JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

Original Paper

Health Belief Model Perspective on the Control of COVID-19

Vaccine Hesitancy and the Promotion of Vaccination in China:
Web-Based Cross-sectional Study

Hao Chen*, MSc; Xiaomei Li*, MSc; Junling Gao*, PhD; Xiaoxi Liu, BSc; Yimeng Mao, MSc; Ruru Wang, MPH;
Pinpin Zheng, PhD; Qianyi Xiao, PhD; Yingnan Jia, PhD; Hua Fu, PhD; Junming Dai, PhD
Department of Preventive Medicine and Health Education, School of Public Health, Fudan University, Shanghai, China
*
these authors contributed equally

Corresponding Author:
Junming Dai, PhD
Department of Preventive Medicine and Health Education
School of Public Health
Fudan University
No.138# Yixueyuan Road, Xuhui District
Shanghai, 200032
China
Phone: 86 021 54237358
Email: [email protected]

Abstract
Background: The control of vaccine hesitancy and the promotion of vaccination are key protective measures against COVID-19.
Objective: This study assesses the prevalence of vaccine hesitancy and the vaccination rate and examines the association between
factors of the health belief model (HBM) and vaccination.
Methods: A convenience sample of 2531 valid participants from 31 provinces and autonomous regions of mainland China were
enrolled in this online survey study from January 1 to 24, 2021. Multivariable logistic regression was used to identify the
associations of the vaccination rate and HBM factors with the prevalence of vaccine hesitancy after other covariates were
controlled.
Results: The prevalence of vaccine hesitancy was 44.3% (95% CI 42.3%-46.2%), and the vaccination rate was 10.4%
(9.2%-11.6%). The factors that directly promoted vaccination behavior were a lack of vaccine hesitancy (odds ratio [OR] 7.75,
95% CI 5.03-11.93), agreement with recommendations from friends or family for vaccination (OR 3.11, 95% CI 1.75-5.52), and
absence of perceived barriers to COVID-19 vaccination (OR 0.51, 95% CI 0.35-0.75). The factors that were directly associated
with a higher vaccine hesitancy rate were a high level of perceived barriers (OR 1.63, 95% CI 1.36-1.95) and perceived benefits
(OR 0.51, 95% CI 0.32-0.79). A mediating effect of self-efficacy, influenced by perceived barriers (standardized structure
coefficient [SSC]=−0.71, P<.001), perceived benefits (SSC=0.58, P<.001), agreement with recommendations from authorities
(SSC=0.27, P<.001), and agreement with recommendations from friends or family (SSC=0.31, P<.001), was negatively associated
with vaccination (SSC=−0.45, P<.001) via vaccine hesitancy (SSC=−0.32, P<.001).
Conclusions: It may be possible to increase the vaccination rate by reducing vaccine hesitancy and perceived barriers to
vaccination and by encouraging volunteers to advocate for vaccination to their friends and family members. It is also important
to reduce vaccine hesitancy by enhancing self-efficacy for vaccination, due to its crucial mediating function.

(J Med Internet Res 2021;23(9):e29329) doi: 10.2196/29329

KEYWORDS
COVID-19 pandemic; vaccination behavior; vaccine hesitancy; health belief model

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 1

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

the joint prevention and control of COVID-19 [10]. Not only

Introduction have the daily numbers of new cases, close contacts, patients
COVID-19 has spread worldwide, causing more than 88 million who recovered, and patients who died been announced, but the
infections and more than 1.9 million deaths as of January 2021 latest number of vaccinated people as well as side effects and
[1]. Due to the lack of effective treatments, the development psychological changes following vaccination have also been
and use of a new COVID-19 vaccine has become an important announced. In addition to communications at the national level,
strategy to control the epidemic. Since COVID-19 broke out, provinces, municipalities, and autonomous regions released the
according to the World Health Organization (WHO), 60 new latest information about the epidemic through various channels,
coronavirus-inactivated vaccines and more than 10 nucleic acid such as press conferences or short videos, specific to their own
vaccines, vector vaccines, and protein subunit vaccines have situations, to ensure mastery and understanding of information
been developed [2]. Vaccination is recognized as the most regarding the epidemic and vaccines [11,12]. As of May 2021,
successful and cost-effective public health intervention in the 6 months after countries had begun carrying out vaccinations,
world today, and it has made a very large contribution to the global number of COVID-19 vaccinations has exceeded 1.5
improving global health by reducing the incidence and deaths billion doses. Among them, nearly 60% are concentrated in
of many infectious diseases [3,4]. China and the whole world China (420 million doses), the United States (270 million doses),
are experiencing the third wave of epidemics, so it is especially and India (180 million doses). Except for a few countries with
important to establish herd immunity by vaccinating against a vaccination rate exceeding 50% (eg, Israel), most countries
COVID-19 [5]. in the world have a vaccination rate below 20% [13]. According
to a recent study, predicted vaccine coverage of 55% to 82% of
On December 30, 2020, the first homegrown COVID-19 vaccine the population is needed to achieve COVID-19 herd immunity
in China was approved for marketing by the China National [5]. In addition to the supply of vaccines, individuals’
Medical Products Administration, and open volunteer psychological mechanisms of vaccine behavior are particularly
vaccination to the public was announced through official media. critical to vaccination [14]. Therefore, it is of great significance
On January 9, 2021, the National Health Commission promised to explore the possible influencing factors of individuals’
free vaccinations for the Chinese population [6]. As of February vaccination willingness when vaccination rates are low in order
2021, the COVID-19 vaccine in China is suitable for people to improve COVID-19 vaccination willingness and coverage
aged 18 to 59 years; the COVID-19 vaccine is not suitable for in China and other parts of the world.
pregnant women, lactating women, and people with the
following conditions: acute stages of fever, infections and other Although vaccines are currently an effective means of improving
diseases, immune deficiency or immune disorders, serious liver global health, in many parts of the world there are still quite a
and kidney diseases, hypertension, diabetic complications, and few people who question the necessity of vaccination, postpone
malignant tumors with uncontrolled drugs [7]. As of February vaccination, or even refuse vaccination; this is especially true
2021, common adverse reactions to vaccines in China mainly when vaccines first came to market and were met with
include headache, fever, local redness or lumps at the inoculation considerable hesitation and even outright opposition [15]. In
site, and cough, as well as loss of appetite, vomiting, and 2012, the WHO established the Strategic Advisory Group of
diarrhea in some people [8]. In first month of COVID-19 Experts (SAGE) working group to address and define vaccine
vaccinations, up to January 26, 2021, 22.8 million doses of the hesitancy and its scope [16]. Vaccination hesitancy was defined
COVID-19 vaccine were administered in China, and less than as the refusal or delay of vaccination when vaccination services
5% of the vaccine-eligible population among them, the main were available [17], and vaccination hesitancy was listed among
group, was at high risk of infection in all regions [9]. the 10 threats to global health in 2019 [18]. Vaccine hesitancy
Considering the occupational exposure risk of COVID-19 is reflected in many factors, including confidence in the efficacy
infection, some populations with priority for vaccination were and safety of the vaccine and in the health service system
those with occupations at border ports, in key places such as providing the vaccine, such as the reliability and competence
international and domestic transportation, and in key industries of the health service system and the professionals involved in
such as medical and health care as well as basic social operation the vaccination service [19]. In the first month after vaccines
services. These populations are mass vaccinated on the basis of became available to all vaccine-eligible members of the Chinese
individual willingness [6]. According to the director of the population, a nationwide cross-sectional study reported the
National Health Commission, National Bureau of Disease prevalence of COVID-19 vaccination hesitancy to be 35.5%.
Control and Prevention, all residents could be vaccinated in an After an instance of illegal marketing of vaccines, 32.4% of
orderly manner where there is an ample supply of vaccine and parents became hesitant of vaccines [20]; rapid sociocultural
where vaccination units are health service centers, township changes have also contributed to vaccine hesitancy [21,22]. A
health centers, or general hospitals located in their respective study on COVID-19 vaccine hesitancy of Italian college students
jurisdictions. Local governments have been required to make showed that among the 735 students who answered questions
public in a timely manner the vaccination sites and units that about their vaccination intentions, more than 1 in 10 students
can administer vaccines in their respective jurisdictions, showed hesitancy [23]. An investigation during Israel’s
including their locations and service hours [8]. From the mandatory quarantine revealed that nurses and medical workers
beginning of 2020 to February 2021, the State Council showed high levels of vaccine hesitancy [24]. According to a
Information Office has held regular press conferences to invite literature review, 68.4% of the global population is willing to
experts from relevant departments to brief the population on receive the vaccination [25].

