Original article
Self-Reported Reactogenicity After Different COVID-19 Vaccination Regimens
An analysis of registry-based data
; ; ; ; ; ; ; ; ;
Background: In Rhineland–Palatinate, most COVID-19 vaccinations are centrally registered by the Rhineland–Palatinate Division of Vaccine Documentation, which includes self-reported vaccination reactions (SRVR) and their level of perceived intensity. We analyzed the occurrence of SRVR reported between 12/2020 and 12/2021 in relation to the different vaccination regimens involving BioNTech/Pfizer (BNT) and Moderna (m1273) mRNA vaccines and AstraZeneca (ChAd) and Johnson & Johnson (Ad26) viral vector vaccines.
Methods: Using sex-specific logistic regression models, we analyzed the occurrence of all local and systemic SRVR, as well as the occurrence of local and systemic SRVR that were self-rated as “severe” by the vaccinated persons, in relation to the vaccine of the first vaccination and the vaccination regimen of the second vaccination (BNT/BNT, ChAd/ChAd, m1273/m1273, ChAd/BNT, ChAd/m1273). Vaccination with BNT or the BNT/BNT regimen formed the reference category for the estimated odds ratios (OR) with respective 95% confidence intervals.
Results: Of all those vaccinated, 40.7% provided valid information on SRVR after the first vaccination and 33.8% after the second vaccination. As a result, 887 052 individuals were included in the analyses. Their median age was 60 years, and 58% were women. The most common vaccination regimen was BNT/BNT (67.3%). The most common SRVR were pain at the injection site and fatigue. Self-reported reactogenicity after the first vaccination was lowest for BNT. Self-reported systemic reactogenicity was notably higher after vaccination with a vector vaccine. After the second vaccination, self-reported reactogenicity was lowest after a ChAd/ChAd regimen and highest after an m1273 second vaccination.
Conclusion: With overall acceptable tolerability, differences in self-reported reactogenicity were evident depending on the particular COVID-19 vaccines and vaccination regimens in question.
Vaccine development was an early priority to mitigate the public health effects of the COVID-19 pandemic (1). Vaccinated persons have a lower risk of hospitalization and death, and are less likely to suffer from severe illness (2). As of 27 July 2022, about 76% of the 4.1 million residents of the German federal state Rhineland–Palatinate have been fully vaccinated, and 63% have received an additional booster dose (3). The European Commission has so far granted conditional marketing authorizations for five COVID-19 vaccines: the mRNA vaccines BNT162b2 (BNT; Comirnaty, BioNTech/Pfizer) and mRNA-1273 (m1273; Spikevax, Moderna), the vector vaccines ChAdOx1 nCoV-19 (ChAd; Vaxzevria, Oxford/AstraZeneca) and Ad26.COV2-S (Ad26; Janssen, Johnson & Johnson), and, since the beginning of 2022, the protein-based vaccine NVX-CoV2373 (NVX; Nuvaxovid, Novavax) (1).
After the approval of the vaccines, the recommendations for vaccination were changed several times. Along with intermittent supply shortages of certain vaccines, this led to the administration of heterologous COVID-19 vaccination regimens.
In general, the approved vaccines are safe (4, 5). Regarding systemic and local adverse events following heterologous vaccination regimens, recent studies report tolerable reactogenicity (4, 6, 7). However, no consensus has been reached regarding the relative tolerability of these regimens compared with different homologous regimens (4).
Real-world data could yield evidence on the reactogenicity of different homologous and heterologous vaccination regimens using the vaccines authorized in Europe. Based on the data of a statewide vaccination campaign, this study compared the self-reported reactogenicity of various COVID-19 vaccines and vaccination regimens.
We investigated:
- 1. The occurrence of systemic and local self-reported vaccination reactions (SRVR) in general.
- 2. The occurrence of systemic and local SRVR classified by vaccinated persons as “severe,” in relation both to the first vaccine (BNT, ChAd, m1273, Ad26) and to different vaccination regimens after the second vaccination (BNT/BNT, ChAd/ChAd, m1273/m1273, ChAd/BNT, ChAd/m1273).
