HHS Public Access Author manuscript Author Manuscript Vaccine. Author manuscript; available in PMC 2021 October 07. Published in final edited form as: Vaccine. 2020 October 07; 38(43): 6730–6734. doi:10.1016/j.vaccine.2020.08.048. County-level correlates of missed opportunities for HPV vaccination in Indiana: An environmental scan Sharon C. Enujiokea, Rivienne Shedd-Steeleb, Joanne Daggyc, Heather Burneyd, Lisa Robertsone, Katharine Headf, Gregory Zimetg,* a.Indiana 100th University School of Medicine, Department of Pediatrics-Adolescent Medicine. 410 W. Street. Ste 100. Indianapolis, IN 46202, USA. Author Manuscript b.Indiana University Simon Cancer Center. 535 Barnhill Dr. Indianapolis, IN 46202, USA. University School of Medicine, Department of Biostatistics. 410 W. 10th St. Indianapolis, IN 46202, USA. c.Indiana University School of Medicine, Department of Biostatistics. 410 W. 10th St. Indianapolis, IN 46202, USA. d.Indiana e.Indiana Immunization Coalition. 6239 S East St. Indianapolis, IN 46227 USA. f.Indiana University-Purdue University Indianapolis, Department of Communication Studies, 425 University Boulevard, Indianapolis, IN 46202, USA. g.Indiana Author Manuscript 100th University School of Medicine, Department of Pediatrics-Adolescent Medicine. 410 W. Street. Ste 100. Indianapolis, IN 46202, USA. Abstract Background: The goal of this study was to examine variability across the 92 Indiana counties in missed opportunities for HPV vaccination and to assess county-level correlates of missed opportunities. Methods: The Indiana immunization registry provided county level data on 2017 missed opportunity rates for adolescents ages 11-18. A missed opportunity was an encounter when a patient eligible for HPV vaccination received one or more other recommended vaccines, but not HPV. Potential county-level correlates of missed opportunities included race, income, insurance Author Manuscript * Corresponding author. gzimet@iu.edu. Declaration of Competing Interest Sharon Enujioke has no interest to disclose. Rivienne Shedd-Steele has no interest to disclose. Joanne Daggy has no interest to disclose. Heather Burney has no interest to disclose. Lisa Robertson has no interest to disclose. Katharine J. Head has no interest to disclose. Gregory Zimet received travel funding from Merck to present research at a scientific conference and received honoraria from Sanofi Pasteur and Merck for consultations related to adolescent vaccination. Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Enujioke et al. Page 2 Author Manuscript status, population density, education, primary care providers per capita, smoking rates, mammography screening, diabetes monitoring, and Pap testing. Results: The missed opportunity rate ranged from 31% to 85% across Indiana counties. Higher population density, mammography screening, income inequality, and diabetes monitoring were associated with fewer missed opportunities. Conclusions: We found wide variability in missed opportunities across counties, which were associated with population density and county-level participation in other health-related behaviors Keywords Vaccination; HPV Vaccination; Health Status; Adolescence; Immunization Surveillance Author Manuscript Human papillomavirus (HPV) is the most common sexually transmitted infection in the United States. An estimated 14 million persons are newly infected every year. HPV is also the primary cause of genital warts and is a major contributing factor in at least six cancers: cervical, vulvar, vaginal, penile, anal, and oropharyngeal. [1] Author Manuscript Currently only one HPV vaccine, the 9-valent vaccine (9vHPV) is available in the U.S. and is licensed by the Food and Drug Administration (FDA) for use in males and females 9 through 45 years of age [2]. The Center for Disease Control’s (CDC) Advisory Committee on Immunization Practices (ACIP) recently issued an updated set of recommendations for HPV vaccination.[3] The guidelines continue to indicate that HPV vaccine should be routinely administered to females and males ages 11 and 12 years and that vaccination can start as early as age 9 years. They also recommend routine catch-up vaccination for all females and males from ages 13 through 26 years, with a shared clinical decision-making approach recommended for individuals 27 through 45 years of age.