Parents’ intention to vaccinate their preschool children against COVID-19: Combining the health belief model and the theory of planned behavior

ABSTRACT

The vaccination rate of COVID-19 in preschool children is low, and parents’ intention to vaccinate their children is also low due to multiple factors. This study aimed to establish an integrated model based on the Health Belief Model (HBM)and Theory of Planned Behavior (TPB), to explore the factors influencing parents’ intention to vaccinate their preschool children with the first and second doses of COVID-19 vaccines. A total of 1264 parents of preschool children from 10 kindergartens participated in this study. Hierarchical multiple logistic regression was used to analyze the intention separately. For the integrated model with the first dose of vaccine of COVID-19, introducing the HBM variable in model 1 explained 33.98% of the variance (F = 398.71, p < .001), then upon adding the TPB variable in model 2, the explanation of variance increased to 41.93% (F = 491.94, p < .001) and parents’ intention were associated with their perceived barriers, cues to action, and subjective norms. For the integrated model with the second dose of vaccine of COVID-19, introducing the HBM variable in Model 1 explained 23.76% of the variance (F = 68.82, p < .001), then upon adding the TPB variable in model 2, the explanation of variance increased to 26.67% (F = 77.24, p < .001), and parents’ intention was associated with cues to action and subjective norms. The combination of the two theories improves the explanatory power of parents’ intention to vaccinate their preschool children against COVID-19, and provides a basis for the development of effective interventions for vaccination of COVID-19 for preschool children.

Introduction

The proportion of COVID-19 cases among children has been increasing in the total infected population.¹ In July 2021, children under the age of 18 in the United States accounted for 14.2% of the total number of infections, and in December 2022 it was 18.2%.² The hospitalization rate of children with COVID-19 is low.³ Studies have shown that most of the hospitalizations of infected children are due to concurrent multisystem inflammatory syndrome.⁴ Although the hospitalization rate of children infected with COVID-19 is low, a large number of deaths are still reported.⁵

The COVID-19 vaccine has demonstrated good safety and efficacy for preschool children.⁶ A study compared the effectiveness of partially and fully vaccinated in children 5 to 11 years against the omicron variant of the virus. The effectiveness of the vaccine against omicron infection in fully vaccinated and partially vaccinated children was 65.3% (95% CI, 62.0 to 68.3) and 24.3% (95% CI, 19.5 to 28.9), respectively.⁷ Besides, vaccination can reduce the risk of hospitalization related to COVID-19 in preschool children. For every 100,000 unvaccinated children aged 5 to 11 years, an average of 19.1 were hospitalized for omicron infection, and for vaccinated children of the same age, an average of 9.2 were hospitalized after diagnosis.⁸ Currently, the vaccination rate for children against COVID-19 is relatively low.⁹ According to data from the US CDC on COVID-19 vaccination rates among children in various states nationwide, as of November 30, 2022, only 10% of US children under the age of five had received one dose of the vaccine, and 31% of 5 to 11 years old children had completed two doses of the COVID-19 vaccine.¹⁰ As of January 4, 2023, 23.82% of children under the age of three in Hong Kong have received one dose of the COVID-19 vaccine.¹¹ Since November 2021, COVID-19 vaccination efforts targeting children aged 3 to 11 years have been progressively promoted in various regions of mainland China. The grassroots healthcare service center in Minhang District, Shanghai, is responsible for providing free two-dose COVID-19 vaccinations to preschool children in the area.¹² Vaccination in this campaign is not mandatory, and the education department will collaborate with the health department to provide guidance to schools and kindergartens for effective communication and notification. Parents are encouraged to take the initiative in bringing their children to designated vaccination locations for immunization. Numerous studies indicate that parents have a low intention to vaccinate their children against COVID-19, and several factors influence parents’ intention to vaccinate their preschool children, including parents’ sociodemographic characteristics,¹³ children’s susceptibility to COVID-19,¹⁴ and external factors.^15,16 To explore these factors, some studies have employed theoretical models for explanation. Using a single theoretical model to explain children’s COVID-19 vaccination intention has limitations, prompting researchers to explore combining various theories to construct a more comprehensive theoretical system.^17–19

The Health Belief Model (HBM) is one of the most widely used models in vaccine behavior, particularly in the study of vaccination behavior for epidemics such as influenza,²⁰ hepatitis B²¹ and in research on COVID-19 vaccine behavior.^22,23 It emphasizes the role of psychological perception in health behavior change.²³ In the context of vaccine decision-making, parents’ intentions to vaccinate their preschool children against COVID-19 can be explained by perceived susceptibility, perceived severity, perceived benefits, perceived barriers, and cues to action.²³ In contrast, Theory of Planned Behavior (TPB)is another theoretical model used to predict an individual’s behavior in terms of intention.¹⁷ The theory states that when individuals perceive a behavior as positive (attitudinal), know that significant others want them to perform the behavior (subjective norms), and perceive that the behavior is under their own volitional control (perceived behavioral control), then the individual will perform the behavior.¹⁷

