CI week 1 assignment

Patient Education and Counseling 101 (2018) 2233–2240

Contents lists available at ScienceDirect

Patient Education and Counseling

journal homepage: www.elsevier.com/locate/pateducou

Health insurance literacy and awareness of the Affordable Care Act in a vulnerable Hispanic population

Suad Ghaddara,*, Jihyun Byunb, Janani Krishnaswamic

a Department of Health and Biomedical Sciences, The University of Texas Rio Grande Valley, Edinburg, USA b School of Human Ecology, The University of Texas at Austin, Austin, USA c Department of Pediatrics and Preventive Medicine, The University of Texas Rio Grande Valley, Edinburg, USA

A R T I C L E I N F O A B S T R A C T

Article history: Received 2 March 2018 Received in revised form 25 August 2018 Accepted 29 August 2018

Keywords: Health insurance literacy Affordable Care Act

Objective: The Patient Protection and Affordable Care Act (ACA) has allowed millions of Americans to obtain coverage. However, many, especially minorities, remain uninsured. With mounting evidence supporting the importance of health insurance literacy (HIL), the purpose of this cross-sectional study is to examine the association between HIL and ACA knowledge. Methods: We conducted 681 in-person interviews with participants at a community health event along the Texas-Mexico border in 2015, after the conclusion of the ACA’s second enrollment period. To assess HIL, we used the Health Insurance Literacy Measure, reflecting consumers’ confidence to choose, compare, and use health insurance. We assessed ACA knowledge through the following question: “How much would you say you know about this health reform law?” Logistic regression was used to examine the association between HIL and ACA knowledge after controlling for several covariates. Results: Almost 70% of participants knew nothing/very little about the ACA. Multivariate analyses revealed that no/very little ACA knowledge was associated with low levels of confidence “choosing health insurance plans” (OR:0.55; 95%CI:0.40-0.75) (full sample) and “comparing plans” (OR:0.56; 95%CI:0.32- 0.96) (U.S.-born sub-sample). Conclusion: No/little ACA knowledge is associated with lower levels of HIL. Practice Implications: Promoting HIL is an essential step towards improving healthcare access.

© 2018 Elsevier B.V. All rights reserved.

1. Introduction

The United States’ health care system has faced an array of challenges, including high costs, health inequities, and high uninsured rates, among others. The Patient Protection and Affordable Care Act (ACA), a monumental health reform effort also known as Obamacare, aimed to address several of the system’s shortcomings, most importantly the high rate of the uninsured which stood at 17% of the population (51 million people) in 2009 [1]. The ACA’s passage in 2010 has allowed more than 20 million Americans to obtain coverage under its provisions [2]. These include expanding Medicaid (government health care coverage for low-income individuals) eligibility in certain states, providing subsidies for qualifying individuals to purchase private health

* Corresponding author at: Department of Health and Biomedical Sciences, The University of Texas Rio Grande Valley, 1201 W. University Dr., Edinburg, TX, 78539, USA.

E-mail addresses: [email protected] (S. Ghaddar), [email protected] (J. Byun), [email protected] (J. Krishnaswami).

https://doi.org/10.1016/j.pec.2018.08.033 0738-3991/© 2018 Elsevier B.V. All rights reserved.

insurance plans in the health insurance marketplaces, prohibiting barriers to enrollment based on pre-existing conditions, and increasing the cut-off age for young adults to stay on a parent’s plan to age 26. Many individuals, however, remain uninsured, especially among minority populations and particularly among Hispanics. Despite considerable outreach efforts and correspond- ing major enrollment gains, 28% of non-elderly Hispanics remain uninsured [2]. In comparison, and for the same time period, only 9% and 15% of non-elderly, non-Hispanic whites and blacks were, respectively, uninsured [2]. Several reasons and persistent challenges to the lack of health coverage remain, including low population awareness of the ACA law, its enrollment guidelines and provisions [3,4]. However, health insurance literacy may be an even more fundamental factor influencing coverage gaps [5–10].

Health insurance literacy (HIL) is defined as “the degree to which individuals have the knowledge, ability, and confidence to find and evaluate information about health plans, select the best plan for their own (or their family's) financial and health circumstances, and use the plan once enrolled [11].” Similar to the evidence supporting an association between low health literacy and poor health outcomes [12], research is starting to reveal that poor

2234 S. Ghaddar et al. / Patient Education and Counseling 101 (2018) 2233–2240

familiarity with a health insurance program, such as Medicare, is associated with a less effective utilization of healthcare services, and consequently poorer health outcomes [13]. Evidence is mounting supporting the importance of HIL in determining health insurance status, healthcare utilization, and health behaviors, among others [6,14,15]. However, few studies have explored HIL within the ACA context, or assessed HIL in Hispanic communities [14]. The purpose of this study is to assess health insurance literacy in a vulnerable Hispanic community and to examine whether ACA knowledge is associated with levels of health insurance literacy.

