Spontaneously reported adverse events following COVID-19 basic and booster immunizations in the Netherlands

Abstract

Introduction

The rapid roll-out of novel COVID-19 vaccines made near real-time post-marketing safety surveillance essential to identify rare and long-term adverse events following immunization (AEFIs). In light of the ongoing booster vaccination campaigns, it is key to monitor changes in observed safety patterns post-vaccination. The effect of sequential COVID-19 vaccinations, as well as heterologous vaccination sequences, on the observed post-vaccination safety pattern, remains largely unknown.

Methods

The primary objective of this study was to describe the profile of spontaneously reported AEFIs following COVID-19 vaccination in the Netherlands, including the primary and booster series. Reports from consumers and healthcare professionals were collected via a COVID-19 vaccine-tailored online reporting form by the National Pharmacovigilance Centre Lareb (Lareb) between 6 January 2021 and 31 August 2022. The data were used to describe the most frequently reported AEFIs per vaccination moment, the consumer experienced burden per AEFI, and differences in AEFIs reported for homologous and heterologous vaccination sequences.

Results

Lareb received 227,884 spontaneous reports over a period of twenty months. Overall, a high degree of similarity in local and systemic AEFIs per vaccination moment was observed, with no apparent change in the number of reports of serious adverse events after multiple COVID-19 vaccinations. No differences in the pattern of reported AEFIs per vaccination sequence was observed.

Conclusion

Spontaneous reported AEFIs demonstrated a similar reporting pattern for homologous and heterologous primary and booster series of COVID-19 vaccination in the Netherlands.

1. Introduction

The global COVID-19 pandemic in 2020 incentivized the rapid development of novel vaccines against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). By December 2020, Pfizer-BioNTech’s Comirnaty® was approved by the European Medicines Agency as the first vaccine against SARS-CoV-2, followed by Moderna’s SpikeVax® and AstraZeneca’s Vaxzevria® in January 2021, Janssen’s Jcovden® in March 2021, Nuvaxovid® by Novavax in December 2021, and Valneva’s COVID-19 Vaccine® in June 2022.

The rapid roll-out of vaccines made near real-time, post-marketing safety surveillance essential to monitor novel safety signals. Ongoing worldwide surveillance is key to update safety information, allowing detection of more rare and/or potential long-term adverse events following immunization (AEFIs) and identifying risk groups for developing AEFIs, which is difficult to estimate in randomized clinical trials with limited follow-up time [1]. The post-marketing surveillance of COVID-19 vaccines led to multiple, new safety signals such as pericarditis and myocarditis [2], vaccine-induced thrombotic thrombocytopenia [3], [4] and heavy menstrual bleeding [5]. In the Netherlands, post-marketing surveillance is performed by the National Pharmacovigilance Centre Lareb (Lareb), using spontaneous reports from consumers, healthcare professionals, and marketing authorisation holders (MAHs) [6].

In light of the ongoing COVID-19 pandemic and discussions on the long-term need of booster vaccines to maintain immunogenicity, it is key to monitor changes in the safety profile of multiple (homologous and/or heterologous) vaccination cycles. Additionally, limited information is available on the regular use of COVID-19 vaccinations in the general population. Safety evaluations of booster doses in clinical trials showed a comparable safety profile to the initial mRNA and Janssen vaccines [7], [8], [9], [10], [11], [12], [13]. The impact of heterologous regimens compared to homologous vaccination series remains more ambiguous [10], [11], [12], [13], [14], [15], [16], [17], [18]. In a prospective, community-based study in the United Kingdom, Menni and colleagues recently showed that local and systemic AEFIs were reported more frequently after the third (i.e. booster) vaccine dose when this was a heterologous sequence when compared to a homologous schedule, [12]. It therefore remains imperative to continuously monitor the safety profile of the COVID-19 vaccines. The aim of this report is to describe the safety profile of the COVID-19 booster vaccination observed in the Dutch population in relation to the initial vaccination series using spontaneous reports received by Lareb during a period of 20 months.

