North Carolina parents are more likely to get their teenage daughters started on the three-dose human papillomavirus (HPV) vaccination if a doctor has recommended it than otherwise; the greater their perception of barriers to vaccination, the less likely they are to do so.1 Other predictors of vaccine initiation, according to results of a 2007 telephone survey, are parents’ belief about whether their daughter’s health insurance covers it and perceptions of the vaccine’s potential harms. Predictors of vaccine initiation appear to be similar for whites and blacks, and for urban and rural residents, but their relative importance may vary among groups.

Although several studies have examined correlates of HPV vaccine acceptability, few have explored predictors of initiation. To help fill that gap, researchers surveyed parents of 10–18-year-old females living in areas of North Carolina with high rates of invasive cervical cancer. In all, 889 parents completed the survey, which asked about a variety of beliefs and situational factors that may be linked to parents’ decision about whether to get their teenage daughter vaccinated against HPV.

The majority of respondents were female (94%), white (70%), married (84%) and aged 40 or older (72%). Four in five had at least a college education, and close to two-thirds reported that their annual household income was $50,000 or more; the sample was about evenly split between urban and rural residents. Twenty-eight percent had a 10–12-year-old daughter, 33% a 13–15-year-old and 39% a 16–18-year-old. Only 12% said that their daughter had had at least one HPV vaccine injection.

Most parents considered the vaccine at least moderately effective, cited some barriers to obtaining it and thought that it could cause some harm (mean scores, 2.5, 1.6 and 2.1, respectively, on scales of 1–4). Twenty-nine percent said that they had had cervical cancer, and 18% reported that they or someone important to them had had genital warts; 22% said that a doctor had recommended getting their daughter vaccinated.

The researchers used bivariate logistic regression to identify potential predictors of vaccine initiation. They then conducted multivariate analyses that controlled for parents’ demographic characteristics, daughters’ age and all situational variables that were significant in the bivariate analysis. At the multivariate level, the odds that a teenager had had at least one injection were sharply elevated if a doctor had recommended it (odds ratio, 5.6); the odds also were higher if parents believed that the daughter’s health insurance covered the vaccination or if they thought that it did not than if they answered “maybe” or “don’t know” when asked about coverage (16.2 and 10.0, respectively). The likelihood of vaccine initiation was inversely associated with parents’ perceptions of barriers to obtaining the vaccine (0.3), harms it may cause (0.1) and their daughter’s risk of getting cervical cancer (0.2).

In exploratory subgroup analyses, results for white parents were similar to those for the sample overall. One additional variable was significant for whites: The more effective parents considered the vaccine, the greater the likelihood that their daughter had had at least one dose (odds ratio, 3.4). Analysis of data from black parents was limited by small numbers, but the results suggest that for this group, too, a doctor’s recommendation to have a teenage daughter vaccinated is a key predictor of initiation and that perceptions of barriers to or potential harm from the vaccine are related to reduced odds of initiation. (The insurance variable was omitted from the analysis for blacks.)

For both urban and rural parents, having had a doctor recommend HPV vaccination strongly predicted that a teenage daughter had had at least one injection (odds ratios, 4.0 and 10.3, respectively), as did believing that the teenager’s insurance covered the vaccine (14.3 and 16.8). Likewise, the odds of injection fell for both as perceptions of harm increased (0.1 for each group) and as perceptions of the daughter’s likelihood of getting cervical cancer increased (0.2 for each). Barriers to getting the vaccine were associated with a reduced risk of teenagers’ having done so in urban areas only (0.3).

The researchers point out that their study has limited generalizability and that the subgroup analyses must be considered “tentative”; they also acknowledge that they identified correlates only of vaccine initiation and not of completion of the regimen. Nevertheless, they interpret the findings as evidence that parents’ beliefs that are linked to teenage daughters’ receipt of the HPV vaccine are “potentially modifiable targets” for efforts to increase vaccination rates. Moreover, according to the investigators, the subgroup findings “suggest that interventions can target many of the same beliefs across subgroups but also may need to be slightly tailored to maximize effectiveness.”

D. Hollander


1. Reiter PL et al., Parents’ health beliefs and HPV vaccination of their adolescent daughters, Social Science & Medicine, 2009, 69(3):475–480.