Material and methods
Our analysis is based on comprehensive data covering a period of 1 year (December 2020 to December 2021) from the statewide COVID-19 vaccination campaign of the German federal state Rhineland–Palatinate. The Division of Vaccine Documentation of the Rhineland–Palatinate Cancer Registry is responsible for recording all individual vaccinations carried out by the public health service in Rhineland–Palatinate. In the period considered, a total of 3 210 638 first and second vaccinations were registered. Together with the vaccination status, vaccination date, and vaccine, person-related characteristics, relevant previous illnesses, and SRVR were documented.
Reporting of SRVR was voluntary, involved no payment, and was based on a SRVR card (eFigure). The vaccinated persons could either report one or more SRVR, or state that they had not experienced any SRVR after vaccination. For each reported SRVR, the vaccinated persons could specify the perceived intensity on a scale from “1 – mild” to “5 – severe”. Some SRVR could be marked in a list provided on the card. Other SRVR that were frequently added as free text were subsequently coded, and the remaining SRVR were grouped under the category “other.” The vaccinated persons were asked to bring the SRVR card completed after the first vaccination with them to the second appointment or to mail it free of charge after the first or second vaccination. Alternatively, SRVR could be reported via an online form.
Statistical analysis
Characteristics of the analyzed population
The results of plausibility checks and the non-responder analysis are reported, along with descriptive characteristics of the persons included for analysis. Absolute number and percentage are stated for categorical variables, median and interquartile range for metric variables.
Reactogenicity of different vaccination profiles
Regarding the first research objective, four binary logistic regression analyses were performed to investigate the occurrence of systemic and local SRVR after first and second vaccination. Regarding the second research objective, four binary logistic regression analyses were performed to investigate the occurrence of systemic and local SRVR classified by the vaccinated persons as “severe” on the five-point scale after first and second vaccination. In both cases, the independent variable of interest was firstly the vaccine used for the first vaccination (four categories: BNT, ChAd, Ad26, m1273) and secondly the vaccination regimen for the first and second vaccinations (five categories: BNT/BNT, ChAd/ChAd, m1273/m1273, ChAd/BNT, ChAd/m1273). All analyses were performed separately for men and women, to investigate possible differences, and were adjusted for age as a continuous variable. Because most persons received a BNT/BNT regimen, vaccination with BNT and the BNT/BNT regimen were the reference category after first and second vaccination, respectively. Odds ratios (OR) and 95% confidence intervals (95% CI) were calculated based on those persons who provided valid SRVR information after the respective vaccination.
Sensitivity analysis 1: SRVR without date of onset
Reports of local or systemic SRVR that did not include a date of onset were excluded. However, because these reports could be allocated to either the first or the second vaccination, they were included in a sensitivity analysis.
Sensitivity analysis 2: SRVR with date of onset later than 8 days after vaccination
To ensure a plausible time window between vaccination and reported SRVR onset, we excluded SRVR for which the period from vaccination date to date of symptom onset was stated as more than 8 days (8, 9). The SRVR thus excluded were included in a further sensitivity analysis. All analyses were performed using the statistical environment R (10).
Results
Characteristics of the study population
Of the 954 238 vaccinated persons that provided valid SRVR information at least once, 7.0% were excluded from the analysis—mostly due to missing or invalid information (Figure 1). SRVR with a reported onset before vaccination, in the future, or more than 8 days after vaccination were excluded. The dataset that forms the basis of this analysis contains information on 887 052 persons. Of all vaccinated persons registered, 40.7% provided valid SRVR information after the first and 33.8% after the second vaccination (Table, eTable 2). Of all persons that provided valid SRVR information after the first vaccination, 54.4% reported having experienced SRVR and 4.9% classified SRVR as “severe” (Table). Of all persons that provided valid SRVR information after the second vaccination, 51.6% reported SRVR and 5.1% classified SRVR as “severe.” After first vaccination and after second vaccination, persons who reported SRVR were younger and a higher proportion of them were female than those who stated that no event had occurred (eTable 2). The most frequently documented local SRVR was pain at the injection site (42.6% after first vaccination, 38.7% after second vaccination), while the most frequently occurring systemic SRVR were fatigue (24.5%, 27.4%) and headache (17.5%, 19.0%). The SRVR most frequently reported as “severe” were chills (15.1% of all reports of chills) and aching limbs (14.5%) after the first vaccination, and generalized weakness (13.2%) and vomiting (14.9%) after the second vaccination.
