Coronavirus Parties, Tupperware and Tinder

I’m old enough to remember being sent to play with other children who had measles and chickenpox.  That was before we had vaccines for either.  We’d have a happy time picking off each other’s chickenpox scabs, leaving our generation marginally scarred for life.  It was an understandable practice – neither disease had a high child mortality rate – it was far more dangerous in adults, so keeping up herd immunity this way had pretty good odds.  

We’re about to come out of lockdown and enter the “New Normal”, whatever that may be.  It means that as far as Covid-19 is concerned, we’re back in the pre-vaccination world.  Throwing technology at the problem appears to be the first choice of most Governments, but we should think about whether there are some pre-vaccination strategies which are worth revisiting.

Right now, in the UK and most of the rest of the world, the preferred answer appears to be Bluetooth tracing apps.  You can read more about them in my previous articles.  The big question is how helpful are they?  We definitely need to trace an individual’s contacts once they realise they’re infected.  That’s good old-fashioned epidemic containment, which generally means a lot of detective footwork.  Bluetooth tracing apps are meant to simplify that.  However, Singapore’s experience with their Bluetooth TraceTogether app looks as if it’s not been as successful as they had hoped.  At its peak, it was only being used by around 17% of residents, and it’s suggested that figure has fallen to around 10%.  They’re moving to their new SafeEntry check-in system, where you need to scan a barcode whenever you enter a tube station, hospital, municipal building or workplace.  It’s simpler, easier to use, and doesn’t need phones, as you can also scan your ID card.  On the flip side, the Government knows exactly where everyone has been.  It’s a step which might be difficult in the UK, as we don’t have ID cards, and we also don’t much like the idea of the Government recording our movements.  But when you hear any Minister talking about tracing apps, you can be pretty sure they’ve got a SafeEntry variant lined up as Phase 2.

In the meantime, the Government is trialling their Bluetooth tracing app on the Isle of Wight.  The Isle of Wight’s a lovely holiday destination, where you feel that you’ve gone back in time.  Now,  devoid of its tourist trade, it’s about as different from somewhere like London as you could possibly find.  Which means the trial may well be irrelevant.  Matt Hancock, our Minister for Health assured us that “Where the Isle of Wight goes, Britain follows” (forgetting that it won’t be Britain, as Scotland’s developing their own app).  Anyway, I’m not sure I want the new normal to be plastic dinosaur parks and sand in my ice cream.

To see why the trial is irrelevant, let’s look at the average number of contacts someone is likely to have during their day.  A quick recap on Bluetooth tracing is that your phone will detect and log anyone else with the tracing app on their phone if you’re within 2 metres of them for more than 15 minutes.  If you exhibit symptoms and are tested positive, all of those contacts you’ve logged will be told to self-isolate and get tested.  The more people your phone has logged, the more will get sent home and need to be tested.  If you’ve logged 15 – 30 people that’s eminently manageable.  Once it gets to more than 50 it starts to overwhelm the tracing and testing infrastructure.

The figure below estimates those numbers for someone with a family on the Isle of Wight, a similar person in London and someone living alone in London.  It shows the likely number of contacts that they will be close to for each hour of the day.

At night, you’ll be in bed with your partner, and your kids when you wake up.  On the Isle of Wight you’ll probably self-isolate in your car to get to work, whereas in London you’ll use public transport.  For the Londoner that will add around 50 contacts who could infect them.

At work I’m assuming around 7 nearby colleagues on the Isle of Wight and 10 in London.  At lunchtime, assuming that sandwich shops and take-aways are open, the Londoner may add 20 contacts because of the density of population and length of queues, compared with 10 on the Isle of Wight.  We see the same difference in contacts on the home commute.  I’m also assuming that the singles may well meet up with friends in the evening as they equate the new normal with the old normal. 

You can see the difference that mass transport and higher population density makes.   Our Isle of Wight worker will come into contact with around 14 people for long enough to be a risk each day, most of whom will be the same people each day of the week.  In contrast, the Londoner will be in close proximity to about 140, of whom the majority will be different each day.  

The figure above shows how that adds up during the course of a working week.  By the end of the week the Isle of Wight worker might have 20 significant contacts on their phone who would need tracing if they contracted symptoms; the Londoner would have over 600. 

It’s a broad brush example, but it makes two points:

• A trial on the Isle of Wight won’t represent large cities, and
• There are some large peaks in cities that we need to address, otherwise we’ll be telling tens or hundreds of thousands of people to self-isolate every week.

The biggest peaks for London come from mass transit commutes.  This is really difficult to control, as in normal times, the tube carries around 5 million passengers each day.   The easy step is to encourage as many people as possible to continue to work from home.  If they’re going to do that effectively, they need to be able to work uninterrupted, so that means reopening the schools.  We can try to stagger working hours, but the chances are that if you use public transport, you’ll still be in close contact with around 50 people for half an hour.  Unfortunately, a busy tube carriage or bus is about the worst possible environment for a Bluetooth tracing app.  On one hand, the metal box you’re in reflects Bluetooth signals, confounding the ability to estimate distance.  On the other, human bodies absorb the Bluetooth signals really well.  What this means is that someone standing three or four metres away holding their phone up could appear closer than the person standing next to you with their phone in their back pocket.   There’s probably not a lot we can do to change that.  Ian Levy, the Technical Director of National Cyber Security Centre has written a good article about the NHS tracing app, but acknowledges that “signal strength is a proxy for distance”.  The reality is that the enclosed space, air recirculation and constant movement as people get on and off mean that if anyone is infected in your carriage, you’re probably breathing in coronavirus.  So distance guestimation is not a lot of use. 

