72 Comments
Dec 15, 2022Liked by Katelyn Jetelina

This particular piece of writing needs to be an op-ed in major new outlets!

Every time I read one of your entries, I am reminded of how gifted you are at capturing the nuance in risk assessment that is so important for moving humanity forward with COVID's presence in the world. It's not binary--there is way more gray than black and white.

You are a gifted communicator, with an ability to convey complexity and--with a topic as polarized as covid responses--evaluate subjectively without judging others. I wish more people had access to and took the time to read your stuff, especially stuff like this.

Expand full comment

Agreed! A very helpful post.

The concept of excess mortality parallels the underestimation of long Covid risks.

For example, according to models published in The Economist:

“Globally, the model estimates that the total number of excess deaths is two to four times higher than the reported number of confirmed deaths due to COVID-19”

https://ourworldindata.org/excess-mortality-covid

In other words the total death count from Covid is probably 2-4 times what is reported globally. I think that is part difficulty tracking, and part increased risk of death after Covid, long haul “symptoms” or not.

Encouragingly, if you look at some of the graphs of excess mortality in 2022, most countries like UK and Germany are still hovering above expected, but not egregiously so like the massive spikes pre-vaccine era 2020.

https://ourworldindata.org/grapher/excess-mortality-p-scores-projected-baseline-by-age?country=DEU~GBR

I would do a post on this but it’s late, and you do a better job!

Also, I agree about that study you mentioned and that was widely reported on this week about “long Covid” as a cause of death listed on a few thousand death certificates not being a valid indicator. I would hesitate to even list “long Covid” on a death certificate, because the cause and effect would be really hard to determine for someone dying of a heart attack for example. I honestly didn’t even know Long Covid was something we could list...

Expand full comment
founding

"The individual risk of death from a COVID-19 infection is now close to the flu thanks to vaccines, immunity, and treatment.” Hey Katelyn, that’s not quite right in my opinion. The last numbers I saw from the CDC were 468 deaths a day or about 170,000 a year. Even if you look at the roughly 300+ it was averaging before the recent bump up, it would be near 110,000 a year. The CDC site showed the average deaths per year from flu in the 10 years prior to COVID, was about 30,000 with the worst of those years being 51,000. Don’t hold me precisely to those numbers but I am pretty sure I have them about right.By my arithmetic that means the “current” (468) COVID death rate is more than triple the risk of flu in its worst year and five and a half times the average flu rate. Using the lower number of 300 for COVID, of the last month or two, then the comparable relationships are over two, and about three and a half, times the numbers of flu deaths. I think that’s a lot more risky - not “close to". I make such a point of this because (1)for a long time I have been using that relationship as a benchmark for when I’d give up my extreme precautions and try to live a “normal” life again. I figure I could handle two times. I am older and thus the the death rates are doubtless higher in my bracket for both diseases, but, if the numbers are both understated than the relationship probably roughly holds and (2) I have seen/heard your comment or ones similar to it often enough to be bothered by it, since I think it makes light of what I consider to be justifiable cause for my (“paranoid") relatively extreme precautions.

PS Notwithstanding the above difference in opinion - one person’s “close to” can be another’s “much more” - I continue to respect, admire and appreciate the great work you are doing in communicating/educating us all.

Expand full comment

David very good points, I would point out though that we have never searched for the flu like we did Covid. Previously we used high level estimates to take educated guesses on the number of people who died from the flu. We are doing more PCR tests searching for Covid per day than an entire year of Flu PCR tests.

When Patricia Dowd died in February of 2020 becoming one of the first deaths ascribed to Covid, she died of a heart attack. It was only after she passed away they ran a PCR test to discover she was positive for Covid.

This would have never happened prior to Covid. We had never done autopsies on people who died of heart attacks to run PCR tests in search of a viral infection to reclassify the death.

This is a massive change in how we count deaths so I don't think we can accurately compare our current method of diagnostic scrutiny to the previous method of using models to estimate.

