Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Jul 26.
Published in final edited form as: Med Care. 2016 Jun;54(6):570–577. doi: 10.1097/MLR.0000000000000544

Racial/Ethnic Disparities in Influenza Vaccination of Chronically-ill U.S. Adults: The Mediating Role of Perceived Discrimination in Healthcare

William K Bleser 1, Patricia Y Miranda 2, Muriel Jean-Jacques 3
PMCID: PMC6060271  NIHMSID: NIHMS981444  PMID: 27172536

Abstract

BACKGROUND:

Despite well-established programs, influenza vaccination rates in U.S. adults are well below federal benchmarks and exhibit well-documented, persistent racial and ethnic disparities. The causes of these disparities are multifactorial and complex, though perceived racial/ethnic discrimination in healthcare is one hypothesized mechanism.

OBJECTIVES:

To assess the role of perceived discrimination in healthcare in mediating influenza vaccination disparities in chronically-ill U.S. adults (at high-risk for influenza-related complications).

RESEARCH DESIGN:

We utilized 2011–2012 data from the Aligning Forces for Quality Consumer Survey on health and healthcare (n=8,127), nationally-representative of chronically-ill U.S. adults. Logistic regression marginal effects examined the relationship between race/ethnicity and influenza vaccination, both unadjusted and in multivariate models adjusted for determinants of health service use. We then used binary mediation analysis to calculate and test the significance of the percentage of this relationship mediated by perceived discrimination in healthcare.

RESULTS:

Respondents reporting perceived discrimination in healthcare had half the uptake as those without discrimination (32% vs. 60%, p=0.009). The change in predicted probability of vaccination given perceived discrimination experiences (vs. none) was large but not significant in the fully-adjusted model (−0.185, 95% CI: −0.385, 0.014). Perceived discrimination significantly mediated 16% of the unadjusted association between race/ethnicity and influenza vaccination, though this dropped to 6% and lost statistical significance in multivariate models.

CONCLUSIONS:

The causes of persistent racial/ethnic disparities are complex and a single explanation is unlikely to be sufficient. We suggest re-evaluation in a larger cohort as well as potential directions for future research.

Keywords: chronic disease, discrimination, health disparities, influenza, vaccinations

Introduction

Adults at highest risk for influenza-related complications and mortality include those aged 65+ years and those of any age with chronic illnesses, including asthma, chronic lung disease, heart disease, and diabetes.1 Adult influenza vaccination rates in the U.S., however, remain well below federal benchmarks of optimal coverage, especially among high-risk adult populations;2,3 e.g., the Healthy People 2020 target is 70% uptake,4 though between 2004–2012 uptake among high-risk working-age adults peaked at 45.6%.2 Influenza results in up to 200,000 hospitalizations, 49,000 deaths5,6 and an estimated $87 billion in total economic burden to the U.S. each year.7 Chronically-ill adults share a disproportionately large portion of this burden.8

Furthermore, influenza vaccination uptake in U.S. adults exhibits well-documented, persistent racial/ethnic disparities. These disparities remain after adjusting for many sociodemographic characteristics9 and despite influenza vaccines’ continued safety record,10 increasing affordability and accessibility,11 and strong evidence that influenza vaccination is the most effective means of preventing influenza and influenza-related outcomes in both general and chronically-ill adult populations.2,12–17 For example, uptake for Blacks and Hispanics aged 65+ was at least 10 percentage points lower than Whites for most years of the previous two decades.9,18 Similar disparities have been documented among working age adults,2 including those with high-risk conditions.3,19 Many conditions that increase the risk of severe influenza illness or death are more prevalent among racial/ethnic minorities (e.g., asthma, diabetes, HIV/AIDS, and hypertension),20 making racial/ethnic disparities in influenza vaccination of even greater consequence, placing inequitably higher mortality on U.S. minority populations.21

The causes of racial/ethnic disparities in influenza vaccination are likely multifactorial and complex.22 Potential explanations include circumstances pertaining to many minorities: lower healthcare access; greater distrust of physicians; potential unconscious racial bias in providers which may result in differential provision of vaccination; minorities may see providers who are generally less inclined to vaccinate; and resistant attitudes/beliefs towards influenza shots.23,24 Research elucidating these mechanisms is limited.21,22

Broadly, racial discrimination and racism are thought to contribute to racial/ethnic disparities in health.25 Measures of perceived discrimination have been used to examine pathways through which they can affect health. How perceived discrimination, as it is experienced in daily life26 and specifically within healthcare settings,27 affects health disparities has received increasing empirical attention across multiple health outcomes. However, it is understudied as a mechanism potentially explaining racial/ethnic disparities in influenza vaccination in U.S. adults. The few studies that have examined this association report mixed results and are generalizable only to limited settings.28–30

This study examines the role of perceived racial/ethnic discrimination in healthcare as a mediator of the association between race/ethnicity and influenza vaccination. It addresses gaps in the literature by: (1) using a nationally-representative sample of high-risk, chronically-ill adults; and (2) applying mediation analysis to assess a potential mechanism (perceived discrimination) hypothesized to explain influenza vaccination disparities.

Methods

Data Source and Sample

Data come from the Consumer Survey (CS) of the Aligning Forces for Quality (AF4Q) initiative. AF4Q was a Robert Wood Johnson Foundation (RWJF) initiative to improve healthcare quality, reduce racial/ethnic health disparities, and provide models for national reform in 16 geographically, demographically, and economically diverse targeted communities representing 12.5 percent of the U.S. population.31 The CS is part of an independent, scientific evaluation of AF4Q. Additional details about the AF4Q program and evaluation have been published elsewhere.32,33 The CS uses random digit dial design to generate a representative sample of adults 18+ years living in AF4Q communities, as well as a national comparison sample of respondents not in AF4Q communities, with at least one of five chronic conditions – asthma, depression, diabetes, heart disease, and hypertension. Survey questions focus on patient engagement in and knowledge regarding their own health, skills and willingness to manage their health, ability to be an effective healthcare consumer in the context of physician visits, and other related topics.33 Two rounds were administered – during the first year of AF4Q from (July 2007-August 2008), and from July 2011-November 2012 of new respondents and first-round respondents who agreed to be re-interviewed.34 AF4Q evaluation research is reviewed and approved by the Pennsylvania State University Institutional Review Board, including all data analysis or use.

