Facebook HPV vaccine campaign: insights from Brazil

ABSTRACT

The vaccination coverage rate has fallen in recent years in Brazil. Although Brazil is a country influenced by social networking and analysis of social media generated content has emerged as an important area of research, this is the first study to investigate how the HPV vaccine is portrayed on the Facebook page of the Ministry of Health and to interpret the user-generated content. Our work evaluated the posts published by the Ministry of Health as well as posted comments, shares and reactions of the population related to the HPV vaccination campaign on the specific Facebook page between January and July 2018.The data were analyzed in three stages: (i) overview of published posts; (ii) analysis of posted comments; and (iii) sub-group analysis of posted comments. This study makes important contributions from the point of view of using social media as a public health communication tool.

KEYWORDS:

• Behavior
• HPV vaccine
• social media
• Facebook
• health communication
• Brazilian Ministry of Health

Introduction

Human papillomaviruses (HPVs) are a group of more than 200 related viruses, and most can easily be spread through direct sexual contact. HPV is the most common viral infection of the reproductive tract¹ and it is associated with a variety of clinical conditions that range from innocuous lesions to cancer.² Cervical cancer is the most serious HPV-related disease. Nearly all cases of cervical cancer can be attributed to HPV infection.^3,4 Cervical cancer is the second most common cancer in less developed regions (445,000 new cancer cases in 2012), accounting for 7.5% of all female cancer deaths worldwide.⁵ In Brazil, the National Cancer Institute estimates 16,370 new cases of cervical cancer for each year of the 2018–2019 biennium, with an estimated risk of 15.43 cases per 100,000 women.⁶

Brazil has one of the highest prevalence rates of HPV co-infection (75.4%) in high-grade lesions,⁷ which provides a tremendous opportunity for cancer prevention through HPV vaccination. The HPV vaccine was incorporated in 80 countries by the end of 2017⁸ and was introduced gradually in Brazil from 2014. Since 2017, the vaccination program has been expanded to include girls aged 9 to 14, boys aged 11 to 14 and priority groups for the female and male population aged 9 to 26.⁹

Vaccination coverage rate in Brazil

The National Immunization Program (NIP) is an international benchmark in Brazil.⁹ However, the high coverage rate, which has always been its main characteristic, has fallen in recent years, leading to an outbreak of vaccine-preventable illnesses in the country. Despite the free delivery and administration of vaccines in Brazil, HPV vaccination coverage with one or more doses decreased by 23% from 2014 to 2015¹⁰ and vaccination rates were below the target set by the Ministry of Health in the first half of 2017: 46.2% for girls and 20.2% for boys.¹¹ At least nine factors have contributed to reducing the vaccination coverage rate in Brazil, and possibly all of them have a joint effect.¹² However, only one study has attempted to quantitatively measure the importance of each factor by a random-digit-dial survey of parents.¹³

In addition to the operational difficulty of maintaining vaccinations in schools and the fear of vaccine adverse events,¹³ there are many other barriers to Brazilian HPV vaccination, such as the inherent difficulty in the vaccination of adolescents, low perceived susceptibility of HPV infection and concerns that the HPV vaccine promotes early adolescent sexual behavior.^12,14 It is also important to note that the influence of fake news circulating on social media¹⁵ and the incipient actions of anti-vaccine groups¹⁶ cannot be discarded as causes of the fall in vaccinations.^17-19 People that were exposed to stories about harm resulting from the HPV vaccine often delay or refuse vaccination.^15-17 In Brazil, the local media has helped to maintain a continuous negative effect on the population, giving undue priority to bad and negative news and revisiting past problems that have already been resolved and addressing them as if they continued to exist and had a causal relationship with the HPV vaccine.²⁰

Online information channel of the HPV vaccine

Online information is an important channel of vaccine-critical content. Previous studies specifically examining online content related to HPV vaccine have found both positive and negative information on general websites,^17,21 as well as on social media, especially on blogs,^18,22 YouTube,^23-26 Facebook,^27,28 and Twitter.^29,30 Although several studies have provided an analysis of online information on HPV vaccination,³¹ there has been limited research on how public health officials use online platforms to communicate their message through existing networks and engage in ongoing dialogue with users.²⁷ This issue is particularly important in Brazil, where messages and recommendations created and disseminated by the Ministry of Health should instill trust and fully clarify that the adverse events involving teenage patients are not associated with the vaccine. Although analysis of social media generated content has emerged as an important field of study, this is the first such study conducted in Brazil, produced at a critical time of low coverage rates of the HPV vaccine.

While there are many social media platforms in the Brazilian Ministry of Health, the focus of this paper is its Facebook page on vaccination, especially the HPV vaccine. Facebook is the most popular social media site among Brazilian people, used by more than 50% of the population.³² Brazil is strongly influenced by social networking,³³ which makes Facebook an appropriate social media platform to reach and engage Brazilian people in discussions on the HPV vaccine. Statistics in December 2018 showed that the Ministry of Health’s Facebook page has the fifth largest number of Facebook users of official government pages, with 2,126,357 followers.³⁴

2018 HPV vaccine campaign

The HPV vaccine is part of the routine schedule available through the NIP. As with other vaccines, the Brazilian Ministry of Health runs an HPV vaccination advertising campaign to mobilize the population and remind people of the need for HPV vaccination. The slogan of the 2018 HPV vaccine campaign was “Do not miss the new season of HPV vaccination”. It involved several advertising pieces and had a limited disclosure period. The campaign film, included on the Facebook page, mixed real images and animation and showed two adolescents, a boy and a girl, fleeing from a virus in a scenario inspired by famous series that can be identified by both the young audience and their parents. The escape of the teenagers ends when they enter a public health office and are vaccinated. The new vaccination campaign also provided information on the expansion of the age group eligible for the booster vaccine against meningitis: adolescents aged 11 to 14.

