Demographical risk factors of adverse events following immunisation of CoronaVac in Jambi, Indonesia

Background and purpose: Slow uptake of COVID-19 vaccination might be associated with the adverse events after vaccination. This study aims to evaluate the association between demographical factors and COVID-19 vaccine adverse events in Jambi, Indonesia Methods: We conducted a cross-sectional study with a total population sampling. Data were collected with a structured questionnaire. We included adults vaccinated for the first dose with CoronaVac in Putri Ayu Public Health Center between March 15th and June 3rd, 2021, accounted for 522 persons. Statistical analysis was done using IBM SPSS 26.0. Multivariate logistic regression analysis was used to analyze the predictive models. Results: This Out of 522 respondents included, immune reactions (34.1%) were the most common adverse effects reported by the respondents. Multivariate analysis showed that respondents with no family members or friends exposed to COVID-19 are less likely to experience puncture site reaction with an OR of 0.28 (95%CI: 0.09-0.92; p=0.036). Those with no history of mental disorders were also less likely to experience the adverse event with an OR of 0.12 (95%CI: 0.024-0.62; p=0.011). As for other reactions, those in the 26-35 years old category are more likely to experience other adverse event reactions with an OR of 11.62 (95%CI: 1.5-89.92; p=0.019). Conclusion: Based Younger respondents tend to experience more other reaction of COVID-19 vaccine adverse events than the older groups. We also found that those with no family or friends who contracted COVID-19 and no history of mental disorders are less likely to experience puncture site reaction adverse events.


INTRODUCTION
3][4] Vaccination has been the primary modality in controlling this new disease internationally, including in Indonesia. 5y February 6 th , 2022, only 47% of Indonesians are fully vaccinated against COVID-19.Meanwhile, the United Arab Emirates is currently first amongst other nations in terms of fully vaccinated people, with 94% of people already receiving complete vaccination. 6The Indonesian government has set a goal to vaccinate around 67% of Indonesians by the end of 2021. 7However, by December 31st, 2021, only 41.26% received the full dose, and 58.57% received at least one shot. 8Although it is well known that vaccines are highly effective and safe, one possible explanation for this low number is the adverse events following immunisation (AEFI), even minor ones, can increase vaccine hesitancy levels and reducing vaccine uptake. 9dverse events following immunisation are divided into two categories, local and systemic.Local adverse events include pain, redness, and swelling on the injection site, while systemic adverse events include fatigue, muscle pain, fever, and nausea. 10Adverse events such as pain, headache, and fatigue are classified into very common AEFI, meaning more than 10% of vaccine recievers will experience this AEFI.In contrast, burning feeling at the injection site, vomiting, edema, and dizziness are uncommon, emerging in 0.1%-1% vaccine receivers. 11everal studies have shown an association between demographical factors such as age and gender with AEFI.Riad et al. 12 found an association between the younger age group and COVID-19 vaccine adverse events, while Hoffmann et al. 13 found that females were more likely to experience vaccine adverse events than males.There are limited studies including in Indonesia, that investigating the role of other demographic factors (family and self comorbidities, smoking status, and history of mental disorder) and their relationship with COVID-19 vaccine adverse events, especially CoronaVac as the first vaccine to be used in Indonesia. 14Identifying these factors is essential to improve vaccination strategies, such as educating the general population that certain factors increase or decrease the risk of AEFI, increasing their understanding, thus improving vaccine acceptance. 15he aim of our study was to evaluate the association of various demographical factors with vaccine reactions after the administration of the first vaccine to be used in Indonesia, the CoronaVac (Sinovac Life Sciences, Beijing, China).By discovering the factors associated with vaccine adverse events, this study might provide insight into providing a pattern of vaccine adverse events.

