Medication Adherence in People of Culturally and

Adherence

Medication Adherence in People of Culturally and Linguistically Diverse Backgrounds: A Meta-Analysis Elizabeth Manias and Allison Williams

edication adherence involves the extent to which a person’s medication-taking behavior conforms to health professionals’ advice. When people are not adherent to their medications there is a greater likelihood that therapeutic goals are not achieved, which leads to increased mortality and morbidity. Estimates for nonadherence range from 4% to 92%.1 Medication nonadherence costs approximately $100 billion a year in the US and leads to numerous, serious adverse events.2,3 The proportion of people of culturally and linguistically diverse (CALD) backgrounds has always been relatively high in the US and other developed countries. In 2000, about 33% of the US population identified themselves as belonging to a racial or ethnic minority group. By 2050, it is estimated that individuals of CALD backgrounds will account for almost half of the US population.4 Several studies have consistently demonstrated that individuals of CALD backgrounds have lower rates of medication adherence than the general population.5-7 In a retrospective cohort study of 56,561 people with hypertension and dementia, differences in medication adherence rates were observed for various drug classes.5 Hispanic and African American people had significantly lower medication

M

BACKGROUND: Medication adherence is of particular importance for people of culturally and linguistically diverse (CALD) backgrounds due to language difficulties, lack of social and organizational supports, lack of access to healthcare resources, and disengagement with the health-care system. OBJECTIVE: To evaluate the impact of interventions to improve medication adherence in people of CALD backgrounds through a systematic review and meta-analysis. METHODS:

A search was performed using the following databases: Cochrane Database of Systematic Reviews, Cumulative Index to Nursing & Allied Health Literature, EMBASE, Journals@Ovid, PsychInfo, PubMed, Science Direct, Scopus, and Web of Science. Databases were searched from January 1978 to October 2009.

RESULTS: Forty-six articles reviewed were assessed as being relevant, which included 36 randomized controlled trials, 2 observational cohort studies, and 8 quasi-experimental studies. The most common method for assessing medication adherence was self-reporting measures, such as the Morisky Scale and its modifications. Few studies used combinations of adherence measures, and adherence involving a medication event monitoring system (MEMS) was used in only 6 studies. Individuals of CALD backgrounds were recruited with people of non-CALD backgrounds and subsequent analyses tended to be undertaken of the whole sample. Twenty studies showed statistically significant improvements in medication adherence, 15 of which were randomized controlled trials. Six of the successful interventions involved delivery by a bilingual person or the use of translated materials and 4 involved the use of a conceptual model. Metaanalyses demonstrated modest improvements in medication adherence. CONCLUSIONS: Relatively little high-quality work has been conducted on adherence-enhancing interventions for people of CALD backgrounds. Greater attention needs to be given to examining the needs of specific CALD population groups. Future researchers should consider rigorously testing interventions that take into account the enormous diversity and differences that exist within any particular CALD group. KEY WORDS: culturally and linguistically diverse background, intervention, medication adherence, meta-analysis, systematic review.

Ann Pharmacother 2010;44:964-82. Published Online, 4 May 2010, theannals.com, DOI 10.1345/aph.1M572

Author information provided at end of text.

964

n

The Annals of Pharmacotherapy

n

2010 June, Volume 44

Downloaded from aop.sagepub.com by guest on October 11, 2013

theannals.com

adherence rates (66% and 64%, respectively) compared with white people (69%) for the use of angiotensin-converting enzyme inhibitors, acetylcholinesterase inhibitors (63% and 60%, respectively, compared with 70%), and memantine (78% and 75%, respectively, compared with 83%).5 In a prospective cohort study of 122 patients with mental health problems, Hispanic and African American people had lower medication adherence rates (77% and 68%, respectively) than whites (90%).6 In a case-control study of 764 people with diabetes, multivariate analyses showed that black people were more likely to demonstrate poor control of hemoglobin A1c, systolic blood pressure, low-density lipoprotein cholesterol, and total cholesterol levels due to missed medication doses (adjusted odds ratio 1.96) compared with white people (adjusted odds ratio 1.13).7 Black people were also significantly more likely than white people to report lack of knowledge about medication dose and purpose and difficulties with asking health professionals about medication-related problems.7 The Agency for Healthcare Research and Quality has regularly found that extensive health disparities, in terms of quality of care and access to resources, continue to persist for individuals of CALD backgrounds.4 In acknowledgment of these health disparities, the Office of Minority Health in the US has released 14 National Standards on Culturally and Linguistically Appropriate Services (CLAS) to facilitate culturally competent care.8 In the US, it is expected that the standards are integrated throughout organizations in partnership with the particular communities being served. In considering the range of effects produced by interventions addressing medication adherence on general populations, a recent systematic review found that only small improvements occurred and the results were inconsistent.9 For acute conditions, 5 of 10 interventions showed an effect on medication adherence while, for chronic conditions, 36 of 83 interventions were associated with improved adherence. It is important to explore the nature of interventions involving people of CALD backgrounds since the proportion of these individuals is increasing. Clearly, there is a need to educate these individuals and health professionals about appropriate ways to improve treatment outcomes. A major way to achieve this aim is to examine adherence interventions aimed specifically at CALD populations. The purpose of this study was to systematically review the research literature and to conduct meta-analyses on interventions undertaken with people of CALD backgrounds to improve medication adherence. Methods DATA SOURCES AND SELECTION

Search Strategy

Databases searched included the Cochrane Database of Systematic Reviews, Cumulative Index to Nursing & Allied theannals.com