However, whether using the HBM model alone or the TPB model alone, there are inherent limitations in investigating the influencing factors of parents’ intention to vaccinate preschool children against COVID-19.²³ HBM lacks a measure of the influence of important others on health behavior, such as the impact of teachers or elderly people living with preschool children on the behavior of vaccinating against COVID-19.¹⁹ TPB is considered to lack more comprehensive cognitive variables related to its constituent variables, such as the lack of measures of the perceived benefits and barriers of vaccinating preschool children against COVID-19, and the model does not explain how general motives serve as an information source to guide social cognitive processes.²⁴ Therefore, the integration of these two models not only provides a more comprehensive elucidation of health behaviors but also offers additional insights for the management of health behaviors.²⁴ The combination of the HBM model and the TPB model has been used in other studies of health behaviors. For example, one study constructed the integrated model to explain myopia prevention behavior.²⁴ In addition, some researchers have also proposed the comprehensive application of TPB and the HBM to explain the influencing factors of COVID-19 vaccination in adults.¹⁹

This study is based on an integrated model of HBM and TPB to explore the factors influencing parents’ intention to vaccinate preschool children against COVID-19, aiming to explore effective intervention measures to improve the vaccination rate of preschool children.

Materials and methods

Study design and data collection

This cross-sectional study recruited parents of preschool children from ten kindergartens in Minhang District, Shanghai, in October 2022. According to the calculation method for sample size in cross-sectional studies, with a vaccine intention rate of 28.9%,¹³ a significance level of 0.05, and an allowable error of 0.0289, the sample size required was at least 984 participants. Considering an inefficiency rate of 20%, we aimed to recruit no fewer than 1,181 participants. We used cluster sampling to recruit parents from ten kindergartens in Huacao Town, Minhang District. There are 3396 preschool children in all kindergartens in Huachao Town. We created an online survey questionnaire using the Questionnaire Star platform. The survey questionnaire was released on WeChat, a Chinese social media platform, to the WeChat groups of each kindergarten on October 30, 2022. Kindergarten teachers forwarded the survey link to the WeChat groups of their respective classes and briefly introduced the purpose of the study to parents, who then filled out the survey online. As of November 6, 2022, we received 1,898 completed questionnaires. The survey questionnaire first asked parents about their preschool children’s COVID-19 vaccination status. Of the 1,264 parents who had not vaccinated their preschool children or had not completed their children’s vaccination, 634 reported that their children had completed the full course of vaccination, and the remaining 1,264 parents were surveyed about their intention to vaccinate their preschool children against COVID-19. To ensure the quality of the survey responses, we set up two quality control questions. The questionnaire was considered valid only when both quality control questions were answered correctly. In addition, the questionnaire was set up to be submitted only after all questions were answered, and each account could only submit one questionnaire. Before the survey began, all participants were informed that they would be providing data for a scientific study and that all data would be kept confidential and anonymous and used only for research purposes. This study obtained the participants’ consent and their responses to the questionnaire. The study was approved by the Institutional Review Board of the School of Public Health, Fudan University (International Registration Number: IRB00002408 and FWA00002399).

Measurement

Socio-demographic factors

Data were collected from each participant on the age of the parents, the child’s kindergarten year, the relationship between participants and child, the education level of the parents, monthly household income per capita, whether the child had allergic diseases and whether the child had congenital diseases.

HBM variables

The HBM scale developed by the investigators was used to measure five dimensions of perceived susceptibility, perceived severity, perceived benefits, perceived barriers, and cues to action.¹⁸ In this scale, three items measuring perceived susceptibility were developed referring to corresponding items in previous studies “The chance of my children getting the SARS-CoV-2 in the next few months is high” that measure the participants’ level of cognitive awareness of the likelihood of their children being infected with COVID-19 and the possibility of serious consequences such as severe illness or death.²⁵ Perceived severity measures parents’ concerns about the consequences of their children being infected with COVID-19.²³ Perceived benefits refers to previous entries “I believe that COVID-19 vaccine will have high efficacy in preventing significant suffering and complications of the disease”, measures participants’ awareness of the effectiveness of the COVID-19 vaccine for children, including four items: reducing the risk of infection, avoiding severe illness or death, etc. Perceived barriers include four items that measure participants’ degree of doubt about the COVID-19 vaccine, three of which were adapted from a study on adults’ intention to receive the COVID-19 vaccine,¹⁹ including the speed of vaccine development, the quality of the vaccine, and the side effects of the vaccine. The cues to action dimension has seven items, four of which were adapted from similar studies,²⁶ and three of which were self-developed, emphasizing the influence of external factors such as health departments, professionals, kindergartens, and people around them (family, friends, etc.) on parents’ intention to vaccinate their preschool children against COVID-19. Participants were asked to rate all HBM items on a 5-point Likert scale (from ‘‘1 = strongly disagree” to ‘‘5 = strongly agree”), and the item scores were averaged to obtain the total scores for each dimension. Higher scores indicate a higher level of a specific dimension. The Cronbach’s alpha coefficient of the HBM scale in this study was 0.85, indicating good internal consistency. The items included in the HBM model are presented in the Appendix.

