Thursday Throughput: Distant Star Edition

Michael Siegel

Michael Siegel is an astronomer living in Pennsylvania. He blogs at his own site, and has written a novel.

Related Post Roulette

51 Responses

  1. Kazzy says:

    Re: Covid deaths

    I’ve been wondering a lot how useful it is to compare any Covid data between populations have very different demographics. It would seem to me that mitigation obviously makes a huge difference — as noted, vaccine rate seems to have been a real difference maker in 2021 — but it seems like we sometimes get a handwave at demographic data but is anyone actually attempting to account for it?

    My layperson perspective makes me wonder how much A) population density and B) population age impact case counts, case severity, and deaths. Like, I don’t want to make light of any death… but it would seem to me that we shouldn’t necessarily look the same at an 80-year-old with lung cancer dying of Covid as we should a 30-year-old who was otherwise healthy. Like, is there a way to account for “lost years”… figuring out what the general life expectancy was in a population and then how many people died and how short of that they were due to Covid? That’s a very analytical approach to a very sensitive and emotionally wrenching subject, but if we are really going to do an analysis of what did and didn’t make a difference with regards to Covid outcomes, it seems like we have to consider all possible factors.Report

    • pillsy in reply to Kazzy says:

      Like, is there a way to account for “lost years”… figuring out what the general life expectancy was in a population and then how many people died and how short of that they were due to Covid?

      There is a way. There are several, in fact, and they’re quite commonly used in health economics. It’s an especially useful set of tools if you want to figure out how to spend money in your healthcare system, because you have limited amount of money to spend, and all else being equal, you want to spend it where it will do the most good.

      Probably the most common way of measuring this is the QALY [1], or “Quality Adjusted Life Year”, where you add up the number of years people you expect people live, and adjust downward by a factor (the eponymous adjustment) that reflects how much suckier their life is due to being sick.

      So an otherwise healthy 30 year old can be expected to live about 50 years, and most of those years will have little downward adjustment for quality of life. 80 year olds generally have not-so-great health to begin with, and will die in the next couple of years.

      So when the 80 year with lung cancer dies, you lose a lot fewer QALYs than when an otherwise healthy 30 year old does.

      Taking this into account is pretty routine when (among other things) designing vaccination programs, because it can indicate whether the incremental benefit of (say) a third dose of Mening B vaccine is worth the cost of buying and administering it. You divide the QALY gain by the incremental cost and you get, effectively, the unit cost of the QALYs you’re buying.

      In the US, spending less than $50k/QALY is usually regarded as a good deal, spending more than $150k/QALY is a pretty bad deal, but in practice because of our rather unique and bizarre healthcare system most decisions aren’t made that way. Vaccination programs are something of an exception, as the ACIP [2] often considers the information.

      Now to do any of this, you need to know how much disease the vaccine is going to prevent, and how serious that disease will be, or at least come up with a good estimate. This means knowing how well the vaccine works, knowing how the infectious disease spreads, and knowing what kind of burdens that disease creates. For most infectious diseases, you do this with a more-or-less complicated mathematical model, which accounts for as much information as you have.

      Information you almost certainly have is what the age distribution of the population you’re interested is. Information you may or may not have is how the disease is spread by contact, how people’s behavior determines who they have contact with, and how many people have natural immunity to the disease. If you don’t have it, you either take your best guess, or take a whole bunch of pretty good guesses to get a range of possibilities.

      Then you run your model, do some pretty simple arithmetic with the results to find the total costs and QALYs, and write your report.

      A lot of this probably won’t be done with COVID yet, because it’s still an emergency which is killing a lot of people, the vaccines are pretty cheap by the standard of things that healthcare systems buy, so it’s almost surely a good deal. If we ever get to a point where we have endemic COVID and we’re trying to figure out booster schedules, or decide whether to buy Pfizer of Moderna, that could change.

      [1] Pronounced “kwallee”.

      [2] Which is the part of the CDC that makes recommendations about vaccination schedules.Report

      • JS in reply to pillsy says:

        IIRC, both Moderna and Pfizer noted that normally they’d have charged around 150 to 200 dollars per dose for an mRNA vaccine, which would recoup their R&D, facilities, and give them a generous profit.

        The US government, which foot the bill for their development costs, testing costs, and facility creations is paying about 20 bucks a dose, which is roughly the cost to produce and ship.

        Worth noting that this is a massively good deal for both companies, as when COVID is over they will have excellent new manufacturing facilities, real world application data that is literally priceless, and a public that has already by and large accepted the new technology.