Health Literature, EMBASE, Journals@Ovid, PsychInfo, PubMed, Science Direct, Scopus, and Web of Science. These databases were searched from January 1978 to October 2009. The term culturally and linguistically diverse background is not a recognized search term for several databases. A preliminary search of the National Library of Medicine was undertaken to determine appropriate MeSH terms to employ for the review.10 The following MeSH terms were identified and used in combination for the search: ethnic groups, healthcare disparities, medication adherence, cultural diversity, transients and migrants, emigrants and immigrants, intervention studies, and outcome assessment. Of the databases searched, MeSH terms are associated with PubMed; however, these terms were helpful in identifying relevant and appropriate papers in the other databases. The search results were limited to articles published in English. Inclusion and Exclusion Criteria

Controlled and uncontrolled, prospective and retrospective, and randomized studies that had medication adherence as a primary or secondary outcome measure were examined for the review. Patients were considered if they were prescribed medication for an acute or chronic condition. Caregivers who had responsibility in managing the medication needs of patients, such as parents of young children, were also considered. Only studies that involved evaluation of an intervention of a psychosocial or educational nature were examined. Examples of interventions assessed in studies included education (eg, information, feedback, mailed material), technology (eg, electronic timer devices), facilitators, scheduled appointments, home-based visits, telephone consultations, and support through disease self-management. Thus, interventions that involved manipulations of medications, such as comparisons between once-daily or twice-daily regimens, were not included. Interventions could be directed at individuals or groups, and at patients or caregivers. Studies were considered if they focused on people of CALD backgrounds in some way. For instance, they were included if a model was used that involved CALD characteristics, the background involved some emphasis of the adherence problem existing in people of CALD backgrounds, there was targeted recruitment of individuals from CALD backgrounds, the intervention addressed CALD issues, or the results involved analysis of CALD subgroups or controlled for these subgroups. Articles were excluded if the research involved the mere reporting of CALD group participation in the demographics, without some acknowledgment of the particular needs of these people. Case studies and epidemiologic studies that did not involve the testing of an intervention were excluded. Research disseminated through “gray” literature such as conference papers and unpublished reports was also not considered.


n


n

965

E Manias and A Williams

Study Selection

We independently screened all abstracts identified in the search to create a group of potentially relevant studies. The full articles of all potentially relevant research were then obtained and independently examined to determine whether they met the inclusion criteria. References cited in potentially relevant work and literature reviews were explored for additional studies. If there was any uncertainty or disagreement about whether certain studies met the inclusion criteria, we arrived at a negotiated agreement. There was uncertainty or disagreement on 4 studies, which was easily resolved by negotiation. Data Extraction

A structured data extraction form was developed for documenting information from each selected study. This information included the population studied, setting, sample size, research design, type of intervention used, methods used to assess adherence, main results, and significance of intervention. We independently extracted the data and resolved any discrepancies through discussion. Quality Ratings

Methodological quality was independently assessed based on 8 criteria: (1) study design, (2) specification of patient sample, (3) power analysis, (4) specification of disease, (5) specification of therapeutic regimen, (6) duration of follow-up, (7) definition of adherence, and (8) adherence measurement.11 Raw scores for the 8 criteria were combined into an overall mean score standardized from 0 to 100. When a discrepancy occurred, we discussed this issue to achieve resolution by a consensual process.

95% confidence intervals that did not include 1.0 for risk ratio estimates and did not include 0.0 for standardized mean difference estimates were considered statistically significant. Results IDENTIFICATION AND SELECTION OF STUDIES

The initial literature search identified 994 abstracts from which 72 full-text articles were retrieved for closer examination according to the inclusion criteria. Twenty-six of these papers were excluded, mainly because the studies did not involve the evaluation of an educational or psychosocial intervention or because they did not examine medication adherence. In total, 46 studies met the inclusion criteria for the systematic review (Figure 1). TYPES OF STUDIES AND SAMPLE CHARACTERISTICS

The final sample of 46 articles included 36 randomized controlled trials,12-47 2 observational cohort studies,48,49 and 8 quasi-experimental studies50-57 (Table 1). Sample sizes ranged from 16 to 3456 participants and, apart from 2 studies,36,50 all research was undertaken in the US. The percentage of CALD groups participating as a proportion of the total sample ranged from 26% to 100%. The most common chronic conditions targeted included hypertension (n = 14) and HIV (n = 12). Population groups involving the treatment of acute conditions included parents and caregivers whose children were prescribed liquid medications after attending an emergency department47 and mothers whose children were prescribed antibiotics for treatment of

Meta-Analysis

Review Manager (RevMan, Cochrane Collaboration, version 5.1) was employed to conduct the meta-analyses, which were specified a priori using a random effects model. Studies that provided results for control and intervention groups and those that had either binary or continuous adherence outcomes following the conduct of a particular intervention were included. The Mantel-Haenszel risk ratio summary estimate was used for studies that presented adherence outcomes as binary data, and the standardized mean difference estimate was used for studies that presented adherence outcomes as continuous data. Separate metaanalyses were conducted on studies whose total sample population comprised individuals from a CALD background. Meta-analyses were carried out employing an intention-to-treat principle. Tests for heterogeneity were calculated, using χ2, τ2, and I2 tests. Forest plots were created to illustrate the results graphically and to identify the weighting of each study. Findings with p values 65 y), medications for chronic illness Black (others: white)

Uncontrolled hypertension Hispanic, Latino

Adults, diabetes African American (others: white, NR)

Women, HIV / AIDS African American, Hispanic, Caribbean, Native American, Pacific Islander (others: white)

HIV / AIDS African American, Latino (others: white, NR)

Prospective cohort

Hirsch (2009)48

83 / 76 34.9% / 40.8%

444 92%

53 100%

21 / 18 46.7% / 56.3%

92 / 82 93% (no breakdown)

1353 / 2103 55.4% / 53.5%

436 / 219 88.1% / 88.1%

Hypertension Hispanic, Latino, Asian or Pacific Islander, black, African American (others: white, NR)

US

RCT

35 / 30 100% / 100%

8/9 87.5% / 88.9%

140 / 139 63.1% (no breakdown)

