As of the middle of February 2021, around 1.3% of the world’s population has been infected with coronavirus disease 2019 (COVID-19), and over 2.3 million people have died (1). One person has been infected every 7.7 s, on average, since the start of 2021. To combat the pandemic, global efforts to produce COVID-19 vaccines have gained momentum and generated more than 240 vaccine candidates in clinical or preclinical stages, updated to February 2021 (2). The United States (US) Food and Drug Administration (FDA) issued Emergency Use Authorization for Pfizer-BioNTech (BNT162b2) and Moderna (mRNA-1273) COVID-19 vaccines separately on December 11th and 18th, 2020. The head-to-head battle between human beings and COVID-19 has entered a new stage.

Other than traditional whole-pathogen vaccines, Pfizer-BioNTech and Moderna adopted new generation technology and commercialized messenger RNA (mRNA) vaccines. Such vector-based vaccines only incorporate specific antigens from the pathogen, namely the spike protein fragment of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Theoretically, vector-based vaccines process better safety profiles (3); both BNT162b2 and mRNA-1273 phase 3 randomized controlled trials (RCTs) have demonstrated the low incidence (0.5–0.6%) of serious adverse events (AEs) in mRNA vaccine groups (4, 5). Most of the AEs reported by the mRNA vaccine recipients were self-limited local and systemic reactions, such as injection site pain (71–84%), fatigue (33–51%), and headache (37–39%) (4, 5).

The past months have witnessed the first large-scale application of mRNA vaccines in the human world. Concerns still exist since the occasional reports of severe or concerning AEs like anaphylaxis (6, 7), virus infection in patients with immune diseases (8), facial nerve palsy (9, 10), and even death (9). As a complement to clinical trials, real-world AE monitoring is crucial to expand our understandings of suspected AEs following immunization (AEFIs). Until December 2020, more than 1.8 million doses of the COVID-19 mRNA vaccine had been administered in the US, and around 0.2% of AE reports had been submitted to the Vaccine Adverse Event Reporting System (VAERS) (7); the reported cases were fairly sufficient to conduct certain informative analyses. We aimed to evaluate the AEs following the mRNA vaccines and further analyzed the potential associations between the clinical backgrounds and some certain AEFIs, including anaphylactic response, concerning neurological disorders and death.

Our study focused on the AEFIs and completed a comprehensive collection until recently to demonstrate the severity of AEs and possible connection with basic clinical features. We confirmed the favorable safety profiles of mRNA vaccines observed in a series of clinical trials (4, 5, 12).

According to the SOC terms, we organized all AEs reported to VAERS to narrow the heterogeneity in their original descriptions. Our classification signaled that general disorders, dominated by fatigue, pain, and chills, took over almost half of the AEFIs among the reported AE cases. Nervous system disorders represented by headache and dizziness, and some other reactions like nausea, vomiting, rash, pain in extremity, and injection site pain, were also frequent within the people who reported AEs. These findings reassuringly echoed previous clinical studies results, which demonstrated that local reactions were common; Simultaneously, systemic side effects, such as fatigue, myalgia, arthralgia, and headache, were noted in about 50% of participants (4, 5).

Severe allergic reaction, including anaphylaxis, is a crucial issue related to a newly licensed vaccine. Anaphylaxis could be life-threatening, typically onsets within minutes to hours after vaccination (13). We found 157 VAERS reports of possible anaphylactic or anaphylactoid reactions during our analytic period; however, the definitively confirmed diagnoses could be fewer. Unlike some common symptoms such as fatigue and headache, the diagnosis of anaphylaxis requires professional judgment, such as case-by-case interpretation with Brighton Collaboration case definition criteria (14). In an early safety monitoring of the BNT162b2 vaccine, 21 cases were determined to be true anaphylaxis among 175 anaphylactoid reaction reports screened in the VAERS database; the anaphylaxis rate was as low as 11.1 cases per million vaccine dose administered (7). Similar to the previous studies (6, 7), we found COVID-19 vaccine recipients with immune response were more likely with a history of allergies. It is vital to screen contraindications, prepare necessary supplies in the vaccination site, and implement adequate postvaccination observation to prevent severe allergies.

