Using Immunization Registry Data to understand disparities in age 9 HPV vaccine initiation in a Midwestern state, US

Abstract

Background/Objectives: Initiating the HPV vaccine series at ages 9/10 compared to 11+, results in on-time completion. This evidence is strong enough that several organizations in the United States, including the American Cancer Society and the American Academy of Peditrics, have updated their recommendation language to indicate conversations about HPV vaccination should begin at earlier ages. However, since little is known about factors associated with early initiation, our goal was to explore sociodemographic factors associated with early initiation using immunization registry data.

Methods: Immunization registries are state-wide databases that are able to work bidirectionally with electronic medical record systems to keep immunization history up-to-date. Iowa, a rural, midwestern state, has a robust registry that captures nearly 100% of the immunizations that occur in the state. We used data from the Iowa Immunization Registry for children born between 2004 and 2013 to explore sociodemographic factors (e.g., gender, race/ethnicity, rurality, and insurance type) associated with early initiation. We then used logistic regression to model the factors associated with likelihood of early HPV vaccine initiation. Results: Of the 255,833 children who initiated the series in our data set, 3.2% (n=8,355) initiated at ages 9 or 10, 173,831 (67.9%) initiated at ages 11 or 12, and the remaining 73,647 (28.8%) initiated after 12. We observed significant differences between the children who initiated at ages 9/10 compared to later across all sociodemographic categories. Males were less likely to initiate early (OR: 0.77, 0.74;0.81) compared to females. Compared to white children, racial and ethnic minority children were more likely to initiate the series early; for example, black children had an odds ratio of 1.99 (CI: 1.87, 2.13) and Hispanic or Latino children had an odds ratio of 1.67 (CI: 1.57, 1.79). Those living in the most rural areas (OR: 0.76, CI: 0.72, 0.81) and those with either no insurance (OR: 0.84, CI: 0.80, 0.89) or public insurance (OR: 0.53, CI: 0.50, 0.56) were less likely to initiate early. Conclusions: Given the recent focus on early initiation for the series, our results contribute to a growing understanding of who is initiating the series earlier and where disparities are occurring. In some ways, these results echo patterns previously seen in analyses of initiation; males, racial/ethnic minorities, and adolescents living in rural areas are less likely to initiate the series in general compared to females, non-Hispanic white, and urban-dwelling adolescents. Whereas, national data has shown that adolescents with public insurance have higher initiation and completion rates, we found that public insurance was associated with a lower likelihood of early initiation. Immunization registry data are highly valuable in providing these nuanced, population-level details about immunization uptake and we can now use this data to inform development of tailored interventions to better promote early-initiation of HPV vaccination. Our future steps will include geospatial analysis so that we can not only develop tailored messaging based on these identified sociodemographic characteristics but also based on geographic areas where early initiation is lagging.