> They could predict with a probability of around 90 percent that individual patients would later become seriously ill or die from the Covid-19 infection. The prediction of a less severe course was somewhat less precise. That is, in individual cases, a predicted severe course did not become evident.
I wonder if this would correlate with date of infection vs date of testing. As in, a high expression of hFwe-Lose in the first N days after infection is a reliable indicator, but if the testing is done at a later stage, the baseline hFwe-Lose expression is higher in all patients and therefore no longer an adequate predictor. I wonder if there is a second biomarker that can provide an indication how long the infection has been progressing, and if they could correlate those two biomarkers?
The linked paper (https://news.ycombinator.com/item?id=29263350) doesn't seem to mention this at all, perhaps because it's impossible to determine. Looking at the sensitivity-vs-selectivity charts, it may simply be that this biomarker by itself casts too wide a net.
Last night I was thinking about why some people are fine and others are not. I had a random thought that is probably nonsense but here it is: when I get a cold I noticed that if I sleep on my back the mucous all drips down my throat and I end up with a chest cold too. Once I determined that it could be related I never sleep on my back if I am sick. I have literally not had a cold move to my chest since. I started wondering if anyone had done studies on sleep position with sickness and maybe Covid to see if there is any correlation.
Yes, but for very different reasons. Admitted patients were frequently at the point of difficulty breathing. When lying on your back, gravity makes your lung capacity smaller, when on your stomach, gravity makes it larger.
The increased survival was simply due to getting more air at a critical time.
That would make sense. I wonder if, laying on the stomach is actually freeing the phrenic nerve, allowing the diaphragm to work better, and therefore providing more oxygen.
Oh, there are many reasons. This is what makes medicine interesting, I guess. Think of the Scott Alexander article from yesterday: Ivermectin works pretty well... because it rids you of parasites which may act when you have severe Covid and are put on immunosupressors.
For some reason the virus also plays havoc with your blood sugar, so having even slight diabetes is a huge risk.
I was talking with a Covid nurse a couple of weeks ago (female, small thing) and she said a big part of the reason she's changing jobs is... the size of the patients she has to move around. Obesity is another huge risk factor - whether through diabetes, independent, or both.
I’m prone to severe micro clotting. (Factor 5)
I’m so sensitive that I have visual changes and one sided numbness 24-48 after going off blood thinners.
I had my dosage and timing down exactly to the minimum of what I needed to not have symptoms.
When I got covid I found I had to double up blood thinners for about six months to keep symptoms under control.
The second time I got covid I didn’t have to adjust anything.
So my option is tiny clots is a major part of covid, and some people are more sensitive then others.
Interesting that the biomarker does not interact specifically with SarsCov2, in fact it is a general stress biomarker. Therefore , even though this will be used to prognose higher-risk-patients, it's not clear if one could use it to treat covid by manipulating the marker.
Yeah this medication is usually not available in large quantities, so you can't give it to everyone. So any type of marker is going to be very helpful to you because you can target it to those whose outcomes you can improve the most with it.
You can treat it with monoclonal antibodies (or even that and remdesivir, if you are willing to hospitalize the person right away, before they develop inflammatory symptoms), so I'm not sure why you would even want to pursue something like that even if you could.
As a relative layperson here, this feels retrospectively obvious, no?
Per my shallow understanding (I only read the introduction to the paper), the marker hFwe-Lose is expressed by cells that are struggling, thus close to death. It's the precursor for apoptsis.
It makes sense to me that COVID affecting lung tissue would cause more apoptsis, thus more hFwe-Lose pushed out either through breath or mucus.
Is the 'news' here the discovery of the marker hFwe-Lose?
Per the paper (https://www.embopress.org/doi/full/10.15252/emmm.202013714) the importance of this finding appears to be the discovery that hFwe-Lose concentration has greater predictive power for patient risk stratification than other methods (which notably include analysis of other biomarkers in addition to things like patient age), at least for patients who will become seriously ill.
In other words, working out who is likely to become seriously ill early on is important for physicians to provide the correct level of care; they have a number of tools to perform this asssessment already; the news is that hFwe-Lose concentration seems to be better than the rest at predicting severe outcomes.
Personally I don't think it's blindingly obvious that biomarker levels early in the course of a disease should be so very highly predictive of patient outcomes late in the course of a disease, but I am also a layperson.