110 / 111 100% / 100%

8/8 62.5% / 87.5%

56 / 63 94.6% / 96.8%

20 / 17 85% / 93%

Intervention 1: 75, 100% Intervention 2: 60, 100% Control: 54, 100%

106 / 82 24% / 23%

Gerin (2007)19

HIV-positive African American (others: white, NR)

Ethnic minority groups or poor background, hypertension Asian, African American (others: white)

Children, asthma Island or mainland Puerto Rican

Elderly, hypertension Black (others: white)

Children, asthma; caregivers Latino, African American (others: NR)

Children, HIV infection; caregivers Hispanic, African American (others: white)

Newly diagnosed type 2 diabetes Hispanic, Latino

Type 2 diabetes Non–English speaking (others: Australian, other English speaking)

Population

Elderly, hypertension Latino, Hispanic, black, African American

US

US

US

US

US

US

US

Australia

Country

Fernandez Pre-post controlled US intervention (2008)51

RCT, pilot study

Dilorio (2003)18

RCT

Burrelle (1987)15

RCT

RCT

Bonner (2002)14

Cooper (2009)17

RCT

Berrien (2004)13

RCT

RCT

Babamoto (2009)12

Canino (2008)16

Parallel group, multisite intervention

Design

Armour (2004)50

Reference

Intervention / Control (n) CALD Background (%)

45 / 55

36 48 22

45 / 51

Sex (% male) Intervention / Control

77 / 78

57.7 / NA

74.2 / NA

54.6 / NA

54.2 / 56.2

37 (no breakdown)

46.0 / 46.7

54.2 / 53.4

72 / 73

46.8 / 38.4

61.3 (no breakdown)

NR

68 / 70

74.7 / 73.7

37 / NA

34 / NA

43 / NA

33 / 25

NA

76 / 81

22 / 21

54 / 36

62.5 / 44.4

34 (no breakdown)

Caregivers: NR Children: 69 / 62

12.5 / 37.5

Caregivers: 37.6 / 35.7 11 / 11 Children: 10 / 9.1 57 / 44

Caregivers: NR Children: 9.9 / 11.2

51 50 51

64 / 65

Mean Age (y) Intervention / Control

Table 1. Characteristics of Studies on Interventions to Improve Adherence in People of CALD Backgrounds

6 mo (phase 1) + 6 mo (phase 2) / 4, 6, 8, 10, 12, 14 mo

NR / NR

NR / 6 mo

1 mo / 1, 3, 6, 9 mo

Every 3 mo for 9 mo / 3, 6, 9 mo

10 wk / 15 mo

NR / 24 mo

12 mo / 3, 12 mo

6 weekly sessions, then 2 monthly sessions (18 wk) / 6, 14 wk

8 wk / 8 wk

30 min at baseline, 10–15 min at 2 wk, 3, 6, 9, 12 mo / 3, 12 mo

18 days / 4 mo

8 wk / 8, 16 wk

3 mo / 3 mo

3 mo / 3, 6–11 mo

6 mo / 6 mo

9 mo / 9 mo

Intervention / Follow-Upa


theannals.com

theannals.com


n


RCT

RCT

Saunders (1991)36

Schaffer (2004)37

US

56 / 59 100%

74 / 77 66% / 74%


109 47.7%

Asthma Intervention 1: 10 African American, Hispanic, Asian, Pacific Islander Intervention 2: 12 Intervention 3: 11 (others: white) Control: 13, 28% (no breakdown)

Hypertension Urban black South African

HIV infection, alcohol problems Hispanic, black (others: white)

US

South Africa

HIV-positive African American, Hispanic, Latino (others: white, NR)

Women, osteoporosis Hispanic, black (others: white)

142 / 140 48% / 41%

22 100%

Elderly (≥65 y), hypertension or hyperlipidemia African American CHF Black, Asian (others: white)

Intervention 1: 25, 96% Intervention 2: 25, 80% Intervention 3: 25, 100% Control: 25, 92%

96 / 99 93% / 94%

16 / 17 26% / 35%

132 / 140 46% / 34%

124 / 124 70% / 71%

95 / 95 100%


1367 98% (no breakdown)

Intervention 1: 199, 96.9% Intervention 2: 204, 96.6% Intervention 3: 196, 93.9% Control: 195, 93.9%


Hypertension Black (others: white)

HIV-positive, on HAART African American, Hispanic, Asian (others: white, NR)

HIV-positive, on HAART Black, Hispanic (others: white)

Type 2 diabetes Black, Hispanic (others: white, NR)

Type 2 diabetes Hispanic (others: white, NR)

Hypertension African American

Adults, HIV-positive or AIDS Black, African American, Latino, Hispanic, Asian, Pacific Islander (others: white)

Hypertension Hispanic, black (others: NR)

Adolescents on anti-tuberculosis medication Asian, African American (others: white, NR)

Hypertension African American (others: white)

US

US

US

US

US

US

US

US

US

US

US

US

US

US

37 (no breakdown)

40–59 (~65%)

42.5 / 43.2

40.8 (no breakdown)

73.2 / NA

80.1 / 78.4

76.4 / NA

50 51.3 49 51.4

35.7 / 37.7

38 / 38

60 / 61

56 / 53

53.7 (no breakdown)

39 (no breakdown)

53.5 (no breakdown)

15.2 15.4 15.2 15.4

54 (no breakdown)

33 (no breakdown)

30 / 34

78 / 84

88 (no breakdown)

NA

32 / 41

36 / NA

24 32 28 16

61 / 69

75 / 94

95 / 99

39 / 44

15 (no breakdown)

88 (no breakdown)

41 (no breakdown)

51 50 53 52

30 (no breakdown)

