key: cord-0722601-65njhj41 authors: Canlas, Ferddie Quiroz; Nair, Sarachandran; Paat, Isabelo D. title: Exploring COVID-19 Vaccine Side Effects: A Correlational Study Using Python date: 2022-12-31 journal: Procedia Computer Science DOI: 10.1016/j.procs.2022.03.102 sha: b7ecafdd08138c2e4a9ab67f01b5e4e0352ae1df doc_id: 722601 cord_uid: 65njhj41 The COVID-19 pandemic had a great impact on the socio-economic stability of every country. To curb the effect and risk of transmission, governments implemented various measures including the mandatory vaccination of their citizens. However, despite these efforts, many people are still hesitant to take the vaccine because of various reasons and biases. This paper attempts to explore the perceptions of the people who have undergone vaccinations regarding the various side effects to provide inputs to vaccine manufacturers and assist people in making informed decisions in selecting the appropriate vaccine for them. The study further explored the correlation and association of age, weight category, diet category, blood type, and sleeping patterns with the severity of the selected vaccine side effects. The results revealed that vaccine side effects are associated with the vaccine type. Age, gender, weight category, diet category, blood type, and sleeping patterns have significant relationships to one or more side effects. The onset of the COVID-19 brought about by the Novel Coronavirus in December 2019 led governments of every country across the globe to impose restrictions on various activities in the society [1] . Government agencies The onset of the COVID-19 brought about by the Novel Coronavirus in December 2019 led governments of every country across the globe to impose restrictions on various activities in the society [1] . Government agencies had imposed lockdowns and temporary closures of establishments to curb the spread of the virus. This action had an adverse impact not only on the economy in general but on almost all sectors of society [2] . Governments find vaccines as the viable solution to curb the impacts of the pandemic specifically in the economy of the country. Initiatives were carefully undertaken to ensure that citizens could avail vaccination either for free or through personal initiatives. Despite these efforts, it is observed that the rate of vaccination is low as many have apprehensions about taking the vaccines for various reasons. BBC reported that people have different biases from vaccine brands and possible severe side effects, especially for older people. Although health practitioners have emphasized the risks of side effects are less significant than that of the risks of the disease itself [3] . The Supreme Council of the Sultanate of Oman, to encourage its citizens and expatriates to take the vaccine and to slow down the rate of virus transmission had issued effective September 1, 2021, a decree mandating the presentation of vaccination certificates in entering private and public premises such as schools, restaurants, shopping malls, etc [4] . Notwithstanding, the rate of vaccination is observed not to be in a full swing as people are hesitant due to the risks involved. This paper attempted to explore the perceptions of the people who have undergone vaccinations regarding the various side effects to provide inputs to vaccine manufacturers and assist people in making informed decisions in selecting the appropriate vaccine for them. The study further explored the correlation and association of age, weight category, diet category, blood type, and sleeping patterns with the severity of the selected vaccine side effects. Ultimately, the analyzed data can be used as a training dataset for the design of prediction algorithms. There are studies conducted across different countries about the side effects of COVID 19 vaccines. Most of the studies focus on AstraZeneca and Pfizer brands, also Sinopharm and Moderna. There are few involving all vaccine types. Parameters used were the common local symptoms (arm pain, swelling, redness, etc. ) and systemic symptoms(headaches, flu, etc.). Using random forest, a study in Jordan [5] revealed that side effects significantly vary based on the vaccine type. Similarly, in Sweden [6] age and gender play roles in the severity of the side effects. In Saudi and the United Arab Emirates, side effects are common to males than females who took Pfizer or AstraZeneca [7] . Older people who have previous infections have difficulty breathing after taking Pfizer [8] . Similarly, AstraZeneca takers who had previous infections experience severe pains [9] . Females who took either Pfizer or Sinopharm were reported to experience adverse side effects [10] . The t-test and chi-square test show that Pfizer is perceived to have severe side effects as compared to the other brands [11] especially in the second dose [12] and varies depending on the BMI and health status of an individual [13] . As compared to Sinopharm, it has a shorter side effects duration [14] . The mRNAbased vaccines have been reported as the source of local side effects while viral vector-based vaccines are on the systemic ones affecting the younger age group and females [15] . The study is confined to the Sultanate of Oman. Respondents possessing the criteria for the study were initially selected using purposive sampling [16] . They were asked to fill out a Google Form containing the survey questions. Later, using the snowball technique, the initial respondents were asked to send the survey link to their friends who took the same vaccine. Sampling goes in the same fashion until the deadline is set. The data preparation phase eliminated the outliers in the data, thus resulting in a sample size of 362 respondents. Although currently there are eight (8) vaccines approved. Most of the respondents took the two brands (Pfizer and AstraZeneca) the government initially approved. Using a 5% level of significance, the Chi-square test of independence/association and Spearman test of correlation ( Fig. 1) determined if there is a significant relationship between the group of variables used in the study: the independent variables (Table 1 ) and dependent variables (Table 2 ) [17] [18] . Statistical analyses (both descriptive and inferential) were performed using Python in the Anaconda Spider and libraries such as Pandas for loading and data preparation, NumPy for arrays and matrices, Matplotlib for data visualization, Seaborn for heatmaps and data summarization, SciPy for statistical analysis, and Sklearn for clustering. Instead of the actual count, the frequency of the responses in each category was expressed in percentage for better understanding. The level of severity [19] was interpreted by mapping the mean values to the verbal interpretations (4.20 -5.00, Worst; 3.40 -4.19, Very Severe; 2.60 -3.39, Severe; 1.80 -2.59, Moderate; 1.00 -1.79, Mild; and 0.00 -0.99, Not experiences at all). The interpretation of the Spearman test was taken from the study of Schobert et al [20] . Respondents were segmented into two based on the vaccine type taken. The normal distribution, test of skewness, interquartile range, and box plotting (Fig. 2) helped to identify the data outliers that will potentially affect the inferential statistics [21] . Extreme outliers were removed, and the minor ones were replaced with the median value as it is not affected by the outliers [22] . The Chi-square tests (Table 3) show the association between local and systemic side effects and vaccine brands. Pfizer has moderate to worst side effects during the second dose. The Spearman correlation shows that vaccine brand has a negative weak correlation(p=0.001,r=0.253) to the duration of the side effects. AstraZeneca tends to have a longer duration of side effects as compared to Pfizer. The tests show that females(F) have greater chances to experience moderate to worst side effects, both for the first (Table 4a ) and second (Table 4b) doses as compared to the males(M). Similarly, blood types B-,AB+, O-, underweight(UW), overweight(OW), and people with an unbalanced diet were found to experience the same. Table 5 shows that age has a negative weak (WN) to moderate (MN) correlation to the side effects' level of severity. The severity affects the younger respondents during the first and second doses. Average sleep per day has no significant relationship with the vaccine side effects except for a very weak negative correlation(r=-0.230,p=0.005) with nausea or vomiting during the second dose. Table 5 shows the result of the test, emphasizing the significant variables. Duration of the side effects has nothing to do with age and sleeping pattern. AstraZeneca has a lesser number of perceived side effects as compared to Pfizer. Mostly, underweight (UW), overweight (OW), and people with unbalanced diets tend to experience moderate to worst side effects. Gender and blood type were not associated with any side effects. For brevity, Table 5 shows the variables having significant associations. Although age has a negative weak correlation (r=-0.362,p=0.030) to fever during the first dose, together with average sleeping hours per day, has no significant relationship with the rest of the side effects. The study revealed that vaccine side effects are associated with the vaccine type. It was identified that Pfizer has many known side effects. Age, gender, weight category, diet category, and blood type have significant relationships to one or more side effects. On the contrary, the sleeping pattern variable has no significant relationship to the other variables. To draw more reliable conclusions, it is highly recommended to involve more respondents and vaccine brands. UNESCO COVID-19 education response: how many students are at risk of not returning to school? Advocacy paper Why some people don't want a Covid-19 vaccine Vaccination mandatory to enter malls, restaurants in Oman: Supreme Committee Side Effects and Perceptions Following COVID-19 Vaccination in Jordan: A Randomized, Cross-Sectional Study Implementing Machine Learning for Predicting Severity of Side Effects Adverse events of special interest for COVID-19 vaccines -background incidences vary by sex, age and time period and are affected by the pandemic Evaluation of Post-Vaccination Symptoms of Two Common COVID-19 Vaccines Used in Abha, Aseer Region, Kingdom of Saudi Arabia Minor to Moderate Side Effects of Pfizer Vaccine Among Saudi Residents: A Retrospective Cross-Sectional Study Adverse Effects of the COVID-19 Vaccine Reported by Lecturers and Staff of Kabul University of Medical Sciences Side effects and perceptions following Sinopharm COVID-19 vaccination Side effects of BNT162b2 mRNA COVID-19 vaccine: A randomized, crosssectional study with detailed self-reported symptoms from healthcare workers Reactogenicity Following Receipt of mRNA-Based COVID-19 Vaccines 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 Qualitative Assessment of Early Adverse Effects of Pfizer-BioNTech and Sinopharm COVID-19 Vaccines by Telephone Interviews Side Effects of mRNA-Based and Viral Vector-Based COVID-19 Purposive sampling in a qualitative evidence synthesis: a worked example from a synthesis on parental perceptions of vaccination communication Possible Side Effects After Getting a COVID-19 Vaccine Coronavirus disease (COVID-19): Vaccines safety What is a pain scale, and how is it used? Correlation Coefficients: Appropriate Use and Interpretation 6 ways to test for a Normal Distribution -which one to use? 7 Ways to Handle Missing Values in Machine Learning