2. Materials and Methods
2.1. Study Design
We conducted a web-based online survey (Umfrageonline.com (last accessed 10 October 2021, enuvo, Zurich, Switzerland). Participation was possible from 23 February 2021 until 29 March 2021.
2.2. Target Population
Our primary target population were frontline HCWs and therefore we primarily addressed emergency medical services (EMS) personnel through their medical directors. We contacted 270 of 401 medical directors of EMS in cities and counties in Germany. Contact data for the remaining 131 medical directors were unavailable. We asked all medical directors to forward the survey link and a flyer with QR code to the rescue stations. On 31 March 2021, 73,333 employees were subject to social insurance contributions in the emergency medical services sector in Germany [21
Since many medical directors work in both COVID-19 crisis teams and hospitals, it is possible for them to get the survey out to hospitals and nursing facilities as well. Therefore, personnel from hospitals had the opportunity to participate in our survey as well. Often physicians from hospitals also work in EMS. Depending on their qualifications, paramedics and nurses can also work in both hospitals and EMS. In order to consider activity in more than one area, it was possible to select more than one answer option.
In the German EMS system, prehospital emergency care is primarily provided by EMS personnel, supported by an on-scene emergency physician in life-threatening situations. The competence of EMS personnel includes a range of Advanced Life Support (ALS) treatments, which EMS personnel must perform until an emergency physician arrives on the scene. At that point, paramedics (and other medical personnel on scene) act under the direct medical supervision of the physician [22
]. Other (non-life-threatening) emergencies are treated independently by the EMS personnel without a physician being on site. EMS personnel have direct contact with all patient groups and are therefore classified as frontline HCWs [17
Questions referring vaccines, demographic data (sex, federal state), profession, number of contacts with COVID-19 cases, and their own medical history were formulated closed as multiple-choice questions (Q1, 3–11, 24, 25, 27, 28). Answers about age and sick leave were given in free text fields (Q2, 22, 23). In each case for the 1st and 2nd vaccination, the information on the administered vaccine was given. The medical history included information on previous diseases and long-term medication use. They were standardized in an Excel table and checked for plausibility in each case. To assess the extent of side effects and attitudes toward future vaccinations, questions on these topics were answered on visual analog scales (one-point steps from 0 = not applicable to 100 = fully applicable; Q12–21, 26, 29). The following side effects were documented: local, systemic, neurological, allergic, and other. The questionnaire was developed by the core study group (medical directors and experts in public health and statistics) and tested by 5 medical directors in emergency medical services. Proposed changes have been adopted.
2.4. Statistical Methods
We express values of the visual analog scale (0–100) as percentages, e.g., a value of 31 is reported as 31%. We report absolute counts and percentages for categorical data. For continuous data, we report means and standard deviations or medians and interquartile ranges where appropriate. To assess the association between independent variables and adverse effects of vaccination we used log-binomial regression to estimate crude and adjusted (age, sex, and profession) risk ratios (RR) with 95% confidence intervals (95%CI) for the occurrence of sick leave.
We are calculating and reporting confidence intervals to assess the precision of our estimates because our goal is estimation and not significance testing. We wish to avoid publication bias by preferential reporting of significant results. Instead, we judge the value of our estimates by their precision and validity [23
Statistical analyzes were performed using IBM® SPSS® Statistics Version 27.0 (IBM Corporation, Armonk, NY, USA) and SAS 9.4 (Cary, NC, USA).
2.5. Ethical Consideration
Participants were informed at the beginning of the survey that participation was voluntary and anonymous. They were also informed that the results would be analyzed and published. Participation could be interrupted at any time. The study is consistent with the Declaration of Helsinki. The ethics committee of the University of Duisburg-Essen has approved our study (21-9929-BO).
Of the 4620 people who opened the online questionnaire, 711 people thereafter decided not to fill-in the questionnaire and had to be excluded. Another four people were excluded because they did not report the type of vaccination, leaving 3905 people for the analysis of side effects after the first vaccination. Among these, 3657 (94%) also filled in the questionnaire after the second vaccination.
The median ages differed by the following professional groups: physicians and others (41 years), EMS personnel (34 years), and nurses (39 years). Age was not associated with sex (both median age of 38 years, diverse excluded) (Table 1
). BioNTech/Pfizer (76.7%) was the most frequent vaccine, followed by AstraZeneca (15.8%) and Moderna (7.45%) (Table 2
). A total of 48.3%, 51.5%, and 0.2% self-identified as male, female, and diverse, respectively.