2. Materials and methods

2.1. Vaccination strategy in the Netherlands

A rapid and large-scale vaccination program was rolled out in the Netherlands in January 2021 and included two dose vaccination with Pfizer-BioNTech [19], Moderna [20], or AstraZeneca [21], or a single dose vaccination with Janssen [22], to complete the initial series (Fig. 1 ). High-risk patients and healthcare-professionals were given priority for vaccination, followed by fixed age-groups, starting with the oldest population first. Patients with a severe immune system disorder were invited to receive an additional third vaccination with Pfizer-BioNTech or Moderna from October 2021 onwards. Within seven months, 84.4 % of the Dutch population older than twelve years was vaccinated with the initial vaccination series [23]. In November 2021, a booster campaign with existing dose Pfizer-BioNTech and Janssen and half-dose Moderna was initiated (vaccination with AstraZeneca was discontinued in the Netherlands for shelf-life reasons). From the second half of December 2021 onwards, COVID-19 vaccination became available for children from five to eleven years of age. In February 2022 onwards, the campaign for a second booster vaccine was initiated.

Fig. 1.

Vaccination strategy in the Netherlands from January 2021 to September 2022. The vaccination campaign commenced on 6 January 2021 with two immunizations with Pfizer-BioNTech, Moderna, AstraZeneca, Nuvaxovid (from December 2021) or one immunization with Janssen (or Pfizer-BioNTech, Moderna, or AstraZeneca in case of COVID-19 in the medical history). For immunocompromised patients, a third vaccination was available following the initial series. The booster vaccination campaign commenced in November 2021, followed by a second booster available as of February 2022. In the Netherlands, these were both performed with same dose Pfizer-BioNTech or half-dose Moderna.

2.2. Definitions

We refer to ‘initial series’ as the initial two doses of Pfizer-BioNTech, Moderna, or AstraZeneca, or a single dose of Janssen (Fig. 1). In the Netherlands, individuals that had a history of COVID-19 infection within the 6 months prior to vaccination required only a single dose of Pfizer-BioNTech or Moderna to complete the initial series. As of 4 June 2021, any history of COVID-19 infection, regardless of the interval between infection and vaccination, required a single immunization with Pfizer-BioNTech or Moderna only. For immunocompromised patients, the initial series was expanded with a third vaccination with Pfizer-BioNTech or Moderna. For most, however, the third vaccination corresponded with their first booster, consisting of either same-dose Pfizer-BioNTech, Janssen, or Novavax [24], or half-dose Moderna. We cannot distinguish between a third vaccination for immunocompromised patients and the first booster vaccine in the Lareb COVID-19 vaccine reporting form.

In this report, ‘consumer’ refers to the person undergoing COVID-19 vaccination and reporting one or more AEFI. The term solicited adverse event includes all AEFIs prespecified on the form. Initially, these were the most common AEFIs labelled in the Summary of Product Characteristics (SmPC): fever, chills, headache, nausea, myalgia, arthralgia, malaise, and fatigue. From 8 March 2022 on also menstruation disorders (heavy menstrual bleeding, polymenorrhea, menstruation delayed, amenorrhea, intermenstrual bleeding, vaginal haemorrhage, and postmenopausal haemorrhage) and lymphadenopathy were included as solicited AEFI because of the high number of reports [19], [20], [21], [22]. Unsolicited adverse events include all other reported AEFIs, which could be described as free text in the reporting form.

AEFIs are coded as serious when they meet the Council for International Organizations of Medical Sciences criteria (CIOMS) criteria [25].

2.3. Data source

This study is based on spontaneous reports collected by Lareb between 6 January 2021 and 31 August 2022 through individual case safety reports (ICSRs). The new bivalent mRNA-vaccines were introduced as booster vaccines in September 2022 and were not included in this study. Lareb directly receives AEFIs from consumers and healthcare professionals. Reporters must be at least 16 years old, but can report for third persons who are not capable of reporting (for example AEFIs experienced by persons younger than 16 years old). When the consumer is not capable of reporting AEFIs, the consumer can indicate that the ICSR was filled in by somebody else. ICSRs received via MAHs were not included in this report.