This study uses the 9,737 second round respondents. The overall response rate for this round was 39.7% (American Association of Public Opinion method) to 42.1% (Council of American Survey Research Organizations method). This was restricted to those in the original 15 study markets and national comparison sample so sample weights could be applied to obtain national representativeness (n=9,527). This was further restricted to those of non-Hispanic White, non-Hispanic Black, and Hispanic race/ethnicity, and not missing on influenza vaccination nor perceived discrimination in healthcare, equating to a total eligible analytic sample of 8,850.

Measures

Since the decision to receive an influenza vaccine is one to receive a preventive health service, Andersen’s Behavioral Model of Health Services Use35(fig1.1) is employed to conceptually-ground the selection of determinants of health services utilization as covariates. Andersen’s Model identifies factors predisposing, enabling, and creating need for health services and how health services use is influenced by health behaviors and processes of medical care; it has been used extensively in studies across various healthcare sectors, and in the context of many outcomes.35(pp3−10),36

Influenza vaccination.

The outcome variable is self-reported receipt of the seasonal influenza vaccine within the previous 12 months (dichotomous), the same as used in the National Health Interview Survey.37 Though subject to recall bias, the annual nature of influenza vaccination likely decreases such bias relative to other vaccines chronically-ill patients may receive.

Race/ethnicity.

Race/ethnicity, a pre-disposing factor of health services use, is the independent variable of interest. Respondents were asked, “What race or races do you consider yourself to be?” and selected all that apply: White (Caucasian); Black or African American; Asian; American Indian or Alaska Native; Native Hawaiian or Other Pacific Islander; Hispanic; or Other (specify). Less than 4% of respondents selected >1 race; in these cases were classified in a new “Multiracial” category. Respondents were also asked, “Are you of Hispanic or Latino origin or descent?” In the rare case that respondents indicated Hispanic race but not Hispanic ethnicity, they were reclassified as Hispanic ethnicity. These two variables were recoded to generate one mutually-exclusive race/ethnicity variable: non-Hispanic White (no other race indicated), non-Hispanic Black or African American (no other race indicated), or Hispanic (of any or multiple races).

Perceived discrimination in healthcare.

The mediating variable is self-reported perceived discrimination in healthcare, a process of care factor of health services use. Respondents were asked, “Do you think there was ever a time when you would have gotten better medical care if you had belonged to a different race or ethnic group?” This question has been used in the California Health Interview Survey and the Commonwealth Fund Health Care Quality Survey, and very similar valid and reliable measures have been used elsewhere.27...

Covariates.

Other covariates represent determinants of health services use from Andersen’s Model and others relevant to influenza vaccination. This includes fifteen total measures. Six factors pre-dispose individuals to health services use: (1) sex; (2) education; (3) age (classified as 18–64 vs. 65 or older given the latter is a high-risk populations for influenza-related complications8); (4) employment status; (5) Patient Activation Measure (PAM) stage, a scale reflecting engagement in one’s own health;38,39 and (6) AF4Q sample market. Two factors are enabling conditions facilitating or impeding health services use: (7) poverty status; and (8) health insurance status. Two factors represent perceived or evaluated needs or conditions requiring health services use: (9) self-rated health; and (10) a summative index of the total number of chronic illnesses the respondent has. Five variables represent health behavior level factors (i.e., personal health practices, processes of medical care, and use of personal health services): (11) tobacco smoking status; (12) a diet and exercise index representing the number of recommended practices to which each individual conforms: exercises regularly, and eats at least 5 servings of fruits or vegetables most days of the week; (13) trust the individual has in information from their doctor (not specific to vaccination); (14) number of visits made to healthcare providers to treat their conditions, previous 3 months; and (15) rating of all care received from all healthcare professionals, previous 12 months.

Statistical Analyses

To make all models comparable, analyses were restricted to respondents not missing any covariates. This “complete case analytic sample” contains 8,127 respondents; 723 respondents (8.17% of the eligible analytic sample) were considered missing. Missingness across the variables was low; the largest contributors were poverty (3.87% missing) and diet and exercise (1.32%); missingness in the remaining variables ranged from 0.01–0.64%. Missingness was not associated with influenza vaccination, discrimination, nor race/ethnicity. However, the 723 missing respondents were significantly more likely to be older, retired, and non-smokers.

We first calculated bivariate statistics to examine the association between influenza vaccination with all covariates individually using design-based F-tests for categorical variables and adjusted Wald tests for continuous and ordinal variables. We then conducted logistic regression of influenza vaccination onto race/ethnicity in several models to examine disparities. First, in unadujsted analyses, we regressed influenza vaccination onto race/ethnicity alone (Model 1) and then added in perceived discrimination in healthcare (Model 2) to see how it is associated with influenza vaccination and how it changes racial/ethnic disparities in influenza vaccination. Models 3 and 4 add to Models 1 and 2, respectively, all other determinants of health services use to generate multivariate results.

Lastly, we test for mediation. Mediation occurs when the effect of an independent variable (race/ethnicity) on a dependent variable (influenza vaccination) is transmitted by a mediating variable (perceived discrimination in healthcare). If path “a” is the effect of the independent variable on the mediator (the effect of race/ethnicity on perceived discrimination; i.e. racial/ethnic disparities in influenza vaccination), and path “b” is the effect of the mediator on the dependent variable adjusting for the independent variable (the effect of perceived discrimination on influenza vaccination status adjusting for race/ethnicity), the recommended way to assess mediation is by multiplying the regression coefficients of paths “a” and “b” together and testing if the a*b cross-product is significantly different from zero.40–44 This process is sometimes referred to as the four-step process and the test as the Sobel test (e.g., see Baron & Kenny,45 James & Brett,46 and Judd & Kenny47). However, when analyzing mediation with a binary mediator and/or dependent variable (“binary mediation”), the aforementioned coefficients calculated via logistic regression are not always correct and require transformation; the recommended approach is a modified Sobel test that calculates a standardized a*b cross-product suitable for logistic regression.40 We calculated the correct mediation coefficients as recommended above, the corresponding percentage of mediation occurring, and tested whether or not it was different from zero.44,48 Using bootstrapping (500 repetitions) to obtain correct standard errors,43,49,50 we calculated the percentage of the unadjusted (Model 1 vs. 2) and adjusted (Model 3 vs. 4) relationship between race/ethnicity and influenza vaccination mediated by perceived discrimination in healthcare and tested if the mediation was statistically significant. All statistical analyses used Stata/SE 13.1 statistical software,51 and Stata’s svy commands to obtain nationally-representative estimates adjusting for complex survey design.

