Parental Acceptance, Parental Hesitancy, and Uptake of Seasonal Influenza Vaccination among Children Aged 6–59 Months: A Systematic Review and Meta-Analysis

This systematic review and meta-analysis summarises the literature on parental acceptance, parental hesitancy, uptake, and the associated factors of seasonal influenza vaccination (SIV) among children aged 6–59 months. Studies were sourced from the following platforms: PubMed, Web of Science, MEDLINE, and EMBASE databases. A random-effects model was used to evaluate the prevalence and 95% confidence intervals (CI) of parental acceptance, parental hesitancy, and SIV uptake in the last flu season and lifetime among children. A total of 36 studies were included for analysis. The overall prevalence was 64% for parental acceptance (95% CI: 51–75%), 34% for parental hesitancy (95% CI: 22–48%), 41% for SIV uptake in the last flu season (95% CI: 33–50%), and 46% for SIV uptake in a lifetime (95% CI: 20–74%). Associated factors of parental acceptance/hesitancy and uptake included the age of the children or parents, parental education level, household income level, ethnicity, and other modifiable factors, including perceived benefits, perceived barriers, perceived severity, perceived susceptibility, and cues to action related to SIV. Meta-regression analyses revealed regional differences in parental acceptance (Americas: 79% vs. Asia: 60%). The results provided implications informing us of the development of intervention programs targeting parents to improve SIV coverage among young children.


Introduction
Seasonal influenza is a common respiratory tract infection and a public health problem, particularly in children [1]. It is a significant cause of illness in young children aged under five years, who are at a greater risk of complications, which results in an estimated 870,000 hospitalisations and 10,200 deaths across the globe every year [2][3][4][5]. Children aged 6-59 months are more susceptible to seasonal influenza and have more severe consequences than the older groups [6,7]. For instance, children aged 6-59 months had higher rates than those aged 5-59 years in infection (83.7 vs. 10.2-16.4 per 100,000 individuals), hospitalisation (6.7-10.9 vs. 1.7-3.2 per 10,000 individuals), and mortality of seasonal influenza (0.2 vs. 0.02-0.04 per 100,000 individuals) [6,7]. Therefore, effective measures, such as vaccination, to prevent younger children from seasonal influenza infection are of vital importance [6,7].
Seasonal influenza vaccination (SIV) is considered one of the most effective ways to prevent influenza illness and severe complications from influenza virus infection [4,8,9]. It was shown that vaccinated young children had fewer respiratory illnesses and primary care consultations than their unvaccinated counterparts, and SIV could help reduce hospitalisation by 70% [10,11]. It was reported that not only did SIV protect the children, terms. Additionally, the Boolean operators "OR" was used to link each term in each set and "AND" was used to link each set of the terms. The complete search strategy is described as follows: (1) willing* OR accept* OR refus* OR hesitan* OR inten* OR reject* OR declin* OR uptake* OR factor* OR determinant* OR reason* OR facilitator* OR barrier* OR attitude* OR perception* OR view* OR predict* OR enabl* OR knowledge OR perspective*; (2) parent* OR guardian* OR caregiver*; (3) influenza OR flu; (4) vaccin* OR immuni*; (5) child* OR infant* OR toddler* OR pediatr* OR preschool* OR kindergart*; (6) 1 AND 2 AND 3 AND 4 AND 5.

Inclusion and Exclusion Criteria
Peer-reviewed original studies published in English-language journals using quantitative, qualitative, or mixed-method methodology were included in this systematic review. The eligible studies should report parental acceptance, parental hesitancy, uptake, and the associated factors of SIV in children aged 6-59 months. The articles were searched with no restriction on country or region. An article was excluded if it was (1) the development of a study protocol/scale/measurement tool; (2) unclear about whether the participants were parents/caregivers of children; (3) the modelling or projection for prediction of SIV uptake among children; or (4) other grey papers, such as government documents/reports, newspapers, textbooks, book chapters, and preprints. As government documents/reports and newspapers were not written for scientific purposes, they were excluded from consideration. In view of the fact that grey papers were generally published without peer review, there are concerns regarding the quality and reliability of these publications.