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 2

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

Recent studies of factors associated with COVID-19 vaccination and other countries where vaccinations are available to the
have identified a number of demographic, cognitive, and domestic population [45].
psychosocial factors, including age, gender, educational level,
In summary, we explored whether HBM constructs were
insurance status, attitudes toward the vaccine, confidence in
associated with vaccine hesitancy and vaccination at the time
government information, perceived susceptibility to COVID-19,
when COVID-19 vaccination became available to the public in
and perceived benefits and side effects of the vaccine [26,27].
mainland China. A previous study identified that vaccine
In the current age of Web 2.0, the spread of false news about
intention and willingness were important predictors of
vaccine safety and validity on social media, such as that
vaccination behavior, with more than 50% of the explained
COVID-19 vaccination can affect individuals’ reproductive
variance in influenza [46] and HPV [34] vaccinations. However,
function, influence vaccination willingness and confidence [28].
a gap seems to exist between intention and vaccination behavior
Several typical behavioral theories, such as the health belief
[47], such as the willingness of students to receive the HPV
model (HBM), the theory of planned behavior (TPB) [29], and
vaccine predicting less than 10% of actual vaccinations [34].
the diffusion of innovation theory (DIT) [14], have been used
Our first hypothesis (Hypothesis 1) was that vaccine hesitancy
to explain COVID-19 vaccination intent combined with
was negatively associated with COVID-19 vaccination behavior.
demographic, cognitive, and psychosocial factors. The HBM
In particular, we examined our major hypothesis (Hypothesis
is a widely used theory that proposes a variety of psychological
2), which was that the HBM constructs of perceived barriers,
factors that affect people's health protective behaviors, such as
self-efficacy, and cues to action would predict vaccine hesitancy
attitudes, beliefs, and intentions [30-32]. The HBM assumes
and vaccination behavior. As in a previous study, self-efficacy
that health-related actions depend on the simultaneous
is defined as the confidence in one’s ability to facilitate decisions
occurrence of three factors [33]: (1) the presence of sufficient
to carry out a health behavior such as vaccination, which is
motivation (or health concern) to make the health problem
useful only to the extent that one feels one can adequately
salient or relevant, (2) the belief that a person is vulnerable to
implement the steps needed to perform the behavior [48].
serious health problems or the sequelae of that illness or
Evidence based on the HBM poses several mechanisms
condition is often referred to as a perceived threat, and (3)
regarding how self-efficacy is associated with vaccine intention
believing that following a specific health recommendation will
and behaviors. Self-efficacy was able to mediate the relationship
help reduce the perceived threat at a subjectively acceptable
between perceived barriers to HPV vaccination and HPV
cost. The TPB assumes that an individual's behavioral posture,
vaccine intentions among young women [49]. A similar
activity attraction, and behavioral control jointly affect and
mediation effect was found in the association between perceived
direct the individual's behavior [34]. The DIT aims to
severity and susceptibility and the intent to receive the Zika
disseminate innovation awareness, technology, or innovative
vaccine [50]. It was also suggested that self-efficacy could
ideas related to the masses, so that patients can develop
influence the path from cues to action (eg, physician
innovative thinking or health awareness. In recent years, the
recommendation, family members recommendation, media
DIT has been gradually introduced into medical and health
coverage, and public health communication) to HPV vaccine
industries, mainly for the guidance of health education strategies
uptake [51] and acceptance of the H1N1 vaccine [52]. The
[14]. The HBM has been one of the most widely used theories
aforementioned studies suggested our third major hypothesis
in understanding health and illness behaviors, and due to its
(Hypothesis 3), which was that self-efficacy of the COVID-19
design, it has been previously used in vaccination studies to
vaccine would mediate the influence of other HBM constructs
identify behavior relationships [35,36]. When compared with
on vaccine hesitancy and vaccination.
other models that explain behavior and resulting actions, the
HBM was specifically developed to focus on preventative health
research [35-38], which has been modified since its early use
Methods
in the 1950s to be more inclusive and encourage interventions Study Design and Participants
that improve health behaviors [39]. Thus, the HBM was chosen
as the preferred model to investigate intention and behavior From January 1 to 24, 2021, we used convenience and snowball
regarding COVID-19 vaccination. There are six main sampling to recruit a sample of 2580 participants from 31 out
components of the HBM: perceived susceptibility, perceived of a total of 34 provinces and autonomous regions in China,
severity, perceived benefits, perceived barriers, self-efficacy with each area consisting of at least 30 participants; we then
for health protective behaviors, and cues to action [40]. Previous conducted a web-based cross-sectional study. A digital
studies, including those on H1N1 [41], hepatitis [42], human questionnaire link was sent to a WeChat “Friends circle,” a
papillomavirus (HPV) [43], and measles [44], have identified function that can be used to share personal photos or public
HBM factors as important predictors of vaccination intentions. website links in one’s “Moments” to make them visible to
Therefore, it is necessary to explore the possible influence of friends on platforms such as Twitter and Facebook. This
these factors on people's willingness to vaccinate against questionnaire link, on the Wenjuanxing platform, could then
COVID-19 in order to improve individual immunity and slow be forwarded or shared by participants with friends in their
the epidemic. Although the aforementioned studies suggested WeChat contact list whom they considered appropriate for this
that there were associations between HBM constructs and survey; their friends were also encouraged to send the link to
vaccine acceptance or hesitancy, relatively few studies have their friend networks. The snowball sampling process continued
focused on COVID-19 vaccination behavior, especially in China until a sufficient sample size was reached. The first page of the
questionnaire contained an electronic consent form. Each

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 3

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

respondent received a small monetary reward of ¥5 (a currency Measurements

exchange rate of ¥1=US $0.15 is applicable) after authentically
completing the questionnaire, which took approximately 5 to Vaccine Hesitancy and Vaccination
10 minutes. To prevent repeated entries from the same Vaccine hesitancy was assessed with a one-item self-report
individual, who may attempt multiple entries for the enrollment measure that quantified the demand for, and acceptance of,
reward, additional measures were adopted: (1) the same IP vaccination: “How willing would you be to get the COVID-19
address was only allowed to be used once to fill in the vaccine?” The respondents were asked to answer the question
questionnaire, which was a built-in function of the Wenjuanxing using the following 7-point scale recommended by the SAGE
platform, and (2) participants were only allowed to fill in working group on vaccine hesitancy: “accept all [vaccines],”
questionnaires after logging in to their WeChat accounts—they “accept but unsure,” “accept some,” “delay,” “refuse some,”
needed to register with this platform with a personal identity “refuse but unsure,” and “refuse all” [17]. Vaccine hesitancy
card—and each WeChat account could only be used once to fill was defined as any response on the scale except for “accept all”
in the questionnaire. The minimum sample size was calculated or “accept but unsure.” Vaccination was assessed by asking the
to be 1100 by using the following formula: participants to answer “yes” or “no” to a single question: “Have
you gotten the COVID-19 vaccine?”
Health Belief Model
where the latest reported prevalence of COVID-19 vaccination Items derived from the HBM were adopted from a previous
hesitancy (p) was 35.5%, based on research that was conducted study or modified to measure the participants’ beliefs about
in China, nationwide, from January 10 to January 22, 2021 [53]. COVID-19 vaccination. Five essential dimensions of health
The type I error () was .05; thus, z1-/2=1.96, the precision (d) beliefs were measured as follows: (1) perceived susceptibility
was 0.04, and the design effect (deff) was 2 [54]. The inclusion to COVID-19 in the future (three items; eg, “I was vulnerable
criteria for participants’ enrollment were as follows: (1) aged to infection with SARS-CoV-2”), (2) perceived severity of
18 to 59 years, (2) able to understand the questionnaire by COVID-19 infection (four items; eg, “It would be very harmful
themselves, and (3) could use online services, such as mobile for me if I got COVID-19”), (3) perceived benefits of
phones, computers, and tablet computers. The questionnaires COVID-19 vaccination (three items; eg, “COVID-19 vaccination
of participants who met the following exclusion criteria were can protect me from infection with SARS-CoV-2”), (4)
discarded: (1) aged less than 18 years (n=16) or more than 59 perceived barriers to COVID-19 vaccination (six items; eg,
years (n=32) and not eligible for vaccination until April 2021 “The COVID-19 vaccine might have side effects, such as fever
in China and (2) returned invalid questionnaires (n=32). or soreness in the arm”), and (5) self-efficacy for COVID-19
Questionnaires were deemed invalid if the following occurred: vaccination (five items; eg, “I believe I can deal with side effects
(1) participant gave one or two wrong answers to two quality of the COVID-19 vaccine with doctors’ help”). Cues to action
control questions, including “Where is capital of China?” and refer to external recommendations that might affect individuals’
“What’s three plus five?”; (2) occurrence of a logic check result health-related behaviors. In this study, the Cronbach α
error, which occurred when the participant selected both “no coefficients indicating internal consistency (ie, reliability) were
disease” and “any type of disease” in response to the question .78 for the total HBM factors, .84 for perceived susceptibility
“Do you have any type of the following diseases or diagnosed to COVID-19, .80 for perceived severity of COVID-19 infection,
medical histories”; and (3) participant took less than the .83 for perceived benefits of COVID-19 vaccination, .80 for
minimum time of 3 minutes to complete the questionnaire. perceived barriers to COVID-19 vaccination, and .82 for
Cognitive interviewing with 5 subjects was done to refine the self-efficacy for COVID-19 vaccination. The sampling adequacy
questionnaires through the web-based platform WeChat. for the HBM factor scale was excellent
Participants were required to respond to each item by answering (Kaiser-Meyer-Olkin=0.82). Inter-item correlations were
three questions: (1) “What does ‘……’ mean to you?”, (2) “Can sufficiently large for principal component analysis (PCA)
you repeat this question in your own words?”, and (3) “When (Bartlett test of sphericity: 2210=23,122.6, P<.001). The PCA
you think about ‘……’ what comes to your mind?” We also revealed five factors, which in combination explained 68.58%
asked participants to answer three questions for the overall of the variance, and each factor accounted for 24.23%, 20.55%,
survey, including the following: (1) “Are there additional 10.32%, 8.16%, and 5.32% of the explained variance,
questions you believe should be asked?”, (2) “Are there respectively. An examination of the factor loadings after rotation
questions you believe should be deleted?”, and (3) “Are there suggested, as expected, that factor 1 (perceived barriers to
questions you believe should be modified?” The entire COVID-19 vaccination) had six items with loading factors
questionnaire was tested and modified to appropriately conduct between 0.74 and 0.79, factor 2 (self-efficacy for COVID-19
the survey. Finally, 2531 participants were included in this vaccination) included five items with loading factors between
study. All participants consented to written ethics approval 0.71 and 0.80, factor 3 (perceived severity of COVID-19
before the survey was conducted. This study was approved by infection) included four items with loading factors between 0.67
the Institutional Review Board of Fudan University, School of and 0.85, factor 4 (perceived benefits of COVID-19 vaccination)
Public Health (IRB00002408&FWA00002399), and approval included three items with loading factors between 0.68 and 0.85,
expired on March 3, 2021. and factor 5 (perceived susceptibility to COVID-19) included
three items with loading factors between 0.78 and 0.89.

