently.
However, we emphasise that it is not at all certain that suppression will succeed long term; no public health intervention with such disruptive effects on society has been previously attempted for such a long duration of time. How populations and societies will respond remains unclear.
Miscellaneous
The leader of the study team, epidemiology professor Neil Ferguson, himself seems to have caught the virus,
Opposing Views
Note that not everyone agrees with scenario #3. Others think that something like scenario #2 with aggressive testing and contact tracing will get the job done with less disruption to life. They point to S. Korea as an example of this approach.
Prof. Bedford (like many others) proposes testing and isolation (similar to S. Korea) as the strategy, rather than aggressive suppression -
Here is some selected text from the twitter chain —
- These transmission routes (mild and maybe even asymptomatic infections) can be reduced by a huge rollout of testing capacity. If someone can be tested early in their illness before they show symptoms, they could effectively self isolate and reduce onward transmission compared to isolation when symptoms develop. 12/19
- This rollout of testing could be achieved through at home delivery of swabs with centralized lab-based processing combined with drive-through testing facilities. There are logistics involved in getting a result quickly, but it's really just logistics, which can be solved. 14/19
- The second, related, strategy is using cell phone location data combined with data on known positive cases to alert possible exposures to self isolate and get tested. (This technique was also employed by S. Korea)
- This strategy targets testing capacity at most likely cases and serves to detect exposure events early, when isolation is most valuable. This cell phone location based approach is outlined carefully here: https://github.com/BDI-pathogens/covid-19_instant_tracing…. 16/19
- A third, supporting, strategy: as the epidemic proceeds get serological assays run on as many people as possible to systematically identify individuals who have recovered and are highly likely to possess immunity. 17/19
- Individuals who have serological evidence of recovery and are no longer shedding virus can fully return to the workforce and keep society functioning (especially important for those at the clinical front lines). 18/19
Epilogue
I am sure I have oversimplified the report and the thoroughly detailed report contains much more subtle analysis and messages. Feel free to add your insights to the comments section.
You may differ with the study’s projections, conclusions and its recommendations for action. We can nitpick about the various factors that were not taken into account in the simulations. We can argue that scenario #3 is impractical in this country. But people are taking this report seriously, including some in the trump administration, although the administration have not yet taken any major responsible steps yet. The study itself is quite comprehensive and is based on the best data we have available.
Adding more testing kits, ICU beds, healthcare staff, etc. will help but it will get overrun until we use suppression techniques to bring R0 to below 1.
Countries like China, S. Korea, Italy, France and Spain are already past implementing mitigation steps and are close to deploying the suppression recommendations. I would presume that we are headed in that direction as well, so we may as well prepare for it physically and mentally.
Granted, trump and the GOP will take steps reluctantly, use the occasion to demonize Democrats, China, immigrants and all his “enemies”, spend trillions of tax and borrowed dollars to shore up businesses and their rich buddies, take credit for every little improvement of the situation and blame Democrats for any bad news. How this will play out politically, how elections will be effected, will Democracy survive — these are questions that will be discussed across the punditry class over the next few months, writing and talking from the confinements of their homes. We need to prepare for that too.
If implemented, scenario #3 will be unlike anything the world has seen. The effects on the world economy and people’s lives has not been calculated yet. Let’s get prepared for scenario #3, even though we all harbor hopes that #2 will get the job done.
Further Reading
- Impact of non-pharmaceutical interventions (NPIs) to reduce COVID19 mortality and healthcare demand — www.imperial.ac.uk/...
- COVID-19: Imperial researchers model likely impact of public health measures - www.imperial.ac.uk/...