Data Extraction
Two independent reviewers (P.S.-f.C. and J.K.) carried out data extraction and reference searching parallelly. Five papers were randomly selected for initial data extraction. The two reviewers discussed the results of these five papers, and any discrepancies arising were resolved through consensus. After that, they continued to work on the data extraction of the rest of the papers. After finishing data extraction of all papers, they discussed all results, and consensus was achieved through discussions. A data extraction form was used, which included reference ID, first author, publication year, study year, title, country, study design, participants, sample size, parental acceptance or hesitancy rates, uptake rates in the previous flu season or lifetime among children aged 6-59 months, and other relevant findings on associated factors, attitude, and perception.

Quality Assessment
The information from the included articles was assessed using the Mixed-Method Appraisal Tool with detailed descriptions of the rating [31]. For quantitative studies, there are four assessment criteria: (1) whether the research question can be addressed using the sampling strategy; (2) whether the sample is representative of the population under study; (3) whether appropriate measurements are used, e.g., clear origin, reliability, and validity; (4) whether there is an acceptable response (at least 60%). For qualitative studies, there are also four assessment criteria: (1) whether the sources of qualitative data are relevant to address the research question; (2) whether the process for analysing qualitative data is relevant to address the research question; (3) whether appropriate consideration is given to how findings relate to the context, e.g., the setting, in which the data were collected; (4) whether appropriate consideration is given to how findings relate to researchers' influence, e.g., via their interactions with participants. For mixed-method studies, in addition to the aforementioned criteria for the quantitative and qualitative studies, there are other three criteria: (1) whether the mixed-method research design is relevant to address the qualitative and quantitative research questions; (2) whether the integration of qualitative and quantitative data (or results) is relevant to address the research question; (3) whether appropriate consideration is given to the limitations associated with this integration, e.g., the divergence of qualitative and quantitative data. For each criterion, "Yes" or "No" is given as the rating [31]. Two authors (P.S.-f.C. and J.K.) independently assessed the quality of the included studies. Similar to data extraction, five papers were randomly selected for initial quality assessment. The two reviewers discussed the results of these five papers, and any discrepancies that arose were resolved through consensus. After that, they continued to work on the quality assessment of the rest of the papers. After finishing the quality assessment of all papers, they discussed all results, and consensus was achieved through discussions.

Statistical Analysis
Frequencies and percentages were used to summarise the characteristics of included studies. Random-effects models were used to generate the pooled uptake rates of children aged 6-59 months in the last flu season and lifetime, parental acceptance, and parental hesitancy. Q-test and I 2 test were employed to assess heterogeneity among studies. The parental hesitancy rate was computed using the formula (100-parental acceptance rate) for studies that only provided information on the parental acceptance rate. Likewise, the parental acceptance rate was computed using the formula (100-parental hesitancy rate) for studies that only provided information on the parental hesitancy rate.
In order to determine if study characteristics could explain the variability between studies, subgroup analyses and meta-regression were conducted. This included study year (2010 or before, 2011-2015, 2016-2019, and 2020 or after), region (Asia, Americas, Europe, and Oceania), and sample size (≤1000 and >1000). In order to test whether the uptake of SIV in the last flu season and lifetime among children aged 6-59 months, parental acceptance, and parental hesitancy varied according to the selected study characteristics, univariate meta-regression was performed. Study characteristics were then included in the multivariable meta-regression if they were significant (p < 0.05) in the univariate analysis.
The presence of publication bias was visualised using funnel plots, and Egger's linear regression test was performed. Meta-prop package (method = Inverse and summary measure = PLOGIT) was adopted in R Studio (version 4.2.1, https://cran.r-project.org/ bin/windows/base/old/4.2.1/, accessed on 1 May 2023) to conduct the meta-analyses.
Additionally, other key findings on associated factors of parental acceptance, parental hesitancy, and uptake of SIV in the last season and lifetime among children aged 6-59 months, as well as parental attitudes and perceptions, were summarised thematically.

Identification of Studies
The initial search returned 3901 citations, and 2242 remained after excluding duplicates ( Figure 1). After that, we further removed 2179 articles after screening for titles and abstracts. We performed full-text screening on 63 articles, of which 27 articles were excluded because they did not meet the selection criteria. The initial search returned 3901 citations, and 2242 remained after excluding duplicates ( Figure 1). After that, we further removed 2179 articles after screening for titles and abstracts. We performed full-text screening on 63 articles, of which 27 articles were excluded because they did not meet the selection criteria.

