Rates of human papillomavirus vaccine uptake amongst girls five years after introduction of statewide mandate in Virginia

Abstract

BACKGROUND:

The Commonwealth of Virginia enacted statewide school-entry human papillomavirus vaccine mandate in 2008 requiring all girls to receive the vaccine before starting the 6th grade. The mandate, one of very few in the country, has been in effect for 5 years. This study assesses the impact that it has had on the rates of human papillomavirus uptake.

OBJECTIVE:

The purpose of this study was to evaluate the uptake of the human papillomavirus vaccine among girls seeking well-child care 5 years after the introduction of a statewide mandate in Virginia in October 2008.

STUDY DESIGN:

This prospective cohort study used the Clinical Data Repository at the University of Virginia to identify girls 11–12 years old who was seen for well-child care from January to December 2014. Billing and diagnosis codes were used to establish human papillomavirus vaccine administration. Those girls who were identified through the Clinical Data Repository were then contacted by advance letter followed by a representative from the University of Virginia Center for Survey Research who invited the responsible parent or guardian to complete a 50-item telephone questionnaire. Questionnaire results were used to inform objective findings and to assess parental attitudes that were related to human papillomavirus vaccination. Findings were compared against those of Pierce et al (2013), who evaluated human papillomavirus vaccination levels in a similar cohort of patients in 2008, before mandate enactment, to assess relative change attributable to vaccine mandate.

RESULTS:

Nine hundred eight girls were identified through the Clinical Data Repository; 50.9% of the girls received at least 1 dose of human papillomavirus vaccine. White race and private insurance coverage were found to be associated negatively with human papillomavirus vaccine uptake (relative risk, 0.74 and 0.71; 95% confidence interval, 0.64–0.85 and 0.62–0.81, respectively). Black race and public insurance coverage were found to be associated positively with vaccine uptake (relative risk, 1.35 and 1.39; 95% confidence interval, 1.17–1.55 and 1.22–1.58, respectively). In comparison with the previous study, there has been no change in human papillomavirus vaccine uptake or distribution of uptake after the introduction of the statewide mandate for human papillomavirus vaccination.

CONCLUSION:

The statewide human papillomavirus vaccine mandate has had no impact on the overall rate of human papillomavirus vaccination, nor has it diminished the previously described racial or payer disparities in vaccine uptake in school-aged girls being seen for well-child care in the state of Virginia.

In April 2007, Virginia became the first state to enact a law mandating human papillomavirus (HPV) vaccination of girls before entry into the 6th grade. The mandate became effective in October 2008; however, given the timing of when the mandate went into effect, it did not change school admission requirements until the 2009 school year. It allowed for parental opt-out on philosophic, medical, and/or religious grounds after a review of prepared materials. In the time since enactment of the mandate in Virginia, 42 states and territories have considered legislation regarding the administration, education, and/or funding related to HPV vaccination. At present, Virginia, Rhode Island, and the District of Columbia mandate HPV vaccination.¹

The Food and Drug Administration approved the first HPV vaccine in 2006; despite nearly a decade of availability, vaccination rates for HPV lag behind those of other childhood vaccinations. Data from the 2014 National Immunization Survey–Teen demonstrate that 60% of female adolescents aged 13–17 years have received 1 dose of the vaccine; only 39.7% of them have completed the series nationally.² Although this represents an increased rate of uptake since initial licensure, it falls short of the Healthy People 2020 goal of 80%.³

School-entry vaccination mandates, in general, have proved beneficial at increasing vaccine uptake by eliminating disparities among those vaccinated with regard to socioeconomic status, race, ethnicity, and insurance coverage.⁴ However, unlike many other childhood vaccinations, HPV vaccination has been the subject of much criticism and debate. Much of the discussion has focused on the potential barriers to vaccination, whether from the standpoint of both patients and providers or because of the legally allowed exemptions to vaccination in locales that require its administration.⁵