There are a lot of biomarkers expressed by “struggling” cells, most of which will not correlate to disease severity, so yes, the news is that this one in particular is predictive.
It appears this provides a way to truly distinguish who should be protected from SARS-Cov-2 infection and who shouldn’t, given a cheap test can be manufactured.
The problem then is, that we and the individuals affected know which group they are belonging to. This makes a very alienable experience. Imagine being one of these people, virtually every other person is radioactive to you.
What would be the societal and political implications of this? Lockdowns only for people who have this marker? I can imagine people wouldn’t want to make this test as it makes them turn paranoid, could we force people to make this test?
“ Lockdowns only for people who have this marker? I can imagine people wouldn’t want to make this test as it makes them turn paranoid, could we force people to make this test?”
I disagree here. People would want to know if they are at risk. For example in our local school system there has been 40 teachers who have tested positive and 230 students since school started this fall (2600 students in the district). Out of the confirmed positive cases we have had 0 severe cases. 0 hospitalizations with an average illness of around 3-5 days.
The amount of school activities missed due to contact tracing, exposure protocols has been immense. The constant stress of a scary pandemic weighs heavily on the community. Imagine if all students, staff and community members could be screened to see if they were vulnerable. Precautions could be made for this select group.
The overall health of the community and school system would improve drastically utilizing another tool to put the scope and danger of this pandemic in perspective.
Has been more or less spread across population. If only a known subset is affected, those will become disadvantaged and possibly even discriminated during employment.
> virtually every other person is radioactive to you.
While you're not wrong, this behaviour is already present - just randomly distributed. Some/many people withdraw, interact with fear, maybe eventually get fed up and take risks, based on guessing-games.
If this research proves fruitful, it will reduce the guesswork a lot, and people can try to find more practical and constructive approaches.
> this behaviour is already present - just randomly distributed
And this is exactly the reason why governments didn’t select for any factor in the first lockdowns and why the discussion about whether it is fair to exclude group X has merit - because while the effectiveness of such measures is high, the efficiency is not.
This would change the situation drastically. A government just tests the whole population on that marker and knows with a very high degree of confidence who is vulnerable.
This doesn't appear to be a permanent (DNA) biomarker, so pre-infection screening may not be effective. Depending on how quickly this marker can fluctuate, a full-population test might not even indicate who will be vulnerable three months from testing.
If only certain identifiable individuals are at risk of severe covid, I expect there would be no lockdowns, and that those individuals would be given the warning to lock themselves down as they see fit, for their own protection.
This is like observing structural damage to a bridge after an explosion. If there's a lot of structural damage, the bridge is likely to collapse later on.
If there happens to be structural damage before the explosion, sure, you can probably make similar inferences about danger after the explosion.
But the absence of structural damage before the explosion is not great evidence that it'll look like that afterwards. This is a biomarker of damaged lungs. Covid can be the thing that causes the damage on a clean slate.
edit: so to clarify, you can't test this to confidently state that someone will be fine with covid prior to infection. you might be able to identify some people that will definitely struggle but its likely that this information wouldn't be too surprising to them.
I'm saying the absence of it is unlikely to be a reliable indicator of poor outcomes. The presence of it may be useful. My point is that this is not some general trait that some people have some people don't. This is a cell marker that tells the larger system that the cell should be killed. The cells only start to express this when they're doing poorly.
Many people start to do poorly only after they get the disease. I would also reckon that known comorbidities are likely to be a more reliable, significantly easier to measure predictive feature for estimating risk before they get covid.
It's not like this is a new thing. Those of us at high risk for COVID because of other conditions have been living in that reality for a while now. More so before the vaccine was available.
> What would be the societal and political implications of this? Lockdowns only for people who have this marker?
I don't think this would be necessary. Vulnerable people can decide for themselves what precautions they wish to take. Lockdowns are to force _others_ to take precautions to protect the vulnerable.
My gut reaction to general lockdowns, early on in this pandemic, was that they were driven more by an inviolable ethic of egalitarianism than by a rational assessment of risk, especially since the most dangerous comorbities seemed to correlate with demographics that, historically or presently, suffer from prejudice.