TPB variables

The self-designed TPB questionnaire was used to measure three dimensions: attitude, subjective norm, and perceived behavioral control.²⁶ The attitude dimension includes two items adapted from a study on the general population’s intention to vaccinate against COVID-19.¹⁷ The items are “I think the COVID-19 vaccine has a good protective effect on children” and “I think it is necessary for children to be vaccinated against COVID-19 to prevent infection.” The dimension measures the parent’s attitude toward the safety and necessity of the COVID-19 vaccine. The subjective norm dimension includes three items and measures the degree to which important others, including family, friends, and kindergarten teachers, support the vaccination of children against COVID-19. The perceived behavioral control dimension is measured using a single item: “I can decide whether to vaccinate my child against COVID-19.” Participants were asked to rate all TPB items using a 5-point Likert scale (from ‘‘1 = strongly disagree” to ‘‘5 = strongly agree”), and the item scores were averaged to obtain the total score for each dimension. Higher scores indicate a higher level of the specific dimension. The Cronbach’s alpha coefficient of the TPB questionnaire in this study was 0.87, indicating good internal consistency. The items included in the TPB model are presented in the Appendix.

Intention

We evaluated the intention to vaccinate against COVID-19 by asking parents whether they were willing to vaccinate their children.²⁷ For parents who had never vaccinated their children against COVID-19, we asked about their intention to vaccinate their children for the first dose (referred to as “intention 1” below). For parents who had already vaccinated their children with the first dose of the COVID-19 vaccine, we asked about their intention to vaccinate their children with the second dose (referred to as “intention 2” below). The answer options were 1 = strongly unwilling, 2 = unwilling, 3 = willing, 4 = strongly willing.

Statistical analysis

Descriptive statistics were used to summarize the socio-demographic characteristics of the study subjects. Qualitative variables were expressed as numbers and percentages, and quantitative variables were expressed as means and standard deviations. Cronbach’s alpha was calculated to test the internal consistency of each dimension in the HBM and TPB models. Correlations among study variables were tested using chi-square test, t-test, and Pearson correlation analysis. Following previous studies,^13,27 we transformed the responses regarding intention into binary variables (1 = willing to vaccinate, 0 = unwilling to vaccinate), with parents who were strongly willing or willing to vaccinate their children classified as “willing to vaccinate,” and those who were strongly unwilling or unwilling to vaccinate their children classified as “unwilling to vaccinate.” Hierarchical multiple logistic regression was used to analyze the intention of preschool children to receive the first and second doses of the COVID-19 vaccine separately. After controlling the significant socio-demographic variables in the single-factor correlation analysis, input HBM variables in model 1, including perceived susceptibility, perceived severity, perceived benefits, perceived barriers, and cues to action. Then input TPB variables in model 2, including attitude, subjective norms and perceived behavioral control.TPB First, HBM variables can explain additional variation after considering sociodemographic factors. Second, the increase in variance after considering individual factors and HBM variables can be explained by TPB variables. Multicollinearity was checked using the variance inflation factor (VIF) and tolerance tests. Tolerance values and VIF values for all variables did not violate the assumption of multicollinearity (tolerance values > 0.1 and VIF values < 10). All statistical analyses were performed using IBM SPSS Statistics 20 software (IBM Corp.), with a two-tailed significance level set at 0.05.

