Renz "Whisteblower" data from CMS falsely claims death rate higher for SARS-CoV-2 vaccine than flu

Back in October, I write a blog post responding to the lawyer Renz's "whistleblower report" claiming that The USA government's Center for Medicare and Medicaid Service's (CMS) database showing that >50,000 Medicare recipients had died within 14 days of receiving the SARS-CoV-2 vaccine was evidence of vaccine-caused deaths. I demonstrated in that blog post that given the high death rate of the >65yr old Medicare population, these number of deaths fell within what was expected as the background rate of death in the population and so did not provide any evidence at all that the SARS-CoV-2 vaccines had caused any excess deaths in this population.

Well, in a recently updated slide deck, Renz is at it again. He presents the following table comparing SARS-CoV-2 vaccine death rates in 2021 up against Influenza vaccine death rates in 2018-2020 suggesting the death rates are >2x as high and again claiming this means the SARS-CoV-2 vaccines are causing high numbers of excess deaths.

This table is cited by WaynetheDBA and SteveKirsch as providing evidence for their dramatically high estimates of SARS-CoV-2 vaccine-caused deaths in the USA (Kirsch's estimate is now up to 388k vaccine-caused deaths).

Well, if we look lower in Renz's presentation, he presents tables split out by 1st and 2nd dose:

Ignoring the inconsistency that the total vaccine deaths don't add up (50,429 vs. 52,030), the key fact here is that these ~50k deaths are split between 1st and 2nd dose, and so cover a period of 28 days, not 14 days.

Since the influenza shot consists of a single shot (for adults), the reported deaths from 2018-2020 cover a period of 14 days. Thus, the fact that we see ~2x the rate of deaths after SARS-CoV-2 shots covering a 28 day period than the rate of deaths after influenza shots covering a 14 day period should not be surprising.

Here is an updated version of Renz's table splitting out the deaths and estimated death rates by 1st and 2nd dose, so as to cover the same 14-day period of time corresponding to the influenza data.

Since Renz does not give the total # of Medicare beneficiaries who returned for second SARS-CoV-2 shots, I estimate this number. From USAFacts, we see that ~86% of those 75yr+ returned for 2nd shots, and ~90% of those 65-74yr returned for 2nd shots, so we estimate that 86% of Medicare beneficiaries returned for 2nd shots.

From this, we see the death rates within 14 days of SARS-CoV-2 vaccine are essentially equivalent to the death rates within 14 days of Influenza vaccine.

Although not all vaccinees have their vaccination status recorded in the CMS system, since they may have gotten their vaccines freely elsewhere and not reported it, this does not affect this calculation, since the denominator is those in the CMS system whose vaccines were documented, and the numerator is those in the CMS system whose vaccines were documented who died within 14 days of the shot.

Thus, we see that there is ZERO evidence of any excess deaths in the Medicare beneficiary population within 14 days after receiving SARS-CoV-2 vaccines based on the rates seen in Medicare beneficiaries within 14 days after receiving influenza shots in 2018-2020. On the basis of these data, if one wants to claim that there are excess deaths within 14 days of SARS-CoV-2 vaccine that are caused by the vaccine, they would have to also believe that there are equivalent numbers of excess deaths caused by Influenza vaccines within 14 days of the shot.

CMS data much more suitable for determining vaccine-caused deaths than VAERs By the way, these CMS data provide a much more rigorous way to assess potential excess deaths after vaccination than VAERs data, given the fact that VAERs data are not verified, have no control group, and suffer from high levels of reporting bias that are difficult (if not impossible) to estimate and account for. The relative completeness of the CMS data covering the USA Medicare population, lack of voluntary reporting bias, and the existence of a natural control (the corresponding death rates for the same time frame after Influenza vaccination in recent years), these data are much more suitable for evaluating the potential of vaccine-caused deaths than VAERs.

When accounting for the fact that SARS-CoV-2 is given in two doses and influenza vaccines in 1, these data do not provide any evidence that SARS-CoV-2 vaccines cause any excess deaths relative to what is experienced after influenza vaccines in 2018, 2019, or 2020.

Of course, an important caveat to keep in mind is that the SARS-CoV-2 and influenza vaccines may be given to different subsets of the population. However, we see that the total number of beneficiaries receiving Influenza and SARS-CoV-2 shots are within 10-15% of each other, so the populations are likely to be reasonably comparable.