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 4

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

External Cues to Action Statistical Analysis

External cues to action were assessed based on four cues used Frequencies were first calculated for all variables, and the
in previous surveys [36,55]: recommendations from authorities, prevalence and 95% CIs of vaccine hesitancy and vaccination
recommendations from friends or family, vaccination of were determined according to the participants’ demographics,
authorities, and vaccination of friends or family. Participants health-related characteristics, and HBM factors. Multivariable
were asked to state their level of agreement with each of the logistic regression analyses were used to explore the
statements, with a score of 1 for positive responses (strongly demographic and health-related characteristics (Table 1) as well
agree or agree) and a score of 0 for neutral or negative responses as the HBM factors (Table 2) associated with vaccine hesitancy
(neither agree nor disagree, disagree, or strongly disagree). The and vaccination. We then ran the multivariable logistic
Cronbach α coefficient for cues to action was .82. The sampling regression again to determine the HBM factors associated with
adequacy for the cues to action scale was excellent vaccine hesitancy and vaccination after controlling for covariates
(Kaiser-Meyer-Olkin=0.75). Inter-item correlations were (ie, demographic and health-related characteristics), with a
sufficiently large for PCA (Bartlett test of sphericity: 26=2829.1, significance level of P<.05. Odds ratios (ORs) with 95% CIs
P<.001). The PCA revealed a single factor, which in were calculated for each independent variable and were
combination explained 59.72% of the variance, and an visualized in forest plots (Figures 1 and 2). All of the analyses
examination of the factor loadings after rotation suggested, as were performed using SAS software, version 9.4 (SAS Institute
expected, that the single factor included four items whose Inc), and all tests were two-tailed with a significance level of
loading factors were between 0.65 and 0.84. P<.05. We used the forest plot package in R software, version
3.5.3 (The R Foundation), to generate the forest plots. We used
Demographic and Health-Related Characteristics Mplus, version 8.4 (Muthén & Muthén), to establish structural
Demographic characteristics in this study included gender, age, equation modeling (SEM) and to assess the standardized
educational level (high school degree and below, bachelor’s structure coefficients (SSCs) among the HBM factors of vaccine
degree, or master’s degree and above), marital status (married hesitancy and vaccination. The mean- and variance-adjusted
or not married [including unmarried, divorced, and widowed]), weighted least squares method was employed as the method of
occupation (medical worker or nonmedical worker), region estimation because the analyses included categorical endogenous
(urban or rural), monthly salary (<¥6000, ¥6000-¥10,000, or variables (ie, vaccine hesitancy and vaccination), and the link
>¥10,000), and family members with backgrounds in medical was the probability unit in the current model [57]. We freed
work or with medical education (yes or no). Health-related covariances between error terms based on their modification
characteristics included self-rated health and self-reported indices during the estimation process to improve model fit. The
chronic diseases having been diagnosed by doctors. Self-rated most common indices and acceptable reference values included
health was evaluated by a single question: “How is your the magnitude of χ2 divided by its degrees of freedom (χ2/df
perceived health in general?”; responses included “excellent,” <5), the comparative fit index (CFI >0.90), the Tucker-Lewis
“very good,” “good,” “general,” or “poor” [56]. We listed 16 index (TLI >0.90), and the root mean square error of
common chronic diseases, such as hypertension and diabetes, approximation (RMSEA <0.08), which were used to determine
and categorized the number of reported chronic diseases into whether the data fit the model [58].
0, 1 or 2, and 3 or over.

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 5

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

Figure 1. Associations between the health belief model and vaccine hesitancy.

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 6

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

Figure 2. Associations between the health belief model and vaccination rate.

(1609/2531, 63.6%), were nonmedical personnel (2034/2531,

Results 80.4%), lived in urban areas (2262/2531, 89.4%), reported good
Participant Characteristics health (2020/2531, 79.8%), and did not have chronic diseases
(1617/2531, 63.9%). Slightly less than half of the participants
Our analysis included 2531 participants aged between 18 and reported monthly salaries lower than ¥6000 (1128/2531, 44.6%)
59 years (mean 33.92 years, SD 8.94); 58.7% (1486/2531) of and had family members with medical personnel backgrounds
the participants were female. Most of the participants were (1056/2531, 41.7%) (Table 1).
married (1660/2531, 65.6%), had a bachelor’s degree

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 7

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

Table 1. Distribution of vaccine hesitancy and vaccination rate by participant demographics and health-related characteristics.
Characteristics Participants Vaccine hesitancy Vaccination rate
(N=2531), n (%)
Vaccine hesitan- Vaccination, n
cy, n (%) ORa (95% CI) P value (%) OR (95% CI) P value
Age (years)
18-29 926 (36.6) 412 (44.5) 1 75 (8.1) 1
30-39 993 (39.2) 467 (47.0) 1.11 (0.93-1.33) .27 110 (11.1) 1.41 (1.04-1.92) .03
40-49 410 (16.2) 163 (39.8) 0.82 (0.65-1.04) .11 55 (13.4) 1.76 (1.22-2.54) .003
50-59 202 (8.0) 78 (39.6) 0.79 (0.58-1.07) .13 24 (11.9) 1.53 (0.94-2.49) .09
Gender
Male 1045 (41.3) 422 (40.4) 1 116 (11.1) 1
Female 1486 (58.7) 698 (47.0) 1.31 (1.11-1.53) .001 148 (10.0) 0.89 (0.69-1.15) .36
Marital status
Married 1660 (65.6) 725 (43.7) 1 187 (11.3) 1
Not married 871 (34.4) 395 (45.4) 1.07 (0.91-1.26) .42 77 (8.8) 0.76 (0.58-1.01) .06
Educational level
High school degree and be- 204 (8.0) 83 (40.7) 1 9 (4.4) 1
low
Bachelor’s degree 1609 (63.6) 725 (45.1) 1.20 (0.89-1.61) .24 150 (9.3) 2.23 (1.12-4.43) .02
Master’s degree and above 718 (28.4) 312 (43.5) 1.12 (0.82-1.54) .48 105 (14.6) 3.71 (1.84-7.47) <.001
Occupation
Nonmedical personnel 2034 (80.4) 929 (45.7) 1 102 (5.0) 1
Medical personnel 497 (19.6) 191 (38.4) 0.74 (0.61-0.91) .004 162 (32.6) 9.16 (6.97-12.04) <.001
Region
Urban 2262 (89.4) 1016 (44.9) 1 242 (10.7) 1
Rural 269 (10.6) 104 (38.7) 0.77 (0.60-1.00) .05 22 (8.2) 0.74 (0.47-1.17) .20

Monthly salary (¥)b

<6000 1128 (44.6) 478 (42.5) 1 68 (6.0) 1
6000-10,000 787 (31.1) 338 (43.0) 1.02 (0.85-1.23) .83 100 (12.7) 2.27 (1.64-3.13) <.001
>10,000 616 (24.3) 303 (49.2) 1.31 (1.08-1.60) .007 96 (15.6) 2.88 (2.07-3.99) <.001
Family members with medical backgrounds
No 1475 (58.3) 667 (45.2) 1 88 (6.0) 1
Yes 1056 (41.7) 453 (42.9) 0.91 (0.78-1.06) .25 176 (16.7) 3.13 (2.28-4.17) <.001
Self-reported health
Good 2020 (79.8) 849 (42.0) 1 230 (11.4) 1
Poor 511 (20.2) 271 (53.0) 1.56 (1.28-1.89) <.001 34 (6.7) 0.55 (0.38-0.81) .002
Number of chronic diseases
0 1617 (63.9) 688 (42.3) 1 178 (11.0) 1
1 639 (25.4) 300 (47.0) 0.84 (0.70-1.01) .045 64 (10.0) 1.11 (0.82-1.50) .49
2 and above 288 (11.2) 136 (49.5) 1.13 (0.86-1.49) .49 22 (8.0) 0.78 (0.47-1.30) .34
Vaccine hesitancy
Yes 1120 (44.3) N/Ac N/A N/A 32 (2.9) 1

No 1411 (55.7) N/A N/A N/A 232 (16.4) 6.69 (4.58-9.77) <.001

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 8

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

a
OR: odds ratio.
b
A currency exchange rate of ¥1=US $0.15 is applicable.
c
N/A: not applicable.

