Demographic characteristics:
Of 3509 participants enrolled in the study, 1672 (47.6%) received at least one dose of the Covid-19 vaccine while 1837 (52.3%) remain unvaccinated. Of 1672 participants with at least one dose of Covid-19, 316 (18.9%) were partially vaccinated and 1356 (81.1%) were fully vaccinated. Males represented 1662 (47.4%) and females represented 1847 (52.6%) of the sample. The age group 20 to 29 years represented a majority in the sample 1380 (39.3%). Participants in the age category 30 to 59 years were lower in the vaccinated group 689 (41.2%) compared with the unvaccinated group 819 (44.6%) (p-value < .001). Regarding residency, people who live in rural areas were lower in the vaccinated 341 (20.4%) compared with the unvaccinated group 552 (30.0%) (p-value < .001). In contrast, Syrians who live in cities or who took the vaccine outside Syria were higher in the vaccinated group 1160 (69.4%) and 171 (10.2%) compared with the unvaccinated group 1244 (67.7%) and 41 (2.2%) (p-value < .001) respectively. Further data about participants residing outside Syria are available in the supplementary (table 1 Supplementary). There were higher proportions of healthcare workers in the vaccinated group 660 (39.5%) compared with the unvaccinated group 380 (20.7%) (p-value < .001). Participants with a history of known allergy, chronic co-morbidity, and smoking were lower in the vaccinated group 208 (12.4%), 369 (22.1%), and 727 (43.5%) compared with the unvaccinated group 288 (15.7%), 459 (25.0), and 902 (49.1%) (p-value = 0.006, 0.042, and 0.001) respectively. Participants who received an influenza shot last year were higher in the vaccinated group 406 (24.3%) compared with the unvaccinated group 87 (4.7%) (p-value < .001) (Table 1).
Unvaccinated participants:
The 1837 unvaccinated participants reported the following reasons for not taking the vaccine: concern about the vaccines side effects 815 (44.4%), unconvinced of the vaccine benefits 762 (41.5%), will not contract Covid-19 as previously contracted the virus 400 (21.8%), medical exemption 94 (5.1%), and vaccine unavailability 84 (4.6%).
Covid-19 vaccine side effects by socio-demographic characteristics:
Of the 1672 vaccinated participants, 788 (47.1%) had mild side effects, 374 (22.3%) had moderate side effects, 366 (21.9%) had no side effects, and 144 (8.6%) had severe side effects. Regarding gender, a higher proportion of females suffered from severe side effects 104 (72.2%) compared with males 214 (58.5%) (p-value < .001). Reported severe side effects was higher among participants in the age categories 16-20 years 16 (11.0%), 20-29 years 60 (8.3%), 30-39 years 29 (9.5%), and 40-49 years 34 (15.0%) compared with the age categories 50-59 years 4 (2.6%) and 60≤ years 1 (0.9%) (p-value < .001).
Participants with chronic co-morbidities reported higher numbers of mild side effects 156 (42.3%) compared with no side effects 78 (21.1%) (p-value < .001). Of the co-morbidities reported, diabetes mellitus was associated with higher numbers of mild side effects 27 (32.1%) compared with no side effects 21 (25.0%) (p-value < .001), while allergies, respiratory disease, and hematological disease were associated with higher numbers of severe side effects 60 (28.8%), 13 (23.2%), and 4 (36.4%) compared with no side effects 31 (14.9%), 10 (17.9%), and (0.0%) (p-value < .001), (p-value = 0.001), and (p-value = 0.001) respectively.
The majority 591 (81.3%) of smokers reported post-vaccine side effects compared with no side effects 136 (18.7%) (p-value = 0.048). A history of Covid-19 infection was associated with higher numbers of vaccine side effects 583 (82.6%) compared with no side effects 123 (17.4%) (p-value = 0.001). Participants who received an influenza vaccine (flu jab) reported higher numbers of mild side effects 192 (47.3%) and no side effects 133 (32.8%) compared with moderate side effects 60 (14.8%) and severe side effects 21 (5.2%) (p-value < .001) (Table 2).
Covid-19 vaccine side effects and complications by vaccine brand:
Vaccine brands available in Syria at the time of the study were AstraZeneca-Oxford 552 (33.0%), Sputnik light 294 (17.5%), Pfizer-BioNTech 280 (16.7%), Sputix v 203 (12.1%), Sinophram 140 (8.4%), Sinovac 93 (5.6%), Johnson & Johnson 58 (3.5%), and Moderna 52 (3.1%). The majority of symptoms started within 12 to 24 hours 614 (47.0%) while the minority started after 48 hours 36 (2.8%) (p-value = 0.001) (Table 3).