2.4. Data collection

A consumer-friendly, COVID-19 vaccine-tailored, online reporting form was designed for the Dutch public and healthcare professionals to collect spontaneously reported AEFIs as previously described [6]. For all AEFIs, additional information was asked including latency time, seriousness according to the CIOMS criteria [25], consumer experienced burden of the reaction [26], additional investigations, treatment, duration time, and outcome. With respect to the consumer, additional information was asked concerning consumer demographics, medical history, COVID-19 history, previous COVID-19 vaccinations, concomitant drugs, and vaccine batch number. The processing and assessment of ICSRs has been described elsewhere [6] and was done following standardized terminology coding using the Medical Dictionary for Regulatory Activities (MedDRA, version 23.1–25.0) and the G-standaard, the drug database used by all healthcare parties in the Netherlands [27]. During the processing of the ICSR (automatically or by the Lareb-assessor), a mandatory field was filled indicating whether the reported AEFI was labelled in the SmPC or not. This information was extracted to determine the percentage of AEFIs reported in the SmPC.

2.5. Analysis of the total number of ICSRs and most frequent reported AEFIs

The total number of ICSRs was calculated as the cumulative sum of all unique ICSRs reported by a consumer and healthcare professionals in relation to COVID-19 vaccination received between 6 January 2021 and 31 August 2022. The total number of administered vaccinations in the Netherlands per timepoint was collected from the National Institute for Public Health and the Environment website [28]. Information was provided up to 15 May 2022. Reporting rates for dose 1 and 2 were determined per vaccine brand administered using CIMS data (the number of administered vaccines specified per brand provided to Lareb by the National Institute for Public Health and the Environment until the 18th of April 2022). The frequency of solicited and unsolicited AEFIs was expressed as the percentage of AEFIs of the same MedDRA coding Preferred Term (PT) with respect to all reported AEFIs in the same time period. The absolute number of PTs reported was also described. For local AEFIs, injection site inflammation was automatically coded when both injection site pain and injection site warmth were reported and was treated as a separate PT. In some ICSRs, previous COVID-19 vaccinations were unintentionally also reported as the suspect drug instead of the past drug therapy. This could lead to double counting of the reported AEFI. In these situations, the AEFI was only assigned to the most recent suspect vaccination. As a result of this data pre-processing step, 192 ICSRs not containing any vaccination date were excluded from the AEFI frequency calculation.

2.6. Analysis of time between vaccinations

For each ICSR, information concerning the date and the type of previous COVID-19 vaccinations was requested. This information was used to calculate the time between the reported dates of the vaccinations per consumer. ICSRs for which the date of vaccination was unknown were excluded. When vaccination dates of previous COVID-19 vaccinations were equal to the vaccination date of the reported suspect COVID-19-vaccine in one ICSR, only one vaccination date was used for the determination of the time between vaccinations. The time interval between the first possible vaccination date in the Netherlands (6 January 2021) and the cut-off point of our study (31 August 2022) was 602 days. All time intervals greater than 602 days were eliminated. In some cases the COVID-19-vaccine was reported as a suspect drug as well as the past drug therapy, when this time interval was zero, the ICSR was excluded.

2.7. Analysis of the consumer experienced burden per ICSR

Reporters were asked to score the burden of each AEFI on a scale from 1 to 5, where 1 stands for “not at all”, 2 stands for “a little”, 3 stands for “quite”, 4 stands for “a lot of”, and 5 stands for “very much”. If a person reported more than one (solicited/unsolicited) AEFI per ICSR, the highest score was used per ICSR. This information was used to estimate the pattern of experienced burden per vaccination.

2.8. Analysis of reported AEFIs per vaccination sequence

The vaccination sequence was established for a maximum of three vaccinations from all ICSRs with three or more registered vaccines with a valid vaccination date. Two vaccines was sufficient to be considered a vaccination sequence if the first vaccination was Janssen. A total of 73 different vaccination sequences could be established in the included dataset. The number of ICSRs of the eight most common vaccination sequences was determined from the Lareb databank. The number of ICSRs received after the eighth vaccination sequence was only one-tenth of the eighth sequence. Because of this large gap, the eight most applied sequences were included and all later sequences were combined as ‘other’. Per vaccination sequence, the frequency of (solicited and unsolicited) AEFIs reported after the first booster was presented, as described above.

A chi-squared tests was performed for the reports on lymphadenopathy for the Pfizer-BioNTech and Moderna boosterdose comparing all possible homologous and heterologous vaccination sequences of Pfizer-BioNTech and Moderna combinations.

