Health Nerd Profile picture
Feb 21, 2022 13 tweets 4 min read Read on X
It feels like the entirety of the Australian media has been making errors about COVID-19 deaths in the last few days, so here's a brief explanation of the issue 🧵1/10
2/10 The Australian Bureau of Statistics recently released a report looking at registered/reported deaths as of Jan 2022. The numbers don't match crude COVID-19 reporting because this is a more complete system abs.gov.au/articles/covid…
3/10 This report is based on death registry systems, which themselves are based largely on death certifications performed by (mostly) doctors. These are assigned ICD-10 codes based on the reported causes of death and comorbidities
4/10 The report found that 2,639 people had died and had COVID-19 mentioned ANYWHERE on their death certificate

Of those, 2,556 (96.9%) were DIRECTLY CAUSED by COVID-19
5/10 So, 96.9% of the COVID-19 deaths in Australia were FROM COVID-19, and 3.1% were 'with' COVID-19. Most of the 'with' deaths were caused by either cancer or chronic disease of some kind
6/10 Of the 2,556 deaths 'due to' COVID-19, there were two main categories:

1. COVID-19 as either the only CoD or as part of the causal sequence only
2. COVID-19 plus comorbidities
7/10 Importantly, the 'causal sequence' is NOT COMORBIDITIES

As the ABS report explains, these are conditions CAUSED BY COVID-19 (i.e. COVID-19 pneumonia)
8/10 So 30.5% of the deaths caused by COVID-19 in Australia had COVID-19 or the disease progression as the ~only~ CoD. Not 8% as you may have heard from whichever incorrect media source
9/10 Of the remaining deaths, there were a variety of comorbidities listed. Importantly, however, the median age was very similar to the median age of death of Australians generally. The comorbidity proportions were a bit lower than you might expect, but not drastically so
10/10 So, there we have it. 2,556 people died in this time period FROM COVID-19. Another 83 died 'with' COVID-19. The people who died were not especially old nor sick compared to people who die of any other disease
11/10 If you've heard the number 83, or that 92% were horribly sick already, you've been misinformed. The ABS report shows very well that most Australians recorded as dying from COVID-19 would not have died at that point without the infection 🤷‍♂️
12/10 Also, probably worth noting that the comorbidity percentages really weren't that high in the group of people who died. Somewhere around 1 in 4 Australians has at least of the conditions they listed
13/10 Small addendum - the actually interesting/depressing stuff in this report, like the clear socioeconomic gradient in the COVID-19 deaths, has largely been ignored, which is terrible

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Health Nerd

Health Nerd Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @GidMK

Mar 4
The final large published trial on ivermectin for COVID-19, PRINCIPLE, is now out. Main findings:

1. Clinically unimportant (~1-2day reduction) in time to resolution of symptoms.
2. No benefit for hospitalization/death. Image
Now, you may be asking "why does anyone care at all any more about ivermectin for COVID?" to which I would respond "yes"

We already knew pretty much everything this study shows. That being said, always good to have more data!
The study is here:

For me, the main finding is pretty simple - ivermectin didn't impact the likelihood of people going to hospital or dying from COVID-19. This has now been shown in every high-quality study out there.pubmed.ncbi.nlm.nih.gov/38431155/
Read 8 tweets
Feb 20
Fascinating study.

What's particularly interesting is a finding that the authors don't really discuss in their conclusion. These results appear to show that gender affirming care is associated with a reduction in suicide risk 1/n
2/n The paper is a retrospective cohort study that compares young adults and some teens who were referred for gender related services in Finland with a cohort that was matched using age and sex. The median age in the study was 19, so the majority of the population are adults. Image
3/n The study is very limited. The authors had access to the Finnish registries which include a wide range of data, but chose to only correct their cohorts for age, sex, and number of psychiatric appointments prior to their inclusion in the cohort.
Read 11 tweets
Oct 26, 2023
These headlines have to be some of the most ridiculous I've seen in a while

The study tested 18 different PFAS in a tiny sample of 176 people. Of those, one had a barely significant association with thyroid cancer

This is genuinely just not news at all Image
Here's the study. I'm somewhat surprised it even got published if I'm honest. A tiny case-control study, they looked at 88 people with thyroid cancer and 88 controls thelancet.com/journals/ebiom…
Here are the main results. There was a single measured PFAS which had a 'significant' association with the cancer, the others just look a bit like noise to me Image
Read 7 tweets
Oct 11, 2023
A new study has gone viral for purportedly showing that running therapy had similar efficacy to medication for depression

Which is weird, because a) it's not a very good study and b) seems not to show that at all 1/n
Image
Image
2/n The study is here. The authors describe it as a "partially randomized patient preference design", which is a wildly misleading term. In practice, this is simply a cohort study, where ~90% of the patients self-selected into their preferred treatment sciencedirect.com/science/articl…
3/n This is a big problem, because it means that there are likely confounding factors between the two groups (i.e. who is likely to choose running therapy over meds?). Instead of a useful, randomized trial, this is a very small (n=141) non-randomized paper
Read 15 tweets
Oct 6, 2023
This is SO MISLEADING

The study showed that COVID-19 had, if anything, very few long-term issues for children! As a new father, I find this data very reassuring regarding #LongCovid in kids 1/n Image
2/n The study is here, it's a retrospective cohort comparing children aged 0-14 who had COVID-19 to a matched control using a database of primary care visits in Italy
onlinelibrary.wiley.com/doi/10.1111/ap…
3/ The authors found that there was an increased risk of a range of diagnoses for the kids with COVID-19 after their acute disease, including things like runny noses, anxiety/depression, diarrhoea, etc Image
Read 13 tweets
Sep 20, 2023
This study has recently gone viral, with people saying that it shows that nearly 20% of highly vaccinated people get Long COVID

I don't think it's reasonable to draw these conclusions based on this research. Let's talk about bias 1/n Image
2/n The study is here. It is a survey of people who tested positive to COVID-19 in Western Australia from July-Aug 2022 medrxiv.org/content/10.110…
3/n This immediately gives us our first source of bias

We KNOW that most cases of COVID-19 were missed at this point in the pandemic, so we're only getting the sample of those people who were sick enough to go and get tested
Read 12 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(