Comparing the video information about the HPV vaccine: a content analysis between Bilibili and YouTube
Article information
Abstract
User-generated content (UGC) platforms have become one of the new channels for the dissemination of health care information. Papillomavirus (HPV) vaccine is one of the hot topics discussed on these platforms. This study took YouTube and Bilibili as examples to compare the concrete information about HPV vaccine, including source types, regions of source, tones and vital elements of the health belief model (HBM). The significant relationships between these primary pieces of information are also validated. Based on a sample of 250 videos, this study found that most of the vaccination videos in Bilibili were produced by individual bloggers, while the sources of YouTube tended to be more diverse. The United States is the leading region of YouTube videos, while in addition to Mainland China, the emergence of other countries on Bilibili is related to overseas study or medical tourists. Aligned with previous studies, a statistical association exists between attitudes and the number of likes on YouTube. As for HBM elements, the proportions of self-efficacy and perceived benefits show the opposite results on the two video sites. Surprisingly, despite there was a significant relationship between the sources of Bilibili and perceived barriers, individual bloggers tended not to mention relevant barriers of HPV vaccine. This study not only provides a novel comparative perspective, but also offers a reliable reference for video promotion of related stakeholders of HPV vaccine.
Introduction
As a virus that spreads mainly through sexual contact, it is generally accepted in the medical community that the high-risk human papillomavirus (HPV) is the primary cause to induce cervical cancer. According to the World Health Organization (2019), in 2018, approximately 570,000 women were diagnosed with cervical cancer worldwide, of whom about 311,000 died. Centers for Disease Control and Prevention (2020) reported that in the United States, around 80 million people, or 25 percent, are infected with HPV; around 14 million people (encompassing adolescents) are infected with HPV annually; about 35,000 people each year are influenced by cervical cancer triggered by HPV. In China, an estimated 30,000 people die of cervical cancer every year (Chinese Center for Disease Control and Prevention, 2020). According to Liu et al. (2020), the total prevalence of HPV of 83561 women in Mainland China is 16.18%, and the incidence of HPV is relatively high regarding female health examinees. As such, HPV is a life-threatening virus, especially for women.
The HPV vaccine is a safe and effective way to prevent cervical cancer, and progress in prevention strategies varies among countries (World Health Organization, 2020). The United States and China are typical examples. In June 2006, the U.S. Food and Drug Administration (FDA) approved a Merck HPV vaccine, Gardasil, to be administered to young females. Shortly after the HPV vaccine was licensed, some states, such as Virginia and the District of Columbia, stepped up legislation to promote vaccination in 2006 and 2007, but it was strongly opposed by public opinion (Head et al., 2018). Currently, the vaccine could protect against more HPV strains, with doses reduced to two doses. In addition, U.S. citizens can pay vaccination costs via private insurance and public programs (Kaiser Family Foundation, 2018). As for China, due to the complexity of Mainland China’s approval procedures for overseas vaccines, a total of four-phase clinical trials of vaccine drugs are needed. Also, concerning about the discrepant performance of the vaccine in the Asian population, the release delayed until 2017 in China. On April 28, 2018, the State Food and Drug Administration (CFDA) conditionally approved the listing of the 9-valent vaccine in China; in May 2020, the first 2-valent vaccine produced by China entered the market, and immediately became a hot pick (Xinhuanet, 2020). In short, the United States has extensive public health coverage for HPV vaccine; in contrast, the vaccine in China is still in its developmental stages, including people’s awareness and health care system waited for maturity.
For the public, the Internet has become a new way to obtain health information (Pandey et al., 2003; McMullan, 2006). Especially, the wide array of video platforms have become crucial channels for various stakeholders to generate health-related content (Gabarron et al., 2013; Huang et al., 2016). As one of the most popular video streaming websites in the world, YouTube Help (2020) allows users to “discover, watch and share originally-created videos”. When we input the keyword “HPV vaccine”, it can be seen that there are thousands of videos about the HPV vaccine. In China, people also rely on the Internet for health information. As a highly concentrated video platform for the Chinese young generation, Bilibili (2020) has an average number of monthly active users (MAU) of 128 million in the third quarter of 2019. When we search the same keywords, we can see more than 1000 videos about the HPV vaccine. Among them, the highest amount of views reached 1.307 million.