2.9. Data analysis and visualization

Plots were generated using GraphPad Prism 9 (version 9.5.0), R (version 4.1.3), and RStudio (version 2022.07.1 + 554).

3. Results

3.1. Reporter characteristics

Between 6 January 2021 and 31 August 2022, more than 35 million vaccines were administered to the Dutch population (Fig. 2 , and non-cumulative in Supplementary Fig. 4). Lareb received 227,884 spontaneous ICSRs related to COVID-19 vaccination (Table 1 ). The total amount of AEFIs reported was 1,096,233, an average of 4.8 AEFIs per ICSR. The reporting rate for dose 1 was 0.95 % and for dose 2 it was 0.62 % (Supplementary Table 2). The majority of the ICSRs were related to the first vaccination (52 %). Overall, the monthly number of ICSRs received decreased over time and per additional vaccination, with a peak from May 2021 to July 2021 and in January 2022. The number of additional vaccinations administered over time showed a similar pattern. The month during which most ICSRs were received was January 2022, possibly as the result of the booster campaign initiated in November 2021 (booster 1). Indeed, the majority of ICSRs received in January 2022 concerned the third vaccination (which for most of the Dutch population was their first booster). The majority of ICSRs were consumer reports concerning females between the ages of 21 and 65 years old. Most reporters had no, or unknown history, of COVID-19 (Table 1). Due to the limited amount of ICSRs (n = 46) concerning the fifth vaccination, these were excluded from further analysis.

Fig. 2.

Cumulative number of spontaneous ICSRs and total number of vaccinations administered over time. Each bar represents the cumulative number of ICSRs received by Lareb over time (left y-axis), each colour represents the cumulative number of ICSRs per vaccine moment. The black line represents the cumulative number of vaccines administered to the Dutch population (right y-axis). ICSR: individual case safety report.

Table 1.

ICSR characteristics.

	Vaccination 1	%	Vaccination 2	%	Vaccination 3	%	Vaccination 4	%	Vaccination 5	%
TOTAL ICSR	118,207	51.87	68,470	30.05	39,364	17.27	1797	0.79	46	0.02
Non-serious	115,333	97.57	66,533	97.17	38,772	98.5	1724	95.94	43	93.48
Serious*	2874	2.43	1937	2.83	592	1.5	73	4.06	3	6.52
TOTAL AEFI	560,636		324,509		201,751		9030		307
Non-serious	554,787	98.96	320,841	98.87	200,720	99.49	8905	98.62	304	99,02
Serious*	5849	1.04	3668	1.13	1031	0.51	125	1.38	3	0,98
Reporter
Consumer patient/ Non-health care professional	113,896	96.35	66,528	97.16	38,850	98.69	1751	97.44	43	93.48
Health care professional	4311	3.65	1942	2.84	514	1.31	46	2.56	3	6.52
Consumer sex
Female	92,034	77.86	54,118	79.04	31,066	78.92	1241	69.06	32	69.57
Male	26,154	22.13	14,348	20.96	8296	21.08	556	30.94	14	30.43
Unknown	19	0.02	4	0.01	2	0.01
Consumer age**
<5***	78	0.07	30	0.04				0.00
5 to 11	70	0.06	12	0.02	1	0.00
12 to 20	5036	4.26	2868	4.19	777	1.97		0.00
21 to 65	98,625	83.43	55,085	80.45	32,552	82.69	741	41.24	18	39.13
66 to 80	7849	6.64	5708	8.34	4275	10.86	879	48.91	22	47.83
>80	1461	1.24	997	1.46	342	0.87	91	5.06	3	6.52
Unknown	5088	4.30	3770	5.51	1417	3.60	86	4.79	3	6.52
Covid-19 history
Yes	16,549	14.00	6551	9.57	3818	9.70	468	26.04	10	21.74
Suspect	8912	7.54	4711	6.88	2206	5.60
No/unknown	92,746	78.46	57,208	83.55	33,340	84.70	1329	73.96	36	78.26

^*When an ICSR contained both non-serious and serious AEFI, the report was considered as serious.

^**When a ICSR contained multiple ages, the youngest age was included.

