We gratefully acknowledge Charlene Wong for her comments on an earlier draft of the manuscript, and Tim McNeel and Will Waldron of Information Management Services, Inc. for their programming support.
Rebecca Anhang Price prepared this manuscript while an SAIC-Frederick, Inc. contractor to the Applied Cancer Screening Branch, Behavioral Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, Bethesda, Maryland.
The findings and conclusions in this report are those of the authors and do not necessarily reflect the official position of the Department of Health and Human Services, the National Institutes of Health, the Centers for Disease Control and Prevention, or the US Government, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government.
The Centers for Disease Control and Prevention recommends catch-up administration of human papillomavirus (HPV) vaccines to girls and women ages 13 to 26 who have not been vaccinated previously. In response to debate regarding catch-up vaccination of young adult women, this study examined whether 18- to 26-year-old women most likely to benefit from catch-up vaccination were aware of the HPV vaccine, and initiated the vaccine series by the end of 2008.
We used data from the 2008 National Health Interview Survey to assess HPV vaccine awareness and use, and reasons for not vaccinating, among women aged 18-26 years (n = 1583). Sociodemographic, health care access, and health history factors associated with vaccine initiation were assessed using multivariate logistic regression.
Overall, 11.7% of women aged 18-26 years reported receiving at least 1 dose of the HPV vaccine by the end of 2008. In multivariate analyses, younger age, history of previous HPV infection, unmarried status, health insurance, flu shot in the past year, and receipt of 1 or more recommended lifetime vaccines were significantly associated with HPV vaccine initiation. Two-fifths (39.6%) of unvaccinated women were interested in receiving the HPV vaccine (n = 1327). Primary reasons for lack of interest in the vaccine were belief that it was not needed, not knowing enough about it, concerns about safety, and not being sexually active.
Quadrivalent and bivalent human papillomavirus (HPV) vaccines, which were approved by the US Food and Drug Administration for girls and women aged 9-26 years in June 2006 and October 2009, respectively,1, 2 protect against HPV 16 and 18, carcinogenic strains of HPV that are responsible for ≈70% of cervical cancers.3-5 Eleven- and 12-year-old girls are the priority population for HPV vaccination because the vaccines are most efficacious when received before initiating sexual activity.6 Also at that age, clinical examinations are advised to allow for administration of other vaccines and preventive services.7, 8 The Centers for Disease Control and Prevention Advisory Committee on Immunization Practices (ACIP) also recommends “catch-up” HPV vaccination of girls and women aged 13-26 years who have not been vaccinated previously or have not completed the 3-dose series.6 Catch-up recommendations promote vaccination of young women who missed the opportunity to be vaccinated by age 12, either because they entered adolescence or adulthood before HPV vaccines were introduced or because they were not aware of, interested in, or able to access vaccines before they had passed preadolescence.
There is some debate regarding the value of catch-up vaccination among young adult women. Benefits and cost-effectiveness of HPV vaccination decline with age,9 as more women become sexually active and infected with HPV as they enter their twenties.10, 11 Although HPV vaccines can protect against types of HPV with which a woman has not yet been infected, the vaccines provide no therapeutic benefit to those already infected with HPV vaccine types.12-15 In addition, routine cervical cancer screening is currently recommended for adult women regardless of HPV vaccination status.16-18 Screening is highly effective at detecting the cervical abnormalities that can result from persistent HPV infections.
In contrast to clinical practice guidelines from the American Academy of Pediatrics and American College of Obstetrics and Gynecology,19, 20 and to the ACIP guidance, which recommends HPV vaccination for all young adult women although clinicians cannot assess the extent of benefit for individual women,6 the American Cancer Society's guidelines state that there is currently insufficiently evidence for or against universal vaccination of women ages 19-26 years in the general population.21
The present study used data from the 2008 National Health Interview Survey (NHIS), a nationally representative survey of a broad range of health topics, to examine whether 18- to 26-year-old women most likely to benefit from catch-up vaccination are aware of the HPV vaccine, and have received initial and subsequent doses in the 3-dose series. We characterized women as potentially receiving the greatest HPV vaccine benefit if they 1) had no previous history of HPV or abnormal Pap test results, 2) were likely to face continued risk of HPV infection due to sexual activity with new partners, 3) did not participate in routine cervical cancer screening, and 4) were members of racial and ethnic minority, socioeconomic, or health care access subgroups that traditionally experience higher burdens of cervical cancer morbidity and mortality.22-27 Because previous studies have documented a relationship between use of the HPV vaccine and women's use of other vaccines,28, 29 we also assessed whether HPV vaccination differed according to previous receipt of other recommended vaccines.