Past research has shown that people who search for cancer information in a positive way (active and target-oriented information search) such as using YouTube, can influence decision making (Niederdeppe et al., 2007). Since it is difficult for most consumers to accurately assess the professionalism of online health information, the quality of video content is extremely important. At present, most of the research on vaccine promotion in China is still in the observational stage, and there is a lack of exploration on this issue from the perspective of communication (Li & Zheng, 2020), while the United States already has a more mature framework theory to support the analysis (Wilson, 2016). As such, there is still a gap in the study of comparing HPV vaccine videos produced by different UGC (User-generated content) video platforms. This study firstly summarized HPV vaccine-related studies on different media platforms, and secondly analyzed the similarities and differences of related content published on YouTube and Bilibili by applying key factors of the health belief model (HBM). This research aims to provide a reliable reference for video promotion of related stakeholders of HPV vaccine and medical institutions.
Literature Review
In this part, the content and standpoints of report contents about HPV vaccines in traditional and new media platforms are briefly summarized. Five important predictors of the health belief model: perceived susceptibility, perceived severity, perceived benefits, perceived barriers and self-efficacy are interpreted in detail. We also raised six research questions in this section.
HPV vaccine contents on different media channels
In recent years, a lot of scholars take the HPV vaccine as an example to analyze the health-related content spreading on different media platforms (Calloway et al., 2006; Abdelmutti & Hoffman-Goetz, 2009). On the traditional media platforms such as newspapers and television, people can learn much about the HPV vaccine from different perspectives. In a content analysis of the U.S newspaper about the HPV vaccine, Quintero Johnson et al. (2011) found that most U.S newspaper articles showed that sexual contact is the main route of transmission, and how the HPV leads to the cervical cancer and the vaccine’s preventive effect on the cervical cancer. Approximately 50% of the articles discussed whether the legislation about children taking the vaccine is needed and only less than half of all articles provided professional information about the virus and the specific characteristics of the vaccine itself. Compared with the U.S, in Germany and Spain, the newspaper can also have a strong impact on disseminating health information about HPV. However, until 2012, only 6% of German newspapers and none of the Spanish newspapers reported the effectiveness of the vaccine accurately (Bodemer et al., 2012). There is also a different situation appearing in Japan. Morimoto et al. (2015) noted that because of several adverse vaccine events happening in Japan, most reports in the Japanese newspapers mentioned that the government stopped recommending the HPV vaccine to ordinary people. However, negative news about the vaccine seemed not to affect people’s decisions on vaccination because 60% of the people who read the adverse news still chose to vaccinate.
Regarding the TV broadcast reports about the HPV vaccine, Bond (2010) discovered that the information provided on TV is similar to that in the newspaper. 96.3% of the U.S TV broadcast reports claimed HPV is the direct cause for cervical cancer; 88.9% reported the vaccine’s influence on cervical cancer. Moreover, there were also 85.2% of reports mentioning the problems led by the vaccine injection in the aspects of ethics and politics at the same time.
Except for traditional media channels, more and more HPV-related content can also be found on some social media platforms and video websites. On Instagram, the information about the basic effects of the HPV vaccine was mostly positive, but the description from the general population, such as some vaccine consumers was more negative (Kearney et al., 2019). Compared with Instagram, on Twitter, most Tweets were written by the consumers who have taken the vaccine and 51% showed a positive attitude (Keim‐Malpass et al., 2017). Furthermore, on the video platform YouTube, Keelan et al. (2007) analyzed 153 videos and found that 48% of the videos were positive toward vaccination, 32% were negative, and 20% showed an unclear attitude. Ekram et al. (2019) showed in their study that the information about pro-vaccine provided in the videos was more accurate than the anti-vaccine ones. The anti-vaccine videos can offer misleading information more easily. Moreover, in the comment section of YouTube, most audiences talked about the side effect of the vaccine and their worries about whether the vaccine is harmful to their health.