^***ICSRs concerning children of mothers vaccinated during pregnancy. AEFI: Adverse event following immunization. ICSR: individual case safety report.

3.2. Time between vaccinations

In the Netherlands, the primary vaccination campaign lasted approximately seven months. It was advised to maintain at least four to six weeks between vaccinations of the initial series (including the third vaccination in immunocompromised patients), and to maintain at least three months between the initial series and the booster vaccines or a COVID-19 infection in the medical history. The first booster vaccination became available approximately six to nine months after the primary vaccination round. The reported time between vaccines is shown in Fig. 3 . The median reported time between dose one and two (the initial series), is 35 (range 1–579) days. There is an additional small peak visible after 180 days (Supplementary Fig. 1), likely representing consumers who completed the primary vaccination round with just one dose (due to a COVID-19 infection in the medical history, or primary vaccination with Janssen). As expected, after the primary vaccination round, the time interval between vaccinations is much more diverse between consumers, with a median of 183 (range 1–575) days between vaccination 2 and 3, and 110 (range 1–513) days between vaccination 3 and 4, in line with the vaccination strategy employed.

Fig. 3.

Time interval between successive vaccinations. From top to bottom, the percentage of ICSRs per number of days between consecutive vaccinations is shown. ICSRs in which the date of vaccination was unknown were excluded. For an ICSRs where multiple vaccinations had the same date of administration, only one vaccination was used. The maximum time-interval between vaccination allowed was 602 days, all ICSRs with greater time intervals were excluded. When the time interval was zero, the vaccination was excluded. ICSR: individual case safety report.

3.3. Most commonly reported solicited and unsolicited AEFIs per vaccination moment

We next compared the most commonly reported solicited and unsolicited AEFIs per vaccination moment (Fig. 4 ). With respect to local adverse events (Fig. 4A), later vaccination moments showed a slightly increasing trend in the relative number of spontaneously reported injection site adverse events compared to earlier vaccination moments. Overall, these differences were very small (e.g. injection site erythema increased from 1.84 % after the first vaccination to 2.76 % after the fourth vaccination). With respect to generalized solicited AEFIs (including fever, chills, headache, nausea, myalgia, arthralgia, malaise, fatigue, heavy menstrual bleeding, polymenorrhea, menstruation delayed, amenorrhea, intermenstrual bleeding, vaginal haemorrhage, postmenopausal haemorrhage, and lymphadenopathy), the majority of AEFIs showed a comparable reporting pattern per vaccination moment. Remarkably, a relatively high number of AEFIs regarding lymphadenopathy was reported after vaccination 3 (3.67 % compared to 0.53 % after vaccination 1 and 0.65 % after vaccination 4).

Fig. 4.

Twenty most reported solicited and unsolicited AEFIs per vaccination moment. The most common solicited (A) and unsolicited (B) AEFIs are shown. Solicited adverse events were injection site reactions, fever, chills, headache, nausea, myalgia, arthralgia, malaise, fatigue, and as of 8 March 2022 also heavy menstrual bleeding, polymenorrhea, menstruation delayed, amenorrhea, intermenstrual bleeding, vaginal haemorrhage, postmenopausal haemorrhage, and lymphadenopathy. Unsolicited adverse events were all other reported events. Data is shown as the percentage of the total AEFIs reported for a specific vaccination moment (including all solicited and unsolicited AEFIs). AEFI: adverse event following immunization.

With respect to unsolicited AEFIs, the most commonly reported AEFIs were dizziness, dyspnoea, and diarrhoea. Comparable to the solicited AEFIs, the different vaccination moments showed a comparable reporting pattern, with a slight decrease for some AEFIs (including dizziness and paranesthesia) and an increase for others (including vomiting, cough, hyperhidrosis, and herpes zoster). Heavy and intermenstrual bleeding were not reported after the fourth vaccination, likely because the population that had received the fourth vaccination up to 31 August 2022 included older and vulnerable populations only. Overall, the number of reports received was very low for all AEFIs. Approximately 80 % of AEFIs per vaccination moment were known and listed in the SmPC at time of reporting (Supplementary Fig. 2). This tendency remained over time.