        It was a good deal for the government because, well, they got a pair of solid vaccines out of it to arrest a massive pandemic, now have a new technology for public health, and the price they paid was effectively about the actual cost of trialing, manufacturing, and distributing the vaccines. They’re out any funds spent on vaccine development that didn’t make the cut, and obviously don’t get to keep the facilities they paid for — but they only paid 20 bucks a dose (roughly) for the vaccine.Report

        • pillsy in reply to JS says:

          Yup.

          Also $100-200/dose pretty common price for novel vaccines, and one which will frequently be highly cost effective for diseases with much lower mortality and morbidity than COVID.Report

      • Kazzy in reply to pillsy says:

        Thanks for all this info.

        Aside from vaccine distribution and the like, is anyone or will anyone apply this type of analysis to broader Covid data?

        Like, saying this group of 100 people had 8 deaths and this group of 100 people had 12 deaths and concluding that the latter’s Covid response was “worse” feels too simplistic. If Group B lost 12 octagenerians and Group A lost 8 30-somethings, that seems pretty important if we’re trying to determine what worked and what didn’t and how we can apply that knowledge going forward… right? Or am I crazy?Report

        • pillsy in reply to Kazzy says:

          Aside from vaccine distribution and the like, is anyone or will anyone apply this type of analysis to broader Covid data?

          I haven’t seen anybody do that yet, and I’m not sure it would be workable given the type of COVID challenges we’ve been dealing with. Similar approaches have been taken to, say, try to predict where in the US you would expect to see COVID cases most overwhelm the healthcare system, but you’re looking at different endpoints here.

          Mostly I focused on cost-effectiveness analyses of vaccination programs not because they’re the most relevant kind of analysis here, and more because I used to do them for a living, and would like to do them again some day.Report

    • DensityDuck in reply to Kazzy says:

      ” it would seem to me that we shouldn’t necessarily look the same at an 80-year-old with lung cancer dying of Covid as we should a 30-year-old who was otherwise healthy.”

      welcome to Sweden, here’s your herfengerfer

      (this was pretty much the attitude Sweden took and they had a much higher death rate among older citizens than anywhere else)Report

  2. Oscar Gordon says:

    ThTh2: I’m not sure how much of an impact it had, but alongside the conservative squawking about vaccine panics was liberals complaining that we shouldn’t get a booster until the rest of the world had 2 shots.Report

    • CJColucci in reply to Oscar Gordon says:

      Which influenced who, when push came to shove?Report

      • Oscar Gordon in reply to CJColucci says:

        That’s the question. If the vaccine scary stuff didn’t turn you off, did the global equality guilt trip?Report

        • fillyjonk in reply to Oscar Gordon says:

          I admit I was slightly influenced by the guilt trip thing, though in the end I decided my forgoing a booster in a part of the country where a plurality of people refused the first shot would not help the Global South much. I would technically be due for shot 4/booster 2 in May; am considering getting it up when visiting my mom because booster 1 laid me flat for a day and it would be nice to have someone to fuss over me and feed me soup if that happens.

          but my feeling is if my fellow citizens are gonna be jerks and refuse to mask and distance during the next wave, well, let me be as protected as I can beReport

    • Kazzy in reply to Oscar Gordon says:

      The equity issue gave me pause throughout. I was eligible early as a teacher but wasn’t in the classroom actively so figured I should wait. Then I read medical ethicists saying folks should take available shots when they’re eligible. So I did. I basically was on a wait list and when an empty seat popped up, I went in.

      The liberal scolds didn’t help but I wonder how impactful they were. I don’t know anyone who didn’t get a shot because of equity concerns. I do know folks who didn’t due to conservative talking points.Report

  3. veronica d says:

    ThTh6 — Yikes! I will assume that isn’t an OSHA approved workplace.

    You know they make tools to do stuff like that.

    The reminds me of a curious fact about welding gloves. My old roommate taught me some basic welding. One of the first things he taught me, besides wear a mask, was to make sure your gloves were loose fitting enough that you could swing your arms sharply and they’d come off. The reason was it wasn’t unheard of to heat up your gloves to the point where they start burning your hands. If this ever happened, you would need to get them off quickly, and fumbling around trying to pull them off was a pretty bad idea.

    I never had to do this, but it was definitely the sort of thing you wouldn’t think of if you didn’t have someone teaching you.Report

  4. Jaybird says:

    I admit to having questions about the second booster at this point. (Not 50 yet… but I will be sooner than I expected to.)

    Like, is the flu shot I got last year the same as the flu shot I got two years ago the same as the flu shot I got three years ago?

    I was under the impression that each flu shot is reformulated yearly.