MATERIALS AND METHODS
We analyzed data from the 2008 National Health Interview Survey. The NHIS is an annual, in-person household survey that collects health information on the civilian, noninstitutionalized population in the United States.30 The survey employs a complex, stratified, multistage sample and provides nationally representative estimates when appropriately weighted. Hispanics, African Americans, and Asians are oversampled so as to ensure adequate representation and stable estimates for these racial and ethnic groups. Further details of the survey are located at http://www.cdc.gov/nchs/nhis.htm. In 2008, 21,781 adult sample persons were surveyed, representing a response rate of 62.6%. The sample under study consists of adult women ages 18-26 (n = 1583). Because off-label use of HPV vaccines among women ≥26 years of age is very limited (≈1%), we excluded these women from our analyses.
The outcomes of primary interest, awareness of the HPV vaccine, and receipt of a first vaccine dose (vaccine “initiation”), were measured via 2 questions, “A vaccine to prevent HPV infection is available and is called the HPV shot, cervical cancer vaccine, or GARDASIL. Before this survey, have you ever heard of the HPV shot or cervical cancer vaccine?” and “Have you ever received the HPV shot or cervical cancer vaccine?” We also analyzed receipt of all 3 doses of the vaccine (vaccine “completion”), interest in receiving the vaccine (among those not previously vaccinated), main reasons for not vaccinating (among those not interested in vaccination), willingness to pay $360-$500 for the vaccine (among those interested but not previously vaccinated), and willingness to receive the vaccine for lower cost (among those interested, not previously vaccinated, and not willing to pay $360-$500).
Covariates were selected to measure subgroups who we identified as most likely to benefit from catch-up HPV vaccination in adulthood.
Previous exposure to HPV was measured as whether the respondent reported past HPV infection, or past abnormal Pap test results. Those with previous exposure to HPV were considered to have less to gain from HPV vaccination.
Likelihood of future risk of HPV infection was measured indirectly by marital status (married or all other, including not currently married and living with a partner). Unmarried women were grouped, regardless of whether they were living with partners, because our preliminary analyses showed that unmarried women with and without partners exhibit similar patterns of HPV vaccine use. Unmarried women were considered to be at higher risk of future HPV infection than married women because they were considered more likely to have multiple sexual partners, a risk factor for HPV infection.
Cervical cancer screening participation was categorized as recent (defined as Pap test within the past 3 years) or not (Pap test over 3 years ago or never).
Groups at high risk for cervical cancer were identified by their race and ethnicity, education, income, immigration status, and health care access. Race/ethnicity was categorized as non-Hispanic white, non-Hispanic black, Hispanic, non-Hispanic Asian, and all other. Education was categorized by highest level of school completed (high school graduate or less, some college, and college graduate or more). Family income was categorized according to percentage of the federal poverty line (300% or more, 200% to <300%, 100% to <200%, and <100%). Birth in the United States was represented by a dichotomous variable (yes or no). Health care access was assessed by type of health insurance coverage (uninsured, public/nonmilitary, and private/military) and health care use, measured by whether respondents had seen a physician or obstetrician/gynecologist in the past year.
Receipt of non-HPV vaccines was measured in terms of recommended lifetime vaccinations (ie, receipt of hepatitis A, hepatitis B, and/or tetanus vaccines) and receipt of annual influenza vaccination. We included both variables, because lifetime vaccinations may have been administered when the respondents were minors, under the health care direction of parents or guardians, while the recent decision to receive a flu shot may more closely reflect the respondents' independent decision making regarding vaccination.
