Cervical cancer, a serious infectious disease mostly caused by high-risk human papillomavirus (HPV) types, is the fourth most common reason worldwide for tumor-related hospital admission in women [1,2]. In 2018, in China, cervical cancer resulted in 48,000 deaths and 106,000 cases [1]. Previous studies found that young women were at higher risk of HPV infection, and showed elevated HPV infection problems owing to their age and the sexually transmitted nature of the virus [3-5]. Female college students, most of whom were ready for sexual debut and had relatively low awareness of disease prevention, similarly demonstrated HPV infection problems right after starting college [4,6].

At the end of 2020, China initiated a program for cancer prevention jointly issued by 10 ministries, including the National Health Commission, and joined in the action of the “global strategy to accelerate the elimination of cervical cancer” advocated by the World Health Organization [7-9]. Since the approval of preventive HPV vaccines in mainland China in 2016, the HPV vaccination rate has remained poor, highlighting the importance of understanding the reasons for vaccination refusal and the ways to promote HPV vaccination among young populations [6,10].

Improving the HPV vaccination rate and vaccination willingness is the most salient way to eliminate cervical cancer [11,12]. Most published studies on the willingness to undergo HPV vaccination among Chinese college students were conducted before the availability of HPV vaccines in China [13-15]. Information about the actual rate of HPV vaccination and willingness to undergo HPV vaccination after the approval of HPV vaccines is scarce. Interventions guided by health behavioral theories were widely applied in previous studies [16-19]. A review conducted in 2018 identified 70 scientific articles providing supportive evidence that education guided by health behavioral theories was effective in promoting HPV vaccination [20]. However, only a few interventions on the perceptions of HPV vaccination have been published in China, and they involved limited sample sizes and did not involve health behavioral theories [21-23].

Currently, in China, there is a free HPV vaccination program in pilot regions among school girls less than 14 years old [24], but there are no effective strategies to improve HPV vaccine coverage among female college students who are at higher risk of HPV infection and have been suggested to be a “catch-up” population for HPV vaccination in many other countries. The information-motivation-behavioral skills (IMB) model is one of the commonly used theories on health promotion. To our knowledge, no previous study has been performed to evaluate IMB model–based online education for HPV vaccination among female college students from different regions in mainland China, with a follow-up duration of 3 months.

This study has the following 3 objectives: (1) to evaluate the feasibility and acceptability of the IMB model–based online intervention; (2) to examine whether and how this intervention improves HPV vaccination; and (3) to identify the barriers and facilitators of HPV vaccination among female college students in mainland China.

The protocol for this study was published before the interventions were carried out [25]. Briefly, this study was a multicenter 2-arm cluster randomized trial. First-year female college students from 7 universities from different geographical locations in mainland China were recruited and randomly assigned 1:1 to either an IMB model–based intervention group or a waitlist control group based on their class. The intervention group received online education to promote the willingness to undergo HPV vaccination for 7 consecutive days, which was guided by the IMB model and conducted on a communication platform called DingTalk (Alibaba Group). The intervention group was compared with the waitlist control group in terms of the perceptions of HPV vaccination and willingness to undergo HPV vaccination. Recruitment of the study participants began in February 2020, and data collection was completed in July 2020. The whole process of participant recruitment, data collection, and intervention was conducted online, which coincided with the outbreak of COVID-19 in China, when Chinese college students were all self-isolated at home. The CHERRIES (Checklist for Reporting Results of Internet E-Surveys) checklist was used to guide the reporting of our web-based survey [26]. The CONSORT-EHEALTH (Consolidated Standards of Reporting Trials of Electronic and Mobile Health Applications and Online Telehealth) checklist is presented in Multimedia Appendix 1 [27].

Participating students were recruited through notices on campus or advertisements in social media groups, posted in advance by partner teachers. Interested participants scanned the QR code on the notices or advertisements to fill in their class name and provide informed consent regarding the research objectives, requirements, procedures, benefits, and other study-related information. Potential participants were assessed for eligibility before the baseline survey. The eligibility criteria for the study were as follows: (1) female sex; (2) age ≥18 years; (3) first-year college student; (4) no vaccination contraindications; and (5) accessibility to computers or smartphones. Enrolled participants were given a link to either the intervention group or the control group that had already been set up in advance in DingTalk.

