The questionnaire was developed based on a literature review. A panel of experts confirmed the content validity of the questionnaire. The questionnaire was developed in the English language and then translated to Arabic, which is the only official language in Jordan (98% of the Jordanian population are Arabs and the remaining 2% use Arabic for their daily interactions) (13). It was then translated back into English by different translators, and finally, compared by a third translator. Face validity was conducted in a pilot study that included 30 participants who assessed the questionnaire's clarity, and no significant modifications were required.

The final Arabic version of the questionnaire consisted of six branched sections. The first section collected participants' demographic information, including marital status, smoking habits, obesity status, education level, household average monthly income, health status, and whether participants worked in or studied a health-related field. The second section gathered information about the participants' experience with COVID-19. The third section assessed the attitudes of participants toward COVID-19, while the fourth evaluated their knowledge about COVID-19, including symptoms, transmission methods, preventive measures, and treatment availability. The fifth section asked about the preventive measures against COVID-19 used by the participants. The sixth and final section asked about the participants' willingness to be vaccinated against COVID-19 (once available), and the participants who responded “No” or “Not sure” were asked to give their reasons.

The degree of Covid-19 risk affecting participants was determined according to the Centers for Disease Control and Prevention (CDC) (1) categories. The high-risk group included smokers, obese, pregnant, or who had at least one of the following conditions (Type 2 diabetes mellitus/Chronic Obstructive Pulmonary Disease (COPD)/Cancer/Kidney Failure/Heart diseases/Organ transplantation/Sickle Cell Anemia). The medium-risk group included those who did not fit for the high-risk category but were overweight or had at least one of these conditions (Type 1 diabetes mellitus/Hypertension/Bone marrow transplant/ Cerebrovascular diseases or stroke/ Cystic Fibrosis/Asthma/Taking steroids or immunosuppressant drugs/Hepatic diseases/ Thalassemia/Lung fibrosis). The low-risk group included all other participants that do not fit the previously mentioned criteria.

Two scores were calculated: the knowledge score and the practice score. For knowledge, the maximum possible score was 21, as for each right answer, one point was granted (the score was calculated based on items in Table 3). The participants were divided into two groups based on their knowledge scores. The high-knowledge group included participants that scored more than the total sample median (median = 17), and the low-knowledge group included participants that scored below the total sample median.

Practice scores were calculated for those who had not been infected with the COVID-19 virus. One point was added for each answer representing a scientifically proven protective measure against COVID-19, and the maximum possible score was 10. After calculating the practice scores, the participants were divided into two groups. The high protective practice group included those who scored more than the sample median (median = 7), and the low protective practice group included those who scored below the median.

Kish formula (14) was applied to determine the least required sample size with a confidence interval level of 95% and a 4% margin for error. The estimated sample size was 600. Categorical variables were presented, such as frequency and percentages, and continuous variables were presented as means and standard deviations (SD). Crosstabulation with χ2 was applied to evaluate the association between intent to be vaccinated and participants' demographic characteristics, knowledge level about COVID-19, and protective practice against COVID-19. Binary logistic regressions were conducted on participants who answered “No” to “Have you ever been infected with COVID-19?” to evaluate variables associated with knowledge level and protective practice against COVID-19 level. A multinomial logistic regression was used to evaluate the variables related to the participants' intent to be vaccinated when a vaccine against COVID-19 becomes available.

An analysis of variance (ANOVA) with LSD post hoc test was applied to measure the difference in the perceived degree of seriousness of COVID-19 between participants with different responses to “Have you ever been infected with COVID-19?” All statistical analysis was conducted using SPSS version 25.

The participants' perception of the seriousness of COVID-19 among participants was not associated with COVID-19 infection history. The degree of seriousness estimations were not different between those who were not infected, those who suspected their infection, and those who were infected (means = 6.49, 6.52, and 6.57, respectively). High adherence to quarantine was reported by those infected, which contradicts the results of other studies (17, 18), which reported poor adherence to quarantine during different pandemics. However, the degree of quarantine adherence among those who were certain about their infection was significantly higher than those who only suspected that they were infected. This indicates the importance of COVID-19 testing in decreasing the disease spread by increasing the adherence of infected patients to quarantine.

Even though this study was conducted in October 2020 when the number of COVID-19 cases increased rapidly in Jordan, only 36.8% of participants intend to be vaccinated once the vaccine was available. This percentage is much lower than the percentage reported by a global survey that included participants from 19 countries (19) (71.5%) and by studies conducted in Ecuador (20) (97%), the United States (21) (57%), France (22) (76%), China (23) (91.3%), and Saudi Arabia (24) (64.7%). The peak rise in COVID-19 cases happened much earlier in other countries than in Jordan, which may contribute to the higher vaccination hesitancy among the Jordanian population compared to other populations, as the disease is new to Jordan. Although the percentage of the population who need to be vaccinated to achieve herd immunity against COVID-19 is not yet well-known, in general, 50–90% (25) of the population needs to be immune either naturally or by vaccines to achieve herd immunity. Should this high hesitancy toward vaccination continue among the Jordanian population, it might be difficult to achieve herd immunity. Several sample characteristics had a significant negative impact on the participants' intention to be vaccinated, including; being female, married, and having a postgrad degree compared to university students. Lower vaccination intention among female participants was also observed in studies conducted in France (22), China (23), and Europe (26). However, other factors reported in other studies, like income (22, 26) and age (22), were not significant predictors in this study. The recognition of these factors could help develop targeted awareness campaigns directed to the population to increase the vaccination rate once the vaccine is available.