The relationship between age and outcomes was also assessed, because younger members of the study sample may have been educated about, or received vaccines, while still in adolescence or in college settings. Respondents were categorized into 2 age groups: 18-20 years and 21-26 years.
We used cross-tabulations to assess the relationship between the outcomes and covariates of interest; for the outcome of HPV vaccine initiation, we conducted multivariate logistic regression. Respondents missing 1 or more covariates were excluded from multivariate models (n = 90). The number of covariates in the final model was constrained by the small sample of vaccinated respondents (n = 179). As a result, we built a parsimonious model that included the independent variables described above, with the exception of: household income and educational status, because these covariates are in flux for women in the study age group; Pap test history because it showed little variability by vaccination status; and birth in the United States, which was highly correlated with Hispanic ethnicity. A multivariate model including separate categories for Hispanic women born in the United States and those born outside the United States did not fit the data significantly better than a simpler model that categorized all Hispanic women together; consequently, we present the simpler model. Analyses were adjusted for the complex survey design using SUDAAN version 9.0.1.
Awareness of HPV Vaccine
Overall, 68.5% of women aged 18-26 years reported ever having heard of the HPV vaccine (Table 1). In bivariate analyses, white women were substantially and significantly more likely than black, Hispanic, or Asian women to have heard of the vaccine (82.5% vs 59.9%, 49.1%, and 45.7%, respectively). Women born in the United States were more than twice as likely as those born outside of the country to have heard of the vaccine (76.7% vs 36.9%). Awareness of the vaccine was also significantly higher among those with private health insurance rather than no insurance or public coverage, and among those who had seen a physician or obstetrician/gynecologist in the past year or received recommended lifetime vaccinations (P<.01). Small differences in awareness by marital status did not reach statistical significance.
Table 1. Sample Characteristics, Awareness, and Use of HPV Vaccine by Sociodemographic, Health Care Access, and Health Behavior Variables (National Health Interview Survey 2008)
Women Aged 18-26 Years (n = 1583)
Ever Heard of HPV Vaccine (n = 1509)
Ever Had HPV Vaccine (n = 1503)
Wgtd % (95% CI)
Wgtd % (95% CI)
Wgtd % (95% CI)
Abbreviation: CI, confidence interval; GED, general equivalency diploma; HPV, human papillomavirus; HS, high school; OB/GYN, obstetrician/gynecologist; Wgtd, weighted.
Overall, 11.7% (95% confidence interval [CI], 9.8-13.8] of women aged 18-26 years reported receiving at least 1 dose of the HPV vaccine (Figure 1). Vaccine initiation decreased significantly with increasing age. Women aged 18-20 years were significantly more likely to be vaccinated than women aged 21-26 years (20.9% vs 7.9%, P = .0000) (Table 1). In bivariate analyses, women who were white, married, born in the United States, covered by any health insurance, or who had visited any physician or an obstetrician/gynecologist in the past year, received 1 or more recommended lifetime vaccines, or received an annual influenza shot were significantly more likely than their counterparts to be vaccinated (P<.01) (Table 1). Recency of Pap testing and history of previous HPV infection or abnormal Pap test results were not significantly associated with HPV vaccine initiation.
In multivariate analyses, younger age, unmarried status, health insurance, influenza shot in the past year, and receipt of recommended lifetime vaccines remained significantly associated with HPV vaccine initiation (Table 2); history of HPV diagnosis or an abnormal Pap test was also significantly associated with initiating the HPV vaccine series in the multivariate model. In the controlled model, disparities in HPV vaccine use were no longer apparent across racial and ethnic groups.
Table 2. Odds Ratios for Initiation of HPV Vaccine from Multivariate Logistic Regression Model (National Health Interview Survey 2008)
Lifetime vaccines include hepatitis A vaccine ever, hepatitis B vaccine ever, tetanus shot in the past 10 years.
Previous exposure to HPV
Ever had abnormal test or been told you have HPV
Future risk of HPV infection (indirect measure)
Subgroups at higher risk of cervical cancer
Seen physician or OB/GYN in past year
Other vaccination behavior
One or more recommended lifetime vaccines*
Influenza shot or nose spray in past year
More than half (53.8% [95% CI, 45.5-61.8]) of those initiating HPV vaccination completed all 3 doses, for a total completion rate of 6.2% [95% CI, 4.9-7.7]. Rates of 3-dose completion varied little by sociodemographics, health care access, or health behaviors.
