Hacker News Reader

Comment by dmje | Hacker News Reader

UK citizen here.

I just don't think this "collapse" is the case.

My gut feel is that everyone has stopped doing lateral flow tests and reporting those tests: noone wants to be isolating now the summer holidays have started, and many people are now ignoring track and trace because they're scared of the possible "pingdemic". I know many people who have stopped scanning in to venues. I reckon many suspect they may have Delta but because the holidays have started and they've been cut loose from gov mandates of mask wearing etc, that the "crowd responsibility" factor has almost totally dissipated.

There may be evidence otherwise but I suspect actual figures are far higher than is being reported.

View on HN

> My gut feel is that everyone has stopped doing lateral flow tests and reporting those tests: noone wants to be isolating

Okay that's easy to check: data should show a reduction in positive tests roughly on par with the reduction in number of tests.

Recent peak value of tests per day was 16.42 per 1000 citizens on Jul 17th. At that time there were 630.57 positive test results per million citizens.

On July 28th (latest data) tests declined to 13.46 and positive cases to 443.45. That's 18% fewer tests and 30% fewer positive cases, ballpark twice as much.

I don't think your gut is telling you the full story.

All values are 7-day rolling averages to avoid weekday spikes or anything. Source: https://ourworldindata.org/explorers/coronavirus-data-explor... (use the selector to switch between tests, positive cases, or other metrics)

UK resident here: I don't know of anyone who has gotten covid in many months, and nearly everyone I know is vaxxed.

I believe this Delta collapse. It's exactly what you'd expect given our pace of vaccinations.

If we're going by anecdotal evidence, I know many more people that get the flu every year than that got covid-19. But I don't think my (on average) responsible acquaintances with (on average) workable-from-home jobs tell the whole story.

You're right though that the numbers somewhat suggest that we're managing to beat the Delta variant. Is it due to vaccinations? The vaccination rate has been slowing for months now, yet the UK still had that spike. The difference in people that decided to get vaccinated between early June (start of positive cases spike) and July 21st (when the decrease set in) is 9%. (from 59% to 68%). It doesn't seem to me that a few percent more vaccinations explains everything. A percentage in the upper 60s, for first vaccinations no less (not even a full vaccination), should not be enough to beat Delta from what I heard about the R0.

(Data source is the same OWID link, you can configure it to show vaccination data.)

I'm not saying it's due to vaccinations, I'm saying it's what you would expect given our pace of vaccinations. All we have are correlations.

The reason I say this is that we have a relatively small population of unvaccinated people, and so with the high R0 of Delta, you'd expect it to rip through these unvaccinated pockets quickly and burn out, which is what we're seeing.

I'd expect a few more Delta mini-waves similar to this in the coming months.

The lateral flow tests results can be self-reported according to the NHS website, so you can't definitely draw conclusions from that data.

Cases don't matter, hospital admissions also show sign of wave going down. And that doesn't depend on people's mood.

A UK person I know said they won’t install the “pingdemic” app because they would have to isolate every five minutes; instead they just use the camera app and pretend. And that’s a vaccinated person who’s pretty cautious. I would assume it’s a widespread sentiment,

i think schools closing, euros ending, the nhs covid19 app ping and holidays starting are actually contributing to a lower number. scotland’s early school holiday should correlate at least part of this hypothesis.

a couple of q’s thou:

1. where do you report lateral flow tests? i do mine at home every day as my work sent out boxes of them.

2. if a lateral flow is positive, aren’t you supposed to do another and then a pcr test?

> where do you report lateral flow tests?

https://www.gov.uk/report-covid19-result

> if a lateral flow is positive...

I imagine that that report result site will refer you for a PCR test, although thankfully I've never personally been in that position.

That's exactly what happens. False positives with lateral flow tests are not unknown. In fact, if the prior probability of having covid is less than about 0.5%, you still have <50% probability of having covid even with a positive lateral flow.

ONS sampling disputes this.

For anyone else not from the UK, ONS seems to be some official body doing some sort of, well, sampling.

Do you have a link to their current stats? I found this page[1] but it's not very accessible, not sure this is what you mean.

[1] https://www.ons.gov.uk/peoplepopulationandcommunity/healthan...

