Parental willingness of HPV vaccination in Mainland China: A meta-analysis

ABSTRACT

This meta-analysis aimed to systematically review and analyze parental awareness of human papillomavirus (HPV) and its vaccine, as well as parental willingness of the HPV vaccine in China. The literature search selected studies that met the following criteria: study published between 2009 and 2023, study design involving parents with at least one child aged ≤ 18 years, sample sizes exceeding 300, availability of data on parental willingness of the HPV vaccine or sufficient information to calculate effect sizes, and studies published in either English or Chinese. Studies that did not meet one of the above points were excluded. From an initial pool of 660 papers, 33 studies were included, encompassing a total sample size of 92,802. The analysis revealed that the pooled awareness rates of HPV and the HPV vaccine among Chinese parents were 45.0% (95% CI: 36.1–54.0%) and 41.4% (95%CI: 30.7–52.5%), respectively. The overall parental willingness for vaccinating children against HPV was 61.0% (95% CI: 53.5–68.3%). Both parental awareness and willingness of the HPV vaccine in China were found to remain low when compared to other countries.

KEYWORDS:

• Chinese
• parental willingness
• parental intention
• HPV vaccine
• meta-analysis

Introduction

Human papillomavirus (HPV) infection is associated with around 4.5% of new cancer cases and 91% of cervical cancer cases globally^1,2, with the percentages being higher in lower-middle-income countries.² In 2020, cervical cancer ranked second in both incidence and mortality rates among Chinese women, resulting in 110,000 new cases and 59,060 deaths.³ Cervical cancer incidence can be significantly reduced through HPV vaccination.⁴ Completing HPV vaccination before the age of 17 reduces the incidence by 88%.⁵

In 2018, the World Health Organization (WHO) Director-General announced a global call for action to eliminate cervical cancer, and the Global Strategy for cervical cancer elimination was later adopted by the World Health Assembly.⁶ As a part of the Strategy, WHO recommends that HPV vaccine be included in all national immunization programs (NIP) and proposed that 90% of girls worldwide be fully vaccinated by the age of 15 in 2030.⁷ By November 2023, 140 (72%) countries have introduced the HPV vaccine into their NIPs,⁸ rendering it accessible for the general public. As a result of both inaccessibility and low awareness, the HPV vaccination coverage among Chinese female adolescents (9–18 years old) is likely extremely low, although no national coverage estimates are available.

In China, as the vaccine must be paid out-of-pocket, parents become the main decision makers in the vaccination of their children.⁹ Their attitude and perception toward the vaccine greatly affects the nationwide vaccination coverage of key populations¹⁰ or even indirectly influenced the development and implementation of vaccine-related policies. However, China did not have comprehensive systematic data available associated with the intention of parents to vaccinate children. In addition, in recent years, the willingness of the HPV vaccine reported in studies from different areas varied greatly, and few reviews were conducted. Therefore, this study addressed this gap by conducting a meta-analysis of past studies to analyze quantitatively the willingness of the HPV vaccine among Chinese parents and to summarize the actual HPV vaccination coverage of children in mainland China. This project would provide the necessary data to help researchers, policymakers, and public health workers to understand the landscape of HPV vaccination and advocate for the inclusion of the vaccine in the NIP.

Methods

Inclusion and exclusion criteria

This meta-analysis incorporated studies that adhere to specific criteria to ensure the quality and relevance of the data. Firstly, the study type was confined to studies published between 2009 and 2023, correlating with the WHO’s recommendation of HPV vaccination since 2009.¹¹ Secondly, the target population consists of parents from mainland China who had one or more children aged 18 or younger. Thirdly, the study must have a sample size greater than 300. Fourthly, these studies should either provide data on parental willingness of the HPV vaccine or sufficient information to calculate effect sizes. Lastly, the studies need to be published in either English or Chinese. Study that did not meet one of the above five points was excluded. In instances where multiple studies examined the same population, preference was given to the study of higher quality as described in Methodological quality assessment.

