Confused about the ONS survey

[Redolent]

Why does it show that cases are steady, when government testing shows an increase? Is it missing local spikes?

If we are seeing rising infections especially in deprived or minority areas, maybe English may not always be a first/fluent language, would this affect the tests that are sent back for processing? Or is this a rubbish theory?

I think that the explanation (not sure if it's accurate) is that extensive government testing is being done in areas with outbreaks. That means that more of the people with covid in those areas are being tested and reported. So there might be more positive tests, but not necessarily more actual cases.

If the ONS test is like the one my DD did, it's a sample of 20,000 over the country with a really low number coming back positive (e.g. 15 out of 20,000). That gives a wider population incidence.

So it is plausible that positive test results increase, without that meaning there's actually more out there.

@Justajot

I think that the explanation (not sure if it's accurate) is that extensive government testing is being done in areas with outbreaks. That means that more of the people with covid in those areas are being tested and reported. So there might be more positive tests, but not necessarily more actual cases.

If the ONS test is like the one my DD did, it's a sample of 20,000 over the country with a really low number coming back positive (e.g. 15 out of 20,000). That gives a wider population incidence.

So it is plausible that positive test results increase, without that meaning there's actually more out there.

That partly makes sense.

But is it possible that some of the households most likely to be covid-positive are the least likely to conduct and return self-swabs? (Language barriers etc) Or does this not factor at all into the methodology?

To add to that, I’d be curious to know what the average age is of an ONS respondent, and how that matches up with the average age of those currently being infected (young adults). Are young adults as likely to go through this process?

I see what you mean, but that won't have changed since the ONS started their monitoring. I assume there's a lot of methodology to the ONS approach, but no idea what. Given the low numbers of positive tests in their samples, I'd guess there's quite a range in the actual estimate. The public might get the headline figure of 1 in 1500 have covid, but the detail might be something like 1 in 1350 to 1 in 1650 have it, with a confidence interval of x.

There's also been quite a fuss about the new cases actually not being in minority ethnic communities. I think that some groups in society have relaxed as they think it's unlikely to impact them, but it is that kind of approach that leads to outbreaks.

The government testing numbers are increasing because more tests are being done. And particulalry because more tests are being done in places where we know there are more cases.

Really the fact that cases found is increasing is a good thing. It doesnt necessarily mean there are more cases, just that we are finding them. And if we are finding them those people will get treatment if needed and isolate from others.

Another important statistic is the test positive rate. I.e. the percentage of tests done which come back positive. This has not increased and is at low level. I read somewhere that WHO recommend a positivity rate of under 5 to keep control of the virus. Ours is currently around 0.6

Yes the ONS household survey format is very much geared to the older (self included!) population, official on Her Majestys Service envelope, official letter format with large text! Very reassuring and not a scam.

Hopefully, they will get younger members of the household participating through the incentive of shopping vouchers or perhaps there's to be a Social media campaign to engage younger people to take part?

@BameChange123

Yes the ONS household survey format is very much geared to the older (self included!) population, official on Her Majestys Service envelope, official letter format with large text! Very reassuring and not a scam.

Hopefully, they will get younger members of the household participating through the incentive of shopping vouchers or perhaps there's to be a Social media campaign to engage younger people to take part?

I’m hoping for the latter too. I saw this thread about the issue a few days ago...young people simply not wanting to volunteer.

www.mumsnet.com/Talk/am_i_being_unreasonable/3995234-to-be-peeved-that-my-kids-refuse-to-be-tested-for-Covid

The latest results also state that there is “
‘.. no evidence from this survey .. a difference in COVID-19 infection rates between regions’. How can this be the case?

Isn't "test positivity" the key statistic for knowing if virus is increasing? Coz increase in that would mean more actual prevalence.