[3] The ACIP currently recommends two doses of vaccine 6-12 months apart if the first dose is administered before age 15 years, with a three-dose recommendation for those who get the first dose at age 15 years or older. [4] The Healthy people 2020 goal is for 80% of adolescents to have completed their HPV vaccination series. In the United States, the HPV vaccination completion rate in 13-17 year olds is 51.1%, lagging behind the Healthy People 2020 goal. [5] Author Manuscript Although Indiana HPV vaccination rates have shown steady progress over the years, initiation of vaccination continues to lag in comparison to the national average. In 2017, when data for the current study were collected, Indiana had a 59.3% initiation rate among adolescents ages 13-17 years, ranking 40th in the United Stated.[6] In comparison, Indiana ranks in the top ten for administration of the tetanus, diphtheria, pertussis (Tdap; 93.1%) and meningococcal ACWY (MenACWY; 95%) vaccines. The discrepancies in vaccination rates between HPV and Tdap/MenACWY are significant, as Tdap and MenACWY are usually administered at 11-12 years of age, the same age that the ACIP recommends initiating HPV vaccination.[6] There could be several reasons for the discrepancies. For instance, Indiana code IC 20-34-4-2 mandates Hepatitis A, Tdap, and MenACWY for middle school entry (ISDH), but not HPV vaccine.[7] Another reason could be a weak recommendation (or no recommendation) for HPV vaccination from a health care provider (HCP).[8] A missed Vaccine. Author manuscript; available in PMC 2021 October 07. Enujioke et al. Page 3 Author Manuscript opportunity is classified as a health care encounter when a patient eligible for HPV vaccination received one or more other recommended vaccinations, but not the HPV vaccine.[9] Author Manuscript Indiana is made up of 92 counties that vary greatly with respect to population density, access to healthcare, and other sociodemographic factors. A study of HPV vaccination rates across the state of Utah showed that there was substantial variability in missed opportunities, suggesting this may be the case for other states.[10] The purpose of the present study was to similarly document county level variability in missed opportunities for adolescent HPV vaccination in Indiana and to examine county level factors that may be associated with missed opportunities across the 92 counties. The findings from this study can provide important insight into different county characteristics associated with vaccination rates, which could be useful when developing geographically-targeted interventions to improve adolescent HPV vaccination rates. MATERIALS AND METHODS This study examined missed opportunities across Indiana’s 92 counties for initiating HPV vaccination amongst adolescents 11-18 years of age during the 2017 calendar year. We included age 18 as the upper limit, because the Vaccines for Children (VFC) program covers vaccination through age 18 years. The data were extracted from the Indiana State Department of Health (ISDH) Indiana Immunization Information System (IIS), also known as CHIRP (Children and Hoosier Immunization Registry Program). The Indiana University Institution Review Board made the determination that this study was exempt from review. Author Manuscript MEASURES The outcome variable of interest, a missed opportunity for HPV vaccination during the 2017 calendar year, was defined as a clinical visit during which an HPV-vaccine-eligible adolescent between the ages of 11-18 received one or more vaccines (i.e., Tdap, MenACWY, Hepatitis A), but not the first dose of HPV vaccine. Since the influenza vaccine is seasonal, receipt of the influenza vaccine was not counted as a missed opportunity. We retrieved county-level adolescent immunization data from CHIRP, which is a secure web-based immunization registry program that permanently stores immunizations records. It is administered by the ISDH, which mandates that all immunization providers enter vaccine administration information for all individuals under the age of 19. No patient-level information (other than age group) or information on race/ethnicity was available through the CHIRP database. Author Manuscript We derived county-specific socio-demographic and health data from two sources: the 2010 U.S Census (https://www.census.gov/programs-surveys/decennial-census/decade.2010.html) and the 2018 County Health Rankings report (https://www.countyhealthrankings.org/), a program supported by the Robert Wood Johnson Foundation. From the measures available, we selected measures that focused on sociodemographic characteristics and health-related factors, including population density per square mile, percentage rural, primary care providers and mental health providers per capita, dental providers per capita, race, rates of Vaccine. Author manuscript; available in PMC 2021 October 07. Enujioke et al. Page 4 Author Manuscript mammography screening (i.e., percentage of female Medicare enrollees ages 67-69 that receive mammography screening), median