In addition to these new media platforms which are widely popular in developed countries, some SNS platforms with certain regional characteristics such as Sina Weibo (China’s version of Twitter) also play a role in promoting the HPV vaccine. Although a previous research paper claimed that compared with American people, some Chinese people lack the knowledge of the HPV and the HPV vaccine, they also show their interests in the vaccine. Some people taking the HPV vaccine shared where and how they took the vaccine and the post-immunization on Weibo. Also, some people who want to take the vaccine asked for detailed information about the vaccine such as the side effect and the problems about public health insurances or commercial insurances. If they post these questions on Weibo, they can get replies from other people who have been vaccinated. In other words, Weibo has been a vaccine information exchange platform (Zhang et al., 2013). As a highly concentrated video platform for the Chinese young generation Bilibili plays an important role in disseminating HPV vaccine-related information among Chinese youth, nevertheless studies about this issue seem to be very rare, especially studies using English.
In sum, albeit a number of studies about vaccine-related YouTube videos have been conducted, studies of vaccine-related Bilibili videos are still very limited. The aim of this research is to fully understand how the HPV vaccine is portrayed on both YouTube and Bilibili and compare the difference between vaccine-related videos on these two platforms. Thus, we propose the following research questions to explore this issue:
RQ1: What are the types of sources of HPV vaccine information on YouTube and Bilibili?
RQ2: From which regions are the sources used by YouTube and Bilibili?
RQ3: What are the video’s attitudes (negative, positive, or neutral) in the videos on YouTube and Bilibili?
RQ4: Is there an association between the tones and the number of likes on YouTube and Bilibili?
Health Belief Model
The health belief model (HBM) has been widely applied as an analytical framework of content analysis to explore different health contents (Chen et al., 2020), and this study also adopted the same approach. HBM was first proposed by social psychologists Hochbaum, Rosenstock and others at the United States Public Health Service (USPHS) in the 1950s (Health Behavior and Health Education, 2020). The model intended to predict why individuals could not participate in disease prevention and detection programs (Hochbaum et al., 1952; Rosenstock, 1974). HBM includes several key indicators that explain people’s reactions to disease screening, control, prevention, etc. These factors contain perceived susceptibility, perceived severity, perceived benefits, perceived barriers and self-efficacy. The definitions of these five key predictors are shown as below:
Perceived susceptibility is the individual conviction about how likely he/she will get a disease condition. According to the HBM, the more a person feels that the disease is unlikely to affect them, the less likely that he/she will act on the desired behavior (Carpenter, 2010).
Perceived severity deals with the feelings about the seriousness of the consequence led by not acting on preventing the health problem. The combination of susceptibility and severity consists of people’s perception of a health condition (Carpenter, 2010).
Perceived benefits refer to an individual’s belief in the perceived benefits of taking action to reduce the threat of disease. It suggests that people who perceive susceptibility and severity of a health issue are not expected to take action unless they also perceive the potential positive consequences of reducing disease threat (Zhang et al., 2017).
Perceived barrier proposes that potential adverse impacts of taking the desired action may prevent individuals from adopting recommended behavior, such as “it could be good for my health, but it might be troublesome, time or energy-consuming, and it also could be a big expense” (Rosenstock, 1974).
Self-efficacy refers to individual beliefs about how confident they feel that they can take action. The more confident people feel, the more likely they will undertake the desired action in a disease condition (Rosenstock et al., 1988).
Despite there are non-negligible limitations of HSM that have drawn criticism from scholars, such as the lack of utility in regard to predicting women’s preventive health behaviors (Tanner-Smith & Brown, 2010), or the lack of consideration of personal habitual behaviors and external factors that may affect the actual action (Boston University School of Public Health, 2019), it is still a robust model since most of its elements can be measured directly in the content analysis (Chen et al., 2020). In particular, HSM has been utilized to analyze information about HPV vaccines across various kinds of media platforms. Madden et al. (2012) analyzed the top search results related to HPV vaccines from four search engines and suggested that healthcare professionals should add information about perceived severity and perceived susceptibility when producing relevant media contents. Shapiro et al. (2017) conducted cross-sectional research on HPV vaccine concerns of tweets in Australia, Canada and the United Kingdom. They found that Twitter users in the three countries had similar types of concerns, and the most frequent concerns were perceived barriers. In addition, Li and Zheng (2020) analyzed HPV-related coverages on Chinese social media platform Zhihu. They found that most articles described the relation among cervical cancer, perceived severity and the benefits of HPV vaccines.