Results

Participants’ sociodemographic characteristics

Table 1 summarizes the sociodemographic characteristics of the participants. A total of 1,098 valid questionnaires were received for this study, with a valid response rate of 86.87%. According to the survey,863 preschool children had never received the first dose of the COVID-19 vaccine. Among them, more than two-thirds of the parents (67.1%) were between 30 and 40 years old, 50.3% of the preschool children attended primary classes in kindergarten, and most of the parents who completed the questionnaire (74.9%) were mothers. More than half of the parents (58.3%) had a college or undergraduate education, 44.6% of the households had a monthly per capita income more than 10,000 yuan, 84.6% of the children had no allergic diseases, and 98.8% of the children had no congenital diseases. A total of 361 parents were willing to vaccinate their preschool children with the first dose of COVID-19 vaccine, while 502 parents were unwilling to vaccinate their children with the first dose, resulting in an intention rate of 41.8%. Univariate correlation analysis of all sociodemographic variables found that parents’ willingness to vaccinate their preschool children with the first dose of COVID-19 vaccine was significantly correlated with parental age (p < .001), children’s kindergarten year (p = .028), and parental education level (p < .001), monthly household income per capita (p < .001), and whether children have allergic diseases (p < .001) were significantly correlated.

Table 1.

Cross‑tabulation of socio‑demographic factors associated with COVID‑19 vaccination among children.

	Intention 1				Intention 2
	Total n = 863 n(%)	Willing n = 361 n(%)	Unwilling = 502 n(%)	p	Total n = 235 n(%)	Willing n = 163 n(%)	Unwilling n = 72 n(%)	p
Parental age				<.001				.005
<30 years	231(26.8%)	123(53.2%)	108(46.8%)		74(31.5%)	62(83.8%)	12(16.2%)
30–40 years	579(67.1%)	216(37.3%)	363(62.7%)		149(63.4%)	94(63.1%)	55(36.9%)
>40 years	53(6.1%)	22(41.5%)	31(58.5%)		12(5.1%)	7(58.3%)	5(41.7%)
Kindergarten year				.028				.021
primary class	434(50.3%)	194(44.7%)	240(55.3%)		13(5.5%)	13(100%)	0(0%)
junior class	230(26.7%)	100(43.5%)	130(56.5%)		106(45.1%)	76(71.7%)	30(28.3%)
senior classes	199(23.0%)	67(33.7%)	132(66.3%)		116(49.4%)	74(63.8%)	42(36.2%)
Relationship with child				.775				.413
father	213(24.7%)	88(41.3%)	125(58.7%)		59(25.1%)	45(76.3%)	14(23.7%)
mother	646(74.9%)	272(42.1%)	374(57.9%)		170(72.4%)	114(67.1%)	56(32.9%)
others	4(0.4%)	1(25%)	3(75%)		6(2.5%)	4(66.7%)	2(33.3%)
Parental educational level				<.001				.024
high school or below	333(38.6%)	189(56.8%)	144(43.2%)		127(54.1%)	97(76.4%)	30(23.6%)
college or undergraduate education	503(58.3%)	166(33%)	337(67%)		103(43.8%)	64(62.1%)	39(37.9%)
postgraduate or above	27(3.1%)	6(22.2%)	21(77.8%)		5(2.1%)	2(40.0%)	3(60.0%)
Monthly household income per capita (CNY¥)				<.001				.248
<5000	119(13.8%)	57(47.9%)	62(52.1%)		42(17.9%)	30(71.4%)	12(28.6%)
5000–9999	359(41.6%)	175(48.7%)	184(51.3%)		110(46.8%)	81(73.6%)	29(23.4%)
≥10000	385(44.6%)	129(33.5%)	256(66.5%)		83(35.3%)	52(62.7%)	31(37.3%)
Allergic disease				<.001				.258
yes	133(15.4%)	21(15.8%)	112(84.2%)		19(8.1%)	11(57.9%)	8(42.1%)
no	730(84.6%)	340(46.6%)	390(53.4%)		216(91.9%)	152(70.4%)	64(29.6%)
Congenital diseases				.446				.505
yes	10(1.2%)	3(33.3%)	7(66.7%)		1(0.4%)	1(100%)	0(0%)
no	853(98.8%)	358(42.0%)	495(58.0%)		234(99.6%)	162(69.2%)	72(30.8%)

As shown in Table 1, there were 235 children who had only received one dose of the COVID-19 vaccine. More than half of the parents (63.4%) were between 30 and 40 years old, 49.4% of the children attended senior classes in kindergarten, and most of the parents who completed the questionnaire (72.4%) were mothers. The education level of the parents was concentrated in high school or below (54.1%), and 46.8% of the households had a monthly per capita income between 5,000 and 9,999 yuan. 91.9% of the children had no allergic diseases, and 99.6% of the children had no congenital diseases. A total of 163 parents were willing to vaccinate their preschool children with the second dose of COVID-19 vaccine, while 72 parents were unwilling to vaccinate their children, resulting in an intention rate of 69.4%. A univariate correlation analysis of all sociodemographic variables found that the second dose of COVID-19 vaccine among preschool children was associated with parental age (p = 0.005), children’s kindergarten year (p = 0.021) and parental education level (p = 0.024).

