Race-specific trends in HPV vaccinations and provider ... - Public Health

3 downloads 2686 Views 581KB Size Report
Results: Provider recommendations to vaccinate and HPV vaccination uptake have increased ... E-mail address: [email protected] (A.M. Burdette). Available ...
p u b l i c h e a l t h 1 4 2 ( 2 0 1 7 ) 1 6 7 e1 7 6

Available online at www.sciencedirect.com

Public Health journal homepage: www.elsevier.com/puhe

Original Research

Race-specific trends in HPV vaccinations and provider recommendations: persistent disparities or social progress? Amy M. Burdette a,*, Noah S. Webb a, Terrence D. Hill b, Hanna Jokinen-Gordon c a

Department of Sociology and Pepper Institute for Aging and Public Policy, Florida State University, USA School of Sociology, University of Arizona, USA c University of Arkansas for Medical Sciences, USA b

article info

abstract

Article history:

Objectives: Although racial and ethnic differences in HPV vaccination initiation are well

Received 6 April 2016

established, it is unclear whether these disparities have changed over time. The role of

Received in revised form

health provider recommendations in reducing any racial and ethnic inequalities is also

10 June 2016

uncertain. This study addresses these gaps in the literature.

Accepted 12 July 2016

Study design: Repeated cross-sectional design.

Available online 1 September 2016

Methods: Using data from the National Immunization Survey-Teen (2008e2013), we estimated a series of binary logistic regressions to model race-specific trends in (1) provider

Keywords:

recommendations to vaccinate against HPV and (2) HPV vaccine initiation for males

Health disparities

(n ¼ 56,632) and females (n ¼ 77,389).

Race and ethnicity

Results: Provider recommendations to vaccinate and HPV vaccination uptake have

Vaccination HPV

increased over time for adolescent males and females and across all racial and ethnic groups. Among girls, minority youths have seen a sharper increase in provider recommendations and HPV vaccination uptake than their White counterparts. Among boys, minority teens maintain higher overall rates of HPV vaccine uptake, however, Hispanics have lagged behind non-Hispanic Whites in the rate of increase in provider recommendations and HPV vaccinations. Conclusions: Our results suggest that racial and ethnic disparities in provider recommendations and HPV vaccinations have waned over time among males and females. While these trends are welcomed, additional interventions are warranted to increase overall rates of vaccination across race, ethnicity, and gender. © 2016 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.

Human papillomavirus (HPV), which can cause cervical cancer, genital warts, and other anogenital cancers, is the most common sexually transmitted infection in the USA.1,2 Strains

of HPV are responsible for 99.7% of cervical cancer cases. Cervical cancer, the second most common malignant disease among women in the USA, is responsible for approximately

* Corresponding author. Department of Sociology, Florida State University, 526 Bellamy Building, Tallahassee, FL 32306-2270, USA. E-mail address: [email protected] (A.M. Burdette). http://dx.doi.org/10.1016/j.puhe.2016.07.009 0033-3506/© 2016 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.