household income, income inequality, diabetes monitoring (i.e., percentage of Medicare enrollees with diabetes ages 65-75 that receive HbA1c monitoring), rates of smoking, and parental education. See Table one for a complete list of measures and the source of each measure. Although the mammography screening and diabetes monitoring measures involve older populations, we included them as variables that might indirectly reflect county-level access to care and utilization of health resources. STATISTICAL METHODS Author Manuscript Descriptive statistics were reported for all covariates and outcomes at the county level. County-level variables considered as potential covariates in this analysis include population density per square quarter mile, percentage of individuals with certain educational attainment levels (i.e., less than a high school diploma, only a high school diploma, some college, or Bachelor’s degree or higher), percentage of smokers, percentage of women with mammography screening, median household income, percentage of racial groups (i.e., African Americans, Hispanics, Non-Hispanic Whites), percentage of the county that is rural, income inequality ratio, dentists per capita, mental health providers per capita, percentage of individuals who engage in diabetes monitoring, and total number of primary care providers per 100,000. Author Manuscript Generalized linear mixed models (GLMM) with a binomial distribution and logit link were used to determine county-level covariates that were independently associated with the proportion of missed opportunities for HPV vaccination, while adjusting for the patient-level covariate of age group (11-12 years old vs 13-18 years old). [11] A random intercept for county was included to account for correlation of outcomes from patients in the same county. For model-building univariable GLMMs were fit, which included age group and each candidate covariate. Odds ratios and associated 95% confidence intervals are reported. Covariates significant at the 0.05 level in univariable models were then included in a multivariable model. The intraclass correlation (ICC) of the outcome was estimated from a model which included only the county-level random intercept and from the full model. From the estimates of the ICC the percent of variability in the county-level proportions of missed opportunities explained by the patient age group and county-level covariates was estimated. Author Manuscript To determine if the effect of the important county-level covariates differed by age group, all two-way interactions between age group and the county-level covariates were then included in the model. The logistic regression model with the final selected county-level covariates was then fit separately by age-group. Odds ratios and the associated 95% profile-likelihood confidence intervals are reported. All analysis was conducted with SAS V9.4 (Cary, NC). RESULTS Descriptive statistics of the 92 counties in Indiana is outlined in Table 1. Table 2 provides summary statistics for the proportion of missed opportunities for HPV vaccination for the state of Indiana. There were significant missed opportunities, with 33% of counties having Vaccine. Author manuscript; available in PMC 2021 October 07. Enujioke et al. Page 5 Author Manuscript greater than 66% missed opportunities for receiving HPV vaccination. The mean proportion of missed opportunities for adolescents’ ages 11-18 years was 60%, with 57% missed opportunities in the 11-12 age group and 61% missed opportunities in the 13-18 age group (Table 2). The proportion of missed opportunity was significantly higher in children 13-18 years of age vs 11-12 years of age, OR = 1.12, 95% CI = [1.10, 1.14], p-value < 0.0001. From the logistic regression models adjusting for only age group (Table 3), increases in population density (p=.02), mammography screening (p<.0001), income inequality (p=.03), and diabetes monitoring (p=.0006) are significantly associated with less missed opportunities. Author Manuscript Based on the intracluster correlation (ICC) of the intercept-only model (ICC = .07), 7% of the variability in missed opportunities was due to variability between counties. From the full model including interaction terms (ICC = 0.052), only 5.2% of the total variability was due to variability between counties, thus patient age group and county-level factors explained 25.7% of the county-level variability in missed opportunities. The addition of the interaction terms provided an improved model fit compared to the model including only main effects based on the likelihood ratio test (p < 0.0001), thus the effect of these county-level factors on missed opportunities varied by