3.4. Reported burden of ICSRs per vaccination moment

The reported burden on the daily life of consumers for solicited and unsolicited AEFIs was comparable over time (Fig. 5 ). For all vaccination moments, the reported burden for the majority of ICSRs were scored as ‘quite’ burdensome. The percentage of serious ICSRs was 2.4 % overall and comparable per vaccination moment (range 1.5 – 4.06 %) (Table 1). The reported serious ICSRs per vaccination moment were also comparable, with pulmonary embolism, dyspnoea, and cerebral infarction being the three most commonly reported serious AEFIs.

Fig. 5.

Percentage of reported ICSRs per burden level. Reporters were asked to score the burden of each AEFI on their daily life. If a person reported more than one (solicited/unsolicited) AEFI per ICSR, the highest score was used. AEFI: Adverse event following immunization. ICSR: individual case safety report.

3.5. Vaccination sequence

In total, Lareb assessed 43,759 ICSRs of people receiving two vaccinations (Janssen + booster) or three vaccinations. Per vaccination sequence, we calculated the percentage of specific AEFIs after the last vaccination of this sequence. Since Pfizer-BioNTech was the most commonly used vaccine in the Netherlands, it is the most frequent vaccine used in homologous as well as in heterologous sequences (Supplementary Table 1). Overall, there was a comparable trend in the reporting pattern of solicited adverse events and we could not identify a clear pattern of reported AEFIs pertaining to a specific vaccination sequence, although lymphadenopathy was reported more often after a Pfizer-BioNTech booster (Supplementary Fig. 3 and Supplementary Fig. 5 for results of the Chi-squared test for the booster sequences).

4. Discussion

Vaccination is one of the most effective tools available against infectious diseases. The rapid introduction of the COVID-19 vaccines and the quick alterations of vaccination strategies required accurate and ongoing safety monitoring of vaccine safety profiles. This overview provides insight into the observed safety profile of the monovalent COVID-19 booster vaccinations in relation to the initial vaccination series in the Netherlands, including homologous and heterologous vaccination schedules, based on spontaneous reporting of AEFIs. Overall, we observed a high level of similarity in the local and systemic AEFIs reported per vaccination moment, with no apparent change in the relative amount and type of serious AEFIs after multiple COVID-19 vaccinations, which were very low for all vaccines. We additionally did not observe any changes in the experienced burden per vaccination moment, nor differences in reported AEFIs per vaccination sequence. A previous cohort event monitoring study has shown there are differences in the frequency of reported AEFI per vaccine brand and per vaccination moment for some brands [29], [30].

Although a prior COVID-19 infection is generally associated with higher post-vaccination reactogenicity [15], contradictory effects of previous COVID-19 vaccinations on the vaccine safety profile have been published [7], [8], [9], [10], [11], [12], [31], [32], [33], [34], [35], [36], [37]. In previous cohort event monitoring studies in the Netherlands, Pfizer-BioNTech was associated with the lowest number of reactogenicity and systemic AEFI’s. Systemic AEFIs were experienced more often after a first dose of AstraZeneca and Janssen, or after the second dose of Moderna [29], [30]. Despite the already normalized heterologous (booster) vaccination campaigns combining mRNA- and vector-based vaccines over time [38], our knowledge on the reactogenicity of heterologous vaccination schedules is still expanding. Heterologous vector/mRNA booster vaccines evaluated in trial settings were reported to induce a broader immune responses [16], [38] but its effect on reactogenicity remains uncertain. Some studies have shown induction of greater systemic reactogenicity in heterologous regimens compared to homologous groups [12], [14], [15], whereas other studies do not observe a significant difference [11], [13], [16], [17], [18]. Of note, safety follow-up time in these studies is generally short and differences among studies, including participant age, sex, previous COVID-19 infection, and immunization intervals, may have contributed to the ambiguous results [29], [30], [39]. In this study, we observed no obvious differences between the AEFI pattern of homologous and heterologous vaccination schedules based on spontaneous reports.

A strength of our study is the use of real-world data to assess AEFIs for different vaccination moments in a diverse population. Spontaneous reporting can be used to identify serious adverse events in real-time that are too rare to be measured effectively in clinical trials, which are limited by shorter follow-up times, inclusion bias, and limited group sizes. Spontaneous reporting additionally allows for further analysis of the course, experienced burden, and contributing factors of reported AEFIs.