Search strategy

A comprehensive literature search was conducted using PubMed, Science Direct, Wan Fang Data, Chongqing VIP, and Chinese National Knowledge Infrastructure (CNKI) databases. The following MESH terms and keywords were used: ‘human papillomavirus vaccine’/‘HPV vaccine’/‘cervical cancer vaccine’ and ‘parental’/‘parents’/‘mother’ and ‘intention*’/acceptance*/‘willingness*’/‘perceive*’/‘attitude*’ and ‘China’/‘Chinese.’ Additionally, the reference lists of relevant reviews were hand-searched to identify other studies of interest.

Study selection and data collection

Two reviewers independently screened all records by title and abstract, and later in full-text, against the inclusion criteria. The reviewers then completed the extraction of studies with a standardized form which included the author’s name, year of publication, survey area, sample size, proportion of female parents, age/grade of the children, education proportion of parents, awareness proportion of HPV and its vaccines, willingness to receive the HPV vaccine. To understand the general situation and trends in HPV vaccination among domestic female adolescents and female parents, as well as the differences in parents’ attitudes toward vaccinating themselves and their children, we also extracted the parents’ willingness proportion to vaccinate themselves or their spouses and vaccination proportion for female children and female parents. In cases where discrepancies occur, a decision was made with a third-party expert. As not all information was available for all included articles, in case where missing information was pertinent to the literature review, we reached out to the authors of the original research for addition.

Methodological quality assessment

The Agency for Healthcare Research and Quality (AHRQ) of the United States of America has considered bias factors from multiple perspectives and recommended quality evaluation criteria for observational studies.^12,13 The scale consists of 11-item checklist. For each question, 1 point was awarded if the answer was “yes,” and 0 if “no” or “unclear.” The total score determined the quality of the study: 0–3 points were categorized as low quality, 4–7 as moderate quality, and 8–11 as high quality. The methodological quality of each study was graded by two reviewers independently.

Meta-analysis

Heterogeneity was assessed through Higgins I² estimate. In cases where I² ≤ 50%, the included literatures were considered homogeneous, and the fixed/common effect model was used for combined analysis. In cases where I² > 50%, heterogeneity was concluded, the random effect model was used to calculate the combined effect size and generate the forest plot, and subgroup analysis was conducted. Several factors including survey time, article language, sample size, gender, age, education, HPV and vaccine awareness, and current residence were analyzed to attempt to explain the heterogeneity. Subsequently, potential publication bias was assessed through funnel plot. R 4.2.3 software was used to conduct the meta-analysis.

Results

Search results and quality assessment

Based on the search terms as listed in the methodology, a total of 660 studies were identified in the search performed in February 2023. After filtering based on the inclusion and exclusion criteria and excluding duplicate records, 33 studies were included in the meta‐analysis, including 28 cross-sectional studies and 5 interventional studies (Figure 1).^14–46 For intervention studies, we only extracted the pre-intervention data, as these data would better represent the baseline situation in the local area (the pre-intervention part was scored as cross-sectional studies). Finally, eleven studies were considered high quality; twelve were of moderate quality.

Figure 1. Flowchart of the study selection procedure.

Study characteristics

A total of 12 English studies and 21 Chinese studies were included, with the majority of them conducted in eastern China. Four out of 33 were multi-center trials. The combined sample size of all included studies was 92,802, with most participants being women. The characteristics of the included studies are detailed in Table 1.

Table 1. Basic information in each study.