Getting vaccinated doesn't guarantee you will not have any complications. It increases the chances that a person will have zero or only mild symptoms, depending on which strain encountered, yet hospitalization and/or death are not completely out of the question i.e. to think you are invincible because you have received one of the various available vaccinations (each of which has varying degrees of effectiveness against various strains), is probably not a good strategy.
No, but it drops the risk to levels comparable with many other activities that normal people take on a regular basis. Which is really the best that you can ask for in life.
If you are old, most diseases and activities are very dangerous to you. Your chance of dying in a car accident are highest of any age group, including 16-17yo new drivers.
Just being 80 gives you a 10% annual chance of death, which climbs dramatically as you get towards 90.
Your risk tolerance automatically goes up as you age, even if you don't notice.
You're right, the vaccines don't have 100% efficacy. However, they are the only way out of the pandemic (so that we can safely transition to an endemic state, much like the flu--still annoying and sometimes lethal, but not mandating the current constraints).
There are quite a few people who aren't getting vaccinated because they have doubts AND because they feel healthy and do not think they are at risk of a severe course of COVID-19 infection.
If their physician could tell them with confidence that the virus is very likely dangerous to them, that would maybe push them to getting vaccinated despite their doubts.
This stuff is hard. There are people that simply won't be convinced, no matter what. But anything to help sway the more reasonable "skeptics" out there is a win, I think.
I wonder what the exit condition for the pandemic mode is, or even how we define "out of the pandemic". Vaccines have delivered a great success in blunting the sharp edge. Compare cases vs. deaths. I picked 4 countries with varied vaccination levels: Singapore (>85%), Austria (>60%), Israel (>60%) and Romania (>35%). Case spikes occur regardless of vaccination levels, death spikes are near extinct in areas with high vaccination. Yet there are still broad restrictions: mask mandates, social distancing, wfh, vaccine passports, partial lockdowns.
Caveat: there is a lot more to this data than meets the eye, but public health policies appear to be driven by broad numbers, so for the purpose of the "how does the pandemic mode end" conversation these curves are a reasonable starting point.
There is also a non-zero chance of getting hit by a bus if you leave your house. Might as well not go outside.
Forgive the joke, but this feels like a glass-half-empty take on something we're not fully understanding. I think we should look at numbers before determining whether or not the vaccine is a good strategy. Life involves risks.
I think if the average human being applied the precautionary principle to everything in their life as deeply as politicians have applied it during the pandemic, their life would be utterly miserable.
Granted the data about vaccines efficacy was last updated end of July 2021, yet according to those numbers, I'm fairly sure the chances of getting hit by a bus are considerably lower. Also you can more easily avoid getting hit by a bus by looking both ways before your cross the street (freak bus accidents aside), whereas avoiding an invisible virus is slightly more difficult.
Thanks for the link. Getting hit by a bus is a bad example for sure. I just wanted to paint a picture about there being risks in life.
According to "usafacts.org" [1] there's a (very) roughly 1.6% rate of death, not controlling for age. The informationisbeautiful data shows that I'm >90% protected from extreme cases with 2 doses. That feels pretty safe to me given how low the rate of an extreme case is to begin with. (I'll admit I only measured death though)
You're right, though, it's not invincibility. That's a pretty high bar for anything.
That condition seems to be fairly widespread right now already.
It's even encouraged under the moniker of "social distancing" (which itself is a telling misnomer; "physical distancing" would've been a much more appropriate and less jarring term but somehow that never stuck), even though there have been better alternatives (getting vaccinated, getting tested when meeting with a larger group of people, wearing high-quality masks) for quite some time now.
On the contrary - looks like a reasonable reaction to me. You know the risk for you is very high, so you protect yourself better against the infection. The best tool we have is avoiding the infection by increasing physical distance from the aerosols in the air exhaled by others.
If you're vaccinated and believe in vaccines, it's safer than ever in this pandemic to go out, especially if you avoid indoor areas. I am concerned about some people who haven't really left their homes since March 2020. I know some of such people, and they are really paranoid. They are vaccinated, but even if they are outdoors, they say that "everyone around is spitting on them". It's very unhealthy and the more they stay in isolation, the harder it is for them to convince themselves it's relatively safe to go outside.