These data also provide incontrovertible evidence that proportion of deaths reported to VAERs is much higher in 2021 than in 2018-2020.

Incidentally, given the MUCH higher reported death rates in VAERs in 2021 relative to 2018-2020, these CMS data provide convincing evidence that indeed a MUCH higher proportion of deaths after vaccination are being reported to VAERs in 2021 than were reported in previous years. Ironically, these data are used by WaynetheDBA and SteveKirsch to justify very high estimates of the underreporting factor (URF) for VAERs, the proportion of events occurring after vaccination not reported to VAERs, that allow them to extrapolate VAERs numbers to their outrageous vaccine-caused death estimates.

As I have mentioned previously, the key assumption driving Kirsch's incredibly high numbers of vaccine-caused deaths (388k by his most recent calculations) is a very low estimate of background death rate, i.e. the number of expected deaths in the vaccinated population after vaccination even if the vaccines caused none of them. He does this by explicitly assuming the VAERs reporting is not any higher this year during the pandemic than previous years, and claiming

all VAERs reports more than 2x the number reported in previous years must be caused by vaccines (and then inflating this number by his assumed 44x underreporting factor). He never considers the life table based background death rates in the vaccinated population that scientists use in computing background death rates to detect safety signals in VAERs.

When evaluated carefully, these data demonstrate that this is a false assumption, and when it comes to deaths after vaccination, clearly a much higher proportion are being reported to VAERs in 2021 than previous years.

Unwittingly, their use of Renz's CMS data undercuts their entire argument.

The rates of death within 14 days after SARS-CoV-2 or Influenza vaccination from these CMS data are not higher than baseline background death rates for a 14-day period in the USA >65yr old population.

By the way, just in case anyone starts getting alarmed and worries that maybe Influenza vaccines are causing excess deaths, we can compare these rates from the CMS tables with overall background death rates in the USA >65yr population:

We can see the 2018 annual death rates per 100k split out by age group are as follows:

And from USA population numbers by group:

From this, we can compute that among the 65yr+ population eligible for Medicare, we have 58.4% 65-74yr, 29.7% 75-84yr, and 11.9% 85yr+

Taking a weighted average of the respective age-specific death rates, we see that the combined 65yr+ cohort in the USA has annual death rate of 3950 per 100k, which would be 151.5 per 100k for any given 14 day period. Given that the 14-day death rates after SARS-CoV-2 or Influenza vaccination in the CMS cohort is ~100-110 per 100k, the post-vaccination death rates are about 1/3 lower than background death rates, so there is no evidence on the basis of these CMS data for any vaccine-induced excess deaths from either SARS-CoV-2 or Influenza vaccines.

Note that the lower-than-background death rate is likely due to selection bias, i.e. what has been called the healthy vaccinee effect, with the vaccinated subset having lower death rates than the overall cohort because (1) individuals in hospice or otherwise at high risk of imminent death may be unlikely to receive vaccination, either SARS-CoV-2 or Influenza (and BTW this effect is temporal and would be strongest on day zero and gradually decrease over time after the bias from selection wanes and the cohort's death rate reverts towards background levels), as well as other factors including the potential that people >65yr who follow strong public health recommendations to be vaccinated may also tend to follow other public health recommendations and have inherently lower death rates, as has been observed in other data (BTW this effect is not temporal, but an inherent difference between cohorts that can endure)

Appendix: Response to Steve Kirsch

Steve Kirsch has written another substack post personally attacking me today.

I will respond here, since he only posted on his substack, to which I cannot post responses unless I pay him a monthly fee. I am not interested in paying him a monthly fee. So hopefully anyone interested in my response will find it here.

This post is supposedly in response to this blog post here, although he never addresses my points:

1. The very CMS data that he cites, when adjusting for their denominator problem, suggests that the death rates after SARS-CoV-2 vaccination are the same (or a little lower) than the death rates seen after Influenza vaccination in 2018, 2019 or 2020.

2. As Steve has emphasized repeatedly, it is clear that more deaths after vaccination are reported to VAERs this year than previous years, by a lot.

3. These two together show that, whatever the underreporting factor (URF) for death is, it is much lower than it was in prior years (2018, 2019, 2020 in particular when flu shots were the main ones given to adults in the USA).

4. If Steve was concerned about finding the true levels of vaccine-induced deaths, he would put more focus on rigorous data sources like these CMS data than VAERs, given its known limitations, instead of completely ignoring these sources and basing his argument almost entirely upon his VAERs projections and anecdotal reports.