Based on the five major indicators of health behavior control defined by the HBM, this research aims to examine to what extent do Bilibili and YouTube videos about the HPV vaccine present messages related to these five factors. As a result, the following research questions are proposed:
RQ5: What are the elements of HBM factors included in the HPV vaccine videos on YouTube and Bilibili?
RQ6: Is there an association between the sources and the HBM factors on YouTube and Bilibili?
Research Method
Sample
From April 10 to April 15, 2020, we searched the videos of HPV vaccine on YouTube and Bilibili, and then collected and analyzed the relevant data. On YouTube, we input “HPV, HPV vaccine, HPV vaccination, human papillomavirus, human papillomavirus immunity” as the keywords. On Bilibili, we searched for the same words in Chinese. After getting the search results which are ranked by relevance, we retained the top 30 results of each keyword as our sample. We removed the repetitive videos that were searched by different keywords and the less informative videos. This method is consistent with the process in previous studies (Ache & Wallace, 2008). 250 of the videos were finally included in the final sample (YouTube: 137; Bilibili: 113).
Coding procedure
Our coders acquired and analyzed the features of the video content of the two comparative websites. Our coders kept a detailed record of the original link, regions. As for those videos on YouTube, since YouTube is an international video platform, we recorded the countries or regions where these videos came from. For the videos on Bilibili, since all the videos are in Chinese and mainly from personal bloggers, we recorded the countries and regions where people intend to be vaccinated. We also recorded the number of subscribers to the account where the video was posted, the total number of views and the number of people who like or dislike the video.
Besides, we encoded the sources and semantic tendencies of the video content. We have nine categories of sources, including government agencies: e.g., European Medicines Agency (EMA) or Chinese Center for Disease Control and Prevention, non-profit organizations (e.g., the University of Chicago), for-profit companies (e.g., Lilac Doctor or Merck), individual consumers or independent media (i.e., videos recorded by unofficial ordinary individuals), news organizations (e.g., CNA or CNTV), medical staff (e.g., Top Doctors the UK or Omnicuris), hospitals or medical institution (e.g., The Children’s Hospital of Philadelphia or Cancer Council Victoria), medical personnel and others (i.e., videos do not belong to any of the above types) (Briones et al., 2012; Li et al., 2013). In terms of semantic tendency, we divided it into four tendencies: recommended injection, non-recommended injection, neutral attitude and unstated attitude.
We also classified and coded five dimensions according to the HBM. There were eight constructs for “barriers”: price (i.e., It’s too expensive to accept), age (i.e., below or above the age at which the vaccine is effective), efficiency (i.e., worry about the efficiency and safety of the vaccine), relatives and friends (i.e., The lack of support from relatives and friends), degree of understanding (i.e., This vaccine had never been heard of or little is known about it), resource status (i.e., People don’t know where to get vaccinated / there is a shortage of vaccines), others and no mention. Regarding “susceptibility”, “severity”, “benefits” and “self-efficacy”, depending on whether the video content had these characteristics, we all coded 1 for present and 2 for absent. About perceived susceptibility, it refers to people’s perception of risks towards negative outcomes; as for perceived severity, it means the perception of the serious consequence; in terms of perceived benefits, it refers to the advantages of HPV vaccination; regarding self-efficacy, it means people’s confidence of their capability to succeed.
Coding measurement
Two authors participated in the coding process. They all participated in the whole process of this study and are proficient in Both Chinese and English. In order to improve the reliability of coding as well as to test the codebook, all coders used the codebook to practice, and precoded over 10% of the samples, namely 27 videos to build intercoder reliability. To calculate intercoder reliability Krippendorff’s alpha (Krippendorff, 2018) was used, which ranged from 0.83 to 1.0 and thus indicated satisfactory inter-coder reliability.
