Saturday, September 19, 2020

Can we avoid lockdown?

I considered briefly the magnitude of the response needed to address the COVID-19 pandemic, and concluded that - roughly - a factor of 10 reduction in social activity was required.

The question then becomes - how to achieve that?

Clearly, lockdown works (although the fact that our decline was slower than it should have been indicates that it wasn't working optimally - lacklustre leadership, being two weeks late into lockdown, and a significant selfish minority who clearly thought the restrictions weren't for them, contributed in part.

There are massive negative economic and societal consequences of lockdown, so it would be better to avoid it if you can. Can we?

What are the options?

  • Working from home: this dramatically eliminates both the work contact and public transport contact elements. It's a sufficiently large component that if you don't do it, you can't win. So, where possible, people should work from home. There's still going to be a set of people for whom this isn't possible.
  • Avoid public transport: yes, this helps a bit, but has negative impacts. Remember, we have a climate disaster looming on the horizon. If the majority work from home, then that reduces the occupancy - and thus the risk - for those that need to travel.
  • Eliminate large-scale events: the sheer number of people, and their closeness, surely indicates that live theatre, sport, concerts, and the like, simply can't be contemplated for a while.
  • Rethink large-scale events: large scale events have a couple of problems, first you're putting a huge number of people in direct contact at the event, and you're having a lot of mixing on the way in and out (and many will visit other social venues as part of the event, making things even worse). But if you only had isolated boxes - and many venues already have some provision like this, after all - with no way for people in different boxes to mix, then you can get to a point where it's not really any worse then the groups meeting to watch the event on TV in a bar or at home.
  • Shut down bars and restaurants: there's a massive economic and social hit from this. This is a smaller version of the question about avoiding total lockdown.
  • Rethink bars and restaurants: how far can you eliminate the risk by having isolated groups (bubbles), table service only, one way routes, and ramping up the restrictions? I think you can actually go a long way from the norm. Having visited a few bars and restaurants, most are doing a pretty good job. The fundamental problem again is those in the selfish subclass who won't obey the restrictions and ruin things for everyone. You're going to have massively reduced occupancy (simply by spacing people out) which reduces the risk. Which also reduces the risk because the reduction in capacity means people won't be able to go out as often.
  • Testing: if you test everyone continuously, and just barricade the infected, then you should allow everyone else to get back to normal, right? Not so fast. The point here is that a single negative test doesn't mean anything. Worse, any contact means you need to get retested - over and over. That's the sort of thing you have to do in hospitals, it's simply not feasible to cover the whole population at the required density.
  • Early closing: I've seen the suggestion that bars should close early. I'm not at all sure that this will help much, as it just pushes people into packing more densely into the hours that places are opening. If anything, you want to expand opening hours (presumably early opening) to spread customers out as much as possible. The only way I can see this helping is if it discourages people from going out at all.
  • Ban pub crawls: this seems to me like a no-brainer, to be honest. Restricting people to a single venue massively reduces the mixing effect. (And it seems likely that there's a strong correlation between pub-crawlers and selfish superspreaders.)
  • Track and trace: essential, but only works if you massively reduce your interactions in the first place. If you try and carry on as normal, then everyone ends up potentially coming into contact with an infected person and you have to lock everybody away as a result.
  • Social distancing: taken as a given, but not enough alone unless you push it well beyond 2 metres. Doesn't matter how far apart you are if you have to share the same door handle, though.
  • Face coverings: again, unlikely to be enough on its own. Really, you should be thinking about face coverings, social distancing, and similar precautions as being an extra level of protection if you can't avoid an activity at all (such as shopping for groceries). You shouldn't take face coverings as an excuse to enable you to partake in riskier activities.
  • Redefined infrastructure: take the things that we have to do and eliminate risk. Examples include automatic doors (you don't touch anything that someone else has), foot or knee operated taps rather than turning by hand. Individual lifts, and likewise travel compartments rather than massive open carriages.
There's no silver bullet here. Saying "if we all just did X" won't work, no single action is enough. This has to be a concerted attempt, all the options need to be pursued together.
I think we could avoid lockdown, but it requires a huge amount of work. You need to go fo the big hitters - working from home as much as conceivably possible, scrapping all large events entirely. Then you need to layer on every single measure and precaution you can. And you need a high level of adherence to make it work.

Sadly, we have a government that hasn't the gumption to do the big things or the eye for detail to do all the little things, and the great british public has far too many gormless prats to expect the measures necessary will be followed.

Friday, September 18, 2020

The accuracy of COVID-19 statistics

Nobody really knows how many people have died in the UK due to COVID-19. There are currently 3 numbers that might give a bit of a clue:

  • The daily statistic of those who've died after a positive test
  • The number who have had COVID-19 mentioned on the death certificate
  • The excess deaths reported compared to a normal year

The last one is more amenable to statistical analysis, but is also subject to a variety of errors: the baseline varies from year to year, and lockdown may increase some types of deaths, while decreasing others (for the latter, consider the reduction in pollution). Still, it's a reasonably well defined number.

The second one is highly tricky, because - absent lots of tests and post-mortems, it's sometimes going to be tricky apportioning the cause of death.

The first one has attracted a lot of attention, and it's the headline number you see in the news. It has the advantage that it's quite well defined (you know, definitively, who has had a test, the outcome, and whether they died). The disadvantage is that it doesn't give you any clue as to whether COVID-19 was actually the cause of death or not - perhaps they got run over by a bus.

Early on, this didn't make much difference. But, over time, the probability that someone would die from another cause obviously increases. So, early in August, the statistic was changed to add a 28-day cut off.

