VA Study Published in JAMA Shows that COVID *and* Flu Shots Don’t Reduce Your Risk of Hospitalization

VA Study Published in JAMA Shows that COVID *and* Flu Shots Don’t Reduce Your Risk of Hospitalization


Of course, the study was designed to show a different effect. But it “accidentally” revealed that the COVID and flu shots don’t reduce your risk of hospitalization from the virus they are designed for

Executive summary

I just love it when a paper designed to show a pro-narrative result accidentally shows data proving that neither the COVID nor the flu shots reduce your risk of hospitalization from their respective diseases for the elderly. A double whammy.

The ridiculous CDC claim

You can’t make this stuff up. You really can’t.

For the record, here’s what the CDC said:

About the JAMA paper

Risk of Death in Patients Hospitalized for COVID-19 vs Seasonal Influenza in Fall-Winter 2022-2023.

This paper was published in JAMA and had over 100,000 views.

This paper looked at the risk of death in VA patients hospitalized for COVID vs. flu between October 1, 2022, and March 2, 2023. This is virtually all men, and the groups had a similar average age (71 vs. 73).

BASELINE characteristics were determined from the people they enrolled. Between October 1, 2022, and January 31, 2023, they enrolled all individuals who were admitted to the hospital with either COVID or the flu (but NOT both).

Note: the study excluded those hospitalized with both infections. However, many people in the study had taken BOTH the COVID and flu vaccines.

The BASELINE characteristics were the stunning result in my case because the people hospitalized with COVID shouldn’t have the same vaccination profile as the people hospitalized with the flu. More on this below.

But the main point of the paper is to show that if you were hospitalized for one OR the other (not both), that you were somewhat more likely to die from COVID than flu. I don’t disagree with that conclusion.

The paper looked ALSO looked at the risk of dying from COVID for the unvaccinated vs. vaccinated:

Here are the findings for DEATH:

  1. The paper found that there failed to be a statistically significant death benefit from 1 or 2 doses vs. the unvaccinated. It didn’t point this out. This is a big deal: if you were vaccinated with the primary series vs. unvaccinated, they found NO STATISTICALLY SIGNIFICANT DEATH BENEFIT. At the very best, the benefit was 2.32/1.66 which works out to 40% fewer deaths, but the numbers were simply too small to say it was better (the error bars overlapped by 50%).
  2. For boosted vs. the unvaccinated, the 95% error bars almost touched, so scientists can say that it’s 95% likely that the boosted people died less in the hospital if they got COVID (but there is a 5% chance that there also is no difference). But remember that if you get boosted and are still alive, there is selection bias.

What the study did not show

The study showed if you were hospitalized for COVID, that the boosted were less likely to die than the unvaccinated.

Does that mean you should be vaccinated?

No! The predicate is highly restrictive: “if you survived all the COVID vaccine shots and are still alive and you got COVID and ended up in the hospital, …”

The question an elderly person should be asking is whether people lived longer who got the COVID shots vs. people who didn’t.

In particular, this study did NOT track a randomly selected cohort from Jan 2020 and then did a time series analysis to see how the vaccinated performed against the unvaccinated in terms of death. They never seem to want to do these studies for some reason. I wonder why? Hmmmm….

I complained about this here:

No one was aware of any such study as you can see from the comments. I was not surprised.

If you want to prove the vaccine works or not, that’s the way to do it. Why hasn’t anyone done this? Could it be that such a study would expose a result that no journal would publish?

The stunning result they completely “forgot” to mention: there was absolutely NO HOSPITALIZATION benefit to being vaccinated

Here is the table from the paper:

Table from the paper. It shows the vaccination status breakdown of the “control group” (column 1) is the same as the group that got COVID (column 2). If the vaccine worked, the BASELINE characteristics of the people in Column 2 should be skewed to the unvacated. The BASELINE vaccine status breakdowns were virtually identical between the two groups. The biggest difference should have been in the Boosted group but it showed just a tiny difference, a fraction of 1%. That is a huge problem. It shows that the vaccination breakdown of the people who got flu was the same as the people who got COVID. In short, the vaccine hasn’t made a difference.

For our purposes, the flu group is essentially the “control” group. It shows you the baseline vaccination status breakdown of the overall VA population. This is because the COVID vaccine doesn’t protect you from the flu.

The COVID group (second column) shows you the vaccination status breakdown of VA members who were hospitalized with COVID.

If the vaccine worked as advertised above (10x reduction in hospitalization risk), we’d expect to see about 70% unvaccinated people being hospitalized for COVID.

But we don’t. It’s not even close. The breakdown is nearly exactly the same (this is why the standard mean deviation (SMD) numbers are so small).

In plain English, what this means is that the COVID vaccines did virtually nothing to reduce your chance of hospitalization from COVID.

This is a big whoops.

For some reason, the paper failed to note this issue. So I thought I would point it out for you in case you didn’t notice.

The second big reveal: the flu vaccine doesn’t work either!

