Thread by @DRTomlinsonEP

'Myth busted'🧴: FACT CHECK

We've all learned that some prefer narrative control & advertising, to an unbiased reporting of the observable nature of reality.

W/o preconceptions re @timspector & team's approach, this arresting take from @apsmunro made me take a closer look.
/1

https://twitter.com/apsmunro/status/1446171545546723329

'there is barely any difference at all between Delta & Alpha in symptom duration, severity'

Methods: 'children were considered to have COVID-19 if proxy-reported with relevant symptoms'

Fact check🧴: This is selection bias.
It's impossible to conclude that Delta is no worse.
/2

E.g. with made up figures:

If the truth was that Covid19 symptoms occurred in 1% & 99% of kids w Alpha & Delta respectfully, the study protocol would be blind to this.

'there is barely any difference at all between Delta & Alpha in symptom duration, severity' = ADVERTISING.
/3

'there is barely any difference at all...in symptom duration, severity'

We now know this is an impossible to make statement, but what does 'barely any difference at all' mean?

Fact check🧴: 14 of the 16 most frequently occurring symptoms were more prevalent for Delta, all ages.

'there is barely any difference at all...in symptom duration, severity'

Statistically significantly greater odds of 9 symptoms occurring with Delta #COVID19 compared to Alpha (red bars) in children.

/5

Significantly greater odds of headache, rhinorrhoea, sore throat, anosmia/dysosmia, fever, dizziness, chills or shivers, eye soreness, hoarse voice.

'there is barely any difference at all...in symptom duration, severity'

Fact check🧴: Selection bias & now SELECTIVE REPORTING.

'there is barely any difference at all...in symptom...severity'

We've covered the prevalence & odds ratios, but symptom severity was not reported in terms of its intensity (e.g. of headache, fever etc).

Fact check🧴: Selection bias, selective reporting & now 'false assertion'.

'there is barely any difference at all...in symptom...severity'

But what *can* be concluded regarding symptom severity?

Hospitalisation rates:

Alpha 2.0%
Delta 2.2%

That's a 1:50 chance of hospitalisation in children with #COVID19 symptoms in this study.

Seems important!

What do @timspector & team say about this (pic)?

Miss it? 🤦‍♂️

Check out the last line under 'Findings':

'Few children presented to hospital...'

I always get suspicious when people use inexact terms: in their defence it was only 30 of 1,400 kids.

But is this narrative control?

But maybe I'm getting carried away.

1:50 symptomatic children hospitalised, but maybe this ends up being such a very, very small number because so many children have ASYMPTOMATIC infection.

😳 (pic)

Fact check🧴: Using @apsmunro's '50%', that's 1:100 likely to be hospitalised.

Oh dear.

This looks like a disaster UK-wide.

Some of the data I used in this letter to my MP is looking way too optimistic.

I'm going to need to double the hospitalisation data for the current situation of unmitigated UK transmission within schools.

https://twitter.com/DRTomlinsonEP/status/1446853026329792516?s=20

Last:

risk of "long covid"

Others have remarked on the quotations. I'll limit my comments to the data.

👉The protocol was not designed to determine symptom duration beyond 28 days.

Fact check 🧴: It is scientifically invalid to make any comment re long covid from these data.

'Turns out there is barely any difference at all between Delta & Alpha in symptom duration, severity or risk of "long covid"

🧴Invalid statement via selection bias, selective reporting & 'false assertion'.

Tweeps, is this advertising & narrative control?

https://twitter.com/apsmunro/status/1446171545546723329?s=20

@RCPCHPresident @rcpch_trainees @RCPCHtweets FYI