Worldwide Estimation of Parental Acceptance of COVID-19 Vaccine for Their Children: A Systematic Review and Meta-Analysis

Currently, the best method to well control the spread of COVID-19 without severe mental health problems is to reach herd immunity. Therefore, the vaccination rate of the COVID-19 vaccine is critical. Among the populations, children are the vulnerable ones to get vaccinated; therefore, it is important to assess parents’ and guardians’ willingness to have their children vaccinated. The present systematic review and meta-analysis synthesized evidence to estimate the parents’ acceptance rate of COVID-19 vaccination toward their children. Additionally, factors explaining the acceptance rate were investigated. Four academic databases (PubMed, Scopus, Web of Science, and ProQuest) together with Google Scholar were searched, and the references of the included publications were searched as well. Using the PECO-S framework (population, exposure, comparison, outcome, and study design), observational studies of cross-sectional, cohort, or case-control studies were included. The outcome was parents’ or guardians’ willingness to let their children be vaccinated. The studies included in the present review were restricted to English and peer-reviewed papers published between December 2019 and July 2022. A total of 98 papers across 69 different countries with 413,590 participants were included. The mean age of the parents was 39.10 (range: 18–70) years and that of their children was 8.45 (range: 0–18) years. The pooled estimated prevalence of parental acceptance to vaccinate their children with the COVID-19 vaccine was 57% (98 studies, 95% CI: 52–62%, I2: 99.92%, τ2: 0.06). Moreover, data collection time was a significant factor explaining parental willingness in the multivariable meta-regression, with a 13% decrease in parental willingness by each month increase in time, explaining 11.44% of variance. Qualitative synthesis results showed that parents’ COVID-19 vaccine knowledge, trust in theCOVID-19 vaccine, and facilitators in vaccination (e.g., low cost, good vaccine accessibility, and government incentive) were significant factors for higher willingness, while mental health problems (e.g., having worries and psychological distress) were significant factors for lower willingness. Given that the acceptance rate was relatively low (57%) and does not achieve the requirement of herd immunity (i.e., 70%), governments and healthcare authorities should try to elevate parents’ knowledge and trust in the COVID-19 vaccine, facilitate in vaccination, and reduce their mental difficulties to improve the overall vaccination rate among children.

the diverse subgroups when synthesizing data to examine the variables related to the acceptance of the COVID-19 vaccine, especially for children. Moreover, the latest studies recommend new vaccines against COVID-19 for children and adolescents [40,41], which results in difficult decision making for parents and caregivers to vaccinate their children. For this reason, the aim of the current research was to draw a comprehensive and related picture of various factors and attitudes related to this decision. The information obtained will help to provide a better understanding for further research as well as for health authorities and professionals to respond to potential problems in an adequate and targeted manner.

Study Aims
This study aimed primarily to estimate the prevalence of parental acceptance/willingness to vaccinate their children with the COVID-19 vaccine. The secondary aims were: • Assessment of heterogeneity and its possible sources for estimated pooled prevalence of parental acceptance/willingness to vaccinate their children with the COVID-19 vaccine; • Moderator analyses to determine influential variables sources for estimated pooled prevalence of parental acceptance/willingness to vaccinate their children with the COVID-19 vaccine; • Determining influential factors for parents to accept COVID-19 vaccination for their children.

Protocol and Registration
The study protocol was registered in the PROSPERO, International prospective register of systematic reviews under decree code of CRD42022333337 [42]. The findings of this systematic review are reported based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) reporting guideline [43].

Systematic Review Questions
The systematic review question was formulated using PECO-S framework. The PECO-S is a framework to formulate search questions assessing associations between exposures and outcomes in various fields of health [44]. The PECO components are P for Population, E for Exposure, C for Comparison, O for Outcome, and S for Study design. The present study was designed to answer the following main research question: What is the worldwide estimated prevalence of parental acceptance to vaccinate their children with the COVID-19 vaccine?

Eligibility Criteria
The eligibility criteria based on PECO components were set as follows: (1) Population: parents or children's guardian with no limitation regarding their demographic characteristics; (2) Exposure: COVID-19 pandemic; (3) Comparison: other populations other than children; (4) Outcome: Frequency or prevalence of COVID-19 vaccination acceptance (and/or no hesitance) or willingness to receive COVID-19 vaccines for children; and (5) Study design: observational studies including cross sectional, cohort, or case-control design.
Other eligibility criteria include being published between December 2019 and July 2022, using English language, published as a peer-reviewed paper, reporting data on frequency or prevalence of parents or children's guardian acceptance for their children's COVID-19 vaccination.

