We began by reviewing concepts from health behavior and message design theories. Next, we conducted two phases of focus groups with parents to identify their health beliefs about HPV infection in their sons and to solicit feedback on message designs and social marketing best practices; we revised initial message designs according to this feedback. Finally, we conducted intercept interviews with parents individually to document reactions to message designs. We defined parents as mothers, fathers, grandparents and any other individuals who said they were responsible for boys’ care. The institutional review board at the University of North Carolina at Chapel Hill approved the study.
The focus group and intercept interview discussion guides drew from the health belief model,23
which outlines factors that influence perceptions of risk associated with a disease. We first explored parents’ perceptions of susceptibility to HPV; severity of STDs and HPV-related cancers; benefits of vaccination to prevent HPV-related disease; and possible barriers, including concerns about safety, efficacy, cost and access. To explore parents’ ability to have their sons vaccinated (i.e., accessing or affording health care services), we asked questions about cues to action (external events that prompt a desire to make a health change) and self-efficacy (individuals’ belief in their ability to make such change).24
We used message design theories of gain framing25
(highlighting positive outcomes of behavior compliance and thereby avoiding a consequence) and emotional relevance26
(arousal leading to action) to ensure the our approach was persuasive enough to encourage HPV vaccination.27
We designed and tested variations of both text and images to determine which combination seemed most relevant to parents.
After creating preliminary message designs, we included questions derived from social marketing principles that measured how parents perceived the value of having their sons vaccinated.28,29
Questions covered the value of the recommended HPV vaccine, its price (in terms of cost, safety and efficacy), promotion (posters, brochures, doctors’ recommendations, radio public service announcements) and where to get it (doctors’ offices, pharmacies).
For the focus groups, we targeted black parents, since they are less aware of HPV or the HPV vaccine than are white parents,30
and since blacks are at disproportionately higher risk for STDs and HPV-related cancers.6
We focused on 11–12-year-old boys because HPV vaccine is believed to be most effective in this presumably sexually uninitiated age-group. We selected south central North Carolina because the region’s rates of cervical cancer and STDs exceed those of the state overall.31,32
Participants were recruited through flyers and announcements at the focus group locations—three churches and a middle school in rural Sampson County. The county director of public health helped to identify potential sites for the focus groups and introduced us to key local leaders. The churches had black congregants, and the pastors announced the focus groups from the pulpit on Sundays and encouraged parents to attend. The middle school offered a program on parenting black youth one Saturday morning, and with endorsement from the principal, we recruited parents for a focus group after the program. Staff members who recruited participants were black. We conducted five focus groups with 29 parents from August 2009 to February 2010. Parents received a $35 gift card for participating in the 90-minute discussion. Discussions were audiotaped and transcribed verbatim.
The first two focus groups, comprising 16 parents, met in a church on a Wednesday evening and Saturday morning, and primarily discussed health beliefs. For example, we asked about susceptibility to HPV (“Do you think boys are at risk for HPV infection?” “Any reasons for this?”) and severity of HPV infection (“Do you think parents see their sons at risk for genital area cancers, such as in their penis?”). We asked parents their thoughts on the benefits of HPV vaccine (“Do you think it is a good idea for an 11–12-year-old boy to get the HPV vaccine?”) and possible barriers to getting their sons vaccinated (“Do you know where to go?” “How much does it cost?”). We asked whether fathers would be involved in their sons’ HPV vaccination decision (“Who in your family generally makes decisions about vaccinating your son?” “Is it generally the mother or the father?” “How would you say you go about making decisions about getting vaccines for your son?”). We also asked about trusted sources of information (“Who do you think fathers/mothers trust for information about the HPV vaccine?” “Any reasons for this?”). At the end of the focus groups, we asked participants to write down a message that they thought would persuade parents to get their sons vaccinated.
We analyzed transcripts from the first two focus groups for common themes related to the health belief model that could inform message development. From these themes, we drafted nine headline messages and developed accompanying images for posters. We asked the last three focus groups, comprising 13 parents, for feedback on these preliminary message design concepts (see box).
Each message was illustrated on a poster with a different image. Eight designs featured a preteenage boy; they varied by the race of people depicted (black only and black with other ethnicities), whether they showed one or both parents, whether they included other people (friends, future wife), setting (outdoors, church, retail) and color scheme (blue and green, brown and yellow, or black). One design featured only an influential figure (doctor, nurse, clergy) prominently advocating HPV vaccination.