ID	First Author	Year	Survey area	Sample size	Female (%)	Age/grade of the child	With university or above education（%）	Awareness of HPV（%）	Awareness of HPV vaccine（%）	Willing to vaccinate children（%）	The vaccination proportion for girls (%）	Willing to vaccinate themselves (%)	The female parents’ vaccination proportion (%)	Quality Score
1	Li Man	2011	Beijing	383	55.4	Junior high school students (12–17 years old)	–	21.4	13.3	42.8	–	70.8	–	5
2	Shao kai Zhang	2013	National multi-center	2895	62.8	Junior high school students	42.4	25.1	15.6	36.2	–	–	–	8
3	Hu Haishan	2014	Guangzhou	542	68. 5	1st grade of junior high school students	46.8	32.1	22.9	26.7	–	56. 5	–	5
4	Zhang Hui	2014	Wuhan	341	63.3	1st and 2nd grade junior high school students	41.1	22.0	14.7	36.7	–	–	–	4
5	Wei Wang	2015	Jinan	360	52.2	Junior high school students	74.6	22.6	10.2	40.8	–	–	–	4
6	Yu Yang	2016	Weihai	1578	100.0	9–17 years girls	39.8	19.3	–	26.5	–	–	–	9
7	Zhang Yuying	2016	Xiamen	2463	100.0	junior high school girls	24.9	42.5	20.0	58.0	–	33.0	–	7
8	Hu Tajing	2018	Shanghai Songjiang	3262	100.0	primary and junior high school students	37.3	38.9	27.6	56.2	–	59.0	–	9
9	Pu Chen	2018	Chengdu	368	65.2	1st grade of junior high school students	12.0	41.3	43.2	61.1	–	63.8	–	4
10	Han Yaofen	2018	Xiamen	2307	100.0	primary and junior high school girls	25.3	42.5	21.4	61.9	–	–	–	6
11	Deng Jingjing	2019	Suzhou	37032	91.7	primary and junior high school students	38.1	–	–	78.6	2.6	–	–	8
12	Wu Zhifang	2020	Shanghai	927	66.7	4th through 9th grade girls	40.3	56.4	47.3	60.7	–	–	–	8
13	Fu Shilan	2020	National multi-center	2198	63.4	Junior high school students	43.5	–	14.9	38.0	–	72.3	–	8
14	Zhong Xiaohong	2020	Guangzhou	1076	82.1	9–15 years students	34.8	44.3	61.1	86.0	–	–	12.5	6
15	He Jialu	2020	Zhejiang	673	75.9	9–14 years girls	58.3	61.4	61.4	84.4	–	–	–	6
16	Wang Binghan	2020	Zhejiang	846	82.2	9–14 years students	27.3	59.2	68.4	87.2	3.1	91.2	13.6	7
17	Yulan Lin	2020	Fujian	3586	–	6–14 years girls	9.9	–	–	83.3	–	–	–	6
18	Shan Zhu	2020	Shandong	995	91.5	12–16 years girls	11.5	–	–	59.9	–	–	–	8
19	Zhao Chunyan	2020	Tongzhou	1227	100.0	Primary and junior high school girls	59.9	59.0	57.9	71.2	1.9	–	–	6
20	Qiu Lirong	2020	Taiyuan and Bethune, Shanxi	700	92.7	9–14 years girls	63.1	–	–	46.1	0.1	–	3.9	7
21	Song Quanquan	2021	Guangyuan City, Sichuan	936	73.1	9–14 years girls	79.0	–	90.4	25.4	–	–	–	7
22	Yu Huang	2021	Zhejiang	1125	87.0	12–18 years girls	22.3	89.1	54.3	35.4	4.4	–	–	6
23	Liang Hui	2021	Zhejiang	1466	78.7	4th through 9th grade girls	28.4	19.7	16.8	49.9	–	–	–	8
24	Yan Xie	2021	Zunyi	1074	73.6	junior and senior high school girls	22.7	28.2	38.0	73.9	–	–	5.3	5
25	Shi Jinjing	2021	Henan; Zhejiang; Sichuan; Heilongjiang	4947	86.1	9–14 years girls	21.2	60.8	54.9	45.6	1.4	50.1	2.6	8
26	Qi Guo	2022	Minhang district of Shanghai	1031	80.1	9–14 years girls	63.6	–	–	78.4	4.5	62.7	18.6	7
27	Zhuoying Huang	2022	Shanghai	1041	83.0	≤18 years girls	53.0	–	–	83.0	–	–	–	8
28	Zheng Zezhou	2022	Haikou	493	100.0	Kindergarten, primary, and junior high school students	48.3	–	–	69.0	–	–	–	5
29	Yun Xuexia	2022	Guangzhou	7391	–	12–14 years girls	–	–	–	90.2	–	–	–	6
30	Rong Zhang	2022	National multi-center	1584	76.0	Junior high school students	19.2	34.0	27.9	69.0	–	73.5	–	6
31	Zhaonan Zhang	2022	Henan; Zhejiang; Sichuan; Heilongjiang	5215	86.1	9–18 years girls	21.3	61.0	55.1	46.4	1.3	52.9	–	8
32	Zhong Wenjiang	2022	Shanghai Qingpu	1309	87.9	4th grade girls	72.7	79.9	86.6	83.3	–	–	19.1	6
33	Hong Xie	2023	National multi-center	1431	83.3	9–18 years adolescents	38.0	75.5	84.7	87.6	–	84.9	–	7