Results

Table 1 contains descriptive statistics of the complete case analytic sample of chronically-ill U.S. adults in 2011–2012 overall, and by influenza vaccination status, race/ethnicity, and experiences of perceived discrimination in healthcare. Overall, 58.2% of respondents reported receipt of the influenza vaccination within the previous 12 months and 7.3% reported ever experiencing a time when they perceived they would have gotten better medical care if they had belonged to a different race or ethnic group. Column 2 of Table 1 shows the unadjusted correlations of determinants of health services with influenza vaccination uptake. Uptake was significantly lower across several covariates, individuals: who ever experienced perceived discrimination in healthcare; without a four-year college degree; of working age; not retired; below the poverty line; without health insurance; currently smoking; and reporting less trust in information from their doctor.

Table 1.

Sample Summary Statistics by Influenza Vaccination Status, Race/Ethnicity, and Experiences of Perceived Discrimination in Healthcare: AF4Q Consumer Survey, Chronically-ill U.S. Adults (2011–2012)

Total Outcome Independent Variable of Interest Moderator
Characteristic Complete
case sample
N = 8,127		 Vaccinated,
prev. 12 months
N = 5,197		 Non-Hispanic
White race
N = 5,505		 Non-Hispanic
Black race
N = 2,112		 Hispanic
ethnicity
N = 510		 Discrimination,
prev. 12 months
N = 597
Influenza vaccination (prev. 12 months) 58.2% -- 60.3% 43.0% 62.1% 32.4%
Race/ethnicity
    Non-Hispanic White 74.8% 77.5% -- -- -- 36.5%
    Non-Hispanic Black 13.5% 9.9% -- -- -- 28.2%
    Hispanic 11.8% 12.6% -- -- -- 35.4%
Perceived racial/ethnic discrimination in healthcare 7.3% ** 4.1% 3.6% 15.3% 21.9% --
Female sex 56.6% 54.7% 53.3% 66.5% 65.7% 62.2%
Education: obtained 4-year college degree 24.7% * 29.8% 29.7% 6.3% 14.0% 7.1%
Age (years)
    18-64 74.5% *** 68.0% 71.0% 86.8% 83.2% 93.4%
    65 and older 25.5% *** 32.0% 29.0% 13.2% 16.8% 6.4%
Employment status
    Full-time (30+ hours/week) 39.9% * 38.0% 41.5% 30.1% 40.8% 20.5%
    Part-time (<30 hours/week) 11.4% * 11.3% 13.2% 8.2% 4.2% 11.4%
    Retired 21.2% * 27.1% 23.4% 17.1% 12.0% 12.7%
    Other 27.4% * 23.6% 21.9% 44.5% 43.0% 55.4%
Patient Activation Measure stage, mean (SD) 3.2 (0.8) 3.2 (0.8) 3.2 (0.8) 3.1 (0.8) 3.1 (0.9) 2.8 (1.1)
Household income above federal poverty line 67.6% * 73.2% 77.9% 30.2% 45.2% 28.4%
Currently insured 85.4% * 90.3% 90.9% 69.3% 68.5% 57.6%
Self-rated health, mean (SD) 3.0 (1.0) 3.0 (1.0) 3.1 (1.0) 2.6 (0.8) 2.4 (0.9) 2.0 (1.2)
Number of chronic illnesses, mean (SD) 1.4 (0.9) 1.5 (0.9) 1.4 (0.9) 1.3 (0.8) 1.6 (0.8) 1.6 (0.8)
Current tobacco smoker 18.4% *** 11.6% 15.7% 38.3% 12.5% 34.9%
Diet and exercise score, mean (SD) 1.3 (0.7) 1.3 (0.7) 1.2 (0.7) 1.3 (0.7) 1.5 (0.7) 1.1 (0.8)
Trust in physician score, mean (SD 1.3 (0.5) * 1.3 (0.5) 1.3 (0.5) 1.3 (0.5) 1.4 (0.5) 1.4 (0.5)
Visits to providers, prev. 3 months, mean (SD) 1.1 (1.2) 1.1 (1.2) 1.0 (1.2) 1.4 (1.4) 1.4 (1.2) 2.1 (1.3)
Rating of all care, prev. 12 months, mean (SD) 8.1 (2.1) 8.3 (1.8) 8.0 (2.0) 8.7 (1.6) 7.6 (3.0) 5.5 (3.2)
Open in a new tab

Column percentages and means are weighted to be nationally-representative of chronically-ill U.S. adults. Due to missing values and rounding, numbers in the column percentages in this table may not add up to the total sample size. Sample size values (N) are unweighted to show actual number of respondents in cells. For the outcome column, bivariate statistical tests were conducted to detect, by influenza vaccination uptake, significant differences among group percentages (design-based F tests) and among means (adjusted Wald tests). In this column, *0.05>p≥0.01; **0.01>p≥0.001; ***p<0.001

Patient Activation Measure stage ranges from 1–4 where 4 indicates higher patient activation.

Self-rated health ranges from 1–5 where 5 indicates excellent health.

Number of chronic illnesses rages from 1–5 (asthma, depression, diabetes, heart disease, and/or hypertension).

Diet and exercise index ranges from 0–2 where higher scores indicate better exercise and diet practices.

Trust in physician score ranges from 1–3, where 1 is “a lot” and 3 is “not at all.”

Number of visits to healthcare providers in the previous 3 months to treat conditions is categorized as 0, 1, 2, 3, or 4+ times.

Rating of all care received from all healthcare professionals in the previous 12 months ranges from 0–10 where 10 is the best possible score.

Table 2 shows marginal changes in predicted probability of self-reported influenza vaccination associated with changes in covariates from logistic regression analyses. In Model 1, the predicted probability of vaccination was significantly lower among non-Hispanic Blacks relative to non-Hispanic Whites (−0.173, [95%CI: −0.332, −0.014]). In Model 2, having experienced perceived discrimination was significantly associated with lower predicted probability of influenza vaccination (−0.273 [−0.493, −0.052]); the racial/ethnic disparities were similar to Model 1 but lost statistical significance.

