Are the Vaccines Killing People?

CDC data have been sending vivid signals.

Top line result: As of the end of February, the vaccines may have claimed at least one life for every 12,800 doses, and the rate is increasing.

A few days ago, I posted an essay in which I had updated results from earlier analyses of mortality data from the Centers for Disease Control (CDC).  The headline results involved signals buried in the data.  These were results that I had not even made a point of looking for.  Those results illuminated the prospect that the vaccines really are killing people at an appreciable rate.  In this short essay I will suggest that the data are signaling the prospect that the vaccines are killing people at a rate on the order of 1-per-12,000 doses.  There is even some suggestion that this toll is disproportionately concentrated on very young people and that the fatality rate, as well as the absolute toll itself, will both markedly increase as lagged effects of the vaccine phenomenon continue to manifest themselves.

Most of my analyses to date have focused on “excess mortality”—that is, on deviations in all-cause mortality from average, benchmark rates.  One reason is that such analyses do not depend on fine categorizations of causes of death.  Attributing causes of death can be a tricky business.  Moreover, attributing causes when hospitals may have perverse incentives to attribute illnesses to COVID makes the business of attributing causes even more fraught. Nursing homes might not be too keen on attributing fatalities at their sites to COVID.  That would be bad for business.  But the CARES Act provides hospitals with a host of “supplemental awards” and other payments for COVID cases.  Critics have called these various payments “bounties.”[1]  Defenders, meanwhile, blithely observe that, “Under the CARES Act, the government will pay hospitals 20% more for Medicare patients with COVID-19,” whatever that means.[2]

Focusing on all-cause “excess mortality” gets away from all such controversies and from the potential, poor accounting of COVID fatalities.  It gets away from all poor accounting, but, rather, concentrates the analysis on the top-line result least susceptible to manipulation: total mortality. To put it coarsely: some very small number of people may yet go “missing in action,” but, generally, fatalities are hard to hide.  Total mortality may yet yield signals that are so coarse and discrete that one would not require privileged access to granular data and over-engineered statistical techniques to discern them.  That said, I did afford myself access to the publicly available data from the CDC on “deaths by select causes.”

Such data do indicate that the COVID phenomenon is very complex and has complicated the attribution of causes of death.  Deaths by flu and pneumonia largely disappeared for much of 2020/2021.  “Diseases of the heart” mostly increased, and fatalities associated with obesity, mostly diabetes, had the appearance of increasing sharply.  It has been long understood that there are tight mechanisms that render obesity a great driver of COVID infection and death; perhaps fatalities attributed to diabetes should have been attributed to COVID?

The diabetes matter is a puzzle, but one discrete signal that came out of the CDC data is clear once one uncovers it: fatalities attributed to “Symptoms, signs and abnormal clinical and laboratory findings, not elsewhere classified (R00-R99)” have deviated spectacularly from underlying benchmarks rates.  These rates exhibit seasonality, ranging from about 570 per week in June to 680 per week in December, but, by May 2021, these rates started to increasingly increase.  By the end of 2021, fatalities attributed to “abnormal … finding, not elsewhere classified” were exceeding 3,000 per week.  What is going on?

It is hard not to suspect that the campaign to get everyone vaccinated—treated with the mRNA gene therapies—has something to do with it.  In this short note I complement the results on “abnormal” fatalities with a graphical presentation of the total counts of doses of vaccines administered to the population.  I then close by outlining what a proper study of the effects of the vaccines on mortality might look like.

First let me graphically reprise the results I presented a few days ago.

These weekly data range from the first full week (ending Saturday) in 2014 through January 1, 2022.  Fatalities attributed to “abnormal… findings” exhibit the same kind of seasonality that most major causes of fatality exhibit.  And then they take off in the spring of 2021.

I generate the benchmark by regressing the weekly count fatalities on a sine wave and a linear time trend for the years 2014 through 2019.  I then project that benchmark on 2020 and 2021.  The difference between benchmark mortality and actual mortality constitutes a measure of excess mortality, the number of deaths in excess of what one might expect week to week in any given season.  The red area indicates excess mortality.[3]  This graph is much the same as one of the graphs I presented a few days ago.