Interest and Willingness to Pay
Of those women who did not receive the HPV vaccine (n = 1327), 39.6% indicated interest in receiving it. As shown in Table 3, the most common primary reasons for lack of interest in the HPV vaccine were belief that it was not needed (35.9%), not knowing enough about it (17.1%), concerns about safety (12.7%), and not being sexually active (10.3%). Of women not interested in the vaccine, only 1.8% cited expense as the primary reason for their lack of interest. However, among respondents interested in the vaccine, 75.0% reported that they were not willing to pay $360 to $500 (ie, full price). Virtually all (97.8%) of those unwilling to pay full price reported that they would receive the vaccine if it were available for free or at a much lower cost.
Table 3. Reasons for Disinterest in HPV Vaccine Among Unvaccinated Women Aged 18-26 Years (National Health Interview Survey 2008)
Wgtd % (95% CI)
Abbreviations: CI, confidence interval; HPV, human papillomavirus; Wgtd, weighted.
Unvaccinated respondents were asked for their open-ended response regarding one main reason that they were not interested in the HPV vaccine. Responses then were coded into the categories shown.
35.9 (31.2, 40.8)
Don't know enough
17.1 (14.0, 20.8)
Worried about safety
12.7 (9.8, 16.2)
Not sexually active
10.3 (7.3, 14.3)
Doctor did not recommend
5.3 (3.9, 7.2)
3.6 (2.4, 5.5)
Already have HPV
2.7 (1.7, 4.4)
1.8 (0.9, 3.4)
Don't know where to get it
0.2 (0.0, 0.9)
Spouse/Family against it
0.1 (0.0, 0.5)
6.3 (4.5, 8.6)
Our analysis of nationally representative survey data revealed limited HPV vaccine initiation among young adult women in the United States. Previous national estimates reported that 10%-11% of 18- to 26-year-old women had received at least 1 dose of the vaccine by the end of 2007.28, 31 We estimate that, in 2008, 12% of this age group received the vaccine. This minimal increase in vaccine uptake among young adult women stands in contrast to coverage among adolescents aged 13-17 years, which increased substantially, from 25% in 2007 to 37% in 2008 and 44% in 2009, according to estimates from the National Immunization Survey-Teen.32-34
Higher vaccine coverage among adolescents than among young adults may be due in part to differences in recommendations regarding adult vaccination. A previous study identified provider recommendation as a driving factor in HPV vaccination among 19- to 26-year-old women.35 In keeping with American Cancer Society guidelines, some clinicians may be consulting with patients on a case-by-case basis to determine whether HPV vaccination is likely to be beneficial. In addition, public financing of vaccines for uninsured and underinsured minors through the Vaccines for Children program may be helping to boost overall HPV vaccination rates among adolescents, and to reduce disparities in vaccine coverage by insurance, income, and race or ethnicity in this age group.33 Most federal and state vaccine financing programs do not extend to adult women, unless they initiate the HPV vaccine series before turning age 19.36 Thus, in the vaccine's initial 2 years on the market, it is not surprising to observe that young adult women exhibit lower overall vaccination rates than adolescents, and that uninsured young adult women are significantly less likely than privately or publicly insured young women to receive the vaccine. Uninsured women are at high risk for poor cervical cancer screening participation,37 and consequently, for cervical cancer morbidity and mortality.38 Uninsured women would, therefore, greatly benefit from catch-up HPV vaccination, especially if they have not initiated sexual activity. Discretionary use of Section 317 funds to pay for HPV vaccines for uninsured adults, Merck's Vaccine Patient Assistance Program, and state-level efforts to bundle free HPV vaccination with cervical cancer screening services may help to expand vaccine coverage among uninsured young adult women.36, 39 Such efforts may be particularly important given that the majority of women interested in receiving the HPV vaccine are not willing to pay full price, but would be willing to obtain the vaccine free or at much lower cost.