We acknowledge several limitations of our study related to the inherent biases of spontaneous reporting [40]. Our study describes the number of reported AEFIs post-vaccination, but does not provide information on absolute incidences of these reported AEFIs among the general population. In addition, spontaneously reported data is reported retrospectively and not collected systematically as in clinical trials. This highlights the complexity of interpreting multifaceted data such as the vaccination timing and sequence in individuals, which was further dependent on vaccination type, previous COVID-19 infection, and medical history. Indeed, we observed great heterogeneity in median time between vaccinations, of which the impact on the safety profile remains largely unknown and likely varies per vaccine type [33], [41], [42], [43].

Spontaneous reporting is additionally subject to shifts in the population undergoing vaccination. For instance, during the initial roll-out of vaccines in the Netherlands, vaccination was highly dependent on age, occupation, and health-status, thus influencing spontaneous reporting of specific demographic groups. As such, younger people were more frequently vaccinated with Janssen or Pfizer-BioNTech, which may explain the relatively higher number of reports after these vaccines and also the higher number of reports concerning menstruation disorders after the Pfizer-BioNTech booster [44]. Similarly, herpes zoster was reported relatively more frequently after the primary AstraZeneca sequence, which was used from the start of the Dutch vaccination campaign and firstly applied to the elderly and immunocompromised.

The extraordinary media attention following first approval of novel vaccines possibly boosted reporting behaviour (also referred to as the Weber effect) [45], [46] and we observed a decrease in reporting after later immunizations. Media attention may have additionally influenced spontaneous reporting of vaccines that were authorized later, such as Janssen, and boosted reporting of newly identified vaccine adverse reactions, including menstrual disorders and thrombosis. Future studies could examine the influence of media attention on the number of submitted spontaneous reports further. It was beyond the scope of this article to discuss the content of reported serious AEFIs, but all serious adverse events were analysed extensively and we would like to refer to the previously published reports in which we show that pulmonary embolism, dyspnoea, and cerebral infarction were the most commonly reported serious AEFIs in line with the observations presented here, and to the overview of safety signals and news on AEFIs after Covid-19-vaccination (Supplementary Table 3) [6]. In future, more rare, long-term, and severe adverse events in relation to multiple vaccinations and homologous versus heterologous vaccination sequences should be examined further. Insight into rare, severe, and long-term AEFIs is important information for vaccine acceptance, which will be particularly relevant if COVID-19 vaccination is administered as a seasonal booster. Finally, a similar safety profile analysis concerning spontaneous reported adverse events of the new bivalent mRNA-vaccines could be performed as these were not included here.

In conclusion, we present an overview of spontaneously reported AEFIs in the Netherlands after one or multiple COVID-19 vaccinations. We observed a similar reporting pattern for the primary and booster series of COVID-19 vaccination and for all vaccination sequences. These findings are highly relevant in light of ongoing COVID-19 vaccinations and discussions on the need of implementing seasonal COVID-19 vaccination campaigns [47]. In light of these discussions, a more complete understanding of the safety profile of COVID-19 booster vaccinations is key. Changes in safety profiles over time can inform vaccination strategies employed by countries, as well as influence the vaccination acceptance in the general population. Therefore, as booster vaccination campaigns continue, it will be key to continue monitoring changes in observed safety patterns through post-marketing surveillance.

5. Funding sources

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

6. Ethical approval and consent

Ethics approval was not needed for this study. No approval or consent to participate was needed for this study.

7. Availability of data and material

The datasets for this article are not publicly available because of the data protection policy of pharmacovigilance centre Lareb. Requests to access the datasets should be directed to the corresponding author and will be granted on reasonable request.

Authors contributions

All authors attest they meet the ICMJE criteria for authorship. The original study protocol was designed by SB, ES, LR, and FH. The query and datasets were established by DS and JS. Data analysis was performed by DS, JS, and SB. The design of the article was determined by SB, ES, LR, and FH. All authors contributed to the final data analyses, to manuscript drafting and revision. All authors approved the final version to be published and agree to be accountable for all aspects of the work.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

The datasets are not publicly available because of the data protection policy of pharmacovigilance centre Lareb. Requests directed to the corresponding author will be granted upon reasonable request.