Information Sources
Academic databases including PubMed, Scopus, Web of Science (WoS), and ProQuest were systematically searched from the beginning of December 2019 to the end of July 2022. To have a more comprehensive search, reference lists of the included publications and medRxiv were independently searched.

Search Strategy
The main search terms included COVID-19, vaccine, parents, and children. The search strategy was developed using Boolean operators (AND, OR). The advanced search attributes of each database were considered and customizing the search syntax was adapted.

Study Selection
In the first step, the title and abstract of all retrieved papers during the electronic and manual search processes were evaluated based on the inclusion criteria. This was followed by examination of the full texts of the potentially relevant articles based on the abovementioned criteria. These processes were performed independently by two reviewers. Initial disagreements about the selection of studies were resolved through discussions.

Data Collection Process and Data Items
Data were extracted and recorded in pre-designed Excel datasheets by two reviewers independently. The following data were abstracted from each study: first-author name; country in which the study had been conducted as well as its income level and development status based on World Bank data; sample size; data collection date; parents and children ages; country location based on WHO regions; type of study; quality of study; and raw data to calculate prevalence of parents' willingness of their children to be vaccinated against COVID-19.

Risk of Bias in Individual Studies
The Newcastle-Ottawa Scale (NOS) was used to assess risk of bias within included studies. This checklist evaluates the methodological quality of observational studies in the following three sections: selection, comparability, and outcome [45,46]. The maximal acquirable score on the NOS checklist is 9 for each study. Studies with less than five points were classified as having a high risk of bias [45]. Methodological quality status was not considered as an eligibility criterion. However, the effect of methodological quality on the pooled effect size was assessed in the subgroup analysis and meta-regression.

Summary Measures and Data Synthesis
The selected summary measure of the present study for meta-analysis was the frequency or prevalence of the acceptance of the COVID-19 vaccine and their 95% confidence intervals (CIs). Numerical evidence regarding the prevalence of the COVID-19 vaccine acceptance was quantitatively synthesized using STATA software version 14. Meta-analysis using a random-effects model was conducted to consider both within-study and betweenstudy variances [47]. Severity of heterogeneity was estimated using the I 2 index [48].
Contributing factors influencing acceptance of COVID-19 vaccines were pooled using a meta-synthesis approach due to methodological heterogeneity of variables and measures.

Risk of Bias across Studies
Funnel plot and Begg's test were used to assess publication bias [49]. Meta-trim with the fill and trim method was used to correct probable publication bias [50]. The Jackknife method was used for sensitivity analysis and probable single study effect on pooled effect size [51].

Additional Analyses
To investigate moderators for COVID-19 vaccine acceptance, subgroup analysis and meta-regression were conducted.

Study Screening and Selection Process
The initial search in four academic databases as well as Google Scholar resulted in retrieval of 8816 records: PubMed (n = 2553); Scopus (n = 1538); WoS (n = 1664); and ProQuest (n = 3061). After removing duplicates (n = 2986), the remaining manuscripts were screened based on their titles and abstracts. Finally, 145 papers appeared to be potentially eligible and their full texts were reviewed. In this process, 98 studies were pooled in the meta-analysis. The search process based on the PRISMA flowchart is illustrated in Figure 1.

Additional Analyses
To investigate moderators for COVID-19 vaccine acceptance, subgroup analysis and meta-regression were conducted.

Study Screening and Selection Process
The initial search in four academic databases as well as Google Scholar resulted in retrieval of 8816 records: PubMed (n = 2553); Scopus (n = 1538); WoS (n = 1664); and ProQuest (n = 3061). After removing duplicates (n = 2986), the remaining manuscripts were screened based on their titles and abstracts. Finally, 145 papers appeared to be potentially eligible and their full texts were reviewed. In this process, 98 studies were pooled in the meta-analysis. The search process based on the PRISMA flowchart is illustrated in Figure 1.   Half of the studies (49 out of 98) were conducted in developed countries with high income (61 out of 98). The smallest sample size was 50 (from the U.S.), and the largest sample size was 227,740 (from Latin America and the Caribbean). Most study respondents (68.1%) were mothers. The mean age of parents was 39.10 (range between 18 and 70 years) and that of their children was 8.45 years (range between 0 and 18 years). Almost all studies used a cross-sectional design, with only two studies adopting a longitudinal design. The first study was conducted during February 2020 in China and the last one was conducted during January 2022 in Saudi Arabia. Table 1 provides the summary characteristics of all included studies.