Although HPV vaccine initiation is higher among adolescents than among adults, 2008 NHIS data indicate that the dose completion rate is higher among adults than among adolescent girls (58% vs 41%).40 This high completion rate among young adults may be reflective of the higher motivation of women who seek and complete the vaccine series of their own volition, rather than in response to prompts from parents or school-based programs. Examination of the drivers and barriers to vaccine completion is a priority for future research, as vaccine efficacy is unknown for receipt of fewer than 3 doses.6
We found preliminary evidence that unmarried women (ie, women with a higher risk of future HPV infection through sexual contact with new partners) were much more likely to be vaccinated than married women, despite the fact that married and unmarried young adult women were equally likely to have heard of the HPV vaccine. This finding suggests that some young adult women or health care providers may be appropriately assessing future risk of HPV infection and the potential benefits of HPV vaccination.
Nearly half of respondents who were not interested in the HPV vaccine reported that their lack of interest was because they did not need the vaccine, or were not sexually active. In the absence of information on the sexual history of respondents, these results must be interpreted with caution. It is possible that women who are in long-term, monogamous relationships, or those who are not planning to become sexually active in the foreseeable future, are making accurate assessments of their low future risk of HPV infection, and hence their limited potential benefits from HPV vaccination. Alternatively, however, these stated reasons for lack of interest in the HPV vaccine may indicate that women who have not yet initiated sexual activity do not know that the HPV vaccine is more efficacious if administered before they become sexually active. If this latter scenario is the case, then public education targeted to this age group regarding the benefits of HPV vaccination is warranted.
Although recent cervical cancer screening participation was associated with awareness of the HPV vaccine, neither recent nor ever screening was correlated with likelihood of HPV vaccination. Because the majority of women in our study had been screened recently, or were younger than the age at which clinical guidelines definitively recommend cervical cancer screening initiation, we may not have been able to detect differences between screened and unscreened subgroups of women. However, we did find evidence that women's other vaccination behavior was associated with receipt of the HPV vaccine. Consistent with previous studies, we found that those who had received an influenza vaccine in the past year or had received 1 or more recommended lifetime vaccines (hepatitis A or B and/or tetanus) were substantially more likely than those who had not to receive the HPV vaccine.28, 29
Our study has several limitations. First, unlike the National Immunization Survey, self-reported NHIS data are not validated against provider immunization records. Consequently, there may be response biases; however, given how new and widely discussed the HPV vaccine is, it seems likely that respondents would remember and report their HPV vaccine status accurately. Second, NHIS 2008 did not collect data on sexual history, which limited our ability to evaluate whether HPV vaccines reached subgroups of women likely to benefit from vaccination. The NHIS could be used to examine this issue if questions on age of sexual initiation and timing of vaccine dose administration were included in future rounds of data collection.
Third, because NHIS 2008 data were collected within 2 years of the introduction of HPV vaccine, older members of the cohort under study never had the opportunity to vaccinate as adolescents. Therefore, we do not expect our results to generalize to the HPV vaccination behavior of future groups of young adult women. Future research should monitor whether the patterns of use we observe continue as the vaccine has been on the market for a more extended period. In addition, as data become available, vaccine dose completion rates should be examined to identify factors that promote and/or impede completion. Current estimates of dose completion likely underestimate true completion rates, as insufficient time has elapsed for many of the respondents to have received second and third doses of the vaccine.
HPV vaccine coverage among young adult women was low in 2008 and was largely driven by high rates of vaccination among women 18-20 years of age, who may have been vaccinated as minors. Catch-up was higher among insured than uninsured young adult women. Because uninsured women are at greater risk of cervical cancer morbidity and mortality, they should be a high priority for catch-up vaccination. Coordinated public vaccine financing programs such as Vaccines for Children seem to be effective at promoting vaccine coverage among uninsured adolescents, and may have potential for expanding catch-up vaccination among young adult women who stand to benefit the most from HPV vaccines.
This study was funded in part with federal funds from the National Cancer Institute, National Institutes of Health (Contract No. HHSN261200800001E). The Cancer Control Module of the National Health Interview Survey was supported by the Centers for Disease Control and Prevention, and the National Cancer Institute, National Institutes of Health.