Table 2.

Marginal Changes in Predicted Probability of Influenza Vaccination from Weighted, Logistic Regression Models: AF4Q Consumer Survey, Chronically-Ill US Adults (2011–2012)

Characteristic Model 1
N = 8,127
ΔProb. (95% CI)		 Model 2
N = 8,127
ΔPROB. (95% CI)		 Model 3
N = 8,127
ΔPROB. (95% CI)		 Model 4
N = 8,127
ΔPROB. (95% CI)
Race/ethnicity (ref: non-Hispanic White)
    Non-Hispanic Black * −0.173 (−0.332, −0.014) −0.142 (−0.308, 0.023) −0.066 (−0.214, 0.081) −0.046 (−0.194, 0.102)
    Hispanic 0.018 (−0.167, 0.203) 0.070 (−0.082, 0.221) 0.087 (−0.064, 0.239) 0.112 (−0.031, 0.255)
Perceived racial/ethnic discrimination in healthcare -- * −0.273 (−0.493, −0.052) -- −0.185 (−0.385, 0.014)
Male sex (ref: female) -- -- 0.008 (−0.096, 0.112) 0.008 (−0.095, 0.111)
Education: obtained 4-year college degree -- -- * 0.145 (0.017, 0.274) * 0.146 (0.018, 0.273)
Aged 65 or older (ref: 18-64 years) -- -- 0.124 (−0.001, 0.248) 0.115 (−0.009, 0.240)
Employment status (ref: full-time: 30+ hours/week)
    Part-time (<30 hours/week) -- -- 0.058 (−0.104, 0.221) 0.065 (−0.093, 0.223)
    Retired -- -- 0.084 (−0.066, 0.234) 0.093 (−0.055, 0.241)
    Other -- -- 0.006 (−0.126, 0.139) 0.013 (−0.117, 0.144)
Patient Activation Measure stage -- -- −0.040 (−0.110, 0.031) −0.038 (−0.106, 0.030)
Household income above federal poverty line -- -- 0.046 (−0.076, 0.167) 0.041 (−0.079, 0.162)
Currently insured (ref: not) -- -- 0.137 (−0.028, 0.302) 0.128 (−0.030, 0.286)
Self-rated health -- -- −0.034 (−0.094, 0.026) −0.036 (−0.096, 0.024)
Number of chronic illnesses -- -- 0.031 (−0.037, 0.098) 0.033 (−0.033, 0.099)
Not a current tobacco smoker (ref: current smoker) -- -- * 0.144 (0.032, 0.255) * 0.143 (0.033, 0.253)
Diet and exercise score -- -- −0.002 (−0.074, 0.071) −0.003 (−0.073, 0.068)
Trust in physician score -- -- ** −0.137 (−0.228, −0.046) ** −0.139 (−0.227, −0.050)
Visits to providers, prev. 3 months -- -- 0.018 (−0.018, 0.053) 0.022 (−0.014, 0.058)
Rating of all care, prev. 12 months -- -- 0.025 (−0.003, 0.052) 0.019 (−0.008, 0.046)
Open in a new tab

Predicted probabilities are calculated using dy/dx marginal effects post-estimation commands following logistic regression models listed at the top of each column (Models 1 and 2 are unadjusted logistic regression models without and with the mediator variable, respectively; Models 3 and 4 add to Models 1 and 2, respectively, all covariates and survey market). Logistic regression models (and the resulting predicted probabilities) are weighted to be nationally-representative of chronically-ill U.S. adults adjusting for complex survey design. *0.05>p&≥0.01; **p<0.01

Patient Activation Measure stage ranges from 1-4 where 4 indicates higher patient activation.

Self-rated health ranges from 1-5 where 5 indicates excellent health.

Number of chronic illnesses rages from 1-5 (asthma, depression, diabetes, heart disease, and/or hypertension).

Diet and exercise index ranges from 0-2 where higher scores indicate better exercise and diet practices.

Trust in physician score ranges from 1-3, where 1 is “a lot” and 3 is “not at all.”

Number of visits to healthcare providers in the previous 3 months to treat conditions is categorized as 0, 1, 2, 3, or 4+ times.

Rating of all care received from all healthcare professionals in the previous 12 months ranges from 0-10 where 10 is the best possible score.

Results from Model 3 were nearly identical to Model 4 so only the latter is reported here. In Model 4, those experiencing perceived discrimination in healthcare had a large decrease in predicted probability of influenza vaccination (−0.185), though this was not statistically significant. Associated with higher predicted probability of influenza vaccination were obtaining a four-year college degree (0.146 [0.018, 0.273]), and not currently smoking (0.143 [0.033, 0.253]), and associated with lower predicted probability of vaccination was having less trust in information from one’s doctor (−0.139 [−0.227, −0.050]). In both models, there were no statistically significant racial/ethnic disparities, though non-Hispanic Blacks had the lowest predicted probabilities of vaccination (−0.046 to −0.066 lower probability than predicted for non-Hispanic Whites).

Table 3 shows the results of the binary mediation analysis. In sum, perceived discrimination in healthcare significantly mediated 16.2% of the unadjusted relationship between influenza vaccination and race/ethnicity. Adjusting for determinants of health services, the amount of the relationship between influenza vaccination and race/ethnicity mediated by perceived discrimination in healthcare dropped to 6.0% and lost statistical significance.

Table 3.

Binary mediation results for testing the mediation of racial/ethnic disparities in influenza vaccination by perceived discrimination in healthcare: AF4Q Consumer Survey, Chronically-Ill US Adults (2011–2012)

Binary mediation results Unadjusted for covariates
N = 8,127		 Adjusted for covariates
N = 8,127
a 0.3442 0.3368
b −0.0567 −0.0145
a*b (indirect effect of X on Y) −0.0195 −0.0049
95% confidence interval of a*b (−0.0286, −0.0114) (−0.0146, 0.0036)
Total effect of X on Y −0.1203 −0.0808
Percent total effect mediated (percent indirect of total effect) 16.2% 6.0%
Open in a new tab

Results weighted to be nationally-representative of chronically-ill U.S. adults adjusting for complex survey design Results presented are transformed binary mediation coefficients a, b, a*b (the indirect effect; i.e., the mediation effect), the 95% confidence interval of a*b (to see if it is different from zero), and the percent indirect effect of the total effect (i.e., the percent of the relationship between race/ethnicity and influenza vaccination mediated by experiences of perceived discrimination in healthcare).