Methodological Quality Appraisal
Most studies (58 out of 98) were categorized as being low-quality (or high risk of bias) studies. The total score of methodological quality is provided in (Table 1), with details in (Figure 2). The main methodological problems were: (1) description of the response rate or the characteristics of the responders and the non-responders not having been reported (94 out of 98 studies); (2) explanation regarding sample size estimation and justification not having been reported (74 out of 98 studies).

Methodological Quality Appraisal
Most studies (58 out of 98) were categorized as being low-quality (or high risk of bias) studies. The total score of methodological quality is provided in (Table 1), with details in (Figure 2). The main methodological problems were: (1) description of the response rate or the characteristics of the responders and the non-responders not having been reported (94 out of 98 studies); (2) explanation regarding sample size estimation and justification not having been reported (74 out of 98 studies).

Pooled Prevalence of COVID-19 Vaccine Willingness
The pooled estimated prevalence of parental acceptance to vaccinate their children with COVID-19 vaccine was 57% (98 studies, 95% CI: 52-62%, I 2 : 99.92%, τ 2 : 0.06). Figure 3 provides the forest plot regarding the pooled prevalence of parental acceptance to vaccinate their children with the COVID-19 vaccine. The probability of publication bias was assessed using Begg's test (p < 0.001) and funnel plot. Based on the asymmetric funnel plot (Figure 4), publication bias seems probable. The fill-and-trim method was used to correct probable publication bias, but no study was imputed, and publication bias was ruled out. Sensitivity analysis (based on the one-out or Jack-knife method) showed that the pooled effect size was not affected by a single-study effect. The probability of publication bias was assessed using Begg's test (p < 0.001) and funnel plot. Based on the asymmetric funnel plot (Figure 4), publication bias seems probable. The fill-and-trim method was used to correct probable publication bias, but no study was imputed, and publication bias was ruled out. Sensitivity analysis (based on the one-out or Jack-knife method) showed that the pooled effect size was not affected by a single-study effect.

Predictor Variables of Parental Willingness
Predictors of COVID-19 vaccine willingness were assessed using subgroup analysis ( Table 2) and univariable and multivariable meta-regression (Tables 3 and 4). Country income level, country location in WHO's regions (i.e., AMR, EUR, WPR, EMR, and SEAR), and data collection method were significant moderators (p = 0.01) of parents' willingness to vaccinate their children with the COVID-19 vaccine. High-income countries had the lowest prevalence of parental willingness (52%) compared to low-and upper-intermediate-income countries (62 and 65%, respectively). The lowest prevalence of parental willingness was observed in countries located in EMR compared to other regions (45% in EMR vs. 58% in AMR, 62% SEAR, 52% EUR, and 67% WPR). Data collection method was another significant variable influencing the pooled estimated parental willingness (p = 0.02). Studies collected data using phone interview had the lowest prevalence of willingness (35%). In the uni-variable meta-regression, data collection time was the only significant variable (p = 0.001) in predicting parental willingness. Multivariable meta-regression revealed that both data collection time (13% decrease in willingness by each increase in month) and country income level (7% decrease by increasing level of country income) were only significant predictors of parental willingness, which explained 11.44% of variance. None of the examined variables affect the heterogeneity.

Predictor Variables of Parental Willingness
Predictors of COVID-19 vaccine willingness were assessed using subgroup analysis ( Table 2) and univariable and multivariable meta-regression (Tables 3 and 4). Country income level, country location in WHO's regions (i.e., AMR, EUR, WPR, EMR, and SEAR), and data collection method were significant moderators (p = 0.01) of parents' willingness to vaccinate their children with the COVID-19 vaccine. High-income countries had the lowest prevalence of parental willingness (52%) compared to low-and upper-intermediate-income countries (62 and 65%, respectively). The lowest prevalence of parental willingness was observed in countries located in EMR compared to other regions (45% in EMR vs. 58% in AMR, 62% SEAR, 52% EUR, and 67% WPR). Data collection method was another significant variable influencing the pooled estimated parental willingness (p = 0.02). Studies collected data using phone interview had the lowest prevalence of willingness (35%). In the univariable meta-regression, data collection time was the only significant variable (p = 0.001) in predicting parental willingness. Multivariable meta-regression revealed that both data collection time (13% decrease in willingness by each increase in month) and country income level (7% decrease by increasing level of country income) were only significant predictors of parental willingness, which explained 11.44% of variance. None of the examined variables affect the heterogeneity.