1.12-4.43), had good self-rated health (OR 1.82, 95% CI

Distribution of Vaccine Hesitancy and Vaccination by 1.23-2.63), were not vaccine hesitant (OR 6.69, 95% CI
Participant Characteristics and Health Belief Model 4.58-9.77), had a monthly salary between ¥6000 and ¥10,000
Factors (OR 2.27, 95% CI 1.64-3.13), had a monthly salary over ¥10,000
Overall, 44.3% (1120/2531; 95% CI 42.3%-46.2%) of the (OR 2.88, 95% CI 2.07-4.17), or had family members with
participants were classified as vaccine hesitant: 1.4% responded medical personnel backgrounds (OR 3.13, 95% CI 2.28-4.17).
“refuse all,” 5.3% responded “refuse but unsure,” 3.7%
According to the multivariable regression analyses including
responded “refuse some,” 18.8% responded “delay,” and 15.1%
the HBM factors (Table 2), the participants were more likely
responded “accept some.” Overall, 55.7% (1411/2531) of the
to be vaccine hesitant if they had high perceived susceptibility
participants were classified as vaccine accepting: 25.1%
to COVID-19 (OR 1.34, 95% CI 1.07-1.69) or had high
responded “accept but unsure” and 30.6% responded “accept
perceived barriers to vaccination (OR 1.84, 95% CI 1.56-2.17).
all.” Only 10.4% (264/2531; 95% CI 9.2%-11.6%) of the
The participants were less likely to be vaccine hesitant if they
participants had been vaccinated for COVID-19, while the
had high perceived benefits of vaccination (OR 0.34, 95% CI
majority (2267/2531, 89.6%) had not been.
0.23-0.50), had high self-efficacy for vaccination (OR 0.26,
According to the multivariable logistic regression analyses 95% CI 0.20-0.34), agreed with recommendations from
including participant characteristics (Table 1), the participants authorities (OR 0.47, 95% CI 0.38-0.58), agreed with
were more likely to be vaccine hesitant if they were female (OR recommendations from friends or family (OR 0.19, 95% CI
1.31, 95% CI 1.11-1.53), were nonmedical personnel (OR 1.35, 0.14-0.24), agreed with the vaccination of authorities (OR 0.46,
95% CI 1.10-1.64), had poor self-rated health (OR 1.56, 95% 95% CI 0.36-0.60), or agreed with the vaccination of friends or
CI 1.28-1.89), or had a monthly salary over ¥10,000 (OR 1.31, family (OR 0.77, 95% CI 0.66-0.91). The participants were
95% CI 1.08-1.60). The participants were more likely to have more likely to have been vaccinated if they had high
been vaccinated if they were 30 to 39 years old (OR 1.41, 95% self-efficacy for vaccination (OR 3.39, 95% CI 1.92-6.00),
CI 1.04-1.92), had a bachelor’s degree (OR 2.23, 95% CI agreed with recommendations from authorities (OR 2.89, 95%
1.12-4.43), had a master’s degree and above (OR 3.71, 95% CI CI 1.75-4.78), agreed with the vaccination of authorities (OR
1.12-4.43), were medical personnel (OR 3.71, 95% CI 2.94, 95% CI 1.62-5.31), or agreed with the vaccination of
friends or family (OR 5.05, 95% CI 3.77-6.76).

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 9

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

Table 2. Distribution of vaccine hesitancy and vaccination by health belief model (HBM) factors and cues to action.
HBM factors and cues to action Participants Vaccine hesitancy Vaccination
(N=2531), n (%)
Vaccine hesitan- Vaccination, n
cy, n (%) ORa (95% CI) P value (%) OR (95% CI) P value
Perceived susceptibility
Low 2191 (86.6) 948 (43.3) 1 231 (10.4) 1
High 340 (13.4) 172 (50.6) 1.34 (1.07-1.69) .01 33 (9.7) 0.91 (0.62-1.34) .64
Perceived severity
Low 292 (11.5) 126 (43.2) 1 42 (14.4) 1
High 2239 (88.5) 994 (44.4) 1.05 (0.82-1.36) .69 222 (9.9) 0.66 (0.46-0.93) .02
Perceived benefits
Low 125 (4.9) 86 (68.8) 1 12 (9.6) 1
High 2406 (95.1) 1034 (43.0) 0.34 (0.23-0.50) <.001 252 (10.5) 1.01 (0.95-1.07) .76
Perceived barriers
Low 1622 (64.1) 630 (38.8) 1 219 (13.5) 1
High 909 (35.9) 490 (53.9) 2.08 (1.77-2.45) <.001 49 (5.0) 0.33 (0.24-0.47) <.001
Self-efficacy
Low 352 (13.9) 252 (71.6) 1 13 (3.7) 1
High 2179 (86.1) 868 (39.8) 0.26 (0.20-0.34) <.001 251 (11.5) 3.39 (1.92-6.00) <.001
Recommendations from authorities
Disagree 393 (15.5) 236 (60.1) 1 17 (4.3) 1
Agree 2138 (84.5) 884 (41.4) 0.47 (0.38-0.58) <.001 247 (11.6) 2.89 (1.75-4.78) <.001
Recommendations from friends or family
Disagree 367 (14.5) 283 (77.1) 1 28 (7.6) 1
Agree 2164 (85.5) 837 (38.7) 0.19 (0.14-0.24) <.001 236 (10.9) 1.48 (0.99-2.23) .06
Vaccination of authorities
Disagree 290 (11.5) 177 (61.0) 1 12 (4.1) 1
Agree 2241 (88.5) 943 (42.1) 0.46 (0.36-0.60) <.001 252 (11.2) 2.94 (1.62-5.31) <.001
Vaccination of friends or family
Disagree 1488 (58.8) 696 (46.8) 1 66 (4.4) 1
Agree 1043 (41.2) 424 (40.7) 0.77 (0.66-0.91) .002 198 (19.0) 5.05 (3.77-6.76) <.001

a
OR: odds ratio.

were occupation as medical personnel (P=38.4%, 95% CI

Influencing Factors of Vaccine Hesitancy and 34.2%-42.7%; OR 0.74, 95% CI 0.59-0.93), high perceived
Vaccination benefits of vaccination (P=43.0%, 95% CI 41.0%-45.0%; OR
We included the participant characteristics and HBM factors in 0.51, 95% CI 0.32-0.79), high self-efficacy for vaccination
the vaccine hesitancy logistic regression, and the influencing (P=38.4%, 95% CI 48.7%-57.4%; OR 1.46, 95% CI 1.18-1.80),
factors are shown in Figure 1. The risk factors for vaccine agreement with recommendations from authorities (P=41.4%,
hesitancy were female gender (P=47.0%, 95% CI 44.4%-49.5%; 95% CI 39.3%-43.4%; OR 0.74, 95% CI 0.57-0.98), and
OR 1.12, 95% CI 1.01-1.44), monthly salary over ¥10,000 agreement with recommendations from friends or family
(P=49.2%, 95% CI 45.2%-53.1%; OR 1.45, 95% CI 1.16-1.80), (P=41.4%, 95% CI 39.3%-43.4%; OR 0.74, 95% CI 0.57-0.98).
poor self-rated health (P=53.0%, 95% CI 48.7%-57.4%; OR
We included the participant characteristics, the HBM factors,
1.46, 95% CI 1.18-1.80), high perceived susceptibility to
and vaccine hesitancy in the vaccination logistic regression,
COVID-19 (P=50.6%, 95% CI 45.3%-55.9%; OR 1.30, 95%
and the influencing factors are shown in Figure 2. The promoting
CI 1.01-1.67), and high perceived barriers to vaccination
factors for vaccination were occupation as medical personnel
(P=53.9%, 95% CI 50.7%-57.2%; OR 1.63, 95% CI 1.36-1.95).
(P=32.6%, 95% CI 28.5%-36.7%; OR 6.52, 95% CI 4.51-9.41),
Additionally, the protective factors against vaccine hesitancy

