Increasing HPV Vaccination among Low-Income, Ethnic Minority Adolescents: Effects of a Multicomponent System Intervention through a County Health Department Hotline

Abstract

Background: Introduction of the human papillomavirus (HPV) vaccine in 2006 was a game-changing advance in cancer control. Despite the vaccine's potential cancer prevention benefits, uptake remains low. We utilized a randomized design to evaluate a multicomponent intervention to improve HPV vaccine uptake among low-income, ethnic minority adolescents seeking services through a county health department telephone hotline.

Methods: Hotline callers who were caregivers of never-vaccinated adolescents (11-17 years) were randomized by call-week to intervention or control conditions. The intervention included brief telephone and print education, delivered in multiple languages, and personalized referral to a low-cost/free vaccine provider. Participants completed baseline (n = 238), 3-month (n = 215), and 9-month (n = 204) telephone follow-up surveys.

Results: HPV vaccine initiation rates increased substantially by 9-month follow-up overall, although no differences were observed between intervention and control groups (45% vs. 42%, respectively, P > 0.05). We also observed significant improvements in perceived HPV risk, barriers to vaccination, and perceived knowledge in both study conditions (P < 0.05).

Conclusions: A low-intensity county hotline intervention did not produce a greater increase in HPV vaccination rates than routine practice. However, 44% of unvaccinated adolescents in both conditions received at least one dose of the vaccine, which can be viewed as a successful public health outcome. Future studies should evaluate more intensive interventions that address accessing and utilizing services in complex safety net settings.

Impact: Study results suggest the need for investigators to be aware of the potential priming effects of study participation, which may obscure the effect of low-intensity interventions.

Figure 1.. CONSORT diagram.

Of the 2786 caregivers assessed for eligibility, 238 were eligible and randomized by week of call; 138 to the intervention condition and 100 to the control condition. Ninety-four percent of randomized participants provided data at at least one follow-up period and were included in the primary outcome analysis.

Figure 2.. Changes in caregiver attitudes and beliefs, by study group (n=204)

This figure shows predicted probabilities of each outcome at baseline and the absolute difference (in percentage points) between baseline and 9-month follow-up (n=204). Predicted probabilities were calculated from mixed models that were adjusted for caregiver gender, household income, and clustering by week of enrollment, with exception of one item (child more likely to think it is okay to have sex if vaccinated), which was calculated from a model that adjusted for clustering by week of enrollment only. There were no significant differences in change over time between study groups.