Premandate data were collected in Virginia in 2008 that demonstrated that the rate of HPV vaccine uptake among girls aged 11–12 years who sought well-child care within a large university system was significantly greater among non-white girls than white girls. Furthermore, girls with private insurance were significantly less likely to have been vaccinated than girls with public insurance, regardless of race.⁶

Using the same protocol as the 2008 study, this study sought to evaluate the effect that the statewide vaccination mandate has had on vaccine uptake 5 years after enactment. Additionally, a telephone survey of responsible adults who accompanied dependents for well-child care over the study period was used to assess parental attitudes regarding vaccination. This potentially could be used to help explain objective findings that are related to vaccine uptake or refusal.

Materials and Methods

The protocol for this prospective cohort study has been described previously.⁶ The University of Virginia Clinical Data Repository was searched at 3-month intervals, capturing all girls aged 11–12 years who sought well-child care from January through December 2014. Visits occurred at University of Virginia–supported Family Medicine and Pediatrics practices, which encompass urban, rural, and suburban practice settings. The primary endpoint was receipt of ≥1 dose of HPV vaccine ascertained from abstracted billing data. Final collection of HPV vaccine uptake and number of injections for identified girls was collected 6 months after the last-identified well-child visit. Additional collected demographic variables included age, race, and insurance status. Additionally, vaccination status for meningococcal, tetanus/diphtheria/pertussis, and varicella immunizations was collected for each identified child.

After collection of described objective data, an advance letter was sent to the parents and/or guardians of identified girls, because all were considered eligible. This letter informed them about and invited them to participate in a forthcoming telephone questionnaire about a recent well-child visit. Parents and/or guardians were then contacted by telephone within 1 week of the mailing. Questionnaires were completed by telephone by trained interviewers through the University of Virginia Center for Survey Research. Conversion calling techniques were used to increase response rate per protocol at the Center for Survey Research. The questionnaire contained 50 items and took approximately 10 minutes to complete. Telephone consent was obtained. Included in the questionnaire were previously validated questions on HPV vaccination attitudes and behaviors that had been designed with the use of constructs from the Health Belief Model.⁷ Furthermore, additional questions addressed the status of other childhood vaccinations, relationship and trust with provider, and usual health information sources. Finally, the questionnaire assessed knowledge of HPV and HPV vaccination with the use of previously studied questions on the subject. The questionnaire was piloted in 2009 on a subset of parents whose data were not included in the final data analysis.

Data analysis included frequency of distribution of race, insurance status, and vaccination status in the original population identified by the Clinical Data Repository and in the cohort that participated in telephone questionnaire with the use of SPSS software (version 22; SPSS, Inc, Chicago, IL) and SAS software (version 9.4; SAS Institute Inc, Cary, NC). Primary outcome of ≥1 dose of HPV vaccination was converted to a binary variable. Chi-squared test was used for categoric variables with a probability value of <.05 considered to be statistically significant. Relative risk and 95% confidence intervals were calculated for all cohort data in 2×2 tables. Likert scales that were used in the questionnaire to assess vaccination attitudes were dichotomized to agree or disagree with the attitude. Demographic, knowledge, and attitude data from the telephone questionnaire were used in multivariable logistic regression. All factors that were associated significantly with HPV vaccine uptake or refusal (P < .1) were retained in the model. The final model was fit with backwards stepwise elimination of nonsignificant variables. Sensitivity analysis was conducted to verify the relationship between the variable and the outcome of HPV vaccine uptake.

After completion of primary data analysis as described, the 2014 cohort was compared with the 2009 cohort to determine effect of statewide vaccine mandate on rates of vaccination. Proportion vaccinated was calculated according to race and insurance status for both 2009 and 2014 cohorts. Proportions for each then were compared, and the probability values were calculated; a probability of <.05 was considered significant.