Correlation analysis between variables and vaccination intentions in the HBM and TPB models

Table 2 presents the Cronbach’s alpha coefficients, means and standard deviations, and inter-variable correlations for each dimension in the HBM and TPB models related to Intentions 1 and Intentions 2. The first dose vaccination intentions are significantly correlated with each dimension in both HBM and TPB models, with substantial differences observed in perceived benefits (t = −12.93, p < .001), cues to action (t = −17.95, p < .001), attitude (t = −16.17, p < .001), and subjective norms (t = −20.73, p < .001) between groups. In the case of the second dose vaccination, perceived susceptibility and perceived severity dimensions are not significantly correlated with vaccination intentions, while all other variables show significant correlations with vaccination intentions.

Table 2.

Correlation analysis between variables and vaccination intentions in the HBM and TBP models.

	Cronbach alpha	Mean±SD	Intention	Perceived Susceptibility	Perceived Severity	Perceived Benefits	Perceived Barriers	Cues to Action	Attitudes	Subjective Norms	perceived behavioral control
Intention			1
Perceived Susceptibility	.71	2.46 ± 0.73	−4.71*	1
		2.44 ± 0.74	−1.14	1
Perceived Severity	.42	3.03 ± 0.74	−2.70*	.50*	1
		3.03 ± 0.81	−1.01	.41*	1
Perceived Benefits	.88	3.17 ± 0.65	−12.93*	.25*	.19*	1
		3.31 ± 0.73	−5.80*	.18*	.30*	1
Perceived Barriers	.69	3.25 ± 0.58	5.00*	−.03	.13*	−.13*	1
		3.11 ± 0.6	2.61*	−.06	.16*	−.23*	1
Cues to Action	.91	3.14 ± 0.7	−17.95*	.21*	.16*	.62*	−.07*	1
		3.36 ± 0.68	−6.99*	.14*	.27*	.62*	−.13	1
Attitudes	.89	3.11 ± 0.69	−16.17*	.24*	.12*	.69*	−.19*	0.69*	1
		3.36 ± 0.71	−6.34*	.17*	.17*	.62*	−.30*	0.72*	1
Subjective Norms	.85	3.17 ± 0.67	−20.73*	.18*	.09*	.61*	−.24*	0.71*	0.79*	1
		3.5 ± 0.69	−6.48*	.06	.15*	.56*	−.32*	0.63*	0.77*	1
Perceived Behavioral Control		3.7 ± 0.8	−2.73*	−.07	.05	.10*	.07*	0.03	0.12*	0.20*	1
		3.69 ± 0.71	−3.24*	.03	.16*	.42*	−.11	0.47*	0.52*	0.65*	1

1. *Indicates p < .05.

2. The first row of each dimension represents the variable correlation in intention 1, and the second row of each dimension represents the variable correlation in intention 2.

Logistic regression analysis of factors related to intention to vaccinate preschool children against COVID-19

As shown in Table 3, logistic regression analysis was used to explore the relationship between sociodemographic variables, HBM variables, TPB variables, and parents’ intention to vaccinate preschool children with the first and second doses of the COVID-19 vaccine.

Table 3.

Logistic regression analysis of factors related to intention to vaccinate preschool children against COVID-19.

	Intention 1(N = 863)								Intention 2(N = 235)
Variables*	Model 1				Model 2				Model 1				Model 2
	OR	95%CI		p	OR	95%CI		p	OR	95%CI		p	OR	95%CI		p
Perceived Susceptibility	1.50	1.11	2.03	.009	1.34	0.96	1.87	.081	1.18	0.70	2.00	.536	1.29	0.74	2.24	.364
Perceived Severity	0.88	0.65	1.19	.407	0.99	0.72	1.37	.966	0.85	0.51	1.41	.521	0.84	0.50	1.41	.512
Perceived Benefits	1.80	1.21	2.67	.004	1.00	0.63	1.61	.988	1.97	1.03	3.78	.04	1.90	0.95	3.80	.07
Perceived Barriers	0.41	0.28	0.59	<.001	0.60	0.40	0.91	.016	0.74	0.38	1.45	.38	0.97	0.47	2.03	.944
Cues to Action	10.52	6.57	16.85	<.001	4.80	2.85	8.08	<.001	4.15	1.95	8.83	<.001	3.14	1.28	7.73	.013
Attitudes					1.06	0.59	1.89	.853					0.81	0.30	2.16	.672
Subjective Norms					11.89	6.16	22.92	<.001					3.46	1.36	8.82	.009
Perceived Behavioral Control					0.82	0.61	1.11	.199					0.70	0.34	1.44	.333
R2	0.3398				0.4193				0.2376				0.2667
F	398.71				491.94				68.82				77.24

Results in the model controlled for sociodemographic variables that were significant in the univariate analysis.