age group. Author Manuscript From the final logistic regression model (Table 4), the two-way interaction terms between each of the four factors and age group were all statistically significant (all p < 0.01). Thus, a separate logistic regression model was fit to each age group including the main effects of population density, mammography screening, income inequality, and diabetes monitoring. For both age groups, lower population density, lower income inequality ratio, and lower percentage of individuals who engage in diabetes monitoring were associated with higher risk of missed opportunities. The odds ratio for population density was smaller for the 13-18 year-old age group indicating a more pronounced association for this age group, whereas the association of income inequality ratio and diabetes monitoring was less pronounced in the 13-18-year-old age group (see Table 5). Author Manuscript For the 13-18 year-old age group, a higher percentage of women with mammography screening was associated with lower risk of a missed opportunity (OR = 0.90, 95% CI =[0.89, 0.92]) which is what we would expect. However, for the 11-12 year-old group, the association was reversed (OR = 1.04, 95% CI =[1.01, 1.07]). In univariable results for the 11-12 year old group, we found that mammography screening was in the direction we would expect (OR = 0.95, 95% CI = [0.92,0.98]) but not when including the three variables of population density, income inequality ratio, and diabetes monitoring. Thus, we ran additional logistic regression models for this age group by examining the inclusion of each of these variables one at a time with mammography screening. We found that if diabetes monitoring only is included, mammography screening is no longer statistically significant (OR = 1.00, 95% CI =[0.96, 1.03], p-value = 0.766). Thus, mammography screening and diabetes monitoring are highly related in this age group. Vaccine. Author manuscript; available in PMC 2021 October 07. Enujioke et al. Page 6 Author Manuscript DISCUSSION The findings from this study indicate wide variability in missed opportunities for HPV vaccination across the 92 counties in Indiana, results similar to those reported by Kepka et al., who found extensive missed opportunities across Utah. [10] Although Indiana does not require HPV vaccine for entry into middle school (6th grade), it does require students to receive Tdap, MenACWY, and Hepatitis A vaccines. Therefore, there should be ample opportunities to improve HPV vaccine delivery at the appropriate target age of 11-12 years old. Author Manuscript Keeping in mind that we were only looking at county-level factors, the association between increased mammography screening and diabetes monitoring with less missed opportunities, particularly in the older age group, may reflect county-level issues related to engagement with, and orientation to, health. Even though health care providers per capita was not related to missed opportunities, in sparsely populated counties, it may nonetheless be challenging to access health care. In support of this idea is the finding that counties with lower population density had higher missed opportunities, a result also reflected in NIS-Teen data [5], which show lower vaccine coverage in rural vs. urban areas. Although 2018 National Immunization Survey-Teen data have shown racial/ethnic disparities in HPV vaccinations (NIS-Teen Supplementary Table 1) [5], this study did not find variability in missed opportunities related to race/ethnicity, which may simply be a reflection of the fact that our analyses could only consider race as a county-level variable (e.g., percentage of African Americans). A lower missed opportunity rate was associated with greater income inequality, which might be attributed to income inequality being higher in urban areas.[12] Author Manuscript The findings from this study are consistent with the framework of Bronfenbrenner’s SocialEcological Model, which considers how factors at the community and societal level can influence relationship-level factors and individual behaviors.