On the other hand, there are people taking lots of unnecessary risks for no ones benefit. As a typical example, in the 12+ years I've been working at our research institute, nobody has ever organized an institute-wide joint lunch and nobody ever expressed any interest in it. On the contrary, many colleagues hate these kinds of events. But now that everyone is vaccinated (or at least assumed to be, nobody checks it), they have organized a 4 hour indoor meeting plus a joint lunch that you can hardly avoid. Nevermind perhaps waiting another year for organizing this. It has to be now to bolster team spirit, etc.
The people who do this think they're geniuses, and literally say things like "we should push people to go to international conference again instead of giving Zoom talks", and so on.
In reality, everybody just projects their own preferences and fears onto others. The people pushing for hitherto unknown social events are extraverts, the people who prefer to stay home and give Zoom talks are introverts.
The bottomline is that there is nothing wrong with being an introvert and avoiding large crowds is quite reasonable even when you're fully vaccinated. (Less so if you recently got a booster shot, but e.g. in our country it will take a while until every adult has received one.) Infection numbers in our country are going up rapidly, with an R-value close to 2. Then, close to Christmas when measures have to be announced, everybody will act surprised again, of course.
I expected that a test for the risk of suffering a severe case would be done after a positive Covid test, not before. That is to decide the further treatment and not to make people take more (or less!) measures of precaution.
The demand of negatives to live their live without what is perceived as unacceptable measures (remember when asking people to wear masks was too much asked?) will force politics to draw this tool out of the toolbox.
They used a type of PCR to analyze nasal swabs for their prediction of severity test, so it's not as easy as doing a lateral flow test yet (that's the type of test we use as rapid Covid tests). Maybe it is not much more difficult or expensive than doing the PCR tests that we conduct for Covid though? There seem to be different types of PCR and I know little about that, so take that with a grain of salt. Maybe someone knowledgable can chime in?
Yeah, maybe then testing for this marker could be done concurrently with the PCR test for Covid, and the tested person can receive a warning that they are likely to get a severe case - or the presence of this biomarker may even be used to give them priority for treatment with monoclonal antibodies or other treatment methods? The next question is of course: how do vaccines get into play? AFAIK they protect you from getting a severe case most of the time, even if you have this marker/lung damage?
Anyway, that's one way having this marker might turn out to be useful - otherwise it only confirms what was already known, that patients with preexisting lung damage are more likely to have a severe case. Of course, patients or their doctors may not be aware of the lung damage or underestimate it, while the marker is more objective...
Intuitively, it dosent seems to be a good idea. We should hear what someone with strong Bayesian statistics knowledge think about this. But running tests on many people to do individual prognostic is usually a terrible idea.
Not knowledgeable but I do know there was a grad student (post doc?) who created their own PCR test for SARS-Cov-2 in their living room using frozen peas as their cooling medium. Obviously this guy was a specialist with access to the right materials but it goes to show that a lot is possible.
also a great example of why commercializing research is so difficult and expensive - to get something from looking like it's barely holding together with duct tape to being a sellable product is no small feat.
Yey, yet another reason to help people make informed decisions.
Ignoring policy for the moment.
It would be cool to know in advance if say whooping cough (for example) would hit me hard or not. If the price of detection could be driven down it would be better than having to store or produce expensive vaccines or treatments for people who won't significantly benefit from either.
Hope that tech like this actually takes off :)
Is there a service where you can have relevant parts of your genome sequenced without running the risk of the data being retained and shared against your will?
Depends on what you are asking about. If you are asking about how I would technically go about looking up things: I have no idea. I don't know what the software space looks like around this. Anyone?
If you are asking about the medical context, at least for me in particular, I have a close relationship with the local hospital and access to people who have the knowledge which would make this data useful. (In fact, the same people who advised me to be very careful about sending off my DNA overseas - advice I'm very glad I took)
I wonder if this would correlate with date of infection vs date of testing. As in, a high expression of hFwe-Lose in the first N days after infection is a reliable indicator, but if the testing is done at a later stage, the baseline hFwe-Lose expression is higher in all patients and therefore no longer an adequate predictor. I wonder if there is a second biomarker that can provide an indication how long the infection has been progressing, and if they could correlate those two biomarkers?
The linked paper (https://news.ycombinator.com/item?id=29263350) doesn't seem to mention this at all, perhaps because it's impossible to determine. Looking at the sensitivity-vs-selectivity charts, it may simply be that this biomarker by itself casts too wide a net.