According to the distinct features of the HPV vaccine videos on the two platforms, we displayed descriptive statistics, including the type of sources, regions, tones, and whether the HBM factors were present or absent. In terms of the analysis of inferential statistics, One-way ANOVA tests and Chi-square tests were used via IBM SPSS Version 25 to verify the relationship between two categorical variables (the sources and the HBM factors) or one categorical variable and one numerical variable (the tones and the numbers of likes).
Results
RQ1: The type-sources of information in HPV vaccine videos
As shown in Table 1, video sources displayed great differences between the two platforms. For Bilibili coded videos, 62.8% (n=71) came from the consumer-generated content, followed by medical workers (20.3%, n=23), medical centers/hospitals, professional associations (16.8%, n=19). While YouTube sources are more diverse, news sources account for the largest proportion (34.3%, n=47), the second largest type was governmental agencies (23.3%, n=32), the third largest source was medical centers/hospitals, professional associations (15.3%, n=21), followed by advocacy groups (12.4%, n=17), nonprofit/academic organizations (8%, n=11), consumer-generated content (4.3%, n=6), medical workers (2.2%, n=3).
RQ2: The regions of information in HPV vaccine videos
Since most of the videos of Bilibili were released by individual bloggers, the description of their injection locations becomes the core of this question. However, the majority of sources on YouTube were groups or organizations, so our focus puts on the countries where the video came from. For the Bilibili videos, the largest proportion was from Mainland China, followed by Hongkong, South Korea, the United Kingdom, Australia, and the United States. For the YouTube videos, the largest proportion was from the United States, followed by the United Kingdom, India, Singapore, Australia, New Zealand, Kenya, China, Canada, and the Netherlands.
RQ3: Attitude tendencies in HPV vaccine videos
Both the largest proportion of the two platforms were positive evaluations (YouTube: 60% n=82; Bilibili: 78.8%, n=89). Bilibili had no negative evaluation, and contents that not mentioned account for 13.2% (n=15); neutral contents account for 8% (n=9). YouTube had the neutral, negative, and not mentioned account for almost the same ratio (n=19, n=18, n=18, respectively).
RQ4: The association between the tones and the numbers of like
One-way ANOVA analysis was conducted to test the relationship between attitudes and the amounts of like from HPV vaccine videos. The result showed that there were significant differences between the tones and the number of likes on YouTube (F=2.938, p =0.043 < .005). Post hoc comparisons using LSD tests displayed statistically significant differences (p < .005) between these three subgroups: the mean value of the number of likes with a positive attitude (M=42, SD=70.51) and that with a negative attitude (M=185.4, SD=296.15), p=0.008; the mean number of likes with a negative tone and that with a neutral tone (M=27, SD=35.57), p=0.025; the mean number of likes with a negative tone and that with an unstated attitude (M=4.67, SD=6.43), p=0.026. In sum, videos with a negative attitude were more likely to receive “like” compared with the other three tones. While there was no significant difference on Bilibili (F=1.849, p =0.18).
RQ5: HBM factors included in the videos
In terms of perceived susceptibility, 60.5% of videos (n=83) on YouTube mentioned this factor, and 89.3% of videos (n=101) on Bilibili mentioned this element; in terms of perceived severity, the figures were 70% on YouTube (n=96) and 86.7% on Bilibili (n=98).
With regard to self-efficacy, the two platforms presented an opposite condition. Only 25.6% of videos (n=29) on Bilibili mentioned self-efficacy; however, 81.7% of YouTube videos (n=112) contained this HBM element.
In terms of perceived barriers of HBM, 47.7% of videos (n=54) on Bilibili talked about this factor, and 56.2% of videos (n=77) on YouTube mentioned that. The top four categories on Bilibili were: don’t know where to vaccinate /vaccine shortage, age, never heard or know little, others. In China, the authority suggested that the best age for injecting HPV 9-valent vaccine is between 16 and 26 years old (Xinhuanet, 2019), so age becomes a major concern. In addition, we found some new barriers in the coding process, including pregnancy, lactation, and menstruation. Whereas the top four obstacles of YouTube were different from those of Bilibili, namely, never heard or knew little, family’s opposition, safety, don’t know where to vaccinate.
In terms of perceived benefits of HBM, all videos on Bilibili mentioned the advantages of HPV vaccine (n=113), while only 21.8% of contents (n=30) on YouTube mentioned the benefits (as shown in Table 2).