Overview of Included Studies
A total of 36 studies published from 2004 to 2023 were finally analysed and summarised [14][15][16][17][18][19][20][21][23][24][25][26][27][28] (Table 1). Most of the studies were performed in Asia (n = 15, 42%) and the Americas (n = 14, 39%), followed by Europe (n = 5, 14%) and Oceania (n = 2, 6%). These studies consist of 33 quantitative studies, 2 qualitative studies, and 1 mixedmethod study. The total sample size is 68,567. Regarding study years, 14 studies were conducted before 2011, 9 were between 2011 and 2015, 10 were between 2016 and 2019, and 3 were in 2020 or after. Regarding the overall quality of the included studies, over 75% of them satisfied each criterion, except for the response rates in quantitative studies and issues related to researchers' influence in qualitative studies. Less than half of the quantitative studies had a response rate higher than 60%. All qualitative studies did not have appropriate consideration given to how findings relate to researchers' influence. (Supplementary Table S1A

Overview of Included Studies
A total of 36 studies published from 2004 to 2023 were finally analysed and summarised [14][15][16][17][18][19][20][21][23][24][25][26][27][28] (Table 1). Most of the studies were performed in Asia (n = 15, 42%) and the Americas (n = 14, 39%), followed by Europe (n = 5, 14%) and Oceania (n = 2, 6%). These studies consist of 33 quantitative studies, 2 qualitative studies, and 1 mixedmethod study. The total sample size is 68,567. Regarding study years, 14 studies were conducted before 2011, 9 were between 2011 and 2015, 10 were between 2016 and 2019, and 3 were in 2020 or after. Regarding the overall quality of the included studies, over 75% of them satisfied each criterion, except for the response rates in quantitative studies and issues related to researchers' influence in qualitative studies. Less than half of the quantitative studies had a response rate higher than 60%. All qualitative studies did not have appropriate consideration given to how findings relate to researchers' influence (Supplementary Table S1A-C).   • Facilitators included caregivers who had a bachelor's degree or above, an annual household income of RMB <20,000, confidence in the importance, safety, and effectiveness of influenza vaccine, knowing other caregivers were vaccinating their children, and receiving positive influence from healthcare workers and family members. • Barriers included poor access, including conflicts between caregivers' availability and vaccination service schedules and inconvenient transportation to the vaccination site.  • Being a mother, having a child aged ≥48 months, and the absence of IV experience in family members were the significant background factors of parental intention in the ever-vaccinated group, while being a mother, younger, and more educated were significant factors of parental intention in the never-vaccinated group. • Perceived susceptibility, perceived benefit, perceived barrier, cue to action, subjective norm, and having family members vaccinated were associated with parental intention for ever-vaccinated children's IV, while only perceived benefit and subjective norm were significant for never-vaccinated children. • Associated factors: perceived doctor's recommendation, relatives and friends thinking that child should have a flu shot, the belief that a child with asthma should receive a flu shot, the belief that receiving an influenza shot is a smart idea, the belief that unvaccinated child will get the flu, and belief that flu shot prevents flu, worrying that child will get the flu from the flu shot, and the belief that the flu shot is more trouble than it is worth. • Facilitators included knowledge and awareness of IV, trust in the vaccine, access to vaccine information, physician recommendation, access to vaccination, perceived severity, social norms, the belief that vaccinating a child protects the wider family, and the belief that vaccinating a child is a social responsibility.

•
Barriers included a lack of knowledge about the flu, the belief that one's child is not at risk from the flu, and the belief that catching the flu is inevitable for children, inconvenient appointments, lack of access due to a vaccine shortage, and lack of awareness about their child's eligibility, and the fear of side-effects and frustration with or distrust of medical professionals. • More children who were 6 to 23 months than those who were 24 to 59 months of age were vaccinated. • Children with chronic medical conditions were more likely to be vaccinated than healthy children who were 24 to 59 months of age.