Expand full comment
founding

As I understand you, the point is that flu deaths were poorly registered and were probably understated and that COVID deaths are much more accurately reported due to increased scrutiny. With all due respect, given the frequent comments about the likelihood of COVID deaths being significantly undercounted as measured by the “excess deaths” methodology, I don’t think we can assume that COVID figures are much more accurate than flu estimates. I don’t think the question is resolvable given the passage of time and evolution of detection processes that you note, but my fallback is the “official” figures reported by the CDC. Those figures strongly support my point that COVID has been and IS much more dangerous than flu. If we reach my “comfort point” of two times flu deaths I will be a very happy man. We got there in the summer of 2021, roughly, and I hope that new, more dangerous, more transmissible variants won’t keep us from getting there again. If mask mandating (or at least social disdain for those who don’t consider the safety of others, or worse, have contempt for it) in some critical /indispensable locations - like grocery stores - were put in place, I think it would be a very achievable goal. As long as people keep repeating that COVID (not long COVID about which we know so little) is no worse than the flu, than people like me who want to be safer will be subject to scorn and/or pity as misguided at best and looney at worst. And selfish unthinking behavior will continue to be accepted as perfectly okay.

Expand full comment

We actually got back to near zero excess deaths in the US this past spring/early summer in 2022. Latest CDC death report just dropped and still going through it, but still likes we slid back to trend ~6% more deaths than expected (when accidental deaths which lag the longest come in may push up another %).

To my original point, I don't know if Flu deaths were overcounted, undercounted, or what the precise way to count them should be. I am saying that how we count Covid and how we previously estimated ILI burden are drastically different so reconciling them is going to be a challenge, only trying to be helpful in suggesting caution comparing them if that is your benchmark for how comfortable to be.

Expand full comment

This is the MOST HELPFUL piece about the long Covid risk that I have seen. I have a pre-existing condition that predisposes me to long Covid (ME/CFS), so I have been extra cautious. Last night I dreamt that I had accidentally pretty much given up mask wearing. While I will not do that, this research will likely help me feel more comfortable taking more risks, such as traveling. I am so grateful to you, Katelyn, for putting together this information (and all the previous Covid pieces as well) for all of us.

Expand full comment

Your Long COVID article is the first clear presentation of the just-now-becoming-available studies. Your calculations comparing Long COVID risk to driving is especially apt. Most people have a clear understanding of the risks of driving. That the chance of having Long COVID is about twice that of having a very serious auto accident is extremely useful information

Expand full comment

Thanks so much for this! Do we know if the risk is the same with every subsequent infection? Or does it change with reinfection?

Expand full comment
author

Ahh I knew I was forgetting something in this post. Good question. We don’t know. The only thing we know is a reinfection is more risky than no infection, but if the risk changes with each subsequent reinfection among someone is unknown

Expand full comment

Most of what I've been reading of late (Disclosure: I'm pretty sure I've someone in the household with PASC/Long COVID, so it's a topic on my reading notifications) suggests that subsequent infection increases the risk by as much as 3x, as do certain comorbidities, including both the usual suspects, as well as autoimmune manifestations.

Expand full comment
Dec 15, 2022·edited Dec 15, 2022

Is this from the VA study? I’m not anywhere close to an expert but it doesn’t seem to me those results are universally applicable since the population was older and in worse health? Editing to add Katelyn's breakdown of that study - https://yourlocalepidemiologist.substack.com/p/epidemiology-of-reinfections

Expand full comment

Not solely from the VA study. That said, the VA study was so large as to make its flaws and weaknesses something to consider in evaluation but not a study to discount.

Expand full comment

Thsnk you for this update. Is there any evidence on the effect of chronic anticoagulation on the acute disease or long Covid-19? So many older folks take Eliquis or Warfarin for atrial fibrillation or other conditions.

Expand full comment

Great question. I think the risk gets higher, but would love to see Katelyn's analysis of the data.

Expand full comment

Thank you so much for this! For Hubby and I, long COVID has always been our primary concern. Are we any closer to figuring out what factors can predict long COVID?

Expand full comment

Ok. Two questions, both from my son the data scientist who is "updating his priors"...

(As a young-ish adult with a friend group of pro-vax/booster youngsters, none of whom even really suffered much when they came down with Covid, he was at first very skeptical about even the existence of long covid (thought maybe psychosomatic), and we've had risk-assessment discussions where he's expressed concern that I'm over-weighting long covid risk compared to risks associated with isolation.)

(1) Where does the 18% come from when you say "If we take into account 3% of infections lead to long Covid and, of those, ~18% will have disease so severe that they are unable to work. So, the annual risk of severe long Covid (unable to work) is 1 in 370."

(2) Also the 3% number seems to be coming from the NHS data, which is "symptoms longer than 2 weeks" which may feel like "Long Covid" for HCW but I think most laypeople are thinking more like "more than 4 weeks" or having sequelae that bring on an entirely new disease (I have a friend who suddenly has diabetes, otherwise fit and average BMI).