Discussion

This study contributes to the very few examining the association of perceived discrimination in healthcare and influenza vaccination and to our knowledge is the only examining perceived discrimination as a mediator of racial/ethnic influenza vaccination disparities. The CDC reports estimates of influenza vaccination uptake among high-risk working-age adults (18–64) and those over 65. Among high-risk working-age adults, uptake was 46.7% in the 2010–2011 season and 45.2% in the 2011–2012 season, and for adults 65 and older, the comparable uptake was 66.6% and 64.9%, respectively.52,53 Uptake in our sample, which spanned these two seasons, was slightly higher (53.1% in working age adults and 73.1% among those 65 and older). The unadjusted racial/ethnic disparities in uptake in our sample, though not statistically significant, are comparable to the magnitude noted over the previous two decades in U.S. adults.2,3,9,18,19

The three determinants that remained significantly associated with higher predicted probability of vaccination in the fully-adjusted model were attainment of a four-year college degree, not being a current smoker, and having greater trust in information provided by one’s doctor. Higher levels of education are generally associated with increased uptake of influenza vaccination in the literature.54 Tobacco smoking is linked to increased influenza severity, complications, and mortality,55,56 suggesting smokers might have increased uptake via increased perceived vaccine need. However, in line with this study, smokers are generally less likely to be vaccinated, the reasons of which are not well understood57 but may relate to influenza vaccines’ reduced efficacy in smokers.58–60 Last, a trustful physician-patient relationship is necessary for physicians to successfully recommend vaccines and distrust of doctors is frequently listed as a reason of vaccine hesitancy.61 Because trust in one’s physician and discrimination are conceptually related, we performed sensitivity analyses. The correlation between trust and discrimination was low (r=0.07) and upon re-running our main analyses removing the trust variable, none of our findings changed, suggesting that they function as two distinct constructs. That these three findings remained significant despite this study’s power limitations (see below) suggests they may be the strongest factors associated with influenza vaccination in chronically-ill U.S. adults and warrant further investigation regarding their role in disparities.

This study has three limitations. The first relates to how the survey was sampled. The AF4Q CS is representative of 15 healthcare markets (~1/8th of chronically-ill US adults) with a national comparison sample representative of the remaining 7/8ths. Though weights can be used to generate national-representativeness, it is not explicitly designed as a national probability sample, presenting design effect limitations. Design effect statistics (DEFF) from main analysis coefficients (Model 4) ranged from 4.6–37.4; corresponding root design effect statistics (DEFT) indicate standard errors are approximately 2–6 times larger than a corresponding simple random sample. While not ideal, it is not uncommon to have DEFF coefficients in this range, especially among cluster-sampled surveys62 like the CS. The implication is the loss of statistical power in nationally-representative estimates. Further, CS response rates were relatively low, though in accordance with general declining trends in survey response rates.63 This issue was previously examined and no significant differences were found in CS respondents’ demographic characteristics nor prevalence of chronic illnesses compared to corresponding years of the National Health Interview Survey,63 a related survey with higher response rates. Moreover, other data sources studying racial/ethnic discrimination in healthcare cannot be used to provide reliable national estimates,27 making this study a contribution despite design limitations. The second limitation relates to key variable categorization. There are differing measures of racial/ethnic discrimination in the literature. It is generally recommended that such measures ideally have both a setting and a timeframe.28,64 Though our measure does not specify a timeframe, it does specify a setting (healthcare visits), allowing us to address our study aims. There are more nuanced measures in the literature27 (e.g., asking about exact actions perceived to be discriminatory, or frequency/intensity of experiences), but since only 7.5% of respondents reported ever experiencing perceived discrimination in healthcare, it is to our benefit to have a broader capture. Moreover, two very similarly worded questions from other sources have been tested for reliability and validity in minority populations,27 increasing confidence in our measure. Future studies should have a specific timeframe (typically, the past 1–2 years) and specific mechanisms of discrimination but therefore will likely need a sample quite larger than ours. Small sample size among races other than White and Black and among Hispanics prevented further stratification of the race/ethnicity variable, though our categorization allows us to make comparisons to national estimates.2,3,9 We performed a sensitivity analysis to examine potential effects of ethnic heterogeneity by separating Hispanics into those of Mexican origin (largest Hispanic subgroup, also noted as using less preventive services compared to other Hispanics65) and Hispanics of other origin; our results did not change. The last limitation relates to temporality. Our results are associative and not causative given this study is cross-sectional. The CS has a sub-section of panel respondents but two factors made them impractical to utilize. First, race/ethnicity is not time-varying, preventing fixed effects analyses. Second, the prevalence of respondents ever experiencing discrimination is already small; any change in its prevalence in the few years since the prior round would provide insufficient variation to use other longitudinal methods. Further, the possibility of bidirectionality in our findings is very unlikely. Because race/ethnicity is a fixed characteristic, it can neither be caused by vaccination status nor discrimination. It is theoretically possible that a respondent could opt to not be vaccinated, causing their healthcare provider to treat them differently, causing the respondent to attribute this perceived difference in care to their race/ethnicity. However, Andersen’s model suggests this scenario to be unlikely: pre-disposing factors (race/ethnicity) precede process of care factors (discrimination in healthcare), which precede health services utilization (vaccination).

Racial and ethnic disparities in healthcare exist broadly in the U.S. and have significant implications. They erode trust in health professionals, hinder the ability of disadvantaged racial/ethnic groups to advance economically and professionally, threaten public health efforts to improve the nation’s health, worsen national healthcare expenditures, and generally tear at the nation’s social fabric pertaining to perception of racial discrimination.66 Racism and racial discrimination is thought to be one mechanism underlying persistent racial/ethnic disparities in health and shown to be associated with numerous health and health services outcomes, such as higher mortality, blood pressure, body mass index, weight gain, and mental health outcomes and worse levels of cancer screening, diabetes care, general care-seeking behavior, and unmet healthcare needs.27 This study found that experiences of perceived discrimination in healthcare significantly mediated 16% of unadjusted racial/ethnic disparities in influenza vaccination. Though this mediation dropped to 6% and loss statistical significance in fully-adjusted models, the magnitude of the adjusted association between discrimination and vaccination was larger than that of other significant findings. This suggests that lack of significance is likely due to the low prevalence of discrimination and that this effect may be significant in future, larger samples. Clearly the causes of influenza vaccination disparities are complex as they have remained for decades, and from this study, the role discrimination plays is not entirely clear. Nonetheless, the overemphasis on main-effect multivariate models in health disparities research and subsequent lack of focus on the processes contributing to these disparities is a limitation of the literature.67 Though this study raises numerous questions for further research, understanding the mechanisms of persistent influenza vaccination disparities among chronically-ill adults is of particular importance since this population is at higher risk for influenza-related complications and because they have higher potential for gains.