Well, he doesn't directly respond to any of these points at all (as he has NEVER responded directly to the many points I have brought up suggesting he is using erroneous arguments and flawed logic in constructing his narratives in my long critiques of his written documents here and here). He says all of these points are "off topic" and leading to a "rat hole":

He says my points here are a distraction from the "real" issue of estimating the URF.

No, Steve, the real issue is not estimating the URF, it is evaluating the veracity of your dramatic claims -- which right now is that 388,000 people in the USA have deaths caused by the SARS-CoV-2 vaccines. As I will explain below, the URF is a secondary factor to this key question, and the issues I bring up in this post get at the key fundamental assumptions you make about background death rates that really drive your conclusions.

As I describe above, you get your estimate of vaccine-caused deaths in a simple fashion:

1. Assuming ALL VAERs reports of deaths in 2021 above 2x the typical number in pre-pandemic years MUST be caused by the vaccine. Using this number as the "background reporting rate", you only filter out a small number of VAERs-reported reported deaths as background based on what was reported to VAERs in pre-pandemic years.

2. Then you multiply this number of "assumed causal deaths" by an URF estimate, 41x before, and 44x now to extrapolate that number up to your (implausibly) high estimates vaccine caused deaths in the USA, which you recently claimed to be 388,000.

So your estimate depends COMPLETELY on your explicit assumption that there is NO higher reporting to VAERs in 2021 than previous years (and neglects to perform any real calculation of background death rates in the vaccinated population, which are the scientific way to adjust for background death rates).

The tables from Renz's post of CMS Medicare data that YOU shared in your blog post as evidence for your URF, suggest that this assumption is patently false.

Given we know that the death reports to VAERs are orders of magnitude higher in 2021 than ANY previous year (and the highest number of reports are deaths within 14 days of vaccination), if the CMS validated death rates in the 65yr+ USA population within 14 days of SARS-CoV-2 vaccination in 2021 are equivalent to the CMS validated death rates in the 65yr+ USA population within 14 days of Influenza vaccination in 2018, 2019, or 2020, then it MUST be true that a MUCH higher proportion of deaths within 14 days of vaccination have been reported to VAERs in 2021 than in previous years.

If these CMS data are accurate, the only conclusion is that your pivotal assumption that VAERs underreporting was just as high or higher in 2021 than previous years must be false. If your assumption were right, then we would see MUCH higher death rates in a population data source like these CMS data, but we don't. They are not higher at all.

This has EVERYTHING to do with your URF argument, and most importantly, your pivotal assumption that all VAERs death reports above reporting levels from pre-2021 years must be vaccine-caused deaths. Your unwillingness to engage with this question and respond to it, and instead try an ad hominem attack against me (again) in which you never address the substantive issues I bring up, says a lot.

As I went through your original 47-page document providing "support" for your DarkHorse podcast claims and your 32-page document making your case for SARS-CoV-2 vaccine deaths line by line, you also never responded to any of those substantive points. You complained my document was "too long" (ignoring the fact that any point by point response to a 47 page or 32 page argument is bound to be long), and claimed your followers would not have the attention span or interest in reading a written response to my critique.

Instead, you repeatedly challenge me to a Zoom debate, privately and publicly, calling me "chicken" like a junior high schoolboy, offer me thousands of dollars, even threatening to sue me for "defamation" for critiquing your claims unless I agreed to debate you. What would a verbal debate accomplish if you are not willing to respond to any of my substantive points? Why would I think that you would use logical scientific arguments and fair debate in that setting? What would that accomplish when after 6 months you still have not responded to any of my substantive critiques? I am not holding my breath that you will respond to any of these fundamental critiques I raise about your approach for estimating number of vaccine-caused deaths. I can't confirm that the CMS data posted by Renz are accurate, since I (sadly) do not have access to the full data set myself. But giving him the benefit of the doubt that they were indeed accurately extracted from CMS, these data are MUCH more appropriate for assessing the potential of deaths caused by vaccines than VAERs.

VAERs is an open reporting system, and it is well known that it is not possible on the basis of VAERs alone to determine causation -- there is no control group against which to compare. Your purported arguments of causation are not scientifically rigorous or valid, as anyone working in causal inference can tell you. Not only can we not determine causation from VAERs, but from VAERs we cannot even accurately estimate the actual death rate AFTER vaccination in the population, much less which deaths were FROM vaccination, since we do not know the denominator.