For parents’ intention to vaccinate preschool children with the first dose of the vaccine, significant sociodemographic variables in univariate analysis were used as predictor variable. In model 1, introducing the HBM variable explained 33.98% of the variance (F = 398.71, p < .001), and upon adding the TPB variable in model 2, the explanation of variance increased to 41.93% (F = 491.94, p < .001).The results showed that perceived barriers (OR = 0.60, p = .016), cues to action (OR = 4.80, p < .001), and subjective norms (OR = 11.89, p < .001) were significantly associated with parents’ intention to vaccinate preschool children with the first dose of the COVID-19 vaccine.

For parents’ intention to vaccinate preschool children with the second dose of the vaccine, significant sociodemographic variables in univariate analysis were used as predictor variable. In Model 1, introducing the HBM variable explained for 23.76% of the variance (F = 68.82, p < .001), and upon adding the TPB variable in model 2, the explanation of variance increased to 26.67% (F = 77.24, p < .001). The results showed that cues to action (OR = 3.14, p = .013), and subjective norms (OR = 3.46, p = .009) were significantly associated with parents’ intention to vaccinate preschool children with the second dose of the COVID-19 vaccine.

Discussion

This study comprehensively used the Health Belief Model (HBM) and the Theory of Planned Behavior (TPB) to explore the factors influencing parents’ intention to vaccinate their preschool children against COVID-19. The final integrated model showed that for the intention to receive the first dose of the COVID-19 vaccine, perceived barriers, cues to action, and subjective norms were related to parents’ intention. For the intention to receive the second dose, cues to action and subjective norms were related to parents’ intention. In the integrated model for the first vaccine dose, as variables are added, the variance explained by the model increases from 33.98% to 41.93%. Similarly, in the integrated model for the second vaccine dose, the variance explained increases from 23.76% to 26.67%. Both HBM variables and TPB variables are significantly associated with vaccination intention across dimensions in the integrated model. The combination of the two theories improved the explanatory power of parents’ intention to vaccinate their preschool children against COVID-19.

In previous studies, the intention of parents to vaccinate their children against COVID-19 was 89% in the UK,^8,28 0% in New Zealand,²⁹ 65% in the US,²⁹ and 42% in Turkey.³⁰ In this study, the intention of parents to vaccinate their preschool children with the first dose of COVID-19 vaccine was 41.8%, and the intention to vaccinate with the second dose was 69.4%. The intention of parents to vaccinate their preschool children with the second dose of COVID-19 vaccine is higher than that of the first dose, which may be related to the fact that parents have already established initial trust in the vaccine after vaccinating their children with the first dose.²³ And completing the full vaccination course can better reflect the effectiveness of the vaccine.⁸ Paying attention to the factors that influence vaccination intention for different doses can help take targeted intervention measures.

In the univariate analysis of sociodemographic variables and vaccination intention, similar to previous studies, parental age, children’s kindergarten year, and parental education level are all related to vaccination intention.²² It is noteworthy that a higher level of parental education is associated with a lower proportion of individuals willing to vaccinate their children against COVID-19 among the study participants. This differs from previous research results, and the reason for this difference may be that parents who have more knowledge about COVID-19 vaccines are more cautious and skeptical about their credibility, safety, and effectiveness.¹⁹ Children with a history of allergic diseases have a very low intention among parents to vaccinate them with the first dose of COVID-19 vaccine, which is consistent with previous research.²² Allergic reactions in children to vaccines are a major barrier for parents to vaccinate them, and vaccine allergy is one of the contraindications for COVID-19 vaccination. This factor is not significant in the second dose of vaccine, possibly because children who have received the first dose rarely have allergic diseases.

For both parents who haven’t vaccinated their children against COVID-19 and those who have administered one dose of the COVID-19 vaccine to their children, the integrated models reveal that cues to action and subjective norms are significantly associated with parental vaccination intention. Professional support for vaccination on social media,¹⁷ official guidelines issued by health departments,^17,18 and the infection of people around them (family, friends, neighbors, colleagues, etc.) with COVID-19¹⁸ are all reasons that greatly increase parents’ intention to vaccinate their children against COVID-19. Objective factors can better promote changes in intention. Healthcare providers are a key determinant of vaccination behavior,³¹ and research has also shown that when people trust the national health department and medical experts, vaccination acceptance is higher.³² Public health intervention programs should consider taking action for external interventions, such as providing vaccines in more places, focusing more on raising awareness of the benefits of vaccination and awareness of the severity of the disease.^18,25 Furthermore, the support of important others such as family members, friends, and preschool teachers significantly increases vaccine intention. Compared to adults getting vaccinated themselves, parents are often more cautious about vaccinating their preschool children,^25,27 so seeking relevant experience sharing is a major need for them. Individuals should be encouraged to share their positive thoughts and experiences about getting vaccinated against COVID-19 with family and friends, and preschool teachers should provide more information on childhood vaccination to encourage parents to vaccinate their preschool children against COVID-19.