2. Methods

2.1. Study setting and data collection

Data for this study was collected from participants at Operation Lone Star (OLS), an annual public health emergency preparedness exercise along the Texas-Mexico border. The event is a collabora- tion between various organizations, including local departments of health, the Texas Department of State Health Services, the U.S. military, institutions of higher education, and a myriad of community organizations and volunteers. The event also provides free primary, dental, and vision healthcare services to community residents. In 2015, OLS events and services took place during the week of July 27–31 at five locations across the South Texas Border from Brownsville to Laredo. Data collection for this study took place at one of the locations in Hidalgo County (home to over 800,000 people) [16] which was attended by almost 3000 county residents (children and adults) over the course of the week. As in other Texas-Mexico border counties, the overwhelming majority of the population is of Hispanic or Latino origin (92%) [16]. The county is characterized by high poverty rates (a third of the population lives below the federal poverty level) and low educational attainment (36% of individuals 25 years and over do not have a high school degree) [16]. Lack of healthcare coverage is a main challenge with 43% of individuals 18–64 years old being uninsured in 2015 [17].

We employed a convenience sampling design. Data was collected in-person by trained student interviewers, some of whom were bilingual (English and Spanish). Students approached OLS attendees, who were waiting to receive health services at various stations, with information about the study and invited them to participate. Based on the participant’s preferred language, interviews were conducted in either English or Spanish. After completing the anonymous interview, participants were provided with educational material about diabetes and a bottle of water. All

Table 1 Health Insurance Literacy Measure18.

Selecting health insurance scales Scale 1. Confidence: Choosing a health plan How confident are you that . . . ? Six statements on which respondents rate their level of confidence choosing a hea 1: Not at all confident, 2: Slightly confident, 3: Moderately confident, 4: Very confi

Scale 2. Comparing health plans When comparing health insurance plans, how likely are you to . . . ? Seven statements on which respondents indicate the likelihood of a behavior relat 1: Not at all likely, 2: Somewhat likely, 3: Moderately likely, 4: Very likely

Using health insurance scales Scale 3: Confidence: Using a health plan How confident are you that . . . ? Four statements on which respondents rate their level of confidence about using h 1: Not at all confident, 2: Slightly confident, 3: Moderately confident, 4: Very confi

Scale 4: Being Proactive When using your health insurance plan, how likely are you to . . . ? Four statements on which respondents indicate the likelihood of being proactive w 1: Not at all likely, 2: Somewhat likely, 3: Moderately likely, 4: Very likely

study procedures were approved by the Institutional Review Board at The University of Texas-Pan American (now The University of Texas Rio Grande Valley).

2.2. Measurements

The survey instrument included questions assessing socio- demographic characteristics, knowledge of the ACA, health insurance literacy, ehealth literacy, and health status, among others. The survey instrument was translated to Spanish. We used existing Spanish translations when available (e.g., Census ques- tions). For those items where no Spanish translation was available, a Spanish native speaker translated the survey items. These were in turn reviewed and modified by a Spanish professor with broad academic knowledge and experience in the linguistic usage of both Spanish and English in the region as well as a deep cultural understanding of the target population.

2.2.1. Dependent variable: ACA knowledge We assessed ACA knowledge by the question “How much would

you say you know about this health reform law?” Response options included: nothing, very little, just some, a fair amount, or a great deal. We recoded the survey responses into a dichotomous variable (1, nothing/very little knowledge; 0, otherwise). The question mirrored that used in a nationally-representative sample [8] allowing us to compare our results to other studies.

2.2.2. Independent variable: health insurance literacy We utilized the Health Insurance Literacy Measure (HILM) [18]

to assess health insurance literacy. The HILM is a valid and reliable measure of “consumers’ ability to select and use private health insurance.” It includes 21 items divided into two scales: selecting insurance and using insurance, each of which encompasses two subscales (Table 1). We administered the using health insurance scales (Scales 3 and 4) only to those individuals who reported having healthcare coverage.

2.2.3. Covariates Our analysis controlled for several covariates.

2.2.3.1. Sociodemographic characteristics. We used standard self- report U.S. Census measures to assess a range of sociodemographic characteristics. These included gender, age, ethnicity (being of Hispanic or Latino origin), country of birth, educational attainment, income, and health insurance status.

lth insurance plan. dent

ed to their behavior when choosing a health plan

ealth insurance. dent

hen using their health insurance.

S. Ghaddar et al. / Patient Education and Counseling 101 (2018) 2233–2240 2235

2.2.3.2. Health literacy. We used two health literacy measures assessing different aspects of the concept. For health literacy, we used the Single Item Literacy Screener [19] which helps identify limited reading ability, an important aspect of health literacy. Participants responded to the question: “How confident are you filling out forms by yourself?” Response options included: not at all, a little bit, somewhat, quite a bit, extremely. We considered those with “extreme” or “quite a bit” of confidence to have adequate levels of health literacy. We assessed eHealth literacy using eHEALS, an 8-item scale, designed to “measure consumers’ combined knowledge, comfort, and perceived skills at finding, evaluating, and applying electronic health information to health problems’’ [20]. Respondents indicated their level of agreement on a 5-point Likert-type scale (1 “Strongly disagree” to 5 “strongly agree”). Higher scores on the summation of responses reflect higher levels of eHealth literacy. The reliability and validity of eHEALS has been established in both English and Spanish [20,21]. Cronbach’s α for the scale was 0.96 for our Spanish-speaking subsample (N = 495) and 0.94 for our English-speaking subsample (N = 172).

2.2.3.3. Health status. Given that poor health and the presence of chronic conditions may represent unmet healthcare needs and, thus, may generate more interest in health coverage as well as awareness of coverage options, we included two measures for health status. We assessed general health status using a validated

Table 2 Participant characteristics by ACA knowledge.