Excluded persons and non-responders
Persons excluded because of implausible or missing information about SRVR differed slightly from the persons included in the analysis, but no major differences were observed with regard to the vaccination regimens (eTable 1). The analyzed population was on average older than the non-responders and included a higher proportion of women (eTable 2). Furthermore, there were marked differences in response rates between the vaccine and vaccination regimen groups (Table).
Reactogenicity of different vaccination profiles
First vaccination
Self-reported reactogenicity was lower after vaccination with BNT than after other vaccines (Figure 2a). Self-reported systemic reactogenicity was particularly high for the vector vaccines. Women reported higher reactogenicity after vaccination with ChAd than men. There was an increased probability of systemic SRVR classified as “severe” among persons vaccinated with a vector vaccine (Figure 2b).
Second vaccination
The probability of SRVR was higher for the ChAd/BNT, ChAd/m1273, and m1273/m1273 vaccination regimens than for the homologous BNT regimen (Figure 2c). Self-reported reactogenicity was particularly high after a second vaccination with m1273. In contrast, the probability of SRVR was lower after a homologous ChAd regimen than after a homologous BNT regimen. Accordingly, the odds of SRVR classified as “severe” by the vaccinated persons were lower after a homologous ChAd regimen than after a homologous BNT regimen (Figure 2d). Persons who received a second vaccination with m1273 had higher odds of reporting SRVR that they classified as “severe.”
Sensitivity analyses
The majority (65.5%) of those excluded from analysis had to be ruled out because there was no onset date for any of their SRVR (Figure 1). Inclusion of these SRVR did not alter the results (data not shown). Likewise, the effects described were not changed by inclusion of the SRVR that occurred later than 8 days after vaccination (data not shown).
Discussion
This study used comprehensive registry-based data to compare the self-reported reactogenicity of different COVID-19 vaccines and vaccination regimens. Around 5% of the persons included reported SRVR that they rated as “severe” on a five-point scale. This corresponds to the previously reported tolerable reactogenicity of the licensed COVID-19 vaccines (4, 6, 11).
Persons who reported having experienced SRVR were generally younger and included a higher proportion of women than those who reported not having experienced any SRVR. These results are consistent with previous research (8, 11, 12, 13). An altered inflammatory response has been suggested as a possible reason for reduced reactogenicity with increasing age (14). Possible explanations for higher rates of reactogenicity in women include a stronger immune response (11, 13, 15), more pronounced self-awareness, and a higher compliance in reporting the occurrence of vaccination reactions. After the first vaccination, we observed elevated self-reported systemic reactogenicity for the vector vaccines. This finding is in line with the results of other authors (16, 17, 18). After the second vaccination, self-reported reactogenicity was higher for the heterologous regimens than for the BNT/BNT and ChAd/ChAd regimens. These results are also consistent with previous research (6, 7, 12, 16, 19) but contradict some studies reporting comparable reactogenicity profiles for homologous and heterologous regimens (17, 20). Comparatively high reactogenicity after ChAd as first vaccine but comparatively low reactogenicity after homologous second vaccination with ChAd has been repeatedly shown (12, 16, 17, 18, 20).