In the next graph, I complement the mortality data with counts of doses of the vaccines administered per 100,000 people in the population.  I also indicate doses per 100,000 young people aged 12 – 17.[4]

The yellow area indicates the cumulative number of doses (per 100,000) administered to young people aged 12 – 17.  The blue area indicates the cumulative number of doses (per 100,000) administered to anyone in the population.  So, by the end of the year, every 100,000 young people aged 12 – 17 had received about 120,000 doses.  Some people will have received only one dose.  A smaller number will have received more than one dose.  Some number will have received zero doses, but, on average, one of these young people will have received about 1.2 doses.

The blue area indicates that every 100,000 people in the general population will have received more than 150,000 doses.  So, a person randomly selected from the population will, on average, have had about 1.5 doses.

A striking feature, of course, of these vaccine data is that they align with the increase in fatalities attributed to “abnormal” causes.  The United States started to roll out vaccines in December 2020, and, early on, most doses went to older people.  But the vaccine data show that most people only started getting doses well into 2021.  Indeed, as the New York Times reported on May 13, 20201,[5]

The first mass campaign to inoculate children against the coronavirus officially began in the United States on Thursday [May 13], after the federal government recommended making the Pfizer-BioNTech vaccine available to children ages 12 to 15. Teenagers 16 and older became eligible in most states last month [April 2021].

So.  Do these data prove that the vaccines are causing an appreciable number of fatalities, especially among younger people?  At the very least, I’d say that these data send a vivid signal that something very bad is happening and that the people who have privileged access to granular data should investigate the role, if any, of the vaccines.  Those people would be the health authorities over at the CDC, the NIH and such.

Will anyone conduct a robust study?  That is hard to imagine, but what would such a study look like?  Ideally, a study would work out of health data specific to all individuals, whether vaccinated or not.  A study could compute a person-specific likelihood of dying in a given interval of time—over the course of a given week, say.  The likelihood of any one person dying may be very low, but, with a lot of data, one might be able to compute a serviceable “hazard rate” for each person.  By each person I mean that the hazard rate would vary across measurable attributes of each person.  These individual attributes would surely include age, weight, sex, and indications relating to co-morbidities.  We might measure these attributes over time, because people are dynamic creatures.  They change.  Indeed, over a given week, some number of them will die.  Knowing an individual’s attributes at the time of death helps us calibrate hazard rates.  Basically, hazard rates are the formulas we use for assessing any one person’s (usually very, very small) likelihood of dying in a given week.

Attributes, of course, should include vaccination status.  When did an individual receive a given dose?  How much time has expired since the last dose?  Ultimately, do the data reveal any statistically significant effect of vaccinations on hazard rates?  If, for example, the vaccines are saving “millions” (as Donald Trump himself, suggested), then the data might reveal that, overall, the vaccines are inducing lower hazard rates; the vaccines are enabling people to live longer.  In contrast, the vaccines might be inducing adverse effects.  The vaccines might be inducing both adverse effects and benefits.  If so, do adverse effects dominate?  And, if so, does this show up in the data?  Basically, have the vaccines increased hazard rates, and are such increases both sizable and statistically significant?

A great thing about calculating hazard rates is that they accommodate the prospect that a person might die from any cause, much less an “abnormal” one, for reasons that have nothing to do with vaccines.  The Vaccine Adverse Effects Reporting System (VAERS) features a staggering number of reports relating to the vaccines, and, indeed, one could surely find reports of people who suffered a heart attack and died the day after getting a COVID jab.  But, conceivably, there had always been a small likelihood that that person would have suffered a heart attack even absent the jab.  An analysis of hazard rates accommodates this by modeling the prospect that anyone of us under any conditions might suffer some type of serious event.  The question is: do the vaccines make certain events more likely?  Or even less likely?  Do hazard rates discernibly increase or decrease post-jab?