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 10

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

monthly salary between ¥6000 and ¥10,000 (P=12.7%, 95% goodness-of-fit statistics, SEM showed a better fit to the data
CI 10.4%-15.0%; OR 2.05, 95% CI 1.38-3.04), monthly salary than the regression models (χ2/df=4.62; RMSEA=0.05; CFI =
over ¥10,000 (P=15.6%, 95% CI 12.7%-18.5%; OR 2.15, 95% 0.95; TLI = 0.91), and all of the paths were statistically
CI 1.40-3.30), family members with medical personnel significant (P<.05). The findings suggested that a mediating
backgrounds (P=16.7%, 95% CI 14.4%-18.9%; OR 1.51, 95% effect of self-efficacy, influenced by perceived barriers
CI 1.07-2.13), a lack of vaccine hesitancy (P=16.4%, 95% CI (SSC=−0.71, P<.001), perceived benefits (SSC=0.58, P<.001),
14.5%-18.4%; OR 7.75, 95% CI 1.01-1.67), agreement with agreement with recommendations from authorities (SSC=0.27,
recommendations from friends or family (P=10.9%, 95% CI P<.001), and agreement with recommendations from friends or
9.6%-12.2%; OR 3.11, 95% CI 1.75-5.52), and agreement with family (SSC=0.31, P<.001), was negatively associated with
the vaccination of friends or family (P=19.0%, 95% CI vaccination (SSC=−0.45, P<.001) via vaccine hesitancy
19.6%-21.4%; OR 4.88, 95% CI 3.41-6.99). Additionally, a (SSC=−0.32, P<.001). Additionally, perceived barriers
lower vaccination rate was associated with higher perceived (SSC=0.53, P<.001) and perceived benefits (SSC=−0.21,
barriers to COVID-19 vaccination (p=5.0%, 95% CI 3.5%-6.4%; P<.001) were directly associated with vaccine hesitancy.
OR 0.51, 95% CI 0.35-0.75). Perceived barriers (SSC=−0.20, P<.001) and recommendations
Structural Equation Modeling of Vaccination from friends or family (SSC=0.14, P<.001) were directly
correlated with vaccination behavior.
We used SEM to examine the underlying psychological
mechanism of vaccination behavior (Figure 3). Based on the
Figure 3. The paths among vaccine hesitancy, vaccination, and health belief model factors. The numbers on the lines are the standardized structure
coefficients.

vaccine hesitancy, while perceived barriers and

Availability of Data and Materials recommendations from friends or family were directly correlated
The data that support the findings of this study are available with vaccination.
from the School of Public Health, Fudan University. The data
were used under license for this study and are not publicly In this study based in China, the prevalence of vaccine hesitancy
available. The data are, however, available from the authors was 44.3% (95% CI 42.3%-46.2%), and the vaccination rate
upon reasonable request and with permission from the School was 10.4% (95% CI 9.2%-11.6%), representing high vaccine
of Public Health, Fudan University. hesitancy and low vaccination behaviors. Vaccine hesitancy
has been universally reported in recent research, with over half
Discussion of participants (53%) across 19 countries showing vaccine
hesitancy, which is similar to our results [59] and in accordance
Principal Findings with the decline in vaccine acceptance (from >70% in March
2020 to <50% in October 2020) reported by a recent review
The findings of our study suggest that five HBM constructs—in
[60]. Undoubtedly, eliminating vaccine hesitancy would be
the absence of perceived barriers, a high level of perceived
beneficial to voluntary vaccination behaviors, as seen in this
benefits, and self-efficacy—as well as individuals’ agreement
study, which showed that the vaccination rate was nearly 8 times
with recommendations from authorities and friends or family
higher among the participants who were accepting of vaccines
were negatively associated with COVID-19 vaccine hesitancy
compared to those who were vaccine hesitant. In the SEM
and positively associated with vaccination behavior.
results, vaccine hesitancy was also strongly negatively
Furthermore, psychological mechanisms were found to mediate
associated with vaccination behaviors (Hypothesis 1 confirmed).
the relationship between perceived barriers, perceived benefits,
Therefore, the control of vaccine hesitancy and the promotion
recommendations from authorities and friends or family, and
of voluntary vaccination still seem to be challenges in the
vaccination uptake behavior via vaccine hesitancy. Self-efficacy,
context of the COVID-19 pandemic.
perceived barriers, and perceived benefits were correlated with

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 11

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

In this study, female participants showed more COVID-19 decision-making process relies on a trade-off between benefits
vaccine hesitancy, which is consistent with previous findings and risks [66]. In addition to cues to action, this result was
in the literature [61,62]; a possible reason for this finding is that consistent with a previous study showing that compliance with
women are more likely to be concerned about side effects [63] recommendations from health workers may also be correlated
and take nonpharmaceutical protective measures (eg, masking with confidence in vaccine efficacy [73], because they can share
and maintaining social distance) [64], while men are more personal knowledge about being immunized and motivate
inclined to adopt medical intervention [65]. Medical personnel vaccine uptake efficacy [75].
showed less vaccine hesitancy and a much higher vaccination
In addition to the direct and mediating effect of self-efficacy,
rate in this study, which may be inconsistent with the general
some HBM constructs were directly associated with vaccine
argument that health workers have strong negative attitudes
hesitancy and vaccination behavior. Perceived barriers were
toward vaccines, with strong skepticism about their safety and
both positively correlated with vaccine hesitancy and detrimental
effectiveness, especially regarding the influenza vaccine [66,67].
to vaccination, as measured by the safety, side effects, and
Another finding seems unexpected; that is, that the participants
inaccessibility of the COVID-19 vaccine, in which safety may
with higher monthly salaries were associated with both vaccine
influence self-efficacy as aforementioned, while inaccessibility
hesitancy and a higher vaccination rate; in other words, even
would hinder the perceived convenience of COVID-19
though these individuals were vaccine hesitant, they were still
vaccination behavior directly. With a more specific formulation,
vaccinated. Vaccine hesitancy was not only a direct determinant
a controlled before-and-after trial study showed that arranging
of vaccination but also a perceived barrier. Participants with
time and transportation were key predictors of both intention
higher salaries were more likely to have higher socioeconomic
and behavior regarding influenza vaccination [76]. A previous
status [68], so they could more easily access social resources;
survey also found that the side effects and safety of influenza
that is, they had lower barriers to obtaining vaccines, which
vaccination were the most common reasons for vaccine
could then increase the vaccination rate among this group.
hesitancy [77]. Perceived benefits were associated with vaccine
Although some of the HBM factors were not directly associated hesitancy, which was measured by preventing the self and one’s
with the vaccination rate, perceived benefits of vaccination, family from being infected after COVID-19 vaccination. From
perceived barriers to vaccination, self-efficacy for vaccination, an altruistic motivation perspective, people could be vaccinated
and recommendations from authorities were correlated with to protect not only themselves but also their loved ones; in other
vaccine hesitancy (Hypothesis 2 partially confirmed), which words, there could be more willingness to receive the vaccine
was consistent with previous research among the Malaysian if individuals believe that it helps reduce the transmission of
public [36] and the Chinese general population [69]. In all HBM COVID-19 [78]. Recommendations from family were found to
constructs associated with vaccine hesitancy and vaccination, be directly associated with vaccination behavior in this study.
self-efficacy for COVID-19 vaccination was an important An online survey in Canada showed that respondents reported
predictor of vaccination behaviors, via vaccine hesitancy. This that encouragement from both colleagues and employers was
result is similar to the findings of previous studies on influenza beneficial to their vaccination decision-making process [55].
vaccination, according to which self-efficacy is a key factor of Another finding implied that a recommendation from a spouse
willingness, which in turn predicts behavior [46,70]. or a friend is an important cue to action in determining
Self-efficacy also plays a mediating role between vaccine willingness to accept the Zika virus vaccine [79]. However,
hesitancy and other HBM components, including perceived perceived susceptibility and severity were not enough to reduce
barriers, perceived benefits, and recommendations from vaccine hesitancy and promote vaccination behavior. A review
authorities and friends or family, and it indirectly influences indicated that perceived barriers were the most powerful single
vaccination uptake. This finding was supported by the HBM predictor of preventive health behavior across all studies and
hypothesis (Hypothesis 3 partially confirmed) that HBM behaviors, and perceived severity was the least powerful
constructs and cues to action may not share a juxtaposition or predictor [71].
parallel relationship, but self-efficacy functioned as a serial
From the perspective of the HBM on understanding vaccination
mediator [71]. Hilyard et al noted that public self-efficacy for
behavior, it is valuable that self-efficacy is an important and
COVID-19 vaccination could be promoted by enhancing the
direct predictor of COVID-19 vaccine hesitancy because it can
perceived benefits of vaccination, confidence in overcoming
also mediate the influences from cues to action, perceived
possible side effects (ie, perceived barriers), and
barriers, and perceived benefits. Furthermore, vaccine hesitancy
recommendations from authorities, such as the Obamas’
was strongly correlated with vaccination behavior but was not
modeling of H1N1 vaccine acceptance for their daughters [52].
the only determinant, since perceived barriers and
In this study, self-efficacy was measured as a specific domain
recommendations from friends or family were also associated
with confidence in the safety of the COVID-19 vaccine, a low
with vaccination behavior directly and in combination.
prevalence of side effects of the COVID-19 vaccine, and success
in dealing with side effects. Vaccine safety or side effects, which In practice, it is valuable for other nations to know what the
are regarded as contributing to the development of disease, are Chinese vaccine hesitancy and vaccination statuses were at the
of paramount importance to individual efficacy when deciding beginning of the critical period when COVID-19 vaccination
whether to vaccinate [72,73] and are even relevant aspects that became available to the public, free of charge. This finding
help explain the antivaccine movement in Europe [74]. A study indicates that health authorities or doctors may be less effective
argues that a perceived risk-benefit balance may influence in motivating people to action, while it may be useful to
confidence in vaccine uptake; in other words, a combined advocate for more volunteers to engage in motivating their
https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 12
(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

