US Covid Deaths Reaching 1M: 997041 in 2022/18/03 18:51 MDT

[bob012345]

Being a confirmed skeptic myself, I would always question results, https://www.thelancet.com/journals/lanam/article/PIIS2667-193X(22)00081-3/fulltext does, as I said earlier cover things very well (just my opinion though). there is much more to the paper than the snippet quoted,

Findings:
"Groups with lower health insurance coverage had significantly higher mortality as well as greater case counts and hospitalization. Early in the pandemic, they were also less likely to be tested for COVID-19. Applying our regression estimates, we estimate that had there been full health insurance coverage of the population, there would have been 60,000 fewer deaths, 26% of the total death toll in the period of this analysis."

Have you read the entire Lancet piece? It's pretty well done with plenty of details as to methods and findings.

I skimmed through it. I noticed this;

Strikingly, we found that low insurance coverage remains associated with increased hospitalization despite both patient and provider financial disincentives for hospitalization of the uninsured.

It seems to me this seriously undercuts their argument. One cannot make claims about how many people would not have died if the situation were different or if the U.S. did not have a market-driven healthcare system. I do not consider that valid science.

 

[Oldman too]

I skimmed through it. I noticed this;

Strikingly, we found that low insurance coverage remains associated with increased hospitalization despite both patient and provider financial disincentives for hospitalization of the uninsured.

It seems to me this seriously undercuts their argument. One cannot make claims about how many people would not have died if the situation were different or if the U.S. did not have a market-driven healthcare system. I do not consider that valid science.

Skimming is good, but perhaps we should desist with the personal points of view and let the interpretation be discussed in a larger audience. I personally feel that Lancet is a reputable publication that wouldn't have put a marginal or questionable paper out without addressing your particular concerns, perhaps you could contact them about... "I do not consider that valid science."

 

[Vanadium 50]

I'd expect case count to be unaffected or lower (due to less testing) and hospitalizations lower due to lack of insurance.

If patients have more than one condition - and multiple conditions is a strong predictor of Covid severity and probability of hospitalization - there may be an incentive to call it Covid for reimbursement purposes.

 

[bob012345]

Skimming is good, but perhaps we should desist with the personal points of view and let the interpretation be discussed in a larger audience. I personally feel that Lancet is a reputable publication that wouldn't have put a marginal or questionable paper out without addressing your particular concerns, perhaps you could contact them about... "I do not consider that valid science."

That is fine. I'm not saying Lancet is not reputable.

 

[Godot_]

[...] We may all be living with it for a very long time so the emphasis should be on learning to live with it and each other regarding different views on handling it.

https://www.nytimes.com/2022/05/16/health/covid-reinfection.html

a) It's not "We may", it's "We will all be living with it..."

b) IMHO the emphasis should rather be put on developing good early and acute treatment options, especially effective medications to cut infection and symptoms in the early stage. Paxlovid is a decent first attempt, but there should (hopefully) be better options in the future. AND doctors should be as well aware of as willing to prescribe these medications in cases of early CoViD-19, high chance of "less severe disease" be damned.

After all, while long CoViD is apparently less frequent with the Omicron strains, it's still there. So as the case numbers will be much higher, the absolute number of long CoViD cases will be higher, too. And the impairment by long CoViD can be damning and disabling for a long while - dead neurons are dead. Those severe cases will become a significant burden for the healthcare and welfare systems, and the loss of QALYs shold be considered too, from a national economy viewpoint.

All of this will, however, also require adequate testing... On a different note, some of you might find these two online publications from the Institute of Health Metrics and Evaluation (IHME) interesting:

https://www.healthdata.org/special-...covid-19-and-scalars-reported-covid-19-deaths

https://www.healthdata.org/news-rel...obally-more-double-what-official-reports-show

They're not the freshest, and I concede I didn't delve into possible peer-reviewed publications for a followup, but IHME is considered a highly reliable source.

 

[mfb]

Do we have long Covid stats for the US?

The UK estimates 2 million, or more than 3% of the population, having symptoms now without an infection in the previous four weeks (as of May 1).

The illness is impacting the day-to-day activities of 1.4 million people, with nearly 400,000 reporting that their ability to go about their day-to-day activities has been ‘limited a lot’.

