COVID-19 updated review of strategies

I received some positive feedback on my previous survey of COVID-19 strategies. So even though we are all overloaded with information about COVID-19, here is an update for anyone interested. Disclaimer: I have no qualifications to write on this topic, but am quoting the most reliable sources I can find.

The USA is now considered the epicenter of the epidemic. The charts from Worldometers.info of COVID-19 cases in the USA this month is the most terrifying thing I have ever seen. Till now it’s almost perfect exponential growth close to 10X per week. Look how cases go from 100 to 100,000 in ~3 weeks (mind you there is also an increase in testing which might be exaggerating the growth somewhat).

These are both the same graph but the one on the right has a “logarithmic scale”, simply meaning that each line on the right represents a 10X growth. So the straight line on the right means exponential growth - 10X every week or so.

Hopefully urgent social distancing will start to dampen the exponential spread in the US. Anyone can see that if the US reopens for Easter, the exponential growth is on track to reach 100 million people within 3-4 weeks, and closing for the two weeks till Easter and then reopening will at the very most buy 2 weeks extra reaching the whole population by May.

There is still a lot of uncertainty around the mortality rate but best estimates seem to be around 0.5-1%. But that’s only when hospitals are functioning. When hospitals fail the rate is likely to be higher. One indication is the shocking 12% mortality rate in Italy. This number is very uncertain in both directions (on the one hand many cases are unreported, but on the other hand many of the 57k diagnosed cases have yet to play out and some will still die). Another indication is that an estimated 4.9-11.5% of COVID-19 cases require an ICU, and presumably many of those may die without ventilation. Again the actual proportion could be less as not all cases are diagnosed, but it seems likely that the mortality rate without hospitals is at least 2%, and perhaps significantly more. Fortunately this has not been fully tested yet although several countries are heading there.

Any way you look at it, if the US opens its businesses and churches during April, 29–74% of the population will contract COVID-19 within a few short weeks, that is 100 to 250 million people (although possibly less if there is natural immunity). Hospitals will be completely overwhelmed and it's likely that several million people will die. Most will die from temporary trouble with breathing, where a relatively inexpensive ventilator could have saved their life had those been prepared in time.

Hopefully the Federal Government and/or the States will continue to mandate and further tighten social distancing way beyond Easter to avoid such an unimaginable catastrophe. 

I touched in my last article on age discrimination. There has been talk about the cost of the cure (economic shutdown) being worse than the disease. Taken to an extreme, it seems absurd to say that economic shutdown is worse than several percent of the population dying. 

But for young people the choice is not so obvious. The mortality for healthy under 40s (while hospitals are working) is 0.1-0.2% (and possibly less given unreported cases that don’t need a hospital). I think many young people would choose to take those odds, over a certainty of months of isolation and unemployment, which also come with real health and psychological risks.

In countries where the hospitals are not yet overwhelmed, I suggest we should have an urgent discussion about giving young healthy people (under age 40) the option for groups of them to intentionally contract COVID-19 while isolating themselves from others for the duration of their incubation and disease. The World Health Organization was arguably a little quick to warn against this. If younger people choose to take a small risk and contract COVID-19, that can be rational, and we will then have more people available to get back to work, to take jobs operating ventilators, and delivering food. We will also be a step closer to herd immunity.

There is possibly a big advantage with intentional infection that you can control the initial dose which may be correlated with a milder disease.

Obviously this needs thought and, sure, it feels wrong, but desperate times call for desperate measures and this idea has real merit so I don't think it should be ruled out so quickly. At the end of the day, it is the virus that introduced age discrimination, not us.

* * *

Overall it seems to me countries seem to be following three different strategies which can be characterized in terms of the basic reproduction number R? which is the expected number of people which each sick patient transmits the disease to.

Squashing to zero R?<1 (China model, especially Wuhan) 

This involves extreme lockdown of the population in their homes for several weeks to reach zero cases. After that, policing the borders very strictly and opening up social freedoms and the economy. 

This approach minimizes deaths and is relatively quick, but requires a particularly traumatic few weeks locking down people and shutting the economy completely. And it’s a very risky strategy. We know that 99% of people have an incubation period under 14 days but what if one in a million incubates for two months. Also a single visitor from outside can trigger another whole wave of the pandemic at any time.