[13] We found, for instance, that county-level measures (e.g., population density and engagement in mammography screening) were associated with missed opportunities for HPV vaccination. Although parental beliefs and health care provider recommendation are strong proximal determinants of HPV vaccination, findings from this study suggest that it may be important to consider the ways in which these proximal factors influence, and are influenced by, larger community characteristics. Author Manuscript A strength of this study was the availability of comprehensive vaccination data from the Indiana IIS, which mandates that all vaccine providers submit vaccination records for all persons under 19 years of age. These data enabled us to accurately assess missed opportunities for HPV vaccination at the county level. A limitation of this study was the focus on a single state in the U.S., which limits generalizability of the results to other states. Also, we were constrained by the availability of only county level data, which necessarily limited our ability to identify a full range of potential correlates of missed opportunities. Furthermore, we were not able to obtain a county-level breakdown of missed opportunities Vaccine. Author manuscript; available in PMC 2021 October 07. Enujioke et al. Page 7 Author Manuscript by race or gender. Finally, we could not track follow-up data for individual adolescents to determine if they had received HPV vaccine at a separate clinical visit. CONCLUSION Author Manuscript This study evaluated missed opportunities for adolescent HPV vaccination across the 92 counties in Indiana. We found wide variability in missed opportunities from county to county, which were associated with population density and county-level participation in other health related behaviors. Findings suggest that interventions designed to improve HPV vaccine coverage may need to consider and target smaller geographic areas. The use of the IIS data set can help identify counties that have high missed opportunities for HPV vaccination, therefore indicating a need for a deeper understanding of the barriers present in each of those counties. This study also reinforces the need to address rural-urban disparities and suggests there is a need to better understand the difference across counties in orientation towards health and access to healthcare. Acknowledgments Funding/Support This study was supported by the National Cancer Institute under Award number P30CA08270918S4. References Author Manuscript Author Manuscript 1. Markowitz LE, Dunne EF, Saraiya M, et al. Human papillomavirus vaccination: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR. Recommendations and reports : Morbidity and mortality weekly report. Recommendations and reports 2014;63(Rr-05):1– 30 2. US Food and Drug Administration. FDA Approves expanded use of Gardasil 9 to include individuals 27 through 45 years old. 2018 Available at: https://www.fda.gov/newsevents/newsroom/ pressannouncements/ucm622715.htm [accessed January 13, 2020]. 3. Meites E, Szilagyi PG, Chesson HW, Unger ER, Romero JR, Markowitz LE. Human Papillomavirus Vaccination for Adults: Updated Recommendations of the Advisory Committee on Immunization Practices. MMWR. Morbidity and mortality weekly report 2019;68(32):698–702 doi: 10.15585/ mmwr.mm6832a3. [PubMed: 31415491] 4. Meites E, Kempe A, Markowitz L. Use of a 2-dose schedule for Human Papillomavirus VaccinationUpdated recommendations of the Advisory Committee on Immunization Practices. . MMWR Morbidity and mortality weekly report 2016:1405–08 [PubMed: 27977643] 5. Walker TY, Elam-Evans LD, Yankey D, et al. National, Regional, State, and Selected Local Area Vaccination Coverage Among Adolescents Aged 13-17 Years - United States, 2018. MMWR. Morbidity and mortality weekly report 2019;68(33):718–23 doi: 10.15585/mmwr.mm6833a2. [PubMed: 31437143] 6. Walker TY, Elam-Evans LD, Yankey D, et al. National, Regional, State, and Selected Local Area Vaccination Coverage Among Adolescents Aged 13-17 Years - United States, 2017. MMWR. Morbidity and mortality weekly report 2018;67(33):909–17 doi: 10.15585/mmwr.mm6733a1. [PubMed: 30138305] 7. Indiana Code Title 20. Education § 20-34-4-2. 2018 Available at: https://codes.findlaw.com/in/ title-20-education/in-code-sect-20-34-4-2.html [accessed January 13, 2020]. 8. Sturm L, Donahue K, Kasting M, Kulkarni A, Brewer NT, Zimet GD. Pediatrician-Parent Conversations About Human Papillomavirus Vaccination: An Analysis of Audio Recordings. The Journal of adolescent health : official publication of the Society for Adolescent Medicine 2017;61(2):246–51 doi: 10.1016/j.jadohealth.2017.02.006. [PubMed: 28455129] Vaccine. Author manuscript; available in PMC 2021 October 07. Enujioke et al. Page 8 Author Manuscript 9. Oliveira CR, Rock RM, Shapiro ED, et al. Missed opportunities for HPV immunization among young adult women. American journal of obstetrics and gynecology 2018;218(3):326.e126.e7 doi: 10.1016/j.ajog.2017.11.602. [PubMed: 29223597] 10. Kepka D, Spigarelli MG, Warner EL, Yoneoka Y, McConnell N, Balch A. Statewide analysis of missed opportunities for human papillomavirus vaccination using vaccine registry data. Papillomavirus research (Amsterdam, Netherlands) 2016;2:128–32 doi: 10.1016/ j.pvr.2016.06.002. 