METHODS
We conducted a cross-sectional study with total sampling.We included adults vaccinated for the first dose with CoronaVac in Putri Ayu Public Health Center (PHC) between March 15th and June 3rd, 2021, accounted for 522 persons.Primary data from respondents were collected directly through a structured questionnaire.This study extended our previous research with a longer duration of sampling and respondents. 16Age, sex, ethnicity, religion, marital status, comorbidities, highest education attained, income, health insurance, history of mental ORIGINAL ARTICLE problems, and smoking status were all included in the questionnaire.COVID-19-related questions included previous exposure or close contact with COVID-19 patients, the impact of COVID-19 on income, whether respondents had experienced COVID-19-related symptoms, and any COVID-19 tests that were done previously.
After filling up the questionnaire, we recorded data on height and weight before administering vaccines to calculate body mass index (BMI) and blood pressure to screen for hypertension.The Asia-Pacific classification of BMI would be used to classify BMI. 17 After vaccination, we observed the participants for 30 minutes and asked them whether they experienced any AEFI.
Respondents were included if they were adults (>18 years) who were vaccinated with CoronaVac (Sinovac Life Sciences, Beijing, China) in Putri Ayu PHC, one of the biggest PHCs in Jambi City, Indonesia.PHCs (puskesmas) are government-run public health clinics in Indonesia promoting primary prevention.PHC was chosen for our study because it was the first and only place where COVID-19 vaccines were given out.The data was collected between March 15th and June 3rd, 2021.COVID-19 vaccines were administered in Indonesia in four phases.Our study fell in the middle of the second period, where the target population was public service personnel and the elderly (over 60 years old). 18However, we experienced a lot of unused doses on the field for various reasons, such as refusal, not showing up, or being contraindicated for the jab.Residents around the PHC were approached to acquire the vaccination to reduce the number of doses that could potentially go to waste.
Our exclusion criteria were broadly categorised into three: refusal to participate in the study for any reason, contraindicated to COVID-19 vaccine administration, and receiving the second vaccination.We initially followed the advice of the Indonesian Society of Internal Medicine (released on March 18th, 2021), which was the first to recommend who should be vaccinated. 19As a result, pregnant women and children were excluded from this study because vaccine guidelines for these groups were issued on June 22nd, 2021, and November 2nd, 2021, respectively. 20,21Respondents with primary immunodeficiency, acute and active infections (including SARS-CoV-2 infection or three months post-infection), blood pressure of 180/110 mmHg, unstable or uncontrolled chronic conditions, such as diabetes mellitus or heart failure, and those with a fatigue, resistance, ambulation, illness, and loss of weight (FRAIL) score of >2 were all considered contraindicated to COVID-19 vaccination. 19A total of 522 adults >18 years were included in the study as been shown in Figure 1.
Income was classified into five categories by the Indonesian Ministry of Health.Poor people had monthly household expenses of less than IDR 1,416,000 ($99); vulnerable people had monthly household expenses of between IDR 1,416,000 and IDR 2,128,000 ($99-$148); aspiring middle-class people had monthly household expenses of between IDR 2,128,001 and IDR 4,800,000 ($148 to $334); middle-class people had monthly household expenses of between IDR 4,800,001 and IDR 24,800,000 ($334 to $1671), and upper-class people had monthly household expenses above IDR 24,000,000 (~$1671). 22espondents were divided into groups according to their stance on COVID-19 immunisation.Respondents were grouped as "vaccine acceptance" if they replied yes to the question "Are you sure that you are ready to be vaccinated before arriving at Putri Ayu PHC?", "vaccine refusal" if they answered no, and "vaccine hesitance" if they answered maybe. 16Adverse events following immunisation were grouped into immune reaction if the participants experience any fever, myalgia, arthralgia; puncture site reaction if participants experience pain, redness, swollen, and stiffness around the injection site; vagal reaction if they experience syncope, dizziness or shortness of breath; and other reaction if they felt uncontrolled sleepiness and excessive hunger after vaccine administration.) was used for statistical analysis.The Kolmogorov-Smirnov test was performed to determine normality, the data had a normal distribution if the p-value was larger than 0.05.The mean and standard deviation implied that the data were regularly distributed, whereas the median and range implied that the data were not.The Chi-square test was used to do bivariate analysis.Multivariate logistic regression analysis was performed to find a prediction model with the fewest confoundings.The receiver operating characteristic (ROC) curve will compute the area under the curve (AUC).An AUC of 1.0 corresponds to a perfect result, >0.9 to a high level of accuracy, 0.7-0.9 to a moderate level of accuracy, 0.5-0.7 to a low level of accuracy, and 0.5 to a chance result. 23A p-value of >0.05 from the Hosmer-Lemeshow test would indicate a good calibration. 24he Ethics Committee of the Faculty of Medicine, Universitas Pelita Harapan, approved this study (No. 155/L-LKJ/ETIK/VI/2021).