This is a more friendly presentation of the same data, though note there’s a lag in the ONS data as it’s from ongoing sampling rather than self reported tests. At time of writing this is up to 30 July (next update 6 August):

https://www.ons.gov.uk/peoplepopulationandcommunity/healthan...

Does this help? https://coronavirus.data.gov.uk/

It really makes me wonder why the authorities don’t do proper randomized surveys? It is obvious that the testing numbers will go up and down as the population goes scared and then goes don’t care.

UK Citizen here, my parents have been part of the ONS random survey for about 18 months now, every 2 (or so) they have a visit at their doorstep from the ONS who test both of them, they let them know the results via the NHS app and they get £25 of vouchers for each visit.... They have to be honest about who they have been in contact with, where they have been (E.g. supermarket) and if they wore a mask...

This data goes into modelling for how the government should make decisions about easing lockdowns etc. The UK has some of the best genomic sequencing in the world which is why we are able to identify variants so quickly.

It's just a shame it took the government 1 year to get ready and lied to us in the meantime.

*every 2 (or so) weeks.

The UK is one of the few countries in the world that does have proper randomized studies of Covid-19 prevalence, and has done since close to the start of the pandemic. Unfortunately the results lag behind the actual number of new infections by long enough that we wouldn't expect the drop in cases to cause a corresponding drop in the numbers coming of those studies just yet.

there is some sort of on going testing with a set group that I've been told was randomly choosen.

My sister/ husband and neice (uk based) are currently being tested once a month since the outbreak (mid 2020 ish) and they where told that they was picked at random,

All of them have had covid - and probably would not have been discovered without this testing - since all their cases have been asymptomatic

Cause they’re already at maximum capacity with higher priority projects, I would imagine.

Several municipalities in the US did do randomized sero surveys earlier in the pandemic.

I can’t provide a citation but largely these were abandoned because they were influence by a couple factors

* participation was obviously still voluntary and largely mirrored the same trends you describe

* the data was not short term actionable and took up testing infrastructure at the time

* it was dangerous - I.e., public health workers were threatened, assaulted, and in at least one case shot at.

> My gut feel is that everyone has stopped doing lateral flow tests and reporting those tests

Is that where this data comes from?

I'm not sure it is. I think it's separate sampling.

Shouldn’t we see this reflected in hospital admissions?

They're a lagging indicator, as are deaths. So, eventually.

If you stare at the government's site https://coronavirus.data.gov.uk/ hospital admissions seem to have levelled off, and maybe deaths are levelling off but it's too early to be sure.

People get infected (and begin testing positive) first, then some of them get sick (a few more will seek a test and be positive) and then some of those seek hospital admission or call for advice and the advice for their symptoms is "get to hospital" and almost all of the people who eventually die get admitted to hospital first, in some cases dying after weeks of attempted care.

I think there's room for some pessimism without going so far as your parent poster, cases may be down but not as much as it seems. Lots of people seem to have gone off around the country on holiday which seems crazy to me, but clearly didn't make things much worse. People can choose not to seek a test while they're alive, but corpses have no say in what the post mortem data says.

Looking at the official data[0], the number of daily hospital admissions is still increasing: https://i.imgur.com/XkbImVg.png

But hospital admissions usually lag behind the number of infections. If in a week or two admissions are still going up while the number of new infections drops, then this theory may be correct.

[0] https://coronavirus.data.gov.uk/

Hospital admissions usually are a bit delayed, but yes, presumably people won't be holding off on being admitted to the hospital the way they would hold off on getting tested if GP is right. (Personally I'm not sure how many more people will now, as compared to a few weeks ago, decide that testing is more of a nuisance than helpful for protecting your loved ones, but then I'm not from the UK so what do I know.)

And if we don't see a rise in hospitalizations that is any worse than a typical flu season, then it really doesn't matter that no one is testing.

At least provide data if you're going to be like "it's just a flu". I'm willing to believe you mean well if you are only talking about people that are fully vaccinated 14+ days ago, but in general this sounds like an excuse.

I didn't say that it is or is not. I'm saying that if we see numbers reach the point that it's no worse than the flu, it becomes a big nothing burger.

Reply on news.ycombinator.com