Sociodemographic variables n % Know som a fair am

ACA Knowledge 681 31% Interview language 681 English 177 26 44 Spanish 504 74 26 Of Hispanic/Latino origin 667 Yes 662 99 31 No 5 1 20 Country of birth 666 U.S.-born 172 26 41 Foreign-born 494 74 28 Gender 672 Male 133 20 38 Female 539 80 29 High school graduate 667 Yes 328 49 39 No 339 51 24 Income < $20K 653 Yes 548 84 27 No 105 16 54 Uninsured 680 Yes 79 12 43 No 601 88 29 Self-rated health status 647 Poor/fair 367 57 27 Good/very good/excellent 280 43 37 Diabetes diagnosis 644 Yes 111 17 39 No 533 83 30 Adequate health literacy 666 Yes 246 37 39 No 420 63 26 Political affiliation 651 Yes 219 34 38 No 432 66 27

n Mean (SD) Mean (SD)

Age (range 18–80) 666 38.78 (12.51) 37.35 (12.03)

eHeals (range 8–40) 667 21.30 23.91 (9.58) (10.03)

question from the Behavioral Risk Factor Surveillance System (BRFSS) [22] asking respondents to rate their health (excellent, very good, good, fair, poor). We recoded the health status question as a dichotomous variable (1, fair or poor health; 0, otherwise). We also checked whether participants had a diabetes diagnosis using the BRFSS question that asked whether they had ever been told by a health professional that they had diabetes.

2.2.3.4. Political affiliation. Given the politicized nature of the health reform debate in the U.S., we asked about the political affiliation of participants. We expect that those with any type of affiliation (Republican, Democrat, Independent) will be more likely to know about the ACA relative to those with no affiliation.

2.3. Data analysis

We analyzed data using SPSS (Version 24) [23]. Descriptive analyses generated participant characteristics. We conducted bivariate tests (two-sided chi-square and t tests, where appropri- ate) to examine the association between ACA knowledge and different variables. To assess the internal reliability of the HILM scales in English and Spanish, we used the Cronbach’s alpha coefficient; a coefficient above 0.80 for basic research tools reflects adequate internal consistency [24]. We ran logistic regressions to examine the association between ACA knowledge and health

e, Know nothing or very little (%) p ount, or a great deal (%)

69% <.001

56 74

.592 69 80

.001 59 72

.058 62 71

<.001 61 76

<.001 73 45

.012 57 71

.005 73 63

.060 61 70

<.001 61 74

.006 62 73 Mean p (SD) 39.43 .048 (12.68) 20.15 <.001 (9.15)

2236 S. Ghaddar et al. / Patient Education and Counseling 101 (2018) 2233–2240

insurance literacy controlling for the different covariates. We used a Type I error rate of 0.05.

3. Results

3.1. Sample characteristics

We interviewed 681 attendees at OLS. Table 2 presents participant characteristics by ACA knowledge (know some, a fair amount, or a great deal vs. know nothing or very little). Almost 70% of participants knew nothing or very little about the ACA. The majority of the interviews were conducted in Spanish with only 26% conducted in English. Participants were overwhelmingly of Hispanic or Latino origin (99%) and female (80%). Almost three- quarters were foreign-born, primarily in Mexico. Low educational attainment and poverty were characteristic of the population served at OLS with only half having a high school degree or its equivalent and 84% reporting annual household incomes below $20,000. Lack of healthcare coverage represented a serious problem among participants with 88% being uninsured. More than half (57%) reported fair or poor health and 17% had a diabetes diagnosis. Two-thirds did not have a political affiliation. The average age was 39 years. eHEALS scores ranged from 8 to 40 with an average score of 21. Knowing nothing or very little about the ACA was associated with speaking Spanish, being foreign-born, not having a high school degree, reporting an annual household income below $20,000, being uninsured, having fair or poor health, having inadequate health literacy, and not being affiliated with a political party. No or little ACA knowledge was also associated with older age and lower levels of eHealth literacy.

3.2. Health Insurance Literacy Measure

3.2.1. Reliability analysis Table 3 reports the internal consistency findings for the HILM

scales for the overall sample, the English-speaking, and the Spanish-speaking samples and compares them to those from the HILM development study [18]. Internal consistency, as measured by Cronbach’s alpha coefficient, was high. The internal reliability of the scales was comparable to that reported by Paez et al. [18].

3.2.2. Health insurance literacy and ACA knowledge Table 4 presents the means and standard deviations for the four

HILM scales along with t-tests examining the association between ACA knowledge and health insurance literacy. ACA knowledge was significantly associated with health insurance literacy as measured by Scales 1 and 2 of the HILM. Scales 3 and 4 were administered only to the insured sub-sample given that the questions refer to

Table 3 Reliability measures.

Scale Cronbach’s α N Cronbach’s α Paez et al.18

Scale 1. Confidence: Choosing a health plan 0.89 512 0.93 English 0.89 145 Spanish 0.88 367

Scale 2. Comparing health plans 0.92 556 0.96 English 0.92 151 Spanish 0.92 405

Scale 3: Confidence: Using a health plan 0.85 73* 0.93 English 0.89 40 Spanish 0.78 33

Scale 4: Being Proactive 0.89 77* 0.80 English 0.89 42 Spanish 0.88 35

* Scales 3 and 4 were administered to only those with healthcare coverage.

health insurance usage. The relationship between these two subscales and ACA knowledge was not significant, potentially due to the much smaller sample size.