RESULT
Of the 522 respondents, a roughly equal distribution of male and female was found (52.1%/47.9%),with most of them falling in the 36-45 years old (21.1%).The median BMI falls under the overweight category with 23.8 (13.9-61.3),with 217 respondents (41.6%) belonging to the overweight group.Two hundred respondents (38.3%) belong to the aspiring middle-class group, with the majority of respondents (77.4%) reporting no changes in income as a result of the epidemic.Muslims (64%) and Melayu (33.3%) are our respondents' most common religion and ethnicity, respectively.Our respondents are mostly married (73.8%) and live with other family members (92.7%), particularly their children or grandchildren (65.1%).Most of our responders are highly educated, with 235 (45%) having a Bachelor's degree or higher.According to the findings, 468 people (89.7%) have at least one insurance, 473 people (88.7%) have no comorbidities, and 426 people (78.3%) have family members with no comorbidities.The number of smokers and those with a history of mental problems is relatively low, with 24 (4.6%) and eight (1.5%), respectively.
Even though the majority (75.3%) of respondents state that no known close connections are positive for COVID-19 and that the majority (92.5%) of respondents have never experienced COVID-19 symptoms, 238 (45.6%) of respondents have taken at least one COVID-19 test.The majority of respondents (60.7%) are selfsufficient in registering their vaccination schedule, while 300 respondents (57.5%) do not get permission for immunisation.Regarding vaccination attitudes, 443 people (84.9%) said they are vaccine acceptors.
Multivariate analysis showed that respondents with no family or friends that are exposed to COVID-19 are less likely to experience puncture site reaction with an OR of 0.28 (95%CI: 0.09-0.92;p=0.036), and those with no history of mental disorder were also less likely to experience this adverse event with an OR of 0.12 (95%CI: 0.024-0.62;p=0.011).As for other reactions, those in the 26-35 years old category are more likely to experience other reactions with an OR of 11.62 (95%CI: 1.5-89.92;p=0.019) (See Appendix Table A3).The Hosmer-Lemeshow test yields a p-value of 0.9 for puncture site reaction and 0.95 for other reaction with an AUC value of 0.743 (95%CI: 0.658-0.828;p<0.0001) and 0.793 (95%CI: 0.717-0.869;p<0.0001),ORIGINAL ARTICLE respectively (Figure 2).These results indicate that this model has good calibration and good discrimination.

