Impact of Health Literacy on Medication Adherence Among Black Medicaid Beneficiaries with Hypertension in Delaware: A Cross-Sectional Study

Abstract

Racial and ethnic minorities are disproportionately affected by limited health literacy. Therefore, this study assessed census block health literacy level and medication adherence in Delaware among Black individuals with hypertension (HTN) receiving health care through Medicaid. This was a cross-sectional study of Black Delaware Medicaid beneficiaries (18–64 years old) from the 3 counties in Delaware (Kent, New Castle, and Sussex) from 2016 to 2019. The primary outcome was medication adherence (full adherence = 80%–100%, partial adherence = 50%–79%, and nonadherence = 0–49%) as a function of health literacy. Health literacy scores were categorized as below basic (0–184), basic (184–225), intermediate (226–309), and proficient (310–500). The results of the study showed that 18,958 participants (29%) had ≥1 HTN diagnosis during the study period. Mean area health literacy score for participants without HTN was significantly higher than participants with HTN (234.9 vs. 233.7, P < 0.0001). Men had lower odds of adherence compared with women (odds ratio [OR]: 0.83, 95% confidence interval [CI]: 0.75–0.92, P < 0.001). Increased time enrolled in Medicaid decreased full adherence. Participants 21–30 and 31–50 years of age are significantly less likely to have full adherence in comparison with participants 51–64 years of age (P < 0.0001). Participants living in an area with basic level of health literacy reported lower medication adherence than those living in an area with an intermediate level of health literacy (OR: 0.72, 95% CI: 0.64–0.81, P < 0.001). In conclusion, men, younger adults, increased time enrolled in Medicaid for the study period, and basic health literacy were significantly associated with low adherence to medication among 3 census blocks in Delaware.

Introduction

Medication adherence is the tendency in which patients follow medical therapy treatment guidelines as advised by their health care provider. Hypertension (HTN), or high blood pressure, can be attributed to nonadherence to antihypertensive medication. HTN is a major risk factor for heart disease and stroke, which are leading causes of death in the United States.¹ In 2019, HTN was a primary or contributing cause of death for 516,955 persons in the United States, making HTN a leading risk factor for cardiovascular mortality.¹ Black individuals account for the third largest racial/ethnic group in the United States (44 million) and in 2017, the highest prevalence of HTN across all races was among Black adults (54%).²

Medication adherence is a growing problem among minorities and individuals of lower socioeconomic status.^3–6 Minorities and individuals of lower socioeconomic status bear an unequal burden of heart disease and often report poorer health outcomes related to cardiovascular disease, including higher rates of HTN and diabetes among Black individuals.⁵ Sixty-six percent of Black individuals with HTN are aware and are receiving treatment for this chronic disease; however, their blood pressure control is lower (28.9%) compared with Whites (35.4%).^7,8

It is estimated that almost 40% of American adults have limited health literacy,⁹ and racial and ethnic minorities are disproportionately affected by limited and low health literacy.^10,11 Health literacy is the degree to which individuals can get access, understand, and use information and services to make individual health-related decisions.¹² There is a higher proportion of Black individuals (24%) that report below basic health literacy skills compared with White individuals (9%).¹³ Health literacy is a barrier to medication adherence and managing chronic conditions such as HTN and individuals with limited health literacy report being less knowledgeable about disease management, healthy behaviors, and they are less likely to utilize preventive services.^14,15

In 2018, 34% of Delaware Medicaid participants were Black.¹⁶ Black individuals are disproportionately represented in Medicaid compared with the state population and Black individuals made up ∼23% of Delaware residents in the same year.¹⁶ A previous study examined health literacy from a Medicaid population in Ohio and found 11.3% had marginal and 22.2% had inadequate literacy.¹⁷ Similarly, Copeland et al reported inadequate literacy (54.5%) among 112 Medicaid beneficiaries.¹⁸ Few studies have examined the impact of health literacy on medication adherence among Black individuals with HTN that receive health care through Medicaid. To address this gap, the authors conducted a cross-sectional study to assess area health literacy level and medication adherence among Black individuals with HTN. The authors hypothesize that low census block health literacy may be associated with low hypertensive medication adherence.