Sorting that out is a homework assignment for anyone sitting on top of actionable data.  Will anyone do it?

Let’s return to the CDC’s publicly available data.  Excess mortality deriving from “abnormal” causes totaled about 22,000 from the March through December 2021.  It totaled at least another 21,000 in January and February of 2022. That makes for at least 43,000 fatalities in excess of baseline rates.  Meanwhile, the United States had administered about 510,000 doses of the vaccines by the end of 2021.  That number rose to over 550,000 doses by the end of February 2022.  Thus, at the end of 2021, the ratio of doses to fatalities was about 23,500.  By the end of February, the ratio of doses to fatalities had declined to 12,800.  As more data come in, it is hard not to expect that the number of doses to fatalities from “abnormal” causes to decline further. 

Is one death for every 12,800 doses a big number?  By the standards of vaccine campaigns in the United States, it is through the roof.  The reader can decide for him or herself.  Would 1-in-6,000 prove to be a big number?

About 62,000 very young Americans perished in the Vietnam War, and by the time the toll of the COVID pandemic had exceeded 60,000, some observers exclaimed that COVID was “worse than Vietnam!”  Note, however, that before the pandemic, as many as 60,000 Americans would die in a given week.  Every week was “like Vietnam” by the standards of these people.  When the official COVID toll exceeded 400,000, those same observers exclaimed that the COVID toll was “worse than World War II!”  Note, however, that before the pandemic, about 2.8 million Americans would die every year.  That’s “like World War II” seven times over—every year.  Note, moreover, that the toll of the pandemic has been concentrated on the very elderly and the infirm—on just the people least likely to live out the next year.  Harsh, but not as harsh as the tolls of Vietnam and the Second World War.  The toll of the wars was concentrated on young Americans in their physical primes.  Those tragedies were obviously far more extreme.  To compare COVID to America’s wars is grotesque.

A few days ago, Anthony Fauci re-emerged from the fog of news from Ukraine to declare that Americans should get their “booster” shots, because, “if, in fact, we do see a turn around and a resurgence, we have to be able to pivot and go back to any degree of mitigation that is commensurate with what the situation is.”[6]  So, we have to get booster shots, because the first two or three shots worked so well.  Weren’t the vaccines supposed to definitely resolve the COVID matter?

Suppose all Americans were to get four shots—the original two shots of a two-shot sequence and two “boosters”.  That would make for more than 1.3 billion doses.  Suppose, also, that one person dies for everyone 12,000 doses.  The data suggest that that would make for about 110,000 vaccine-induced fatalities (attributed to “abnormal” causes) of mostly younger people.  That’s nearly twice as bad as an actual Vietnam War.  Suppose the rate proves to be 1-in-6,000.  Then we end up with nearly four Vietnam Wars.  And then we have to account for fatalities from other causes, if any.

A proper study of underlying hazard rates could clear things up, but a high-level view of the data does not look good.

---

[1] https://aapsonline.org/bidens-bounty-on-your-life-hospitals-incentive-payments-for-covid-19/

[2] https://www.denverpost.com/2020/05/20/coronavirus-covid-medicare-payments-hospitals/

[3] I recover mortality data from two CDC sources: https://data.cdc.gov/NCHS/Weekly-Provisional-Counts-of-Deaths-by-State-and-S/muzy-jte6 and https://data.cdc.gov/NCHS/Weekly-Counts-of-Deaths-by-State-and-Select-Causes/3yf8-kanr.

[4] I recover these vaccine data from https://data.cdc.gov/Vaccinations/COVID-19-Vaccinations-in-the-United-States-Jurisdi/unsk-b7fc.

[5] https://www.nytimes.com/2021/05/13/us/coronavirus-vaccine-children-ages-12-15.html

[6] https://www.realclearpolitics.com/video/2022/03/17/fauci_we_cant_just_say_were_done_with_covid_time_to_get_your_booster_shot.html