friends or family members. Although the antivaccine movement Limitations

that occurred in other nations was not popular in mainland There are some potential limitations to this study. First, due to
China, vaccine hesitancy and refusal were not rare occurrences the convenience sampling and snowball recruitment methods
without mandatory vaccination in this study. Moreover, it is that were part of the online survey process, selection bias, such
essential to reinforce the publishing of information regarding as the participation of fewer respondents with low education
the safety and validity of COVID-19 vaccines and incentives attainment and fewer older adults (aged over 50 years), may
of vaccination completion, which could then promote public have affected the generalizability of the results. Second, vaccine
confidence in overcoming vaccination barriers and in achieving hesitancy was measured by a single item derived from a
benefits after vaccination. definition from the SAGE working group, which may promote
In summary, there was a high prevalence of vaccine hesitancy more accurate measurement tools in future research.
and low vaccination behavior in China during the first month Furthermore, the vaccination rate in this study may not reflect
(January 2021) when vaccinations became available to the future trends because only some participants had received the
vaccine-eligible population. The HBM framework is a useful vaccine in a timely manner, vaccinations were available to the
framework to guide the development of future campaigns to public for only 1 month, and there were no incentives except
reduce vaccine hesitancy and promote COVID-19 vaccination. to receive a free vaccination before participating in the study.

Acknowledgments
This work was supported by the National Key R&D Program of China (grants 2018YFC2002000 and 2018YFC2002001) and
the National Natural Science Foundation of China (grant 71573048).

Authors' Contributions
JD, JG, HF, PZ, and YJ designed the study and obtained the data. HC and X Li undertook the analysis, supervised by JD, JG,
and HF, and wrote the manuscript. X Liu translated the questionnaire. HC, X Li, X Liu, YM, and RW performed the survey. All
authors read the final manuscript and agreed with the content.

Conflicts of Interest
None declared.

References
1. COVID-19 Weekly Epidemiological Update. Geneva, Switzerland: World Health Organization; 2021 Jan 12. URL: https:/
/www.who.int/publications/m/item/weekly-epidemiological-update---12-january-2021 [accessed 2021-08-13]
2. WHO Coronavirus (COVID-19) Dashboard. Geneva, Switzerland: World Health Organization; 2021. URL: https://covid19.
who.int/ [accessed 2021-08-13]
3. Dubé E, Laberge C, Guay M, Bramadat P, Roy R, Bettinger JA. Vaccine hesitancy: An overview. Hum Vaccin Immunother
2013 Aug;9(8):1763-1773 [FREE Full text] [doi: 10.4161/hv.24657] [Medline: 23584253]
4. Rappuoli R, Mandl CW, Black S, De Gregorio E. Vaccines for the twenty-first century society. Nat Rev Immunol 2011
Nov 04;11(12):865-872 [FREE Full text] [doi: 10.1038/nri3085] [Medline: 22051890]
5. Schaffer DeRoo S, Pudalov NJ, Fu LY. Planning for a COVID-19 vaccination program. JAMA 2020 Jun
23;323(24):2458-2459. [doi: 10.1001/jama.2020.8711] [Medline: 32421155]
6. The new coronavirus vaccine is free for all! Health insurance fund and public finance share the burden. Government of the
People's Republic of China. 2021. URL: http://www.gov.cn/fuwu/2021-01/09/content_5578430.htm [accessed 2021-08-13]
7. Experts: China's vaccine adverse reaction monitoring system will continue to track the relevant situation after COVID-19
vaccination. Government of the People's Republic of China. 2021. URL: http://www.gov.cn/xinwen/2021-01/03/
content_5576402.htm [accessed 2021-08-13]
8. Who can be vaccinated? Where can I get the vaccine? What to do if there is an adverse reaction? Authorities respond to
seven questions about the novel coronavirus vaccine. Government of the People's Republic of China. 2020. URL: http:/
/www.gov.cn/xinwen/2020-12/19/content_5571273.htm [accessed 2021-08-13]
9. National Health Commission. As of January 26, 22.767 million doses of COVID-19 vaccine had been administered. People's
Daily Online. 2021. URL: http://health.people.com.cn/n1/2021/0127/c14739-32014069.html [accessed 2021-08-13]
10. The State Information Office of the People's Republic of China held a press conference on the joint prevention and control
of novel coronavirus pneumonia. The State Information Office of the People's Republic of China. 2020. URL: http://www.
scio.gov.cn/xwfbh/xwbfbh/wqfbh/42311/42478/index.htm [accessed 2021-08-13]
11. Local press release. The State Council Information Office of the People's Republic of China. 2021. URL: http://www.
scio.gov.cn/xwFbh/gssxwfbh/index.htm [accessed 2021-08-13]
12. Shandong held a press conference on the epidemic prevention and control work and the next steps. The State Council
Information Office of the People's Republic of China. 2020. URL: http://www.scio.gov.cn/xwFbh/gssxwfbh/xwfbh/shandong/
Document/1672973/1672973.htm [accessed 2021-08-13]
https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 13
(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

13. Novel coronavirus (COVID-19): Global vaccination status inquiry. Win.d. URL: http://wx.wind.com.cn/WindSariWeb/
sari/messageVaccin.html?lan=en [accessed 2021-08-13]
14. Mo PK, Luo S, Wang S, Zhao J, Zhang G, Li L, et al. Intention to receive the COVID-19 vaccination in China: Application
of the diffusion of innovations theory and the moderating role of openness to experience. Vaccines (Basel) 2021 Feb
05;9(2):129 [FREE Full text] [doi: 10.3390/vaccines9020129] [Medline: 33562894]
15. Siddiqui M, Salmon DA, Omer SB. Epidemiology of vaccine hesitancy in the United States. Hum Vaccin Immunother
2013 Dec;9(12):2643-2648 [FREE Full text] [doi: 10.4161/hv.27243] [Medline: 24247148]
16. Immunization, vaccines and biologicals. World Health Organization. URL: https://www.who.int/teams/
regulation-prequalification/eul/immunization-vaccines-and-biologicals [accessed 2021-08-13]
17. MacDonald NE, SAGE Working Group on Vaccine Hesitancy. Vaccine hesitancy: Definition, scope and determinants.
Vaccine 2015 Aug 14;33(34):4161-4164 [FREE Full text] [doi: 10.1016/j.vaccine.2015.04.036] [Medline: 25896383]
18. Ten threats to global health in 2019. World Health Organization. URL: https://www.who.int/news-room/spotlight/
ten-threats-to-global-health-in-2019 [accessed 2021-08-13]
19. SAGE Working Group on Vaccine Hesitancy. What influences vaccine acceptance: A model of determinants of vaccine
hesitancy. World Health Organization. 2013 Mar 18. URL: https://www.who.int/immunization/sage/meetings/2013/april/
1_Model_analyze_driversofvaccineConfidence_22_March.pdf [accessed 2021-08-13]
20. Yu WZ, Ji SS, Liu J. Continuous monitoring of parental confidence on vaccination following the Shandong illegal vaccine
selling event. China Vaccines Immun 2016 Dec 6;24(02):230-236.
21. Larson HJ, Cooper LZ, Eskola J, Katz SL, Ratzan S. Addressing the vaccine confidence gap. Lancet 2011
Aug;378(9790):526-535. [doi: 10.1016/s0140-6736(11)60678-8]
22. Larson H, Paterson PB, Erondu N. The globalization of risk and risk perception. Drug Saf 2012 Dec 13;35(11):1053-1059.
[doi: 10.1007/bf03261991]
23. Barello S, Nania T, Dellafiore F, Graffigna G, Caruso R. 'Vaccine hesitancy' among university students in Italy during the
COVID-19 pandemic. Eur J Epidemiol 2020 Aug;35(8):781-783 [FREE Full text] [doi: 10.1007/s10654-020-00670-z]
[Medline: 32761440]
24. Dror AA, Eisenbach N, Taiber S, Morozov NG, Mizrachi M, Zigron A, et al. Vaccine hesitancy: The next challenge in the
fight against COVID-19. Eur J Epidemiol 2020 Aug;35(8):775-779. [doi: 10.1007/s10654-020-00671-y] [Medline: 32785815]
25. Wang W, Wu Q, Yang J, Dong K, Chen X, Bai X, et al. Global, regional, and national estimates of target population sizes
for COVID-19 vaccination: Descriptive study. BMJ 2020 Dec 15;371:m4704 [FREE Full text] [doi: 10.1136/bmj.m4704]
[Medline: 33323388]
26. Neumann-Böhme S, Varghese NE, Sabat I, Barros PP, Brouwer W, van Exel J, et al. Once we have it, will we use it? A
European survey on willingness to be vaccinated against COVID-19. Eur J Health Econ 2020 Sep;21(7):977-982 [FREE
Full text] [doi: 10.1007/s10198-020-01208-6] [Medline: 32591957]
27. Guidry JP, Laestadius LI, Vraga EK, Miller CA, Perrin PB, Burton CW, et al. Willingness to get the COVID-19 vaccine
with and without emergency use authorization. Am J Infect Control 2021 Feb;49(2):137-142 [FREE Full text] [doi:
10.1016/j.ajic.2020.11.018] [Medline: 33227323]
28. Ward JK, Peretti-Watel P, Bocquier A, Seror V, Verger P. Vaccine hesitancy and coercion: All eyes on France. Nat Immunol
2019 Oct;20(10):1257-1259. [doi: 10.1038/s41590-019-0488-9] [Medline: 31477920]
29. Sherman SM, Smith LE, Sim J, Amlôt R, Cutts M, Dasch H, et al. COVID-19 vaccination intention in the UK: Results
from the COVID-19 vaccination acceptability study (CoVAccS), a nationally representative cross-sectional survey. Hum
Vaccin Immunother 2021 Jun 03;17(6):1612-1621 [FREE Full text] [doi: 10.1080/21645515.2020.1846397] [Medline:
33242386]
30. Abraham C, Sheeran P. The health belief model. In: Ayers S, Baum A, McManus C, Newman S, Wallston K, Weinman J,
et al, editors. Cambridge Handbook of Psychology, Health and Medicine. 2nd edition. Cambridge, UK: Cambridge University
Press; 2007:97-102.
31. Ajzen I. The theory of planned behavior. Organ Behav Hum Decis Process 1991 Dec;50(2):179-211. [doi:
10.1016/0749-5978(91)90020-t]
32. Mckenna SP. Predicting health behaviour: Research and practice with social cognition models. Saf Sci 1996
Dec;24(3):229-230. [doi: 10.1016/s0925-7535(97)81483-x]
33. Rosenstock IM, Strecher VJ, Becker MH. Social learning theory and the Health Belief Model. Health Educ Q
1988;15(2):175-183. [doi: 10.1177/109019818801500203] [Medline: 3378902]
34. Juraskova I, O'Brien M, Mullan B, Bari R, Laidsaar-Powell R, McCaffery K. HPV vaccination and the effect of information
framing on intentions and behaviour: An application of the theory of planned behaviour and moral norm. Int J Behav Med
2012 Dec;19(4):518-525. [doi: 10.1007/s12529-011-9182-5] [Medline: 21879340]
35. Coe AB, Gatewood SB, Moczygemba LR, Goode J, Beckner JO. The use of the health belief model to assess predictors of
intent to receive the novel (2009) H1N1 influenza vaccine. Innov Pharm 2012;3(2):1-11 [FREE Full text] [doi:
10.24926/iip.v3i2.257] [Medline: 22844651]