168

p u b l i c h e a l t h 1 4 2 ( 2 0 1 7 ) 1 6 7 e1 7 6

4000 deaths per year.3,4 Over 25% of females between the ages of 14 and 59 years are infected with at least one strain of HPV.5,6 The virus is also prevalent among sexually active males.7 Recent estimates suggest that oral HPV infection, the cause of a subset of oropharyngeal squamous cell carcinomas (OSCCs),8 is more common in men than women.9 In 2006, a vaccine was approved for use among females between the ages of 9 and 26 years.10 HPV vaccine guidelines were later extended to males in October 2009, with a routine vaccination recommendation beginning in October 2011.11 Research suggests that cervical cancer screening, diagnosis, treatment, and survival vary widely according to sociodemographic characteristics.12,13 African American and Hispanic women exhibit higher rates of cervical cancer than their nonHispanic White counterparts.14,15 Rather than reducing social disparities in HPV infection and cervical cancer mortality, the HPV vaccine could potentially increase racial and ethnic disparities if the vaccine is unequally distributed in the population. African American and Hispanic females appear to be less likely than non-Hispanic Whites to have initiated16e19 and completed17,20e22 the three-dose HPV vaccine series. Although social differences in HPV vaccination uptake among boys have been understudied, there is some evidence to suggest that nonHispanic White parents are less likely to have vaccinated their sons than parents of other racial and ethnic groups.23,24 Receiving a recommendation from a health provider to vaccinate against HPV is a key predictor of receiving the vaccine.25e27 Drawing on the fundamental cause theory of health disparities, Polonijo and Carpiano (2013)21 find important social inequalities in both HPV vaccine knowledge and receipt of a health provider's recommendation to vaccinate. Fundamental cause theory postulates that health inequalities emerge and persist in society because of the unequal distribution of health-beneficial resources. Those who are more socially advantaged in terms of knowledge, money, status, and social connections tend to have greater access to healthpromoting innovations.28 Although fundamental social cause theory was initially applied to understanding socio-economic disparities in health, race and ethnicity are strongly associated with access to resources and have thus been theorized to have similar effects on health.29 Consistent with this theoretical perspective, Polonijo and Carpiano (2013)21 find that racial/ethnic minority parents are less likely to know about the HPV vaccine. Furthermore, African Americans are less likely to receive a recommendation from a health professional to vaccinate their daughters. While recent longitudinal studies of females suggest that rates of vaccine uptake have increased across racial and ethnic groups,30,31 it remains unclear whether racial and ethnic disparities in HPV vaccination uptake have changed over time. It is possible that while all racial and ethnic groups are increasing vaccination, some racial and ethnic groups are increasing at faster or slower rates than other groups. Increasing disparities in HPV vaccination may forecast increasing racial and ethnic disparities in cervical cancer, and consequently may inform screening programs and other interventions. Conversely, decreasing disparities in HPV vaccination uptake may indicate that recent public health interventions focused on minority youths, particularly African American females,32 have been effective.

There is some concern that racial and ethnic trends in provider recommendations remain unexplored. While research has noted both gender and racial and ethnic variations in receiving a recommendation to vaccinate,21,26 scholars have yet to examine whether these disparities have changed since the introduction of the vaccine. Scholars also have yet to examine whether racial and ethnic trends in HPV vaccination uptake are explained by variations in provider recommendations. Understanding the mediating role of provider recommendations may inform new interventions focused on eliminating remaining disparities in HPV vaccination. If racial and ethnic variations in vaccination are driven by variations in provider recommendations, this would suggest that public health interventions should focus on healthcare providers, rather than adolescents and their parents. In this paper, we use data collected from a large, nationally representative sample of parents of adolescent children to examine racial and ethnic trends in HPV vaccination uptake and provider recommendations. Specifically, we address the following research questions: 1) Have HPV vaccination uptake and health provider recommendations increased over time among Black, Hispanic, and non-Hispanic White adolescent girls and boys? 2) Do trends in HPV vaccination uptake and health provider recommendations vary by race and ethnicity among adolescent girls and boys? 3) To what extent are the gender- and race-specific trends in HPV vaccination uptake explained by variations in provider recommendations? Given gender differences in the timing of the HPV vaccination recommendation,9,10 as well as previous research noting gender differences within racial and ethnic groups in vaccination initiation,24,30 all analyses are presented separately for males and females.

Methods Data Our study employs data from the 2008e2013 National Immunization Survey-Teen (NIS-Teen). The National Center for Immunization and Respiratory Diseases and the National Center for Health Statistics collected the telephone survey data from a random digit sample of parents of adolescent children.33 The NIS-Teen survey implemented a dual-frame sampling design with independent landline and cell phone samples. Response rates for these samples averaged 56.4% and 23.1%, respectively. The primary aim of the NIS-Teen is to estimate vaccine coverage for adolescents aged 13e17 years based on parental reports. Within recruited households, the parent or guardian who reported knowing the most about the adolescent's health was administered a survey asking about the child's vaccine history. With only a few rare exceptions (e.g. emancipated minors), teens under 18 must have parental consent in order to receive the HPV vaccine. Thus, parental reports are preferable for this age group.

p u b l i c h e a l t h 1 4 2 ( 2 0 1 7 ) 1 6 7 e1 7 6

Measures HPV vaccine uptake Our focal dependent variable reflects whether or not the respondent indicated that his or her adolescent child had ever received at least one dose of the HPV vaccination series (1 ¼ HPV vaccine uptake 0 ¼ did not initiate the HPV vaccine).

Provider recommendation Our proposed mediating variable reflects whether the respondent indicated that his or her adolescent child had ever received an HPV vaccination recommendation from a doctor or some other healthcare professional (1 ¼ has received a recommendation to vaccinate 0 ¼ has not received a recommendation to vaccinate).