30–60 min / 3 mo, 6 mo

6 mo / 6 mo

3 mo / 1, 2, 3, 6, 12, 13 mo

1 session / 2, 12 wk

12 mo / 3, 6, 9, 12 mo

90 days / 90 days

12 wk / 14 wk

12 wk / 12 wk

4 wk / 5, 8, 12, 16, 24 wk

12 wk / 4, 16, 28 wk

12 mo / 12 mo

12 mo / 12 mo

12 mo / 12 mo

7 wk / 7 wk, 5 mo

10 min (for 1 intervention, others NR) / 6, 12 mo

6 mo / 6, 12 mo

18 mo / 18–24, 54–60 mo

CALD = culturally and linguistically diverse; CHF = congestive heart failure; COPD = chronic obstructive pulmonary disease; HAART = highly active antiretroviral therapy; NA = not applicable; NR = not reported; RCT = randomized controlled trial. a Follow-up periods relate to time from the point of assessment of baseline measures. (continued on page 970)

RCT

Samet (2005)35

RCT

Safren (2001)34

Open-label clinical trial

Rawlings (2003)31

Pre-post intervention

RCT

Rathbun (2005)30

Robbins (2004)55

RCT

Piette (2001)29

RCT

RCT

Piette (2000)28

Rich (1995)33

RCT

Ogedegbe (2007)27

Pre-post intervention, single group

RCT

Murphy (2002)26

Resnick (2009)54

RCT with 4-group design

Morisky (2002)25

RCT

RCT with 4-group design

Morisky (2001)24

Rehder (1980)32

Sequential randomized factorial trial

Morisky (1985)23

Medication Adherence in People of Culturally and Linguistically Diverse Backgrounds

n

969

970

n


n

RCT

RCT

RCT

Pre-post intervention

RCT

RCT

Vivian (2002)43

Walker (2000)44

Webb (1980)45


Westberg (2005)57

Wyatt (2004)46

Yin (2008)47

Caregivers, children on liquid medications Asian, Native Hawaiian, Pacific Islander, Latino black, Native American, American Indian, Alaskan Native (others: white, NR)

Women, HIV-positive Latina, African American (others: European American)

Pharmacy customers Vietnamese, Hmong, Laotian, Somali, Spanish, Cambodian, Native American, African American, African (others: European American)

Hypertension, low income, rural background Black

Hypertension African American

Hypertension African American (others: white, NR)

Schizophrenia African American, Hispanic (others: white, NR)

HIV infection Latino

Low income, schizophrenia Mexican, Guatemalan, Salvadoran

Hypertension or COPD Black, Asian, Hispanic, Latino (others: white, NR)

HIV or AIDS Black (others: white)

Mother-child couplets, acute otitis media Hispanic

Population

124 / 121 97.6% / 95.9%

67 / 80 93.9%

91 71%

Intervention 1: 37 Intervention 2: 31 Control: 55, 100%

43 / 40 100%

27 / 29 85.1% / 69%

Intervention 1: 32, 61.3% Intervention 2: 34, 70.6% Control: 29, 55.2%

41 / 40 100%


Hypertension 63 / 70, 38.1% / 34.3% COPD 43 / 55, 9.3% / 16.4%

22 / 21 68.1% / 80.9%

31 / 31 100%

100 / 100

43 65 62

88 / 93

6 mo / 6 mo

9 mo / 3, 6, 9, 12, 15 mo

5 wk / 6 mo

NA

47 / NA

19 23 22

Caregivers: 31.1 / 29.6 13 / 7 52 / 64 Children: 3.7 / 3.4

39 (no breakdown)

26–55 (n = 45) 56–65 (n = 37) 66–85 (n = 21)

55.5 55.7 55.5

1.5–3 min / 3–5 days

11 wk / 11 wk, 3, 6 mo

NR / 10 mo

3 mo 9 wk / 1, 2, 3 mo

48–98 (no breakdown) 13 (no breakdown) 6 wk / 3 mo

64 / 65.5

40 37 39

41.9 / 39.6

6 mo / 6 mo

3 mo / 3 mo

10 min / 10 days

Intervention / Follow-Upa

64 (no breakdown) 12 mo / 12 mo

100 / 100

69.3 / 69.3 30 (no breakdown)

98 / 93

77 / 81

Mothers: NA Children: NR

Sex (% male) Intervention / Control

66.3 / 67.3

37 (no breakdown)

Mothers: 24 / 25.6 Children: 2.5 / 2.7

Mean Age (y) Intervention / Control

CALD = culturally and linguistically diverse; CHF = congestive heart failure; COPD = chronic obstructive pulmonary disease; HAART = highly active antiretroviral therapy; NA = not applicable; NR = not reported; RCT = randomized controlled trial. a Follow-up periods relate to time from the point of assessment of baseline measures.

US

US

US

US

US

US

US

US

RCT

RCT

Telles (1995)41

US

Velligan (2008)42

RCT

Solomon (1998)40

US

US

RCT

Smith (2003)39

US

Country

van Servellen (2003)56 Quasi-experimental, repeated measures

Randomized, post-test only, controlled group design

Design

Schwartz-Lookinland (1989)38

Reference

Intervention / Control (n) CALD Background (%)

Table 1. Characteristics of Studies on Interventions to Improve Adherence in People of CALD Backgrounds (continued)