Materials and Methods

Study design and population

The study was a longitudinal cross-sectional study of Medicaid data for the state of Delaware. The study population comprised Black Medicaid beneficiaries between the age of 18 and 64 years from the 3 counties in Delaware (Kent, New Castle, and Sussex) during the period 2016–2019, identified from the Delaware Medicaid Eligibility Files. The Delaware Medicaid Encounter and Claims data were used to identify HTN, and included only beneficiaries with ≥1 month of comprehensive eligibility. HTN was defined as a beneficiary having ≥1 medical claim with an HTN diagnosis in any setting (inpatient or outpatient) based on the diagnosis code [I10: Essential (primary) HTN] from the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM).¹⁹

Beneficiaries with either a stroke (ICD10 I60–I63) or a heart attack (I519 and I21) in the year before their first HTN diagnosis were excluded from the analysis. The purpose of this exclusion criteria was to exclude any patients who had evidence of a previous serious cardiovascular-related disease, especially those that are linked to uncontrolled HTN, to focus on capturing full patient journey of patients with HTN that are in a less severe state of their disease. There were no additional exclusion criteria in this analysis.

Outcome variables

The primary outcome was medication adherence as a function of health literacy. Prescription claim files were used to identify HTN drugs based on almost 4000 National Drug Codes listed in the National Committee for Quality Assurance Health care Effectiveness Data and Information Set 2017 Final National Drug Codes Lists.²⁰ HTN drugs included renin angiotensin receptor antagonists, angiotensin converting enzyme inhibitors, angiotensin II inhibitors, alpha-beta blockers, beta-blockers, calcium channel blockers, diuretics, selective aldosterone receptor antagonists, and antihypertensive combinations. A medication adherence variable was constructed based on days of fills of the prescription and controlled for the proportion of days covered by Medicaid during the period. This categorization for medication adherence has been used in previous studies using Medicaid data.^21–23 The following categories were assigned for adherence: full adherence = 80%–100%, partial adherence = 50%–79%, and nonadherence = 0–49%.

The addresses of all beneficiaries in the study population were geocoded to census block groups using Esri ArcGIS Version 10.8.1,²⁴ with an 87.7% geocoding success rate. Failure to geocode was due to post office boxes, incomplete addresses, or unknown addresses. The census block group of each beneficiary was linked to a measurement of health literacy developed by the University of North Carolina at Chapel Hill.²⁵ The University of North Carolina provides health literacy estimates for every census block group in the United States, the estimates can be downloaded from a publicly available website.²⁵ The health literacy estimates range from 0 to 500 with a higher score indicating higher health literacy. Scores are categorized into 4 categories: below basic (0–184), basic (184–225), intermediate (226–309), and proficient (310–500).

Statistical analysis

Descriptive statistics were performed on patient demographics and reported as N (%). Demographic variables in this analysis included gender, age group, total time of comprehensive coverage during the study period, and county. For HTN patients with an HTN prescription fill, number of HTN prescription fills and days after diagnosis were reported as N (%). The primary outcome variable was medication adherence as a function of health literacy. Chi-square tests were used to compare categorical variables between Black Medicaid beneficiaries with and without HTN.

Binary logistic regression models were used to analyze the outcome variable and the impact of area health literacy of a beneficiary's census block group for participants with full medication adherence, and adjustments were made for covariates (gender, age group, and total time of comprehensive coverage during the study period). Mantel–Haenszel odds ratios (ORs) with 95% confidence intervals (CIs) were reported to quantify the effect.¹⁹ Tests were 2-sided and P < 0.05 was considered statistically significant. Statistical analyses were performed with SAS, version 9.4 (SAS Institute, Inc., Cary, NC).²⁶ The data underlying this article are available in its entirety in this manuscript. This study was approved by the University of Delaware Institutional Review Board.