Facebook is an important social, cultural, and behavioral thermometer. To address the research gaps that involved vaccination campaigns and social media, our study seeks to investigate how the HPV vaccine is portrayed on the Facebook page of the Brazilian Ministry of Health and interpret its user-generated content. This study makes four contributions related to: (i) the importance of examining online discussion groups and correcting unintended errors in communicating vaccination campaigns; ii) monitoring the engagement rate, sentiment polarity, and emotion types of specific audiences to which the HPV vaccine message was directed, iii) determining which types of contents represent the most effective message for positively engage Brazilians in adolescent HPV vaccination campaigns, and iv) evaluating whether the engagement strategies used by the Ministry of Health are in line with what the literature indicates to be effective strategies. Overall this manuscript is of interest to the international community, as it brings important discussions to light regarding public health policies and science communication through the emergent technology of a widely used social media.

Methods

In this study, sociological investigation for behavioral research was used, that is, we applied a non-experimental research method directed to the analysis of social variables to look for behavioral patterns as people engage in interactions³⁵ in a digital setting. We collected empirical data from the Facebook page of the Brazilian Ministry of Health between January and July 2018 by use of Netvizz App*¹. Once the data extraction process was completed, a separate soft copy of the extracted data was exported to a spreadsheet application (Microsoft Excel) for further analysis. To understand fully how the HPV vaccine is portrayed by the Ministry of Health on its Facebook page and interpret user-generated content, this study was divided into three parts: (1) overview of published posts; (2) analysis of posted comments; and (3) subgroup analysis of posts and attached comments. In Part 1, based on factors influencing user activity and user engagement,³⁶ the posts were classified according to type of content format, length, time of day, day of the week, frequency of posting, target audience, and categories of main subjects. Four categories were created based on a manual assessment of the main topics posted by the Ministry of Health: (i) vaccine dosage; (ii) vaccine-guaranteed protection: (iii) vaccination target audience; and (iv) just a call for the new HPV vaccination campaign. The metrics for each post according to number of reactions, comments, and shares were also analyzed.

Part 2 of this study analyzed the posted comments. The comments generated by each post were analyzed regarding the sentiment polarity and the main type of emotion. They were also classified based on manual assessment of the main topics of content. The primary objective of sentiment polarity is to classify a user’s attitude as positive, negative, or neutral.³⁷ In addition to sentiment polarity, it is possible to identify and classify the type of emotion expressed in the positive and negative comments into six categories: joy, disgust, fear, anger, surprise and sadness.^38,39 These two analyses can be done using text analysis techniques through computational language or manually. Since no software works properly to analyze both text and feelings in the Portuguese language, the posted comments were analyzed manually and independently labeled by three annotators. The annotators were instructed to select the appropriate main topic, the emotions for each posted comment based on the presence of words or phrases with emotional content, as well as the overall feeling invoked by the headline.³⁸

Keyword-based emotion detection is an intuitive and straightforward text analysis technique. The idea is to detect patterns similar to emotion keywords and match them. In other words, the first task is to find the word that which expresses the emotion in a sentence. These words are then matched against a standard list of words representing emotions according to a validated and translated keyword-dictionary.⁴⁰ In the subsequent step, sentiment polarity identifies subjective information from the text and from the type of emotion to classify a person’s polarity of attitude to published posts. For example, regarding the safety of the HPV vaccine, one person may express anger whereas another person may express disgust, though anger and disgust are both negative sentiments. If someone writes ‘I am happy with new government vaccination campaign’, then it is understood that the person’s emotion is ‘happy’ and the sentiment behind the emotion is ‘positive’.³⁸ It is important to note that surprise was depicted as a pre-affective state, or as an emotion that can be both positive and negative, depending on the goal conduciveness of the surprising event.⁴¹ Finally, as in part 1 of this paper, after identifying the content of every comment, rating categories were created based on manual assessment of the main topics of each attached comment.

Part 3 of this study evaluated specific subgroups in two distinct analyzes: (i) correlations between the main topic of each post (4 categories, Part 1) and sentiment polarity/type of emotion of their attached comments and (ii) correlations between the main topic of each comment (16 categories, Part 2) and sentiment polarity/type of emotion of their attached comments. To find patterns of response, quantitative results were generated from the classifications, which were analyzed by Microsoft Excel Software. This analysis allowed us to make assumptions about how a person feels about a specific information based on some predefined emotion models, according to psychological emotion theories.⁴⁰ Experts in the social media discipline validated the results so that there was a consensus on the different classifications of each annotator (details on methodology architecture of social media analysis can be found in the supplementary material).

Results

Part 1: overview of published posts

This study examines naturally occurring data between January and July 2018 on the Facebook page of the Brazilian Ministry of Health on HPV vaccination. Although the HPV vaccine is available year-round at vaccination posts, the Ministry of Health’s vaccination campaign has only two annual waves of outreach. Our work aimed to evaluate the first wave of 2018. Generally, the first wave is more representative in terms of engagement with the Brazilian population and vaccination coverage rate. The database consisted of 28 HPV vaccination posts published between February 2 and May 10 and 4,302 user comments, generating an average of 154 comments per post. There were 18,634 shares (average = 666) and 55,756 reactions (average = 1,991), of which 93.12% were likes. All 28 HPV vaccination posts contained 1,362 words (in the Portuguese language), with an average of 48.64 words per post. An analysis on the overall population of posts shows that there is no clear relationship between the length of the post and user engagement. The communication style used was friendly, politically correct, colloquial and sympathetic, but sometimes in an authoritative style, as is typical of a public health organization.