Race and ethnicity Our focal independent variable is the race and ethnicity of the adolescent. Because we are primarily interested in comparing trends in HPV vaccine uptake across racial and ethnic groups, we have limited our subgroup analysis to the three wellrepresented groups in our sample, namely those who are: Hispanic (13%), Black (10%), and non-Hispanic White (69%). No other racial or ethnic group provided sufficient representation to allow for reliable estimates in our subgroup analyses, although we have retained these respondents (8%) in our analyses of the full sample.

Background characteristics Prior research has identified a number of correlates of HPV acceptability and uptake that may help identify barriers to HPV vaccine initiation, including the child's sex, age, maternal education (less than high school, high school degree, some college, college or higher) and household income (living at or below poverty level, greater than poverty but less than $75,000, greater than or equal to $75,000). We also include controls for a number of other potential confounding variables, including insurance status (privately insured, Medicaid, no insurance), the relationship of the respondent to the child, number of children in the household, the respondent's marital status (currently married, other status), maternal age (34 years or younger, 35e44 years, 45 years or older), and region of residence (South, Midwest, West, Northeast).

Analysis Descriptive statistics for all measures are presented in Table 1. The first and fifth columns display information on sample characteristics for the full sample stratified by gender. Columns 2e4 (females) and columns 6e8 (males) provide information on sample characteristics stratified by race and ethnicity. Fig. 1 displays the proportion of the sample receiving a recommendation to vaccinate by year, stratified by gender and race and ethnicity. Fig. 2 displays the proportion of the sample receiving at least one dose of the HPV vaccine by year, stratified by gender and race and ethnicity. Tables 2e5 present our multivariate trend analysis. Tables 2 and 3 examine racial and ethnic trends in provider recommendations, stratified by gender. Model 1 displays race and ethnic variation in provider recommendations, controlling

169

for sociodemographic characteristics. A positive and statistically significant effect for survey year in Model 1 would indicate a positive trend in provider recommendations. Model 2 examines whether the trends in provider recommendations vary by race and ethnicity. A significant effect for the interaction term in this model (‘Year  Race’) would indicate that trends in provider recommendations are inconsistent across race and ethnicity. Models 3e5 provide an alternative method of addressing our first two research questions, by examining within race and ethnic trends in provider recommendations to vaccinate. Tables 4 and 5 examine racial and ethnic trends in HPV vaccination uptake, stratified by gender. Model 1 displays race and ethnic variation in vaccination initiation, controlling for sociodemographic characteristics. Model 2 examines whether the trends in HPV vaccination uptake vary by race and ethnicity. Model 3 addresses our final research question by assessing the extent to which gender- and race-specific trends in HPV vaccination uptake are explained by variations in provider recommendations. A significant reduction in the interaction coefficients in Model 3 would indicate that race-specific trends are explained, at least in part, by variation in receiving a provider recommendation. Models 4e6 provide an alternate method of addressing our second research question, by examining within race and ethnic trends in HPV vaccine initiation. To formally assess mediation, we employ the Sobel statistic34 to test for significant changes in the odds of HPV vaccine uptake across nested models (i.e. before and after adjusting for mediators). All analyses are conducted in Stata 13,35 and all estimates were weighted using the protocol for the dual-frame sample weight detailed in the user guide, which can be found online.33

Results In the pooled sample, approximately 44% of females initiated the HPV vaccine, although the proportion is highest among Hispanic teens (48%) and lowest among African American adolescents (40%). Provider recommendations among females also appear to vary by race and ethnicity, ranging from approximately 52% among African Americans to 61% among non-Hispanic Whites. A slightly different trend exists among adolescent males. Only about 14% of teenage boys initiated the HPV vaccine, ranging from roughly 12% among nonHispanic Whites to almost 24% of Black and Hispanic adolescents. Provider recommendations for males are low in the pooled sample (22%), with a slightly lower percentage of nonHispanic Whites receiving a recommendation to vaccinate than their racial and ethnic minority counterparts. Fig. 1 shows that the proportion of the sample receiving a recommendation to vaccinate has increased steadily since the introduction of the HPV vaccine. While less than half of females received a provider recommendation to vaccinate in 2008, approximately 70% received a recommendation in 2013, although this proportion was higher for non-Hispanic White females (72%) than for Black (65%) or Hispanic (65%) teens. The increase in provider recommendation to vaccinate was even more dramatic in teen boys. While only about 5% received a recommendation to vaccinate in 2010, roughly 43% received

