Significance

What is already known on this subject: HPV vaccine coverage of female adolescents in the IW is poor. A recent regional assessment demonstrated that older teen age, younger parent age, and receipt of other recommended vaccinations were significantly associated with HPV vaccination. This assessment informed regional intervention strategies.

What this study adds: This study demonstrates sub-regional differences in factors associated with HPV vaccination: health care system level factors are significant in the East sub-region; family structure is significant in the Central sub-region; and race/ethnicity is significant in the West-sub-region. Regional HPV vaccination interventions could be improved by considering these sub-regional findings.

Introduction

In 2006, the Centers for Disease Control and Prevention (CDC) recommended a 3-dose human papillomavirus (HPV) vaccine as a cancer prevention strategy for women aged 11–26 years old (Markowitz et al. 2007). A half-decade later, HPV vaccination rates remain low (“National and state vaccination coverage,” 2011; “National and state vaccination coverage,” 2012). Uptake has been particularly poor in the Intermountain West (IW), a region comprised of Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah, and Wyoming. Among adolescent females, in both 2011 and 2012, three of the eight IW states had HPV vaccine initiation rates below the national average, and five had completion rates below the national average (“National and state vaccination coverage,” 2011; “National and state vaccination coverage,” 2012).

The sociodemographic characteristics of the IW region may contribute to these low vaccination rates. Several of the IW states are among the most rural in the US (US Census Bureau 2015a), so individuals may have limited access to health care due to the greater geographic distance from providers (Arcury et al. 2005). Additionally, the IW features a large religious population—specifically a high density of Mormons (Pew Research Center 2014)—and religious beliefs regulating the sexual activity of unmarried women may pose challenges to HPV vaccination (Constantine and Jerman 2007; Zimet et al. 2008). Finally, the IW also has a rapidly growing minority and immigrant population that may lack well-established community and institutional support necessary for navigating the health care system (Grieco et al. 2012).

In spite of their shared characteristics when compared to other regions of the US, the IW states also feature notable within-region variation. For example, the rural population ranges from 5.8% in Nevada to 44.11% in Montana (US Census Bureau 2015a); the Mormon population ranges from 2% in Colorado and New Mexico to 55% in Utah (Pew Research Center 2014); and the Hispanic population ranges from 3.5% in Montana to 47.7% in New Mexico (US Census Bureau 2015b). Furthermore, though the HPV vaccination rates in the IW states are among the lowest in the nation, they too vary considerably. In 2011, HPV vaccine initiation among adolescent females ranged from 45.5% in Idaho to 60.9% in Wyoming, and completion ranged from 25.3% in Colorado to 40.9% in Wyoming (“National and state vaccination coverage,” 2011). In 2012, initiation ranged from 44.3% in Utah to 62.5% in Nevada, and completion ranged from 39.0% in Utah to 46.5% in Montana (“National and state vaccination coverage,” 2012).

This substantial within-region variation suggests that efforts to improve HPV vaccination in the IW region may be improved by considering sub-regional differences. Therefore, building on an assessment of the entire IW region (Lai et al. 2016), we use national survey data to clarify similarities and tease apart differences in the factors related to adolescent females’ HPV vaccine initiation and completion in three sub-regions of the IW.

Methods

Data from the 2011 and 2012 National Immunization Survey-Teen (NIS-Teen) was used. The 2011 and 2012 NIS-Teen surveys occurred January 2011-April 2012 and January 2012-April 2013, respectively (NORC at the University of Chicago [NORC] 2012, 2013). Analysis of the NIS-Teen data was conducted September 2014-March 2015. Analysis of publicly available data is considered exempt by the University of Utah Institutional Review Board.

Participants

The NIS-Teen is a national survey that monitors adolescent vaccination coverage. It consists of a phone survey of parents and legal guardians to collect immunization information about adolescents, followed by a mailed survey of providers to validate the adolescents’ vaccination records (NORC 2012, 2013). For the present analyses, we included respondents (hereafter referred to as parents) to the 2011 and 2012 NIS-Teen surveys who were living in the IW and who had daughters aged 13–17 years old with provider-verified immunization records.

Sub-regions

We defined three sub-regions of the IW based on geographical contiguity, similarity in HPV vaccine initiation and completion rates in the 2011 and 2012 NIS-Teen surveys, and similarity in sociodemographic characteristic (e.g., having large religious populations). The East sub-region includes Colorado, Montana, and Wyoming; the Central sub-region includes Idaho and Utah; and the West sub-region includes Arizona, New Mexico, and Nevada.