theannals.com


come for adherence using the standardized mean difference approach (Figure 3). It was not possible to undertake meta-analysis on 6 studies for which no control group was present,49,52-55,57 on 4 studies for which no clear adherence results were provided for control and intervention groups following conduct of the intervention,20,25,41,45 and on 3 studies that related to ongoing research.17,19,27 For the meta-analysis involving studies with dichotomous outcomes, the random effects model using the Mantel-Haenszel summary estimate was statistically significant with a risk ratio of 0.75 for adherence (95% CI 0.67 to 0.85). The χ2 test for heterogeneity was 129.49, which was significant at p < 0.001, while the τ2 and I2 results were 0.05% and 85%, respectively (Figure 2). For the meta-analysis involving studies with continuous adherence outcomes, the random effects model using the standardized mean difference was statistically significant, with a value of –0.71 for adherence (95% CI –1.17 to –0.26). The χ2 test for heterogeneity was 80.61, which was significant at p < 0.001, while the τ2 and I2 results were 0.52% and 86%, respectively (Figure 3). Two separate meta-analyses were undertaken on a total of 7 studies where 100% of the participants came from a CALD background: 1 meta-analysis involved studies using a dichotomous adherence measure (n = 5), while the other involved studies using a continuous adherence measure (n = 2). For the meta-analysis relating to studies involving a dichotomous adherence outcome, the random effects model was not statistically significant, with a risk ratio of 0.81 for adherence (95% CI 0.63 to 1.04). For the meta-analysis involving studies that employed a continuous adherence measure, the random effects model was not significant, with a standardized mean difference of – 0.22 for adherence (95% CI –1.01 to 0.57). MODELS GUIDING INTERVENTIONAL WORK

No mention was made about a model to guide the research process in 34 studies (Table 2). The Health Belief Model and Motivational Theory were the most common models employed and, in some studies, combinations of models were used. Four of the 20 studies demonstrating significant improvements in medication adherence involved the use of a model.13,14,37,39 QUALITY OF STUDIES

Table 3 shows an overview of the quality of the interventional studies reviewed. Of the 46 studies, there was no power analysis reported in 38. Loss of participant follow-up varied extensively, ranging from 1% to 51%. Reasons for loss to follow-up were mentioned in 29 studies. Lack of blinding of the outcome assessor was an issue for 10 studies and, in 11 studies, no mention was made of whether blinding occurred. Mean quality percent scores for all studies were theannals.com

completed using the method described by Nichol et al.11 Of 46 studies, 23 had a mean percent quality score >60%. Discussion Despite the critical issue of medication adherence in people of CALD backgrounds, relatively little high-quality work has been conducted on adherence-enhancing interventions. A range of interventions were identified in the review with modest efficacy in improving adherence. Many studies were of variable methodological quality due to small sample sizes, lack of power analysis, lack of a blinded outcome assessor, and insufficient description or choice of adherence measures.11 Despite the espoused benefits of using models based on cognitive therapy, behavioral modification, and motivational interviewing in developing tailored interventions for medication adherence,60,61 the majority of studies lacked a strong theoretical basis. Of the studies that involved use of a model, much of the emphasis was based on explicating current literature relating to people of CALD backgrounds or in developing culturally specific interventions. Conversely, models were generally not considered in interpreting results of interventions or in refining attributes of interventions to make them more specific to the needs of individuals from CALD backgrounds. Considering the lack of work in this area, more tailored interventions informed by validated models are warranted. Many studies lacked adequate clarification of the nature and content of interventions provided. In some studies, there were no details about the duration of interventions or about the frequency of contact with participants. This situation makes it very difficult for future investigators to reproduce, build upon, and refine interventions used in past research. Interventions tended to focus on increasing participants’ knowledge about medications as the means by which to improve medication adherence. However, past work has consistently shown that education alone does not lead to shifts in medication adherence.61,62 Our review has also demonstrated that using just bilingual people and translated resources in interventions is not sufficient for enhancing medication adherence. Past research has demonstrated links between communication difficulties, patient decision making, and medication adherence.60 People of ethnic minority backgrounds have been shown to be less verbally expressive, less affective, and less assertive in their decision making and interactions with physicians, compared with other patients.62 Communication styles are influenced by culture, and for some cultures, people’s communication styles can make it challenging for them to pose questions to health professionals, which in turn can lead to a lack of effective information exchange and understanding that facilitates poor medication adherence.7 In considering the US-based interventions aimed at improving


n


n

971

972

n


n


Attendance at Treatment Information on Medications for the Elderly program over 8 wk Model NR

8 asthma education modules delivered during 2 home visits and telephone contacts over ~18 days Model NR

20-min coaching before physician visit, 10-min debriefing session, 10- to 15-min telephone calls at 2 wk and 3, 6, 9, 12 mo post baseline; communication skills CD-ROM program for physicians Model NR

3 motivational interviewing sessions, 2 wk apart; motivational Self-report of missed medications in past 2 wk and 30 days, Antiretroviral General and educational materials Adherence Scale (possible score: 5–30) Model NR

6 weekly and 2 monthly group sessions Model NR

Burrelle (1986)15

Canino (2008)16

Cooper (2009)17

Dilorio (2003)18

Fernandez (2008)51

Monthly 5-min telephone calls for 12 mo, self-monitoring of BP; all study materials available in Spanish and English; case managers and recruiters were English- and Spanishspeaking Model NR

3 group education workshops conducted at 1-mo intervals over 3 mo, delivered by family coordinator fluent in Spanish and English Asthma Self-Regulation Model

Bonner (2002)14

Gerin (2007)19

8 structured home visits over 3 mo, education on knowledge and understanding of disease and medications; Spanishspeaking case manager available to intervention group Health Belief Model

Berrien (2004)13

MEMS of primary medication only

Self-report (Morisky Scale); adherent if responded “no” to all items

Appointment keeping, pharmacy records, self-report (Hill-Bone Adherence Scale), provider report

Self-report of use of rescue and maintenance drugs; drugs brought to clinic

Pill counts

Self-report: 4-item family adherence scale developed by authors; α internal consistency = 0.67

Self-report: 37 items developed by authors (score 0–37); pharmacy refill records (score 0–3)

Self-report: Morisky Self-Reported Medication Behavior Scale

INT 1: community health worker, 10-wk diabetes education program by bilingual workers and follow-up phone calls over 6 mo INT 2: case management, monthly care by nurses Transtheoretical model

Babamoto (2009)12

Risk of nonadherence, 5-item subscale of Brief Medication Questionnaire (range 0–7, 7 = no adherence)