Results

During the study period of January 2014 to December 2014, 908 total patients were identified who met criteria. Comparing demographic data, the postmandate patients in the current study were similar to the premandate patients from 2009. All patients were female and 11–12 years old. Six hundred seventy-seven patients (74.6%) identified as non-Hispanic white; 146 patients (16.1%) identified as African American, and 84 patients (9.3%) identified as any other race or ethnicity. Distribution of race and insurance coverage did not show significant variance (Table 1).

TABLE 1.

Demographics of patients before and after the mandate

Demographic	Before the mandate, n (%)	After the mandate, n (%)	P value
Race			.897
Non-Hispanic white	396 (72.0)	677 (74.6)
African American	115 (20.9)	146 (16.1)
Other	39 (7.0)	84 (9.3)	.770
Insurance
Private	292 (53.1)	487 (53.6)
Public	175 (31.8)	392 (43.2)
Self-pay	22 (4.0)	29 (3.2)

Cuff et al. HPV vaccine uptake 5 years after statewide mandate in Virginia. Am J Obstet Gynecol 2016.

Four hundred forty-two patients (48.7%) received at least 1 HPV vaccination in the series. The potential for a patient to receive at least 1 HPV vaccine in the series was dependent on race and insurance status (Table 2). When compared by race/ethnicity, white girls were approximately 25% less likely to initiate vaccination (relative risk, 0.74; 95% confidence interval, 0.64–0.85); African American girls were 35% more likely to initiate vaccination (relative risk, 1.35; 95% confidence interval, 1.17–1.55). When compared by insurance status, those with private insurance were almost 30% less likely to initiate vaccination (relative risk, 0.71; 95% confidence interval, 0.62–0.81); those with public insurance were almost 40% more likely to initiate vaccination (relative risk, 1.39; 95% confidence interval, 1.22–1.58).

TABLE 2.

Human papillomavirus vaccination^a rate by race and insurance status

Demographic	Vaccinated, n (%)	Relative risk	95% Confidence interval
Race
Non-Hispanic white	329 (48.5)	0.74	0.64–0.85
African American	96 (65.8)	1.35	1.17–1.55
Other	37 (44.0)
Insurance
Private	211 (43.3)	0.71	0.62–0.81
Public	237 (60.5)	1.39	1.22–1.58
Self-pay	14(48.3)

^aVaccination refers to having at least 1 vaccine in the series.

When we compared the 2009 premandate vaccination rates to those of the 2014 postmandate year, there is no significant change in the rates of any group when similar ethnicities and insurance coverages were compared (Table 3). There was approximately a 5% increase in the rate of both white and African American female patients who initiated vaccination and a 15% drop in all other races although none achieved statistical significance. When we compared insurance status, there was approximately a 1–2% shift in vaccination rates from the premandate data, which similarly was not statistically significant.

TABLE 3.

Difference of human papillomavirus vaccination rates by race and insurance status before and after the mandate

Demographic	% Difference	P value
Race
Non-Hispanic white	5.1	.12
African American	4.9	.49
Other	−15.0	.17
Insurance
Private	−1.2	.80
Public	1.6	.79
Self-pay	−1.7	.87

Of the 908 criteria patients, all were considered eligible for participation in the telephone-based questionnaire. Attempts were made to contact the responsible parents and/or guardians of all 908 identified girls; interviews were conducted, and data were completed on 205 to yield a response rate of 22.6%. The most common reasons for telephone questionnaire noncompletion were answering machine without a message, refusal, a nonworking number, a number linked to a fax or data line, and language barrier. All those involved in the questionnaire confirmed that the patient was the appropriate age and that they classified themselves as the patient’s parent or guardian. Most parents were married (64%), employed (63%), and had at least some college education (70%).

Although all but 1 patient had a primary physician for their child, a majority of the respondents (70%) had been with that provider for >5 years, and almost three-quarters of the respondents claimed they received HPV information from their healthcare provider. Almost all of the respondents (91%) went to the physician’s office with the patient; 73% of them could confirm that the physician discussed the HPV vaccination at some point, and 68% of them had heard a recommendation for the vaccine from that same physician.