Acknowledgments

Disclosure of funding

This research was primarily supported by a grant for the Aligning Forces for Quality evaluation provided by the Robert Wood Johnson Foundation. We also acknowledge support from the Department of Health Policy and Administration at Pennsylvania State University, and assistance provided by the Population Research Institute at Penn State University, which is supported by an infrastructure grant from the National Institutes of Health (2R24HD041025–11).

Contributor Information

William K. Bleser, Department of Health Policy and Administration, Pennsylvania State University. 501-A Ford Building, Pennsylvania State University, University Park, PA 16802.

Patricia Y. Miranda, Department of Health Policy and Administration, Pennsylvania State University. 601-G Ford Building, Pennsylvania State University, University Park, PA 16802.

Muriel Jean-Jacques, Division of General Internal Medicine, Northwestern University Feinberg School of Medicine. 750 N. Lake Shore, Rubloff Building 10th Floor, Division of General Internal Medicine and Geriatrics, Department of Medicine, Northwestern University Feinberg School of Medicine, Chicago, IL, 60611.

References

  • 1.Centers for Disease Control and Prevention. People at High Risk of Developing Flu-Related Complications. http://www.cdc.gov/flu/about/disease/high_risk.htm. Published January 8, 2015 Accessed March 13, 2015.
  • 2.Lu P, Singleton JA, Euler GL, Williams WW, Bridges CB. Seasonal Influenza Vaccination Coverage Among Adult Populations in the United States, 2005–2011. Am J Epidemiol. 2013;178(9):1478–1487. doi: 10.1093/aje/kwt158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Centers for Disease Control and Prevention. Influenza vaccination levels among persons aged > or =65 years and among persons aged 18–64 years with high-risk conditions--United States, 2003. MMWR Morb Mortal Wkly Rep. 2005;54(41):1045–1049. [PubMed] [Google Scholar]
  • 4.U.S. Department of Health and Human Services. Immunization and Infectious Diseases. https://www.healthypeople.gov/2020/topics-objectives/topic/immunization-and-infectious-diseases/objectives. Published April 1, 2015 Accessed April 3, 2015.
  • 5.Centers for Disease Control and Prevention. CDC - Seasonal Influenza (Flu) - Questions and Answers Regarding Estimating Deaths from Influenza in the United States. http://www.cdc.gov/flu/about/disease/us_flu-related_deaths.htm. Published June 24, 2011 Accessed March 24, 2013.
  • 6.Thompson WW, Shay DK, Weintraub E, et al. Influenza-associated hospitalizations in the United States. JAMA J Am Med Assoc. 2004;292(11):1333–1340. doi: 10.1001/jama.292.11.1333. [DOI] [PubMed] [Google Scholar]
  • 7.Molinari N-AM, Ortega-Sanchez IR, Messonnier ML, et al. The annual impact of seasonal influenza in the US: measuring disease burden and costs. Vaccine. 2007;25(27):5086–5096. doi: 10.1016/j.vaccine.2007.03.046. [DOI] [PubMed] [Google Scholar]
  • 8.Centers for Disease Control and Prevention. Influenza In: Atkinson W, Wolfe C, Hamborsky J, eds. Epidemiology and Prevention of Vaccine-Preventable Diseases. 13th ed. Washington, DC: Public Health Foundation; 2015:187–208. http://www.cdc.gov/vaccines/pubs/pinkbook/downloads/flu.pdf. Accessed May 26, 2015. [Google Scholar]
  • 9.Centers for Disease Control and Prevention. Racial/ethnic disparities in influenza and pneumococcal vaccination levels among persons aged > or =65 years--United States, 1989–2001. MMWR Morb Mortal Wkly Rep. 2003;52(40):958–962. [PubMed] [Google Scholar]
  • 10.Centers for Disease Control and Prevention. Influenza Vaccine Safety. http://www.cdc.gov/flu/protect/vaccine/vaccinesafety.htm. Published August 29, 2013 Accessed March 27, 2014.
  • 11.Centers for Medicare & Medicaid Services. Affordable Care Act Implementation FAQs - Set 12. http://www.cms.gov/CCIIO/Resources/Fact-Sheets-and-FAQs/aca_implementation_faqs12.html#CoverageofPreventiveServices. Accessed July 14, 2014.
  • 12.Osterholm MT, Kelley NS, Sommer A, Belongia EA. Efficacy and effectiveness of influenza vaccines: a systematic review and meta-analysis. Lancet Infect Dis. 2012;12(1):36–44. doi: 10.1016/S1473-3099(11)70295-X. [DOI] [PubMed] [Google Scholar]
  • 13.Nichol KL, Baken L, Nelson A. Relation between influenza vaccination and outpatient visits, hospitalization, and mortality in elderly persons with chronic lung disease. Ann Intern Med. 1999;130(5):397–403. [DOI] [PubMed] [Google Scholar]
  • 14.Talbot HK, Griffin MR, Chen Q, Zhu Y, Williams JV, Edwards KM. Effectiveness of seasonal vaccine in preventing confirmed influenza-associated hospitalizations in community dwelling older adults. J Infect Dis. 2011;203(4):500–508. doi: 10.1093/infdis/jiq076. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Talbot HK, Zhu Y, Chen Q, Williams JV, Thompson MG, Griffin MR. Effectiveness of influenza vaccine for preventing laboratory-confirmed influenza hospitalizations in adults, 2011–2012 influenza season. Clin Infect Dis Off Publ Infect Dis Soc Am. 2013;56(12):1774–1777. doi: 10.1093/cid/cit124. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Colquhoun AJ, Nicholson KG, Botha JL, Raymond NT. Effectiveness of influenza vaccine in reducing hospital admissions in people with diabetes. Epidemiol Infect. 1997;119(3):335–341. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Udell JA, Zawi R, Bhatt DL, et al. Association between influenza vaccination and cardiovascular outcomes in high-risk patients: a meta-analysis. JAMA. 2013;310(16):1711–1720. doi: 10.1001/jama.2013.279206. [DOI] [PubMed] [Google Scholar]
