Medication Adherence in American Indians With Type 2 Diabetes: An Integrative Review Free

An integrated literature review on medication adherence in American Indians with type 2 diabetes was used to ensure inclusion of all research methodologies, as literature on this topic is limited (7). After consulting with a health science reference librarian, the databases selected to search for articles were PubMed and the Cumulative Index of Nursing and Allied Health Literature. We found 1,012 articles published between 1974 to 2021 by using the search terms (American Indian OR Native American OR Indians, North American OR Alaska Natives) AND (diabetes OR diabetes mellitus) AND (adherence OR compliance OR medication adherence OR medication barriers OR medication facilitators OR self-management OR barrier OR facilitator OR management).

We selected studies if they 1) were peer-reviewed studies, 2) included American Indian participants ≥18 years of age with type 2 diabetes, 3) discussed medication adherence in the results, 4) were conducted in the United States, and 5) were written in English. Studies were excluded if 1) all participants had type 1 diabetes, 2) participants had gestational diabetes, or 3) children or adolescents were included.

The systematic search was conducted using the Preferred Reporting Items for Systematic Review and Meta-Analysis (Figure 1) (8). A total of 1,012 articles were found, and 288 duplicates were then removed. The remaining 724 articles were screened using the title, with 657 irrelevant articles removed. Of the remaining 67 articles, an additional 50 were removed because they did not meet the inclusion or exclusion criteria. The remaining 17 articles were examined independently by two investigators using the Standard Quality Assessment Criteria for Evaluating Research Papers from a Variety of Fields (9). Disagreements were discussed to reach a consensus on final scores. The 17 articles met a minimum threshold of 55%, and none were eliminated from the review.

The characteristics of the 17 studies are presented in Table 1. Five studies used a qualitative design (10–14), five used a mixed-methods design (15–19), and seven used a quantitative methodology (20–26). American Indians made up the totality of participants in 10 studies (10–14,17–19,22,23), and seven studies included participants from other races or ethnicities (15,16,20,21,24–26). Nine studies focused on type 2 diabetes (10–12,15,18–21,23), whereas in seven studies, both type 1 diabetes and type 2 diabetes were included or diabetes type was not specified (13,16,17,22,24–26). One study focused on prediabetes, diabetes, and caregivers (14). Two studies included participants with either or both diabetes and hypertension, but did stratify results related to medication adherence by diagnosis (16,26).

In six of the 17 studies, medication adherence was a primary focus (17,19,23–26). In 11 of the articles, medication adherence was discussed within the context of self-management, but was not the primary focus (10–16,18,20–22).

Various methods were used to examine medication adherence (Table 1). Prescription claims data were used to assess adherence in one study, and good medication adherence was determined to be ≥80% on a scale from 0% (fully nonadherent) to 100% (fully adherent) (24).

Validated self-reported measures (5) assessing adherence were included in six studies (15,17,19,20,23,25). Four studies included the 4-item Morisky Medication Adherence Scale (MMAS-4) (17,19,23,25), which has dichotomous (yes/no) questions that are scored from 0 to 4. We suspect that the MMAS-4 data were likely shared in two of these four studies (i.e., data from the same participants in the same project used in two studies), but only one reported percentages of medication adherence (19,23). The Summary of Diabetes Self-Care Activities (SDSCA) measure was used during interviews in two studies to assess the number of days out of 7 days participants were adherent to their medication (15,20). One study included the Health Behavior Scale (HBS), a measurement tool not validated in participants with diabetes (22).

The remaining studies used several different methods to assess medication adherence and associated factors, including a behavioral assessment survey given as part of an interview (16), a diabetes self-management behavior intake form and lifestyle survey (21), and adherence assessment interviews (11). Interviews were used in two additional studies (12,26), and focus groups were used in three studies (10,13,14). A Talking Circle, a traditional forum in which individuals can discuss, provide, and receive support, was used to assess factors associated with adherence in one study (18).