The most common reported side effects were tiredness and fatigue 1028 (61.5%), pain at the injection site 933 (55.8%), low-grade fever 684 (40.9%), headache 648 (38.8%), and muscle pain 615 (36. 8%). Tiredness and fatigue were higher among AstraZeneca-Oxford 377 (68.3%) and Sputnick light 200 (68.0%) compared with Sinophram 67 (47.9%) and Sinivac 39(41.9%) (p-value < .001). Headache was more common among Moderna 27 (51.9%) compared with Sinovac 23 (24.7%) (p-value < .001). Low grade fever (<39) was more common among Johnson & Johnson 32 (55.2%) compared with Sinovac 26 (28.0%) (p-value = 0.001). High grade fever was more common among AstraZeneca 90 (16.3%) compared with Pfizer-BioNTech 14 (5.0%) (p-value < .001). Chills were higher among AstraZeneca 157 (28.4%) compared with Sinovac 5 (5.4%) (p-value < .001). Pain at the injection site was more common among Sputnik light 190 (64.6%) compared with Sinovac 31 (33.3%) (p-value < .001). Swelling, redness, and/or temperature at the injection site were more common among Moderna 15 (28.8%) compared with Sinovac 5 (5.4%) (p-value < .001). Joint pain were more common among Johnson & Johnson 25 (43.1%) compared with Pfizer-BioNTech 65 (23.2%) (p-value < .001). Myalgia was more common among AstraZeneca 261 (47.3%) compared with Sinopharm 30 (21.4%) (p-value < .001). Diarrhea was more common among Johnson & Johnson 7 (12.1%) compared with Sputnick light 2 (0.7%) (p-value = 0.001). Blurred vision was higher among AstraZeneca 22 (4.0%) compared with Johnson & Johnson 0 (0.0%) (p-value = 0.028). Sweating was more common among Johnson & Johnson 19 (32.8%) compared with Sinopharm 18 (12.9%) and Sinovac 12 (12.9%) (p-value < .001). Cough was more common among Moderna 14 (26.9%) compared with Sputnick light 20 (6.8%) (p-value < .001). Nasal congestion was more common among AstraZeneca 97 (17.6%) compared with Sputnick light 21 (7.1%) (p-value < .001). Runny nose was more common among Johnson & Johnson 8 (13.8%) compared with Sputnick light 10 (3.4%). Sore throat was more common among AstraZeneca 76 (13.8%) compared with Sinopharm 9 (6.4%) (p-value = 0.028). Laziness was more common among AstraZeneca 247 (44.7%) compared with Sinovac 18 (19.4%) (p-value < .001). Insomnia was more common among Sputnick light 87 (29.6%) compared with Sinopharm 24 (17.1%) (p-value = 0.006). Dysrhythmia was more common among AstraZeneca 64 (11.6%) compared with Sinopharm 2 (1.4%). Hypotension or hypertension was more common among Moderna 5 (9.6%) compared with Johnson & Johnson 0 (0.0%) (p-value < .001). Chest pain was more common among Johnson & Johnson 5 (8.6%) compared with Moderna 1 (1.4%) (p-value = 0.002). Dyspnea was more common among AstraZeneca 57 (10.3%) compared with Sinopharm 5 (3.6%) (p-value = 0.012). Anxiety was more common among Johnson & Johnson 8 (13.8%) compared with Sinopharm 0 (0.0%) (p-value < .001) (Table 3).
Regarding Covid-19 vaccine symptom severity, the Sinopharm vaccine was associated with a higher percentage of no side effects 51 (36.4%) compared with Johnson & Johnson 7 (12.1%) (p-value < .001), while most of the mild side effects were associated with Sputnik light 168 (57.1%) compared with AstraZeneca 215 (38.9%) (p-value < .001). Moderate side effects were most reported among participants who received Johnson & Johnson 21 (36.2%) compared with Sinovac 8 (8.6%) (p-value < .001). Severe side effects were most associated with AstraZeneca 83 (15.0%) compared with Sinopharm 5 (3.6%) (p-value < .001) (Table 3). Post-vaccination side effect severity varied across the first and second doses (Figure 1). Shockingly, severe side effects were higher after the second dose of most vaccines, including AstraZeneca, Pfizer-BioNTech, Sinopharm, and Moderna (Figure 1 B, D, H, and L).