The results of meta-analysis

The Higgins I² showed that I² > 50%, the random effect model was used to calculate the combined effect size and generate the forest plot. The meta-analysis found that though parents have been increasingly willing to vaccinate their children against HPV in the past decade, there was a low level of parental awareness of HPV and the HPV vaccine in China. The pooled parental awareness of HPV was 45.0% (95%CI: 36.1–54.0%, Figure 2), and that of HPV vaccine was 41.4% (95%CI: 30.7–52.5%, Figure 3). The parents’ pooled willingness to vaccinate children against HPV was 61.0% (95%CI: 53.5–68.3%, Figure 4), similar to 65.1% (95%CI: 55.3–74.2%, Supplemental Figure S1) that to themselves or their spouses. In contrast, the female children vaccination proportion against the HPV was only 3.7% (95%CI: 0.8%–8.5%, Supplemental Figure S2) and that among female parents was 9.8% (95%CI: 5.2–15.7%, Supplemental Figure S3).

Figure 2. Forest map of HPV awareness among Chinese parents.

Figure 3. Forest map of HPV vaccine awareness among Chinese parents.

Figure 4. Forest map of Chinese parents’ willingness to vaccinate girls against HPV.

Subgroup analyses

As high heterogeneity was found among these groups (all p < .05, Table 2), we compared differences between the characteristics of parents. Sensitivity analysis was first performed by excluding 3 studies with a score of 4, there were no significant changes before and after the comparison (I² = 99.8%, Z _p = 0.2405 > 0.05, Table 2).

Table 2. Subgroup meta-analysis of Chinese parents’ willingness to vaccinate girls against HPV.

Subgroup	No. studies	No. of Sample Size	No. with willingness	Heterogeneity Test Result (%): I^2	Effect Model	Integrated Intention Rate (95% CI)	Z/χ^2-test
							Z/χ^2-value	P-value
Excluding low-score (4) articles
No (Total)	33	92802	63446	99.8	Random effects model	61.0% [53.5%2–68.3%]	1.4	.241
Yes	30	91733	62949	99.8	Random effects model	62.5% [54.5%–70.1%]
Survey time
≤2016 years	7	8562	3476	98.8	Random effects model	38.1% [30.1%–46.3%]	4083.3	<.001
2016–2020 years	13	55197	41016	99.6	Random effects model	68.3% [59.1%–76.8%]
>2020 years	13	29043	18954	99.8	Random effects model	65.7% [53.3%–77.1%]
Language
Chinese	21	70887	50620	99.8	Random effects model	61.0% [51.7%–69.9%]	1284.2	<.001
English	12	21915	12826	99.7	Random effects model	61.0% [47.8%–73.5%]
Sample size
<1000	12	7564	4172	99.3	Random effects model	54.0% [41.6%–66.2%]	2437.6	<.001
1000–3000	15	23805	13989	99.7	Random effects model	63.6% [52.4%–74.0%]
>3000	6	61433	45285	99.9	Random effects model	68.2% [51.0%–83.2%]
Gender
Male	12	5145	2738	96.3	Random effects model	57.7% [49.0%–66.2%]	5.8	<.05
Female	12	15576	8591	99.6	Random effects model	57.4% [43.3%–70.9%]
Age
≤40	10	14965	9387	99.5	Random effects model	63.2% [52.7%–73.1%]	231.4	<.001
>40	10	7366	3836	98.5	Random effects model	58.2% [48.4%–67.8%]
University or above Education
Yes	18	11004	6722	99.1	Random effects model	61.7% [50.0%–72.7%]	27.0	<.001
No	18	23848	13864	99.6	Random effects model	63.1% [52.9%–72.6%]
Awareness of HPV
Yes	11	9729	6733	99.0	Random effects model	73.2% [64.9%–80.8%]	645.2	<.001
No	10	8993	4589	98.6	Random effects model	55.4% [45.9%–64.7%]
Awareness of the HPV vaccine
Yes	10	8025	5799	99.0	Random effects model	73.0% [63.3%–81.6%]	893.8	<.001
No	10	9073	4522	98.2	Random effects model	52.3% [44.4%–60.2%]
Current residence
Rural	8	11062	7876	99.7	Random effects model	60.8% [39.9%–79.9%]	10.1	<.05
Urban	8	10325	7553	99.5	Random effects model	71.8% [60.6%–81.8%]