The largest group of participants were emergency services personnel (39.4%), followed by nurses (36.5%) and physicians (12.7%). People vaccinated with BioNTech were slightly older than people vaccinated with other vaccines (median age, 38 years vs. 36 years). While EMS was dominated by men (78.7%), the nursing service and others were dominated by women (80.1% and 71.6%, respectively). Among physicians, men predominated only slightly (54.5% vs. 45.5%). While men made up 52% of the participants in BioNTech, this proportion was only 25% in Moderna.
The need for medication after the first and second vaccinations depended on the vaccine. While after the first vaccination with BioNTech, Moderna, and AstraZeneca in 10.0%, 23.0%, and 68.7%, respectively, this need was 37.6%, 70.0%, and 31.6%, respectively, after the second vaccination.
Among the reactions elicited (local, systemic, allergic, neurologic, and other reactions), increased VAS scores were observed essentially only for local and systemic reactions. For the first vaccination, local reactions were strongest for Moderna, while systemic reactions were strongest for AstraZeneca. At the second vaccination, the expression of local reactions for BioNTech and Moderna was nearly the same as at first vaccination, while significantly fewer local reactions were reported for AstraZeneca at second vaccination.
At the first vaccination, systemic reactions were much less for BioNTech and Moderna than for AstraZeneca, while at the second vaccination, systemic reactions were markedly more for BioNTech and Moderna than for AstraZeneca (Table 3
After the first vaccination, the intensity of local and systemic vaccine reactions (VAS: visual analog scale) tended to be lower the older the age was for all types of vaccinations, with the following exception: for the systemic reaction after BioNTech vaccination, there was virtually no relationship between the intensity of systemic reactions and age at vaccination. Moreover, after the second vaccination, local, and systemic reactions tended to be lower the older the person was, with the exception of Moderna (Table 4
shows the number and percentage of sick days after the first and second COVID-19 vaccinations. Here it is shown that the number of sick days increases after the second vaccination.
Overall, 13.4% were unable to work for at least one day after the first vaccination. The percentage of HCWs who were unable to work for at least one day after the first vaccination differed by vaccine as follows: BioNTech 4.4%, Moderna 13.1%, and AstraZeneca 55.0%.
Compared with BioNTech, age-, sex-, and profession-adjusted sick leave after the first vaccination was 2.9 times more likely with Moderna (95%CI: 2.0–4.0) and 11.6 times more likely with AstraZeneca (95%CI: 9.7–14.0). In contrast, after the second vaccination, age-, sex-, and profession-adjusted sick leave was 2.0 times more likely with Moderna (95%CI: 1.8–2.3) and 0.7 times less likely with AstraZeneca (95%CI: 0.6–0.9) (Table 6
Willingness for future vaccination of any kind showed a U-shaped distribution on the VAS scale with stronger occupations on the edges of the scale, resulting in large interquartile ranges. Vaccination propensity differed markedly depending on the COVID-19 vaccine. After the second vaccination, individuals vaccinated with BioNTech or Moderna had markedly stronger vaccination readiness (median and interquartile range [IQR]: BioNTech 89 [47–100], Moderna 92 [47–100]) than individuals vaccinated with AstraZeneca (56, [16–99]).
Based on our results, we can conclude that especially local and systemic reactions occur after both the first and second vaccinations. However, the mRNA vaccines have less intense reactions after the first vaccination than after the second vaccination. In contrast, the local and systemic reactions are less pronounced in the AstraZeneca vector vaccine during the second vaccination. This difference between mRNA and vector vaccine is also reflected in the proportion of HCWs with sick leave for at least one day. After the second vaccination with the mRNA vaccines, sick leave was higher. In contrast, sick leave was lower after the second vaccination with the AstraZeneca vector vaccine compared to the first vaccination.
Vaccination reactions appear to be weaker the higher the age. The following results are also evident in older studies: B-cell responses are lower at older ages. Older people are also no longer able to respond quickly and persistently to new antigens [25
]. However, a recent study shows that vaccine efficacy in subjects over 65 years of age was no lower than in younger subjects [26
]. Otherwise, older persons may have a different awareness of health and disease due to increased life experience as well as experience with diseases. Accordingly, it could also be that the existing symptoms after vaccinations could be subjectively perceived as less.
