Is There an Association Between Maternal Pap Test Use and Adolescent Human Papillomavirus Vaccination? Implications and Contribution

Abstract

Purpose

To identify the association between mother’s recent receipt of a Pap test and daughter’s uptake and completion of the three-shot human papillomavirus (HPV) vaccination series.

Methods

We used cross-sectional data from the 2008 to 2010 Behavioral Risk Factor Surveillance System from 9 U.S. states and Puerto Rico and logistic regression models to examine the association between mother’s receipt of a Pap test in the past 3 years and daughter’s uptake and completion of the three-shot HPV vaccination series among adolescent girls aged 9-17 years (N = 4,776).

Results

Approximately one-quarter of adolescent girls began the HPV vaccination series, and 13.6% completed the three-shot series. Uptake and completion were more likely among girls whose mothers had obtained a Pap test within the past 3 years—for HPV uptake, odds ratio: 1.342, 95% confidence interval: 1.073-1.692; for HPV completion, odds ratio: 1.904; 95% confidence interval: 1.372-2.721—but the relationship between mother’s recent Pap test and vaccine uptake was explained by the mother’s use of a personal doctor and obtaining a routine physical examination in the past year.

Conclusions

HPV vaccination uptake and completion were more likely among adolescent girls whose mothers obtained a recent Pap test. Interventions designed to educate mothers on the importance of HPV vaccination and to facilitate relationships between physicians and mothers may prove successful at increasing HPV vaccination among adolescent girls.

Implications and Contribution

This study addresses the paucity of research examining the association between mothers’ cervical cancer screening behavior and actual adolescent HPV vaccination. Because of their important role in providing HPV vaccine education and recommendation to vaccinate, the results of this study can be useful to doctors who provide services to women with adolescent daughters. The human papillomavirus (HPV) is now recognized as the major cause of cervical cancer and is the most common sexually transmitted infection in the United States [1]. HPV types 16 and 18 account for 70% of cervical cancer cases [2]. The prevalence of HPV infection among U.S. women aged 14-59 years is estimated to be 27% [3]. The most efficient and cost-effective mechanism for combating cervical cancer is preventing infection. There are two vaccines targeting HPV approved by the Food and Drug Administration for girls and women aged 9-26 years. The bivalent HPV vaccine (Cervarix, GlaxoSmithKline, Philadelphia, PA) prevents HPV types 16 and 18. The quadrivalent vaccine (Gardasil, Merck & Co., Inc., Whitehouse Station, NJ) prevents HPV types 16 and 18, as well as 8 and 11, which account for 90% of genital warts [4]. Both vaccines are administered as a three-dose series and are routinely recommended for the prevention of cervical cancer in preteen and adolescent girls [5].

Cervical cancer rates in the United States have declined in recent years, largely because of screening and early treatment [1]. However, widespread use of the HPV vaccine has been impeded by various barriers, including lack of parental knowledge about the vaccine, lack of physician recommendation, vaccine cost, and parental fear that vaccination may condone early sexual activity^[6][7]. Previous research indicates that vaccine uptake and completion in the United States are at suboptimal levels. Estimates across a wide range of geographic areas and population subgroups in the United States suggest that uptake among adolescent girls ranges from approximately 9% to 37% and completion ranges from approximately 2% to 28%^{[8][9][10][11][12]}.

A number of previous studies have examined predictors of HPV vaccination among adolescent girls. These studies have consistently found that poverty, age, maternal education, health insurance coverage, and adolescent sexual activity are all significantly associated with vaccination^{[8][9][11][13][14]}. Several studies have also indicated that maternal health beliefs, knowledge of sexually transmitted infection, perceptions of disease risk, and mother-daughter communication predict adolescent vaccination acceptance^{[8][15][16][17]} and that many adolescents make decisions about vaccination with their parents^[18][19]. Studies find relatively high levels of acceptability of HPV vaccination among parents, typically motivated by a desire to protect their children’s health^[20][21][22].