The lunchtime foray is more manageable and has two mitigating options.  The first is to ask people to remain in their offices and arrange for local shops to deliver lunchtime food.  That has the secondary advantage of encouraging more small foodshops to stay open, although that will mean more people on public transport.  The second approach is to persuade everyone to bring in their own lunch.  It’s a pre-vaccination era habit which used to be common for many, with packed lunches gently warming on each desk in their Tupperware boxes. 

I’ve added single Londoners as they may exhibit a different behaviour once they’ve finished work.  Before Covid-19, their evening plan was often to go out with friends.  Once they’re back at work, my guess is that they’re going to find it hard to resist reverting to form.  If they’ve braved public transport to get into work, they’ll not see a major issue with a meeting a few friends afterwards, even if their favourite bars and clubs are still closed.  Social Distancing will also turn into Social Nearnessing, as Tinder becomes their preferred contact tracing app.  And delaying the journey home, they’ll argue, will reduce the crowds on the tube, so it all evens out.

The reality is  that we don’t have any effective mitigations for returning to life as normal in large cities.  Contact tracing could create more problems than it solves, because of the enormous number of contacts that arise from the population density and need for mass transit.  If London and other large cities are to get back to work, we need another approach.

One starting point for that is to look at mortality against age.  Looking at the published stats of those who have died across the UK, Italy, China and Spain, it is predominantly the elderly who are dying.  As the figure below shows, in the UK, over 70s account for almost 90% of deaths.  Under 50s for less than 1%. 

Data from New York City Health, which provides more information about those who died, suggests that around 90% of those under 45 who died with Covid-19 had underlying medical conditions.  The corollary is that it’s probably as safe for everyone under fifty without an underlying heath condition to go back to work, as it used to be to send your child to play with friends who had chickenpox.  However much the older sections of the population may feel that millennials don’t seem to care as much about social distancing, it may not matter.  The sooner they all infect each other, the faster we build up herd immunity (assuming we stay immune, which is a big question).   Let them party and dispense with social distancing; bring back the hospitality and entertainment industries, pack the bars, theatres, nightclubs and restaurants and see where that gets us.  Some will die, but it should be manageable numbers.  If the numbers rise, then we move back into lockdown.  If the numbers start to fall, the lockdown can be relaxed by age and medical condition, based on the predicted number of deaths that the health service could handle.   We could probably use an antibody test to let over 50s who appear to have had Covid-19 back out.  That has a risk if the test is not accurate, but that’s a risk many may be happy to take to escape lockdown.

The flip side is that we would need to keep strict lockdown for the over 50s and anyone younger with an underlying medical condition, whilst everyone else returns to life as normal.  For those at risk, particularly care home staff and residents, we should move to a daily test regime.   It would probably also mean self-isolation or regular testing for any household with older members.  It would be a hugely divisive strategy, but it could be our best hope.

Would it be irresponsible?  It comes down to the fact that managing a pandemic is a numbers game.  Like any numbers game, you need to have enough data if you want to play a winning hand and at the moment we don’t.  There are too many things we don’t know about Covid-19, because we haven’t been doing the testing.  That has to change to the point where we are performing millions of tests, again and again and again to give us confidence about infection rates.  Every front-line worker who has tested positive should be tested weekly to determine whether they can be infected a second time, and if so, how long immunity lasts.  The fact that we’re not doing that speaks volumes about Government policy, showing that it is led by politics and not by the science, despite frequently repeated assertions that policy is following the science. 

The more we know, the better chance we have of a strategy.  Before we had vaccines, the only approach we had was deliberate infection of our own children to continue herd immunity.  Our current policies prioritise preventing deaths, but until we get a vaccine, what we are offering to many is death by a thousand cuts.  Isolating the old and vulnerable with a lockdown policy which may last years can turn into a life that is not worth living.  In the past few weeks, we have lost a few of our elderly neighbours.  I don’t think it was through Covid-19, but we don’t know, as they weren’t tested.  They had long-term medical conditions and decided that they would discretely carry on life, inviting friends around for their last few weeks, as they couldn’t face the isolation of a lockdown on their own. 

That’s a fate that affects many.  Phil Hammond, the GP and broadcaster, has spoken of elderly patients that ask him for a hug at their annual check-up, as it’s the only physical contact they have each year.  In a recent edition of Private Eye he talked about the fear of many in care homes, wondering how many may be just taking to their beds and giving up, rather than existing in this terrifying new world of PPE and isolation.  He reported one resident joking that “If this is all I’ve got left, I’d rather lick a nurse and get it over with”.  What is the humane target to set for a pandemic that will probably be with us for longer than the natural lifespan of many care home residents?  Is it a number of deaths, or a quality of life?

Our short to medium term goal cannot be to defeat Covid-19.  We have no tools to do that until we develop a vaccine, which is a hope, not a strategy.  Instead we need to downgrade Covid-19 to a manageable outbreak that society can live with; something which is no more than an inconvenience to our everyday lives.  Rolling out shiny apps is one option, but it probably won’t work.  Looking back to what we did in the past, along with acquiring meaningful volumes of data might be a more useful approach.