The duration of post-vaccination symptoms were reported as follows, >12 hours 262 (20.1%), 12-24 hours 508 (38.9%), 1-2 days 380 (29.1%), 3 days to 1 week 105 (8.0%), 1-2 weeks 23 (1.8%), and >2 weeks 28 (2.1%) (p-value = 0.011) (Table 3). Participants, whose symptoms lasted over 2 weeks, self-reported the following symptoms, pain at the injection site 20 (1.2%), muscle or joint pain 13 (0.8%), fatigue 11 (0.7%), dysrhythmia 5 (0.3%), menstrual abnormalities 4 (0.5%), headache 3 (0.2%), and nasal congestion 2 (0.1%). Other symptoms include low white blood cell count and low platelet count, swollen legs, leg purpura, and hypertension, which were each reported among 1 (0.1%) participant.
The proportion of participants who took painkillers for their symptoms were higher among AstraZeneca 377 (68.3%), Johnson & Johnson 39 (67.2%), Sputnik Light 182 (61.9%), Sputnik v 122 (60.1%), and Moderna 31 (59.6%) compared with Pfizer-BioNTech 139 (49.6%), Sinopharm 69 (49.3%), and Sinovac 42 (45.2%). The highest number of required hospitalization post-vaccination was among AstraZeneca 34 (6.2%) (Table 3).
Serious medical complications, such as blood clots and low platelet counts were reported by 5 (0.3%) and 2 (0.1%) respectively. Anaphylaxis shock 1 (0.1%), seizures 1 (0.1%), and cardiac infarction 1 (0.1%) were only reported among those who took the AstraZeneca vaccine (Table 3).
Regarding vaccine efficacy, a minority reported having Covid-19 infection after taking the first dose 129 (7.7%) and after taking two doses 117 (7.0%) (p-value < .001). The highest Covid-19 reinfection rate after the first dose was among Sinovac 12 (12.9%), whereas after the second dose was among Sinopharm 21 (15.0%) (p-value < .001) (Table 3).
Multivariate Logistic Regression Analysis:
A multivariate logistic regression analysis was performed to identify the variables including age, sex, smoking, occupation, influenza vaccine, vaccine brand, history of Covid-19 infection, and co-morbidities and their association with the development of side effects versus no side effects post-vaccination. The logistic regression model was statistically significant for the following factors, age <40 (vs. ≥40; OR: 1.866, p-value < .001), females (vs. males; OR: 1.696, p-value < .001), current smoker (vs. nonsmoker; OR: 1.428, p-value = .006), did not receive Influenza shot last year (vs. did receive Influenza shot last year OR: 1.929, p-value < .001), AstraZeneca vaccine (vs. other vaccine brands OR: 1.426, p-value = .021), history of Covid-19 infection pre-vaccination (vs. no history of Covid-19 infection pre-vaccination OR: 1.317, p-value = .034), and co-morbidities (vs. no co-morbidities, OR: 1.438, p-value .029) were significantly associated with post-vaccination side effects (Table 4).
A multivariate logistic regression analysis was performed to identify the variables including age, sex, smoking, occupation, influenza vaccine, vaccine brand, history of Covid-19 infection, co-morbidities, hypertension, and diabetes mellitus and their association with the development of severe side effects versus no, mild, and moderate side effects post-vaccination. The logistic regression model was statistically significant for the following factors, age <40 (vs. ≥40; OR: 2.113, p-value = 0.008), females (vs. males; OR: 2.245, p-value < .001), did not receive influenza shot last year (vs. did receive Influenza shot last year OR: 1.697, p-value = 0.041), AstraZeneca (vs. other vaccine brands; OR: 2.799, p-value < .001), co-morbidities (vs. no co-morbidities; OR: 1.993, p-value = 0.008), and diabetes mellitus (vs. no diabetes mellitus; OR: 2.788, p-value = 0.007) were significantly associated with severe post-vaccination side effects (Table 5).
Vaccine efficacy:
The Kaplan-Meier plot was used to calculate the probability of remaining free of Covid-19 infection over time. Fully vaccinated participants had higher free of infection time 14.8 months (95% CI, 14.6 to 15.0), compared with partially vaccinated 11.1 months (95% CI, 10.7 to 11.4) and previously infected unvaccinated participants 10.5 months (95% CI, 10.3 to 10.7) (p-value < .001) (figure 2 A) (table 1 supplementary). Also, the Kaplan-Meier plot was used to compare the efficacy of vaccination in remaining free of infection between September 2021 and February 2022, during the last Covid-19 waves in Syria. Participants who received any vaccine dose before September 2021 had a higher probability of remaining Covid-19 infection-free between September 2021 and February 2022, followed by previously infected unvaccinated participants, while uninfected unvaccinated participants had the lowest probability of remaining infection-free (p-value < .001) (figure 2 B) (table 2 and 3 supplementary).