Maternal acceptability of vaccination is one of the strongest predictors of adolescent acceptability [21], adolescent vaccination interest, and actual vaccine uptake [16]. Findings from previous research suggest that girls are more likely to discuss topics related to sexuality with their mothers than with their fathers [23], and mothers are more likely than fathers to select their child’s doctor and take children to doctor appointments [24]. Previous research demonstrates that mothers who engage in their own preventive health behaviors are more likely to accept vaccination for their children^[25][26] and that maternal communication about Pap screening is associated with adolescent Pap screening [27]. Further, mother-daughter communication about HPV vaccination is more common among mothers who have been advised by doctors to vaccinate their daughters [19]. However, much less is known about how mothers’ preventive health care practices are related to actual HPV vaccination among girls. Accordingly, the present study focused on the relationship between maternal preventive health care practices and daughters’ vaccination status. Specifically, we examined the extent to which mother’s receipt of a recent Pap test was associated with uptake and completion of HPV vaccination among adolescent girls. We hypothesized that there would be a positive relationship between mother’s receipt of a recent Pap test and daughter’s uptake and completion of the HPV vaccination series. In the only study to examine the relationship between mother’s Pap test use and daughter’s HPV vaccination, Chao et al. found that mothers’ Pap test histories were significantly associated with daughters’ likelihood of vaccination [15]. However, that study was restricted to members of a managed care organization in California. In addition, because individual-level socioeconomic indicators were unavailable, the authors relied on neighborhood-level indicators of educational attainment and household income. We expanded on this previous research using a sample of mother-daughter pairs from various states across the United States and controlling for several individual- and family-level characteristics that may be associated with both mother’s use of Pap test and daughter’s HPV vaccination.

In addition, having a regular doctor or other health care provider and having a routine physical checkup have been identified as the most important predictors of women’s Pap test use^[28][29]. Women with more interactions with health care providers have the opportunity to obtain more information about the link between HPV and cervical cancer and the importance of vaccination [30]. Accordingly, we examined whether these indicators of mother’s access to health care explain the relationship between maternal Pap test use and adolescent HPV vaccination.

Methods

Sample

This cross-sectional study used data from the 2008, 2009, and 2010 Behavioral Risk Factor Surveillance System (BRFSS) [31]. The BRFSS is the world’s largest monthly telephone health survey, tracking adult health behaviors, health care use, and health outcomes from all 50 U.S. states, the District of Columbia, and U.S. territories since 1984. In 2008, questions about HPV vaccination for people aged ≤17 years were included as optional modules for states. One child from each sampled household was randomly selected for participation. The sampled adult within each household answered questions about the child, including his/her relationship to the child, demographic information, and HPV vaccination status. Since the HPV module was introduced in 2008, nine U.S. states (Connecticut, Delaware, Kentucky, Oklahoma, Pennsylvania, Texas, West Virginia, Wisconsin, and Wyoming) and Puerto Rico incorporated the HPV questions into their surveys. The core survey questions and HPV module questions were identical for all 10 geographies, which enabled pooling of the data. More information regarding the BRFSS survey is available at INTER REF http://www.cdc.gov/brfss/technical_infodata/index.htm. A distribution of respondents by state is included in Table 1. Because more than half of our respondents came from Texas, and as we were concerned that the introduction of mandated HPV vaccinations for 11- and 12-year-old girls in Texas in 2007 would bias our results, we ran supplemental analyses where we treated Texas as a fixed effect. Our results were unchanged (models not shown but available on request). We also tested the need for multilevel models and found nonsignificant state-level variance intercepts and intraclass correlation coefficients of <.05; therefore, we elected to present the results of the more parsimonious models. Because we were explicitly interested in the relationship between maternal use of Pap tests and daughter’s HPV vaccination, the analyses were restricted to mother-daughter pairs. We used poststratification weights throughout all analyses to correct for selection, nonresponse, and nontelephone coverage bias. This study was exempted from institutional review board review by the lead author’s institution.

Table 1.

Distribution of sample by state and statewide % HPV uptake, HPV completion, and maternal Pap test

State	Frequency	Percentage (%)	HPV uptake (%)	HPV completion (%)	Maternal Pap test (%)
Connecticut	258	5.4	36.4	26.2	97.1
Delaware	38	.8	24.2	14.3	97.2
Kentucky	159	3.3	34.2	19.7	83.6
Oklahoma	135	2.8	20.0	10.8	83.9
Pennsylvania	911	19.1	34.5	11.2	86.4
Texas	2,748	57.5	23.1	12.3	89.9
West Virginia	119	2.5	26.4	15.9	86.9
Wisconsin	220	4.6	27.2	19.2	100.0
Wyoming	44	.9	33.1	22.0	84.6
Puerto Rico	145	3.0	21.4	12.0	100.0
Total	4,776	100

