Trends in Human Papillomavirus Vaccine Safety Concerns and Adverse Event Reporting in the United States

Key Points Question Does public sentiment of human papillomavirus (HPV) vaccine safety align with spontaneous HPV vaccine adverse event reporting data? Findings This cross-sectional analysis of the 2015 to 2018 National Immunization Survey indicates a 79.9% increase in the proportion of parents who refused the HPV vaccine for their adolescents due to safety concerns. In contrast, estimates from the national vaccine safety surveillance system found that the HPV vaccine adverse event reporting rate per 100 000 doses distributed decreased from 44.7 in 2015 to 29.4 in 2018. Meaning These findings suggest an urgent need to combat safety concerns about the HPV vaccine in the US.


Introduction
The human papillomavirus (HPV) vaccine is effective for the prevention of up to 6 cancers (cervical, anal, oropharyngeal, penile, vaginal, and vulvar). [1][2][3][4] Despite being licensed for over a decade, the HPV vaccine coverage remains suboptimal in the US, with 46% of vaccine-eligible adolescents not up-todate in 2019. 5 Furthermore, the coverage varied substantially across states (from nearly 80% in Rhode Island to only 30% in Mississippi). 5 The coverage among US adults aged 18 to 26 years was also low (only 21.5% in 2018). 6 The HPV vaccine was demonstrated to be safe and effective in trials before its licensure. 1,4 Subsequent analyses of the Vaccine Adverse Drug Event Reporting System (VAERS) also established postlicensure safety of the vaccine. [7][8][9] However, exposure to vaccine misinformation through traditional and social media has created a negative perception of HPV vaccine safety in public. 10,11 According to a recent national study, safety concern was the top reason for parental lack of willingness for initiating HPV vaccination. 12 The unfavorable views regarding HPV vaccine safety are contributing to lack in vaccine confidence at an individual level. At a societal level, the collective sentiment of HPV vaccine hesitancy has had an untoward impact on public health policies. In the past, legislative bills proposing the HPV vaccine mandate were overturned, citing vaccine safety concerns. 13 Despite these repercussions, data documenting HPV vaccine safety perceptions, nationally and across the 50 states, is currently unavailable. Understanding the trends in reasons for failure to HPV vaccinate can provide insights into the extent to which safety-related concerns prevent individuals from receiving the HPV vaccine, and inform the development of interventions to ameliorate this barrier to vaccination.
Data from vaccine adverse event surveillance systems play a critical role in shaping public opinion of vaccine safety. A surge in vaccine safety concerns in the absence of substantive pharmacovigilance data can be indicative of vaccine misinformation in public. Therefore, we performed a parallel assessment of the trends in HPV vaccine safety concerns, as reported in the 2015 to 2018 National Immunization Survey and the trends in nonserious and serious adverse events (AE) reports following HPV vaccination from the 2015 to 2018 VAERS database.

Methods
The institutional review board of the University of Texas Health Science Centre at Houston deemed this study exempt from review and informed consent because it uses publicly available deidentified data. This cross-sectional study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Reasons for HPV Vaccine Hesitancy
To examine trends in reasons for HPV vaccine hesitancy, we used the 2015-2018 National Immunization Survey-Teen (NIS-Teen) data. The NIS-Teen is a nationally representative random-dialdigit telephone survey of adolescents aged 13 to 17 years conducted by the Centers for Disease Control and Prevention (CDC). The survey respondents were adult caregivers most knowledgeable of the adolescent's immunization status. The survey collected information regarding the number of vaccine doses administered. Information on age, sex, race and ethnicity, geographic area of residence, income, and insurance status was self-reported; these data were used in our study to describe the sociodemographic characteristics of the adolescents. Each participant in the NIS was assigned a weight that allowed estimates from the surveyed adolescents to be combined to obtain population estimates (weighted N) that reflected the relative proportions of these groups in the nation as a whole.
For this study, we identified unvaccinated adolescents (received 0 doses of the HPV vaccine) at the time when the survey was administered. Caregivers were then asked if they intend to vaccinate their adolescents in the next 12 months. Parents of unvaccinated adolescents who responded "Not too likely," "Not likely at all," and "Don't know/not sure" were further asked to identify the primary reason for vaccine hesitancy from a list of predefined reasons. If the reason was not listed, the response was solicited in an open-ended manner. In the final data set, all the reasons listed by parents/caregivers were recoded into 28 unique reasons. Additional information regarding the survey methodology and questionnaire is available online. 14

AE Reports
To estimate trends in serious AE reporting, we analyzed the 2015-2018 VAERS database. The VAERS is a national reporting system for monitoring and evaluating vaccine safety in the United States. The system was not designed to examine a causal relationship between the vaccine administered and an AE, but it is useful for signal detection (ie, detecting unusual patterns in AEs). AEs in the VAERS database are voluntarily reported by manufacturers, patients, guardians, health care clinicians, and others. The database contains demographic information of the patient, date of vaccination and date of AE, signs and symptoms of the AE, information regarding the suspected vaccine(s) (including the name of vaccine(s) administered, vaccine type, and manufacturer information), and the outcome of the event. The signs and symptoms reported in VAERS are classified based on a clinically validated standardized methodology, the Medical Dictionary for Regulatory Activities (MedDRA). 15 A detailed description of VAERS database is available on the VAERS website. 16 We identified and included all AE reports listing HPV vaccine (including the quadrivalent