Measures

The outcomes of interest were HPV vaccine initiation and completion, defined by provider-verified receipt of at least one dose or of three doses of the HPV vaccine, respectively. Independent variables were chosen a priori based on existing literature on the correlates of HPV vaccine initiation and completion among adolescent females in the US. Independent variables related to teens included age, race/ethnicity, and receipt of other recommended adolescent vaccinations (influenza, TDAP, and meningitis). Independent variables related to parents included age, marital status, educational attainment, and poverty status. Independent variables related to the health care system included source of health insurance, type of provider practice, and providers’ use of state or local health departments to obtain vaccines.

Statistical Analyses

To combine NIS-Teen data from 2011 to 2012, we used the survey weighting methodology provided by the CDC (NORC 2013). Actual frequency and survey-weighted percentages were reported for categorical variables, and comparisons between teens who did and did not initiate or complete the HPV vaccine were evaluated using a survey-weighted Pearson Chi square test. Multivariable weighted Poisson regression models with robust standard errors were used to estimate adjusted prevalence ratios (PRs) and 95% confidence intervals (CIs) for selected factors. For all tests, the null hypothesis was that no significant differences existed between groups, and statistical significance was defined as p < 0.05; all p values represent two-sided comparisons. Data management was done with SAS Statistical Software version 9.3, and statistical analyses were performed using Stata version 13.1.

Results

HPV Vaccine Initiation

Bivariate analyses (Table 1) revealed that receipt of other recommended adolescent vaccinations was associated with HPV vaccine initiation in all three sub-regions (all p < 0.05). In the East sub-region, parent and teen age were both related to HPV vaccine initiation (both p < 0.02). In the Central sub-region, parents’ marital status was associated with HPV vaccine initiation (p = 0.002). In the West sub-region, parent age (p = 0.002), poverty status (p = 0.049), and teen race/ethnicity (p = 0.014) were all associated with HPV vaccine initiation.

Table 1 Factors associated with HPV vaccine initiation among female adolescents in three sub-regions of the Intermountain West
Full size table

Poisson regression analyses (Table 2) showed that when adjusting for other variables, older teen age and receipt of other recommended adolescent vaccinations were significantly associated with higher HPV vaccine initiation in all three sub-regions (see Table 2). In the East sub-region, providers’ facility type and providers’ use of state or local health departments to obtain vaccines were significantly related to HPV vaccine initiation (see Table 2). In the Central sub-region, non-Hispanic White teens and teens with married parents were significantly less likely to have initiated the HPV vaccination series compared to Hispanic teens and teens with unmarried parents, respectively (see Table 2). In both the East and West sub-regions, younger parent age and lower educational attainment were significantly associated with HPV vaccine initiation (see Table 2). In both of these sub-regions, poverty status was also significantly associated with HPV vaccine initiation, though the direction of the effect was different for each (see Table 2).

Table 2 Multivariable analyses for HPV vaccine initiation among female adolescents in three sub-regions of the Intermountain West
Full size table

HPV Vaccine Completion

In bivariate analyses (Table 3), receipt of influenza or meningitis vaccination was associated with HPV vaccine completion in all three sub-regions (all p < 0.05). In the Central and West sub-regions, receipt of the TDAP vaccination was associated with HPV vaccine completion (both p < 0.01). In the East sub-region, parent and teen age (both p < 0.05), poverty status (p = 0.005), and parents’ marital status (p = 0.025) were related to HPV vaccine completion. In the Central sub-region, parents’ marital status (p = 0.003) and provider facility type (p = 0.015) were related to HPV vaccine completion. In the West sub-region, teen race/ethnicity (p = 0.011) and age (p = 0.012) were associated with HPV vaccine completion.

Table 3 Factors associated with HPV vaccine completion among female adolescents in three sub-regions of the Intermountain West
Full size table

Poisson regression analyses (Table 4) demonstrated that when adjusting for other variables, receiving other recommended adolescent vaccinations was significantly associated with HPV vaccine completion in all three sub-regions (see Table 4). In both the East and West sub-regions, older teen age was significantly associated with higher HPV vaccine completion (see Table 4). In the Central sub-region, teens with unmarried parents were significantly more likely to complete the HPV vaccination series than were teens with married parents (see Table 4). In the West sub-region, Hispanic teens were significantly more likely to complete the HPV vaccination series than were non-Hispanic teens (see Table 4). Certain health care system-level factors were significantly related to HPV vaccine completion: in the East sub-region, providers’ use of state or local health departments to obtain vaccines; in the Central sub-region, providers’ facility type (see Table 4).

Table 4 Multivariable analyses for HPV vaccine completion among female adolescents in three sub-regions of the Intermountain West
Full size table

Discussion

Low HPV vaccination rates in the IW present a serious public health challenge for the region (“National and state vaccination coverage,” 2011; “National and state vaccination coverage,” 2012). Identifying factors that influence HPV vaccination practices regionally and sub-regionally is critical to improving HPV vaccine coverage in the IW. This study builds on an earlier assessment of the entire IW region (Lai et al. 2016) to assess the factors related to female adolescents’ HPV vaccine initiation and completion in three sub-regions of the IW. Results highlight opportunities for improvement and will inform intervention strategies for the region. Although from the time of data collection, the vaccination schedule for younger adolescents has changed to require only two doses of the HPV vaccine (Meites et al. 2016), the variables related to both initiation and completion that are discussed in this paper continue to be relevant.