Outcome variables

Both HPV vaccines licensed by the Food and Drug Administration are recommended for girls beginning at the age of 9 years [32]. Accordingly, we restricted our sample to mothers reporting on girls aged 9-17 years. There were two outcomes of interest for this analysis: HPV vaccination uptake and HPV vaccination completion. HPV vaccination uptake was based on the following question: “A vaccine to prevent the human papilloma virus or HPV infection is available and is called the cervical cancer vaccine, HPV shot, or GARDASIL. Has this child EVER had the HPV vaccination?” The variable was dichotomously coded to indicate whether the girl had ever received the vaccination (yes vs. no). A follow-up question asked, “How many shots did she receive?” Because the HPV vaccine is administered as a three-shot series, the variable was recoded to create a dichotomous variable, indicating whether the girl received all three shots (yes vs. no).

Independent variable

The independent variable was mother’s recent Pap test use, based on the question, “A Pap test is a test for cancer of the cervix. Have you ever had a Pap test?” Women who answered affirmatively were asked, “How long has it been since you had your last Pap test?” with five options ranging from “within the past year” to “5 or more years ago.” The U.S. Preventive Services Task Force recommends that women older than 30 years receive a Pap test every 3 years. Consistent with previous research [15], mother’s Pap test status was coded as a binary variable, indicating whether she had obtained a Pap test within the past 3 years.

Covariates

Based on previous research on predictors of HPV vaccination among adolescent girls^{[6][8][9][11][13]}, regression models were adjusted for the adolescent’s age, mother’s age, mother’s race/ethnicity, mother’s marital status, number of children living in the household, household income, mother’s educational attainment, mother’s employment status, health insurance status, whether the mother reported that there was a time in the past 12 months when she needed to see a doctor but could not because of cost, and metropolitan versus nonmetropolitan residence. To examine the extent to which maternal access to a regular source of care explained the relationship between maternal Pap test use and adolescent vaccination, we introduced dichotomous indicators of whether the mother had a personal doctor and whether the mother had a routine physical checkup in the past year. Although the adolescent’s race/ethnicity was also available, it was strongly correlated with mother’s race/ethnicity; thus, we did not include it in our models. We ran separate models, substituting adolescent race/ethnicity for mother’s race/ethnicity (not shown but available on request), and the results were unchanged.

Data analysis

Descriptive statistics were calculated to assess sample characteristics. Difference of proportions tests were used to identify significant differences in adolescent vaccine uptake and completion between girls whose mothers reported having a Pap test in the past 3 years versus girls whose mothers had not had a Pap test in the past 3 years. Logistic regression models were used to calculate the odds of adolescent HPV uptake and completion (separately). For both dependent variables, the first model included only the independent variable for mother’s recent Pap test. The second model was adjusted for adolescent’s age, mother’s demographic characteristics, socioeconomic status, and family characteristics. In the final model, we added mother’s access to a personal doctor and receipt of a physical examination in the past year. Data were analyzed using SAS version 9.3 (SAS Enterprises, Inc., Cary, NC).

Results

Descriptive statistics for the sample are presented in Table 2. Approximately one-quarter (26.6%) of the sample of 4,776 adolescent girls aged 9-17 years ever started the HPV vaccine, and only 13.6% completed the three-shot series. The overwhelming majority of mothers (90%) have received a Pap test in the past 3 years. It is important to note that because a large proportion of our sample came from Texas, our study sample overrepresents Hispanics. In addition, blacks are underrepresented in our analyses. The implications of this will be discussed at the end of the article. As demonstrated in Figure 1, a significantly greater percentage of girls whose mothers reported receiving a Pap test within the past 3 years began the vaccination and completed the vaccination series compared with girls whose mothers did not receive a Pap test in the past 3 years.

Table 2.

Sample description

Variables	% or mean (SD)
Adolescent ever had HPV vaccination	26.6
Adolescent completed HPV vaccination series	13.6
Independent variable
Mother had Pap test in past 3 years	90.0
Control variable
Adolescent age	13.3 (2.58)
Characteristics of mothers
Age	41.3 (9.66)
Race/ethnicity
White	60.7
Black	8.7
Hispanic	26.6
Other race	4.0
Married	70.6
Number of children in household	1.9 (1.06)
Lives in metropolitan area	82.9
Socioeconomic status
Household income
<$25,000	26.1
$25,000–$49,999	22.8
≥$50,000	51.1
Education
<High school	11.7
High school graduate	54.8
Four-year college graduate	33.5
Employed	62.8
Has health insurance	79.9
Experienced medical cost barrier in the past year	23.3
Mother’s health characteristics
Has personal doctor	82.9
Had a routine physical checkup in past year	80.9

Figure 1.