Results

Patient demographics

Among 65,639 Black Medicaid beneficiaries in the database, 18,958 participants (29%) had ≥1 HTN diagnosis during the study period (Table 1). The majority of participants with HTN were 50–59 years of age. Participants without an HTN diagnosis tended to be younger (10%> 50 years old, mean = 32.5) compared with participants with HTN (44%> 50 years, mean = 45.5, t = −132.8, P < 0.0001). There was a significant higher percentage of female participants with HTN than without HTN (59% vs. 55%, χ² = 80.4 P < 0.0001). Most of the total study population resided in New Castle County. A higher proportion of participants with HTN lived in Sussex County than participants without HTN (16% vs. 12%, χ² = 134.2, P < 0.0001). Participants with HTN had Medicaid coverage during the study period longer than participants without HTN (33.6 vs. 34.6; t-test = −66.1, P < 0.0001). Figure 1 represents the distribution of Black Medicaid participants in the study by census block groups, with and without HTN, respectively.

Table 1.

Patient Demographics

N (%)	Total (N = 65,639)	HTN (N = 18,958)	No HTN (N = 46,681)	P
Age group				<0.0001
18–19	3426 (5)	177 (1)	3249 (7)
20–29	21,751 (33)	2033 (11)	19,718 (42)
30–39	16,366 (25)	3910 (21)	12,456 (27)
40–49	10,897 (17)	4583 (24)	6314 (14)
50–59	9949 (15)	6003 (32)	3946 (8)
60–64	3250 (5)	2252 (12)	998 (2)
Gendera				<0.0001
Female	36,901 (56)	11,175 (59)	25,726 (55)
Male	28,734 (44)	7783 (41)	20,951 (45)
County				<0.0001
Kent	13,602 (21)	3914 (21)	9688 (21)
New castle	43,311 (66)	12,074 (64)	31,237 (67)
Sussex	8726 (13)	2970 (16)	5756 (12)
Months of eligibility				<0.0001
1–6 months	9266 (14)	972 (5)	8294 (18)
7–12 months	9004 (14)	1615 (9)	7389 (16)
13–24 months	12,292 (19)	3198 (17)	9094 (19)
25–26 months	10,551 (16)	3060 (16)	7491 (16)
37–48 months	24,526 (37)	10,113 (53)	14,413 (31)

^a Gender was missing for 4 participants in the cohort.

HTN, hypertension.

FIG. 1.

Distribution of Black Medicaid Participants by Census Block Groups (A) with HTN and (B) without HTN.

Health Literacy and Medication Adherence

All participants lived in census block group with either the basic or intermediate level of literacy (Fig. 2). Participants with an HTN diagnosis had a higher percentage in the basic literacy level compared with participants without HTN (29% vs. 26%). However, the area's mean literacy score for participants without HTN was significantly higher than participants with HTN (234.9 vs. 233.7, t = 10.8, P < 0.0001). Figure 3 presents a map of the distribution of literacy scores within Delaware. Among participants with HTN, 7382 participants (39%) had an HTN prescription fill (Table 2). The mean number of fills was 6.1 (ranging from 1 to 22) and almost 3/4 of participants had 3 or more fills. The mean number of days for a prescription fill after diagnosis for HTN was 118.3, ranging from 0 (same day) to 1438 days. More than half of these participants (60%) had a claim for HTN drugs within 30 days of HTN diagnosis. The average percent of adherence days was 58.2%, ranging from 2% to 100%.

FIG. 2.

Health literacy level among Black Medicaid participants. HTN indicates hypertension.

FIG. 3.

Health literacy estimates range from 0 to 500 with a higher score indicating higher health literacy. Literacy scores are categorized into 4 categories: below basic (0–184), basic (184–225), intermediate (226–309), and proficient (310–500).

Table 2.

Medication Adherence Among Participants with Hypertension with a Prescription Fill

N (%)	HTN (N = 7382)
No. of HTN prescription fills
1	1115 (15)
2	824 (11)
3–5	1683 (23)
6–10	2314 (31)
11 and over	1446 (20)
Days after HTN diagnosis
Same day	1461 (20)
1–30	2946 (40)
31–180	1666 (22)
181–365	509 (7)
366–730	464 (6)
731 and over	331 (5)
Adherence
Nonadherence	2920 (40)
Partial adherence	2012 (27)
Full adherence	2450 (33)

Medication adherence was defined based on days of fills of the prescription and categorized by adherence: full adherence = 80%–100%, partial adherence = 50%–79%, and nonadherence = 0%–49%.