For parents who had never vaccinated their children against COVID-19, the integrated model shows that there is a negative correlation between perceived barriers and parents’ intention to vaccinate their preschool children with the first dose of the COVID-19 vaccine. However, for parents who have already vaccinated their children with one dose of COVID-19 vaccine, perceived barriers are not significant in the model. These perceived barriers include parents’ concerns about the rushed development of the COVID-19 vaccine, the complex process of vaccination, the existence of counterfeit vaccines, and the potential side effects of the vaccine, which would lower parents’ intention to vaccinate their children. This is consistent with the majority of previous research findings.^24,30 The lack of correlation in the second vaccine dose might be related to the trust in the efficacy of the vaccine and the perception of convenience in the vaccination process among parents who have already vaccinated their children with the first dose of the vaccine. Previous findings on intention to vaccine and intention to pay have shown a correlation between vaccine acceptance and vaccine price.^16,33,34 However, in our study, vaccine cost cannot be considered as a perceived barrier to children receiving the COVID-19 vaccination. In mainland China, the policy of providing free COVID-19 vaccinations for preschool children has been implemented. Therefore, parents do not have concerns about the cost of vaccination. Eliminating perceived barriers and increasing parents’ trust in the COVID-19 vaccine requires increasing their knowledge and awareness of the vaccine. Parents’ mistrust of vaccines may stem from a lack of accurate understanding of vaccine safety and negative reporting on social media.²² Providing adequate and appropriate information to the public about COVID-19 vaccines, as well as conclusive evidence on the safety and effectiveness of COVID-19 vaccines, may be a key strategy to increase vaccine uptake and actual vaccination rates.¹⁸

In this study, perceived susceptibility, perceived severity, perceived benefits, behavioral attitude, and perceived behavior control were not significant variables for either the first or the second vaccine dose administration, which is inconsistent with most research results.^17,18 The reason for this insignificance may be related to the COVID-19 epidemic situation at the time of the survey. Our investigation was conducted during the period when Shanghai implemented a dynamic zero-COVID-19 policy, and various measures were taken to fully prevent and control the COVID-19. At this time, parents had not fully realized the susceptibility of preschool children to the COVID-19, had a lack of experience with the severity of the infection in children after being infected with COVID-19, and had not yet had a clear understanding of the benefits of vaccination. They are not clear about their attitude toward COVID-19 vaccination for preschool children, so there is no perceived behavioral control of COVID-19 vaccination for preschool children.³⁵ With the important adjustment of China’s epidemic prevention and control policy, the government has relaxed various control measures, and the current prevalence of omicron strain further highlights the importance of complete vaccination of preschool children before the outbreak.³⁶

There are several limitations to this study. First, participants were recruited from the Minhang district of Shanghai, which is an area with a relatively high level of economic and urbanization status. The similar living environment and economic level of the population may limit the generalizability of the results to other areas, particularly rural areas. Future studies should focus on more rigorous sampling methods to obtain a nationally representative sample and a more balanced sociodemographic population to determine vaccine acceptance and its influencing factors. Secondly, the results of this study are based on self-reported information, which may suffer from information bias. The data of this study were collected through online surveys, which may result in some situations that affect the quality of the data, such as misunderstanding the questionnaire content and not reading the questionnaire when filling it out. Therefore, we set up logic test questions to prevent participants from completing the survey without reading it. Additionally, the questionnaire used in this study was employed for the first time, and its validity has not been previously validated in other studies. However, the results of this study also confirmed partial validity, providing a validity reference for future research, such as experimental studies. Moreover, due to the limitations of the questionnaire, only two items were used to measure perceived severity and the Cronbach’s alpha coefficient of this dimension was low, which may be related to the small number of items in this dimension.^37–39 Furthermore, the perceived behavioral control dimension lacks Cronbach’s alpha coefficient because it has only one item. A low Cronbach’s coefficient means that the item does not measure this dimension well, and reliability has also been described as a basic source of evidence for measuring validity, so a low reliability of this dimension also affects the validity of the questionnaire.^39,40 In addition, the questionnaire lacks items to measure variables such as “whether parents have time to take their children for vaccination” and “negative emotions’ impact” which makes this study lack comprehensiveness and completeness. In future studies, in order to increase the alpha value, we should add more related items to the test that test the same concept and the research design of the questionnaire will be improved and strengthened to obtain more comprehensive results.³⁷ Lastly, this study used a cross-sectional observational design, so causal conclusions cannot be drawn.