HPV vaccine uptake and completion for adolescents whose mothers had a Pap test in the past 3 years versus adolescents whose mothers did not have a Pap test in the past 3 years. Note: ttests for difference of proportions; *** p < .001 (two-tailed); weight percentages.

HPV vaccine uptake and completion for adolescents whose mothers had a Pap test in the past 3 years versus adolescents whose mothers did not have a Pap test in the past 3 years. Note: t tests for difference of proportions; *** p < .001 (two-tailed); weight percentages.

Results of the regression analyses predicting HPV uptake are presented in Table 3. Results of the unadjusted model (model 1) indicate that adolescents whose mothers had a Pap test within the past 3 years were significantly more likely to begin vaccination than girls whose mothers did not have a Pap test within the past 3 years (p < .05). These results held after controlling for adolescent’s age, mother’s demographic characteristics, and socioeconomic status (model 2). The introduction of mother’s access to a usual source of care (personal doctor and routine checkup in past year) in model 3 eliminated the significance of mother’s Pap test, suggesting that mother’s access to and use of a regular source of health care explains the relationship between maternal Pap test use and daughter’s likelihood of HPV vaccine uptake. Indeed, girls whose mothers had received a physical checkup in the past year had significantly greater odds of beginning the HPV vaccination series than those girls whose mothers had not had a checkup in the past year. Adolescent age, number of children in the household, and health insurance status were positively associated with HPV vaccine uptake. The mother’s experience with a medical cost barrier during the past year was also positively associated with daughter’s vaccine uptake.

Table 3.

ORs for the association between mother’s recent Pap test (within the past 3 years) and adolescent completion of HPV series

Variables	Model 1		Model 2		Model 3
	OR	95% CI	OR	95% CI	OR	95% CI
Mother had Pap test within the past  3 years	1.342 □	1.073– 1.692	1.411 □□	1.106– 1.814	1.258	.978– 1.628
Control variables
Adolescent age			1.295 □□□	1.259– 1.333	1.299 □□□	1.262– 1.337
Characteristics of mothers
Age			1.002	.994– 1.010	1.000	.993– 1.008
Race/ethnicity
White (reference)
Black			1.036	.800– 1.334	1.021	.789– 1.316
Hispanic			1.082	.895– 1.305	1.082	.893– 1.308
Other race			.781	.529– 1.128	.810	.548– 1.172
Married			.855	.725– 1.009	.848	.719– 1.002
Number of children in household			1.077 □	1.007– 1.152	1.085 □	1.014– 1.160
Lives in metropolitan area			1.117	.928– 1.349	1.130	.938– 1.365
Socioeconomic status
Household income
<$25,000			1.181	.947– 1.471	1.156	.926– 1.442
$25,000-$49,999			.901	.743– 1.092	.902	.743– 1.093
≥$50,000 (reference)
Education
<High school			.825	.623-– 1.090	.849	.639-– 1.124
High school graduate			.945	.803– 1.114	.946	.803– 1.114
Four-year college graduate  (reference)
Employed			.952	.823– 1.103	.958	.827– 1.110
Has health insurance			1.605 □□□	1.298– 1.991	1.417 □□	1.132– 1.778
Experienced medical cost barrier in  past year			1.386 □□□	1.151– 1.667	1.468 □□□	1.217– 1.770
Mother’s health characteristics
Has personal doctor					1.229	.989– 1.531
Had a routine physical checkup in  past year					1.408 □□	1.158– 1.720

^□p < .05.

^□□p < .01.

^□□□p < .001 (two-tailed).

Table 4 presents the results from the regression analyses predicting HPV vaccination completion. Again, results of the unadjusted model indicate that adolescent girls whose mothers reported having a Pap test within the past 3 years were significantly more likely to have completed the three-shot vaccine series than girls whose mothers had not had a Pap test within the past 3 years (p < .001). These results held after adjusting for adolescent age and mother’s demographic, socioeconomic, and family characteristics (model 2). In model 3, having a personal doctor and having a physical examination were both significantly associated with daughter’s completion of the HPV vaccination series. However, these additional variables only partially explained the positive association between mother’s Pap test and adolescent’s completion of the three-shot series. Adolescent age and number of children living in the household were also positively associated with vaccine completion. Adolescents whose mothers identified as “Other Race” versus white and whose mothers had less than a high school diploma versus a college degree were significantly less likely to complete the vaccination series.