The perceived risk of COVID-19 was a significant predictor of the participants' vaccination intention in this study, reflecting several other studies (22, 23). The higher the perceived risk, the lower the vaccination hesitancy. Therefore, increasing the population's consciousness about the seriousness of the disease is essential in improving their willingness to be vaccinated.

Detecting the causes of vaccination refusal or hesitancy could improve the population's vaccination intentions. It is important to better understand the rationales and reasons for vaccination refusal or hesitancy if we are to remove these barriers.

Concerns about the vaccine were the most common reason behind hesitancy or refusal among the participants. These concerns about vaccine safety and side effects are global, as indicated by studies conducted in the United States (21), Europe (26), and China (23). The rationale behind these concerns is reasonable, as several vaccine candidate trials were paused (27, 28) due to detected side effects. However, the suspension of these studies, once side effects were noted, could be used to assure the rigor of vaccine testing, another concern among the population. Misbeliefs about the association between vaccines and autism or the vaccine's effect on fertility were not common among the participants, implying the dubiety about many popular myths associated with vaccines.

Efficacy is a frequently mentioned concern whenever a new vaccine is developed (29, 30). This concern could be of less importance once the vaccine is available and successful results are published.

Undesirable attitudes were the second most mentioned barriers. Increasing the population's understanding of the vaccines and the related mechanisms of action through different awareness-raising methods could overcome this barrier. About 30% of those with undesirable attitudes were against vaccination in general. Several studies (31, 32) have established approaches to overcome vaccine refusal that could be useful for COVID-19 vaccination. For example, opposing the spread of false information and targeting children and adolescence, who might not have robust emotions about vaccines yet, could increase COVID-19 vaccine acceptability.

The need for additional information was reported as a barrier by 58% of the respondents and 79% of participants in an Indonesian study (33). The role of healthcare providers is influential in this respect, as they provide patients and the general population with much-needed information. An Australian study (34) provided a framework that could be used by healthcare providers to increase confidence in any potential COVID-19 vaccines.

Lack of trust was also a cause of vaccination hesitation or refusal for many participants in this study and an American study (21). A belief in the conspiracy theories associated with COVID-19 among the Jordanian population was observed in the present study. These beliefs have also been reported by another Jordanian study on COVID-19 (unpublished data). Several strategies have been suggested to combat conspiracy theories (35), including the careful dissemination of medical research, social media campaigns, and developing a culture of fact-checking. A report issued by WHO has discussed the behavioral considerations of COVID-19 acceptance and suggested different approaches to increasing vaccine acceptance. These include building an enabling environment and using open communication to address people's beliefs and uncertainty, educating them about the safety and efficacy of the vaccine (36). Social and governmental collaboration will increase public confidence in the COVID-19 vaccine and enable the country to reach herd immunity rapidly.

One of this study's strengths is the large sample size, which decreases the influence of existing bias. Another strong aspect of the present study is that it evaluated the participants' KAP toward COVID-19 and their vaccination intentions and assessed the association between KAP and vaccination intentions.

At the time this study was conducted, the vaccine was not available. The participants' vaccination intentions may be different now, as the vaccine has been made available. More information has now been published, which may be considered a limitation of this study. This study was based on an online questionnaire, meaning the results are subject to recall and selection biases. However, previous studies have shown that web-based research is a cost-effective method that can be used to generate a sample that is representative of the total population with a fraction of the cost (37). It can reach people otherwise unreachable and provides a safe and private environment for the respondents to answer questions accurately and honestly compared with face-to-face interviews (38).

It has been suggested that as the number of Internet users has increased globally, the socio-demographic characteristics of the recruited participants via web-based surveys reflect the general population (39). This can be applied in Jordan as Internet users are estimated to be 67% of all age groups (40). This percentage could be higher when children under the age of 18 are excluded. Another limitation could be that the study sample age was positively skewed. However, the Jordanian population is a young one, and the age group between 20–29 years represents 33.45% of the total population above the age of 19 (41).

Finally, almost half of the sample participants worked in the medical field, which may cast doubt in the sample's representation of the total population. Nevertheless, the percentage of medical field workers in Jordan is significantly higher than in other countries globally. For example, Jordan is ranked fourth in terms of the number of pharmacists for every 10,000 people (16 pharmacists for 10,000) (42). Furthermore, the authors believe that the perception of medical-related staff is particularly interesting, as their opinions influence the general population, and they represent a high-risk group, therefore, their vaccination is a priority. Finally, as indicated by the Kish formula, the smallest required sample size is 600, therefore, a study sample that includes 1,144 participants could provide sufficient data to evaluate each subgroup.