The idea of a cut off is that, very roughly, the number of people who die from COVID-19 after the cut off are offset by those who die of other causes before the cut off. At a nigh level, it makes sense, because otherwise it's going to be wrong, and the error will increase over time.

The question really is whether the correction applied is correct. After all, that 28 days is basically a guess - it's commonly used, so is reasonable for comparison purposes, but it's still a guess.

There are a couple of ways to see if the value for the cut off is reasonable. And for that, we need data. It turns out that we can download the time series from the portal, and have a look at the numbers.

One fortunate thing we have is that the dataset actually includes 3 numbers - raw numbers, with the 28-day cut off, and with a 60-day cut off. As of today, the 28-day cut off removes 6,634 deaths from the starting 44,115, and the 60-day cut off less that half that at 2,695. These are what would come in as deaths from other causes.

First, given the number of positive tests, does that correction look sensible. We know, roughly, that there were ~300,000 positive tests in the first peak. So that's mostly 4-5 months ago. For the 28 day cutoff, that gives a regular remaining life expectancy of 15-19 years to account for the observed deaths. For 60 days, the range is 37-46 years.

The problem with this is that we don't know the demographics of those tested. However (you can look these things up in actuarial tables), for the 28-day number to be accurate, those tested have to be quite old - over 65, whereas the 60 day number would be right for a population of working age - say typically in their 40s. Given that we know that there was little testing in nursing homes, the 28-day life expectancy looks a bit wrong, whereas the 60-day version looks reasonable if you're testing a lot of health workers. It's not definitive, but there's a hint that 28 days is overcorrecting.

Another thing to do is look at the rate of corrected deaths over time. This is what it looks like for 28 days:

and this is for the 60 day data:

There's quite a difference there. Note that our expectation is that the probablity of death from other causes is constant (approximately) over time, so that the overall rate will increase over time (it's constant if there's just a single input at a fixed point in time, but we're testing more people as time goes on so the population is growing - the relatively recent big spike in positive tests hasn't worked its way into the data yet).

From this point of view, the 28-day graph looks a little suspect - it starts to rise 28 days after significant testing, but after day 90 there's a decline. That's plain wrong, indicating that there's a correlation with the time of the test (there are a lot of positive tests in days 30-90 which is when the big first peak was). If they're correlated with the positive test, there's going to be some element of correlation with the cause of the test - namely COVID-19 itself.

By contrast, the 60-day chart has the right shape. The problem is that any large cut off will have the right shape, so it's not telling us anything about the correct number to cut the data at, just that 60 days is beyond it.

The thing is, if you had all the data (inclduing demographics) you could do this properly, and work out what the optimal cut off to minimize errors should be. I just haven't seen that yet. But I'm fairly sure that, while the original quoted numbers overestimated the number of deaths, the new numbers with a 28-day cut off are underestimating the true impact, and they might even be more in error the other way than the original figures were.

Sunday, September 13, 2020

What level of response to Covid-19 do we need?

Short version: quite a lot.

The long version below is a rough attempt for me (a former scientist) to understand the problem and the rough nature of possible solutions.

It starts with the R factor - essentially the number of people that an infectious person passes the disease on to. The more people you infect, the faster the disease grows. If you manage to get to a situation where people infect less than 1 person on average, the disease will decline.

I've seen R factors for the UK of round about 3 in the early stages. That's the effective R factor - even then, before our governmenat finally bothered with a form of lockdown, a significant number of individuals and organisations were taking preventive measures. So the intrinsic R factor could be a bit higher, although 6-9 is the highest range I've seen.

As we've seen, the effective R factor depends on reactive measures and behaviour. So what sort of measures are necessary to get R below 1?

Reember, R is the number of people who can pass the disease on to. So you need to reduce that number from anything between 3 and 9 to below 1. Actually, you need to go a decent way beyond that to give yourself a margin of error. So I'm going to say that you need to reduce the likelihood of passing the infection of by a factor of 10, which basically means reducing direct contact by a factor of 10.

And reducing transmission by a number that large is quite a challenge. If you just carried on and tried to use face coverings and social distansing, I don't think you can get there - they might give you a factor of 2 or 3, which isn't enough.

What level of contact do people have with others? I'm going to try and estimate the number of person-hours of contact you generate in a week. Remember, this is before the pandemic.

  • Work, you might have 40 hours a week in an office or building with 25 people, giving 1,000 contact-hours
  • Commute, You might have 10 hours (1 hour per journey), but cross the paths of 100 people, giving another 1,000 contact-hours
  • Play, a restaurant or bar might be 100 people or so for a couple of hours, but you have to get there and back, and might do it several times a week. And I'm ignoring massive events like theatre, gigs, festivals, sports. So that could be another 1,000 contact-hours

That gives 3,000 contact-hours per week, pretty much evenly spread across the 3 categories. And we want to reduce that by a factor of 10 to 300.

Even if you avoid public transport (which has other negative implications for the next major disaster of climate change heading our way) you don't get anywhere near. Carry on as before and drive to work and you only get from 3,000 to 2,000, nowhere near the 300 desired.

As far as I can see, the only way to even get close is to not go in to the workplace, which cuts out the commute part as well. And, in addition, cut down the play component and take all the face-covering and social-distancing precautions when you do.

I see no possible scenario in which widespread full-time return to the office can possibly be entertained as being safe.

Even if we use the idea of only some people going to the office 1 day a week, you're still using up a very significant portion of your contact budget.

Of course, there are those who do need to go to a workplace. There are those who do need to commute. That's even more reason why those who can work from home should do so, to give a bigger allowance to those who really need it.

You can, very roughly, estimate your own exposure. For example, I went out to Ely and had a guided walk with a group of a dozen people or so around the town today. Socially-distanced, including the travel and an outdoor lunch, that was something near 50 contact-hours.