Previous studies noted some rarer neurological findings, such as facial paralysis (4, 5, 9, 10). The BNT162b2 and mRNA-1273 phase 3 trials reported seven cases with suspected facial paralysis in 36,930 vaccine receivers (incidence rate 0.019%) (4, 5). However, in the scenario of rare and potential AE issues, clinical trials lack enough power to draw a definitive conclusion due to their strict inclusion criteria, limited sample sizes, and relatively short observation periods. The SRS could be a supplementary source for new evidence. Taking advantage of VAERS, we identified 18 cases with facial paralysis, together with other concerning neurological events like syncope (37 cases) and seizure (12 cases). Such concerning neurological AEs after vaccines may be more than chance events, and the possibility and causality bear close monitoring. A VAERS analysis previously reported the possible relationship between facial paralysis (0.05% incidence rate) and influenza vaccine (15). The female prevalence in all-cause facial paralysis was 58.2–65.8% in some previous studies (16–18); however, the 18 AEFI cases with facial paralysis were notably all females in our study. We further analyzed that neurological AEs were more likely to happen in vaccine recipients with emotional distress histories. Some neurological diseases and emotional disorders share mutual risk factors. A case-control study hinted that psychological stress could exacerbate facial paralysis (19), and another retrospective cohort demonstrated a bidirectional association between facial paralysis and anxiety disorders (20). As a public health emergency, the COVID-19 outbreak psychologically impacted a massive population, resulting in moderate-to-severe anxiety in about one-third of respondents in an online survey (21). If there were possible links between vaccination and some neurological AEs, psychological support should be emphasized, especially for patients with histories of emotional disorders.

Within the current VAERS database, the hospitalization rate (2.53%) and cases that resulted in life-threatening AEs (0.97%) were minor among all 3,908 COVID-19 mRNA vaccine recipients. We could not calculate the accurate recovery rate of AEFI cases due to the un-reporting records in the database, and the results labeled unrecovered were possibly not final due to the relatively short analytic period. Overall, the prognosis of AEFIs after the COVID-19 vaccines could be favorable based on some previous studies demonstrating the complete recovery for all anaphylaxis cases (6, 7). The BNT162b2 and mRNA-1273 phase 3 trials separately noted two participants died in vaccine arms without clear causality with vaccines (4, 5). We identified 11 cases (0.28%) that reported death in the VAERS database. Still, we could not set up causality after reviewing the description case by case. However, we found that those who died after the vaccine were significantly older (79.36 ± 10.41-year-old vs. 42.64 ± 12.55-year-old, P <0.01) and had more underlying diseases compared to other vaccine recipients. Since previous clinical trials included limited old participants with relatively more risk factors (4, 5), a more desired vaccine dosing and the procedure is needed for the elderly.

It was demonstrated that systemic reactogenicity was more common and severe after receiving the second dose than the first (4). However, most AEs in our observations occurred after receiving the first dose since most vaccine recipients completed only the first dose of vaccine during our analytic period. A convincing real-world comparison of reaction of the first and second vaccine doses could be available when the reported cases grow in the future VAERS database.

We admit some limitations to our study. First, unlike researchers in clinical trials who report AEs using standardized data collection methods, the quality of information submitted by VAERS reporters varies widely. The organization of SOC terms merits, but it does not work for some vague descriptions. Second, as a passive SRS, VAERS is inherited with reporting biases, over-reporting or under-reporting, and lack of an unvaccinated comparison group (11). Although it is less likely to under-report a concerning AE, some mild or moderate AEs may actually be more frequent than what we assessed in the database. Third, VAERS does not provide denominator data to calculate the incidence rates of AEs (22); therefore, we could not calculate accurate incidence rates by interpreting the VAERS database. Furthermore, the quality and completeness of the reports collected in VAERS are variable, and some records lack valid medical diagnoses, thus making the assessment of causality challenging. Although we made an effort to demonstrate the possible connection between certain AEFIs and clinical backgrounds, the causal relationship could not be confirmed.

We confirmed the AEFIs associated with the COVID-19 mRNA vaccines were generally mild treatable local or systemic reactions based on the VAERS real-world data. We suggest that the vaccine recipients with past allergy histories should be observed closely to prevent severe allergies, and patients with a prior history of anxiety need to be evaluated before vaccination and monitored for neurological symptoms after vaccination. Moreover, elderly persons with multiple underlying diseases need a repeated evaluation and a more desired vaccination procedure. Our findings pave the way for continued AEFI investigation. We recommend assessing the detailed profiles of vaccine recipients who developed severe AEFIs to explore the potential causal relationship between clinical characteristics and symptoms after the vaccine.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

Ethical review and approval were not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

BZ and GC designed the study. GC, XiL, MY, and MS analyzed and interpreted data. GC generated figures/tables and drafted the manuscript. YZ and XuL reviewed and corrected the manuscript. All authors contributed to the article and approved the submitted version.

This work has been made possible through an ISN Sister Renal Centers Grant.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.