3.3. Multivariate logistic regression analyses

To examine the association between ACA knowledge and HIL, we conducted binary logistic regressions controlling for individu- al-level variables associated with ACA knowledge. Table 5 (HILM Scale 1) and 6 (HILM Scale 2) report the results with odds ratios (OR) and 95% confidence intervals (CI). After controlling for individual-level variables potentially associated with ACA knowl- edge, logistic regression analyses revealed that higher levels of health insurance literacy, as measured by HILM (Scales 1 and 2), were associated with decreased odds of no or little ACA knowledge. Scale 1 (confidence choosing insurance) demonstrated a higher level of significance relative to Scale 2 (comparing health plans), <0.001 versus 0.058, respectively. Income was also associated with ACA awareness. Participants who reported annual household incomes below $20,000 were twice as likely to have no or little knowledge of the ACA. A diabetes diagnosis was also a significant variable. Participants with a diabetes diagnosis were half as likely to have no or ACA knowledge. When including Scale 2 (comparing health plans) of the HILM in the model (Table 6), being female was also related to no or little ACA knowledge.

Given the area’s sociodemographic profile, there is a possibility that some of OLS attendees are not U.S. citizens and, therefore, not eligible for purchasing health insurance under the ACA. As a consequence, this group may not be interested in knowing about the ACA. To rule out that possibility, we repeated the analyses for the sub-sample that was U.S.-born (Panel B of Tables 5 and 6), being fully aware that many of those who are foreign-born may still be citizens. The association between ACA knowledge and Scale 1 of HIL continued to retain significance (OR:0.48; 95%CI:0.26-0.87). Scale 2, which was marginally significant for the overall sample, became significant in the U.S.-born subsample (OR:0.56; 95% CI:0.32-0.96).

4. Discussion and conclusion

4.1. Discussion

This study suggests that low levels of health insurance literacy predict no or poor knowledge of the ACA in a low-income, predominantly Hispanic population along the Texas-Mexico border, independent of sociodemographic factors and health status. Overall, awareness of the ACA was very low in an underserved Hispanic community. Even after the conclusion of two ACA enrollment periods, over two-thirds of participants still reported knowing nothing or very little about the health reform law. This lack of awareness was more pronounced among those with lower levels of health insurance literacy, as measured by two scales reflecting confidence choosing health plans and comparing health plans. Those with low income levels were also more likely to have no or little knowledge of the ACA while those who had a diabetes diagnosis were less likely to have no or little ACA knowledge.

The low level of ACA knowledge, though more pronounced in our sample, is consistent with other findings in the literature. In an analysis of a nationally-representative sample conducted a few weeks before the introduction of the health insurance exchanges, 24% of respondents knew a great deal/fair amount about the ACA compared to 14% in our sample. When comparing ACA knowledge among the uninsured, a group that is expected to be more interested in obtaining healthcare coverage, the corresponding awareness numbers were surprisingly lower (17% in the national

S. Ghaddar et al. / Patient Education and Counseling 101 (2018) 2233–2240 2237

Table 4 Health insurance literacy by ACA knowledge.

Total sample Know some, Know nothing or very little (%) p a fair amount, or a great deal (%)

Confidence choosing (n,(%)) 512 (100%) 166 (32%) 346 (68%) Mean (SD) 2.15 (0.78) 2.47 (0.79) 1.99 (0.73) <.001 Comparing plans (n,(%)) 556 (100%) 181 (33%) 375 (67%) Mean (SD) 2.41(0.87) 2.68 (0.91) 2.29 (0.83) <.001 Confidence using (n,(%))* 73 (100%) 33 (45%) 40 (55%) Mean (SD) 2.81 (0.84) 2.91 (0.80) 2.73 (0.87) .355 Being proactive (n,(%))* 77 (100%) 36 (47%) 41 (53%) Mean (SD) 2.92 (0.94) 3.07 (0.90) 2.79 (0.97) .201

* Scales 3 and 4 were administered to only those with healthcare coverage.

Table 5 Logistic regression results: ACA knowledge (dependent variable) and confidence choosing insurance.

Panel A Odds ratio 95% Confidence Interval p Total sample: n = 465

Confidence choosing insurance 0.55 (0.40, 0.75) <.001 Income < $20K 2.00 (1.16, 3.45) .006 Has diabetes 0.53 (0.30, 0.93) .028 Model fit: p-value: <.001; Nagelkerke R Square: 0.18

Panel B Odds ratio 95% Confidence Interval p U.S.-born sub-sample: n = 130

Confidence choosing insurance 0.48 (0.26, 0.87) .016 eHEALS 0.94 (0.89, 1.00) .051 Model fit: p-value: <.001; Nagelkerke R Square: 0.34

Table 6 Logistic regression results: ACA knowledge (dependent variable) and comparing health plans.