DISCUSSION
In total, 23.8% of our respondents reported AEFI after CoronaVac vaccine administration, which is lower than a study in Turkey, with 62.5% of respondents suffering from at least one AEFI after CoronaVac administration. 25The difference can also be observed in the most prevalent AEFI.While most studies find that the most common adverse events are injection site reactions such as pain and tenderness, 26,27 our study finds that immune reactions are the most common.Our results show similarity with another CoronaVac study by Supangat et al. 28 They found systemic symptoms to be more prevalent on the first dose, with the most common systemic symptoms being malaise (36%) and pain in the injection site for the local symptoms (45%).The disparity may be explained by the vaccination used (CoronaVac vs Pfizer-BioNTech®) and numerous demographic factors, but research on these two factors is currently limited.
We found that the variable most frequently associated with AEFI was the age difference in bivariate analysis.We also found that females were significantly associated with immune reactions.Being single was significantly associated with immune and other reactions.Having no jobs was significantly associated with other reactions.Lastly, having more than two comorbidities in a family was significantly associated with puncture site reaction.
Concerning gender analysis, our bivariate results show significant differences in immune reaction adverse events where females are 2.18 times more likely to experience this adverse event.However, on multivariate analysis, gender differences became insignificant.We hypothesised that sex as a confounder to marriage status might explain the change in results.In Indonesia, there are more married women compared to males, and our study shows that there are more married respondents than the ones not married. 29Most married respondents who experience immune reactions may be females who experience the same adverse effect.Thus, females appear to be significantly more likely to experience adverse immune reaction events in bivariate analysis, not multivariate analysis.
In multivariate analysis, we also found similar findings regarding age differences.Respondents in the 26-35-year-old group are more likely to experience other reaction adverse events.These findings are consistent with the Food and Drug Administration (FDA) report that injection sites and systemic reactions were more prevalent in the younger group. 30This finding is explained by the term "immunosenescence," which describes the decreasing functional capacity and reducing the immune system's efficiency with increasing age. 31The immune systems of older people would produce fewer responses than younger people, thus explaining the findings that younger respondents experience more AEFI. 13n multivariate analysis, we also discovered that having no history of family or friends contracting severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was significantly associated with puncture site reaction and that having no history of mental illness was significantly associated with puncture site reaction.However, together with marital status, occupation, and comorbidities in the family, the exact mechanism of these demographic characteristics associated with adverse vaccination effects remains unknown.
Our study is naturally limited by its cross-sectional design that relies on self-assessed and self-reported outcomes.In Indonesia, any vaccinee that experience adverse events after vaccination is recommended to report to the government to confirm whether the adverse events felt were actual or not.It is well known that psychosocial factors, such as expectations, emotions, and attitude, could modulate physical adverse event responses to COVID-19 vaccines. 32This means that when using a questionnaire, respondents who did not feel any AEFI felt that they were experiencing the listed adverse events when filling out the questionnaire with no way to confirm it.There are also possibilities where respondents could downplay the AEFI they were experiencing at the time.Another downside of our study is that we did not list all AEFI, such as diarrhea, headaches, and vomiting. Although we provided an option where the respondents could write their answer, this could prompt the respondent only to answer the listed AEFI.Our study is also limited to AEFI of the first dose vaccine only.Lastly, our study only achieves a 47.7% (1279/2679) response rate, introducing a non-response bias. 34espite some limitations, our study also has its edges.Our study is the first to look into the association between various demographic factors and COVID-19 vaccine AEFI, thus providing valuable data for further studies concerning COVID-19 vaccine AEFI in Indonesia.Our study is also one of the few vaccine studies conducted in person in Indonesia.Therefore, our study does not suffer from limitations due to using web-based surveys such as sampling and responder bias. 35,36The multivariate model of our study has good calibration and moderate discrimination, further enhancing its results and usage in clinical settings

CONCLUSION
Younger respondents tend to experience more COVID-19 vaccine adverse events than older groups.Other demographical variables associated with COVID-19 vaccine adverse are no family or friends who contracted COVID-19 and no history of any mental disorders.Further independent epidemiological studies for COVID-19 vaccine AEFI should be carried out to increase public confidence in vaccine safety and accelerate its uptake process.Upcoming studies should also compare vaccines from other manufacturers and investigate demographical variables' associations with COVID-19 vaccine adverse events.Table A1

Figure 1 .
Figure 1.Flowchart of respondent selection

Figure 2 .
Figure 2. The receiver operating curve to assess discrimination of the multivariate model of other reaction (left)