RQ6: Association between the sources and the HBM factors
Chi-square tests were used to test the relationship between the sources and HBM elements. The results showed that for Bilibili’s videos, there was a significant difference between sources and barriers, χ2 (1, N=113) =37.808, p < 0.01 (as showed in Table 3), but other HBM elements were not significant; for YouTube’s videos, sources showed no significant difference of all HBM factors.
Discussion
The first China-made HPV vaccine has been promoted in May 2020, which to some extent alleviates the vaccine shortage in China and reduces the cost paid by Chinese vaccinators (Xinhuanet, 2020). At the same time, it was reported that with the outbreak of COVID-19 worldwide, the vaccination rate has decreased since local communities taking the social distancing measures and parents’ cancellation of health inspection (Hoffman, 2020). Any medical and health behaviors of individuals or organizations are never in the “injections from a vacuum”, but closely related to concrete social policies, diverse cultural dimensions, and collective awareness. Meanwhile, the content of health communication produced in the form of videos also shapes people’s views and even behaviors on vaccination. Therefore, the media content could be a bridge between social medical conditions and people’s medical behaviors (Boepple & Thompson, 2014; Allem & Ferrara, 2016).
In this study, two representative video platforms of UGC were selected (YouTube and Bilibili), aiming to reflect types of sources, regions and attitudes of HPV vaccine that vary from different media channels. Additionally, a theoretical framework of coding scheme HBM was used to explore relevant medical information on the two video platforms. Despite there have been various separate studies on YouTube’s video clips of HPV messages (Briones et al., 2012; Ekram et al., 2019) or Chinese social media platforms’ HPV vaccine contents (Li & Zheng, 2020; Su, 2020), to our knowledge, this is the first research that compared the videos of YouTube and Bilibili on HPV vaccination. In particular, no international studies have focused on the contents of the HPV vaccine on Bilibili. We filled this research gap from a comparative perspective.
In terms of type-sources, we found that most of the vaccination videos in Bilibili were uploaded by individual bloggers, while the sources of YouTube tended to be more diverse, and the proportion of professional medical institutions was larger. This result may be related to Bilibili’s entertainment orientation since it is well-known as subculture and bullet screen and has a myriad of young users who are creative (Dwyer, 2017). On the one hand, new media is able to provide discussion spaces for the general public, which is not necessarily dependent on traditional channels like the government or public health institutions (Tozzi et al., 2010). On the other hand, the absence of professional institutions would result in negative outcomes such as misinformation or fake news (Kata, 2010). Therefore, we suggest that more government medical departments speak up to provide more authoritative, high-quality information for the audience on Bilibili. YouTube is a good example, as professional organizations like Centers for Disease Control and Prevention (CDC) and the MD Anderson Cancer Center have opened official accounts and published HPV vaccination guidelines.
With regard to RQ2, this study innovatively compared the region categories of the two video sites and found some interesting results. The United States is the main source of YouTube videos, followed by the United Kingdom, India, and Singapore. There is no doubt that the United States has the largest market share and complete health care services. People can have access to HPV vaccine resources easily. Apart from listed developed countries, India’s controversy over the HPV vaccine may be one reason for the large proportion of videos. In 2018, the authority came under fire for approving HPV vaccine provided for cervical cancer patients, with health activists questioning both the technical process and the high cost of the vaccine campaign (The Pharma Letter, 2018). In addition to Mainland China being the most prominent source, the other regions topped the list are largely attributed to studying abroad or medical tourists, such as Hongkong, South Korea and the UK. In recent years, tourism plus vaccination has become a hot topic among Chinese media (Bloomberg, 2018). Compared with the limitation of an officially sanctioned age range in Mainland China, overseas vaccine prices are relatively low, and there is no age limit. Additionally, Chinese students who study abroad may also utilize local medical conditions for injection. For example, in some of the videos we have coded several vloggers who are international students talking about their experiences getting vaccinated in private clinics in the UK. But with the permission of China’s domestic vaccines and the pandemic of COVID-19, this overseas vaccination trend may be decreasing.