HTN, hypertension.

Table 3 shows the OR estimates for participants with HTN and full medication adherence. The majority of participants (40%) had nonadherence, with approximately one third (33%) of the participants on HTN drugs having full adherence. Men had lower odds of adherence compared with women (OR: 0.83, 95% CI: 0.75–0.92, P < 0.001). Increased time enrolled in Medicaid for the study period slightly decreased full adherence. Participants 21–30 and 31–50 years of age are significantly less likely to have full adherence in comparison with participants 51–64 years of age (P < 0.0001). Participants living in an area with basic level of health literacy reported lower medication adherence than those living in an area with an intermediate level of health literacy (OR: 0.72, 95% CI: 0.64–0.81, P < 0.001).

Table 3.

Odds Ratio Estimates for Hypertension Participants with Full Medication Adherence (N = 2450)

Effect	Point estimate	95% Wald confidence limits		P
Male (reference = female)	0.827	0.748	0.915	<0.001
Age (reference = 51 plus)
18–20	0.571	0.301	1.082	NS
21–30	0.358	0.275	0.465	<0.0001
31–50	0.581	0.525	0.644	<0.0001
Total time in study period (months)	0.985	0.982	0.989	<0.0001
Basic literacy (reference = intermediate literacy)	0.718	0.641	0.805	<0.0001

Health literacy estimates range from 0 to 500 with a higher score indicating higher health literacy. Literacy scores are categorized into 4 categories: below basic (0–184), basic (184–225), intermediate (226–309), and proficient (310–500).

NS, not significant.

Discussion

Few studies have examined medication adherence in a Medicaid population of Black individuals with HTN and, furthermore, there is a lack of evidence describing census block health literacy among this population. The main objective of this study was to assess census block health literacy level and medication adherence among Black individuals with HTN receiving health care through Delaware Medicaid. A major finding of the study revealed that less than half of participants received an appropriate prescription for treatment and most patients were nonadherent. Younger patients were less likely to report higher adherence compared with older patients. The study hypotheses indicated that lower census block health literacy was significantly associated with lower census block adherence to medication. Furthermore, Black men lived in areas with lower levels of health literacy and lower medication adherence compared with Black women.

In respect to a Medicaid population of Black individuals with HTN, there are relatively few studies that have evaluated the association between health literacy and medication adherence that can be used to draw comparisons with the study findings. However, a single previous report assessing individual level health literacy and adherence behaviors in African American adults with HTN (N = 121) found no statistically significant association between health literacy and adherence; whereas census block health literacy of Medicaid participants in this study living in an area with lower health literacy had lower adherence to medication.²⁷

The same study also reported those who were younger and reported poor or fair health status were less likely to adhere to medication,²⁷ supporting the study results that younger patients were less likely to have full adherence compared with older patients. Similarly, an integrative review by Du et al summarizing health literacy and outcomes in hypertensive patients (N = 19 studies) reported that persons with low health literacy are likely to have low knowledge of HTN.²⁸ Studies included in this same review were not specific to Black individuals or Medicaid beneficiaries.

Consistent with previous investigations of Medicaid beneficiaries,^29–33 Black individuals were less likely to be adherent to medication compared with White individuals. In this study, age was also significantly associated with medication adherence. The study findings were similar in scope to studies of California and Maryland Medicaid beneficiaries^32,34; however, their findings revealed that younger patients were more likely to be nonadherent to hypertensive medications. Shaya et al specifically found that participants <40 years were less likely to be adherent than patients >60 years of age.³⁴ The study results reported that participants 21–30 and 31–50 years of age were less likely to report high adherence in comparison with participants 51–64 years of age.

