A Mixed-Methods Examination of Factors Related to HPV Vaccination Promotion in Private Dental Settings, Iowa, 2019

Natoshia Askelson, MPH, PhD; Grace Ryan, MPH; Susan McKernan, MS, DMD, PhD; Aaron Scherer, MA, PhD; Eliza Daly, BA; Lejla Avdic, BA


Prev Chronic Dis. 2021;18(3):e26 

In This Article


With the goal of using this formative research to inform intervention development, we conducted a mail-based survey of dental hygienists and individual interviews with dentists and dental hygienists. We reported results in this analysis from participants working in a private practice only. Our intention was to develop an intervention focused on dental hygienists; therefore, we conducted both surveys and interviews with that population. We included dentists in the interviews to understand their perception of dental hygienists' ability to participate in an HPV-focused intervention and their willingness to be supportive of these efforts. Our work was determined not to be human subjects research by the University of Iowa's institutional review board.

Mailed Survey to Dental Hygienists

By using existing research,[11,13–15] we designed a 31-item questionnaire to assess individual characteristics, HPV-related measures, clinic characteristics, and individual willingness to incorporate HPV vaccine promotion into practice (Table 1). HPV-related measures included current vaccine promotion activities, the extent to which hygienists viewed themselves as vaccine promoters, and their willingness to be involved in vaccine promotion activities. We included items about demographics and personal beliefs, because previous studies found associations among religiosity, political beliefs, and perceptions of the HPV vaccine.[16–18] We obtained a list of all licensed dental hygienists in Iowa from the Iowa Dental Board; surveys were mailed to 2,074 dental hygienists in May 2019. We sent a postcard reminder after 10 days to dental hygienists who had not completed the survey, followed by a second survey mailing to all non-completers after 20 days.

We considered 4 primary outcomes related to dental hygienists' willingness to complete the following vaccine-related activities: 1) Participate in continuing education (CE) about HPV and oral cancer, 2) Educate patients about the HPV vaccine, 3) Recommend the HPV vaccine to parents of adolescent patients, and 4) Refer parents of adolescent patients to eligible providers for HPV vaccination (Table 1). We defined a recommendation for the HPV vaccine as the dental hygienists' endorsement of the vaccine to a parent of an adolescent patient, whereas a referral is the act of connecting a parent to a local vaccinating provider.

Most items were analyzed by using the original response categories, but, for several questions, we collapsed response categories as a result of small numbers. Discussing the HPV vaccine as oral cancer prevention with either parents or adolescent patients had responses ranging from 1 for never to 5 for always, but we dichotomized responses for those items. Never discussed became 0, and responses 2–5 for the same discussion became 1 for discussed.

Responses to items measuring practice activities, personal HPV vaccination, knowing someone with an HPV-related cancer, or whether the practice accepted Medicaid were dichotomized into 0 for no, unsure, or I don't know, and 1 for yes.

To explore factors driving dental hygienists' willingness to participate in HPV vaccine promotion activities, we used a hierarchical model-building approach to generate multiple linear regression models. For each outcome, we considered 3 models. Model 1 included individual demographic characteristics; model 2 added variables related to attitudes and beliefs; and model 3 incorporated practice-level characteristics.

Interviews With Dental Hygienists and Dentists

We conducted telephone interviews with dental hygienists (March–July 2019) and dentists (July–August 2019) to assess their willingness to perform specific activities related to HPV vaccine promotion and their preferred methods of receiving educational information. Semistructured interview guides were created using questions written by the research team, existing tools, and adapted closed-ended survey items.[15,19] We sent the dentist interview guide to a dentist for review and then met as a team to revise questions based on the dentist's feedback. For both sets of interviews, we used a multipronged sampling strategy that included predetermined sampling frames and active recruitment at area conferences for dental providers.

Dental Hygienists

We stratified dental hygienists who completed the survey by congressional district (n = 4) and randomly selected 25 dental hygienists from each district (n = 100). Invitations were sent to a new set of 100 randomly selected dental hygienists after an inadequate response rate for the first round of invitations (10%). Dental hygienists were sent an initial letter and follow-up postcard inviting them to sign up for an interview online. We also recruited dental hygienists at 2 local dental professional conferences (n = 5).


Our sampling frame for dentists comprised the alumni list (N = 1,642) provided by the University of Iowa College of Dentistry, from which 76% of the state's dentists graduate.[20] We first eliminated anyone from the list without an email address, leaving 905 potential participants. We originally drew a random sample of 300 dentists from this pool of 905 and conducted recruitment via either email or phone. When this failed to yield a satisfactory number of interviews (n = 12), we mailed recruitment materials to the remainder of the sample (n = 605).

All interviewees who completed an interview were sent a $25 gift card for their choice of a local grocery store or gas station. Interviews were audio-recorded and transcribed by a third-party service. We used the same analytic approach for both sets of interviews. We first developed separate codebooks for the dental hygienist and dentist interviews using a primarily deductive method, in which codes were created based on our research questions and interview guides; however, we also allowed coders to add their own codes if they thought it was warranted after discussion with the coding team. Initially, 2 research assistants coded 2 transcripts from each data collection. To establish reliability between the 2 coders, we used a negotiated agreement process in which the 2 coders met with a third member of the research team to discuss any discrepancies and update the codebooks as needed.[21] The remaining transcripts were divided equally among the original coders and coded using NVivo 12 software (QSR International).