We asked parents to interpret each poster design and message, describe what they liked or disliked and suggest ways to increase its effectiveness. We eliminated two messages that participants considered too wordy (“Know the facts …” and “Six million people become newly infected …”) and one that they deemed too repulsive (“Genital warts are painful and gross.”) We retained the remaining six messages as poster headlines, consolidated images into four designs and further refined 22 informational bullet points for the intercept interviews. The final poster designs featured at least one influential figure or institution in parents’ decision to have their sons vaccinated: a doctor, teenager, church or parent. All of the posters included images of people of different ethnicities, and were designed in various hues of blue and green, preferred colors from the focus groups.
To build on the focus group findings, we conducted intercept interviews with 100 parents of boys aged 9–13 at a university-based pediatric and adolescent health clinic. We recruited these parents as they brought children (not necessarily the son in question) to the clinic, which serves a racially and ethnically diverse population. We focused on the 9–13 age-group to include parents who may have been contemplating their sons’ HPV vaccination or whose sons were just outside the CDC’s target ages of 11–12. Research team members gave parents a flyer about the study as they checked in to the clinic; if eligible and interested, parents were asked for their consent to do the interview after their child’s appointment. Trained interviewers recorded responses to closed-ended and open-ended questions. Participants received a $20 gift card for the 20-minute interview.
Before parents were shown headlines or poster designs, they were asked if they had heard of HPV vaccine for boys and how likely they were to have their sons vaccinated against HPV in the next 12 months. Interviewers next asked parents which headline, design and set of facts would most motivate them to seek HPV vaccine for their sons. The order of headlines, designs and facts was rotated with each participant. Questions covered two main domains: appearance and details (e.g., attention-getting, attractiveness, relevancy, reliability, acceptability); and cognitive processing and response (e.g., comprehension, how persuasive parents found the message, likely actions they would take after seeing it and information needed to make a decision about vaccination). Interviewers also asked an open-ended question about what kinds of information parents wanted to see on the posters.
After parents answered questions about each headline, we asked them to specify which one they found most motivating. We then showed them the four poster designs, in random order, each with this chosen headline. We did this to focus their attention on the design while keeping the chosen headline constant. Each poster design included space at the bottom where additional facts about the vaccine, as determined by results from these interviews, would be placed later. For the interviews, the space was filled with the statement “This is where the facts will go.” We instructed parents to focus on the colors, design layout and photos, and not on this space. We concluded the interview with a written questionnaire asking parents to rate how motivating each of 22 factual statements about HPV infection and HPV vaccine would be in their decision to have their sons vaccinated; scores ranged from 1 (not very) to 7 (extremely). Demographic information was collected, including ages of parents and sons, as well as parents’ gender, race and ethnicity.
Using Atlas.ti and a constant comparative method, two of the authors independently created coding categories from one randomly selected focus group transcript.33
Coding categories drew from the interview guide questions, the health belief model and social marketing principles. After coding one transcript, the authors compared their interpretations and refined coding categories. The same method continued for a second and third randomly selected transcript. The coders reached final agreement about the coding categories and created 47 codes with defined parameters and meanings. The same two coders then independently recoded all five focus group transcripts. Strong interrater reliability was attained: Agreement on the codes was 87–100%; Cohen’s kappa was more than 0.80 for 67% of the codes (mean, 0.83; range, 0.53–1.00).
Interviewers used audio recordings and notes to enter participants’ responses into an online database. From the entered responses, two of the authors created and refined a coding protocol derived from the interview guide questions and the health belief model. In each of three rounds of independent coding, approximately 10% of the codes disagreed. The two coders met after each round to compare and resolve discrepancies.34
Once reliability was medium to strong (agreement on the codes was 92–99%; Cohen’s kappa was (0.66–0.95 and averaged 0.79), the two authors each coded half of the remaining sample.
Chi-square tests were used to assess differences among parents—by gender, race, age, sons’ age and likelihood of vaccination—according to the poster headlines and designs they found most motivating, actions they would take after viewing their favorite poster design and further information they wanted to see on the poster. We calculated mean scores for how motivating the 22 factual statements were, and used an independent sample t test or analysis of variance to compare the means across subgroups. Eleven participants had more than one son aged 9–13, and we randomly selected one about whom parents would answer questions. Statistical significance was assessed at p<.05.