In order to facilitate the management of participants and reduce loss of follow-up, enrolled participants were randomly assigned by class to the intervention and control groups. The investigator in each college, who was blinded to the identity of the participants, used computer software (Excel program, Microsoft Corp) to generate a series of random numbers for both the first-year Arts and Science major classes. Eligible participants were identified as either the intervention group or the control group based on the class name they entered when they scanned the QR code on the recruiting information. Participants, data analysts, and investigators were all blinded to the randomized allocation, and only the research assistants in each center were aware of the allocation.

The study was carried out when COVID-19 broke out worldwide at the beginning of 2020. In order to increase the compliance of the control group and reduce the probability of compromising the blind nature of the study, the control group was given 7 days of information on COVID-19 prevention, which was organized and presented in the same format and platform as the educational materials in the intervention group. At the end of the study, the waitlist control group received the same educational materials as the intervention group.

The internet-based questionnaire surveys were administered at 4 time points (baseline, immediately after the intervention, and 1 month and 3 months after the intervention), and participants were given notebooks and pens with the logo of the research institution as incentives.

Background characteristics, including age, major in college, ethnicity, residence, parental residence, education and marital status, monthly living expenses (RMB), family/friends with any cancer, ever received sexual education, currently in a romantic relationship, sexual debut, and attitude toward premarital sex, were measured in this study. These variables are necessary information, basic information, or information related to HPV vaccination as shown in previous literature [30].

The primary outcome measures were self-reported willingness to undergo HPV vaccination and uptake of HPV vaccination. At baseline and each follow-up, the participants were asked “Are you willing to get the HPV vaccine in the future?” and “Have you been vaccinated against HPV?” with “Yes” and “No” response options. It should be noted that there were some differences in the way we asked questions regarding the willingness to undergo HPV vaccination. The 3-month questioning time frame was in the “future,” while the baseline and other 2 follow-up questioning time frames were “within the study period.” The differences in outcomes are detailed in the Results section.

The secondary outcome measures were the information/knowledge, motivation, and behavioral skills regarding HPV vaccination, which were designed based on the IMB model. Among them, the information part consisted of 11 questions. For example, “HPV is related to the development of cervical cancer,” with answer options “Agree,” “Disagree,” and “Do not know” (Cronbach α=.78) [30-33]. Motivation for vaccination was measured by 19 questions. For example, “Getting vaccinated for HPV will help protect me from HPV infection” (Cronbach α=.71) [31,34-36]. Behavioral skills were measured by 10 questions. For example, “I feel confident in my ability to get vaccinated for HPV, even if it is expensive” (Cronbach α=.88) [33,35]. The answers for these items were measured on a 5-point Likert scale (1=strongly disagree, 2=disagree, 3=neither disagree nor agree, 4=agree, and 5=strongly agree). A description of the baseline results for this study has been published [37].

In the 3rd month after the intervention, the participants were asked about their perceptions of the barriers and facilitators of HPV vaccination, such as the reasons for not receiving the HPV vaccine; choices of the HPV vaccine; opinions on promoting HPV vaccination in China; and willingness to receive the HPV vaccine under different scenarios.

This study was approved by the Institutional Review Board of the Chinese Center for Disease Control and Prevention on October 24, 2019 (approval number: 201918-01).

IBM SPSS Statistics 23.0 (IBM Corp) was used to process the data and conduct statistical comparisons between the intervention and control groups. Independent t tests and chi-square tests were used to compare the distributions of the continuous and categorical variables, respectively. Analyses were conducted based on an intention-to-treat approach, and statistical significance was set at P<.05 (2-sided). The effects of the IMB model–based intervention on knowledge, motivation, behavioral skills, and willingness regarding HPV vaccination were examined using generalized estimating equations (GEEs). We included significant variables in baseline chi-square analysis, and assessed group (intervention and control), time (baseline, immediately after the intervention, and 1 month and 3 months after the intervention), and time × group interaction, with the time × group interaction indicating a differential change by group from baseline to the end of the trial.

