Hacker News Reader

Comment by gambiting | Hacker News Reader

Have you watched the presentation yesterday? Or just reading the news articles today? Because that's precisely not what they said.

What was said was - from epidimiology point of view, there's no point in telling everyone to stay at home right now when the spread is in the early stages. It will have a marginal effect, and the spread will continue anyway. By government's own estimates 80% of people in the UK will get infected, and imposing a quarantine right now won't change that. What would help is setting the quarantine for 14-18 weeks, and they know that people won't abide by that, so instead, they clearly said they will impose quarantine just before UK hits the predicted peak infections - to lower the stress on the NHS and to achieve the maximum effect of the quarantine. Like, it all made sense if you watched the actual conference.

View on HN

I watched a part of the presentation, but it was actually unclear to me what is the peak infections mentioned. I mean, in Italy, the numbers are very small compared to 80% of population, and the health system is already not able to cope, and mortality rate is quite high. Does it mean that UK will wait for their numbers to be 10 times the current Italy numbers before starting the quarantine?

No, it means that UK government knows that it cannot stop the situation getting a lot worse, and a quarantine will be a lot more effective as a tool later on. They seemed to be quite realistic about this, not going " it will be all fine, trust us". They know things will get fucked and a full quarantine is one of the many tools to use against the virus - so let's use it when it's most effective. Otherwise like they said - if we quarantine the entire country right now, it won't stop the spread anyway. So we'll come out of quarantine and then have to do it again in a month - and people will get fed up with it and won't obey a quarantine if it goes on for too long or too frequently. So if the spread is inevitable, let's use the quarantine at the point that prevents the most possible number of infections.

I can see some discussion about trying to figure out which restrictions will work best with the psychology of a given population.

But the idea that the "most effective" time is going to be later doesn't make much sense to me. Marginal rate changes in exponential spread are most likely to have the biggest impacts early. It doesn't much change the total volume of people who get sick, but it very likely changes the volume of people who get sick at once, and that can save lives.

This sounds like saying "we can only imagine quarantine working if things get fucked, so we're going to let things get fucked, so that the quarantine will work except for the fact that things are already fucked oh well too late."

Sweden is making the same decision and they explained it quite well. A quarantine means children stay home from nurseries, daycare and school. This increases the stress on the medical system because some doctors and nurses have to stay home to take care of their children.

In addition, new constellations of people gathering become likely. E.g. children from different schools may play together more in the neighborhoods, opening new pathways for the disease to spread.

So when a quarantine is not expected to help in most cases (because most people are not infected), the drawbacks easily outweigh the benefits.

I think I get it now and I think it's probably not about the stresses of closing school: I think it may actually be a calculated gamble that what you want is the low-risk population widely exposed first, falling ill at a rate just below the level at which you can treat people. If that rate threatens to climb, then maybe they close the schools and try to slow the rate. Otherwise the at-capacity wave is moving people from the "not immune potential vulnerable & vector" population into the "immune and not a vector" population.

If that's the case... it seems like it heavily relies on timing, good data & models, and more precisely timed and effective social cooperation, and I'm not sure I'd want to bet on having all those pieces in place, but it is interesting.

Also something that was mentioned during the conference with the prime minister yesterday - if we close the schools and tell the kids to stay at home, who are some of them going to stay with? Their grandparents. The group that's at the highest risk from the virus.

They are smaller pools in the neighborhood, than in a school. Slowing down the infection rate is the whole game right now. Closing schools is a huge step toward that.

It seems strange a hospital cannot respond to daycare needs in their own population in a situation like this...

It’s not the size of the pool but the novelty of the configuration of people. Every time an infected individual meets a healthy one that’s a new chance to spread the infection. But you can only spread it once to each person, so if you meet the same people over and over that’s okay. When you ask people to change their routines, you also encourage new “lateral” disease spread.

I don’t think hospitals have “emergency daycare” available generally speaking. If it does come down to quarantine, I’m sure efforts will be made. It is still a cost and a negative outcome of the quarantine so my point remains: when the benefit is very small, the costs easily outweigh the benefits.

Now that isn't right. Meeting the same people over and over spreads the disease - we don't all get it at the same moment, and each contact even with the same person is a new opportunity to get infected (because they got it from somewhere else since your last contact).

Social distancing, reduce frequency of contact, stop meeting in large groups. That's the whole ballgame now.

Of course, reducing the frequency of contact with other people in general is good. I don’t think anyone is arguing against that. What I’m saying is that quarantine can cause more novel contacts.

Statistically speaking there is a good chance no-one at your work place has the virus yet. We are still counting infections in the hundreds or the thousands in populations of millions. If all you do is to spend time at home and at work, there are good odds you’re just switching between two disease free spheres. You can’t get sick without meeting sick people. The only thing that will change that steady state is contact with new people (by yourself or your coworkers/family).

The point the government was making in this case, I believe, is that quarantine can in fact cause more novel connections, faster, and perversely increase the velocity of the spread. The people who are at home will not just sit there. They will meet other people and those people may be from outside of their ordinary, currently disease free social sphere. Thus new vectors of infection spread are introduced that would not have existed without the quarantine.

A quarantine can certainly help down the road, it’s a numbers game. But right now they’re doing the math and think it would be a net negative.

Again, not right. New people are not necessary at all. People are not in little bubbles with no communications between them. If there's even one person in common between two groups, then infection can/will in time cross between.

An infection travels like a game of 'telephone' passed from one to the next, right across the world.

Quarantine will initially regroup people, but in much smaller groups. If the disease is not yet widespread, then your new cohort will also likely be disease-free. There may be a tiny spike as an infected person in a large group joins a different, small group. But this is a good thing, since that group is smaller, and fewer people will now be at risk.

No, quarantine is the only effective way forward.

Of course it will in time spread. By having a quarantine right now, that time would be shorter, not longer. That’s the point being made.

No. Smaller groups contain and slow down the spread. It extends our response time.

That does not contradict anything the government said. All else being equal, of course smaller groups, in the sense of people seeing fewer other people, is better. That’s not the argument. Nobody is saying “we should see more people”.

All else is not equal. You pull the quarantine lever, you increase social distancing but you also increase local mixing. Costs and benefits. The right choice at any given time depends on the disease spread. Computer modelling and simulation tells us when there is a net benefit.

The UK are not following quarantine and isolation protocols to mitigate the virus because they are trying to protect the economy post-Brexit.

I mean, maybe, but I'm yet to see proof of that. The reaction yesterday seemed perfectly reasonable.

Ideally, a government would quarantine all probable cases with the aim of extinguishing the disease.

But if the disease is spread enough (inside or outside your borders) so that is impossible, the next best thing is to reduce social interaction just before your health system collapses, and just enough to avoid a collapse. That minimizes both the cost and time of isolation.

I understand that the approach has some degree of internal consistency, but it requires a huge amount of trust in the underlying models, and in our ability to pull the right levers at the right time in the future. I can totally accept that any approach _must_ address what happens for the entire period from now up until immunity or vaccination - we know we can't keep everything shut down until 2021. But the government is hoping to tap the brakes when the outbreak is going at 100mph, compared to controlling acceleration from the standing start of lockdown. This is an immense and shocking gamble. I hope it works, although would mean every other country appears to be making a huge mistake by shutting down now.

Reply on news.ycombinator.com