As for video attitudes, there was no negative attitude towards the encoded videos of Bilibili, and positive attitudes accounting for 78.8%. These findings are consistent with earlier research on the HPV vaccine messages on Chinese social media and Chinese newspapers (Li et al., 2018; Li & Zheng, 2020). It also indicates that both official media and we-media in China generally hold positive attitudes on HPV vaccine. Such a unified tone would boost the public acceptance and enhance the delivery of HPV vaccination promotion campaigns. Regarding YouTube, aligned with previous studies on the content analysis of HPV videos on YouTube (Ache & Wallace, 2008; Tuells et al., 2015), most encoded videos agreed with vaccination. But 18 of the analytic videos hold negative attitudes, which is a striking discrepancy compared to Bilibili.
As for RQ4, there is a significant relationship between attitudes and the number of likes on YouTube, more precisely, people prefer to “like” vaccine videos with negative tones. This is in line with previous findings which analyzed YouTube videos about HPV information (Briones et al., 2012; Covolo et al., 2017). This reaction to the information may reflect dissatisfaction with their past experiences about HPV vaccination. These preconceptions may affect their actual vaccination behavior, such as choosing to refuse vaccination.
In terms of RQ5, the majority of encoded videos on Bilibili mentioned several HBM elements including perceived benefits, perceived susceptibility and perceived severity, which is congruent with past findings of articles of HPV vaccination on Chinese social media Zhihu (Li & Zheng, 2020). However, self-efficacy was often ignored in many videos. This may be associated with the difficulty of obtaining vaccines in Mainland China. The high cost of the vaccines and limited supply both shake people’s confidence in vaccination. But findings on YouTube are strikingly different. 81.7% of videos included self-efficacy and this proportion is much larger than previous research on YouTube content analysis (Briones et al., 2012). Only 21.8% of contents mentioned perceived benefits. Before 2015, there were several sensational news about the death and disability resulted from vaccines included in the encoded videos on YouTube, But on Bilibili, this type of video is completely absent.
In addition, another intriguing finding is the difference of perceived barriers between YouTube and Bilibili. Age restriction is one of the obstacles mentioned frequently in the video of Bilibili. Some novel barriers have been found, such as pregnancy, lactation and menstruation. Nevertheless, the types of barriers obstacles on YouTube also included the family’s opposition and safety, while those were not found on Bilibili’s contents.
Lastly, as for RQ6, there was a significant association between type sources on Bilibili and perceived barriers. The data from crosstabulation indicated that individual bloggers tended not to mention barriers when talking about HPV vaccination. This finding is in contrast to previous research on HPV vaccine posts on Weibo (the largest social media site in China) (Chen et al., 2020), which found that individual users rather than organizational users were more likely to post contents about barriers. The possible explanation is that on Weibo, as an online public sphere in China (Jiang, 2016), netizens were more willing to express their opinions and the barriers can be objectively discussed. However, Bilibili has an obviously entertaining and hedonistic tendency (Yin & Fung, 2017). Compared with susceptibility, severity, benefits, it seems less eye-catching for those who want to click on the video to find ease. In addition, the types of barriers mentioned by the medical centers include prices and age, which may further affect the intention of the audience to inject HPV vaccine. But it is important to note that such media effects are not clear to date, since the majority of analytic videos on Bilibili is created by individual bloggers. Future research needs to further examine the influence of HBM factors of relevant content on user intention and their health behaviors.
In conclusion, this study provides a comparative framework to clarify the content preferences of the two video platforms, which extends the application of HBM theory towards media contents of the HPV vaccine. As for the practical contributions, this study can shed some light on how to provide medical and health knowledge through UGC platforms for video creators, medical professionals, and government agencies.
Limitations and future studies
Here are some research limitations. Firstly, the number of research samples is relatively small on the two platforms, and expanding the sample size may lead to richer and more reliable results; secondly, for inferential statistics, there are still some variables that can be verified, such as the relationship between the number of dislikes and the attitude. Thirdly, the content features of the two platforms can also develop more coding classifications. Finally, for the benefits categories of HBM, we only observe the benefits of physiological vaccination in these videos. In the future, other studies can combine qualitative interviews with exploring more possibilities toward videos of HPV vaccine disseminated through UGC platforms.