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 14

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

36. Wong LP, Alias H, Wong P, Lee HY, AbuBakar S. The use of the health belief model to assess predictors of intent to
receive the COVID-19 vaccine and willingness to pay. Hum Vaccin Immunother 2020 Sep 01;16(9):2204-2214 [FREE
Full text] [doi: 10.1080/21645515.2020.1790279] [Medline: 32730103]
37. Mercadante AR, Law AV. Will they, or won't they? Examining patients' vaccine intention for flu and COVID-19 using the
Health Belief Model. Res Social Adm Pharm 2021 Sep;17(9):1596-1605 [FREE Full text] [doi:
10.1016/j.sapharm.2020.12.012] [Medline: 33431259]
38. Quah SR. The health belief model and preventive health behaviour in Singapore. Soc Sci Med 1985 Jan;21(3):351-363.
[doi: 10.1016/0277-9536(85)90112-1]
39. Orji R, Vassileva J, Mandryk R. Towards an effective health interventions design: An extension of the health belief model.
Online J Public Health Inform 2012;4(3):1-31 [FREE Full text] [doi: 10.5210/ojphi.v4i3.4321] [Medline: 23569653]
40. Rahman M, Berenson AB, Herrera SR. Perceived susceptibility to pregnancy and its association with safer sex, contraceptive
adherence and subsequent pregnancy among adolescent and young adult women. Contraception 2013 Apr;87(4):437-442
[FREE Full text] [doi: 10.1016/j.contraception.2012.09.009] [Medline: 23083528]
41. Myers LB, Goodwin R. Determinants of adults' intention to vaccinate against pandemic swine flu. BMC Public Health
2011 Jan 06;11(1):15 [FREE Full text] [doi: 10.1186/1471-2458-11-15] [Medline: 21211000]
42. Khodaveisi M, Salehi Khah M, Bashirian S, Karami M, Khodaveisi M. The effect of health belief model-based training on
preventive behaviors of hepatitis B in addicts. Int J High Risk Behav Addict 2018 Jan 09;7(2):e58579 [FREE Full text]
[doi: 10.5812/ijhrba.58579]
43. Donadiki E, Jiménez-García R, Hernández-Barrera V, Sourtzi P, Carrasco-Garrido P, López de Andrés A, et al. Health
Belief Model applied to non-compliance with HPV vaccine among female university students. Public Health 2014
Mar;128(3):268-273. [doi: 10.1016/j.puhe.2013.12.004] [Medline: 24529635]
44. Wagner AL, Boulton ML, Sun X, Mukherjee B, Huang Z, Harmsen IA, et al. Perceptions of measles, pneumonia, and
meningitis vaccines among caregivers in Shanghai, China, and the health belief model: A cross-sectional study. BMC
Pediatr 2017 Jun 12;17(1):143 [FREE Full text] [doi: 10.1186/s12887-017-0900-2] [Medline: 28606106]
45. Rabin C, Dutra S. Predicting engagement in behaviors to reduce the spread of COVID-19: The roles of the health belief
model and political party affiliation. Psychol Health Med 2021 Apr 27:1-10. [doi: 10.1080/13548506.2021.1921229]
[Medline: 33906540]
46. Fall E, Izaute M, Chakroun-Baggioni N. How can the health belief model and self-determination theory predict both
influenza vaccination and vaccination intention? A longitudinal study among university students. Psychol Health 2018
Jun;33(6):746-764. [doi: 10.1080/08870446.2017.1401623] [Medline: 29132225]
47. Webb TL, Sheeran P. How do implementation intentions promote goal attainment? A test of component processes. J Exp
Soc Psychol 2007 Mar;43(2):295-302. [doi: 10.1016/j.jesp.2006.02.001]
48. Brewer NT, Rimer BK. Perspectives on health behavior theories that focus on individuals. In: Glanz K, Rimer BK, Viswanath
K, editors. Health Behavior and Health Education: Theory, Research, and Practice. 4th edition. San Francisco, CA:
Jossey-Bass; 2008:149-165.
49. Christy SM, Winger JG, Mosher CE. Does self-efficacy mediate the relationships between social-cognitive factors and
intentions to receive HPV vaccination among young women? Clin Nurs Res 2019 Jul;28(6):708-725 [FREE Full text] [doi:
10.1177/1054773817741590] [Medline: 29134823]
50. Guidry JP, Carlyle KE, Perrin PB, LaRose JG, Ryan M, Messner M. A path model of psychosocial constructs predicting
future Zika vaccine uptake intent. Vaccine 2019 Aug 23;37(36):5233-5241. [doi: 10.1016/j.vaccine.2019.07.064] [Medline:
31375439]
51. Gerend MA, Shepherd JE. Predicting human papillomavirus vaccine uptake in young adult women: Comparing the health
belief model and theory of planned behavior. Ann Behav Med 2012 Oct;44(2):171-180 [FREE Full text] [doi:
10.1007/s12160-012-9366-5] [Medline: 22547155]
52. Hilyard KM, Quinn SC, Kim KH, Musa D, Freimuth VS. Determinants of parental acceptance of the H1N1 vaccine. Health
Educ Behav 2014 Jun;41(3):307-314 [FREE Full text] [doi: 10.1177/1090198113515244] [Medline: 24369176]
53. Wang C, Han B, Zhao T, Liu H, Liu B, Chen L, et al. Vaccination willingness, vaccine hesitancy, and estimated coverage
at the first round of COVID-19 vaccination in China: A national cross-sectional study. Vaccine 2021 May
18;39(21):2833-2842 [FREE Full text] [doi: 10.1016/j.vaccine.2021.04.020] [Medline: 33896661]
54. Wang Q, Xiu S, Zhao S, Wang J, Han Y, Dong S, et al. Vaccine hesitancy: COVID-19 and influenza vaccine willingness
among parents in Wuxi, China-A cross-sectional study. Vaccines (Basel) 2021 Apr 01;9(4):342 [FREE Full text] [doi:
10.3390/vaccines9040342] [Medline: 33916277]
55. Nowrouzi-Kia B, McGeer A. External cues to action and influenza vaccination among post-graduate trainee physicians in
Toronto, Canada. Vaccine 2014 Jun 24;32(30):3830-3834. [doi: 10.1016/j.vaccine.2014.04.067] [Medline: 24837775]
56. Kondo N, van Dam RM, Sembajwe G, Subramanian SV, Kawachi I, Yamagata Z. Income inequality and health: The role
of population size, inequality threshold, period effects and lag effects. J Epidemiol Community Health 2012 Jun;66(6):e11.
[doi: 10.1136/jech-2011-200321] [Medline: 22012964]
57. Muthén LK, Muthén BO. Mplus User’s Guide. 8th edition. Los Angeles, CA: Muthén & Muthén; 2017 Apr. URL: https:/
/www.statmodel.com/download/usersguide/MplusUserGuideVer_8.pdf [accessed 2021-08-17]