    If that’s the case, I’ve got questions about getting a booster for alpha, again. I got the alpha shots, I got the alpha booster. Alpha ain’t the problem anymore. Right?Report

    • fillyjonk in reply to Jaybird says:

      I think it’s more like showing your immune system “that bastard” again so they remember better what he looks like and can beat him up if he shows up.

      though I am seeing some reports that crossing boosters (e.g. get Moderna if you had Pfizer before) seems to enhance immunity, so I’d hope they’d start maybe on a more generalized/variant related shotReport

      • Jaybird in reply to fillyjonk says:

        Yeah, but it’s that bastard’s 2nd cousin that I’m now worried about. I am not against getting a reformulated booster. Not at all!

        I’m just wondering at the utility of getting a booster for the old one given the new one.Report

        • Michael Cain in reply to Jaybird says:

          Cousins is a terrible analogy.

          Think of viruses as simple little machines, or pieces of code. Each one’s got to perform three tricks to be successful. (1) They have to trick some subset of cells in your body into letting the virus through the cell membrane. (2) They have to trick some part of the normal cell mechanisms into making copies of the virus. (3) They have to trigger reactions at a higher conceptual level that reliably transport copies of the virus from host to host.

          The whole purpose of a vaccine for viruses is to prime your immune system to recognize and destroy a particular type of little machine before it can accomplish (1) and (2). Viruses “evolve” because of mutations — bad copies — such that (a) your body doesn’t recognize them and (b) the tricks listed above still work.

          The mRNA vaccines teach your body to recognize parts of the virus’s spike protein. The data suggest to me — a mere systems guy, so take it for what it’s worth — that each dose primes your immune system to do a better job at recognition, to recognize smaller sections of the protein. Or at least to refresh its memory. Nothing that I have read has suggested that Omicron or the newer variants have replaced all of the spike protein, only portions of it. Better recognition may matter.

          Those of you with young robust immune systems probably get little benefit from the extra recognition practice. Us oldsters’ systems benefit (statistically, based on antibody titers) from another round of practice.

          Examples for tricks (1), (2), and (3). The viruses that give your cat(s) a cold have a version of trick (1) that doesn’t work in humans. Some of the Covid vaccines use viruses where trick (2) doesn’t work in humans to expose you to Covid proteins. Ebola is pretty bad at trick (3), because it kills its hosts too quickly and pretty much requires fluid-to-fluid contact for transmission.Report

          • Jaybird in reply to Michael Cain says:

            I don’t know what the best analogy would be for “variant”.

            It seems like the variants in the flu are different enough that my doctor has said to me “some years the shot misses” and shrugged.

            I don’t mind getting another booster and going through another weekend-long hangover (not compared to the potential benefit).

            But I do find myself wondering about divergence between the shot and whatever new variant happens to replace BA.2.Report

            • Michael Cain in reply to Jaybird says:

              We are each a giant bag of protein chemistry state machines. Bazillions of them, all interacting. Bazillions of stochastic state machines. Viruses are little state machines that abuse the privileges the giant bag grants to some of its own machines, in effect fitting through the cracks to replicate. Influenza viruses have found their own “strategy” for getting through the cracks. Measles has another one. HIV another one. Don’t whine because we haven’t figured out how Covid fits in the cracks, and how to stop it, in two years.Report

              • pillsy in reply to Michael Cain says:

                Not just protein! There’s considerable important chemistry that happens in RNA on its own.

                Many lifetimes ago RNA was kind of my field so I gotta stick up for it.Report

              • Jaybird in reply to Michael Cain says:

                I’m not whining because we haven’t figured out how Covid fits in the cracks, but I’ve given explanations to people IRL for why they ought to get the shot and get the booster and one of the questions they asked about the booster was “I don’t get polio boosters!” (okay, that’s not really a question) but I explained that they get flu shots every year because the flu shot is reformulated.

                And I’m being told to get a booster shot for alpha despite the fact that we’ve had covid long enough to adopt and then abandon the Greek alphabet for categorizing variants.Report

              • Michael Cain in reply to Jaybird says:

                Tell the people IRL the truth.

                They don’t get polio boosters because (a) they got a four-dose regimen spread out using optimal timing over a few years when they were kids and (b) outside of Afghanistan and Pakistan, there is no polio virus in the wild to be exposed to. (You might point out that if they are adults planning to travel to Afghanistan or Pakistan, a polio booster is strongly recommended.)

                There is overwhelming evidence that there is so much overlap between alpha and omicron Covid variants that being fully vaxed and boosted for alpha provides excellent protection against omicron putting people in the hospital or ICU. There is consistent lab evidence that for people most at risk, a fourth dose of the alpha vaccines increases their immune system response to omicron substantially.Report

              • Jaybird in reply to Michael Cain says:

                Well, BA.2 is now the dominant strain in the US.

