Disproportionality analysis of anaphylactic reactions after vaccination with messenger RNA coronavirus disease 2019 vaccines in the United States.

One year after the emergence of the novel severe acute respiratory syndrome coronavirus 2 that causes the coronavirus disease 2019 (COVID-19), the US Food and Drug Administration (FDA) granted Emergency Use Authorization (EUA) for 2 novel messenger RNA (mRNA) COVID-19 vaccines. Reports of acute hypersensitivity reactions in the real world after EUA are creating anxiety among potential vaccine recipients and may delay achieving universal vaccination. Although the Centers for Disease Control and Prevention (CDC) and the FDA jointly monitor vaccine adverse reactions through a variety of surveillance systems, such as the Vaccine Adverse Event Reporting System (VAERS) or the CDC's Vaccine Safety Datalink, local health care professionals also play an essential role in monitoring vaccine safety, reporting to VAERS, and providing factual and up-to-date information to increase vaccine acceptance.

In a previous report, we have revealed the simplicity and versatility of the CDC Wide-ranging Online Data for Epidemiologic Research (CDC WONDER) in conjunction with proportional reporting ratios (PRRs) to obtain up-to-date information evaluating the reports on vaccine safety submitted to the VAERS. No disproportionate reporting of severe adverse drug reactions (ADRs), including anaphylaxis, was found to be associated with the use of the MMR vaccine compared with all other vaccines in 30 years. In this report, we used the CDC WONDER web interface and PRRs to evaluate whether the rates of anaphylaxis cases reported in the VAERS database secondary to the 2 currently available mRNA COVID-19 vaccines is disproportionately different from all other vaccines.

We used the CDC WONDER interface to retrieve data from the VAERS national database of all spontaneous ADRs and anaphylaxis reports, including “anaphylactic reaction” or “anaphylactic shock” from the “Symptoms” tab in the site request form, of individuals aged 18 years and older vaccinated with the Pfizer-BioNTech (New York City, New York-Germany) or Moderna (Cambridge, Massachusetts) vaccines in the United States between December 1, 2020, and March 5, 2021, and all ADRs and anaphylaxis reports of individuals aged 18 years and older vaccinated with all other vaccines during the same timeframe. Unknown and non-mRNA vaccines were excluded. We evaluated disproportionate reporting of anaphylactic reactions relative to all other ADRs after the COVID-19 vaccines compared with all other vaccines using the Evans criteria: PRR greater than or equal to 2 (where PRR is a / [a + c] divided by b / [b + d] in a 2 × 2 table), χ2 greater than or equal to 4 with Yates correction (adjustment for low frequencies), and greater than or equal to 3 individual cases. A disproportionate signal requiring further evaluation is detected if the reports meet all 3 criteria.

A total of 112,368 ADR reports were identified for the mRNA COVID-19 vaccines and 771 reports for all other vaccines between December 1, 2020, and March 5, 2021, with 185 reports of anaphylaxis for the mRNA COVID-19 vaccines and 1 report of anaphylaxis for all other vaccines over the same time interval. The PPR was 1.26 (95% confidence interval, 0.18-9.05), and the χ2 with Yates correction was 0.043. According to the Evans criteria, a disproportionate reporting of anaphylactic reactions after the COVID-19 vaccines was not found relative to all other vaccines (Table 1 ).

Table 1

Proportional Reporting Ratios Calculation of Coronavirus Disease 2019 Vaccines and Anaphylactic Reactions

ADRs	COVID-19 vaccines	All other vaccines
Anaphylaxis	185	1
All other ADRs	112,183	770
Total	112,368	771

Abbreviations: ADR, adverse drug reaction; CI, confidence interval; COVID-19, coronavirus disease 2019; PRR, proportional reporting ratio.

NOTE. PRR = a / (a + c) divided by b (b + d) = 1.26, 95% CI 0.18-9.05; χ2 with Yates correction (1 df) 0.043.

We found no disproportionate reporting of anaphylactic reactions with the 2 available mRNA COVID-19 vaccines after EUA was granted. The primary purpose of spontaneous ADRs reporting is to provide early warning of hazards not recognized before marketing an experimental drug because of limitations of clinical trials regarding sample size, duration, and generalizability to real-world practice. In the pivotal phase 3 trials of the mRNA COVID-19 vaccines, participants with a history of severe hypersensitivity reactions to the experimental vaccines or their excipients were excluded. Both trials presented low and similar hypersensitivity-related adverse events in the vaccine and placebo groups. , After the implementation of vaccination in the real world, reports of anaphylaxis after the first dose of mRNA COVID-19 vaccines emerged. The CDC initially reported 21 case reports submitted to VAERS after administering 1,893,360 first doses of the Pfizer-BionNTech COVID-19 vaccine that met Brighton Collaboration case definition criteria for anaphylaxis and 10 reports after 4,041,396 first doses of the Moderna COVID-19 vaccine, corresponding to an estimated rate of 11.1 cases per million doses administered and 2.5 cases per million doses administered, respectively. , In a recent update, after a total of 9,943,247 doses of the Pfizer-BioNTech vaccine and 7,581,429 doses of the Moderna vaccine, the CDC identified 66 case reports that met the case definition criteria for anaphylaxis: 47 after Pfizer-BioNTech vaccine, for an updated reporting rate of 4.7 cases per million doses administered, and 19 after Moderna vaccine, for a stable reporting rate of 2.5 cases per million doses administered. No deaths from anaphylaxis after vaccination with either vaccine were reported. Anaphylaxis after vaccination is indeed a rarely reported event, with an overall reported rate of 1.3 cases per million doses administered for all licensed vaccines.

The process of scrutinizing spontaneous ADR data for hazards is known as data mining and signal generation. Browsing the CDC WONDER website and applying PRRs is a relatively simple and straightforward disproportionality analysis method for signal generation based on comparing reporting proportions in a contingency table between the study drug and all drugs in the spontaneous reporting database combined. , However, this method has some significant caveats. First, the VAERS data consist of unverified reports of health events, both minor and severe, that occur after vaccination, and these reports may include incomplete, inaccurate, coincidental, and unverified information. Second, the request form in the CDC WONDER interface only includes broad and nonspecific terms for the search engine; hence, some submitted reports may not meet the criteria for anaphylaxis, overestimating the database numbers. Finally, PRRs and χ2 values are measures of association, not causality. Establishing causal relationships between vaccines and adverse events requires additional scientific investigation. Ultimately, this method cannot be used for comparative drug safety analysis beyond basic hypothesis generation. Despite these limitations, the CDC WONDER interface and calculation of PRRs are valuable aids for signal generation from spontaneous ADR data that provide clinicians and allergists versatile tools for quick evidence-based decision making after developing new vaccines and treatments.