Primary Findings
The pooled prevalence of parental acceptance, parental hesitancy, and the uptake of SIV in the last flu season and lifetime among children aged 6-59 months are presented in Figure 2A-D (forest plots), the associated factors are shown in Table 2, and parental attitudes and perceptions are presented in Table 1.

SIV Uptake in the Last Flu Season among Children Aged 6-59 Months
Twenty-three studies reported on the SIV uptake in the last flu season (Figure 2A). The overall uptake rate was 41% (95% CI: 33-50%), with a range of 9% to 76%, and had a significant heterogeneity (I 2 = 99%, p < 0.01). The highest SIV uptake was reported in the United States (76%), followed by Argentina (73%), while the lowest uptake was observed in Thailand (9%).
To determine if the study year, area, and sample size could contribute to the observed heterogeneity among studies, meta-regression and subgroup analyses were conducted. Univariate meta-regression revealed that there was no significant association between SIV uptake in the last flu season and study year, region, and sample size. In subgroup analyses, since all heterogeneity analyses were significant (p < 0.01), no variability could be explained (Table 3 and Supplementary Figure S1A-C).
Publication bias among studies for SIV uptake in the last flu season was assessed using the funnel plot and Egger's test. The funnel plot revealed that there was a probable publication bias toward studies with higher uptake rates using visual inspection ( Figure 3A). However, Egger's test showed no significant publication bias (p = 0.05). The adjusted uptake rate was 51% (95% CI: 41-62%) after filling in six missing studies by performing the trim-and-fill analysis.   Acceptance: household income (+); perceived vaccine safety (+); perceived benefit (+); cues to action (+); perceived support from significant others (+)  USA 153 --78% 22% Acceptance: perceived severity (+); perceived vaccine safety (+); perceived importance to vaccinate the child (+); parental education level (+)        Publication bias among studies for SIV uptake in the last flu season was assessed using the funnel plot and Egger's test. The funnel plot revealed that there was a probable publication bias toward studies with higher uptake rates using visual inspection ( Figure  3A). However, Egger's test showed no significant publication bias (p = 0.05). The adjusted uptake rate was 51% (95% CI: 41-62%) after filling in six missing studies by performing the trim-and-fill analysis.
Univariate meta-regression revealed that there was no significant association between SIV uptake in the lifetime and the study year, region, and sample size. In subgroup analyses, however, only the study year had no significant heterogeneity, which means that only the study year explained some of the variability among studies (Table 3 and Supplementary Figure S2A-C).
Concerning publication bias, the funnel plot showed a probable publication bias toward studies with higher uptake rates using visual inspection ( Figure 3B). However, Egger's test found no significant publication bias (p = 0.34). Moreover, no adjusted uptake rate could be generated after performing the trim-and-fill analysis.
Univariate meta-regression revealed that there was a significant association between parental acceptance and hesitancy and study region (p = 0.04, p = 0.02). The Americas had a higher parental acceptance than Asia (79% vs. 60%) and lower parental hesitancy than Asia (17% vs. 40%). In subgroup analyses, since all heterogeneity analyses were significant (p < 0.01), no variability could be explained. (Table 3 and Supplementary Figures S3A-C and S4A-C).
The funnel plot revealed a probable publication bias toward studies with low acceptance rates and high hesitancy rates ( Figure 3C,D). However, Egger's test showed no significant publication bias in parental acceptance and parental hesitancy (p = 0.39 and 0.35). Nevertheless, the adjusted acceptance and hesitancy rates were 46% (95% CI: 30-62%) and 54% (95% CI: 37-71%) after filling in six missing studies in acceptance and hesitancy using the trim-and-fill analyses.