Table 4.

OR for the association between mother’s recent Pap test (within the past 3 years) and adolescent completion of HPV series

Variables	HPV vaccination series completion
	OR	95% CI	OR	95% CI	OR	95% CI
Mother had Pap test within past 3  years	1.904 □□□	111 1.372– 2.721	2 117 □□□	1.489– 3.089	1.872 □□	1.305– 2.751
Control variables
Adolescent age			1.332 □□□	1.282– 1.385	1.337 □□□	1.286– 1.391
Characteristics of mothers
Age			1.018 □□□	1.008– 1.028	1.016 □□	1.006– 1.026
Race/ethnicity
White (reference)
Black			1.209	.875– 1.653	1.210	.875– 1.655
Hispanic			1.087	.849– 1.386	1.106	.863– 1.412
Other race			.472 □	.246-828	.494 □	.257-.869
Married			.996	.804– 1.238	.982	.792– 1.220
Number of children in household			1.119 □□	1.028– 1.217	1.131 □□	1.039– 1.230
Lives in metropolitan area			.970	.767– 1.236	.985	.779– 1.256
Socioeconomic status
Household income
<$25,000			1.183	.888– 1.574	1.148	.861– 1.530
$25,000-$49,999			.828	.640– 1.066	.824	.637– 1.063
≥$50,000 (reference)
Education
<High school			.574 □□	.394-.829	.604 □□	.413-.874
High school graduate			.824	.671– 1.013	.824	.670– 1.014
Four-year college graduate–  (reference)
Employed			.961	.795– 1.164	.967	.799– 1.171
Has health insurance			1.168	.887– 1.547	.965	.720– 1.300
Experienced medical cost barrier–  in past year			1.101	.859– 1.404	1.178	.917– 1.507
Mother’s health characteristics
Has personal doctor					1.500 □□	1.115– 2.037
Had a routine physical checkup in–  past year					1.438 □□	1.107– 1.887

^□p < .05.

^□□p < .01.

^□□□p < .001 (two-tailed).

Discussion

Because parental consent is generally required to provide the HPV vaccine to adolescents, mothers play an essential role in facilitating vaccine uptake among their daughters. Further, because mothers shoulder the majority of responsibility for children’s involvement in the health care system^[23][24], girls’ completion of the three-shot series is highly contingent on mothers’ desire, motivation, and ability to get their daughters to the necessary follow-up visits. Our finding that 27% of adolescent girls aged 9-17 years had initiated HPV vaccination is consistent with previous research^[14][30][33]. Our hypothesis that mother’s recent Pap test receipt would be associated with daughter’s uptake and completion of the HPV vaccination series was supported by the study results. Consistent with Chao et al.’s[15] findings, adolescents whose mothers received a Pap test within the past 3 years were significantly more likely to initiate and complete the HPV vaccination series. We found this to be true regardless of age, mother’s race/ethnicity, socioeconomic status, health insurance coverage, family characteristics, and metropolitan versus nonmetropolitan residence. Our study extended that of Chao et al[15] by also examining the extent to which access to and use of a regular source of care explains the relationship between maternal Pap test use and adolescent HPV vaccination.

Previous research indicates that HPV vaccination is more likely when it is recommended by a health care provider^[19][30] and that regular contact with one’s doctor is the most important predictor of women’s Pap test use^[28][29]. We found that mother’s access to a personal doctor and her receipt of a physical examination within the past year fully explained the relationship between maternal Pap test use and adolescent HPV vaccine uptake and partially explained the relationship between maternal Pap test use and adolescent HPV vaccine completion. This suggests that the mechanisms determining HPV vaccine completion are slightly different from those determining vaccine uptake. HPV vaccine initiation requires only one visit to a health care provider, and initiation may begin during an annual well-child visit or at the time of the girl’s first annual vaginal examination. However, finding the time to transport the child to the clinic or hospital for the next two shots, which are to take place within 213 days of the first dose [10], may be more difficult. The time barrier may be especially pronounced for single mothers, mothers who work full time or multiple jobs, and for families who live far from their health care providers. Increasing opportunities for vaccine delivery can help to increase HPV vaccine completion rates [10]. In addition, scientific advances toward a vaccine that could be administered in one dose would likely reduce vaccine noncompletion rates.