Based on our results, it can also be assumed that a homologous m1273 regimen and a heterologous ChAd/m1273 regimen have similar reactogenicity profiles. Compared with persons on other regimens, persons with these vaccination profiles had a higher probability of SRVR in general and higher odds of SRVR that they classified as “severe.” These results support those described in other publications (6, 11) but contradict any general statement that heterologous COVID-19 vaccination regimens have higher reactogenicity than homologous regimens. Stuart et al. (2022) point to the higher mRNA dosage as a possible reason for the elevated reactogenicity of the m1273 vaccine (6). Moreover, it can be assumed that a substantial proportion of SRVR can be attributed to nocebo effects and thus to fear and negative expectations with relation to vaccination (21). The frequent changes in vaccination recommendations—leading to heterologous vaccination schedules—may have caused feelings of insecurity within the population. Furthermore, the media coverage of adverse events after COVID-19 vaccination can be assumed to have influenced the reported reactogenicity in our study, prominent examples being the coverage of rare thromboembolic events after vaccination with ChAd (22) and rare cases of myocarditis and pericarditis in young persons after administration of an mRNA vaccine (23).
The monitoring and investigation of vaccine reactogenicity remains important, especially regarding third or fourth doses of vaccine, which often follow a heterologous regimen. Furthermore, a two-dose primary course of COVID-19 immunization is of interest, particularly in countries with poor vaccine supply (24).
Limitations
Our study has several limitations. All data on SRVR were provided on a voluntary basis and could not be medically verified. The data were checked for plausibility in order to include only SRVR that were stated by the vaccinated persons to have occurred within the first 8 days after vaccination. However, the SRVR cannot necessarily be causally attributed to vaccination.
The vaccinated persons rated the intensity of their SRVR on a subjective scale with no concrete definition of “1 – mild” or “5 – severe.” The grading is therefore not equivalent to a clinician’s diagnosis, nor is a SRVR rated as “severe” comparable to the “severe adverse events” in clinical vaccination trials.
Of all vaccinated persons, 40.7% provided valid SRVR information after the first and 33.8% after the second vaccination. Moreover, the vaccine and vaccination regimen groups differ in terms of the response rates. Therefore, bias due to self-selection can be assumed. On the one hand, the absence of a vaccination reaction may lower the motivation to make a report. On the other hand, a very severe reaction or a simultaneously occurring illness may lead to non-reporting if they seriously restrict the person concerned. Nevertheless, we assume that our study overestimates the “true” reactogenicity. It seems plausible that experienced reactions are reported more frequently than the absence of reactions and that reactions perceived as “severe” are reported more frequently than “mild” ones. A systematic review by Xing et al. (2021) showed that the proportion of persons who experienced a vaccination reaction in clinical trials was usually less than 30% (25). In the population we investigated, however, approximately 53% of persons who provided valid SRVR information reported a SRVR.
Another limitation lies in the design of the SRVR card. Because the vaccination reactions that were listed on the card were specified most frequently, over-reporting seems plausible, while it can be assumed that other events are subject to under-reporting. The free-text-based “other” SRVR were analyzed not individually but as a combined category. Therefore, very rare and serious events (e.g., myocarditis) were not considered separately.
Besides the vaccination regimens included here, further combinations of first and second vaccine have been administered (e.g., Ad26/mRNA). These regimens could not be considered, however, either because they have not yet been recommended or because the case numbers were too low. This is a field where future studies can generate more evidence.
Strengths
This is the first study in Germany to use comprehensive real-world data to compare the self-reported reactogenicity of licensed COVID-19 vaccines. The threshold for communication of SRVR was low and left room for subjective evaluation by the vaccinated persons. The findings yield insight into the response behavior of a population when using SRVR cards as a public health instrument. Frequently reported SRVR were systematically recorded, enabling us to analyze local and systemic SRVR separately. All analyses were controlled for age and performed for men and women separately to allow detection of any relevant differences. Two sensitivity analyses were conducted to test the validity of our inclusion criteria yielded results consistent with those of the primary analysis.
Conclusion
Differences were evident in self-reported reactogenicity after first and second vaccination with the COVID-19 vaccines approved in Europe. We confirmed greater self-reported reactogenicity for heterologous vaccination regimens than for the homologous BNT/BNT and ChAd/ChAd regimens. Overall, most persons reported they experienced only “mild” to “moderate” SRVR after vaccination or none at all, confirming the tolerability of the vaccines licensed to date.
Conflict of interest statement
The authors declare that no conflict of interest exists.