More than 86% of the respondents confirmed that they were aware of the HPV vaccine, and 85% of them could confirm that the vaccine at least protects against cervical cancer. Although 83% of the respondents believe the vaccine is safe, and 91% of them believe that it is effective; only 14% of them believed that it was required for their child, and 35% of them said that their child does not need it. Only 2 of the 205 respondents believed that the best time for HPV vaccination was when it is required for school; 28 of them (14%) believed the best time for girls to receive the vaccine was before the onset of sexual activity. With regards to other vaccinations, the respondents had their children vaccinated at a very high rate against measles/mumps/rubella (99%), meningococcus (92%), diphtheria/tetanus/pertussis (99%), hepatitis B (98%), and varicella (98%). Very few parents believed that vaccines generally do more harm than good (6%), and even fewer claimed they could not afford it (3%) (Figure).

FIGURE.

Demographic and attitude/belief resultant responses from questionnaire

Percentage of affirmative responses to a selection of statements and/or questions.

HPV, human papillomavirus; MMR, measles/mumps/rubella; P/G, parent/guardian; TDAP, tetanus, diphtheria, pertussis.

Multivariate Logistic regression demonstrated that black patients and those on public insurance were 5.3 and 2.7 times more likely to have initiated the HPV vaccine series, respectively (Table 4). Conversely, parents of patients who expressed safety concerns were significantly less likely to have initiated vaccination.

TABLE 4.

Final regression model: factors found to be independently linked to initiation of the human papillomavirus vaccine series

2014 Cohort	Odds ratio	95% Confidence interval	P value
Black racea	5.3	1.4–20.0	.014
Other racea	1.2	0.39–4.06	.70
Public insurance	2.7	1.38–5.61	.004
Safety concerns	0.83	0.74–0.94	.002
Perceived susceptibility	1.01	0.93–1.2	.26

^aAs compared with non-Hispanic white.

Comment

Statewide school-entry mandate has had no effect on the rate of HPV vaccination in the 5 years since its enactment in Virginia. Furthermore, among a cohort of girls who sought well-child care, HPV vaccine uptake was again noted to be higher among minorities and those with public insurance than white girls or those who were privately insured. These findings are concordant with the premandate vaccination data and with the rates of HPV vaccine uptake, which was defined as ≥1 dose, within the Centers for Disease Control and Prevention (CDC) National Immunization Survey Teen Vaccine Dataset.¹³ Understanding implications of these findings requires a consideration of Virginia law against a broader context of compulsory vaccination in the United States. Additionally, findings of the questionnaire—specifically parental attitudes and knowledge of HPV and HPV vaccine—are instrumental in the evaluation and understanding of the dynamics of HPV vaccination in Virginia.

Vaccine evaluation and approval is regulated by the federal Food and Drug Administration. The vaccine schedule is set by the Advisory Committee on Immunization Practices, a component of the CDC. However, although approval and recommendation fall to the federal government, only the states can require vaccination.⁸ Mandatory or compulsory vaccination falls under the doctrine of police power, which is reserved for the states, and is the rationale for limitation of personal rights when needed in the interest of the common good.⁹

Although constitutional, legal, and historic precedents allow for states to mandate vaccination and to do so as a school-entry requirement, exemptions are permitted. Virginia allows for both medical and religious exemptions for all vaccines recommended as part of the Advisory Committee on Immunization Practices recommended series. However, when the HPV mandate was added to the Code of Virginia, it allowed for an HPV-specific philosophic exemption. The rationale for the exemption reads: “Because the human papillomavirus is not communicable in a school setting, a parent or guardian, at the parent or guardian’s sole discretion, may elect for their child not to receive the human papillomavirus vaccine, after having reviewed materials describing the link between the human papillomavirus and cervical cancer approved for such use by the Board.”¹⁰

Previous studies have documented increased incidence of vaccine-preventable disease in states where philosophic exemption was allowed and noted specifically that the incidence of pertussis was twice as high in such states when compared with those with religious and medical exemptions alone. Furthermore, increasing ease of citing philosophic exemption was also linked to an increasing incidence of disease, with easy and medium ease of exemption associated with >90% and >73% increased incidence of pertussis, respectively.¹¹ The philosophic exemption for HPV vaccination in Virginia is broad, easy to cite verbally, and is largely unenforced. As a result, philosophic exemption is likely a large contributor to the findings of this study.