Can you verify / explain?

If this gives you a headache, please skip it, he'll update his priors some other way. :D

#momAlliance

#waitUntilYourKidsHaveAPhD

Expand full comment
Dec 17, 2022·edited Dec 17, 2022

If 2M people have Long Covid to such an extent they can't work, that would be 1/150 people. If we estimate that these 2M got Long Covid over the last two years, then the estimate of 1/300 annual chance of getting Long Covid seems plausible from that measure.

However, I think most studies show that the second infection is less severe than the first, so it doesn't seem likely we'll continue seeing severe Long Covid at the rate of 1/300 indefinitely.

Expand full comment

Thanks. Where is the source for 3% * 18% of covid sufferers having severe long covid?

I think it’s the NHS data set.

If that’s the case I’m hoping for understanding about 2 weeks of symptoms.

Expand full comment

The data you cite on the comparative risks of death from flu or COVID is from the UK. The US is running at about 150,000 deaths a year from COVID, probably about 3 times the annual deaths from the flu. Thus, your elaborating on the risk of death from the two causes in the US would be most welcome.

Expand full comment

Does it make sense to breakdown odds of long covid based on age and male/female (and possibly underlying health conditions) to help people more fully understand their individual risk assessment?

Expand full comment

Yale recently published a report on case studies showing significant improvement in long covid brain fog by using two FDA approved medications. Worth a read and future report. https://medicine.yale.edu/news-article/potential-new-treatment-for-brain-fog-in-long-covid-patients/?utm_source=YaleToday&utm_medium=Email&utm_campaign=YT_YaleToday-Public_12-15-2022

Expand full comment

@Janet Billups: Thank you for sharing a very interesting study worthy of a large scale comparison with other candidates and placebos. In our area there have been successes treating both brain fog and 2 patients with progressive neurodegenerative disorders with suspected PASC origins using a brief course of HU and carefully monitored titrated dosing with pyridostigmine or ravastigmine. The basis for using these 2 drugs lies in the presumptive targets of the a7NACh receptors and mitochondria whose functionalities may have been compromised by an antibody derived from the antigenic component of the spike protein of the SARS-Cov2 virus. A return of function in one PASC case enabled a Dentist to return to practice (Rx HU 3 Days then Rx Exelon) and the other a stabilizing and partial (to date)reversal of an ALS-like neurodegenerative process when he had been given a prognosis of <6 mos to live in recent months by 2 esteemed Neurologists in 2 countries. Clearly there are safe and readily available repurposed drugs “standing by” ready for trials by scientific literate advocates who understand the serious barriers and possible biases that only new (enormously expensive) Pharma-derived drugs under EUA status are candidates for acute COVID19 or PASC treatments. It is our opinion that, sadly, ROI, economic incentives have perverted the focus in getting efficacious and affordable treatments for this novel virus. It is our contention that the ultimate targets go much deeper than the infectious and immune cascades of immune processes or individual organs. It’s the cellular biological targets that need to be scrutinized and addressed as well as immunomodulatory responses that need more of a focus. Hydroxyurea, unfortunately, will forever be associated with that highly politicized drug hydroxychloroquine. Hopefully the guanfacine, a selective alpha-adrenergic receptor agonist would not have significant interactions with other drugs with CYP3A4 or vasodilatory sensitivities for these PASC patients nor would it be confused with guaifenesin.

Expand full comment

Cancer patient here. It's nearly 4 months since my bivalent booster. I'm wondering about updated boosters for the elderly and immunocompromised. My oncologist hasn't heard anything yet.

Expand full comment

Another wonderful piece, Dr. Jetelina. However, I think without broad agreement and standards on how to make a diagnosis of long COVID, the data (and risks) will remain suspect.

Expand full comment

One thing not mentioned here, but which is apparently another factor among long Covid suffering is the number of people who commit suicide . Apparently some people experience excruciating pain from neuropathy that does not seem to be remediated by any current treatments. There may be other factors as well . But this indicates that there may be much more to some cases of long Covid than just feeling tired and sick.

Expand full comment
author

Yes, I actually considered including that, but could find no data on it. I was disappointed that it was excluded from the mortality study. I’ll keep looking, but I know this is very important to keep in mind

Expand full comment

I heard about it in an interview on NPR with two women who are involved in a self-help group for people with long Covid. It was anecdotal. But they both knew of people who had committed suicide because of the pain that they were dealing with.