The parents’ willingness to vaccinate children against HPV increased from 38.1% before 2016, to 68.3% after 2016, which subsequently remained similar to 65.7% after 2020, and the pooled willingness proportion in studies published in Chinese was 61.0%, same as that of the studies published in English (61.0%), and the willingness raised from 54.0% to 68.2% with the increase in sample size. In addition, there have been slight differences among studies as ages of parents, education levels of parents and localities vary. Specifically, the proportion among parents > 40 was 58.2%, comparing to 63.2% for parents ≤ 40. The proportion of parents with a university education and above (61.7%) was lower than those with below a college education (63.1%). The proportion of parents from rural areas was lower than that from urban areas (60.8% comparing to 71.8%). However, there has not been a marked difference between the willingness of fathers (57.7%) and of mothers (57.4%).

Publication bias

For the publication bias, there were slight asymmetry that could be seen through visual inspection in the contour enhanced funnel plot (Figure 5). Among them, 31 studies were not likely to have such bias, while two studies with small impact were likely to have publication bias (white areas and dark gray areas). Furthermore, the main results were not affected after excluding the two studies (p > .05).

Figure 5. Funnel plot of standard error by logit event rate.

Discussion

To our knowledge, this is the only and most recent analysis of HPV vaccine awareness and willingness among Chinese parents in the last three years. This analysis provided an accurate overview of parental awareness and willingness since 2013, the year which HPV vaccine was widely recommended by the WHO. In total, 33 studies were included, consisting of a sample size of 92,802. The low-score (4) studies and publication bias studies in this paper did not affect our results.

According to the meta-analysis, less than half of Chinese parents were aware of HPV (45.0%) and HPV vaccine (41.4%). The proportion was lower than regions or countries with HPV vaccine included in the NIP such as western Kenya (62.0% for HPV and 64.0% for HPV vaccine)⁴⁷ and the UK (55.0% for both).⁴⁸ Comparing to the awareness of 40% in non-NIP countries such as Iran,⁴⁹ Turkey⁵⁰ and Saudi Arabia,⁵¹ the awareness in China was slightly higher.

At present, Chinese parental willingness proportion of the HPV vaccine remained low at 61.0% (pooled proportion). When comparing with other non-NIP countries, although the proportion was higher than some Middle Eastern countries (Saudi Arabia,⁵¹ Iran⁴⁹ and Serbia),⁵² it was considerably lower than Bangladesh (92–99%),⁵³ Poland (54–87%),^54,55 and Nigeria (72–89%).^56–58 This means that China still has much work to do on the road to advancing HPV vaccination.

Further analysis found variabilities in willingness proportion between parents from rural and urban areas, and among parents from different education backgrounds. The lower proportion among parents from rural area (60.8%, comparing to 71.8% of urban areas) may be due to limited access to health information because of lack of health advocacy campaigns,⁵⁹ or due to competing priorities due to financial constraints, as rural parents tend to be poorer than their urban counterparts.⁶⁰ In addition, consistent with the result of another study in Kenya,⁶¹ parents with lower education levels had higher willingness proportions. This may be attributed to the fact that these parents are more likely to follow advice from medical professionals^62–64 and influenced by other parents who are more accepting.^65,66 In contrast, parents with higher education level may be more confident and thus tend to ignore the advice from their healthcare providers.⁵¹

Further research was needed to understand the low parental willingness proportion in China, as well as disparities between parental education levels and the urban-rural divide. Studies in the UK and the USA cited safety concerns as the main barrier to parental willingness.^48,67,68 While in Poland and India, parents feared that the vaccine may promote early sexual onset and sexual promiscuity among their children.^54,55,69 The findings of this study indicated that the willingness and coverage of HPV vaccination of Chinese female parents were significantly higher than that of their girls, which could also be caused by the reasons mentioned above. In a word, these studies could provide a basis to conduct in-depth qualitative research on how cultures and beliefs affect HPV vaccine perception in China,^70,71 which could be used to develop and improve HPV vaccination advocacy and communication efforts.