Panel A Odds ratio 95% Confidence Interval p Total sample: n = 510

Comparing health plans 0.77 (0.59, 1.00) .058 Income < $20K 1.94 (1.16, 3.25) .011 Has diabetes 0.51 (0.29, 0.88) .015 Female 1.66 (1.02, 2.69) .041

Model fit: p-value: <.001; Nagelkerke R Square: 0.15

Panel B U.S.-born sub-sample: n = 130 Odds ratio 95% Confidence Interval p

Comparing health plans 0.56 (0.32, 0.96) .035 Female 2.85 (1.09, 7.46) .033 Model fit: p-value: <.001; Nagelkerke R Square: 0.31

sample [8] compared to 13% for our uninsured subsample and 10% for our uninsured U.S.-born subsample). In another study focusing on ACA awareness in West Virginia, familiarity with the health insurance marketplace under the ACA improved from 2013 to 2014, yet 29% were still “not at all familiar” and 25% were “not too familiar” in 2014 [3]. The lack of ACA knowledge in our sample is more prominent and concerning given that our data was collected after the conclusion of two enrollment periods in the health insurance marketplace, where one would expect more exposure to have occurred through outreach campaigns, media coverage, and word-of-mouth.

The measurement of health insurance literacy varies across studies. Many studies utilize objective knowledge-based questions on key insurance concepts (e.g., premium, deductible, co-pay, provider networks) through true/false statements, definitions, and/or multiple choice questions [6,8,14,15,25]. Other studies have utilized subjective measures such as rating an individual’s level of confidence in understanding health insurance terms [5] or a limited number of questions from the HILM [3]. However, no consistent number of questions or cut-off points for defining

adequate HIL is available across studies. In contrast, HILM, despite its limitations, represents a multi-dimensional measure that captures multiple domains of health insurance literacy. The only other study to our knowledge that utilizes scales 1 and 2 of the HILM is one where HIL is assessed pre and post an intervention aiming to enhance HIL [26]. Question scores for most of the statements were lower for our sample compared to the 83% White, non-Hispanic sample in the other study.

Few studies specifically examined the relationship between ACA knowledge and HIL. Among the few that did, similar findings were reported where familiarity of the marketplace was positively associated with health insurance literacy [3].

The positive relationship between income levels and ACA knowledge is consistent with other studies where ACA knowledge and HIL were higher among higher income brackets [8]. The finding that a diabetes diagnosis was associated with higher

likelihood of at least some ACA knowledge may reflect unmet healthcare needs among this group, thus, generating more interest in seeking information about healthcare coverage options. It might also indicate more experience with the healthcare system.

4.1.1. Limitations Our study has several limitations. First, similar to other cross-

sectional designs, we merely establish an association, and not causality, between ACA knowledge and HIL. We also cannot establish the directionality of the relationship, as it is likely that knowledge of the ACA will also influence HIL. Second, cultural dynamics may have influenced comprehension and relevance of HILM measures. In Mexico, health insurance is universal and nationally owned, and bears little resemblance to the U.S. market- driven insurance system. However, the relationship between ACA knowledge and HIL maintained significance when considering only U.S.-born individuals, mitigating this possibility. Third, the assessment of HILM’s psychometric properties was conducted in a majority White, non-Hispanic population group and to our knowledge has not been tested in Spanish-speaking groups. Our

2238 S. Ghaddar et al. / Patient Education and Counseling 101 (2018) 2233–2240

reliability analyses, however, reflects high internal consistency of the scales. We also ran the analyses separately for the Spanish- and English-speaking subgroups (not reported) and had similar findings regarding the significance of the association between the two HILM scales and ACA knowledge. Fourth, our ACA knowledge assessment is measured by one subjective knowledge question at a time when there are multiple components and provisions to the ACA. This is especially problematic given that comparisons between subjective and objective measures of ACA knowledge revealed that people were not as knowledgeable as they thought they were [8]. This, however, may be mitigated by the fact that self-assessments of knowledge in our sample were generally low and, therefore, unlikely to be overestimated. Finally, given the large number of uninsured participants and, conse- quently, the small number of those who have used insurance, it was difficult to assess the associations between ACA knowledge and scales 3 (confidence using health insurance) and 4 (using health insurance/being proactive) of the HILM.

4.2. Conclusion

In this study of low-income, predominantly Hispanic individu- als in a U.S.-Mexico border community, little or no knowledge of the ACA was significantly associated with low health literacy levels. This association was significant even after controlling for other factors that could influence ACA knowledge, including sociodemo- graphic factors and health status. This suggests that health insurance literacy exerts an independent effect on knowledge of major healthcare reform policies such as the ACA.

National initiatives, such as Healthy People 2020, the 2011 Department of Health and Human Services Disparities Action Plan, and elements of the ACA, frequently include a goal of addressing widening health disparities in the United States, which are among the largest in the developed world [27,28]. Promoting access to primary care and preventive services for high-risk, vulnerable populations has been posited as one pathway to equitably improve health outcomes, while reducing unnecessary tertiary- and emergency care spending on advanced disease [29]. Moreover, health insurance is a social determinant of health shown to increase financial security, access to preventive and primary care, and treatment for chronic conditions [29,30]. Uninsured popula- tions are more likely to be low-income, non-white, and less likely to report good health [28].

The ability of health policies to exert intended population health benefits depends on participation and adoption by eligible individuals. This is especially true in vulnerable communities facing high burden of disease and disability, such as Hispanics in border communities, who can benefit from policies that influence access and entry into healthcare. While healthcare reform policies such as the ACA may intend to increase healthcare and insurance access for vulnerable communities, those with low health insurance literacy may not actually know about potentially helpful provisions, coverage options, and health reform efforts, and thus remain uninsured [6]. Despite achieving significant gains in coverage after the passage of the ACA, Hispanics are least likely among all ethnic groups to have insurance [2,4,31]. This is especially unfortunate given that Hispanics represent the fastest- growing U.S. minority, are more likely than any other ethnic group to delay needed care due to cost, and experience significant health disparities in prevalence and severity of diseases such as diabetes, colon cancer and cardiovascular disease – diseases which can be potentially prevented, treated and alleviated through regular primary care [28].