 

[Oldman too]

This might be a close approximation.
https://www.cdc.gov/mmwr/volumes/71/wr/mm7121e1.htm
https://www.cdc.gov/mmwr/volumes/71/wr/pdfs/mm7121e1-H.pdf

 

[Godot_]

This might be a close approximation.
https://www.cdc.gov/mmwr/volumes/71/wr/mm7121e1.htm
https://www.cdc.gov/mmwr/volumes/71/wr/pdfs/mm7121e1-H.pdf

...ummm, nope. Really Interesting papers, but it's all about identification of risk factors for long CoViD, not about prevalence.

Re. actual prevalence, I found a meta-analysis from April 2022:
https://pubmed.ncbi.nlm.nih.gov/35429399/ (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9047189/pdf/jiac136.pdf)
which, due to its meta-ness is based on 2021-published data, and hence not the freshest. But still, its sources guide you to studies like Hirschtick et al. 2021, which estimate the general population prevalence based on the prevalence of long CoViD in post-CoViD patients.

The Long-CoViD-prevalence in the US population deferred by the Chen et al. metaanalysis form Hirschtick et al. is 0.47 [0.41; 0.54] for male, 0.56 [0.52; 0.66] for female citizens - in 2021. That'd mean roughly half a percent of roughly 330 million, or a million and two thirds identified cases. In December 2021, publication date - data collection ended Oct 2021. Hence before many of the patients from the big Delta wave even got to the necessary time point to potentially qualify for Long CoViD. Doh... ...dis be tricky.Also, there are quite a bunch of other issues, most central those of underreporting:

The first issue is, that neither all CoViD cases were registered, nor were all acute CoViD patients followed up. So there's quite the uncertainty... ...underreporting. This quite obviously can make quite the difference. Especially if you try to compare data from a "open-to-all"-System like tha NHS with centralized data collection like the British NHS... ...with a pay-per-visit-system for many with fractioned data keeping like the US.

The second issue is that the individual bias to (not) report is pretty variable, too: Harry the hauler might not at all find the odd way stuff tastes worth paying fifty bucks to visit the doc, as might Charlie the couch potato think of his headache or shortness of breath. Ben the binge drinker might well consider his dizziness a warning sign caused by his hobby, whereas Dianne the drama queen will show up in the practice with every minuscule symptom, and Mary Munchausen-Syndrome will anyway take it all and find some extra...

And, as mentioned above already, this pertains to 2021 data, when even Delta hadn't been "washed out". Omicron apparently doesn't cause that many long CoViD sufferers, but we don't know yet.To give _some_ answer to the question:

I'd put my money on the NHS data, and simply multiply by case numbers and underreporting, then use the different population sizes as basis. So, guesstimating the US figure from those estimated 2M ppl in the UK's 22.5M cases* in a 60.8M** population, with an excess mortality based vs. reported factor for CoViD deaths of 1.4-ish would give me a "CoViD-prevalence-estimate-coefficient" of...

22.5M * 1.4 / 60.8M = 0,518

...meaning 52% of the UK population having seen CoViD (let's ignore multiple infections for simplicity's sake - it's a guesstimation). So the Long-CoViD risk on a per case-basis would be 2M estimated Long-CoViD cases divided by 31.5M "true" Cases, or .0635 - 6.35% of (corrected) cases. So that 0.0635 would be the "Long-CoViD-factor"

...with the USA's 84M cases* in a 330M** population with the death underreporting being estimated as 1.6-ish...

84M * 1.57 / 330M = 0.401 ("true" CoViD case rate per US citizen, which sounds about right)

84M * 1.57 * 0.0635 = 8.5M

...or: I'd slipshoddily guesstimate there probably are 8.5 million Long-CoViD-sufferers in the USA.Feel free to add, argue, debunk, debate my train of thought. I'm happy to learn! This should in no way be perceived as scientifically sound, just as a moderately informed approximate number using some sound data sources.
TL;DR:
Millions for sure. How many? Not so sure. Way more than 5, def.* from the JHU CoViD Dashboard, 6/1/22 (date of typing)
** from Wikipedia, 6/1/22 (date of typing)

 

[Oldman too]

I'm happy to learn!

Thanks for the info, very interesting.