  • 18.Setse RW, Euler GL, Gonzalez-Feliciano AG, et al. Influenza vaccination coverage - United States, 2000–2010. Morb Mortal Wkly Rep Surveill Summ Wash DC 2002. 2011;60 Suppl:38–41. [PubMed] [Google Scholar]
  • 19.Egede LE, Zheng D. Racial/ethnic differences in influenza vaccination coverage in high-risk adults. Am J Public Health. 2003;93(12):2074–2078. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Hutchins SS, Fiscella K, Levine RS, Ompad DC, McDonald M. Protection of Racial/Ethnic Minority Populations During an Influenza Pandemic. Am J Public Health. 2009;99(S2):S261–S270. doi: 10.2105/AJPH.2009.161505. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Fiscella K, Dressler R, Meldrum S, Holt K. Impact of influenza vaccination disparities on elderly mortality in the United States. Prev Med. 2007;45(1):83–87. doi: 10.1016/j.ypmed.2007.03.007. [DOI] [PubMed] [Google Scholar]
  • 22.Logan JL. Disparities in Influenza Immunization Among US Adults. J Natl Med Assoc. 2009;101(2):161–166. [DOI] [PubMed] [Google Scholar]
  • 23.Fiscella K Commentary--Anatomy of Racial Disparity in Influenza Vaccination. Health Serv Res. 2005;40(2):539–550. doi: 10.1111/j.1475-6773.2005.00371.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Hebert PL, Frick KD, Kane RL, McBean AM. The causes of racial and ethnic differences in influenza vaccination rates among elderly Medicare beneficiaries. Health Serv Res. 2005;40(2):517–537. doi: 10.1111/j.1475-6773.2005.00370.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Williams DR, Collins C. US Socioeconomic and Racial Differences in Health: Patterns and Explanations. Annu Rev Sociol. 1995;21:349–386. [Google Scholar]
  • 26.Williams DR, Mohammed SA. Discrimination and racial disparities in health: evidence and needed research. J Behav Med. 2009;32(1):20–47. doi: 10.1007/s10865-008-9185-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Shavers VL, Fagan P, Jones D, et al. The State of Research on Racial/Ethnic Discrimination in The Receipt of Health Care. Am J Public Health. 2012;102(5):953–966. doi: 10.2105/AJPH.2012.300773. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Trivedi AN, Ayanian JZ. Perceived discrimination and use of preventive health services. J Gen Intern Med. 2006;21(6):553–558. doi: 10.1111/j.1525-1497.2006.00413.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Hausmann LRM, Jeong K, Bost JE, Ibrahim SA. Perceived discrimination in health care and use of preventive health services. J Gen Intern Med. 2008;23(10):1679–1684. doi: 10.1007/s11606-008-0730-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Zuckerman RB, Tinsley LJ, Hawk H, Cohen B. Perceived reactions to race and health status in the Massachusetts Behavioral Risk Factor Surveillance System Survey. Ethn Dis. 2012;22(4):492–496. [PubMed] [Google Scholar]
  • 31.Aligning Forces for Quality. About Us. Aligning Forces for Quality. http://forces4quality.org/about-us. Published 2013 Accessed March 15, 2014.
  • 32.Scanlon DP, Beich J, Alexander JA, et al. The aligning forces for quality initiative: background and evolution from 2005 to 2012. Am J Manag Care. 2012;18(6 Suppl):s115–s125. [PubMed] [Google Scholar]
  • 33.Scanlon DP, Alexander JA, Beich J, et al. Evaluating a community-based program to improve healthcare quality: research design for the Aligning Forces for Quality initiative. Am J Manag Care. 2012;18(6 Suppl): s165–s176. [PubMed] [Google Scholar]
  • 34.RTI International. Aligning Forces for Quality Consumer Survey 2.0 Final Methodology Report. Research Triangle Park, NC: RTI International; 2013. [Google Scholar]
  • 35.Andersen R, Rice TH, Kominski GF, eds. Changing the U.S. Health Care System: Key Issues in Health Services Policy and Management. 3rd ed. San Francisco: Jossey-Bass; 2007. [Google Scholar]
  • 36.Babitsch B, Gohl D, von Lengerke T. Re-revisiting Andersen’s Behavioral Model of Health Services Use: a systematic review of studies from 1998–2011. Psycho-Soc Med. 2012;9. doi: 10.3205/psm000089. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.National Center for Health Statistics. 2011 NHIS Questionnaire - Sample Adult. May 2012.
  • 38.Hibbard JH, Mahoney ER, Stockard J, Tusler M. Development and testing of a short form of the patient activation measure. Health Serv Res. 2005;40(6 Pt 1):1918–1930. doi: 10.1111/j.1475-6773.2005.00438.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Hibbard JH, Stockard J, Mahoney ER, Tusler M. Development of the Patient Activation Measure (PAM): conceptualizing and measuring activation in patients and consumers. Health Serv Res. 2004;39(4 Pt 1):1005–1026. doi: 10.1111/j.1475-6773.2004.00269.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.MacKinnon DP, Fairchild AJ, Fritz MS. Mediation analysis. Annu Rev Psychol. 2007;58:593–614. doi: 10.1146/annurev.psych.58.110405.085542. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.MacKinnon DP, Fritz MS, Williams J, Lockwood CM. Distribution of the product confidence limits for the indirect effect: program PRODCLIN. Behav Res Methods. 2007;39(3):384–389. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Birkett M, Russell ST, Corliss HL. Sexual-orientation disparities in school: the mediational role of indicators of victimization in achievement and truancy because of feeling unsafe. Am J Public Health. 2014;104(6):1124–1128. doi: 10.2105/AJPH.2013.301785. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Morera OF, Castro FG. Important considerations in conducting statistical mediation analyses. Am J Public Health. 2013;103(3):394–396. doi: 10.2105/AJPH.2012.301047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Mackinnon DP, Dwyer JH. Estimating Mediated Effects in Prevention Studies. Eval Rev. 1993;17(2):144–158. doi: 10.1177/0193841X9301700202. [DOI] [Google Scholar]