In this study and a previous investigation of Medicaid beneficiaries,³⁵ men were less likely to be adherent than women. However, the association of gender on medication adherence has been inconsistent among previous studies, with several studies concluding women reported lower medication adherence than men,^31,36,37 and others finding no influence of gender on medication adherence.^34,38 Khanna et al suggested that factors, including lack of time and limited access to specialists, are important barriers for females enrolled in Medicaid.³¹

This study expands upon published literature on Medicaid populations to show that nonadherence among Black individuals with HTN in Delaware remains a significant problem. This study also provides new insights on the impact of health literacy on medication adherence, an area of research that remains drastically understudied. Most participants with HTN lived in areas with a basic health literacy and lower mean literacy scores compared with participants without HTN. There was a significant association between lower health literacy leading to low medication adherence, exacerbating disparities, and potentially leading to poorer health outcomes for hypertensive Black individuals receiving health care through Medicaid.

This study highlights the necessity of providing literacy assistance and health education to promote healthy behaviors and lifestyle change such as medication adherence. Technology-based interventions to promote health literacy have been established as an effective tool for bolstering literacy to manage HTN.³⁹ These technologies could be used in populations included in this study to provide literacy assistance and improve health outcomes for Black individuals with HTN. Furthermore, future research should examine the association of health literacy on incidence of severe complications from uncontrolled HTN (such as strokes and heart attacks) and replicate this analysis in other minority populations that experience health disparities, both in Medicaid and commercially insured populations.

There are several limitations of this study. This study utilized administrative claims rather than electronic medical records. Diagnosis of HTN relied on at least one billed claim of HTN for an individual. The absence of clinical data limits the ability to clinically confirm cases of HTN and patients with borderline HTN may have been included in this cohort of patients. Owing to the nature of claims, this study was limited in demographic and clinical variables that could be captured for stratification of participants. In addition, claims data do not include other potential confounders that may affect medication adherence and health literacy in this population, including disease severity, comorbidities, and information on social determinants of health (socioeconomic status, access to quality education, access to health care, transportation, etc.).

Thus, the authors were unable to control for these variables in this study. Furthermore, provider level perspectives are not included in claims data, which could help explain challenges in controlling blood pressure and improving health outcomes in this population. Qualitative studies using semistructured interviews of medical providers could complement future research aiming to close this unmet need in Black individuals with HTN.

Participants in this study were hypertensive Black Medicaid beneficiaries in the state of Delaware. Therefore, these results may not be generalizable to other races, insurance types, and may not reflect the experiences of individuals living in other areas of the United States. Specifically, health literacy levels may differ between states and individuals due to varying educational and other socioeconomic factors. Even though this analysis could not control for these socioeconomic factors, limited heterogeneity can be assumed due to the need for participants to qualify for Medicaid under its prespecified program requirements.

In addition, census block health literacy data may not reflect individual-level health literacy and serves only as a proxy for individual health literacy because of ecological fallacy.⁴⁰ Finally, this study was cross-sectional in nature; thus, a causal relationship cannot be concretely determined. However, a major strength of this study is presenting foundational contemporary data on health literacy and medication adherence among hypertensive Black individuals receiving health care through Medicaid in Delaware.

Conclusion

Among 3 census blocks in Delaware, men, younger age, increased time enrolled in Medicaid for the study period, and lower census block health literacy were significantly associated with lower adherence to medication. This study highlights the necessity of providing literacy assistance and health education to promote healthy behaviors and lifestyle change such as medication adherence. Future research should examine the association of health literacy on incidence of severe complications from uncontrolled HTN (such as strokes and heart attacks) and replicate this analysis in other minority populations that experience health disparities, both in Medicaid and commercially insured populations.

Authors' Contributions

Dr. Cuffee was the principal study investigator. Ms. McDuffie conducted the data analysis and drafted the methods section of the article. Dr. Butzner wrote the first draft of the article with input from other authors. All authors were involved in the study design, contributed to the interpretation of the results, writing, and revising the article, and approved the decision to submit the final article for publication.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

The study was conducted and funded by University of Delaware College of Health Sciences. Access to health care claims data was supported by the Institutional Development Award (IDeA) from the National Institute of Health's National Institute of General Medical Sciences under grant no. P20GM103446.

Supplementary Material

Supplementary Material