170

Table 1 e Descriptive statistics (mean/percentage ± SD), National Immunization Survey-Teen. Femalesa 1

2

Full sample (n ¼ 77,389)

a

44.10 59.04 13.16 10.16 68.51 8.16 14.97

Hispanic (n ¼ 10,188) 48.24 53.54

1.40

3 c

14.89

d,e e

1.40

4 d

Black (n ¼ 7864) 39.62 51.84

14.99

c,e e

1.39

5 e

White (n ¼ 53,019) 43.79 61.32

15

c,d c,d

1.40

6

Full sample (n ¼ 56,632)

Hispanic (n ¼ 7512)

13.94 22.13

19.65 23.88

14.99

13.26 9.43 69.05 8.26 14.94

1.41

7 f

g,h h

1.39

8 g

Black (n ¼ 5338) 18.08 23.72

f,h

14.98

1.41

h

Whiteh (n ¼ 39,102) 11.99 21.45

f,g

15.01

1.41

45.79 41.12 13.08

24.86 41.41 33.73

24.9 47 28.1

53.11 40.24 6.65

46.37 40.37 13.35

24.96 40.48 34.56

25.78 47.23 27

53.58 39.4 7.02

7.95 18.2 29.71 44.14 83.09 1.7 76.7

28.31 20.35 26.29 25.06 86.01 1.83 69.39

8.83 23.92 35.16 32.1 87.68 1.72 46.45

3.99 17.01 29.83 49.18 82.47 1.67 82.89

7.6 17.94 28.87 45.59 80.23 1.7 77.3

28.12 21.98 24.45 25.45 84.05 1.81 70.53

8.97 22.63 34.32 34.08 85.5 1.72 47.77

3.57 16.51 29.07 50.85 79.69 1.67 82.95

0.63

0.65

0.66

0.62

0.63

0.63

0.66

7.06 42.33 50.61

12.59 50.18 37.23

14.11 45.41 40.48

4.81 40.52 54.67

6.85 41.68 51.48

12.69 49.67 37.65

12.8 46.67 40.54

4.75 39.29 55.96

36.09 22.27 22.65 18.99

41.72 12.87 31.65 13.76

59.1 17.9 5.01 17.98

32.38 25.39 21.83 20.5

36.41 21.26 23.43 18.9

42.35 11.58 32.44 13.63

58.64 16.8 5.79 18.77

32.84 24.44 22.25 20.47

f,g

0.62

2008e2013, b2010e2013 shown are variable, means/proportions, and standard deviations (SD). Additional superscripts identify significant differences in the dependent variables between designated classes (z-score, P < 0.05).

p u b l i c h e a l t h 1 4 2 ( 2 0 1 7 ) 1 6 7 e1 7 6

Vaccine uptake Provider recommendation Race and ethnicity Hispanic Black Non-Hispanic White Other Age in years Federal poverty level (FPL) 75k Greater than poverty, less than 75k Below federal poverty level Mother's education Less than high school High school only Some college College graduate Mother respondent Number of children in household Married Mother's age in years 35 or younger 35e44 45 Census region South Midwest West Northeast

Malesb

p u b l i c h e a l t h 1 4 2 ( 2 0 1 7 ) 1 6 7 e1 7 6

171

Fig. 1 e Proportion of the sample receiving a recommendation to vaccinate by year.

an HPV vaccination recommendation in 2013. By 2013, Black males were slightly more likely to receive a recommendation to vaccinate (47%) than their Hispanic (43%) or non-Hispanic White (43%) counterparts.

Fig. 2 shows that the proportion of adolescents initiating the HPV vaccine has also increased over time, although not to the same degree as provider recommendations. Approximately 35% of females initiated the vaccine in 2008. By 2013,

Fig. 2 e Proportion of the sample receiving at least one dose of the HPV vaccine by year.

172

p u b l i c h e a l t h 1 4 2 ( 2 0 1 7 ) 1 6 7 e1 7 6

Table 2 e Logistic regression estimates for females predicting provider recommendation, National Immunization SurveyTeen. Full sample (n ¼ 77,389)

Year Racea Hispanic Black Other Age Household incomeb 75k Greater than poverty,