  • 45.Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51(6):1173–1182. [DOI] [PubMed] [Google Scholar]
  • 46.James LR, Brett JM. Mediators, moderators, and tests for mediation. J Appl Psychol. 1984;69(2):307–321. doi: 10.1037/0021-9010.69.2.307. [DOI] [Google Scholar]
  • 47.Judd CM, Kenny DA. Process Analysis: Estimating Mediation in Treatment Evaluations. Eval Rev. 1981;5(5):602–619. doi: 10.1177/0193841X8100500502. [DOI] [Google Scholar]
  • 48.Hale L, Parente V, Dowd JB, et al. Fibrinogen may mediate the association between long sleep duration and coronary heart disease. J Sleep Res. 2013;22(3):305–314. doi: 10.1111/jsr.12020. [DOI] [PubMed] [Google Scholar]
  • 49.Preacher KJ, Hayes AF. SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behav Res Methods Instrum Comput J Psychon Soc Inc. 2004;36(4):717–731. [DOI] [PubMed] [Google Scholar]
  • 50.Preacher KJ, Hayes AF. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behav Res Methods. 2008;40(3):879–891. [DOI] [PubMed] [Google Scholar]
  • 51.StataCorp LP. Stata Statistical Software: Release 13. College Station, TX: StataCorp LP; 2013. [Google Scholar]
  • 52.Centers for Disease Control and Prevention. Final Estimates of Monthly Cumulative Influenza Vaccination Coverage for the 2010–2011 Season - BRFSS and NIS, August 2010 through May 2011.; 2011. [Google Scholar]
  • 53.Centers for Disease Control and Prevention. State-Level Monthly Cumulative Influenza Vaccination Coverage Estimates for the 2011–12 Season, United States, August 2011 through May 2012.; 2012. [Google Scholar]
  • 54.Nagata JM, Hernández-Ramos I, Kurup AS, Albrecht D, Vivas-Torrealba C, Franco-Paredes C. Social determinants of health and seasonal influenza vaccination in adults ≥65 years: a systematic review of qualitative and quantitative data. BMC Public Health. 2013;13:388. doi: 10.1186/1471-2458-13-388. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Vander Weg MW, Howren MB, Cai X. Use of Routine Clinical Preventive Services Among Daily Smokers, Non-daily Smokers, Former Smokers, and Never-smokers. Nicotine Tob Res. 2012;14(2):123–130. doi: 10.1093/ntr/ntr141. [DOI] [PubMed] [Google Scholar]
  • 56.Pearson WS, Dube SR, Ford ES, Mokdad AH. Influenza and pneumococcal vaccination rates among smokers: data from the 2006 Behavioral Risk Factor Surveillance System. Prev Med. 2009;48(2):180–183. doi: 10.1016/j.ypmed.2008.11.001. [DOI] [PubMed] [Google Scholar]
  • 57.Looijmans-van den Akker I, van den Heuvel PM, Verheij TJM, van Delden JJM, van Essen GA, Hak E. No intention to comply with influenza and pneumococcal vaccination: behavioural determinants among smokers and non-smokers. Prev Med. 2007;45(5):380–385. doi: 10.1016/j.ypmed.2007.07.009. [DOI] [PubMed] [Google Scholar]
  • 58.Cruijff M, Thijs C, Govaert T, Aretz K, Dinant GJ, Knottnerus A. The effect of smoking on influenza, influenza vaccination efficacy and on the antibody response to influenza vaccination. Vaccine. 1999;17(5):426–432. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Finklea JF, Hasselblad V, Riggan WB, Nelson WC, Hammer DI, Newill VA. Cigarette smoking and hemagglutination inhibition response to influenza after natural disease and immunization. Am Rev Respir Dis. 1971;104(3):368–376. [DOI] [PubMed] [Google Scholar]
  • 60.Horvath KM, Brighton LE, Herbst M, Noah TL, Jaspers I. Live Attenuated Influenza Virus (LAIV) induces different mucosal T cell function in nonsmokers and smokers. Clin Immunol. 2012;142(3):232–236. doi: 10.1016/j.clim.2011.12.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Yaqub O, Castle-Clarke S, Sevdalis N, Chataway J. Attitudes to vaccination: a critical review. Soc Sci Med 1982. 2014;112:1–11. doi: 10.1016/j.socscimed.2014.04.018. [DOI] [PubMed] [Google Scholar]
  • 62.Shackman G Sample Size and Design Effect. New York State Department of Health; 2001. http://faculty.smu.edu/slstokes/stat6380/deff%20doc.pdf. [Google Scholar]
  • 63.Scanlon DP, Shi Y, Bhandari N, Christianson JB. Are Healthcare Quality “Report Cards” Reaching Consumers? Awareness in the Chronically Ill Population. Am J Manag Care. 2015;21(March 2015 4). http://www.ajmc.com/publications/issue/2015/2015-vol21-n3/are-healthcare-quality-report-cards-reaching-consumers-awareness-in-the-chronically-ill-population/1. Accessed April 6, 2015. [PubMed] [Google Scholar]
  • 64.National Research Council (U.S.). Measuring Racial Discrimination. (Blank RM, Dabady M, Citro CF, eds.). Washington, DC: National Academies Press; 2004. [Google Scholar]
  • 65.Vargas Bustamante A, Chen J, Rodriguez HP, Rizzo JA, Ortega AN. Use of preventive care services among Latino subgroups. Am J Prev Med. 2010;38(6):610–619. doi: 10.1016/j.amepre.2010.01.029. [DOI] [PubMed] [Google Scholar]
  • 66.Smedley BD, Stith AY, Nelson AR, Institute of Medicine Unequal Treatment Confronting Racial and Ethnic Disparities in Health Care. Washington, D.C.: National Academies Press; 2003. [PubMed] [Google Scholar]
  • 67.Shavers VL, Shavers BS. Racism and health inequity among Americans. J Natl Med Assoc. 2006;98(3):386–396. [PMC free article] [PubMed] [Google Scholar]

RESOURCES