The posts were classified according to the type of content format as image, video or text. Of the 28 data published by the Brazilian Ministry of Health on HPV vaccination, only post 10 was not a picture (23 posts) or video (4 posts). The advertising campaign was developed to redirect users from the Facebook advertisement to the government website, and all the published data had a call to action for users to find out more about the HPV vaccine at www.saude.gov.br/vacinahpv. Most of the published data (24 posts) was only for informational purposes, and only four posts called for user engagement. In other words, they suggested sharing the specific post or published content. Information regarding the expansion of the age group eligible for the booster vaccine against meningitis was included in 15 posts.

It is important to note that some posts targeted a specific audience: seven posts were directed at teenagers, seven at parents, three at health professionals, and one at teachers. For example, the post “Do not let your patients miss the new season of vaccination. HPV – children aged 9 to 14. Meningitis C – adolescents aged 11 to 14. More information at saude.gov.br/vacinahpv. # VaccinarÉProteger” targeted health professionals. Ten posts were aimed at the general public and did not mention a specific target audience.

The highest level of user activity and user engagement were found in posts published after 6:00pm (82% of the comments) and from Tuesday to Thursday (84% of the comments) with no more than two posts per day. During the publicity campaign in question, six posts were resubmitted by the Brazilian Ministry of Health on its Facebook page within 1 to 10 days (posts 2 and 6; 4 and 8; 5 and 7; 13 and 23; 14 and 22; 16 and 18). Only one of them increased user activity and user engagement in a secondhand information publication (posts 2 and 6). This post was published on March 13 and 15: “Have you seen the new film from the Ministry of Health? Watch and see if you can identify references to your favorite series. Learn more at saude.gov.br/vacinahpv #VacinarÉProteger”.

The post published on March 15 (post 6) had the highest user activity and user engagement, with 1,478 comments, 42,323 reactions, and 4,851 shares. In an innovative way, this post contained only the video of the 2018 HPV vaccine campaign, without any information about the vaccine in its subtitle. More videos were published later, but they were not new or kept users connected to the HPV vaccination campaign. Posts 5 and 6, published on March 14 and 15 respectively, accounted for over 50% of the comments and shares. Post 5, in particular, targeted parents and had 712 comments and 4,453 shares. Figure 1 provides more detailed information on Posts 5 and 6. (Information on other Ministry of Health posts can be found in the Supplementary Material).

Figure 1. Published post with the highest user activity and user engagement in the 2018 HPV vaccination campaign.

Finally, the analysis of the overall population of post content highlighted one or two subjects, which allowed us to manually classify them into 4 subgroups according to the main topic of the message: (i) vaccine dosage (4 posts), (ii) vaccine-guaranteed protection (4 posts), (iii) vaccination target-audience (17 posts), or (iv) just a call for the new HPV vaccination campaign (3 posts). Table 1 shows the general descriptive statistic analysis of 28 published posts on the Facebook page of the Brazilian Ministry of Health. (Detailed information on each post, their attached comments, reactions and shares can be found in the supplementary material).

Table 1. General descriptive statistical analysis of 28 posts published on the Facebook page of the Brazilian Ministry of Health.

	Posts
	Comments	Reactions	Shares	Engagement Rate
TOTAL	4,302	55,756	18,634	78,692
MIN	3	76	19	117
MAX	1,478	42,323	4,851	48,652
AVG	154	1,991	666	5,427
AMPLITUDE	1,475	42,247	4,832	48,535
STD DEV	322	7,802	1,259	9,030
MEDIAN	26	191	147	391

Part 2: analysis of posted comments

The Netvizz app enabled the downloading of data from the Facebook platform for the extraction of comments and metrics from the 28 HPV vaccination posts. In this study, we evaluated the type of emotions and sentiment polarity of 4,302 attached comments expressing opinions on the HPV vaccination campaign. As shown in Table 2, positive and negative comments were identified and classified according to the specific emotions expressed by users: joy (74.2%), sadness (12.9%), anger (3.5%), disgust (3.2%), fear (2.8%) and surprise (1.8%). In the subsequent step, sentiment polarity identified subjective information from the text and type of emotion to classify a person’s polarity of attitude to published posts. Neutral comments, for the most part, corresponded to users’ comments about vaccines or sharing news, as well as personal objective texts without a clear opinion. According to Table 3, neutral comments were found to represent the majority of all comments, at 48.0%. Positive comments accounted for 39.4%, while negative comments represented 11.8%.

Table 2. Emotion type of positive and negative comments attached to posts published by the Brazilian Ministry of Health on its Facebook Page.

	Posts
	Joy	Disgust	Fear	Anger	Surprise	Sadness	Empty	Total
TOTAL	1659(74.2%)	72(3.2%)	62(2.8%)	79(3.5%)	40(1.8%)	289(12.9%)	35(1.6%)	2236
MIN	0	0	0	0	0	0	0	1
MAX	529	45	44	43	24	148	22	837
AVG	61	7	5	8	5	14	6	80
AMPLITUDE	529	45	44	43	24	148	22	836
STD DEV	114	13	11	13	8	33	8	167
MEDIAN	17	2	1	3	2	2	3	19

Empty: followers comment about miscellaneous subjects not related to the Ministry of Health, to the vaccination campaign or to the HPV vaccine.

Table 3. Sentiment polarity of all comments attached to posts published by the Brazilian Ministry of Health on its Facebook Page.

	Posts
	Negative	Neutral	Positive	Empty	Total
TOTAL	507(11.8%)	2,066(48.0%)	1,694(39.4%)	35 (0.8%)	4,302
MIN	0	0	1	0	3
MAX	283	638	535	22	1478
AVG	18	74	61	1	154
AMPLITUDE	283	638	534	21	1475
STD DEV	54	162	116	4	322
MEDIAN	3	10	16	0	26

Empty: followers comment about miscellaneous subjects not related to the Ministry of Health, to the vaccination campaign or to the HPV vaccine.

It is important to highlight that the publications that targeted teenagers generated a higher rate of positive comments (59.1%), unlike the posts addressed to the general public, health professionals and parents, in which neutral comments prevailed (45.7%, 48.8% and 57.4%, respectively). The post with the most frequent negative comment rate (64.3%) was an image with the following message (post 28): “Vaccination against HPV is the best way to protect adolescents against diseases such as cervical cancer in addition to cancers of the vulva, vagina, penis, anus, mouth, and oropharynx, and also genital warts. Visit saude.gov.br/vivamaissus to learn more and watch the new episode of the web series #VivaMaisSUS”. Some comments on Post 28 expressed joy (35.7%), meaning that some people were positively impacted by it. All the other attached comments had negative sentiments: sadness (14.3%), anger (28.6%), disgust (7.1%), and fear (14.3%).

On the other hand, a total of 14 posts had more than 50% positive attached comments. The post with the most frequent positive comment rate was an image that targeted health professionals, with the following message (post 19): “The 2^nd season of vaccination starts in 6 months. Make sure your patients receive the first dose of the HPV vaccine. More information at saude.gov.br/vacinahpv. #VaccinarÉProteger”. Figure 2 shows more detailed information on Posts 19 and 28. (Detailed information on each post, their attached comments and the type of emotions and sentiment polarity of 4,302 comments can be found in the Supplementary Material.)

Figure 2. Published post with highest rates of positive comments and negative comments in the 2018 HPV vaccination campaign.

We also classified all 4,302 attached comments according to their main subject into 16 subgroups. The most frequent subject involved mentions and personal messages (45.4% of the comments), in which users marked other people in their comments or sent direct messages related to published post to other users. Other comments were related to doubts and information regarding the vaccine (12.5%), content of advertisements in the HPV vaccination campaign (10.3%), protection-guaranteed by HPV vaccine (9.7%), message sharing (7.6%), affirmation that the user had already received the HPV vaccine (4.6%), doubts of the vaccine target audience (2.6%), vaccine availability at public health offices (1.3%), vaccine adverse events (1.2%), other vaccine preventable diseases (1.1%), investment and cost of vaccination campaign (0.8%), vaccination schedule (0.8%), vaccination card (0.5%), political issues (0.4%) and questions about the recommendations of health professionals (0.4%). Overall, it can be stated that the 2018 HPV vaccination campaign generated significant user activity and user engagement. On the other hand, some doubts remained concerning the HPV vaccine and target audience among the Brazilian population, despite all the information conveyed during the Health Ministry’s campaign.

Part 3: subgroup analysis of posts and attached comments

We evaluated the comments attached to the four subgroups of posts published by the Ministry of Health regarding the specific type of emotion (see Table 4, positive and negative comments) and sentiment polarity (see Table 5, all comments). Base on Table 5, positive comments were mainly related to posts classified as vaccine dosage (72.0%). On the other hand, posts classified as HPV vaccine guaranteed-protection generated a large number of negative comments (40.4%) among Facebook users, highlighting the following types of emotions: disgust (8.3%), fear (11.1%), anger (16.7%), and sadness (16.7%). According to Table 4, the most frequent type of emotion found in the comments attached was joy (74.2%) in all subgroups analyzed.

Table 4. Emotion type analysis of positive and negative comments attached to the four subgroups of posts published by the Ministry of Health on its Facebook page.

Subgroups	Joy		Disgust		Fear		Anger		Surprise		Sadness		Empty		Total
Vaccine dosage	54	94.7%	0	0.0%	0	0.0%	2	3.5%	0	0.0%	1	1.8%	0	0.0%	57
Vaccine-guaranteed protection	16	44.4%	3	8.3%	4	11.1%	6	16.7%	1	2.8%	6	16.7%	0	0.0%	36
Vaccine target audience	1,059	88.2%	24	1.9%	14	1.1%	28	2.2%	32	2.5%	134	10.4%	0	0.0%	1,291
Just a call for the new HPV vaccination campaign	530	64.9%	45	5.5%	44	5.4%%	43	5.3%	7	0.9%	148	18.1%	0	0.0%	817
EMPTY	0	0.0%	0	0.0%	0	0.0%	0	0.0%	0	0.0%	0	0.0%	35	100.0%	35
TOTAL	1,659	74.2%	72	3.2%	62	2.8%	79	3.5%	40	1.8%	289	12.9%	35	1.6%	2,236

Vaccine dosage: published post shows that two doses of HPV vaccine need to be taken, the second dose being 6 months after the first. Vaccine-guaranteed protection: published post shows that HPV vaccine helps protect against various types of tumor as well as genital warts. Vaccination target-audience: published post shows the target audience for the HPV vaccine. Just a call for the new HPV vaccination campaign: published post shows a call for people to access the advertising material of HPV vaccination campaign.

Table 5. Sentiment polarity of all comments attached to the four subgroups of posts published by the Ministry of Health on its Facebook page.

Subgroups	Negative		Neutral		Positive		Empty		Total
Vaccine dosage	3	4.0%	18	24.0%	54	72.0%	0	0.0%	75
Vaccine-guaranteed protection	19	40.4%	11	23.4%	17	36.2%	0	0.0%	47
Vaccine target audience	203	7.6%	1,384	51.7%	1,088	40.7%	0	0.0%	2,675
Just a call for the new HPV vaccination campaign	282	19.2%	653	44.4%	535	36.4%	0	0.0%	1470
EMPTY	0	0.0%	0	0	0	0.0%	35	100%	35
TOTAL	507	11.8%	2,066	48.0%	1694	39.4%	35	0.8%	4,302

Vaccine dosage: published post shows that two doses of HPV vaccine need to be taken, the second dose being 6 months after the first. Vaccine-guaranteed protection: published post shows that HPV vaccine helps protect against various types of tumor as well as genital warts. Vaccination target-audience: published post shows the target audience for the HPV vaccine. Just a call for the new HPV vaccination campaign: published post shows a call for people to access the advertising material of HPV vaccination campaign.

We also evaluated the comments regarding the specific type of emotion (see Table 6, positive and negative comments) and sentiment polarity (see Table 7, all comments) according to the sixteen subgroups classified based on the main subject of the attached comments. According to Table 6, message sharing (99.7%) and the affirmation that the user had already received the HPV vaccine (97.9%) involved a emotion of joy, while negative emotions involved concern over adverse vaccine events (fear, 40.0%; sadness 24.0%), criticism of investments and the cost of the vaccination campaign (disgust, 17.6%; sadness 50.0%), availability of the vaccine (anger, 17.9%; sadness, 35.7%), and political issues (sadness, 41.7%; anger, 25%).

Table 6. Emotion type analysis of positive and negative comments attached to the sixteen subgroups classified based on the main subject of the attached comments.

Main Topics	Joy		Disgust		Fear		Anger		Surprise		Sadness		Empty		Total
Mentions and personal messages	54	59.3%	9	9.9%	2	2.2%	11	12.1%	0	0.0%	15	16.5%	0	0.0%	91	4.1%
Vaccine doubt and information	323	64.9%	23	4.6%	12	2.4%	23	4.6%	15	3.0%	102	20.5%	0	0.0%	498	22.3%
Advertisement content	306	83.6%	5	1.4%	4	1.1%	10	2.7%	6	1.6%	35	9.6%	0	0.0%	366	16.4%
Protection guaranteed by vaccine	283	74.1%	12	3.1%	17	4.5%	9	2.4%	3	0.8%	58	15.2%	0	0.0%	382	17.1%
Message sharing	322	99.7%	0	0.0%	0	0.0%	0	0.0%	0	0.0%	1	0.3%	0	0.0%	323	14.4%
Affirmation that user had already received the HPV vaccine	186	97.9%	0	0.0%	1	0.5%	0	0.0%	3	1.6%	0	0.0%	0	0.0%	190	8.5%
Vaccine target audience doubt	84	86.6%	2	2.1%		0.0%	1	1.0%	4	4.1%	6	6.2%	0	0.0%	97	4.3%
Vaccine availability	21	37.5%	4	7.1%	0	0.0%	10	17.9%	1	1.8%	20	35.7%	0	0.0%	56	2.5%
Vaccine adverse event	8	16.0%	3	6.0%	20	40.0%	7	14.0%		0.0%	12	24.0%	0	0.0%	50	2.2%
Other diseases	22	52.4%	2	4.8%	4	9.5%	2	4.8%	2	4.8%	10	23.8%	0	0.0%	42	1.9%
Investment and cost of vaccination campaign	9	26.5%	6	17.6%	0	0.0%	1	2.9%	1	2.9%	17	50.0%	0	0.0%	34	1.5%
Vaccination schedule	23	76.7%	1	3.3%	0	0.0%	0	0.0%	4	13.3%	2	6.7%	0	0.0%	30	1.3%
Vaccination card	14	73.7%	1	5.3%	1	5.3%	0	0.0%	1	5.3%	2	10.5%	0	0.0%	19	0.8%
Political issues	1	8.3%	2	16.7%	1	8.3%	3	25.0%	0	0.0%	5	41.7%	0	0.0%	12	0.5%
Health professionals	3	27.3%	2	18.2%	0	0.0%	2	18.2%	0	0.0%	4	36.4%	0	0.0%	11	0.5%
(Empty)	0	0.0%	0	0.0%	0	0.0%	0	0.0%	0	0.0%	0	0.0%	35	100,0%	35	1.6%
TOTAL	1,659	74.2%	72	3.2%	62	2.8%	79	3.5%	40	1.8%	289	12.9%	35	1.6%	2,236	100.0%

Mentions and personal messages: followers tag other Facebook people to whom they send personal messages related to the published post; Vaccine doubt and information: followers ask about various subjects related to HPV vaccine, especially about vaccine application sites. Advertisement content: followers comment about the Ministry of Health’s new HPV vaccination campaign and compare it with previous editions; Protection guaranteed by vaccine: followers comment about the efficacy of the HPV vaccine in preventing cancer and genital wart. Message sharing: followers share the Health Ministry post on their personal profile; Affirmation that user had already received the HPV vaccine: followers comment about when they received the HPV vaccine; Vaccine target audience doubt: followers comment about target audience to HPV vaccine; Vaccine availability: followers comment about HPV vaccine availability in the application sites; Vaccine adverse event: followers comment about the HPV vaccine safety; Other diseases: followers comment about other vaccine preventable diseases; Investment and cost of vaccination campaign: followers comment about the high investment in the HPV vaccination campaign; Vaccination schedule: followers comment about changing the dose schedule from three to two doses of the HPV vaccine; Vaccination card: followers comment about not having the vaccination card available; Political issues: followers comment about political issues associated with incorporation of the HPV vaccine into the public health system; Health professionals: followers comment about the importance of health professionals’ recommendation on the vaccination of children and adolescents. Empty: followers comment about miscellaneous subjects not related to the Ministry of Health, to the vaccination campaign or to the HPV vaccine.

Table 7. Sentiment polarity of all comments attached to the sixteen subgroups classified based on the main subject of the attached comments.

Main Topics	Negative		Neutral		Positive		Empty		Total
Mentions and personal messages	37	1.9%	1,860	95.3%	54	2.8%	0	0.0%	1951	45.4%
Vaccine doubt and information	159	29.7%	38	7.1%	339	63.2%	0	0.0%	536	12.5%
Advertisement content	55	12.4%	79	17.8%	311	69.9%	0	0.0%	445	10.3%
Protection guaranteed by vaccine	95	22.8%	35	8.4%	287	68.8%	0	0.0%	417	9.7%
Message sharing	1	0.3	3	0.9%	322	98.8%	0	0.0%	326	7.6%
Affirmation that user had already received the HPV vaccine	3	1.5%	10	5.0%	187	93.5%	0	0.0%	200	4.6%
Vaccine target audience doubt	12	10.7%	15	13.4%	85	75.9%	0	0.0%	112	2.6%
Vaccine availability	34	59.6%	1	1.8%	22	38.6%	0	0.0%	57	1.3%
Vaccine adverse event	42	84.0%	0	0.0%	8	16.0%	0	0.0%	50	1.2%
Other diseases	18	39.1%	4	8.7%	24	52.2%	0	0.0%	46	1.1%
Investment and cost of vaccination campaign	25	69.4%	2	5.6%	9	25.0%	0	0.0%	36	0.8%
Vaccination schedule	3	9.1%	3	9.1%	27	81.8%	0	0.0%	33	0.8%
Vaccination card	4	19.0%	2	9.5%	15	71.4%	0	0.0%	21	0.5%
Political issues	11	57.9%	7	36.8%	1	5.3%	0	0.0%	19	0.4%
Health professionals	8	44.4%	7	38.9%	3	16.7%	0	0.0%	18	0.4%
(Empty)	0	0.0	0	0.0	0	0.0%	35	100.0%	35	0.8%
TOTAL	507	11.8%	2,066	48.0%	1,694	39.4%	35	0.8%	4,302	100.0%

Mentions and personal messages: followers tag other Facebook people to whom they send personal messages related to the published post; Vaccine doubt and information: followers ask about various subjects related to HPV vaccine, especially about vaccine application sites. Advertisement content: followers comment about the Ministry of Health’s new HPV vaccination campaign and compare it with previous editions; Protection guaranteed by vaccine: followers comment about the efficacy of the HPV vaccine in preventing cancer and genital wart. Message sharing: followers share the Health Ministry post on their personal profile; Affirmation that user had already received the HPV vaccine: followers comment about when they received the HPV vaccine; Vaccine target audience doubt: followers comment about target audience to HPV vaccine; Vaccine availability: followers comment about HPV vaccine availability in the application sites; Vaccine adverse event: followers comment about the HPV vaccine safety; Other diseases: followers comment about other vaccine preventable diseases; Investment and cost of vaccination campaign: followers comment about the high investment in the HPV vaccination campaign; Vaccination schedule: followers comment about changing the dose schedule from three to two doses of the HPV vaccine; Vaccination card: followers comment about not having the vaccination card available; Political issues: followers comment about political issues associated with incorporation of the HPV vaccine into the public health system; Health professionals: followers comment about the importance of health professionals’ recommendation on the vaccination of children and adolescents. Empty: followers comment about miscellaneous subjects not related to the Ministry of Health, to the vaccination campaign or to the HPV vaccine.

Based on Table 7, the positive comments posted were mainly related to the following subgroups: message sharing (98.8%), the affirmation that the user had already received the HPV vaccine (93.5%), and the HPV vaccine schedule (81.8%). On the other hand, negative posted comments involved concern over adverse vaccine events (84.0%), criticism of investments and the cost of the vaccination campaign (64.9%), vaccine availability (59.6%), and political issues (57.9%).

Discussion

This study investigated how the HPV vaccine is portrayed on the Facebook page of the Brazilian Ministry of Health and analyzed user-generated content. For this purpose, our work evaluated the posts published by the Brazilian Ministry of Health and posted comments, shares and reactions of the population related to the HPV vaccination campaign on the specific Facebook page. This study makes important contributions from the point of view of using social media as a public health communication tool.

First, this study has shown the importance of examining online discussion groups regarding the marketing strategies of the Ministry of Health, as well as individual behaviors and sentiments related to the HPV vaccination campaign. In the digital age, public health institutions need to recognize the importance of using Facebook to communicate their message better through existing networks and engage in ongoing dialogue with the public. In general, there was a low rejection rate or negative repercussions among the Facebook users who left their comments on the Ministry of Health’s posts. This information should be interpreted with caution because Facebook allows echo chambers to emerge, in which well-segregated communities emerge from the users’ consumption habits.⁴² Pro- and anti-vaccination attitudes polarize Facebook users. In other words, the Ministry of Health campaign that uses Facebook to advocate HPV vaccination should be expected to reach more pro-vaccination users.

Generally, sentiment polarity and emotion type analyzes can be used effectively to identify the target public’s preferences, to detect dissatisfaction related to the HPV vaccine, and to correct unintended errors of campaign communication. Moreover, a prevalent sentiment in the posted comments, such as joy, can contribute to the government’s viral marketing efforts by encouraging positive electronic word of mouth with regard to the HPV vaccine campaign. For example, positive posted comments were related to message sharing and the affirmation that the user had already received the HPV vaccine, which provides an incentive for other Facebook users to join the HPV vaccine campaign. The user-generated content impacts attitudes and behaviors of other social media users. Positive comments posted spontaneously by social media users generate higher source credibility, and desired social identification.⁴³

On the other hand, the Ministry of Health should not ignore negative posted comments (11.8%) because they might be used to detect what is not going right in the vaccine campaign. The more insight that the Ministry of Health has from the population, the better equipped it will be to deliver meaningful messages about the HPV vaccine campaign. For example, some Facebook users showed fear of adverse HPV vaccine events, anger related to the availability of the HPV vaccine, surprise (or unfamiliarity) with the vaccination schedule and disgust at investments and the cost of the vaccination campaign. These issues can (and should) be regularly monitored on the Facebook page to respond promptly to each comment and correct any misinformation.²⁷ The Brazilian Ministry of Health has been responding to its Facebook followers. Our work was not intended to measure the quantity and quality of responses from the Ministry of Health, nor the fluctuations of opinions on the HPV vaccines over time. These issues are fundamental for adapting the Ministry of Health’s communication strategies and should be addressed in future research.

The second insight of this study is related to the Facebook audience at which the HPV vaccine message was aimed. This work found that the Brazilian Ministry of Health attempted to target specific audiences with its Facebook postings, although the post with the highest user activity and user engagement was intended for the general public (Post 6). On the other hand, as in other countries, the Brazilian HPV vaccination campaign targets teenagers, and therefore parental acceptance is critical to ensuring uptake, especially since adolescents consider themselves to be passive participants in the decision-making process concerning the HPV vaccine.⁴⁴ Normally, parents are the primary decision-makers regarding the HPV vaccine,^19,44 and social media offers a good opportunity to align narrative communication about protecting adolescents and the HPV vaccination campaign.¹⁹ In this sense, Post 5 exemplifies the importance of the messages of the Ministry of Health that target parents because of the high number of individuals who liked, commented on, or shared this post. Finally, it is important to note that Brazilian Facebook requires everyone to be at least 13 years old before they can create an account, which leads us to believe that the Facebook page is not the main form of communication for the target audience of the HPV vaccination campaign.

From a strategic point of view with regard to communication tools, it is also important that the managers of the Facebook page should attempt to gauge the opinions of the leading members in the reference group of their target audience and use their posts to promote user engagement. Previous research has shown that people who are important to Facebook’s specific network influence other users’ engagement.³³ When the vaccine content is delivered by the leading members of a network or using sucess stories, as in the Ministry of Health video inspired by famous series, Facebook followers tend to have more positive engagement. Health professionals also play a critical role in recommending the HPV vaccine as a routine vaccine,⁴⁵ even on social media. The post with the most frequent positive comment rate was an image that targeted health professionals. As in previous studies, this result suggests that many people recognize an important educational and advocacy role for health professionals.^19,46 Messages about the HPV vaccine as a clearly recommended medical treatment appear to improve vaccination rates substantially, and some recommendation styles may be more effective at increasing vaccine receipt.⁴⁷ The HPV vaccine should become normalized in pediatric practice as a routine vaccine, treated like any other,⁴⁶ and health care providers should be encouraged to simply state, “your child is due for an HPV vaccine today”.⁴⁷

The third insight of this study concerns content published on the Facebook page of the Ministry of Health. Generally, current messaging from government agencies emphasizes the role of the vaccine in preventing cancer, which is an effective message for young women,²⁸ but may not be the most effective message for vaccinating teenagers.²⁷ Moreover, previous studies have shown that there is little evidence to suggest a positive association between hearing stories about the diseases preventable by the HPV vaccine and vaccination behavior.¹⁹ Attempts to increase concerns about communicable diseases may be likely to be counterproductive⁴⁸ especially in a scenario where concerns about adverse events of the vaccine represent one of the main barriers against HPV vaccination.^22,28 Among published posts classified as “vaccine-guaranteed protection”, 40.4% of the attached comments were negative. According the main topic classification, 84% of the comments classified as “vaccine adverse event”, were negative and 40% expressed fear. Despite the importance of HPV vaccine safety communication, no published post by the Ministry of Health campaign addressed this issue. The Brazilian HPV vaccination campaign for pre-adolescent girls and boys should communicate to parents that the safety of the vaccine has been firmly demonstrated. For credible sources, such as the Brazilian Ministry of Health, risk negation communication leads to lower perceived risk.^21,49

On the other hand, the Brazilian population seems to positively recognize the efficacy of the vaccine in their comments (68.8% of comments classified as “protection guaranteed by vaccine” were positive and 74.1% expressed joy) as well as the new vaccination schedule, requiring only 2 doses of HPV vaccine (81.8% of comments classified as “vaccination schedule” were positive and 76.7% expressed joy). Other issues worth highlighting because of the high rate of negative comments were investiment and cost of vaccination campaign, vaccine availability and political issues. In 2014, the Brazilian HPV campaign was launched in high schools, where the first rounds of doses took place with a high coverage rate, but the second round of immunization was transferred to the health centers where it continues until today. Most of the comments classified as “vaccine availability” expressed sadness related to the health centers’ opening hours and the unavailability of the vaccine. School vaccination has meant convenience, availability and access to the HPV vaccine in Brazil. Although this issue should be addressed by the Ministry of Health’s communication strategies, it is important to highlight that easier availability alone may not be sufficient for a substantial increase in HPV vaccine uptake.^50,51 Finally, negative comments on investment and the cost of the vaccination campaign, as well as political issues, are related to the distrust of Brazilians regarding the priorities for public health, unjustified government spending and corruption. There is a relationship between public perceptions of politicization of the HPV vaccine and public support for HPV vaccine policies. The extent to which the population perceives politicization is directly related to public support for the vaccine’s use.⁵²

Last but not least, the fourth insight has to do with whether the strategies used by the Ministry of Health are in line with what the literature indicates are effective strategies. Clearly, the prevalence of images on the Facebook page shows that the Ministry of Health recognizes the importance of visual content sharing on social media. Since one the reasons users use Facebook is for recreation,³³ posting pictures and videos instead of texts is important for maintaining an engaged audience of followers.³³ This study also found that posts with video content (Post 6) generated the highest user activity and user engagement among Facebook followers, but specific research is needed to show whether video content is preferred over the posting of images. Furthermore, oftentimes, images and videos do not allow the disclosure of all the necessary information. Since many people first find health information through Facebook,⁵³ it is an assertive strategy to build a bridge between social followers and public health information websites, as was done by the Brazilian Ministry of Health.

According to determinants of Facebook social engagement,³⁶ this work showed that posts published after 6:00pm had a high rate of user activity and user engagement. All posts were published at a low frequency (not more than 2 posts per day), which is an advantage because maximum engagement occurs when the daily frequency is low.⁵⁴ On the other hand, it is difficult to identify a clear-cut relationship between the day of the week when content is posted and the resulting user activity and user engagement. No analysis was done to indicate statistical significance of the association between posting day and engagement in the case of the Brazilian Ministry of Health, but the most posts were made during the working week and the highest impact on engagement was achieved from Tuesday to Thursday. Neither is the relationship clear between the length of a post and engagement, but the peaks of user activity and user engagement were found in short posts (Posts 5 and 6, with fewer than 40 words).

The results of this study need to be interpreted within several constraints, mainly because our sample was limited to the Facebook page data of the Brazilian Ministry of Health from January to July 2018. Our results may not be generalizable to other online content. It is likely that Internet users are exposed to many sources of HPV vaccine information in the real world and the permanence within specific echo chambers shapes their psychological profiles. Moreover, since it is almost impossible to identify all variations of text representing all possible human emotions, emotion detection is a research field with many unturned stones and possibilities. Detecting the personality of a person, relating emotion to different social and individual parameters, detecting the mood of an individual, predicting a person’s action based on his/her emotion – these are limitations of our study that could be addressed in future research.⁴⁰

Conclusion

This study investigated the Ministry of Health’s communication strategies on HPV vaccination on its Facebook page as well as user activity and user engagement. Analyzing this theme without evaluating the fluctuations of opinions on the HPV vaccines over time makes it difficult to determine whether higher levels of user activity and user engagement are the by-product of an experimental use of social media or simply the outcome of a consolidated and systematic strategic approach to social media.³⁶ Beyond the need to constantly review new HPV vaccination campaigns, future research suggestions involve investigating how to integrate many types of social media and traditional media to increase awareness and promote HPV vaccination. The likelihood of success is substantially increased by the application of multiple interventions. Looking forward, we believe there is a clear need for studies capable of determining the factors that affect the formation and adoption of beliefs regarding Brazilian public health interventions, particularly exploring how social media can be used to support a healthy relationship among all the stakeholders in question.

In conclusion, although Brazil is a country strongly influenced by social networking, it appears that social media analysis as a monitoring tool in public health has yet to become routine practice in Brazilian government health agencies. As people are increasingly accessing information online, health professionals need to move quickly beyond simply offering online information about vaccinations and proceed to monitoring and measuring the consequences of each published item of information, including investing in social listening technologies. Although analysis of social media generated content has emerged as an important field of study, this was the first study conducted in Brazil on a government vaccination campaign, and it was performed at a critical time of low coverage rate of the HPV vaccine.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Authors’contributions

All of the authors contributed to the planning, writing and revision of the manuscript. All the authors have read and approved the final manuscript.

Ethical approval

Not required given the nature of the article.

Acknowledgments

We are grateful to our editor and anonymous reviewer for the tips contributions and recommendations. All comments were very helpful, and we believe that by addressing the comments and suggestions, our revised manuscript was significantly improved.

Supplementary material

Supplemental data for this article can be accessed online at here.

Notes

1 https://apps.facebook.com/netvizz/.