Expand full comment

I've heard this anecdotally, but have not found quantifiable evidence yet.

Expand full comment
Dec 15, 2022·edited Dec 15, 2022

Chronic pain is as an important risk factor in suicidality from chronic illness. But so is perceived burdensomeness (how much of a burden on others patients perceive themselves as being) and "critical expressed emotion" ("critical and hostile interactions" patients have with their support system, like loved ones accusing the patient of malingering).

https://www.psychologytoday.com/us/blog/chronically-me/202109/suicide-and-chronic-illness

For example, for patients with the connective-tissue disorder EDS, suicidality is fairly normal. Some people with EDS do not find it painful. For others, EDS pain is excruciating. For still others, the problem with EDS is low-level malaise that grinds you down but that just doesn't seem real to other people. Many people with EDS go decades if not their whole lifetime not knowing they have it, finding life vaguely more painful and difficult than average, bad enough to tank your performance as a supposedly "healthy" human being, but not bad enough to attract intense medical attention. It's not so hard to see how people in this predicament might conclude they just don't deserve to live.

Withering skepticism directed toward Long COVID could do the same thing for Long COVID sufferers.

Depression itself may just be sickness behavior without a clearly-identified cause, either because there is no physical cause or the physical cause got overlooked. (I was persuaded to spend years of life from my teens on interpreting pain from EDS as "just depression".)

Expand full comment

I wanted to question one statement in your post:

"Of course, the more the virus mutates to become more contagious, the risk of infection (and thus long Covid) increases."

If someone has already had Covid without Long Covid, what are the odds that a future infection will trigger Long Covid? Given that most humans have already had Covid once, and that you mentioned that Omicron is less likely to lead to Long Covid than prior variants, I don't see that it necessarily follows that the risk of Long Covid increases with new virus mutations.

Also it was hard to tell, in the many studies you reported, how many were from unvaccinated populations. e.g. The Lancet article you mention showing Covid leading to a 3x risk of dying was entirely describing infections during the first year of Covid, almost entirely before vaccines and effective treatments. For people becoming infected post-vaccination these risks may be much lowered.

Expand full comment

It should probably be stressed that a key difference between driving risk and covid risk is that there's a sophisticated, well regulated risk management infrastructure - car insurance - that helps to guarantee that you'll have adequate medical care, should you sustain an injury in a car accident. There are even controls in place that help make sure that those who engage in driving behaviors that increase risk to others pay proportionally more for the anticipated harm they might cause to others. For instance, if you get a speeding ticket or if you run a red light, your rates could go up. Insurance companies share info about you, by way of a database run by LexisNexis called CLUE (Comprehensive Loss Underwriting Exchange). Until something of this sort exists for URTI's - which can be thought of as "injuries" resulting from "collisions" between people - driving will always be "safer" than risking getting covid.

Expand full comment
founding

Are you sure about the dog bite numbers? One in 400 requiring reconstructive surgery seems quite high.

Expand full comment

The study she links to says there are 27k reconstructive surgeries due to dog bites needed a year. So that's 1/11,111.

"2018, almost 27,000 individuals required reconstructive surgery owing to dog bites (American Society of Plastic Surgeons 2018)"

That's the number that should be compared with long covid. Hope the Dr. makes a correction.

Expand full comment

I was confused whether “requires reconstructive surgery” was for the entire population or only for those who had already suffered a dog bite. My take is it’s the latter, ie conditional based on preexisting dog bite, and therefore not an apples to apples comparison of risk assessment for the general population.

Expand full comment

I'm sure it's 1 in 400 of reported dog bites, which would have to be semi-serious in the first place to be reported. Could be tens of thousands of minor bites that don't get reported.

Expand full comment

I agree but only in principle since I’m not a plastic surgeon. In the past there was controversy about suturing vs not suturing a fresh dog bite. Like a lot of acute events timing enters the decision making as well as bite location, the patients risk of a serious infection (diabetic, etc). Below is a pretty decent study discussing the pros and cons as well as outcomes for each approach. Sometimes it’s a judgement call esp if its a child, bite on the face, size and depth of the explored wound and such. Any minor closure could be billed as reconstruction by some practitioners I suppose.

https://pubmed.ncbi.nlm.nih.gov/23916901/

Expand full comment