Overall, local and systemic vaccine side effects predominated in our study and were similar to the results from Germany [27
]. Serious adverse effects could not be detected with certainty in our study because the severity of symptoms was only documented by a visual analog scale and was self-reported. A distinction between the usual immune response (usual local and systemic reactions) and severe complications is therefore not possible. Our results are comparable to those of the Paul Ehrlich Institute. The Paul Ehrlich Institute reports 106,835 suspected cases. According to the Robert Koch Institute, over 74 million vaccinations were given by 30 June 2021, including over 54 million vaccinations with Comirnaty, over 6 million vaccinations with Spikevax, over 11 million vaccinations with Vaxzevria, and over 1 million vaccinations with the Janssen COVID-19 vaccine. 49,735 suspected cases were reported for vaccination with Comirnaty, 14,153 suspected cases were reported for Spikevax, 39,398 suspected cases were reported for Vaxzevria, and 3061 were reported for the COVID-19 vaccine by Janssen. In 488 reported suspected cases, the COVID-19 vaccine was not specified. The reporting rate was 1.4 per 1000 vaccine doses for all vaccines combined and 0.1 per 1000 vaccine doses total for reports of serious reactions [28
Accordingly, in our collective, based on the PEI data, with a participant number of almost 4000 HCWs, a serious complication would have been expected in 0.8 participants (if all participants had received the second vaccination dose).
The CDC identified 66 case reports received by the vaccine adverse event reporting system (VAERS) that met Brighton Collaboration case definition criteria for anaphylaxis (levels 1, 2, or 3): 47 following the Pfizer-BioNTech vaccine, for a reporting rate of 4.7 cases/million doses administered, and 19 following the Moderna vaccine, for a reporting rate of 2.5 cases/million doses administered [29
]. Twenty-one (32%) of the 66 case reports noted a prior episode of anaphylaxis from other exposures, including 18 in intensive care, 7 of whom required endotracheal intubation, and hospitalization ranged from one to three days.
Based on our results, it is reasonable to assume that a booster vaccination, especially with an mRNA vaccine, may lead to renewed service absences and should therefore be considered in the duty scheduling of personnel.
It is well known that in recent years, there has been an overall decrease in vaccination readiness in Europe. A recent European survey on the willingness of the population to be vaccinated against COVID-19 shows disappointing results. In turn, concern about possible side effects increased [30
]. Several studies show a low willingness for seasonal influenza and COVID-19 vaccination among HCW and also EMS personnel. Reasons often given are the following: self-determination, sufficient health status, fear of adverse effects, and concerns about safety and efficacy [8
]. Inadequate vaccination compliance was also reflected among EMS personnel [32
]. An effect of the now available personal protective equipment, information about vaccination reactions, and side effects on the willing is possible.
In the case of the AstraZeneca vaccine, serious but rare complications associated with vaccination have been shown to occur in Germany and the EU. Vaccination was then paused for a short time. However, after further investigations, the vaccination was again approved by the European Medicines Agency [34
]. Our study was conducted at the time of inconsistent reports about AstraZeneca. The uncertainty and bad news, such as side effects after the first vaccination, may have led HCWs to overestimate side effects and be more skeptical of future vaccinations in this group. However, the overall future vaccination willingness is still too low to achieve a safe herd immunity, according to our results. In particular, a recent study shows that vaccine efficiency decreases with decreasing antibody levels after the second COVID-19 vaccination with BioNTech and that a so-called booster vaccination will be necessary [37
]. Thus, these results confirm our first study to survey the vaccination willingness of frontline HCWs before the start of the first vaccination [31
]. Apparently, even medically educated people remain skeptical even though they have already been vaccinated. Therefore, a continuous information campaign will remain necessary, especially for these professional groups. Our results may help to show that mainly local and systemic side effects occurred, reflecting a physiological immune response.
Our study has strengths and limitations. We present one of the very first studies that has evaluated the inability to work in HCWs after COVID-19 vaccinations and can therefore help to better calculate employee absences in future vaccinations. Furthermore, our study used a visual analog scale to assess the intensity of adverse effects on a metric scale. Our sample, compared to similar studies, is larger than average. To our knowledge, this is the first study in the world that also evaluates HCWs’ work absence after COVID-19 vaccination.
Because of the study design, the reported symptoms and absences could not be validated. The questionnaire asked whether the patient had taken any medication after the vaccination. However, it was not considered whether medication was taken prophylactically to suppress any symptoms that might occur. In this case, a bias may occur, although the side effects are likely to have been milder. We also cannot differentiate whether the responses are typical vaccine reactions, adverse drug reactions, or vaccine injuries. No data were collected on the participants’ work experience. Work experience could influence the results. The following conditions during the survey could not be controlled: whether the participant was distracted, whether other people were present, influencing the processing, or whether some people participated more than once, it could not be traced. It might be possible that HCWs with side effects after vaccinations tended to participate more than HCWs without side effects.
Most vaccination reactions are physiological (local or systemic) and exist for a short time. Incapacity to work and vaccine reactions were more frequent after the second vaccination for mRNA vaccines. For vector vaccines, incapacity to work and vaccine reactions were less frequent after the second vaccination. Younger participants and women experienced increased sick days and vaccination reactions. The higher the medical education level, the less frequent are sick leaves and adverse reactions after vaccination.