Manuscript received on 27 April 2022, revised version accepted on
29 July 2022.
Corresponding author
Dr. rer. physiol. Emilio A. L. Gianicolo
Institut für Medizinische Biometrie, Epidemiologie und Informatik
Universitätsmedizin der Johannes Gutenberg-Universität Mainz
Langenbeckstr. 1, 55131 Mainz, Germany
emilio.gianicolo@uni-mainz.de
Cite this as:
Pfrommer LR, Schoeps M, Blettner M, Wollschläger D, Herm-Stapelberg NH,
Mittnacht L, Kachel P, Jahn K, von Loewenich FD, Gianicolo EAL:
Self-reported reactogenicity after different COVID-19 vaccination regimens—
an analysis of registry-based data. Dtsch Arztebl Int 2022; 119: 727–34.
DOI: 10.3238/arztebl.m2022.0298
►Supplementary material
eTables, eFigure:
www.aerzteblatt-international.de/m2022.0298
Division of Vaccine Documentation, Cancer Registry Rhineland-Palatinate, Mainz: Melissa Schoeps, MsE, Dr. Nils Herm-Stapelberg, Dr. Lukas Mittnacht, Philipp Kachel
Rhineland–Palatinate Ministry of Science and Health, Mainz: Dr. Klaus Jahn
Institute for Virology, University Medical Center of the Johannes Gutenberg University Mainz: PD Dr. med. Friederike D. von Loewenich
Institute of Clinical Physiology of the Italian National Research Council (IFC-CNR), Lecce, Italy: Dr. rer. physiol. Emilio A. L. Gianicolo
| 1. | European Commission: Safe COVID-19 vaccines for Europeans 2022. www.ec.europa.eu/info/live-work-travel-eu/coronavirus-response/safe-covid-19-vaccines-europeans_en (last accessed on 15 April 2022). |
| 2. | World Health Organisation (WHO): Coronavirus disease (COVID-19): Vaccines. 2022. www.who.int/news-room/questions-and-answers/item/coronavirus-disease-(covid-19)-vaccines (last accessed on 12 July 2022). |
| 3. | Bundesministerium für Gesundheit: Wie ist der Fortschritt der COVID-19-Impfung? Aktueller Impfstatus 2022. www.impfdashboard.de/ (last accessed on 28 July 2022). |
| 4. | Nguyen TT, Quach THT, Tran TM, et al.: Reactogenicity and immunogenicity of heterologous prime-boost immunization with COVID-19 vaccine. Biomed Pharmacother 2022; 147: 112650 CrossRef MEDLINE PubMed Central |
| 5. | European Medicines Agency: Safety of COVID-19 vaccines 2022. www.ema.europa.eu/en/human-regulatory/overview/public-health-threats/coronavirus-disease-covid-19/treatments-vaccines/vaccines-covid-19/safety-covid-19-vaccines (last accessed on 15 April 2022). |
| 6. | Stuart ASV, Shaw RH, Liu X, et al.: Immunogenicity, safety, and reactogenicity of heterologous COVID-19 primary vaccination incorporating mRNA, viral-vector, and protein-adjuvant vaccines in the UK (Com-COV2): a single-blind, randomised, phase 2, non-inferiority trial. Lancet 2022; 399: 36–49 CrossRef |
| 7. | Sablerolles RSG, Rietdijk WJR, Goorhuis A, et al.: Immunogenicity and reactogenicity of vaccine boosters after Ad26.COV2.S Priming. N Engl J Med 2022; 386: 951–63 CrossRef MEDLINE PubMed Central |
| 8. | Menni C, Klaser K, May A, et al.: Vaccine side-effects and SARS-CoV-2 infection after vaccination in users of the COVID Symptom Study app in the UK: a prospective observational study. Lancet Infect Dis 2021; 21: 939–49 CrossRef |
| 9. | Tozzi AE, Asturias EJ, Balakrishnan MR, Halsey NA, Law B, Zuber PLF: Assessment of causality of individual adverse events following immunization (AEFI): a WHO tool for global use. Vaccine 2013; 31: 5041–46 CrossRef MEDLINE |