Overall adherence to medication was assessed in three of the four studies that used the MMAS-4. In one study, 27.7% of participants had high medication adherence (score of 4 of 4), and 20.5% had low adherence (0–1 of 4), and average adherence to diabetes medication was considered moderate (2.54 of 4) (19). Another study using the MMAS-4 found that average adherence to diabetes medication was low (0.90 of 4) (17). In the fourth study, 56.3% of females reported nonadherence to medication (>1 of 4) (25). In the first three studies, a higher score on the MMAS-4 indicated better adherence (17,19,23), whereas in the fourth, a score >1 indicated nonadherence (25). One study using the SDSCA measure found that 92.7% of participants reported being fully adherent to their medication regimen (20), and another study using the SDSCA measure found that most participants reported being adherent to medication on 6 of 7 days (15).

The only study using prescription claims data did not include medication adherence rates but found that American Indians were less adherent to oral diabetes medication than non-Hispanic Whites (24). Authors of another study planned to calculate the adherence rate using pharmacy records but were unable to consistently access pharmacy records (26).

Self-report forms and surveys used in two studies found that most participants took their medication as directed (16,21). Authors of the study using an intake form and lifestyle survey reported that 74% of participants took their medication as directed (21). In contrast, authors of a study using focus groups found that many participants did not adhere to their medication regimen, even when they recognized the benefits of the medication (13). In a study using interviews along with an adherence assessment (11) and another study using the HBS (22), investigators found that participants were more adherent to medication than to their dietary regimen. Also, medication adherence was lower in elders with strong ties to their Indigenous culture (11).

Medication adherence barriers and facilitators were examined in 12 of the 17 studies (10–15,17,18,22,23,25,26) (Table 2). Three of these 12 studies included responses from participants from other racial and ethnic groups (15,25,26). The association between gender and medication adherence was examined in three studies, age and medication adherence was reported in two studies, and education and medication adherence was reported in one study (19,23,25) (Table 2). One of the three studies that looked at gender and adherence included responses from women participants from other minority groups (25). Financial barriers and diabetes fatigue related to treatment were the most common barriers noted (12,13,17,18,26), and knowledge that the medication was effective and assistance from care partners or community health workers were the most noted facilitators (12,15,17,26). Sociodemographic factors associated with lower adherence were female gender, younger age, and lower levels of education (19,23).

The goal of this integrative review was to synthesize existing literature on medication adherence and associated facilitators, barriers, and sociodemographic variables in American Indians with type 2 diabetes. The 17 reviewed studies had a variety of methods and measures, and many included relatively small numbers of American Indians with type 2 diabetes. Findings indicate considerable variability in medication adherence and a number of facilitators and barriers. For the few studies considering sociodemographic variables, findings indicate that women, younger people, and individuals with lower education levels had lower medication adherence than their counterparts.

We found variability in the methods used to measure medication adherence across studies. Medication adherence was examined using prescription refill data in one study (24). Validated self-report measurements within this review included the MMAS-4 (17,19,23,25) and the SDSCA (15,20). Currently, there is no universal gold standard for measuring diabetes medication adherence (27,28). Self-report measures are commonly used because they are low-cost and easy to administer (29). In several studies with self-report measures, including the SDSCA, the HBS, surveys, and an intake form, the majority of participants reported taking their medication as directed most or all of the time (15,16,20–22). The preponderance of self-report measures in most of the reviewed studies may overestimate medication adherence as a result of biases such as recall bias and social desirability (29). In future studies of American Indians’ adherence with type 2 diabetes medications, we recommend using validated measures and incorporating two or more measures when possible (29).

Overall, there was variability in findings regarding medication adherence among American Indians with type 2 diabetes. This variability may have been the result of the various approaches used to measure and report mediation adherence. Low adherence was reported in several studies (13,17,25), in addition to one study reporting lower adherence in American Indians compared with non-Hispanic Whites (24). These findings are consistent with previous studies noting racial and ethnic disparities in medication adherence in other underrepresented racial and ethnic groups (5,6,30,31). Because of the variability in findings, additional research is needed to examine adherence to diabetes medication using validated measures in American Indian communities.