Adherence Measurement

9-mo program delivered by community pharmacists Model NR

Intervention and Theoretical Model

Armour (2004)50

Reference

Self-report baseline: 32.2 3 mo: 34.8 Refill records baseline: NR 3 mo (mean): 2.7

Never forget to take diabetes medications, % baseline INT 1: 69 INT 2: 77 6 mo INT 1: 79 INT 2: 55

Risk of nonadherence baseline: 3.89 9 mo: 2.74

Intervention Group

Systolic BP

Systolic and diastolic BP change

Adherence

Adherence

Symptom-free days

Adherence

NA

Baseline: 24% 6 wk: 56% 14 wk: 52%

Baseline: NR AGAS mean, 8 wk: 26.5 Missed medications in past 2–8 wk: 0.06 Ability to take medication in past 30 days–8 wk: 87.5%

NA

Baseline: 67.3 4 mo: 64.8

Pill counts % baseline: NR 8 wk: 92

Knowledge Prescribed frequency % about asthma baseline: NR 3 mo: 82

Adherence

HbA1c, BMI

HbA1c

Primary Outcome

Table 2. Interventions and Adherence Measurementsa

NA

NA

Yes Yes No

Yes 33.3% 59% 36% 58%

NR No No

NA

No No

NR Yes

NR Yes

NR Yes

NR 23.44 0.70

NA

60.4 64.6

NR 71

NR 40

NR 1.7

No No

Yes Yes

50 50 31.7 31.9

No No

No Yes

Significant Differences (p < 0.05)

67 67

2.81 3.90

Control Group


theannals.com

theannals.com

10 weekly 2-h sessions of cognitive-behavioral stress management Model NR

Counseling by certified diabetes educator every 3 mo for 9 mo Model NR

Jones (2003)20

Krier (1999)21


Prospective observational phase: 6 mo of individualized education, drug dispensed in blister packs, pharmacist follow-up every 2 mo

Lee (2006)22

n

Combination of 3 INTs: 1. 10-min exit interview 2. family member support through a booklet 3. small training group Model NR

4-item self-report measure

Pill counts

Pharmacy records; adherence based on number of pts. on 3 drugs (aspirin, ACE inhibitor, statin)

Self-reported use of medications in pharmacy records

Medication refills

Self-report on adherence using 4-point Likert scale (1 = very adherent to 4 = not adherent)

Self-report (14-item Adherence to Medication Scale (adapted from ACTG Questionnaire for Adherence to AntiHIV Medication)

Pill counts

Adherence

Adherence

BP control



Adherence to follow-up visits HbA1c levels

Adherence

Adherence

Baseline: NR Family member support at 18– 24 mo: 52%

Baseline (phase 2): 61.4% 14 mo (end of phase 2): 95.5%

Baseline (phase 1, run-in phase): 61.2% 8 mo (end of phase 1): 96.9%

Baseline: 127/446 (28.5%) Follow-up (date NR): 281/446 (63.0%)

Baseline: NR 6 mo: 114 adherence problems identified in 258 pts.

Baseline: 70.6% 1 mo: 71.2% 3 mo: 82.7% 6 mo: 88.5% 9 mo: 95.8%

Baseline: 1.7 3 mo: 1.8 6 mo: 1.8 9 mo: 1.5

Baseline: 68% mean adherence combined result 10 wk: low-adherent pts. (≤80% adherence): 30.4% increase in adherence

Baseline: not conducted 24 mo: 56.3%

NR 36%

61.1% 69.1%

NA

NA

NA NA

NA

2.1 2.3 2.0 1.6

19.6% increase in adherence

not conducted 38.1%

NR Yes

Yes (withingroup for phase 1 intervention) No Yes

NA

NA Yes (withingroup comparison)

NA NR (within-group comparison)

NA No Yes Yes Yes (comparisons with baseline of intervention group)

No No No No

Yes: within-group for intervention No: within-group for control No comparison between groups

NA

NA Yes

ACE = angiotensin-converting enzyme; ACTG = AIDS Clinical Trials Group; AGAS = Antiretroviral General Adherence Scale; BMI = body mass index; BP = blood pressure; COPD = chronic obstructive pulmonary disease; HbA1c = hemoglobin A1c; INT = intervention; MEMS = Medication Event Monitoring System; NA = not applicable; NR = not reported; RCT = randomized controlled trial. a Unless otherwise specified, adherence is defined as a percentage of the doses taken at the identified period. To determine significant differences in adherence between groups, intervention and control group results were compared. If only 1 group was present, adherence was compared between baseline and follow-up periods. (continued on page 974)

Morisky (1985)23

Pharmacist-based disease state management service with a bilingual, certified diabetes educator Model NR

Leal (2008)53

RCT phase: same intervention for additional 6 mo Model NR

Medication consultations from pharmacists Model NR

Lam (2008)49

Lai (2007)52 9-mo community pharmacy-based hypertension disease management program; all oral and written communication in Spanish Model NR

Provision of medication therapy management by community pharmacies to pts. with ≥50% prescriptions filled at pharmacy over 2 y Model NR

Hirsch (2009)48



n

973

974

n


INT 1: peer counseling INT 2: parent-participant contingency contract INT 2: combined peer counseling and contingency contract Model NR All INTs 6 mo

INT 1: individualized pt. counseling sessions INT 2: appointment tracking INT 3: home visit All interviews translated into Spanish Health Belief Model, Theory of Planned Action, and Social Support Theory

Alternating series of 5 group and individual sessions over 7 wk using behavioral change strategies, social support, education Model NR

4 motivational interviewing sessions every 3 mo for 12 mo Model NR

Automated telephone calls (5–8 min) and nurse telephone education over 12 mo Spanish-language versions of calls, bilingual interpreter translated messages, intervention nurse competent in Spanish Model NR

Biweekly automated telephone disease management and education calls and nurse follow-up calls based on assessment; Spanish-language version available Model NR

