Explanation why Boris Johnson's decision about pandemic is probably right [UK]

(2020-03-18)

There has been a lot of criticism about Boris Johnson’s coronavirus plan. But scientists think it could work better than WHO plan.
(Disclaimer: I work at Imperial College, but I was not involved in this study.)
Here is my rationale:


It is much too late to stop pandemic from happening in the UK and virtually in any other country on the Planet except Korea, People’s Republic of China, Singapore plus maybe some smaller countries. Regardless what we do now in about 6 weeks (late April / early May) UK will enter a critical period of a large portion of the population infected with many more patients critically ill than beds, ventilators and hospital staff available. The UK is trying to increase the personnel, but there are serious limitations when it comes to ventilators. Majority of retired NHS staff don't want to return to tackle Covid-19 crisis .


The Imperial College team have a top-level expertise in pandemic modelling. A couple of days ago they published a report titled Impact of non-pharmaceutical interventions (NPIs) to reduce COVID- 19 mortality and healthcare demand .

The team considered different scenarios. The main outcome is very well explained by this graph:




The red line is the medical capacity of providing care to critically ill. It is a constant. Other curves show the amount of required capacity. Different curves correspond to different policies. The government's goal is to save as many lives as possible. We should be looking at the number of untreated cases: the surface area between the red line and the part of a curve above the red line. It is proportional to the number of unlucky citizens unable to get treatment.

The strategy of doing nothing (black) is the worst scenario. The strategy of closing of schools and universities alone (green) is surprisingly not much better from doing nothing at all (black).
Other strategies are more adequate. And out of all these frankly horrible scenarios UK government’s policy (blue) is the most reasonable thing to do.

Does it look counter-intuitive at first glance? For me it did. You may think that (obviously) more isolation early on makes more sense. But it is not what the plot shows!
 

The researchers from Imperial College are experts in modelling such events. In such modelling there is one critical parameter: Pandemic dynamic depends strongly on the frequency of human interactions as a function of time (or more precisely the connectivity of graph of human interactions). Some interactions may be unavoidable, like shopping, but most are avoidable social interactions. The Imperial College team have gathered existing data on the connectivity graph in the UK, added parameters of COVID-19, and ran the model multiple times for different policies. They were building, studying and adjusting such models for many years.

UK government’s bet is that a new vaccine won’t be available in this year. But most likely people who recovered from infection will become immune. At least this is what a study on animals shows.

Since NHS capacity cannot be substantially increased, the only way to change the shape of the curve is to manipulate connectivity of social networks. The catch is that the Prime Minister has only a limited control over this parameter. Moreover, his amount of control will change (mostly decrease) as pandemic will progress.

In the case of immediate self-isolation of everybody, this is probably going to happen: Kids and eventually adults will get bored of staying at home and will break the rule by returning to social interactions thus increasing graph connectivity back.


Here are the two scenarios explained not with equations, but in layman’s words: 


1)  Full social isolation now. Families are told now to self isolate from now on. The kids, and eventually adults, will get bored and will eventually increase social interactions. When the critical strike happens around the 1st of May, they won’t be self-isolating themselves as well as in the beginning, resulting in a steeper growth of infections over a course of 3 months.

2) The government's approach: If people are told not to isolate themselves yet, but only told to do so from mid-April, then they won’t become bored of staying at home during the 4 week period till that time. So they will be more obedient and more eager to self-isolate at the critical time. In about 2 weeks from then everybody would be seeing horrendous news about the situation in hospitals. This will scare them off and motivate them even more to hide themselves at homes. The result will be a sudden drop in the connectivity graph. The critical period will be stretched and flattened, exactly as the blue curve shows. In effect this will give more time for the NHS to work on critical cases over May, June, July and the first half of August. In any case it won’t be adequate enough, but at least will give more time. My guess is that 50,000-100,000 people could be saved this way from the current estimate of 500,000 deaths till the end of August. During the drop-down phase in August herd immunity will start to work.



Herd immunity is normally obtained by vaccination. For instance in the case of measles every person who is vaccinated becomes immune, thus invisible for measles virus. Virus cannot infect such person and cannot spread through that person. This effectively makes social interaction graph less connected. Above some proportion of immune people the disease will stop spreading even if not everybody would be immune. The scientists estimate this threshold to be 60% of the population. Herd immunity does not mean that any single person is safe. It means that small outbreaks will extinguish themselves locally. From Autumn all people who evaded infections will be relatively safe, but not 100% safe.




Imagine fire which has burnt most of grass on a field, but some clumps of grass evaded the fire. Now if one of these clumps catches fire, it might get burnt, but it won’t spread fire to other disconnected clumps!

Side note: Different parts of the country have different connectivity. It is very different to live in the middle of Scottish moorland where you have little neighbors and no public transport, than to live in central London. The 60% figure is a global estimate. In London and other big cities it must get higher than that.

What can you do now:

  • You need to have enough food for yourself and for everybody else in your household to sustain a 2-week self-quarantine in case of infection. This is because your family/housemates will likely get infected from each other so they also need to stay at home. In the beginning of May panic is likely too happen. Then there will be a shortage of food. Since food distribution is govenment's critical priority it will be quickly fixed. But you should better prepare to have enough food in such case. The worst case is when he infection time and the time of global panic being adjacent. There will be no time to resupply. You will need a 3-4 weeks of supplies. In a more lucky and probable case you will be fine with using up 2 weeks of supplies and then restocking for the other 2 weeks when possible.
  • Do you know any old people living alone? Offer them help in getting a 2-week supply of food and other necessities.
  • Do you have some medical training and you are young enough not to fear coronavirus? Offer your help to the NHS now.
People are more prone to diseases during cold months. It is yet uncertain if infection rate will increase during next autumn and winter. Flu works this way so we have a flu season. Common cold as well. Some of common cold is caused by different strains of coronaviruses circulating for centuries in Eurasia.


Boris Johnson’s bet is that COVID-19 will become seasonal.



The model predictions for winter 2020/2021 is most favorable to the “blue strategy”. In the case of no intervention (black) almost everybody will be infected during summer and this will be sufficient to induce herd immunity to such levels that during autumn/winter there will be no second outbreak. But all of this at a very high cost. Italian, German etc policy (green) of at total shutdown of schools, case isolation and general social distancing (not only people above 70) will result in much lower number of mostly manageable infections during spring and summer. But there won’t be enough immunity at autumn/winter time, so there will be a return of a pandemic. Remember that this assumes no vaccine ready before autumn.
The orange strategy of case isolation, household quarantine and general social distancing, but without school and university closure looks best, but if you look at the scale and compare it with the first plot you will see as good as the blue curve.

What you should be doing? If you are selfish enough it is safest to live locked in another country over the period of spring/summer or isolate yourself in a remote area anywhere else. Then you may return to the UK (or to any country which developed herd immunity) in early autumn.

So why is WHO recommending a different approach? They might be more optimistic about vaccine being in mass use before late autumn. Plus they have a different model. Around early May we will see if the Imperial College model works. And by early November we will see if a second outbreak in Italy or Germany will happen. If not, the conservative government will be in big trouble.

Popularne posty z tego bloga

Życie na Mauritiusie

Życie w hostelu na Khao San road w Bangkoku