Low health insurance literacy is one potential pathway influencing disparity in ACA-related coverage. Low health insur- ance literacy predicted little or no knowledge of the ACA reform;

low awareness could in turn prevent eligible individuals from accessing ACA enrollment opportunities. Of note, a serious lack of ACA knowledge among participants in our Hispanic sample was found even after the conclusion of two enrollment periods in the health insurance marketplace.

Our study findings are mirrored in other vulnerable communi- ties around the world. Mounting evidence is pointing to inade- quate understanding of health insurance concepts as a barrier to the success and sustainability of various health insurance schemes in low- and middle-income countries [32]. In the Lucknow region in India, low health insurance literacy has an indirect potential impact, distinct from affordability, on the purchase of private health insurance through its contribution to negative perceptions about health insurance [33]. Across West Africa, education, a likely contributor to a better understanding of the benefits of health insurance, is an independent predictor of enrolling and remaining in plans [34,35]. In Ghana, negative beliefs and attitudes towards health insurance decrease the odds of remaining insured for the richest quintile [35]. Such evidence underscores the need for promoting health insurance literacy to ensure that government efforts to expand access, whether through free or subsidized coverage, achieve their goal of equity in the provision of health care.

4.3. Practice implications

The relationship between HIL and ACA knowledge seen in our study highlights the need for healthcare reform policies to more strongly emphasize supporting educational programs. While various assistance programs were established with the ACA to provide outreach and education [36] and were successful in reaching millions [37–39], there appears to specifically be a need for improved, culturally-appropriate outreach efforts in vulnerable settings, such as the low-income, Hispanic border community surveyed in this study. Additionally, this study underscores the need for outreach programs to increase general knowledge of health insurance (e.g., HIL), rather than simply providing a policy- specific education, such as determination of eligibility. Integrating health insurance education within health delivery systems, specifically those serving low-income communities (e.g., federally qualified health centers) will not only promote awareness of health insurance options but also promise to support more effective utilization of healthcare services. Several programs and tools have been recently developed and tested, with promising results in terms of enhancing HIL and assisting with healthcare coverage decisions [25,26,40,41]

Finally, it is possible that increasing general knowledge of health insurance, a key prerequisite to “entry” into healthcare, may serve to increase a patient’s confidence and self-efficacy to enroll in relevant programs enabled by healthcare reform and to navigate the healthcare system. In this manner, HIL can be seen as a type of education that functions similarly to other recognized pathways linking the social determinants of education, health literacy and health. Further research is needed to elucidate these possible mediating mechanisms between HIL and enrollment in healthcare programs.

Funding

This work was supported by the College of Health Sciences and Human Services (now the College of Health Professions) at the University of Texas Rio Grande Valley (previously The University of Texas-Pan American). The College played no role in study design; in the collection, analysis and interpretation of data; in the writing of the manuscript; and in the decision to submit the manuscript for publication.

S. Ghaddar et al. / Patient Education and Counseling 101 (2018) 2233–2240 2239

Declaration of interest

None.

Acknowledgements

This research would not have been possible without the assistance of many individuals and organizations: Hidalgo County Health and Human Services Department (Mr. Eduardo Olivarez, Ms. Lauren Garcia, Ms. Brenda Salázar and Ms. Nancy Treviño); Palmview High School leadership (Dr. Armando Ocaña and Mr. Luis García); Dr. John Gonzalez, Associate Professor of Social Work at the University of Texas Rio Grande Valley (UTRGV), for incorporat- ing OLS data collection within his Research for the Social Services course (SOCW 4311); Social Work students (SOCW 4311, 2015 Summer II) for their assistance with data collection; Ms. Patricia Garcia, research assistant, for securing and organizing health educational materials for distribution to OLS participants; Ms. Alma Arteaga for helping with various administrative tasks; student interviewers and volunteers (Patricia Garcia, Reem Ghaddar, and Elsa Suarez); Ms. Ameera Khan for assistance with data entry; Dr. Elvia Ardalani, Professor of Writing and Language Studies at UTRGV, for assistance with Spanish translations; Dr. John Ronnau, Dean of the College of Health Sciences and Human Services at the time of the study, for his unwavering support throughout all phases of the project.

References

[1] C. DeNavas-Walt, B.D. Proctor, J.C. Smith, Income, poverty, and health insurance coverage in the United States: 2009, Current Population Reports, U.S. Census Bureau, 2010, pp. 60–238.

[2] K. Avery, K. Finegold, A. Whitman, Affordable Care Act has Led to Historic, Widespread Increase in Health Insurance Coverage, Office of the Assistant Secretary for Planning and Evaluation, 2016. (Accessed 2 February 2018) https://aspe.hhs.gov/system/files/pdf/207946/ACAHistoricIncreaseCoverage. pdf.

[3] T.K. Bias, P. Agarwal, P. Fitzgerald, Changing awareness of the health insurance marketplace, Am. J. Public Health 105 (2015) 633, doi:http://dx.doi.org/ 10.2105/AJPH.2015.302844.

