Covid cases increasing

[weepingwillow22]

Is anyone else using the zoe app where around 3.6m people are reporting any or no symptoms of covid?
covid.joinzoe.com/

Until recently the app was showing a nice fall in people reporting symptoms however in the last 10 days they have stabilised and started to increase again.
Peak april 1st 2.1m suspected cases
8th may 230k suspected cases
17th may 252k suspected cases

I realise the data is self reported however the trend should be accurate and reflective of actual cases.

Is anyone else worried about this? We can't rely on government data on cases as test numbers are increasing/unstable and the death data lags by 2 weeks. I was going by this app to show some progress but it seems like we are going in the wrong direction.

I have also looked at the maps in the app and reported symptoms are increasing in London despite the government is saying the r number is very low here.

If this app is correct the death rate should stop falling soon.

@GabriellaMontez This link gives you a brief overview of the modelling. The numbers I cited earlier are ouputs from the model
covid.joinzoe.com/post/science-magazine-covid

I'm not seeing anything that says they altered the numbers to reflect an increase I participants. Are you?

They make an estimate of total numbers based on the results from the people participating, using a model, which I'm fairly sure will have taken into account the number of people submitting results...

This link explains more: covid.joinzoe.com/post/covid-cases-england

Are you 'fairly sure' or does something in that link make it clear?

Because I'd really like to know for sure now.

@Gabriella If you want more detail here is the link to a journal article with the full methodology
www.nature.com/articles/s41591-020-0916-2

They appear to have changed now to reporting new cases each day rather than total symptomatic cases, though I can’t see any explanation of why. Ideally they should have continued to display the old data as well rather than replace it with a new one.

There seems to be a suggestion that the uptick in cases on the app is related to the introduction of loss of taste and smell as a symptom.

I think (hope) that it is just a case of people discovering the app. I've only recently discovered it.

That sounds plausible for the increase.

I wonder why they've changed the data though? Annoying they've not kept the old figures too for comparison.

I'm not clever enough to interpret the info absolutely, I'm basically filling in the gaps with assumptions.

Fair enough ihavent!

I think before an OP makes a scaremongering thread like this 'covid cases increasing' the least they can do is read through.

More people discovering the app would not of itself explain an increase in the percentage rates of people estimated to have symptomatic Covid in each area. That would only happen if for some reason the app had started attracting a higher proportion of new users with symptoms than had previously been the case. I am also struck by the fact that the areas showing higher percentage rates tend to be grouped together (eg Lancashire, West Midlands, South Wales). It seems likely to me that the app is picking up an increase in infection rates in some parts of the country.