1- to 1.5-h visit at start of therapy and 30-min phone follow-up up to 12 wk Model NR

Interactive, educational program of 4 sessions conducted over 4 wk Model NR

INT 1: drug/disease counseling 3 times over 12 wk INT 2: drug container only (with 7-day supply) INT 3: drug/disease counseling and drug container Model NR All INTs 12 wk

60-min sessions 3 times/wk for 12 wk Social Ecological Model

Reference

Morisky (2001)24

Morisky (2002)25

Murphy (2002)26

Ogedegbe (2007)27


n

Piette (2000)28

Piette (2001)29

Rathbun (2005)30

Rawlings (2003)31

Rehder (1980)32

Resnick (2009)54

Adherence

Adherence


Self-efficacy and mastery

2010 June, Volume 44 Self-care, use of specialty services, severity of diabetes

Self-report on 5-point scale from 0 to 4 (4 = adherence most of the time, 0 = no adherence)

Unit dose count of drugs remaining

MEMS caps

NR 77.8% 77.8% 77.8%

NA No Yes Yes NA No (withingroup comparison)

Not conducted 48% 48% 48% NA

≥95% for % of pts. by unit dose count baseline: not conducted INT 1: 60% INT 2: 88% INT 3: 90% Baseline: 3.2 (mean) 14 wk: 3.9 (mean) Decrease in systolic and diastolic BP

NA No

NA No No No

No No

No Yes

Not available

No No No

Adherence

Not conducted 73% 56% 51%

46% 39%

69/124 (56%) 78/124 (63%)

Not available

62% 87% 83%

No differences No differences

NR No No No


Not conducted 74%

Electronic monitoring baseline: not conducted week 4: 86% week 16: 77% week 28: 74%

Medication problems baseline: 56% 12 mo: 45%

Baseline: 69/124 (56%) 12 mo: 55/124 (44%)

Not available

Baseline (mean): 69% 7 wk (mean): 87% 3 mo (mean): 86%

Baseline: NR NR No breakdown for 6 or 12 mo for No breakdown any INT for 6 or 12 mo for control group

Baseline: NR INT 1: 80.3% INT 2: 76.4% INT 3: 84.8%

Intervention Group

Control Group

Baseline: not conducted 24 wk: 70%

HIV-1 RNA levels

Electronic monitoring of 1 antiretroviral Adherence drug (doses consumed/number of doses of prescribed doses); validated self-report

Self-report: any problems with medications reported

Self-report using variation of 3-item Morisky Mean HbA1c scale, any medication adherence levels problems reported

Pill counts, Medication Adherence Selfefficacy scale, Morisky scale

Self-report using Adult ACTG Adherence Baseline Questionnaire

Cronbach’s α = 0.68

Self-report using variation of 6-item Morisky scale

Self-report with variation of 3-item Morisky scale Cronbach’s α = 0.59 Adherence based on those completing 6 mo of TB treatment


Primary Outcome

Table 2. Interventions and Adherence Measurementsa (continued)


theannals.com

theannals.com


n

Volume measure of antibiotic regimen; adherence using 80% missing drugs to determine proportion of adherent pts.

Self-report (pts. asked how many doses of preventive drugs missed in last 2 wk); pharmacy records

Pill counts at clinic visits

MEMS caps (30-day adherence: distributed at baseline and 12 mo; only weak association with self-report data; most data not usable); self-report (3- and 30-day adherence)

Self-report Adherence Questionnaire (pts. document number of pills prescribed and taken each day)

Pill counts; MEMS for minority groups; study; Hispanic breakdown provided for results

Pill counts 30 days after discharge (≥80% of drugs taken correctly)

Adherence

Adherence

Adherence

Adherence

Adherence

Adherence

Survival at 90 days after hospital discharge without readmission

No No No

69.0% 62.0% 64.0%

Baseline: NA 10 days: 67.6%

Baseline INT 1: 32% INT 2: 62% INT 3: 41% 3 mo INT 1: 40% INT 2: 73% INT 3: 53% 6 mo INT 1: 48% INT 2: 77% INT 3: 77%

NA 64.5%

NA No

No Yes Yes

No Yes No 42% 42% 42% 40% 40% 40%

No No No

62% 62% 62%

Yes

NA No

No No No

No No No

NA No No No No (withingroup comparison)

NA Yes

65.0% 63.0% 62.0%

84% 90% 93%

NA

NA 64.9% pts.

Baseline: not conducted Not conducted Newly diagnosed 6 mo: 8/26 3/20 (15%) (31%) took ≥80% of drugs Infrequent attenders 6 mo: 25/37 13/35 (37%) (68%) took ≥80% of drugs

3-day adherence baseline: 58.0% 3 mo: 65.0% 13 mo: 71.0% 30-day adherence baseline: 68.0% 3 mo: 63.0% 13 mo: 67.0%

Baseline: 74% Week 2: 95% Week 12: 94%

Pill counts baseline: 83.0% 3 mo: 83.7% 6 mo: 78.6% 9 mo: 78.5% 12 mo: 85.8%

Baseline: NA 30 days: 82.5% of pts.