[4] G.R. Sanchez, F.I. Pedraza, E.D. Vargas, The Impact of the Affordable Care Act on Latino Access to Health Insurance, RWJF Center for Health Policy, 2015. (Accessed 15 February 2018) http://healthpolicy.unm.edu/sites/default/files/ ACA_2015_Research_Brief_FINAL.pdf.

[5] S.K. Long, G.M. Kenney, S. Zuckerman, et al., The health reform monitoring survey: addressing data gaps to provide timely insights into the Affordable Care Act, Health Aff. (Millwood) 33 (2014) 161–167, doi:http://dx.doi.org/ 10.1377/hlthaff.2013.0934.

[6] M. Hoerl, A. Wuppermann, S.H. Barcellos, S. Bauhoff, J.K. Winter, K.G. Carman, Knowledge as a predictor of insurance coverage under the Affordable Care Act, Med. Care 55 (2017) 428–435, doi:http://dx.doi.org/10.1097/ MLR.0000000000000671.

[7] S. Ghaddar, S. Schembri, Challenges to ACA Enrollment: Perspective of ACA Navigators, American Public Health Association Annual Meeting & Expo, 2017. (Accessed 25 February 2018) https://apha.confex.com/apha/2017/ meetingapp.cgi/Paper/380617.

[8] S.H. Barcellos, A.C. Wuppermann, K.G. Carman, et al., Preparedness of Americans for the affordable care act, Proc. Natl. Acad. Sci. U. S. A. 111 (2014) 5497–5502, doi:http://dx.doi.org/10.1073/pnas.1320488111.

[9] National Academies of Sciences, Engineering, and MedicineHealth Insurance and Insights from Health Literacy: Helping Consumers Understand: Proceedings of a Workshop—In Brief2016, Health Insurance and Insights from Health Literacy: Helping Consumers Understand: Proceedings of a Workshop—In Brief (2016), doi:http://dx.doi.org/10.17226/24613.

[10] Assessing Americans’ Familiarity with Health Insurance Terms and Concepts, The Henry J. Kaiser Family Foundation, 2014. (Accessed 27 February 2018) https://www.kff.org/health-reform/poll-finding/assessing-americans- familiarity-with-health-insurance-terms-and-concepts/.

[11] L. Quincy, Measuring Health Insurance Literacy: a Call to Action – a Report From the Health Insurance Literacy Expert Roundtable, Consumers Union, 2012. (Accessed 24 February 2018) http://consumersunion.org/pub/ Health_Insurance_Literacy_Roundtable_rpt.pdf.

[12] N.D. Berkman, S.L. Sheridan, K.E. Donahue, D.J. Halpern, K. Crotty, Low health literacy and health outcomes: an updated systematic review, Ann. Intern. Med. 155 (2011) 97–107, doi:http://dx.doi.org/10.7326/0003-4819-155-2- 201107190-00005.

[13] R.O. Morgan, C.R. Teal, J.C. Hasche, et al., Does poorer familiarity with Medicare translate into worse access to health care? J. Am. Geriatr. Soc. 56 (2008) 2053– 2060, doi:http://dx.doi.org/10.1111/j.1532-5415.2008.01993.x.

[14] J. Edward, S. Morris, F. Mataoui, P. Granberry, M.V. Williams, I. Torres, The impact of health and health insurance literacy on access to care for Hispanic/ Latino communities, Public Health Nurs. (2018), doi:http://dx.doi.org/10.1111/ phn.12385.

[15] R.T. Braun, Y. Hanoch, A.J. Barnes, Tobacco use and health insurance literacy among vulnerable populations: implications for health reform, BMC Health Serv. Res. 17 (2017) 729, doi:http://dx.doi.org/10.1186/s12913-017-2680-7.

[16] 2012-2016 American Community Survey 5-year Estimates, U.S. Census Bureau, 2016. (Accessed 3 February 2018) https://factfinder.census.gov/faces/ tableservices/jsf/pages/productview.xhtml?src=bkmk.

[17] Small Area Health Insurance Estimates (SAHIE) Program, U.S. Census Bureau, 2015. (Accessed 3 February 2018) https://www.census.gov/programs-surveys/ sahie.html.

[18] K.A. Paez, C.J. Mallery, H. Noel, et al., Development of the Health Insurance Literacy Measure (HILM): conceptualizing and measuring consumer ability to choose and use private health insurance, J. Health Commun. 19 (2014) 225– 239, doi:http://dx.doi.org/10.1080/10810730.2014.936568.

[19] N.S. Morris, C.D. MacLean, L.D. Chew, B. Littenberg, The single item literacy screener: evaluation of a brief instrument to identify limited reading ability, BMC Fam. Pract. 7 (21) (2006), doi:http://dx.doi.org/10.1186/1471-2296-7-21.

[20] C.D. Norman, H.A. Skinner, eHEALS: the eHealth literacy scale, J. Med. Internet Res. 8 (2006) e27, doi:http://dx.doi.org/10.2196/jmir.8.4.e27.

[21] G. Paramio Pérez, B.J. Almagro, A. Hernando Gómez, J.I. Aguaded Gómez, Validation of the eHealth literacy scale (eHEALS) in Spanish university students, Rev. Esp. Salud Publica 89 (2015) 329–338, doi:http://dx.doi.org/ 10.4321/S1135-57272015000300010.

