Executive summary
The US nursing home data shows that the COVID vaccines increased the odds of death from COVID by 6%. The result is highly statistically significant. Emory University infectious disease Professor Carlos del Rio reviewed my work, thought it was flawed, and encouraged me to submit it to a peer-reviewed journal where he presumed it would be rejected. If there is a mistake, I want to find it before I submit it to the journal. If you are a data scientist or epidemiologist who is willing to verify my result, this would be extremely helpful.
Data source
US Nursing home data collected by CMS from 15,410 providers who submit COVID cases and deaths within their facility weekly since week ending 5/24/20. This is the most comprehensive dataset on the single most important demographic for COVID:So if the COVID vaccine was or wasn’t effective for this group, this data would tell the story. This data is also the most credible because there were over 15,000 independent observers. So we can look at the aggregated results while minimizing any systematic errors as might occur if the numbers were collected by a single entity.
Methods
I downloaded the raw data from the site (the .csv files). The data was processed using an R program. This resulted in the elimination of all data submitted by 2.3% of the nursing homes. I looked at various values of N, with N=12 being the most reasonable to detect an effect. The Excel spreadsheet (see the week tab in the top right) allows you to vary this and it didn’t change the result if you moved it later, e.g., by 3 weeks since the vaccination rollout was widespread by 12/27/20.
Additional evidence relevant to the nursing home data
All this data has been available publicly for at least 2 years. The IFR matches numbers found in peer-reviewed studies. But the nursing homes only tested symptomatic patients. Nearly 78% of original strain COVID infections are asymptomatic so that means our 16% should be reduced by a factor of 4.5 to get the true IFR (our IFR was calculated based on symptomatic cases). So this means that certainly by Feb 7, 2021, which is well after half the nursing home population is fully vaccinated (see chart below), we’d see a huge IFR reduction vs. baseline. So our 12 weeks analysis from 12/6/20 included data until March 7, 2021.
Limitations
In all the nursing homes I am aware of, all newly admitted residents and all symptomatic patients are tested for COVID.Some people have suggested that nursing homes were expressly told not to count infections of newly admitted residents The data dictionary doesn’t mention this as a factor. If this were material, they would have mentioned it in the data dictionary. Nor did the reporting instructions (the Interim Final Rule Updating Requirements for Notification of Confirmed and Suspected COVID-19 Cases Among Residents and Staff in Nursing Homes) mention that nursing homes should not count such cases.
They told people the vaccines reduce your risk of death by 10X. It did no such thing.
They told people that the COVID vaccines would reduce the risk of death by 10x. For example,Here’s the death image (lower right in the tweet) showing the 10X reduction:
Reality: It increased your odds of death from COVID by 6%
The IFR was 5% higher in the 12 weeks after the vaccine rolled out. This translates into a 6% increase in the odds of death. If we do a Fisher Exact Test to see if this is statistically significant and to look at the confidence intervals, we see the result is highly statistically significant. It shows that the US government forced people to take a vaccine that would increase their risk of death from COVID.An earlier paper showed that the COVID vaccine also increases your risk of getting COVID.In short, the COVID vaccines were all risk and no benefit.
If the vaccines worked, it would show up in the US Nursing home dataset
If the vaccines worked as promised (a 10X reduction in death as claimed by certain health authorities), the very first place you’d see a huge and clear benefit would be in this data. You see the opposite: a rise in the risk of death on the same variant as before the vaccines were rolled out (so it wasn’t the variant).All in all, the COVID vaccines were the deadliest medical interventions in US history.
The consistency of the data
If you look at the 1 week before vs. after analysis of the dataset, the numbers are nearly identical. But as the weeks increase, the differences are apparent. If you look at the 24 week period before the vaccine rolled out, the second 12 weeks had a lower IFR than the first 12 weeks. This is exactly as expected as the IFR gradually drops over time (more people have been infected once already so on the second time, it’s a lower death risk, dry tinder has been burned). Note also that when averaged over a few weeks or more, the IFR is quite stable (see the various analyses in the weeks tab on the spreadsheet; I compute for 1,4,8,12,16, and 24).If you are in the same variant, the only thing that can cause a rise in the IFR of a virus is human intervention.
This is official US government data and the single most complete and trusted dataset for what happened at US nursing homes before and after the vaccine. The IFR should have dropped like a rock after the vaccines rolled out. But instead the odds of death increased by 6%. To make this as solid as possible, I want to have at least 20 highly qualified peer reviewers. This work has already been reviewed by others including Norman Fenton who praised it.
No comments:
Post a Comment