Statistical Analysis
We examined the frequency distribution of reasons for HPV vaccine hesitancy during each year from 2015 to 2018 using the NIS-Teen data. Based on the frequency, the top five reasons for HPV vaccine hesitancy were identified (nationally and across 50 states and the District of Columbia). The trends in reasons for not vaccinating were examined using linear regression models. A χ 2 test was used to compare the proportions for reasons in 2015 vs 2018 across states. All analyses of the NIS-Teen data were adjusted for strata and weights using the SAS SURVEY procedures to account for the complex survey design.
The crude AE reporting rates for the HPV vaccine (per 100 000 doses distributed) were calculated by dividing the number of reports in VAERS by the number of HPV vaccine doses distributed in the US. 18 Similarly, we examined the reporting rate for serious AE, including those leading to hospitalizations, disability, and death or a life-threatening condition. In the sensitivity analysis, reports of serious AE that were submitted based on online information (personal testimony, blogs, Facebook posts, and tweets) were excluded because these reports were inconsistent in providing details of the patient and the AE. 19 Trends in the reporting rate were examined using Poisson models accounting for the number of HPV vaccine doses distributed during each year.
Statistical significance was tested at P < .05. All analyses were conducted per the analytical guidelines for the NIS-Teen database and VAERS data user guidelines. 14,20 All analyses were  proportion of unvaccinated adolescents whose parents cited "Safety concerns"' as the main reason for HPV vaccine hesitancy increased significantly from 13.0% (95% CI, 12.1%-14.0%) to 23   Not needed or not necessary, P for trend =.28 Safety concerns, P for trend <.001 Not recommended, P for trend =.007 Lack of knowledge, P for trend <.001 Not sexually active, P for trend <.001    Figure 3A and Figure

Discussion
We examined trends in reasons for HPV vaccine hesitancy and HPV vaccine adverse event reports a decline in the number of parents who did not initiate the HPV vaccine during 2010 to 2016 owing to concerns of increased sexual activity. 22 A significant increase in citing safety concerns was also reported during 2010 to 2016 in the study and this upward trend has persisted in recent years based on our analyses. 22 Furthermore, our data provide additional insights into public perceptions of HPV  A, The percentage change in safety concerns in decreasing order. B, The population size of individuals reporting safety as a primary concern for vaccine hesitancy in 2015 and 2018.
a Percentage change was statistically significant. vaccine safety in the US and to our knowledge, is the first to present trends in all 50 states and DC and concurrently evaluate AE reporting.

JAMA Network Open | Public Health
The rise in safety concerns noted in our study may have resulted from several reasons. First, it is possible that misinformation related to unsubstantiated AEs of the HPV vaccine on social media and online blogs is increasing mistrust among parents. 11 At least 2 prior studies have documented a rise in negative content related to the HPV vaccine on social media during 2015 to 2017. 23,24 Exposure and engagement with antivaccine content is positively correlated with HPV vaccine hesitancy (correlation coefficient = 0.18, P = .002). 25 In a 2017 survey study, parents who had reportedly heard stories about HPV vaccine harms from social media were more likely to refuse the HPV vaccine than the parents who had never heard such stories (odds ratio, 8.9; 95% CI, 4.1-19.3). 11 Dunn et al 10 [26][27][28][29] The internet has become a major source for parents seeking vaccine information. 30 Stories about health injuries, disabilities, autism, and even death from receiving the HPV vaccine have been circulating on the social and traditional media that may have misled parents to believe that the vaccine is not safe. 11 Fear tactics are often used by antivaccine campaigners to Death or life threatening P for trend <.001 P for trend <.001 P for trend =.47 P for trend =.31 P for trend =.95 P for trend =.70 The figure illustrates reporting rates of adverse events (per 100 000 vaccine doses distributed) following human papillomavirus vaccination. Trends were examined using Poisson models, adjusting for the number of vaccine doses distributed. dissuade parents from vaccinating. Perceptions that vaccines are unnatural or consist of toxic elements are also often propagated. 31 The rise in safety concerns was consistent in nearly all states and DC. Particularly, the highest rise in safety concerns was observed in California. A previous study reported that exposure to negative sentiments regarding vaccines was disproportionately higher in California. 32 Similarly, in Mississippi (where a more than 200% rise in safety concerns was observed), high distrust in parents regarding the HPV vaccine was documented as the most substantial barrier to vaccine uptake. 33

Limitations
This study had a few limitations. Data on trends in safety concerns were examined using the NIS-Teen. Respondents in the NIS-Teen are parents of adolescents (aged 13 to 17 years) eligible to receive the HPV vaccine; therefore, findings may not be generalizable to other vaccine-eligible age groups. Nevertheless, concerns regarding the safety of the HPV vaccine have also been reported in the young adult age group (aged 18 to 26 years) in previous studies. [45][46][47] Trends in AE reports were examined using the VAERS database. The AEs in the VAERS database are spontaneously reported, and information on critical aspects such as existing disease conditions, co-administered vaccines, and use of medications is often missing or incomplete. Spontaneous AE surveillance systems are also prone to reporting bias; nonserious AEs tend to be underreported whereas reporting sensitivity of serious AEs can vary depending on the outcome. 48 However, a recent study from Australia found that serious AE reporting rates following HPV vaccination are similar between periods of passive vs active surveillance. 49 It is important to note that causal inference on the link between the vaccine and reported AEs cannot be drawn from VAERS data for multiple reasons. 50 Nonetheless, summary data on AE reports generated from VAERS, specifically summaries on serious AEs, can increase HPV vaccine acceptance and trust in the public, which demonstrates the importance of descriptive reports from VAERS. 39 Finally, our study examined the trends in safety perceptions and AE reporting from 2 separate data sources that currently do not have linkage capabilities; therefore, findings should be interpreted within the context of this limitation.

Conclusions
In conclusion, concerns regarding safety are rising among HPV vaccine-hesitant parents despite consistent evidence of the vaccine's safety from prelicensure trials and postmarketing surveillance data. These findings suggest that strategies to combat safety concerns and improve vaccine confidence are urgently needed to expedite the achievement of optimal HPV vaccination coverage in the US.