Contributing Factors of Parental Willingness
Two main categories of contributing factors (i.e., family-related factors and vaccine-related factors) were identified among included studies using a qualitative synthesized approach.
History of COVID infection in parents, children, or family members did not have consistent results among included studies, showing positive association [113,120,127,135], negative association [100,116], or no significant difference [87,88].
Source of information regarding the vaccine influenced parents' decision for their children's vaccination. When they received information from healthcare providers, physicians, or pediatrics, they reported more willingness to vaccinate their children [59,80,84,96,99,128,134,136]. Social media played different roles. In most studies, participants reported that social media and exposure to negative information increased parent hesitancy regarding vaccination [75,93,94,105,106], while some others reported a positive influence of social media for increasing parent acceptance [66,70].
Parents with a higher number of children reported inconsistent decisions regarding willingness to accept their children's vaccination. A higher number of children was associated with less willingness in some studies [38,55,87,105,108], higher willingness in other studies [99,102], or no difference in one study [88].
Parents' higher levels of stress, anxiety, and psychological distress were associated with less intention for children's vaccination [53,78,89,113]. Only one study reported that higher vaccine acceptability was associated with higher levels of anxiety regarding COVID-19 infection [142].
Parents having chronic conditions reported higher prevalence of vaccination willingness for their children in three studies [83,113], while lower willingness [57] and no difference [87] each was reported in one study. Additionally, parents whose children have a history of adverse vaccine reactions and allergies were less willing to vaccinate their children [52].
Parents living in rural and sub-urban areas showed less willingness compared to those in urban areas [105,113,125], except for one study reporting higher willingness of rural parents [52].

Vaccine Related Factors
Parents reported more children's vaccination willingness when they believe that children vaccination is necessary to halt the pandemic and to reach a better national economic situation [65,70,72,77,83,86,97,103,115,120].
Cost of vaccine [56,59,73], vaccine accessibility [59,63,101,123,141], and governmental incentive of giving a green pass after vaccination [101,123] were among of contributing factors for parents' decision-making for their willingness to vaccinate their children.
Parents who perceived the risk of COVID-19 transition from children to adults [60,106,118] and who perceived risk of children's infection and being hospitalized because of COVID-19 [59,92,99,106,115,118,121,127,128,134] reported more willingness to vaccinate their children.
Comparisons between domestic and foreign vaccine preference have been investigated in very few studies, and domestic vaccines were preferred by parents for themselves and their children [56,142].

Discussion
In order to provide thorough estimation regarding parents' willingness to have their children vaccinated, the present systematic review and meta-analysis synthesized data from studies published between December 2019 and July 2022 (98 papers). Apart from using meta-analysis to quantitatively synthesize parents' acceptance rate of having children vaccinated, quantitative (i.e., meta-regression and subgroup analysis) and qualitative approaches have been used to synthesize the factors explaining parents' willingness on their children's vaccination. The synthesized results showed that the pooled estimated rate of parents' willingness was 57% (95% CI = 52-62%). The low willingness to let children get vaccinated concurs with prior results reported by meta-analysis: 60.1% [32]. Although the present meta-analysis also revealed high heterogeneity (I 2 > 90%) as similar to previous meta-analyses [32,33], the 95% CI in the present meta-analysis was narrower (52-62% vs. 25.6-92.2% and 21.6-91.4%). Nevertheless, the parents generally had low willingness to let their children get vaccinated. Consequently, it is important to know the potential reasons increasing or decreasing parents' willingness to let their children get vaccinated. Apart from the quantitative finding, qualitative synthesis in the present review showed that the positive factors on parents' willingness to vaccinate their children were knowledge on the COVID-19 vaccine, trust in the COVID-19 vaccine, and facilitators in vaccination (e.g., low cost, good vaccine accessibility, and government incentive); negative factors were parents' mental difficulties, including worries, anxiety, and psychological distress.
Although having similar point-estimation in the parents' willingness to have their children vaccinated, the present systematic review and meta-analysis had a narrower 95% CI than the two previous systematic review and meta-analyses [36,37]. A potential reason is that the papers included in their meta-analyses [36,37] were fewer than the present one, which resulted in an unstable estimation in the 95% CI. Specifically, Galanis et al. reviewed 44 papers [36] and Chen et al. reviewed 29 papers [37], while the present systematic review and meta-analysis reviewed 98 papers. With a double-size reviewed papers, the present systematic review and meta-analysis is likely to have a more precise estimation in the acceptance rate, especially in the interval-estimation.
Data collection time was a significant factor explaining dropped willingness (13% decreased by each month increased). This can be explained by risk compensation [143,144] and diffusion of responsibility with bystander effect [145,146]. It seems that when time progresses and the percent of vaccinated people increases, parents may feel safe not to let their children get COVID-19 vaccinated as other people have already been vaccinated. This point was pointed in regression analysis based on percent of vaccinated people in countries' national level, which showed that each increased percent in percent of vaccinated people contributes to a 1% decrease in parental willingness. In other words, parents feel that the risk of COVID-19 infection for their children is decreased and they would not like to let their children vaccinated as a type of risk compensation [143,144]. Additionally, when the vaccination rate increases, parents may feel their responsibility of letting their children get vaccinated decreased, which is a phenomenon of diffusion of responsibility and bystander effect [145,146].
Although the parents' willingness to have their children vaccinated, it is important to maintain the vaccination rate across time to adhere to the herd immunity. That is, a decreased vaccination rate may cause another wave of COVID-19 outbreak as documented in the literature. Therefore, governments and health authorities should have appropriate methods to maintain willingness of having children vaccinated among parents. The present systematic review and meta-analysis used the qualitative synthesis to find the importance of parents' COVID-19 vaccine knowledge, trust in COVID-19 vaccine, and facilitators in vaccination (e.g., low cost, good vaccine accessibility, and government incentive) to improve their willingness to have children vaccinated. Moreover, mental health problems (e.g., having worries and with high levels of psychological distress) might reduce parents' willingness to have their children vaccinated. Therefore, government and healthcare authorities should consider building campaigns on COVID-19 vaccine knowledge improvement, COVID-19 vaccine support systems, and psychological distress reduction to elevate the parents' willingness to have their children vaccinated.