The effectiveness and benefit of HPV vaccines among children have been extensively documented.^8,72–74 A modeling analysis found that cervical cancer could be eliminated in several countries if resource constraints and vaccine supply shortages were addressed.⁷⁵ Another study concluded that adequate supply and improved access to HPV vaccine could result in significant reduction in cervical cancer burden and economic loss.⁷⁶ Unfortunately, despite more than half of the Chinese parents being willing to vaccinate themselves, their spouses, and their daughters against HPV, the HPV vaccine coverage in China remained remarkably low due to supply and price issues. The cumulative HPV vaccine coverage for women aged 9–45 in China from 2018 to 2020 was 2.24%, even in large cities like Beijing and Shanghai, the coverage remains only at 8.28% and 7.37%, respectively.⁷⁷ There was no statistical data on the vaccination coverage of Chinese female adolescents, while from this paper, the included studies providing the vaccination proportions of girls were all no more than 5%. In contrast, countries that include HPV vaccine in their NIP was significantly higher: In the United States, there were 76.0% of girls aged 13–17 years had received ≥ 1 HPV vaccine dose in 2022⁷⁸; In France, among girls aged 11–14, the coverage was 38.6% in 2021⁷⁹; In South Africa, among girls in public schools, 67.1% had received at least one HPV vaccine dose in 2019.⁸⁰ The coverage in China was lower even when comparing to other non-NIP countries such as Ghana (4.5%)⁸¹ and Nigra (6.9%).⁸² Price was often cited as a significant barrier to wider vaccine coverage.^6,83 In fact, financial burden and parental willingness were often interlinked: several studies suggested that parents were much more likely to accept the HPV vaccine if it was free; similar findings were reproduced in other studies.^19,39,84,85 Addressing the inaccessibility of HPV vaccine due to cost may therefore increase the parental willingness in China.

Limitation

There were several potential limitations in the research. Firstly, the analysis mainly was cross-sectional studies. Secondly, as the primary focus of this paper was on parental vaccine willingness for children, some studies on vaccine coverage may be overlooked, which may affect our summary of Chinese adolescent vaccination coverage, although a manual search for additional articles suggested the impact was small. Thirdly, there was variability in how different studies define parental vaccine willingness. While some specify that parental willingness was measured among populations aware of the vaccine, others did not make this distinction. Lastly, the majority of included studies were conducted in Eastern China, an area known for its urban and economic development. This geographic focus indicates a need for more studies encompassing diverse regions with varying characteristics to provide a more balanced view.

Conclusions

It revealed that less than half of Chinese parents were aware of HPV and the HPV vaccine, and the willingness of HPV vaccine was less than 70%. Both parental awareness, willingness, and some cities’ vaccination coverage of the HPV vaccine in China were found to remain low when comparing to other countries. Key factors influencing parental willingness included the level of awareness about HPV, educational background, and the urban-rural divide. This underscores the critical role of health education and promotion. Nevertheless, there has been a noticeable increase in parental willingness of the HPV vaccine over the past decade. This trend suggests that the willingness could rise further if the regional government implements the free vaccine strategies. The HPV vaccine should be an emerging priority for China’s NIP as a crucial next step. Our results require validation by further studies. Additionally, future studies should focus on improvements such as clearly defining willingness, elucidating reasons for case exclusions, and explaining methods for handling lost data.

Author contributions

S.S. Tan, J. Yin, X.M. Lian, and Y.L. Qiao contributed to the conception and design of this study.

S.S. Tan, S.M. Wang, X.H. Jia, and C.Y.H. Tong interpreted the data.

S.S. Tan prepared the first manuscript draft.

J. Yin, X.W. Zou, and X.M. Lian revised the draft critically, and all authors proofed reading the final version.

Acknowledgments

We would like to thank all the staffs who had taken part in this study.

Disclosure statement

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

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website at https://doi.org/10.1080/21645515.2024.2314381.

Additional information

Funding

This work was supported by the Bill & Melinda Gates Foundation, Seattle, WA under Grant [INV006373].