                It’s officially a sub-omicron variant so maybe a fourth dose is all I need.

                May they never need to get to pi.Report

              • Kazzy in reply to Jaybird says:

                Wasn’t “Sub-Omicron Variant” a Michael Bay movie?

                I’ve been recommended and/or required to get different boosters as an adult. As a child care provider, I need prove of vaccination/immunity against certain things in order to work. Sometimes that is checked via bloodwork and sometimes my doctor will say, “You don’t NEED this booster — legally or otherwise — but you don’t seem to have any immunity left from your childhood rounds so it wouldn’t hurt to get it again.”

                I think Hep B or C or Omega was the latest one I had to go in for.Report

              • Jaybird in reply to Kazzy says:

                I don’t mind getting the whooping cough or whatever boosters because any side effects I get are normally gone by suppertime.

                The mRNA shots take me out of the game for two and a half days. I’ve lost multiple weekends to them.Report

              • Michael Cain in reply to Jaybird says:

                So, in just over a year the medical researchers and developed countries’ governments provide you with a medical miracle, and you want to engage in speculative whining? IIR your details correctly, you’d need to find a doctor to write a prescription for off-label use for you to even get a fourth dose legally in the US.Report

              • Jaybird in reply to Michael Cain says:

                I’m more engaging in speculative questioning and noticing that someone who got both shots and the booster and is asking “why am I getting a booster for the same stuff as last time” is getting the Ivermectin treatment.

                (I don’t need a fourth dose yet. But it will officially be recommended for me later this year.)Report

              • Michael Cain in reply to Jaybird says:

                …noticing that someone who got both shots and the booster is asking “why am I getting a booster for the same stuff as last time” is getting the Ivermectin treatment.

                Because they’re just looking for an excuse to duck a fourth shot. With no evidence whatsoever they hypothesize a vaccine that could be used as a booster and would work better against the full spectrum of variants as a fourth shot of the current vaccines. You know how much added protection hypothetical vaccines in a hypothetical protocol provide? Zero.Report

              • Jaybird in reply to Michael Cain says:

                It’s more that it’s with minor evidence. I mean, NPR had a report on this yesterday and it’s vaguely noncommittal and touches on some of the stuff that I’m wondering about.

                I know we’re never going to have perfect and I know better safe than sorry but I’m finding myself vaguely irritated at the walkbacks on the strength of vaccination and I am now wondering how much more I’m going to have stuff walked back.

                Fingers crossed we get more and better info by October.Report

              • Michael Cain in reply to Jaybird says:

                Not minor evidence. Solid evidence of minor benefits. And a very consistent story that the more at risk a group is, the larger the probability of benefit. Eg, one of the quoted people says the recommendation shouldn’t be people over 50, it should be people over 60. That’s a value judgement, and pretty fine-grained. If I were 50 and fit, single, would I opt for a fourth dose? Don’t know. Maybe losing a weekend to the side effects isn’t worth it. At 68, with a wife whose dementia is progressing and who is dependent on me, it’s not even close. I’ll take whatever benefit I can get.Report

              • Jaybird in reply to Michael Cain says:

                Oh, believe me, I’m not saying that YOU shouldn’t get one.

                And, heck, if there was one that used BA.2 as a baseline, I’d get in line for that… even at the cost of a weekend.Report

              • Michael Cain in reply to Jaybird says:

                And here’s the deal. We know about the BA.2 set of mutations today. In 30 days Moderna and/or BioNTech might have spec’ed the mRNA for it and confirmed that the current set of lipids are compatible. Call that point X. If they started immediately from there, at X+60 days they would have vaccine, because producing the vaccine in meaningful quantities is batch processes that take 60 days. At approx X+180 days they might get FDA EUA approval, if the field testing went well. But there’s a chance that the field testing won’t go well, and a chance that at X+180 days no one cares about BA.2 any more, and a chance that at X+180 days the FDA might not be doing mRNA EUAs — there are laws. In that case, approval is at X+Y*365 days, Y>1 and possibly Y>>1 because licensing when there’s already a fully licensed vaccine is a years-long process.

                So, your choices for at least the next several months are a fourth dose of the current vaccines, or nothing. The others are imaginary.

                I’ve forgotten Maribou’s exact status. Two unpleasant days to reduce her risk? Not my problem.Report

              • Jaybird in reply to Michael Cain says:

                Her status is shot twice, boosted twice.