[22] Centers for Disease Control and Prevention - Behavioral Risk Factor Surveillance System. https://www.cdc.gov/brfss/index.html, (Accessed 17 February 2018).

[23] IBM, Statistical Package for the Social Sciences, 24 ed., IBM Corp., Armonk, NY, 2016.

[24] J.C. Nunnally, I.H. Bernstein, Psychometric Theory, 3rd ed., McGraw-Hill, New York, 1994.

[25] M.C. Politi, M.D. Kuzemchak, J. Liu, et al., Show me my health plans: using a decision aid to improve decisions in the federal health insurance marketplace, MDM Policy Pract (2016) 1.

[26] V. Brown, M. Russell, A. Ginter, et al., Smart choice health insurance©: a new, interdisciplinary program to enhance health insurance literacy, Health Promot. Pract. 17 (2016) 209–216, doi:http://dx.doi.org/10.1177/ 1524839915620393.

[27] J.O. Hero, A.M. Zaslavsky, R.J. Blendon, The United States leads other nations in differences by income in perceptions of health and health care, Health Aff. (Millwood) 36 (2017) 1032–1040, doi:http://dx.doi.org/10.1377/ hlthaff.2017.0006.

[28] P. Ubri, S. Artiga, Disparities in Health and Health Care: Five Key Questions and Answers, (2016) . (Accessed 27 February 2018) https://www.kff.org/ disparities-policy/issue-brief/disparities-in-health-and-health-care-five-key- questions-and-answers/.

[29] B.D. Sommers, A.A. Gawande, K. Baicker, Health insurance coverage and health – what the recent evidence tells us, N. Engl. J. Med. 377 (2017) 586–593, doi: http://dx.doi.org/10.1056/NEJMsb1706645.

[30] Office of Disease Prevention and Health Promotion, Disparities (Accessed 27 February 2018), https://www.healthypeople.gov/2020/about/foundation- health-measures/Disparities.

[31] H.E. Alcalá, J. Chen, B.A. Langellier, D.H. Roby, A.N. Ortega, Impact of the Affordable Care Act on health care access and utilization among Latinos, J. Am. Board Fam. Med. 30 (2017) 52–62, doi:http://dx.doi.org/10.3122/ jabfm.2017.01.160208.

[32] R. Fadlallah, F. El-Jardali, N. Hemadi, et al., Barriers and facilitators to implementation, uptake and sustainability of community-based health insurance schemes in low- and middle-income countries: a systematic review, Int. J. Equity Health 17 (2018) 13, doi:http://dx.doi.org/10.1186/s12939-018- 0721-4.

[33] T. Mathur, G. Das, H. Gupta, Examining the influence of health insurance literacy and perception on the people preference to purchase private voluntary health insurance, Health Serv. Manage. Res. (2018), doi:http://dx.doi.org/ 10.1177/0951484818760529 951484818760529.

[34] S. Chankova, S. Sulzbach, F. Diop, Impact of mutual health organizations: evidence from West Africa, Health Policy Plan. 23 (2008) 264–276, doi:http:// dx.doi.org/10.1093/heapol/czn011.

[35] C. Jehu-Appiah, G. Aryeetey, I. Agyepong, E. Spaan, R. Baltussen, Household perceptions and their implications for enrollment in the National Health Insurance Scheme in Ghana, Health Policy Plan. 27 (2012) 222–233, doi:http:// dx.doi.org/10.1093/heapol/czr032.

[36] The Commonwealth Fund, HHS Awards $67 Million in Navigator Grants to 105 Applicants (Accessed 11 February 2018), http://www.commonwealthfund. org/publications/newsletters/washington-health-policy-in-review/2013/aug/ august-19-2013/hhs-awards-67-million-in-navigator-grants-to-105- applicants.

[37] K. Pollitz, J. Tolbert, R. Ma, Survey of Health Insurance Marketplace Assister Programs, (2014) . (Accessed 5 February 2018) http://files.kff.org/attachment/ survey-of-health-insurance-marketplace-assister-programs-report.

2240 S. Ghaddar et al. / Patient Education and Counseling 101 (2018) 2233–2240

[38] K. Pollitz, J. Tolbert, R. Ma, 2015 Survey of Health Insurance Marketplace Assister Programs and Brokers, (2015) . (Accessed 5 February 2018) http://files. kff.org/attachment/report-2015-survey-of-health-insurance-marketplace- assister-programs-and-brokers.

[39] K. Pollitz, J. Tolbert, A. Semanskee, 2016 Survey of Health Insurance Marketplace Assister Programs and Brokers, (2016) . (Accessed 5 February 2018) http://files.kff.org/attachment/2016-Survey-of-Marketplace-Assister- Programs-and-Brokers.

[40] J. Giovannelli, E. Curran, Efforts to support consumer enrollment decisions using total cost estimators: lessons from the Affordable Care act’s Marketplaces, Issue Brief (Commonw. Fund) 3 (2017) 1–12.

[41] C.A. Wong, D.E. Polsky, A.T. Jones, J. Weiner, R.J. Town, T. Baker, For third enrollment period, marketplaces expand decision support tools to assist consumers, Health Aff. (Millwood) 35 (2016) 680–687, doi:http://dx.doi.org/ 10.1377/hlthaff.2015.1637.