Conclusion

This is a cross-sectional study conducted in Shanghai, China, which uses a combination of the HBM and TPB models to investigate parents’ intention to vaccinate preschool children against COVID-19. The integrated model constructed in this study can effectively explain parents’ intention to vaccinate their children with the first and second doses of the COVID-19 vaccine. The findings can help guide the vaccination of COVID-19 and promote better vaccination uptake.

Appendix.

Items for assessing measures of the two theoretical behavior models: HBM and TPB
Model	Measures	Items
HBM	Perceived Susceptibility	My child has a high likelihood of contracting COVID-19.
		There is a high likelihood of severe consequences (e.g., severe illness or death) if my child contracts COVID-19.
		Not vaccinating my child against COVID-19 will increase the risk of them getting infected.
	Perceived Severity	I believe the likelihood of my child fully recovering after contracting COVID-19 is low and that they may experience lingering effects.
		I am concerned that my child will be subject to a series of isolation measures if they contract or are suspected of having COVID-19.
	Perceived Benefit	Getting the COVID-19 vaccine can protect my child from contracting COVID-19.
		Getting the COVID-19 vaccine can help my child avoid severe consequences (e.g., severe illness and death) even if they contract the virus.
		Getting the COVID-19 vaccine can enhance the safety of my child’s daily social activities (e.g., daily life, interactions with family and friends, shopping).
		Getting the COVID-19 vaccine can reduce my concerns about the risk of my child contracting COVID-19.
	Perceived Barriers	I believe the process of registering and getting my child vaccinated against COVID-19 is quite cumbersome.
		I worry that the development of the COVID-19 vaccine was rushed and may not ensure its safety.
		I am concerned about the presence of counterfeit COVID-19 vaccines.
		I worry that the potential side effects of the COVID-19 vaccine may affect my child’s health.
	Cues to Action	If healthcare professionals express support for COVID-19 vaccination on social media, the likelihood of me getting my child vaccinated against COVID-19 will increase.
		If official guidelines are issued by health authorities, the likelihood of me getting my child vaccinated against COVID-19 will increase.
		If my child’s kindergarten organizes COVID-19 vaccination for children, the likelihood of me getting my child vaccinated will increase.
		If the kindergarten requires vaccination for children to attend normally, the likelihood of me getting my child vaccinated against COVID-19 will increase.
		If my child faces discrimination for not receiving the COVID-19 vaccine, the likelihood of me getting my child vaccinated will increase.
		If my city experiences a severe COVID-19 outbreak, the likelihood of me getting my child vaccinated will increase.
		If people around me (family, friends, neighbors, colleagues, etc.) become infected with the coronavirus, the likelihood of me getting my child vaccinated against COVID-19 will increase.
TBP	Attitudes	I believe that the COVID-19 vaccine provides significant protection for children.
		I believe that it is necessary for children to receive the COVID-19 vaccine to prevent infection with the coronavirus.
	Subjective Norms	My family supports me in vaccinating my child against COVID-19.
		My friends support me in vaccinating my child against COVID-19.
		My child’s kindergarten teachers support me in vaccinating my child against COVID-19.
	Perceived Behavioral Control	I have the autonomy to decide whether to vaccinate my child against COVID-19.

Funding Statement

This study was supported by the Minhang District Public Health Key Project (Number: MGWXK2023-11) and the Fudan-Minhang Health Consortium Cooperation Project (Grant No. 2021FM10).

Authors’ contributions

QQW conducted the preliminary development of the questionnaire. QQW, JYC, YHY are responsible for data collection and collation. NJ&YXZ contributed to the acquisition and interpretation of data, QQW has analyzed and expounded the data. QQW&NJ critically reviewed the manuscript for important intellectual content. NJ, YXZ, TW, HC, YYL provided advice on the study design. JWW, JYC, LLC were the project coordinator and contributed to the review and revision of the manuscript. All authors read and approved the final manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The datasets used to support the findings of this study are available from the corresponding author on reasonable request https://www.who.int/publications/i/item/WHO-2019-nCoV-Sci_Brief-Children_and_adolescents-2021.1.

Ethics approval and consent to participate

The study was approved by the Medical Research Ethics Committee of the School of Public Health, Fudan University (The international registry no. IRB00002408 and FWA00002399).