Parental Attitudes and Perceptions of SIV
A total of 17 studies reported parental attitudes and perceptions of SIV (Table 1). One revealed that the majority of the parents (about 90%) were aware of SIV [46], but another qualitative study showed that parents lacked knowledge about SIV [49]. The cost of the vaccine was an important issue mentioned in several studies [17,23,24,37,43]. SIV was perceived as safe, effective, and important [14,17,18,[23][24][25]28,37,43,45,49]. However, worries about side effects and doubts about vaccine effectiveness were also reported [17,18,[23][24][25]28,33,37,45,49]. The convenience of taking up SIV was also frequently discussed, with parents in some studies reporting that it was easy to take up SIV [17,24,28,45], but inconvenience was also reported in a few studies [37,49]. The belief that SIV was the social norm (e.g., many parents arranged SIV for their children) was revealed in various studies [18,[23][24][25]49]. Nonetheless, worrying that their children could catch the flu from the vaccine was mentioned by parents [26][27][28]43]. Studies showed that parents were aware that the government recommended SIV for children aged less than 5 years or that physicians gave them the recommendations [14,[23][24][25]27,33,37,45,49]. On the other hand, parents also thought that it was too young for their child to receive the SIV, and the vaccine would have a negative effect in the interactions with other vaccines to be received by the child [23][24][25]. Parents also had a high perceived self-efficacy in receiving SIV, as they believed that they would be able to get their child vaccinated if they desire to do so [23][24][25].

Discussion
To our knowledge, this is the first systematic review and meta-analysis to summarise evidence on SIV among children aged 6-59 months. This study revealed the pooled prevalence of parental acceptance, parental hesitancy, and the uptake of SIV in the last flu season and lifetime among children aged 6-59 months. Then, it assessed factors associated with SIV and parental attitudes and perceptions of SIV. Moreover, a reproducible search strategy with a well-established keyword system was employed for the identification of studies from key databases. All these are contributions to the literature.
Although 64% of the parents were willing to vaccinate their children, only 41% received SIV in the last flu season. This uptake rate is much lower than the target influenza vaccine coverage rate (VCR) of 75% recommended by the WHO [13]. This uptake is also slightly lower than that of their older counterparts (45.1-52.8%) [54,55]. This shows that there is a large room for improvement in the uptake of SIV in young children. Specific barriers to parents' acceptance of SIV in their young children included their concerns that it was too young for their children to receive SIV and that SIV would have a negative effect in interaction with other vaccines to be received by the child [23][24][25]. Future SIV promotion programs should address these specific barriers in order to reduce parents' concerns. For instance, supporting evidence emphasizing the safety of SIV in young children should be included in the promotion programs.
Additionally, meta-regression results showed that parental acceptance/hesitancy varied in different regions. The Americas had a higher acceptance than Asia (79% vs. 60%) and lower hesitancy than Asia (17% vs. 40%). The observed differences could be explained by the factors at the individual level and policy level. At the individual level, this study showed that a higher number of studies conducted in Asia (n = 6) than Americas (n = 3) reported perceived side effects of SIV as an important factor for acceptance of SIV. Additionally, the parental belief that it is too young for their child to receive SIV and SIV would have a negative effect in interaction with other vaccines to be received by the child were only reported in studies conducted in Asia [23][24][25]. Furthermore, these findings may be explained by the different vaccination policies in various countries. In the United States and Canada, SIV is included in the subsidised childhood immunisation scheme and free for children aged 6-59 months [56,57]. However, in countries in Asia, such as China and Singapore, SIV is not part of the subsidised childhood immunisation scheme, and it needs to be paid for out of pocket [40,46]. Health authorities in these Asian countries should consider providing subsidies for children aged 6-59 months to receive SIV.
Associated factors of the uptake/acceptance/hesitancy of SIV could be mainly categorised into individual and interpersonal levels based on the socioecological model [58]. Factors at the individual level included sociodemographics, such as the age of the children or parents, parental education level, household income level, ethnicity, and others. Older parents, lower parental education level, and household income were more consistently found to be negatively associated with uptake and acceptance. Therefore, in order to achieve the VCR (75%) recommended by the WHO [13], for successful vaccination programs in the future, such sociodemographic factors should be considered and incorporated when designing health promotion programs or targeted interventions. Other factors at the individual level also included some modifiable factors, such as perceived benefits, perceived barriers, perceived severity, and perceived susceptibility. Previous studies showed that interventions targeting modifiable factors (e.g., perceived benefits and perceived barriers) are effective in increasing vaccination uptake [59][60][61]. Factors at the interpersonal level included the influences of others (e.g., relatives/friends/spouses/health professionals). The study results consistently found that support from significant others and recommendations from health professionals to vaccinate their children had positive influences on parental decisions. Therefore, involving significant others and health professionals should be one of the core components in future health promotion programs or interventions for SIV among children.