The flowchart for participant recruitment is presented in Figure 1. From February 2020 to March 2020, a total of 4051 female college students were recruited and screened, among whom 83 refused to participate in the study, 101 did not meet the inclusion criteria, and 102 reported that they had been vaccinated with the HPV vaccine prior to the study. Among the participants who met the inclusion criteria, the HPV vaccination rate was 2.6% (102/3867). A total of 3765 participants signed the informed consent and completed the baseline questionnaire (T-baseline). However, 26 participants were not grouped and withdrew from the study. There were eventually 1936 participants in the intervention group and 1803 participants in the control group. A total of 3224 (86.2%) participants completed the postintervention questionnaire (T-postintervention). In the intervention group, 1662 participants completed the IMB model–based intervention, and 1562 participants in the control group received non-HPV–related content in the same period. In addition, a total of 3215 (86.0%) participants (1670 in the intervention group and 1545 in the control group) completed the follow-up assessment 1 month after the intervention (T-1 month). Finally, a total of 3071 (82.1%) participants completed the 3-month follow-up evaluation (T-3 months), including 1582 and 1489 participants in the intervention and control groups, respectively. The proportions of participants in the intervention group and control group who completed the T-postintervention assessment were 85.9% (1662/1936) and 86.6% (1562/1803), respectively, and the difference was not statistically significant (P=.49). The completion rates of the T-1 month assessment in the intervention and control groups were 86.2% (1670/1936) and 86.6% (1562/1803), respectively, and the difference was not statistically significant (P=.62). The completion rates of the T-3 months assessment in the intervention and control groups were 81.7% (1582/1936) and 82.6% (1489/1803), respectively, with no statistically significant difference (P=.49).

The participants with at least one follow-up record were included in our analysis. A total of 3484 respondents were identified, and the mean age was 19.12 (SD 0.73) years. Over half of the respondents (1848/3484, 53.0%) were majoring in Arts in college, and 87.7% (3057/3484) were Han Chinese. Most of the respondents (2207/3484, 63.4%) and their parents (2189/3484, 62.8%) lived in urban areas during the study period. Regarding the educational background of the parents, less than half (1591/3484, 45.7%) of them had a junior high school or lower education, and only 24.6% (856/3484) of the parents had a college degree or above. Most of the respondents’ parents (2819/3484, 80.9%) were married at the time of the research, while 6.9% (239/3484) were divorced, 9.8% (340/3484) were unmarried cohabiting, and 2.5% (86/3484) were widowed. Overall, 71.3% (2483/3484) had monthly living expenses ranging from 1000 to 2000 yuan (US$ 144.35 to 288.71), and about a third (842/3484, 24.2%) of all respondents had relatives or friends with cancer. Moreover, 79.2% (2758/3484) of the respondents self-reported having ever received sexual education, while 16.9% (590/3484) were currently in a romantic relationship. Only 2.8% (97/3484) had sexual debut, while 37.6% (1310/3484) said they could accept premarital sex. As shown in Table 2, both groups were comparable in terms of their sociodemographic characteristics, except previous sexual intercourse, for which there were more participants in the intervention group than in the control group.

Baseline differences were discovered for previous sexual intercourse alone, and as such, all subsequent analyses were adjusted for previous sexual experience (Table 3).

Table 4 shows the main effect of the intervention on the willingness to undergo HPV vaccination from baseline to follow-up. At baseline, the willingness rates to undergo HPV vaccination in the intervention and control groups were 33.33% and 31.96%, respectively, and there was no significant difference between the 2 groups (χ²₁=0.75; P=.39). Immediately after the intervention, the willingness rates to undergo HPV vaccination in the intervention and control groups were 40.39% and 31.56%, respectively, with statistically significant differences between the 2 groups and within the intervention group from T-baseline to T-postintervention (χ²₁=27.11; P<.001; and χ²₁=15.43; P<.001, respectively). At T-1 month, the willingness rates to receive the HPV vaccine in the intervention and control groups were 35.65% and 31.34%, respectively. At this point, the difference between the 2 groups was statistically significant (χ²₁=6.64; P=.01), but there was no significant difference within the intervention group from T-baseline to T-1 month (χ²₁=1.40; P=.24). At T-3 months, the willingness rates to undergo HPV vaccination in the intervention and control groups were 70.59% and 66.17%, respectively, with statistically significant differences between the 2 groups and within the intervention group from T-baseline to T-3 months (χ²₁=6.81; P=.01; and χ²₁=368.59; P<.001, respectively).