| 10. | R Core Team: R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, 2021. |
| 11. | Rosenblum HG, Gee J, Liu R, et al.: Safety of mRNA vaccines administered during the initial 6 months of the US COVID-19 vaccination programme: an observational study of reports to the Vaccine Adverse Event Reporting System and v-safe. Lancet Infect Dis 2022; 22: 802–12 CrossRef |
| 12. | Powell AA, Power L, Westrop S, et al.: Real-world data shows increased reactogenicity in adults after heterologous compared to homologous prime-boost COVID-19 vaccination, March-June 2021, England. Euro Surveill 2021; 26: 2100634 CrossRef MEDLINE PubMed Central |
| 13. | Borobia AM, Carcas AJ, Pérez-Olmeda M, et al.: Immunogenicity and reactogenicity of BNT162b2 booster in ChAdOx1-S-primed participants (CombiVacS): a multicentre, open-label, randomised, controlled, phase 2 trial. Lancet 2021; 398: 121–30 CrossRef |
| 14. | Andrew MK, McElhaney JE: Age and frailty in COVID-19 vaccine development. Lancet 2020; 396: 1942–4 CrossRef |
| 15. | McCartney PR: Sex-based vaccine response in the context of COVID-19. J Obstet Gynecol Neonatal Nurs 2020; 49: 405–8 CrossRef MEDLINE PubMed Central |
| 16. | Shaw RH, Stuart A, Greenland M, Liu X, Nguyen Van-Tam JS, Snape MD: Heterologous prime-boost COVID-19 vaccination: initial reactogenicity data. Lancet 2021; 397: 2043–46 CrossRef |
| 17. | Schmidt T, Klemis V, Schub D, et al.: Immunogenicity and reactogenicity of heterologous ChAdOx1 nCoV-19/mRNA vaccination. Nat Med 2021; 27: 1530–5 CrossRef MEDLINE PubMed Central |
| 18. | Benning L, Töllner M, Hidmark A, et al.: Heterologous ChAdOx1 nCoV-19/BNT162b2 prime-boost vaccination induces strong humoral responses among health care workers. Vaccines (Basel) 2021; 9: 857 CrossRef MEDLINE PubMed Central |
| 19. | Normark J, Vikström L, Gwon Y-D, et al.: Heterologous ChAdOx1 nCoV-19 and mRNA-1273 Vaccination. N Engl J Med 2021; 385: 1049–51 CrossRef MEDLINE PubMed Central |
| 20. | Hillus D, Schwarz T, Tober-Lau P, et al.: Safety, reactogenicity, and immunogenicity of homologous and heterologous prime-boost immunisation with ChAdOx1 nCoV-19 and BNT162b2: a prospective cohort study. Lancet Respir Med 2021; 9: 1255–65 CrossRef |
| 21. | Amanzio M, Mitsikostas DD, Giovannelli F, Bartoli M, Cipriani GE, Brown WA: Adverse events of active and placebo groups in SARS-CoV-2 vaccine randomized trials: a systematic review. Lancet Reg Health Eur 2022; 12: 100253 CrossRef MEDLINE PubMed Central |
| 22. | Ständige Impfkomission: Pressemitteilung der STIKO zum AstraZeneca-Impfstoff (30.03.2021). 2021. www.rki.de/DE/Content/Kommissionen/STIKO/Empfehlungen/AstraZeneca-Impfstoff-2021-03-30.html (last accessed on 15 April 2022). |
| 23. | Ständige Impfkomission: Pressemitteilung der STIKO zur COVID-19-Impfung mit mRNA-Impfstoff bei Personen unter 30 Jahren (10.11.2021) 2021. www.rki.de/DE/Content/Kommissionen/STIKO/Empfehlungen/PM_2021-11-10.html (last accessed on 15 April 2022). |
| 24. | Our World in Data: Coronavirus (COVID-19) Vaccinations, 2022. www.ourworldindata.org/covid-vaccinations (last accessed on 11 April 2022). |
| 25. | Xing K, Tu X-Y, Liu M, et al.: Efficacy and safety of COVID-19 vaccines: a systematic review. CJCP 2021; 23: 221–8 CrossRef MEDLINE PubMed Central |