So given you value these CMS Medicare data enough to highlight them in a recent blog post, why don't you consider what they tell you about vaccine-caused deaths? Do you not recognize that the fact these data are covering essentially the entire 65yr old population, are not subject to reporting or underreporting bias, and have a natural control group (influenza vaccinated in pre-pandemic years) from which to compute background death rates? These data are MUCH more valuable than VAERs in answering your purported question -- finding the "truth" about vaccine-caused deaths. Why do you insist. on only considering VAERs in addressing this question? Is it because using the steps I highlight above you can concoct a super-high and dramatic death estimate that fits your narrative? Why don't you think for a second about what these CMS data say about the (in)feasibility of your 388,000 death estimate?

If there were 388,000 deaths caused by vaccines in 2021 as you now claim, how can it be true that the CMS Medicare database show similar rates of death from influenza in 2020, in 2019, in 2018 as SARS-CoV-2 in 2021? The only way you can believe that is if you think that 388,000 people were also killed by influenza vaccines during those years as well (keeping in mind that in a typical year there are ~2,800,000 total all cause deaths in the USA). Otherwise, a rational and open-minded person seeing these data would think that maybe the assumptions underlying your VAERs-based projections are wrong. How can you say that this issue is off-topic and irrelevant? As I show above, the rate of deaths in Renz's posted CMS data after SARS-CoV-2 vaccination in 2021, and after Influenza vaccination in 2018, 2019, or 2020, are in line with the typical rate of background deaths in the population. So where are these hundreds of thousands of excess deaths supposedly caused by vaccines? How do you account for this? The URF is a secondary issue here. The key issue with your vaccine-caused death estimation is your assumption of an artificially low background death rate in the vaccinated population that allows you to claim nearly all VAERs reported deaths are CAUSED by vaccines. The way you do this is to strongly assume that VAERs reporting rates have not changed during the pandemic, and so any additional reporting in 2021 over previous years is ONLY explained by causation -- deaths caused by vaccines. This is your sleight of hand and the driving factor of your conclusions, and these data show how erroneous this assumption is. The URF is a secondary issue relative to this, and even if you are correct on the URF of 41 or 44, the VAERs death repots are still within background rates if you compare them to life table based death rates in the population, as scientists do. I showed this in a step-by-step transparent presentation in my previous critique to your "2 dead for every 1 saved" claims, -- even assuming an URF of 41 as you claimed back then (or 44 as you claim now), there is still not evidence that the death rate after vaccination is any higher than the expected rate of background deaths in the population based on a 1 month time horizon for reporting (of course you never replied to this or explained what was wrong with my methodology or calculations). The key issue is not the URF. It is whether it is possible to estimate the number of vaccine-caused deaths from VAERs in the first place (it's not).

And it is what other more rigorous data have to say on the question of vaccine-caused deaths. These can serve as a feasibility check or validation of your VAERs-based estimates. Data like these CMS data covering essentially the entire 65yr+ USA population, lacking the reporting bias problem, and with a natural control group. Data like all-cause death data, including the UK all cause deaths data with deaths split out by vaccination status and age group, as I have posted on recently here and here and here. Data like all-cause death data showing total excess deaths in 2020 and 2021 over time, data that for many countries show the lowest excess death rates in the pandemic during the months when most vaccination was done.

Certainly, you must understand the uncertain nature of your VAERs based projections given their strong dependence on your assumptions about the URF and the background death rate, and should welcome these rigorous data to serve as potential validation of your conclusions. If you were serious about getting to the truth of this question you would engage these data sources, not keep making your argument based solely on VAERs data that depends completely on your strong assumption that all of the higher reporting to VAERs is from vaccine-caused deaths. You talk about truth and truth-telling. Truth is unknown, that is why we perform studies and collect data. The data contain the truth. As a data scientist, my goal is to use scientific and mathematical principles to evaluate and aggregate the available data to try to discern the uncertain truth from the data on hand. To interpret the truth as best we can with all available data. To do that, we need to consider all data sources, account for known and potential biases, and use legitimate scientific principles of statistical analysis and inference. That is how we interpret truth. If that process revealed the vaccines were causing hundreds of thousands of deaths, I'd be shouting that from the rooftops. But it doesn't. The available data strongly refute that and show it to be completely infeasible.