Although we found no maternal educational differences in daughters’ vaccine uptake, there was a negative association between mother’s low educational attainment and vaccine completion. Low education may be a marker of low health literacy. Although researchers have generally found that education is not a predictor of parental acceptance of the HPV vaccine^[17][25] or that parents with lower levels of education actually have higher HPV vaccine acceptability for their children^{[6][20][22][34]}, the perceived importance of receiving all three shots may be weaker among mothers with lower educational attainment. If this is the case, educational campaigns targeted toward mothers without a high school diploma should emphasize the risks associated with incomplete vaccination [10]. In addition, mothers with lower levels of education may face some of the time barriers discussed earlier, especially if they have more difficulty taking time off work to take their daughters to obtain the follow-up shots.

Interestingly, we found that experiencing a medical cost barrier in the past year was positively associated with daughter’s HPV vaccine initiation. This may be related to type of health insurance coverage. Having public health insurance, such as Medicaid, has been found to be positively associated with HPV vaccine initiation [14]. The ability to obtain the vaccine at minimal or no cost may encourage mothers with health care cost barriers to have their daughters vaccinated [8]. Other possible explanations may be related to causality issues that we were unable to address with these cross-sectional data. For example, because the survey question asks about a time in the past 12 months when the mother needed to see a doctor but could not because of cost, it may be that the vaccination occurred before the medical cost barriers existed for the mothers. It may also be that cost barriers are more pronounced among mothers who use health care services more frequently. That is, the more one seeks out and uses medical care or pays for necessary medical care for their children, the more cost becomes a perceived barrier. Future research should track maternal health prevention behaviors, family socioeconomic status, and daughter’s reproductive health behaviors over time in an effort to gain a life-course perspective on the relationship between mothers’ and daughters’ reproductive health activities.

Health beliefs and knowledge are widely known to influence health behaviors [35]. Previous research suggests that knowledge about HPV and its relationship to cervical cancer increases parental acceptability of vaccination^[21][34][36]. Although we are unable to directly link maternal knowledge about HPV and acceptance of the HPV vaccine to daughters’ actual uptake and completion of the vaccine series, our findings suggest that more frequent interactions between mothers and health care providers who promote the importance of regular screening and prevention can facilitate the decision to vaccinate their daughters. This suggests that policy interventions aimed at encouraging mothers to find a “health care home” and establish long-term and trusting relationships with their health care providers have tremendous potential to increase HPV vaccination uptake and completion for adolescent girls.

The results of this study should be considered in light of some limitations, including the potential for maternal recall bias about length of time since last Pap test and about daughters’ completion of the HPV vaccine and the inability to include variables that have been identified in previous research as important predictors of vaccination, such as provider recommendation, adolescent sexual history, type of health insurance coverage, and vaccine cost^[6][7][26]. We were also unable to examine the impact of neighborhood socioeconomic status, racial composition, spatial access to health care, and other contextual characteristics that may be associated with adolescent vaccination [37]. In addition, our sample overrepresents Hispanics and underrepresents blacks. Our finding that there were no significant black-white or Hispanic-white differences in HPV vaccine uptake and completion is consistent with recent nationally representative studies on girls in the same age range [8]. Accordingly, we are confident that the racial/ethnic distribution of respondents in this study did not bias our results. Finally, the rate of 90% Pap test completion among the mothers in our sample is high relative to other studies [38]. We believe that this is because our sample comprised mothers with adolescent children living in the household (aged 9- 17 years). This means that our sample overrepresents young- to middle-aged mothers who are more likely to obtain Pap tests than older women and women without children [39]. Despite these limitations, our study has several important strengths, including using measures of actual vaccine uptake and completion rather than intent to vaccinate or approval of vaccination, individual-level controls for socioeconomic status and family characteristics, and use of population-based data from a diverse range of states across the United States.

The most effective way to combat cervical cancer mortality is to prevent infection. Results of our study suggest that increasing the use of Pap tests among mothers, particularly among groups with low health literacy, and facilitating relationships and regular contact between mothers and health care providers may have the secondary benefit of increasing HPV vaccination among adolescent girls. Mothers of adolescent girls are essential targets for public health efforts to increase HPV vaccination. Physicians who come into contact with mothers of adolescent girls should educate those mothers about cervical cancer risk and the role of the HPV vaccine in preventing cervical cancer and should encourage these women to have their daughters vaccinated.