Our analyses yielded two findings that support the results of the region-wide examination (Lai et al. 2016). First, in all three sub-regions, receipt of other recommended adolescent vaccinations was significantly associated with HPV vaccine initiation and completion. This finding underscores the importance of reducing missed opportunities for HPV vaccination by administering the HPV vaccine along with other adolescent vaccinations. It also highlights the opportunity to improve HPV vaccination rates through promotion of the HPV vaccine along with other adolescent vaccinations via parental education and provider training and reminders (e.g., Mayne et al. 2014). Second, older teen age was significantly related to HPV vaccine initiation and completion in all three sub-regions. This finding may reflect parent and/or provider reluctance to vaccinate younger adolescents against a sexually transmitted virus. Interventions should therefore work to educate parents and providers about the rationale for vaccinating younger adolescents (e.g., Mayne et al. 2014; PATH 2008), to leverage the school system to reach younger adolescents (e.g., Eldred et al. 2015), and to use HPV vaccine delivery as the basis for comprehensive adolescent reproductive health services (e.g., Pollack et al. 2007).

Supplementing these all-region findings, four important sub-regional differences emerged in our analyses. First, whereas the region-wide investigation indicated that teens with older parents were less likely to be vaccinated than teens with younger parents (Lai et al. 2016), our study showed this finding to hold for the East and West sub-regions only, and with regards to HPV vaccine initiation only. Therefore, the recommendation to focus educational and persuasive messaging to older parents (Lai et al. 2016) may be less effective in the Central sub-region and should be employed only in the East and West sub-regions. Second, health care system-level factors were associated with HPV vaccine initiation and completion predominantly in the East sub-region. Though the processes underlying this finding merit further investigation, it is clear that intervention efforts in the East sub-region should operate on the health care system level in addition to the patient level (e.g., CDC 2015; Perkins et al. 2015). Third, in the Central sub-region, parents’ marital status was significantly related to HPV vaccine initiation and completion, with teens with married parents less likely to be vaccinated than teens whose parents were unmarried. Although our data cannot unequivocally establish the causes of this finding, it may be that unmarried couples are more understanding of the vagaries of sexual activity, and thus more inclined to vaccinate their daughters against a sexually transmitted virus. This finding suggests that interventions must be sensitive to family structure in this sub-region in order to reach married parents (e.g., Gerend et al. 2013). Finally, teen race/ethnicity was significantly associated with HPV vaccine initiation and completion in the West sub-region, with Hispanic teens more likely to have completed the vaccine than non-Hispanic teens. This finding may reflect different levels of support for vaccines among various racial/ethnic groups, and underscores the need to target non-Hispanic racial/ethnic groups—particularly whites—with HPV vaccine intervention materials in this sub-region (e.g., Lechuga et al. 2011).

In addition to highlighting these opportunities for improvement, this study also contributes to a large body of literature concerning HPV vaccination in the US (Fisher et al. 2013; Holman et al. 2014; Kessels et al. 2012; Rambout et al. 2014), and to research using data from the NIS-Teen (Rahman et al. 2014). To the best of our knowledge, this is the first study to use national survey data to study sub-regional similarities and differences related to HPV vaccination in the US.

Limitations

This study is limited by the methodology of the NIS-Teen. Bias related to the exclusion of households with non-response or without any phones is a possibility. Additionally, adequate provider-verified vaccination data was available for only 54.6 and 56.38% of cellular and 61.5 and 61.97% of landline respondents in 2011 and 2012, respectively (NORC 2012, 2013). Also, this study does not address geographical, political, or sociodemographic variations among individual states within the sub-regions studied. Moreover, comparisons drawn between this study and the prior regional examination are limited by differences in their samples: the regional investigation assesses data from the 2012 NIS-Teen, whereas this investigation relies on NIS-Teen data from 2011, as well.

Conclusions

Identifying the regional and sub-regional factors that influence HPV vaccination is critical to improving HPV vaccine coverage among female adolescents in the IW. This study presents region-wide intervention opportunities and highlights a number of strategies for tailoring efforts to address unique sub-regional needs. Regionally, interventions should promote the HPV vaccine along with other recommended adolescent vaccinations and should focus efforts to vaccinate younger girls. Sub-regionally, interventions should work on the health care system level in the East sub-region, reach out to married couples in the Central sub-region, and focus on non-Hispanics in the West sub-region.