This study also attempted to compare the uptake/acceptance/hesitancy of SIV before and during the coronavirus disease 2019 (COVID-19) pandemic. The interruptions in vaccination service provision (e.g., COVID-19 control measures and changes in the priority of health services) might play a role in the changes in the uptake/acceptance/hesitancy of SIV among children. Nonetheless, it did not show significant differences in the uptake/acceptance/hesitancy of SIV in this study. However, such results should be interpreted cautiously because there was a limited number of studies (n = 3) that were conducted during the COVID-19 pandemic, and thus, the impact of COVID-19 on children's SIV was unclear. On the other hand, individual studies including older children (>59 months old) consistently reported a significant increase in parental acceptance (1.48-3.43-fold higher) after the COVID-19 outbreak [62][63][64]. For instance, one Chinese study found that parents were more willing to vaccinate their children against seasonal influenza than the time before the COVID-19 pandemic (68.4% vs. 35.2%) [62]. In Switzerland, compared to the time before the COVID-19 pandemic, about a two-fold increase in parental acceptance of SIV for their children was observed during the pandemic (17% vs. 7.2%) [63]. A similar trend was also reported among parents in Saudi Arabia [64]. Nevertheless, the above studies included children aged older than 59 months. Whether there is a change in the uptake/acceptance of SIV for children aged 6-59 months is still unclear.
Some studies suggested that certain strategies or interventions could increase SIV among children, including those conducted after the COVID-19 outbreak. One study showed that increasing the frequency of recommendations to receive SIV by health professionals could significantly increase SIV uptake among children [65]. Parental acceptance of SIV could be increased by offering free vaccination [62]. A review study revealed that vaccination reminders and education directed at either parents or healthcare providers, as well as via vaccination-related clinic process changes significantly improved SIV by 60% among children [66]. Several studies were conducted during the COVID-19 pandemic attempting to maintain or enhance children's vaccination. These studies suggested using innovative approaches may be useful strategies for the post-pandemic period, such as using utilizing online technologies to remind parents and update them parents about their children's vaccination status and remind them about the vaccination schedule [67][68][69], providing phone counselling interventions for children's vaccination in parents [70], using telemedicine to provide necessary health information to assist parents in achieving selfefficacy and making decisions for their children's vaccination [67][68][69][70][71][72][73], and using a chatbot real-time consultation messenger service to increase in parents' motivation and self-efficacy to vaccinate their children [73].
Although this study provides valuable insights into parental acceptance/hesitancy and SIV uptake in the last flu season and lifetime among children aged 6-59 months, it has several limitations. First, high heterogeneity existed regarding the pooled prevalence of parental acceptance/hesitancy and uptake of SIV among children aged 6-59 months. The high heterogeneity could not be fully addressed despite subgroup analysis being performed. Nonetheless, these studies were from different countries/regions with different vaccination policies, and various types of study methodologies were used, which made heterogeneity inevitable. Second, data collected in the included studies in this review were self-reported, and recall and social desirability bias might exist. Third, since all studies were cross-sectional, no causal relationship between uptake/acceptance/hesitancy with factors could be established. Fourth, the exclusion of government documents/reports and newspapers without peer review might be justifiable for scientific rigour, but it is acknowledged that this decision might limit the inclusion of valuable information.

Conclusions
This study revealed the pooled prevalence of parental acceptance, parental hesitancy, and the uptake of SIV in the last flu season and lifetime, as well as their associated factors among children aged 6-59 months. The results revealed that the uptake of SIV in the last flu season, SIV uptake in a lifetime, parental acceptance, and parental hesitancy, were 41%, 46%, 64%, and 34%, respectively. Associated factors of the uptake, acceptance, or hesitancy included the age of the children or parents, parental education level, household income level, ethnicity, and other modifiable factors, including perceived benefits, perceived barriers, perceived severity, perceived susceptibility, and cues to action related to SIV. In order to achieve the VCR (75%) recommended by the WHO to protect children from vaccinepreventable diseases, it is crucial to implement tailored SIV interventions or programs for parents that address the factors, parental attitudes/perceptions, and specific local contexts in various regions or countries.