key: cord-0153138-sdzu9pp4 authors: Cormack, Gordon V; Grossman, Maura R title: The Absurdity of Death Estimates Based on the Vaccine Adverse Event Reporting System date: 2022-02-09 journal: nan DOI: nan sha: a16ae10781f93da3c42eb259227e1fdcbf1b1c12 doc_id: 153138 cord_uid: sdzu9pp4 We demonstrate from first principles a core fallacy employed by a coterie of authors who claim that data from the Vaccine Adverse Reporting System (VAERS) show that hundreds of thousands of U.S. deaths are attributable to COVID vaccination. Arguments [1] [2] [3] [4] derived from VAERS [5] that COVID vaccines have killed vast numbers of people fail for one principal reason: They hinge on the false assumption that the reporting rate for adverse events following vaccination is strictly proportional to the actual incidence of such events. Rose [1] introduces the term "underreporting factor," = , which she claims to be invariant for all serious adverse events ("SAEs" [1] [2] [3] [4] , all relying either tacitly or explicitly on the false assumption that URF is constant notwithstanding differences in the type of vaccine, the nature and cause of the event, or the time elapsed from vaccination to event. These authors reject the prominent disclaimer [5] that must be acknowledged to access VAERS data: "The number of reports alone cannot be interpreted or used to reach conclusions about the existence, severity, frequency, or rates of problems associated with vaccines." A number of non-archival works [6] [7] [8] [9] have rebutted these false claims, citing authorities, evidence to the contrary, and sophisticated data analysis. Members of the coterie have in turn dismissed these rebuttals as the product of conspiracy or the work of ivory-tower academics disconnected from reality (See Box 1). The purpose of this article is to show, from first principles, using only basic mathematics and elementary logic that can be understood without specialized training, that the foundational assumption on which the false claims depend is not supported by the data. This analysis may be useful as a tool to help convince some of those who demand "first-hand" evidence that that COVID vaccine is not "a biological product that has killed over 150,000 Americans to date" [3] . reports of death following zoster vaccination. It does not require sophisticated mathematics to see that the reporting rate for deaths following vaccination is orders of magnitude greater for COVID than for influenza or zoster vaccines, while the actual incidence is comparable. A more precise analysis follows. "Reporting rate" is the number of reports of a prescribed event per person per unit time, expressed in this article as reports/million/day. "Incidence" is the number of actual occurrences of the prescribed event per person per unit time, expressed as events/million/day. It follows that, for a given prescribed event and interval of time, = . Causality is an important factor in determining the nature of an event. In particular, the events "death following vaccination" and "death caused by vaccination" are materially different in nature, with different incidence and reporting rates. Suppose, for example, that the actual number of deaths reported in VAERS came from two groups: 90% were from the inoculation-attributable death group and 10% were from the expected death group. Assume there is no overlap between the two groups. In that case, what VAERS shows is not that 1% of actual expected deaths were reported, but rather that 1/10 of one percent of the expected deaths were reported. If that metric is used as the standard to scale up to total deaths, then the number in the actual inoculation-attributable death group is not 100 times the VAERS reported deaths, but rather 1000 times the VAERSreported deaths! This argument is not only mathematically flawed, but precisely backwards. If 90% of reported deaths were vaccine-caused, it would indicate that the reporting rate for vaccinecaused deaths was much greater-and the URF much smaller-than for all-cause deaths. How much smaller is impossible to determine without knowing the actual incidence of vaccine-caused deaths. Kirsch et al. [3] , building on the same flawed reasoning, purport to show clinical evidence that a large proportion of reported events were vaccine-caused. Even if this clinical evidence were compelling, it would show only that the reporting rate for vaccine-caused deaths was higher than for all-cause deaths. they define to be the total number of deaths following COVID vaccination minus the total number of deaths following "the typical vaccine," independent of causation. In estimating the total number of deaths, they rely on the key false assumption that the URF for all-cause deaths is the same for COVID as for "the typical vaccine." Kostoff et al. [2], referring to a plot of reported all-cause deaths as a function of days since vaccination (cf. Figure 1 , below), argue: "If there were no effect from the inoculation, as To calculate reporting rate and incidence, we must specify a time interval. Consider first, the number of reported and actual deaths on the day immediately following vaccination. For the 2021 calendar year, among people at least 65 years of age, there are 409 reports of fatalities the day following COVID vaccination, 7 following influenza vaccination, and 1 following zoster vaccination. These numbers show reporting rates of 4/million/day, 0.2/million/day, and 0.12/million/day, respectively-all far lower than the incidence of 112 deaths/million/day in the general population. VAERS reports provide no evidence of "excess deaths," either relative to incidence in the general population, or relative to other vaccines. Rather, they indicate that all-cause deaths are underreported by a larger factor for other vaccines than for COVID vaccines. Figure 1 shows the decline in reporting rate over the 60 days following all COVID, influenza, and zoster vaccinations from January 1, 2000 to October 31, 2021. We see that there is no material difference in falloff between the COVID vaccine and the vaccines administered in "non-COVID years." The falloff indicates that all-cause deaths are underreported by a larger factor when temporally distant from vaccination. Similar falloff for COVID and non-COVID vaccines By examining reported all-cause deaths vs. known all-cause deaths, we have shown that there is no single URF by which one can multiply the number of VAERS reports to estimate the number of actual events, let alone the number of events caused by COVID vaccination. All of the arguments purporting to quantify COVID vaccine-related deaths from VAERS, and all of the arguments purporting to show causality, fail for the same reason-because they hinge on this false assumption regarding URF. Arguments relying on URF similarly fail to show evidence of elevated adverse events other than death, and hence offer no basis to disregard the clear VAERS disclaimer: "The number of reports alone cannot be interpreted or used to reach conclusions about the existence, severity, frequency, or rates of problems associated with vaccines." Critical appraisal of VAERS pharmacovigilance: Is the US Vaccine Adverse Events Reporting System (VAERS) a functioning pharmacovigilance system?. Science, Public Health Policy, and the Law Why are we vaccinating children against COVID-19? Estimating the number of COVID vaccine deaths in America Declaration of Jane Doe. America's Frontline Doctors v Becerra. ND AL Vaccine Adverse Event Reporting System There's no basis to claim thousands have died from COVID-19 vaccines Antivaccine activists use a government database on side effects to scare the public. Science Insider Toxicology Reports article vaccines kill 5 for every 1 saved. COVID Data Science Are the mRNA vaccines really safe? Evaluating claims by Steven Kirsch on danger of spike proteins. COVID Data Science Author defends paper claiming COVID-19 vaccines kill five times more people over 65 than they save. Retraction Watch Kirsch's callout of John Su and the inevitable clown attacks. Rounding the Earth Vaccine Adverse Event Reporting System. VAERS Table of Reportable Events Following Vaccination Shingles vaccination among adults aged 60 and over: United States