I’m not sure any other country has succeeded in implementing this model and it remains to be seen if China is really virus free.

Continuous management R?≈1 (Taiwan/Singapore/Hong Kong model) This model aims to manage the spread with strong social distancing, avoiding exponential growth entirely, while allowing the economy to function as much as possible.

To make this work the country must have massive random testing every week to inform social distancing decisions and tune up/down. It is probably necessary to use technology/apps to track people in some detail and contain those who are infected and those who were in direct contact with them. This involves a temporary breach of privacy, perhaps leaving home will be illegal without a tracker app; more on that below.

Taiwan, Singapore, Hong Kong have some shops, offices and restaurants open with increased spacing and temperature checks and alco-gel at each entrance. In Singapore even most schools are open. Also here would help to allow young healthy people to choose to contract COVID-19 as having part of the population immune can help reduce R? for everyone. 

Singapore acted early and has amazing control of the epidemic with relatively less draconian measure. One caution is that it is likely that Singapore is only managing this with help from the hot humid weather. In the Northern Hemisphere many more countries will be warm come May and this will help for those that can delay the peak till the summer.

Mitigating and ventilating R?>1 (UK model, apparently also US and most of EU) 

This model accepts that 29–74% of the population will get the virus over the next year until we reach herd immunity. Up to 0.14-0.75% will die from the virus (mortality rate 0.5-1% x herd immunity threshold 29-74%), more if the hospitals are overwhelmed. Life and the economy are disrupted for several months but social distancing is not as extreme as in the other approaches as they are not trying to prevent the virus from growing exponentially.

Even so, significant social distancing is required to taper the exponential growth (R? close to 1) and avoid everyone getting ill at almost the same time. Today, 75% of the cases diagnosed ever were diagnosed in the same two weeks - that is the last two weeks. 

With this approach, after a few months the population has herd immunity and the epidemic is “done”. There can still be cases but they will not trigger a wave of exponential growth. 

What is absolutely critical in this scenario is to prepare massive field hospitals with ventilators and staff. There are efforts going on but it’s not clear that any country is fully prepared for the peak they are facing. As discussed last time, 4.9-11.5% of COVID-19 cases require an ICU and no country currently has enough ICU beds to ventilate that proportion of the population even spread over a year.

For those countries who are not yet overwhelmed, it could help to allow young healthy people to volunteer to contract COVID-19 first, in isolated groups, as above, so they can have jobs helping as the older people get it later.

Again here technology can help to police the required social distancing.

* * *

By the way on the topic of policing social distancing, Noah Harari has written beautifully why education is better than shattering freedom and privacy. But I think he’s wrong. Education can’t dramatically change people’s behaviors quickly enough. Much as I hate it, given the ferocious speed of this pandemic, it seems we need all the social distancing police and phone tracking we can get. We’ll have to deal with the horrible breach of privacy after we’ve saved millions of lives.

In summary, here in Israel, but also in other countries, the evidence suggest governments need to take the following steps taken yesterday:

• A war-like effort right now to build/buy ventilators, tests, masks, alco-gel dispensers, and to train medics (unless seriously going for suppression to zero strategy - then instead all resources have to go into stocking homes and going for complete lockdown) 
• Model out the three strategies and the human cost of the virus and of the economic shutdown and make a strategic decision
• In countries like Israel where hospitals are not yet overwhelmed, consider giving young healthy volunteers an option to get COVID-19 right now in an isolated area, say a camp of tents on the beach. After that, give them bracelets to show they are (probably) immune and allow them to get back to work preferably taking jobs manning hospitals, delivering food, or just to get back to normal jobs and restart the economy.
• Split the testing programs into two: (1) Targeted testing of people with symptoms and people who have been exposed so that they can be isolated if needed (this is already happening) and (2) Randomized testing of the population so we can understand how many people really have COVID-19, get more accurate mortality/hospitalization rates, and monitor R? so we can control social distancing up and down based on real data.

We’re living through some dramatic history. One huge conciliation is that so many homes have internet and video conferencing and we are able to stay connected and support each other, and keep parts of our economies going, from our homes. About twenty years ago we had none of that.

Wishing everyone a safe weekend at home.