Strengths and Limitations
The strengths of the present systematic review and meta-analysis included (i) a comprehensive search of the literature across WHO-defined regions (i.e., included AMR, EUR, WPR, EMR, and SEAR) to cover different ethnic and country populations; (ii) the use of robust methodology in reviewing papers (i.e., using the NOS checklist to evaluate each paper's methodological quality; applying subgroup analysis and meta-regression to examine the moderator effects on parents' willingness of having children vaccinated); (iii) synthesized findings from both quantitative and qualitative approaches to identify all potential factors explaining parents' willingness of having children vaccinated.
There are some limitations in the present systematic review and meta-analysis. First, the COVID-19 pandemic had different severities and progresses across the studied papers because different countries controlled the COVID-19 pandemic with different situations. Therefore, it is hard to control this important confounder when cumulating the empirical data across countries worldwide. Most of the studies did not report data regarding the time window between vaccines availability at national level or time of vaccine approval for different age groups and collecting the data. Second, almost all papers analyzed in the present systematic review and meta-analysis did not use a standardized instrument assessing willingness to children's vaccination. Subsequently, there might have been some measurement biases across the analyzed papers. Third, over half of the papers were at high risk of bias and the findings of the present systematic review and meta-analysis could be affected by the high risk of bias. Therefore, additional studies with low risk of bias are needed to further investigate the issue regarding parents' willingness to have their children vaccinated. Fourth, although the present systematic review and meta-analysis had sought several academic databases (e.g., PubMed) and Google Scholar, the Google's search engine was not used to search for potential grey literature. Therefore, it is possible that some relevant articles might not be included in the present systematic review and meta-analysis. Lastly, the following important information was not able to retrieve from the analyzed studies in the present systematic review and confounded the present findings: definitions of children (and the actual age ranges used for the analyzed studies); available vaccine (brands or types); and the vaccination procedure (e.g., whether the vaccines have been politicized in other countries as they have been in the US).

Conclusions
In conclusion, the present systematic review and meta-analysis updates the current understanding of parents' willingness and hesitancy of letting their children get vaccinated. The willingness of the parents was generally low (mean acceptance rate: 57%; 95% CI: 52-62%), although it was highly heterogeneous (I 2 = 99.92%). Moreover, time appeared to be the primarily significant factor explaining high levels of acceptance. Qualitative synthesized results showed that parents' knowledge on COVID-19 vaccine, trust in COVID-19 vaccine, and facilitators in vaccination (e.g., low cost, good vaccine accessibility, and government incentive) could improve parents' acceptance of children vaccination. In contrast, parents' mental difficulties (e.g., having worries and psychological distress) were barriers to improve their willingness.