As we know, the flu shot doesn’t protect you from COVID. So the % of people who got the flu shot in the COVID hospitalized group is our “control” in this case for looking at the effectiveness of the flu shot.

The baseline characteristics show that the percentage of people who got the flu shot is only slightly lower in the people who were hospitalized for the flu.

This is a second big whoops.

It shows that the flu shot didn’t work.

Is there a confounder that explains this? Not that I can even imagine.

For both of these results, you could legitimately argue that there was not a true control group. I agree with that. You could argue that. After all, there could be some unknown interaction that invalidates our assumptions that:
  1. In people who are hospitalized for the flu, their COVID vaccination status is not the same as the overall VA cohort
  2. In people who are hospitalized for COVID, their flu vaccination status is not the same as the overall VA cohort

But both of those seem extremely unlikely for four reasons:

  1. Why would there be a difference? What is the biological mechanism?
  2. The percentages in the baseline group match up reasonably well with expectations for those in the age 70 age group overall
  3. The vaccines are “supposed” to work and give you a substantial reduction and so when you have cohorts who are 61% vaccinated (flu) and 80% vaccinated (COVID) you’d expect to see a pretty big difference in the groups and you don’t, i.e., you’d have expected the flu group to have a much lower % of the flu vaccine than the COVID group and the COVID group to have a much lower % of the COVID vaccine than the flu group.
  4. The fact that the flu vaccine does nothing for reducing mortality or hospitalization in the elderly is already well established in the peer reviewed literature with a very extensive study that included 170 million episodes of care and 7.6 million deaths. See The Effect of Influenza Vaccination for the Elderly on Hospitalization and Mortality. So we should be surprised to see this inadvertently confirmed in this study. Thank you to John Amor for bringing this paper to my attention.

If someone thinks they have a confounder that explains the effect, they should tell us what it is. Nobody seems to be able to do that.

If I’m wrong, where is their data explaining why there is no difference between the groups?

I don’t see how it is possible for them to produce data showing how the baseline vaccine breakdowns can be so similar between the two groups.

But I’m open to seeing the data, not handwaving arguments.

Right now the data shows the vaccine don’t work.

If you want to dispute that, the burden is on them to produce the data explaining why there is no difference in the vaccine breakdowns between the two groups.

Is the vaccine breakdown in the flu group a reasonable control group?

Some people might claim, without any evidence, that the vaccine breakdown of the flu group is not the same as baseline.

And that appears to be the case.

But the problem is that this argument works against them because it shows that the vaccinated are overrepresented in the group that is hospitalized.

Which means at least one of the following things must be true:

  1. People who opt for vaccination are less healthy. This is unlikely because the vast majority of people who choose to get vaccinated are “health seeking” and there is a “healthy user bias” often associated with vaccines.
  2. People who opt for vaccination are made less healthy by the vaccine itself and more likely to be hospitalized. This is what seems to be happening here.

Let’s look at the numbers.

According to the VA website, as of October 18, 2023, only 5.3M veterans received at least 1 dose of the COVID vaccine (4.9M Pfizer or Moderna + .4M Janssen). To see this, click on the “Download VA data in CSV format” link at the bottom of the VA COVID19 home page.

There are 9 million vets enrolled in VA healthcare.

So the background rate of COVID vaccinated vets (with one or more vaccines) is 5.3/9 = 59%. It would even be less than that for this study since this is the percentage today.

Guess what the percentage has one or more COVID vaccines in the flu group? 81%.

So if they want to claim that this is not the the background vaccination rate, then this is DEVASTATING to the narrative because the vaccinated were overrepresented in those who got hospitalized in both groups.

In plain English, if you got vaccinated with the COVID vaccine, you were nearly 40% MORE likely to be hospitalized for COVID or flu, not less likely.

So they probably won’t bring up the background rate.

Or maybe they’ll argue that the vaccinated rate in those hospitalized is because it was only the sickest, most frail people who got vaccinated. OK, but there’s no proof of that, but let’s say it is true. Then, if the COVID vaccine worked as promised (since it is supposed to save the sickest from death from COVID), why is there no differential in the COVID vaccine breakdown in the flu group vs. the COVID group at baseline? That’s what nobody will explain.


If the vaccines worked, there would be a difference in the vaccine breakdown of the two groups at baseline with respect to each other. There wasn’t. The profile for both vaccines was identical between the groups. This means that neither vaccine worked to reduce hospitalization.

This paper unintentionally shows two important things for the elderly, which is the population most at risk of dying from these infections. It shows that:

  1. the flu shot doesn’t materially reduce your risk of being hospitalized for the flu and
  2. the COVID shot doesn’t materially reduce your risk of being hospitalized for COVID.

These results are likely applicable to everyone else as well, but the paper couldn’t explore that since it used a VA dataset which was primarily elderly men in their 70’s.

The paper was published in the peer-reviewed scientific literature in a “gold standard journal” using “gold standard” fully reported data from the VA.

This is really very embarrassing for mainstream medicine. I just wanted to point that out since nobody else has.

Original source: