Interplay between vaccine-related informational milestones, social media discourse, and vaccine administration rates for human papillomavirus and measles/mumps/rubella vaccines in the United States

ABSTRACT

The digital age has profoundly transformed public health communication. This study investigated the interplay between vaccine-related informational milestones, social media discourse, and vaccine administration trends for human papillomavirus (HPV) and measles, mumps, and rubella (MMR) vaccines from 2011 to 2019. The goal was to understand how vaccine-related public health and policy events align with online discourse and vaccination behaviors. This retrospective study utilized sentiment analysis of Reddit, X, and YouTube posts categorized as positive, neutral, or negative. Medical claims data from the Merative MarketScan Commercial Database were analyzed to track HPV and MMR vaccine administration rates. Vaccine-related informational milestones were compiled from the Immunization Action Coalition, Centers for Disease Control and Prevention, Food and Drug Administration, and Advisory Committee on Immunization Practices. A total of 110 informational milestones were identified. Social media discussions were primarily neutral (59%), with positive sentiments declining from 35% to 24% between 2011 and 2019. HPV vaccine discussions demonstrated a stable positive-to-negative sentiment ratio (2.75:1 by 2019), while MMR discussions surged during specific events, notably the 2015 Disneyland measles outbreak. Informational milestones often coincided with increased social media activity. HPV vaccination rates peaked in August, and MMR rates showed gradual increases with consistent seasonal peaks. Although increased online discussions rarely correlated with immediate behavioral changes, our analysis reveals intricate relationships between informational milestones, social media sentiment, and vaccine administration rates. Public health communication strategies should consider the nature of these events and the timing of announcements to effectively engage the public and promote vaccination.

Introduction

In the digital age, public health communication has undergone a dramatic transformation, enabled by the proliferation of social media platforms.¹ This shift has changed how information about vaccines is disseminated, discussed, and interpreted by the public,² sparking speculation about the impacts of social media on vaccine sentiment over the past few decades. It remains largely unknown how public social media discourse on vaccines shifts perceptions^3,4; moreover, it is unclear the extent to which public opinion and behavior are altered by vaccine-related event milestones and seasonal trends.

The 2010s were marked by significant developments in vaccine policy and public health communication that reshaped the public’s views on vaccination.⁵ In 2011, the Advisory Committee on Immunization Practices (ACIP) expanded its recommendation for human papillomavirus (HPV) vaccination to include males aged 11–12 y, marking a pivotal moment in HPV prevention efforts.⁶ This period also witnessed several high-profile measles outbreaks, including the Disneyland outbreak in 2015, which reignited public debate about vaccination and community health.⁷

Concurrent with these public health developments, social media platforms experienced explosive growth, becoming key arenas for health information sharing and debate. There was a considerable shift toward consumer empowerment in healthcare, with individuals seeking more information and control over their healthcare decisions. As such, online information-seeking became the norm, resulting in overwhelming exposure to a variety of media and messages. A 2018 study found that 80% of internet users search for health information online, with social media playing an increasingly important role in shaping health beliefs and behaviors.⁸ This shift in information-seeking behavior presents both opportunities and challenges for public health communication.⁹

Public agencies, including the Centers for Disease Control and Prevention (CDC) and the World Health Organization (WHO), regularly release information and updates about vaccines. Key types of communication include research updates, recommendation changes, and disease outbreak responses. Whether these official communications have an impact on public discourse and, ultimately, on vaccination rates is unclear and represents a critical knowledge gap.^10,11

Failure to understand public sentiment about vaccination can yield significant business, sociopolitical, and public policy challenges.^12–17 Understanding vaccine sentiment is critical to gain the level of public support needed to maintain high coverage rates.^14,18 Leveraging the power of combined data streams to track public sentiment can provide valuable insights and potentially detect critical shifts in opinion, allowing for timely pivots in communication strategies and public health policies.^19,20 The value of tracking public sentiment toward vaccines has been demonstrated in a few studies,^21–23 but apart from these, there is a dearth of information on the relationship between vaccine-related informational milestones, social media discourse, and vaccine administration patterns.

This study examines the relationships between vaccine-related informational milestones, social media discourse, and vaccination trends for the HPV and measles, mumps, and rubella (MMR) vaccines from 2011 to 2019. The MMR vaccine has been a cornerstone of global childhood immunization programs since the 1970s,^24,25 but has recently faced increasing scrutiny and hesitancy,²⁶ leading to resurgences of measles around the world.²⁷ The HPV vaccine, first approved in 2006, represented a major advance in cancer prevention,^28,29 but has encountered hurdles in public acceptance and policy implementation.^30,31 These vaccines represent different eras in vaccine development and public health communication, and both face unique challenges in the current information ecosystem.^24,32,33 By analyzing social media discourse alongside vaccination rates and key public health announcements, we aim to understand how public health communications and policy changes influence public discourse and, subsequently, vaccination behaviors. This analysis will not only provide insights into the effectiveness of past communication strategies but also inform the development of future approaches to promote vaccine acceptance and uptake.^34,35

Materials and methods

Data collection and processing

Vaccine-related informational milestones

We compiled a comprehensive timeline of vaccine-related events and informational milestones from 2011 to 2019, focusing on United States (US) based activities. This information was sourced from the Immunization Action Coalition’s vaccine timeline,⁵ supplemented by official announcements from key public health agencies (Supplementary Table S1). Informational milestones primarily reflected US regulatory approvals, policy changes, professional associations’ clinical practice guidance, and significant public health announcements related to vaccines, although a few global milestones were included. Milestones were categorized by vaccine type (general, HPV, or MMR).

Social media

The core of this study involved the collection of English-language social media posts from Reddit, X (formerly known as Twitter), and YouTube, focusing on content related to vaccines and vaccination. The data collection phase spanned from January 1, 2011 to December 31, 2019. This period was chosen to encompass a diverse array of public discussions, excluding those influenced by the COVID-19 pandemic. We retrieved posts based on vaccine-specific search queries tailored to each platform, as detailed in our previous paper.³⁶ We collected 86 million discussions from these social media platforms. Top-ranked machine learning algorithms were trained to classify discussions into positive, negative, and neutral sentiments. Discussions related to HPV and MMR vaccines were included in this study.

We systematically gathered content that captured discussions and debates regarding vaccines and vaccination. General vaccine-related posts were those that mentioned any specific vaccine (including HPV or MMR) or vaccines in general. HPV-related and MMR-related posts were those that mentioned each of these vaccines specifically. Discussion volume was defined as the total number of posts of each type in a given time period (week or month).

The sentiment analysis in this study was performed using a previously described dashboard.³⁶ Positive sentiment included posts expressing favorable opinions or experiences regarding vaccines. Neutral sentiment encompassed posts that were informational or contained mixed sentiments. Negative sentiment included posts with adverse opinions or potentially deterring information about vaccines. Detailed definitions, examples, and the full classification framework are described in our previous study.³⁶

Claims data

To track vaccine administration patterns, we accessed anonymized medical claims data from the Merative MarketScan Commercial Database, which includes over 215 million individuals covered by private health insurance in the US.³⁷

HPV vaccine claims were retrieved using Current Procedural Terminology (CPT) codes 90649 and 90651 for individuals aged 11–26 y, per CDC recommendations.³⁸ MMR vaccine claims were identified using CPT codes 90707 and 90710, focusing on children aged 0 to 6 y, per CDC recommendations.³⁹ The extraction process involved filtering the database to include only relevant claims, followed by a thorough verification to ensure accuracy and completeness.

Statistical analysis

We used descriptive statistics, frequencies, and ratios to summarize the sentiment trends and vaccine administration rates. All statistical analyses were conducted using Python 3.10.

For social media discussion volume, we examined the impact of each milestone on HPV-related discussions one week before and after the event, as well as one month before and after, calculating the percentage change from the previous week or month. To understand recurring patterns in the sentiment data, we performed a seasonal decomposition of the time series using the statsmodels library in Python. This technique separated the time series into trend, seasonal, and residual components, allowing us to identify underlying patterns and anomalies in the data. The decomposition helped in distinguishing long-term trends from short-term seasonal effects and irregularities, providing a clearer picture of sentiment dynamics over time.

The monthly claims rates of vaccine administrations were calculated for each target group described above. These rates were determined by dividing the number of vaccines administered each month by the number of individuals in that target group enrolled in the study database for that month. This calculation provided a normalized measure of vaccination uptake, allowing for meaningful comparisons across different time periods and target groups.

Data visualizations were created using matplotlib, pyplot, and seaborn libraries in Python. We generated time series plots and other visualizations to illustrate sentiment trends and patterns.

Ethical considerations

This study used publicly available social media data and did not directly involve human subjects. All claims data were anonymized and aggregated to protect individual privacy, and ethical standards were maintained throughout the research process.

Results

General vaccines

Vaccine-related informational milestones

A total of 110 milestones were identified, including regulatory approvals, new vaccine recommendations, public health announcements, and publication of national survey results (Supplementary Table S1). The CDC accounted for the most milestones (n = 41), followed by the Food and Drug Administration (FDA; n = 37), ACIP (n = 11), WHO (n = 5), and the American Academy of Pediatrics (n = 2). Additional agencies and medical associations contributed 14 milestones in total.

Figure 1 shows the monthly counts of general vaccine-related informational milestones and social media discussion trends from 2011 to 2019. Analyzing the temporal distribution of milestones shows that 2015 and 2018 had the highest activity (18 milestones each), while 2011 and 2012 had the fewest (9 milestones each). The number of milestones also varied month to month: certain months, such as June and October, were particularly active, with 13 and 12 milestones, respectively, likely linked to regulatory review cycles or ACIP meetings, which are held annually in February, June, and October.

Figure 1.

Monthly trends in general vaccine discussions and counts of all vaccine-related informational milestones from 2011 to 2019.

Patterns of social media discussion

Discussion volume

Milestone-heavy months often saw notable fluctuations in social media discussions (Figure 1). In October 2015, discussion volume surged by over 60% compared to the prior month, followed by a 48% drop the next month. Similarly, June 2015 saw a 13% increase in discussions, followed by a further 28% rise in the subsequent month.

Significant vaccine-related informational milestones sometimes coincided with spikes in discussion volume. For instance, a notable surge in vaccine-related posts was observed in February 2015, coinciding with the Disneyland measles outbreak and subsequent discussions. However, not all milestone months showed consistent engagement. For instance, March 2015 saw a 76% drop in discussions despite multiple milestones, followed by a 14% increase the next month.

The total volume of vaccine-related discussions on social media platforms showed a gradual increase over the years (Figure 1). In 2011, the average monthly volume of discussions was around 65,000–70,000 posts. By 2019, this had grown to an average of 190,000 posts per month.

Distribution of sentiment

Throughout the study period, neutral sentiment consistently dominated the discourse, accounting for an average of 59% of all vaccine-related posts (Figure 1). Positive sentiments made up an average of 28% of posts, while negative sentiments comprised 13%.

Positive sentiments showed a gradual decline over the years, starting at around 35% of posts in 2011 and dropping to approximately 24% by 2019. Conversely, neutral sentiments saw a corresponding increase, rising from about 53% in 2011 to 62% in 2019. Negative sentiments remained relatively stable, fluctuating between 10% and 15% throughout the study period.

Sentiment analysis revealed that, out of 572 d where negative sentiment exceeded positive sentiment, five days showed particularly high differences: September 13, 2011; May 15, 2014; August 13, 2014; September 17, 2014; and October 20, 2014. However, no major vaccine milestones were identified around these dates. Only one milestone was identified in proximity to any of these dates: the CDC published ACIP recommendations for PCV13 and PPSV23 pneumococcal vaccines in adults aged 65 and older on September 19, 2014. Although this milestone falls near some of the negative sentiment spike dates, its temporal relationship does not explain the observed sentiment changes.

HPV vaccine

Vaccine-related informational milestones

Our analysis of 9 key HPV vaccine-related informational milestones between 2011 and 2019 (Table 1) demonstrated corresponding variations in both public discussion volume and sentiment. Major milestones included the release of national survey data, approval of the vaccine for males, and expanded use in adults.

Table 1.

HPV vaccine-related informational milestones.

Date	Event
August 2011	A national survey indicated HPV vaccine rates trail other teen vaccines.
October 25, 2011	ACIP recommended all 11- to 12-y-old males get vaccinated against HPV.
December 10, 2014	FDA approved Gardasil 9 HPV vaccine.
March 2015	CDC published ACIP recommendations for 9-valent HPV vaccine.
December 2015	FDA expanded Gardasil 9 licensure to include males aged 16–26 y.
December 2016	CDC published ACIP recommendations on 2-dose HPV vaccine for younger adolescents.
October 5, 2018	FDA approved expanded Gardasil 9 use for adults aged 27–45 y.
October 25, 2018	ADA adopted policy supporting HPV vaccination to prevent oral HPV.
August 16, 2019	CDC updated HPV vaccine recommendations for adults.

ACIP, Advisory Committee on Immunization Practices; ADA, American Dental Association; CDC, Centers for Disease Control and Prevention; FDA, Food and Drug Administration; HPV, human papillomavirus.

Patterns of social media discussion

Discussion volume

The trajectory of HPV vaccine discussions and sentiment over the study period offers a glimpse into the public’s changing attitudes toward this vaccine (Figure 2). Five peak months for social media discussions were identified: September 2011, October 2012, February 2015, October 2015, and October 2017. The volume of HPV vaccine-related discussions showed significant seasonality, with peaks typically occurring in August and troughs in February, indicating lower public engagement during this month.

Figure 2.

Monthly trends in HPV vaccine discussion and claims rates over time with key vaccine milestones marked.

HPV, human papillomavirus.

A significant shift occurred on October 25, 2011, when the ACIP recommended HPV vaccination for all 11- to 12-y-old males. The months following this recommendation saw a sustained increase in HPV vaccine discussions, with average daily mentions remaining 50% higher than pre-recommendation levels throughout the rest of 2011 and early 2012.

Nearly all milestones were associated with increases in discussion volume in the week following the announcement. The most substantial increase was observed following the October 25, 2011, ACIP recommendation for male vaccination, with an 823% increase in HPV-related mentions (from 1,537 to 14,189). Other notable short-term increases included the December 10, 2014, FDA approval of Gardasil 9 (175% increase) and the October 5, 2018, FDA approval for expanded use in adults (125% increase). The only milestone that did not result in increased discussion was the December 2015 FDA expansion of Gardasil 9 licensure for males aged 16–26 y, which saw a slight decrease (−8%) in the week following the announcement.

When examining month-long periods, the October 2011 ACIP recommendation again showed the largest increase (310%), followed by the release in August 2011 of the National Immunization Survey results (150% increase) (Figure 2).

Distribution of sentiment

Sentiment analysis indicated a stable positive perception of the HPV vaccine over the study period (Figure 2). In 2011, the ratio of positive-to-negative sentiment (i.e., the number of posts with positive sentiment versus the number of posts with negative sentiment) was 2.68:1. By 2019, this ratio had slightly improved to 2.75:1.

Seasonal decomposition of positive sentiments revealed a clear seasonal pattern, with peaks typically occurring in August. However, sentiment changes were more varied than discussion volume. In the short term (one week following their occurrence), 5 of the 9 milestones resulted in decreased positive sentiment, three showed improved sentiment, and one remained unchanged. The most significant short-term decrease in positive sentiment occurred after the December 2014 FDA approval of Gardasil 9, with the positive-to-negative sentiment ratio dropping by 25%. Conversely, the largest short-term improvement in sentiment followed the October 2011 ACIP recommendation expanding vaccination to males, with an 8% increase in the positive-to-negative sentiment ratio. Broader sentiment shifts were observed in 2012, 2015, and 2019, each showing substantial increases in positive sentiment compared to previous years.

When analyzing month-long periods, sentiment changes were generally less pronounced than changes seen in shorter designated timeframes. The most substantial improvement in sentiment was observed following the March 2015 CDC publication of ACIP recommendations for the 9-valent HPV vaccine, with a 15% increase in the positive-to-negative sentiment ratio. The largest decrease in positive sentiment over a month-long period occurred after the October 2018 FDA approval for expanded adult use, with a 21% drop in the positive-to-negative sentiment ratio.

Notably, some milestones showed divergent trends between short-term and long-term sentiment changes. For instance, the March 2015 CDC publication showed a modest 3% improvement in the week following the announcement, but a more substantial 15% improvement over the subsequent month, underscoring the complex nature of public response to health recommendations and regulatory changes

Vaccine administration rates

The analysis of HPV vaccine-related data from 2011 to 2019 revealed a complex interplay between social media discussions and insurance claims for vaccination. Figure 2 shows the HPV vaccine claims rates on the same timescale as vaccine-related informational milestones and social media discourse. No consistent correlation was observed between the peak months of social media discussion (September 2011, October 2012, February 2015, October 2015, and October 2017) and the patterns of HPV vaccine claims. In fact, HPV vaccine administration rates showed significant variability over time, with a mean monthly administration rate of 0.91% (SD 0.36%), ranging from 0.41% to 2.01%. The trend in administration rates can be divided into two distinct periods: an increase from 2011 to 2016, followed by a slight decline from 2016 to 2019.

Our analysis suggests that HPV vaccine administration rates were more influenced by seasonal patterns than social media discussions or informational milestones. Strong, consistent peaks were observed in August (mean peak 1.65%, SD 0.22%) and troughs in February (mean trough 0.66%, SD 0.09%) of each year. This seasonal trend persisted regardless of discussion volumes or informational milestones. For example, September 2011 recorded the highest number of social media discussions but had the lowest claims rate (0.60%) among all peak claims months. Conversely, October 2015 exhibited the highest claims rate (1.03%) but had a moderate discussion count. Comparing the claims rates across the same months in different years further emphasizes the importance of seasonal effects. Claim rates in October generally showed an upward trend over the years, regardless of discussion volumes.

MMR vaccine

Vaccine-related informational milestones

There were 5 key MMR vaccine-related informational milestones between 2011 and 2019 (Table 2). Notably, despite a regional outbreak of measles in 2015, both rubella and measles were declared to be eliminated in the Americas during this time.

Table 2.

MMR vaccine-related informational milestones.

Date	Event
June 14, 2013	CDC published recommendations for preventing measles, rubella, congenital rubella syndrome, and mumps.
January 23, 2015	CDC issued measles outbreak advisory linked to Disneyland.
April 29, 2015	PAHO declared rubella eliminated in the Americas.
September 27, 2016	PAHO and WHO announced measles elimination in the Americas.
January 12, 2018	CDC published ACIP recommendations for third MMR dose during mumps outbreaks.

ACIP, Advisory Committee on Immunization Practices; CDC, Centers for Disease Control and Prevention; PAHO, Pan American Health Organization; MMR, measles, mumps, and rubella; WHO, World Health Organization.

Patterns of social media discussion

Discussion volume

Analysis of MMR vaccine-related discussions highlighted several notable peaks in social media discourse throughout the study period: September 2011, October 2012, February 2015, October 2015, and October 2017 (Figure 3). All 5 milestones were associated with increases in social media discussion volume in the week following the event. However, the most significant surge in MMR vaccine-related posts occurred in February 2015, coinciding with the CDC’s health advisory about a multi-state measles outbreak linked to Disneyland. This event triggered an unprecedented volume of discussions, far surpassing any other period in our dataset. There was a 406% increase in MMR vaccine-related mentions in the week after the announcement, and a 718% increase in the month after the announcement (Figure 3). Seasonal decomposition of overall MMR vaccine discussion counts indicated significant patterns, with peaks in discussion volume typically occurring in February and October. Seasonal troughs were observed in December, indicating lower public engagement during this month.

Figure 3.

Monthly trends in MMR vaccine discussion and claims rates over time with key vaccine milestones marked.

MMR, measles, mumps, and rubella.

Other notable increases in social media discussion volume included the September 27, 2016, Pan American Health Organization (PAHO) and WHO announcement of measles elimination in the Americas, which saw an 84% increase over one week and a 52% increase over one month. This announcement was significant, as it marked the Americas as the first region in the world to eliminate measles, culminating a 22-y effort involving mass vaccination against MMR throughout the Americas.⁴⁰ Similarly, the January 12, 2018, CDC publication of ACIP recommendations for a third MMR dose during mumps outbreaks resulted in a 98% increase in discussion volume in the week following and a 47% increase over the month-long period. The smallest increase in discussion volume was observed following the April 29, 2015, declaration of rubella elimination in the Americas, with only a 5% increase in the week following and a 15% increase over the month-long period following.

Distribution of sentiment

Sentiments of social media posts were generally positive or neutral, and neutral sentiments consistently outweighed positive and negative sentiments over the years. Positive sentiment reached its peak in 2015. Negative sentiment, while generally lower, peaked in 2013, 2015, and 2019. Seasonal decomposition of positive sentiments toward the MMR vaccine revealed a clear seasonal pattern, with peaks typically occurring in February and October.

Three of the 5 milestones showed improved positive-to-negative sentiment ratios in the week following the announcements. The most significant short-term improvement in sentiment occurred after the June 14, 2013, CDC publication of recommendations for preventing measles, rubella, congenital rubella syndrome, and mumps, with the positive-to-negative sentiment ratio increasing by 25% (from 12:1 to 15:1). The January 2015 CDC health advisory on a multi-state measles outbreak linked to Disneyland was also associated with substantial improvements in sentiment, showing a 21% increase in the positive-to-negative sentiment ratio over one week and a 29% increase over one month. Interestingly, the April 2015 rubella elimination announcement was the only milestone that resulted in a decrease in the positive-to-negative sentiment ratio, both in the short term (−8% over one week) and long term (−0.4% over one month), despite a modest increase in discussion volume.

Vaccine administration rates

Figure 3 presents the trends in MMR vaccination claims rates on the same timescale as vaccine-related informational milestones and social media discourse. The vaccine administration rate trends demonstrated a gradual upward trajectory over the 9-y study period. The mean monthly vaccine administration rate was 2.06% (SD 0.35%), ranging from 1.35% to 3.08%. Notably, MMR vaccine administration rates also exhibited strong seasonal patterns. Peak rates consistently occurred in August, with a mean peak rate of 2.84% (SD 0.15%). The lowest rates were typically observed in December, with a mean trough of 1.55% (SD 0.12%).

In April 2013, in conjunction with a notable peak in MMR-related social media discussions, vaccination claims rates increased to 2.22% from 1.91% in March. In August 2013, discussions spiked again and claims rates also rose from 2.46% in July to 2.97% in August before decreasing to 2.09% in September. Despite these peaks in discussions and the corresponding rise in claims rates, the overall trends in claims rates suggest a seasonal pattern rather than a direct correlation to increased social media discourse, and the timescale of changes in claims rates is longer than that of discussion volumes. As an example, MMR discussions peaked in March and April of 2019. Correspondingly, claims rates increased from 1.99% in February 2019 to 2.12% in March, 2.63% in April, and further to 2.72% in May.

MMR vaccine claims rates did appear to respond to certain informational milestones. In December 2014, the claims rate was 1.57%. Following the measles outbreak linked to Disneyland in January 2015, the claims rate rose to 2.00% in January 2015 and continued to climb, reaching 2.40% in February 2015, which also coincided with a significant spike in social media discussion volume.

Discussion

While prior studies have discussed the potential impact of vaccine-related information milestones, our study is among the first to quantitatively evaluate their effects using real-world vaccine uptake data. We comprehensively examined the interplay between multiple data sources – including public policy milestones, discussions across three social media platforms, and vaccine claims data – focusing on two major vaccine targets: MMR and HPV vaccines. The analysis incorporated more than 86 million social media posts and medical claims data from 215 million individuals, strengthening the validity of the findings.

Our study examined trends in vaccine-related social media discussions from 2011 to 2019, revealing changing patterns of public engagement with vaccine topics. Social media sentiment analysis revealed that neutral sentiments consistently dominated discussions, with positive sentiments declining over time and negative sentiments remaining stable. These finds align with previous research that noted a similar pattern in their analysis of X data related to HPV vaccination.⁴¹ The shift away from binary pro- or anti-vaccine positions may be attributed to increased public health education efforts, more nuanced media coverage, or a growing recognition of the complexity of vaccine-related issues among social media users.^42,43

For the HPV vaccine, nearly all informational milestones were associated with increases in discussion volume, and there was a remarkably stable positive perception of the HPV vaccine over the study period, with a clear seasonal pattern; the peaks in positive sentiment typically occurred in August. However, HPV vaccine administration rates were more influenced by seasonal patterns than social media discussions or informational milestones. For the MMR vaccine, all informational milestones were associated with increases in social media discussion volume in the week following the event, but seasonal peaks in discussion were also observed. Notable peaks in discussion volume, positive sentiment, and vaccine administration rate were related to the 2015 measles outbreak at Disneyland, but overall, vaccine administration rates were more seasonal than milestone-related.

The disconnect between specific policy milestones and sentiment fluctuations reveals a complex relationship between official health communications and public perception. The lack of major vaccine milestones surrounding the most significant negative sentiment spikes suggests that other factors or uncaptured events may have influenced these spikes. In contrast, the sustained interest following the October 2011 ACIP recommendation indicates that specific policy milestones impacting clinical guidelines can have a lasting impact on public discourse.

Data from the 2011 National Immunization Survey further indicated that the lag in HPV vaccination rates compared to other adolescent vaccines was likely associated with heightened public engagement. Furthermore, the divergence between short-term and long-term sentiment changes, specifically regarding the March 2015 CDC publication, underscores the complex nature of public responses. This delayed reaction suggests that the impact of informational milestones on public discourse may be more subtle and gradual than immediate and dramatic.

Response to informational milestones

The dataset identified several vaccine-related informational milestones. Temporal analysis revealed peaks in milestone activities in 2015 and 2018, reflecting intensified public health efforts (or reporting of such efforts) during these years. Regulatory approvals and recommendations often serve as critical junctures in public discourse. For instance, the October 2011 ACIP recommendation for HPV vaccination in males aligned with a substantial increase in HPV-related social media mentions, demonstrating the potential influence of policy changes on public engagement. Similar findings were reported by Massey et al., ⁴⁴ where an increase in online engagement was observed following major vaccination policy shifts. Disease outbreaks appear to be particularly powerful in driving public engagement. The 2015 Disneyland measles outbreak was associated with an unprecedented surge in MMR-related discussions, highlighting the role of timely public health communication in encouraging vaccine-related discourse.

However, the impact of informational milestones can be significantly influenced by media coverage and competing news stories, sometimes leading to shifts in public sentiment and engagement.⁴⁵ Our analysis revealed significant negative sentiment spikes on specific dates; however, these spikes did not correspond to informational milestones but rather coincided with notable medical news events. For example, on May 15, 2014, a date when negative sentiment spiked, the Center for Infectious Disease Research and Policy released a news scan, and reports emerged regarding an unprecedented trial where a woman’s cancer was reportedly killed by the measles virus.^46,47 Other examples of news stories that could influence public sentiment toward vaccines are politician comments,⁴⁸ lawsuit initiations,^49–53 and product recalls.⁵⁴ The 2019 measles outbreaks in New York City, which were extensively covered in the media, further influenced discourse and fueled legal battles over mandatory vaccination policies.⁵⁵

These examples highlight that public discourse on vaccines is not solely influenced by official informational milestones but also by media narratives, lawsuits, and high-profile public statements.^56,57 Future research should explore the relative impact of these different types of events to better understand how public sentiment evolves in response to both structured public health messaging and broader societal conversations.

Seasonality of social media discourse

Seasonality of social media discourse was observed for both HPV and MMR vaccines, and recurring peaks in positive sentiment were observed during specific months. January and February consistently showed increased positive sentiment, possibly due to new vaccination campaigns, public health announcements, or heightened awareness following the holiday season. Another period of increased positive sentiment was observed in September and October, possibly associated with back-to-school vaccination drives and public health initiatives targeting school-age children. These event- and media-driven temporal effects have similarly been observed in other vaccination studies.^58–60 From a programmatic standpoint, this evidence collectively suggests that vaccine promotion campaigns may be best aligned with seasonal trends in public interest and vaccination rates. Additionally, training for health communicators should raise awareness of these periods of intensified social media dialogue.

Trends in vaccination claims

The analysis of vaccine administration claims, specifically during peak months of social media discussions, revealed a complex relationship between public sentiment and vaccination claims rates. In several instances, claims rates during peak discussion months of the HPV vaccine were lower or only slightly higher than the previous or subsequent months, suggesting that social media discussions alone are not a reliable tool for predicting vaccination uptake, particularly for gradual trends.⁶¹ A delayed response indicates the influence of other factors such as public health campaigns or seasonal trends.⁶²

Changes in vaccination uptake are often subtle and can be influenced by multiple environmental factors, including structured immunization programs, media coverage, and real-world events. In our analysis, only large-scale, highly publicized events showed a measurable impact on both social media activity and vaccination uptake. This underscores the need for multifaceted public health strategies that integrate multiple data sources rather than relying solely on social media metrics to assess vaccination behavior.

The data also reveal that changes in claims rates most often align with general seasonal trends rather than online discussion trends. Factors such as structured vaccination schedules and clinical recommendations have been shown to play a more consistent role in influencing vaccination uptake than social media activity.⁶³

Implications for future communication strategies

The complex nature of public response to health recommendations and policy actions was evident in this study. Future research should continue to explore these dynamics, incorporating a broader range of data sources to provide a more comprehensive understanding of the factors driving public engagement and vaccination behaviors. The shift toward more neutral discussions, alongside the decline in positive sentiment toward vaccination, suggests a need for clearer communication approaches.⁶² Additionally, the varying impact of different types of milestones indicates that public health communicators should tailor their strategies based on the nature of the event. For example, Massey et al.⁴⁴ provides evidence that major policy shifts tend to generate immediate engagement, whereas smaller regulatory updates may have a limited effect on public discourse. The seasonal patterns in vaccine administration rates highlight the importance of timing in promoting vaccination via media and organized outreach.

Limitations

This exploratory study has several limitations. The analysis was limited to select social media platforms, which may not capture the full spectrum of public vaccine discourse or sentiments expressed offline, on other platforms, or in traditional media. Sentiment analysis tools are also prone to errors in detecting sarcasm, mixed sentiments, and nuance, potentially leading to misclassification. This analysis covered the period from 2011 to 2019, prior to substantial shifts in the information and communication landscape that occurred during and after the COVID-19 pandemic. During the study period, vaccine-related discourse was shaped by the prevailing social media platforms, media consumption patterns, and public health communication strategies of that time. Since 2020, the scale, speed, and polarization of vaccine-related information – particularly on social media – have changed markedly, alongside increased public engagement with vaccination topics. As a result, the relationships observed between vaccine-related information milestones, social media discourse, and vaccination trends may not fully reflect the current dynamics. Additionally, this study was not designed to comprehensively track all vaccine-related media events. While significant milestones such as CDC and ACIP recommendations were incorporated, other influential factors – including news reports, medical breakthroughs, lawsuits, and misinformation – were beyond the scope of this analysis. Together, these considerations may affect the generalizability of findings to more recent periods characterized by intensified digital communication, evolving platform algorithms, and heighted public attention to vaccination. The claims data used in this study may contain omissions and miscoding.⁶⁴ This study focused on claims data with commercial insurance through large employers, which limits generalizability to the broader US population and likely underrepresents the uninsured and those insured through other programs.⁶⁵ This study examined monthly vaccine administration rates rather than long-term vaccination coverage. Without longitudinal follow-up or continuous enrollment, gradual trends in vaccination uptake may not have been fully captured. Finally, evolving social media algorithms and shifting health communication strategies further complicate the interpretation of online vaccine discourse trends. Platform policies, algorithmic content ranking, and coordinated messaging campaigns may have affected the visibility and reach of vaccine discussions over time, such that online discussion volume does not directly reflect the general opinion of the lay public.

Conclusion

Public sentiment regarding HPV and MMR vaccines has become more nuanced over time, with neutral sentiments dominating, positive sentiments declining, and negative sentiments remaining stable. These observed trends were impacted by vaccine-related informational milestones such as regulatory approvals and public health announcements in a few instances, which influenced social media engagement and public discourse. There is a complex interplay between social media discussions and vaccination rates, often exhibiting seasonal patterns (e.g., around vaccination campaigns). Public health communication strategies should consider the nature of these events and the timing of announcements to effectively engage the public and promote vaccination.

Biography

Amanda L. Eiden is a Director in MSD’s Outcomes Research team. She conducts research activities to support vaccination for all vaccines and age groups; topics include broad impacts of vaccination, valuation of vaccines, and vaccine confidence. At MSD she has also worked in Global Market Access on the pneumococcal program and Global Policy on the pipeline vaccines. Prior to joining MSD, Amanda’s professional and academic experiences broadly spanned public health and research, with work designed to advance policies and best practices in the United States, Thailand, and Panama. Some notable experiences include quality of life evaluations to understand the impact of clinical interventions in pediatric cerebral palsy patients, leading infectious disease investigations at the Florida State Health Department – Epidemiology, and academic research focused on reducing vector-borne disease and overall pest-related public health impacts. Amanda holds a PhD from the University of Florida, an MPH from the University of South Florida, an MBA from Rutgers University, and a BS from the University of Wisconsin – Stevens Point.

Funding Statement

This work was supported by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA (MSD). The content is the sole responsibility of the authors and does not necessarily represent the official views of MSD or IMO Health.

Disclosure statement

ALE, DW, and PF are current employees of Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA (MSD), who may own stock and/or stock options in Merck & Co., Inc., Rahway, NJ, USA (MSD). Intelligent Medical Objects, Inc., was compensated for activities related to execution of the study.

Data availability statement

The data that support the findings of this study are available from Merative MarketScan. Restrictions apply to the availability of these data, which were used under license for this study. Data are available from the authors upon reasonable request and with permission of Merative MarketScan.

Ethics statement

This study used publicly available social media data and did not directly involve human subjects. All claims data were anonymized and aggregated to protect individual privacy, and ethical standards were maintained throughout the research process.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21645515.2026.2655512