On a one-year followup, people who had covid have a higher risk of a wide variety of cardiac conditions. Yeah, great. The added risk to those who were not hospitalised seems to be low, but still statistically significant. (And I don't like looking at any of those green bars, thanks.) And of course those who had been in ICUs were at much higher risk.
So when those folks die from those conditions, this year or in five years (but five years early) it won't be "with" covid, because they have supposedly recovered from covid. And their death certificates will not read that they died from it. But you could say they died from it at some level, couldn't you? If covid causes an increase in heart attacks, then some of those people - we can't tell which ones, but some group of people in that pile - had a heart attack they wouldn't have otherwise had.
But maybe we shouldn't count those as "real" heart attacks or strokes, because those people had a preexisting condition: covid. It's the same reasoning...
This is what happens when people make up artificial distinctions to try and evade reality. They back themselves into logical corners with reasoning that seemed good at the time.
If I had, say, rheumatic fever that had left me with a irreversible heart condition that led to my death 15 years later, would it be on my death certificate? Or a recovered bone-marrow cancer in remission, but with a compromised immune system - would my death certificate from, say, a staph infection, carry that as an underlying cause? My instinct as a non-medical-professional tells me no.
ReplyDeleteI would just suggest that death certification should be a standardized process (but appears to be anything but) and with those standards, COVID should not be treated as an exception just because it's the most recent high-impact phenomenon. As a layman, the sensible thing to do would seem to be to note underlying causes if they're proximal and significant contributory factors. One thing COVID seems to be showing is that our data-keeping habits in the medical field are far from standardized and often subject to on-the-spot subjective preferences. What is irrelevant to some is considered front-and-center by others.
Serious question, AVI. Don't you have to distinguish between those people in the set (Confirmed Covid, not injected) and those in the set (Confirmed Covid + injected)? Are there differences in the outcomes?
ReplyDeleteThe 'non-hospitalized' group...presumably people who were diagnoses with Covid at a healthcare facility, implying that they felt ba enough to come in rather than just getting a home or clinic test and recovering with minimal problems....which would imply that this population is at least somewhat worse-off than the general Covid-catching population...Reasonable logic?
ReplyDeletePer Gavin's question, why on earth didn't they categorize by vaxxed and unvaxxed?...or maybe I missed in in a quick scan of the paper.
The paper is checking in after a year, so none of these folks were vaccinated when they got Covid, I believe.
ReplyDeleteGraniteDad is correct, the cohort was those diagnosed between March 1st 2020 and January 15th, 2021. It is likely almost none of them were vaccinated. A quick Google suggests <1% of the population was fully vaxxed by the end of the study.
ReplyDelete@David Foster- I don't think you can rule out people who got diagnosed in a clinic. The sample was all those in the VA healthcare system so from the way I read it, all their outpatient/clinic settings were included. Given that they have free testing with far more limited wait times than most places (at least in my area) I would imagine most people who could do that, did. Home tests were quite expensive/not terribly available a year ago IIRC. To be fair though, this does mean people who had asymptomatic COVID/thought they had a mild cold are mostly excluded since they wouldn't have sought a test.
@Aggie - so for bone marrow cancer, I can actually tell you what happens! I've worked in that area for 14 years now. If you have bone marrow cancer and receive a marrow transplant, by federal law (Stem Cell Therapeutic and Research Act of 2005, 2010, 2015 and 2020) your health outcomes must be reported to a national database. There are hundreds of pages of guidance for how to report, but all deaths up to 1 year post-transplant regardless of cause are required to be reported in our outcomes. We can certainly note that it was actually a car accident (a thing that has happened!) but this keeps the comparison across hospitals fair. IIRC, some hospitals have been given exemptions/asterisks in years of large local disasters (like hurricanes) that would have been out of their control, but otherwise the mortality numbers you see do have other things included. The reasoning is as you mentioned: we have a different focus. We want to keep people healthy. If we start to see that one center (or all centers) are losing a bunch of cancer patients to car accidents, we may ask ourselves if they should be advised not to drive for some time after transplant. Perhaps some medication is making them dizzy or impacting vision somehow. Likewise with COVID. If we know the chance of cardiac issues goes up post-COVID, people should be told so they can have a heads up not to that chest pain that might have been starting.
I agree the lack of standardization with COVID deaths is an issue. Over on my blog, I've been tracking all cause mortality instead, to see which states are doing worst on that. Counting every death is the only way to make sure everyone's doing everything the same way.
https://graphpaperdiaries.com/2022/01/02/state-level-excess-mortality-december-29th-2021/
Thank you @ bs king. I posted before this thought occurred to me: What about comorbidities? The CDC has said that, on average, US COVID deaths are associated with 4x comorbidities. The pre-print makes no mention of these facctors, but maybe it handles data that is too early to consider them.
ReplyDeleteThe CDC statement you're referring to is for COVID deaths of people who had been vaccinated. So the CDC is referring to a different set of deaths when making that statement than the set of deaths referred to here.
DeleteOr, not just deaths, I guess, but all kinds of heart problems in this data set.
DeleteYes, this data set actually excluded those who died of COVID.
ReplyDeleteThe comorbidities for the COVID patients in this paper were reviewed in supplementary table 4 and 5. I have not combed over it in depth, but it appears age was a factor (average age was 60 for non-hospitalized, 68 for both the hospitalized and ICU groups). Once they age weighted (table 5) I didn't see any stunning differences in comorbidities. Will admit I was reading it on my phone though, could have missed something.
So when those folks die from those conditions, this year or in five years (but five years early) it won't be "with" covid, because they have supposedly recovered from covid. And their death certificates will not read that they died from it. But you could say they died from it at some level, couldn't you? If covid causes an increase in heart attacks, then some of those people - we can't tell which ones, but some group of people in that pile - had a heart attack they wouldn't have otherwise had.
ReplyDeleteThere's a logic to that, certainly.
It seems as if it applies to potentially... well, potentially almost every death henceforth. Maybe for Joe it was alcoholism, increased by a year or two of limited employment, the crashing of his business, and subsequent increased drinking. Maybe for Sam it was heart disease, heightened by COVID. Maybe for Sally it was depression leading to suicide, which depression was enhanced by isolation for months. Some people died of drug overdoses because they weren't working and didn't know what to do with their time, and the drugs were tainted.
Every one of them could have "...and COVID" on their death certificates; and maybe everyone else could too. Maybe you lost friends and family directly to COVID, and while you had no trouble with it yourself the grief and loss of connections brought on the fatal condition. Maybe the end of regular church services stripped your life of meaning, or foiled a drug rehabilitation that had been going well when you could hold hands and pray with other people. Supply chain issues from global disruptions deprived someone of needed medicines for other conditions.
The same would even follow if they died of an issue caused by the vaccines (granting arguendo that at least some people die of adverse reactions) since the only reason they might have taken a vaccine for COVID is because of COVID's existence.
Something that touches everything ("pan-") ends up being found everywhere.