Many of the barriers found in our review such as knowledge deficits, forgetfulness, and financial issues are consistent with barriers to medication adherence in the general population (5,10,13,17,26,31). Consistent with our findings, prior literature noted that African American participants, a population also facing diabetes disparities, similarly reported concerns regarding adverse effects, as well as apprehension and frustration toward medications as barriers to adherence (32). In our review, participants from urban settings and medically underserved communities and those without access to medication and care provided by Indian Health Service benefits listed cost as a barrier (13,17,26). To fully understand the extent to which cost affects medication adherence, additional research is needed, specifically in reservation-dwelling American Indians who use tribal health care services, which provide medical care and prescriptions with no cost to patients.

Social support or involvement of community health workers, each noted in one included study as facilitators of medication adherence, have resulted in mixed findings in the literature (5,12,26,28,31,33–35). Diabetes empowerment and patient-centered care were both linked to medication adherence in our review (23). There is evidence from previous literature that diabetes empowerment, patient-centered care, and patient-centered medical homes may improve mediation adherence in patients with diabetes and other chronic diseases (36–38). Because of differences in values and beliefs regarding medications, culturally appropriate methodologies such as Talking Circles (18,21) should be considered when designing research studies to better understand facilitators and barriers in American Indians and other racial and ethnic minorities.

Sociodemographic factors were also found to be associated with medication adherence. Although there is no consensus on the effects of sex or gender on medication adherence in the general population (31), some studies have shown that women have lower adherence than men, which is consistent with our findings (19,23,25,39). Shiyanbola and Nelson (25) explained that sociocultural factors such as having the role of caretaker within families may affect adherence to medication in some women with diabetes. Also, congruent with our findings (19,23), younger individuals aged 20–44 years have been found to be less adherent than individuals ≥45 years in the general population, and those with a high school education were less likely to be adherent than those with a college education (39). Future studies should further explore how sociodemographic variables affect medication adherence, specifically in American Indian populations.

Because of the limited published research in this area, this review included only 17 articles. The strengths of this integrative review were the inclusion of all articles that addressed medication adherence within American Indian communities and a rigorous approach to rating the quality of the selected articles. Notable limitations were that few studies measured medication adherence, making it difficult to reach conclusions regarding adherence to diabetes medication in the American Indian population. Recall bias can be an issue in studies with self-report measures, which were used in the majority of the reviewed studies. Other limitations include studies that did not report findings by race (15,20,21,25,26), studies that did not exclusively include participants with type 2 diabetes (13,14,16,17,24–26), and studies that included a small percentage of American Indians (16,24,26). In addition, measures not validated in participants with diabetes were used in some studies to assess adherence, and one study reported having inadequate power because of its small sample size (n <20) (25).

Diabetes empowerment, patient-centered care, and efforts to address barriers such as financial issues, diabetes fatigue, and knowledge deficits may be important strategies to improve medication adherence (11,13,15,17,23,26). Also, the use of community health workers in partnership with pharmacists to address barriers may be a useful strategy (26). Finally, there are many distinct Indigenous tribes in the United States, and culturally responsive practices need to be used while working alongside tribal communities.

Medication adherence in American Indians with type 2 diabetes was low in several studies, but the results were variable. More research is needed to understand medication adherence and associated factors in the American Indian population, specifically in reservation-dwelling American Indians. Future research using validated measures is also needed to better understand medication adherence and its associated factors in this population. It is imperative that researchers and health care providers practice cultural competence to address medication barriers and incorporate facilitators to improve diabetes medication adherence.

This publication was made possible by a graduate fellowship award from the University of Florida (T.N.) and grants R01NR020386 and U54CA233444 from the National Institutes of Health (NIH) National Institute of Nursing Research (NINR), and National Cancer Institute (NCI). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NINR and NCI. The final peer-reviewed manuscript is subject to the NIH Public Access Policy.

No potential conflicts of interest relevant to this article were reported.

T.N. researched data and wrote the manuscript. D.J.W. conceptualized the literature review and reviewed/edited the manuscript. L.S. conceptualized the literature review, evaluated articles, and reviewed/edited the manuscript. T.N. is the guarantor of this work, and as such, had full access to the included articles and takes responsibility for the integrity and accuracy of this integrative review.