                I wasn’t aware that the booster does anything to prevent a person from getting it (and therefore spreading it). I thought it just reduced symptoms at this point.

                (Granted, a year ago, I thought that it prevented you from getting it.)Report

              • Kazzy in reply to Jaybird says:

                I’m not saying everyone should get shots willy-nilly. But if folks are saying, “I never get boosters for other vaccines!” I am living proof that that isn’t always true.Report

            • Brandon Berg in reply to Jaybird says:

              The best analogy would be great great great…[hundreds of greats]…great grandclone. So have yours come over and take a picture together.Report

    • pillsy in reply to Jaybird says:

      You’re correct: flu shots vary year to year, precisely because the flu strains you have to worry about vary year to yer.

      As for the alpha shots, we know that alpha and especially boosted alpha were great against delta, and pretty good against omicron.

      But it’s hard to make good predictions about strains that may come out in the future.

      Still, I’m getting the 4th booster based on past performance, and because everything we know about COVID suggests that repeated challenges, whether from vaccination or natural exposure, improve immune system response, and that’s helped with cross-protection until now.

      It’s not always that way. For instance, getting dengue a second time is more dangerous than getting it the first time, and getting dengue the first time already sucks really bad.Report

      • JS in reply to pillsy says:

        The most recent paper I saw showed that, interestingly enough, “natural” immunization from Omicron (that is, if you weren’t vaccinated and caught omicron) doesn’t seem to protect against any other strain of COVID.

        The vaccine-induced response, however, is still quite robust against omicron (at preventing serious illness, hospitalization, and death). You get more mild breakthroughs with omicron than delta and more from delta than alpha, but protection is quite high.

        Modern vaccines target the hardest to change aspect of a virus (in this case the spike protein. Even omicron still HAS that protein, and the changes are fairly minimal because the spike protein is exactly how omicron infects cells — it’s got to keep that in largely the same shape or else it stops being able to get inside a cell at all) — whereas ‘natural’ immunity will evolve to whatever works. Which may be the spike protein, or might be something far more malleable.Report

    • Brandon Berg in reply to Jaybird says:

      IIRC there were clinical trials for updated boosters that failed to show any improvement over the an additional dose of the original vaccines.

      That doesn’t seem intuitively plausible to me, but I’m not an immunologist, so what do I know?Report

      • pillsy in reply to Brandon Berg says:

        One thing that’s been known to happen, and can cause some very surprising results along these lines, is that the immune response we measure (in terms of antibody titers) is not the immune response that actually provides efficacy against the disease, or is only part of it.

        So we may not be entirely sure why the original vaccines worked as well as they did, and the improvements or updates may not be improving or updating the parts that really do the work.Report

        • Mike Schilling in reply to pillsy says:

          This is the advantage of ivermectin. We know exactly how well it works.Report

        • Brandon Berg in reply to pillsy says:

          One possible explanation, I suppose, is that a freshly boosted alpha vaccination does so well that there’s no room for improvement that can be detected with a reasonably sized trial. If the original vaccine is 92% effective against symptomatic infection with omicron for three months after the third dose, and the other 8% mostly just comes down to a subset of the subjects having immune systems so weak that no vaccine is going to prevent symptomatic infection, then an omicron-targeted booster might only increase that to 93.5%. I haven’t worked out the math, but you probably need a huge trial to get statistically significant results with that small a difference in efficacy.Report

  5. Jaybird says:

    Report

  6. DensityDuck says:

    “Earendel”

    haha, nerds

    “Yeah, but you got the reference without needing it explained, right?”

    um (pulls collar) um, um–Report

  7. DensityDuck says:

    “we shouldn’t be using a horrifying toxin as rocket fuel”

    …you’re joking, right?Report

  8. Kazzy says:

    Jesus H Christmas:

    “ Mr. Adams had announced last week that he planned to stop requiring masks for children under 5 at childcare centers and preschools starting this Monday. But cases have started to rise over the last month, from about 500 daily cases in early March to about 1,250 daily cases now. The city could soon move from a low risk level to a medium risk level, according to a new color-coded alert system.

    Mr. Adams, a Democrat who took office in January, has been almost singularly focused on the city’s recovery from the pandemic and has rolled back some restrictions, including removing a mask mandate for students in kindergarten through 12th grade and proof-of-vaccination requirements for restaurants and gyms and for athletes and performers.”

    Someone explain this logic to me:
    1. Everyone else unmasks regardless of vaccine status.
    2. Vaccine requirements lifted.
    3. Special exemptions for Kyrie Irving!
    4. Cases rise.
    5. Must be those darn kids. Keep ‘em masked.Report