At different time points, the HPV vaccination rates in the intervention and control groups were as follows: T-postintervention, 0.48% and 0.19%, respectively (χ²₁=1.98; P=.16); T-1 month, 0.72% and 0.65%, respectively (χ²₁=0.06; P=.81); T-3 months, 2.21% and 1.75%, respectively (χ²₁=0.86; P=.36).

Table 6 shows that 2060 (68.4%) participants were willing to undergo HPV vaccination and 2788 (90.52%) participants would recommend HPV vaccination to their relatives and friends. The reasons for not undergoing or recommending HPV vaccination included “worried about its side effects” (471/1694, 27.8%), “expensive” (433/1694, 25.6%), “unsure of its safety and effectiveness” (406/1694, 24.0%), “lack of knowledge about the HPV vaccine” (318/1694, 18.8%), “have no sexual behavior” (279/1694, 16.5%), and “possible needle injury” (178/1694, 10.5%).

Among the 3071 complaint participants, most of them (2760/3071, 89.9%) believed that the promotion of the HPV vaccine would be well accepted. The major concerns about the HPV vaccine were “safety” (856/3071, 27.9%) and “price” (786/3071, 25.6%), followed by “side effects” (543/3071, 17.7%) and “efficacy” (507/3071, 16.5%). In addition, there were 250 (8.1%) participants concerned about the “procedure for appointment and uptake of the HPV vaccine,” and only 94 (3.1%) participants did not express any concerns. Regarding the barriers for promoting HPV vaccination, “too expensive” (2330/3071, 75.9%), “not being covered by medical insurance” (1718/3071, 55.9%), “insufficient message from the government” (1660/3071, 54.1%), and “not being included in national immunization programs” (1587/3071, 51.7%) were major concerns. Each of the measures to promote HPV vaccination in China received more than 50% agreement from the participants, including “reasonable price” (2453/3071, 79.9%), “doctors’ recommendation” (2341/3071, 76.2%), “be convenient to schedule vaccination” (1994/3071, 64.9%), and “media publicity” (1793/3071, 58.4%).

Interestingly, the most preferable type of HPV vaccine we found was the domestically produced 2-valent HPV vaccine (1087/3010, 36.1%), followed by the 9-valent HPV vaccine (947/3010, 31.5%), the 2-valent HPV vaccine (552/3010, 18.3%), and the 4-valent HPV vaccine (425/3010, 14.1%), which were all produced abroad.

The willingness rate to receive the HPV vaccine was 94.3% (2839/3010) if the price was affordable. In addition, it increased to 95.1% (2862/3010) if the HPV vaccination was covered by medical insurance, and it increased to 95.9% (2887/3010) if the HPV vaccination was included in national immunization programs, compared with 68.4% (2060/3010) unconditional willingness.

This is the first randomized trial to study the effectiveness of an IMB model–guided education program to improve HPV vaccination among female college students in mainland China. The strengths of this study include its multicenter design, the inclusion of a representative and sufficient sample of the target population, and the application of a health behavioral theory. Owing to online education and participant management, our study was carried out as scheduled even during the period of the COVID-19 outbreak. This provided a unique opportunity to make clear statements about the feasibility and convenience of the study. In addition, most of the possible contaminations were avoided owing to home isolation of the participants during the study period. The quality control measures of the study can be considered strengths. They included a professional research design and pilot trial conducted by researchers with public health and epidemiological backgrounds, daily quizzes and regular reminders to improve the maintenance rate during follow-up, and logical questions to ensure the quality of the available data.

Several study limitations must be considered. First, the effectiveness of the study might have been compromised by cluster randomization and a mismatch in the number of participants in the intervention and control groups. In addition, although the IMB model–based intervention showed relatively desirable results, the adverse impacts of the COVID-19 pandemic could not be ignored, which included too much concern about getting COVID-19 and the possible trauma, as well as difficulties in making an appointment and getting vaccinated against HPV. Moreover, the participants were asked to self-report if they had received the HPV vaccine before and within the study; however, the accuracy of these self-reported data could not be verified through the community medical care system. Future studies with randomization at the individual level, a longer follow-up period, and regular repetition of education will allow better exploration of the effects of the intervention.

Online education based on the IMB model can improve HPV-related knowledge, as well as the motivation, behavioral skills, and willingness regarding HPV vaccination among female college students in the short term; however, its long-term effects may need to be consolidated and maintained by repeated education programs in the future.