Our national experience with the effectiveness of mandated HPV vaccination is limited. As noted, only Virginia, Rhode Island, and the District of Columbia presently mandate vaccination as a prerequisite for school entry. The latter 2 have less experience than Virginia, and their effectiveness is still relatively untested. Unlike Virginia, Rhode Island’s policy lacks philosophic exemption from vaccination, which will likely lead to increased rates of HPV vaccine uptake.¹²

The findings of our questionnaire infer that personal exemption is responsible for the lack of improved uptake of HPV vaccination in Virginia. Although nearly 87% of respondents reported being aware of the HPV vaccine, only 51% reported that their daughter had received the vaccine, and 14% reported in the affirmative to the statement “HPV vaccine is required.” The most common reason cited for having chosen not to vaccinate was safety concerns: 41% of respondents. Only 1.5% of respondents cited religious beliefs as the reason for vaccine refusal. Interestingly, 100% of respondents indicated that they thought their child was “up-to-date on vaccinations.”

These findings are not explained entirely by the presence of a lax exemption. Other factors that include parental education and perceived susceptibility to HPV, physician recommendation, and the cost of vaccination are all almost certainly involved in the parental decision to accept or decline vaccination. Our study had a relatively small proportion of parents (8%) who reported that they thought their daughter was “not at risk.” Likewise, 73% of the parents reported that their physician mentioned the HPV vaccine, and 68% recall their physician recommending vaccination. As noted in our baseline study of rates of HPV vaccination, we believe that physician recommendation is 1 of the strongest modifiable factors in promoting vaccination. The design of this study cannot evaluate provider beliefs and recommendation practices, which remains an area warranting further evaluation.

This study evaluated the rates of HPV vaccine uptake in Virginia 5 years after enactment of state-wide school-entry mandate through the quantification of vaccination during well-child visits at a single academic center. Practice settings were multiple and representative of the state as a whole. Though there was an increase in the objective sample size from the original baseline study (550 samples) to the present (908 samples), there was diminished response to the telephone survey, with only a 22% response rate in the current study vs 44% in 2009. Nonresponders were more likely to be insured privately. This also represented a change from the 2009 study wherein nonresponders were more likely to be insured publically. This may be the result of changes in the insurance dynamic after the enactment of the Affordable Care Act during the period of data collection for this study. The impact of these insurance changes on access to care and subsequent rates of HPV vaccination cannot be measured accurately by this study but does warrant future evaluation. Finally, this study cannot assess the attitudes or vaccination status of those who were note evaluated for care at all or who were evaluated for reasons other than well-child care.

HPV vaccination rates in this study are significantly lower than those reported by the CDC for the state of Virginia. In 2014, >59% of female adolescents aged 13–17 years had initiated HPV compared with the 48% that was observed in our findings among 11- to 12-year-old girls.¹³ Although there is a possibility that our cohort is not representative, the numbers of patients who were sampled are similar. Our respective populations differ only by studied age ranges, which likely accounts for the difference in reported rates of uptake.

Although seen as relatively safe and effective, the HPV vaccine has not earned priority among parents with adolescents. It is clear that, to further increase HPV vaccination rates in the United States, exemption policies must become more strict or patient attitude must change. The findings of this study, although largely related to the construct of the mandate in Virginia, have national implications as numerous other states consider their own legislation concerning education, funding, and administration of the HPV vaccine.