11. Brown HP, Prescott R. Applied Mixed Models in Medicine. New York, NY: John Wiley and Sons, 1999. 12. Bishaw A, Posey KG. A comparison of rural and urban America: Household income and poverty. United States Census Bureau 2016 Available at: https://www.census.gov/newsroom/blogs/ randomsamplings/2016/12/a_comparison_of_rura.html [accessed January 13, 2020]. 13. Bronfenbrenner U Toward an experimental ecology of human development. Amer Psychol. 1977;32(7):513–531. Author Manuscript Author Manuscript Author Manuscript Vaccine. Author manuscript; available in PMC 2021 October 07. Enujioke et al. Page 9 Table 1 Author Manuscript Descriptive statistics for county-level variables (N = 92). Variable Mean ± Standard Deviation Population 2010 Census 70.476 ± 116.748 Density Per Square Mile 2010 Census 11.1 ± 17.60 Percentage with Less than HS Diploma 12.7 ± 4.03 Percentage with HS Diploma 39.8 ± 6.21 Percentage with Some College 28.9 ± 2.81 Percentage with Bachelors or More 18.7 ± 8.02 Author Manuscript Percentage of Smokers 19.7 ± 1.95 Percentage with Mammography Screening 60.2 ± 5.92 Median Household Income ($) 52,606 ± 8,525.2 Percentage of African Americans 2.8 ± 4.46 Percentage of Hispanics 4.0 ± 3.51 Percentage of Non-Hispanic Whites 90.6 ± 7.72 Percentage of Rural 54.5 ± 26.91 Income Inequality Ratio 4.0 ± 0.47 Dentists Per Capita 39 ± 17.2 Mental Health Providers Per Capita 87 ± 71.7 Diabetes Monitoring 85 ± 6.7 Tout PCP Kate per 100,000 102 ± 46.3 Author Manuscript Author Manuscript Vaccine. Author manuscript; available in PMC 2021 October 07. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Table 2 Mean ± Standard Deviation (N = 92) Median (25th–75th Percentile) Minimum Maximum Missed Opportunities 11–12 Years 0.56 ± 0.13 0.55 (0.46–0.65) 27% 94% Missed Opportunities 13–18 Years 0.61 ± 0.11 0.61 (0.51–0.68) 31% 84% Missed Opportunities 11–18 Years 0.60 ± 0.11 0.60 (0.51–0.67) 31% 85% Enujioke et al. The proportion of missed opportunities summarized across counties by age group. Vaccine. Author manuscript; available in PMC 2021 October 07. Page 10 Author Manuscript Author Manuscript Author Manuscript Author Manuscript Table 3 Effect Vaccine. Author manuscript; available in PMC 2021 October 07. * Unit Increase Odds Ratio, 95% CI p-value Density Per Square Mile 2010 Census 100 0.51 [0.29, 0.91] 0.023 Percentage with HS Diploma 10 0.91 [0.77, 1.08] 0.275 Percentage with Some College 10 1.42 [0.99, 2.04] 0.057 Percentage with Mammography Screening 10 0.72 [0.61, 0.84] <0.0001 Median Household Income (thousands) 10 1.03 [0.91, 1.16] 0.638 Percentage of African Americans 10 0.84 [0.67, 1.06] 0.142 Percentage of Hispamcs 10 1.01 [0.75, 1.36] 0.926 Percentage of Non-Hispanic Whites 10 1.10 [0.96, 1.26] 0.161 Percentage of Rural 10 1.04 [1.00, 1.08] 0.063 Income Inequality Ratio 1 0.79 [0.64, 0.98] 0.034 Dentists Per Capita 10 0.95 [0.90, 1.01] 0.112 Mental Health Providers Per Capita 10 0.99 [0.98, 1.01] 0.210 Diabetes Monitoring 10 0.77 [0.66, 0.89] 0.001 Total PCP Rate per 100,000 10 0.98 [0.96, 1.00] 0.093 p < 0.05 * Enujioke et al. Logistic regression with each effect and age group only. * * * Indicates a statistically significant effect. Page 11 Author Manuscript Author Manuscript Author Manuscript Author Manuscript Table 4 Effect Intercept Estimate SE Num DF Den DF F Value p-value 4.221 0.805 - - - - Age Group (13–18) −0.344 0.209 1 85 2.72 0.103 Density Per Square Mile −0.001 0.003 1 85 0.58 0.450 Percent Mammography Screening −0.010 0.008 1 85 4.20 0.043 Vaccine. Author manuscript; available in PMC 2021 October 07. Income Inequality Ratio −0.334 0.105 1 85 5.17 0.026 Diabetes Monitoring −0.024 0.007 1 85 8.48 0.005 Density Pet Square Mile * Age Croup (13–18) −0.003 0.0002 1 85 158.92 <0.0001 Percent Mammography * Age Croup (13–18) −0.013 0.002 1 85 53.68 <0.0001 Income Inequality Ratio * Age Croup (13–18) 0.198 0.019 1 85 109.37 <0.0001 Diabetes Monitoring * Age Croup (13–18) 0.007 0.002 1 85 9.53 Enujioke et al. Final multivariable logistic regression model with age interaction terms. 0.003 Page 12 Author Manuscript Author Manuscript Author Manuscript Author Manuscript Table 5 Age Group (11–12) Effect Unit Increase Age Group (13–18) Odds Ratio, 95% CI p-value Odds Ratio, 95% CI p-value Density Per Square Mile 100 0.79 [0.76 ,0.82] <0.0001 0.60 [0.59 ,0.61] 0.023 Percent Mammography Screening 10 1.04 [1.01 ,1.07] 0.021 0.90 [0.89 ,0.92] <0.0001 Vaccine. Author manuscript; available in PMC 2021 October 07. Income Inequality Ratio 1 0.71 [0.69, 0.74] <0.0001 0.88 [0.87, 0.89] <0.0001 Diabetes Monitoring 10 0.72 [0.69, 0.75] <0.0001 0.77 [0.76, 0.78] <0.0001 Enujioke et al. Multivariable logistic regression separately by age group (N = 92). Page 13