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 15

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

58. Byrne BM, Crombie G. Modeling and testing change: An introduction to the latent growth curve model. Underst Stat 2003
Aug;2(3):177-203. [doi: 10.1207/s15328031us0203_02]
59. Lazarus JV, Ratzan SC, Palayew A, Gostin LO, Larson HJ, Rabin K, et al. A global survey of potential acceptance of a
COVID-19 vaccine. Nat Med 2021 Feb;27(2):225-228 [FREE Full text] [doi: 10.1038/s41591-020-1124-9] [Medline:
33082575]
60. Lin C, Tu P, Beitsch LM. Confidence and receptivity for COVID-19 vaccines: A rapid systematic review. Vaccines (Basel)
2020 Dec 30;9(1):16 [FREE Full text] [doi: 10.3390/vaccines9010016] [Medline: 33396832]
61. Bish A, Yardley L, Nicoll A, Michie S. Factors associated with uptake of vaccination against pandemic influenza: A
systematic review. Vaccine 2011 Sep 02;29(38):6472-6484. [doi: 10.1016/j.vaccine.2011.06.107] [Medline: 21756960]
62. Baumgaertner B, Ridenhour BJ, Justwan F, Carlisle JE, Miller CR. Risk of disease and willingness to vaccinate in the
United States: A population-based survey. PLoS Med 2020 Oct;17(10):e1003354 [FREE Full text] [doi:
10.1371/journal.pmed.1003354] [Medline: 33057373]
63. Freed GL, Clark SJ, Butchart AT, Singer DC, Davis MM. Parental vaccine safety concerns in 2009. Pediatrics 2010
Apr;125(4):654-659. [doi: 10.1542/peds.2009-1962] [Medline: 20194286]
64. Bish A, Michie S. Demographic and attitudinal determinants of protective behaviours during a pandemic: A review. Br J
Health Psychol 2010 Nov;15(Pt 4):797-824 [FREE Full text] [doi: 10.1348/135910710X485826] [Medline: 20109274]
65. Moran KR, Del Valle SY. A meta-analysis of the association between gender and protective behaviors in response to
respiratory epidemics and pandemics. PLoS One 2016;11(10):e0164541 [FREE Full text] [doi: 10.1371/journal.pone.0164541]
[Medline: 27768704]
66. Rey D, Fressard L, Cortaredona S, Bocquier A, Gautier A, Peretti-Watel P. Vaccine hesitancy in the French population in
2016, and its association with vaccine uptake and perceived vaccine risk-benefit balance. Euro Surveill 2018 Apr
23;23(17):1-10 [FREE Full text] [doi: 10.2807/1560-7917.es.2018.23.17.17-00816]
67. Raude J, Fressard L, Gautier A, Pulcini C, Peretti-Watel P, Verger P. Opening the 'vaccine hesitancy' black box: How trust
in institutions affects French GPs' vaccination practices. Expert Rev Vaccines 2016 Jul;15(7):937-948. [doi:
10.1080/14760584.2016.1184092] [Medline: 27140417]
68. Wang K, Wong EL, Ho K, Cheung AW, Yau PS, Dong D, et al. Change of willingness to accept COVID-19 vaccine and
reasons of vaccine hesitancy of working people at different waves of local epidemic in Hong Kong, China: Repeated
cross-sectional surveys. Vaccines (Basel) 2021 Jan 18;9(1):62 [FREE Full text] [doi: 10.3390/vaccines9010062] [Medline:
33477725]
69. Lin Y, Hu Z, Zhao Q, Alias H, Danaee M, Wong LP. Understanding COVID-19 vaccine demand and hesitancy: A nationwide
online survey in China. PLoS Negl Trop Dis 2020 Dec;14(12):e0008961 [FREE Full text] [doi: 10.1371/journal.pntd.0008961]
[Medline: 33332359]
70. Ernsting A, Gellert P, Schneider M, Lippke S. A mediator model to predict workplace influenza vaccination behaviour--An
application of the health action process approach. Psychol Health 2013;28(5):579-592. [doi: 10.1080/08870446.2012.753072]
[Medline: 23259583]
71. Jones CL, Jensen JD, Scherr CL, Brown NR, Christy K, Weaver J. The Health Belief Model as an explanatory framework
in communication research: Exploring parallel, serial, and moderated mediation. Health Commun 2015;30(6):566-576
[FREE Full text] [doi: 10.1080/10410236.2013.873363] [Medline: 25010519]
72. Chantler TEA, Lees A, Moxon ER, Mant D, Pollard AJ, Fiztpatrick R. The role familiarity with science and medicine plays
in parents' decision making about enrolling a child in vaccine research. Qual Health Res 2007 Mar;17(3):311-322. [doi:
10.1177/1049732306298561] [Medline: 17301340]
73. Reynolds D, O'Connell KA. Testing a model for parental acceptance of human papillomavirus vaccine in 9- to 18-year-old
girls: A theory-guided study. J Pediatr Nurs 2012 Dec;27(6):614-625. [doi: 10.1016/j.pedn.2011.09.005] [Medline: 22020360]
74. Fournet N, Mollema L, Ruijs WL, Harmsen IA, Keck F, Durand JY, et al. Under-vaccinated groups in Europe and their
beliefs, attitudes and reasons for non-vaccination: Two systematic reviews. BMC Public Health 2018 Jan 30;18(1):196
[FREE Full text] [doi: 10.1186/s12889-018-5103-8] [Medline: 29378545]
75. Kempe A, O'Leary ST, Kennedy A, Crane LA, Allison MA, Beaty BL, et al. Physician response to parental requests to
spread out the recommended vaccine schedule. Pediatrics 2015 Apr;135(4):666-677 [FREE Full text] [doi:
10.1542/peds.2014-3474] [Medline: 25733753]
76. Payaprom Y, Bennett P, Alabaster E, Tantipong H. Using the Health Action Process Approach and implementation intentions
to increase flu vaccine uptake in high risk Thai individuals: A controlled before-after trial. Health Psychol 2011
Jul;30(4):492-500. [doi: 10.1037/a0023580] [Medline: 21534678]
77. Lau AYS, Sintchenko V, Crimmins J, Magrabi F, Gallego B, Coiera E. Impact of a web-based personally controlled health
management system on influenza vaccination and health services utilization rates: A randomized controlled trial. J Am
Med Inform Assoc 2012;19(5):719-727 [FREE Full text] [doi: 10.1136/amiajnl-2011-000433] [Medline: 22582203]
78. Cerda AA, García LY. Hesitation and refusal factors in individuals' decision-making processes regarding a coronavirus
disease 2019 vaccination. Front Public Health 2021;9:626852 [FREE Full text] [doi: 10.3389/fpubh.2021.626852] [Medline:
33968880]

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 16

(page number not for citation purposes)
XSL• FO
RenderX
 JOURNAL OF MEDICAL INTERNET RESEARCH Chen et al

79. Wong LP, Alias H, Hassan J, AbuBakar S. Attitudes towards Zika screening and vaccination acceptability among pregnant
women in Malaysia. Vaccine 2017 Oct 13;35(43):5912-5917. [doi: 10.1016/j.vaccine.2017.08.074] [Medline: 28886944]

Abbreviations
CFI: comparative fit index
d: precision
deff: design effect
DIT: diffusion of innovation theory
HBM: health belief model
HPV: human papillomavirus
OR: odds ratio
p: prevalence of COVID-19 vaccination hesitancy
PCA: principal component analysis
RMSEA: root mean square error of approximation
SAGE: Strategic Advisory Group of Experts
SEM: structural equation modeling
SSC: standardized structure coefficient
TLI: Tucker-Lewis index
TPB: theory of planned behavior
WHO: World Health Organization

Edited by C Basch; submitted 02.04.21; peer-reviewed by Q Yang, PKH Mo; comments to author 06.05.21; revised version received
26.06.21; accepted 12.07.21; published 06.09.21
Please cite as:
Chen H, Li X, Gao J, Liu X, Mao Y, Wang R, Zheng P, Xiao Q, Jia Y, Fu H, Dai J
Health Belief Model Perspective on the Control of COVID-19 Vaccine Hesitancy and the Promotion of Vaccination in China: Web-Based
Cross-sectional Study
J Med Internet Res 2021;23(9):e29329
URL: https://www.jmir.org/2021/9/e29329
doi: 10.2196/29329
PMID: 34280115

©Hao Chen, Xiaomei Li, Junling Gao, Xiaoxi Liu, Yimeng Mao, Ruru Wang, Pinpin Zheng, Qianyi Xiao, Yingnan Jia, Hua Fu,
Junming Dai. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 06.09.2021. This is an
open-access article distributed under the terms of the Creative Commons Attribution License
(https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic
information, a link to the original publication on https://www.jmir.org/, as well as this copyright and license information must
be included.

https://www.jmir.org/2021/9/e29329 J Med Internet Res 2021 | vol. 23 | iss. 9 | e29329 | p. 17

(page number not for citation purposes)
XSL• FO
RenderX