ACTG = AIDS Clinical Trials Group; BP = blood pressure; HbA1c = hemoglobin A1c; INT = intervention; MEMS = Medication Event Monitoring System; NA = not applicable; NHLBI = National Heart, Lung and Blood Institute; NR = not reported; PRECEDE = Predisposing, Reinforcing, and Enabling Causes on Educational Diagnoses and Evaluation; TB = tuberculosis. a Unless otherwise specified, adherence is defined as a percentage of the doses taken at the identified period. To determine significant differences in adherence between groups, intervention and control group results were compared. If only 1 group was present, adherence was compared between baseline and follow-up periods. (continued on page 976)

Schwartz- One 10-min motivational interview with 2 handouts, written Lookinland in Spanish, and pictorial instructions (1989)38 Motivational Theory

INT 1: 30-min audiotape INT 2: NHLBI booklet INT 3: audiotape and booklet Protection Motivational Theory

Motivational interviewing over 3 mo, initial 60-min appointment and 30- to 45-min home visit; and 2 subsequent 15to 30-min appointments at 1 and 3 mo; medication timer device given; all data collection in English or Spanish Principles of Motivational Enhancement and Health Belief Model

Samet (2005)35

Schaffer (2004)37

Single session involving cognitive-behavioral, problem solving, and motivational interviewing, 1 follow-up telephone call 1 wk later Model NR

Safren (2001)34

Written reminders to pts., pts. retained records used as a focal point for counseling at clinic visits over 6 mo Model NR

Standardized teaching over 12 mo, pictorial and verbal instruction, available in Spanish and English PRECEDE model

Robbins (2004)55

Saunders (1991)36

Nurse-directed, intensive education during hospitalization, dietary assessment by dietitian, consultation with social service, home follow-up visits, telephone contact over 90 days Model NR

Rich (1995)33



n

975

976

n


n


Webb (1980)45

INT 1: three 1-h group sessions over 3 mo INT 2: 3 individual, 1-h counseling sessions for 9 wk Model NR

Weekly programmed telephone messages on hypertension and spirituality over 6 wk Model NR

Walker (2000)44

Bringing drugs to each appointment (maximum of 18)

Home visits to conduct pill counts

Self-report adherence survey; pharmacy records of refills (refilling drugs within 2 wk after scheduled refill date)

Visit to clinical pharmacist–managed hypertension clinic monthly for 6 mo; pharmacist prescribing changes to hypertensive therapy and medication education given Model NR

Vivian (2002)43

Adherence

Baseline: NR 18 mo INT 1: 4.6 INT 2: 5.0

BP knowledge Pills taken/pills prescribed baseline: 0.83 6 wk: 0.88

Changes in BP Baseline: NR Refills within 2 wk–6 mo: 85% Self-report asking if forgot to take medications ≥1 time/wk– 6 mo: 68%

Baseline: NR 9 mo INT 1: 82% INT 2: 84% 12 mo INT 1: 83% INT 2: 85% 15 mo INT 1: 80% INT 2: 82%

Adherence

Unannounced pill counts, pharmacy records

4.8 4.8

0.80 1.10

No No

NR

No No

No No No

Yes 58%

NR 93% 48%

Yes

No

0.32

55%

No

2.16

NR Yes

No

0.29

NR 56%

No

1.82

NR

12 mo: NR Adherence affected by level of acculturation: F = 12.12, df = 1.36, p = 0.001

No differences reported No differences reported

Yes No

23% NR NR

No No

NA Yes


60% NR

Not conducted 37%

Control Group

Baseline: NR

INT 1: Medication adherence targeted Cognitive Adaptation Training (Pharm-CAT) INT 2: Full CAT Each intervention 30–45 min/wk over 9 mo Model NR

Preventing relapse and reducing psychotic exacerbation

Baseline BP group: 61% nonadherence COPD group: NR 6 mo BP group: 63% nonadherence COPD group: NR

Velligan (2008)42

Self-report using a 5-point rating of medication adherence

Systolic or diastolic BP or dyspnea on exertion

Baseline: not conducted 12 wk: 96%

Level of health Doses missed in last 4 days literacy baseline: 2.38 Doses missed in last 24 h baseline: 0.56 Doses missed in last 4 days–6 wk: 1.26 Doses missed in last 24 h–6 wk: 0.29

Behavioral family intervention of weekly sessions for 6 mo, then every 2 wk for 3 mo, then monthly for 3 mo; bicultural clinicians delivered intervention; instructional materials available in Spanish Model NR

Telles (1995)41

Self-report; 4-item Morisky Scale; pill counts during clinic visits

Adherence

Self-report using adherence questions from the Adults ACTG Adherence Baseline Questionnaire

Pharmaceutical care program with scheduled visits at baseline and 1-mo intervals for 5 visits up to 6 mo Model NR

Solomon (1998)40

MEMS, Adherence Confidence Scale

Intervention Group

5-wk instructional support program with 6-mo follow-up with case management; conducted in Spanish by bilingual practitioners Model NR

Individual drug self-management and 3 once-monthly follow-up sessions Bandura Self-Efficacy Model and Social Cognitive Theory

Smith (2003)39


van Servellen (2003)56


Reference

Primary Outcome

Table 2. Interventions and Adherence Measurementsa (continued)


theannals.com

ACTG = AIDS Clinical Trials Group; BP = blood pressure; COPD = chronic obstructive pulmonary disease; INT = intervention; MEMS = Medication Event Monitoring System; NA = not applicable; NR = not reported. Unless otherwise specified, adherence is defined as a percentage of the doses taken at the identified period. To determine significant differences in adherence between groups, intervention and control group results were compared. If only 1 group was present, adherence was compared between baseline and follow-up periods.

1.5- to 3-min program with instructional sheets and pictograms; instructions in English or Spanish Model NR Yin (2008)47

theannals.com

a

NA Yes NA 62.0% Baseline: NA 3–5 days: 90.7% Reduction of dosing errors

11 weekly sessions guided by cognitive behavioral approaches, psycho-education, Spanish-speaking data collectors for both groups Treatment Engagement Model

Self-report: % of pts. who took within 20% of prescribed doses

NR No

NA NR (within-group comparison) NA

NR 73.3% Baseline: NR 11 wk: 75.6% Adherence

NA NA

Wyatt (2004)46

Self-report: pts. asked how many days in past 2 wk drugs were taken on schedule and at right dose

NA

Pharmaceutical care for medication management, anticoagulation, and health education; full-time interpreters in participating clinics Model NR Westberg (2005)57

Pharmacy records; % of non–Englishspeaking background people with medication adherence problems; outcomes met for medication problems

Language services provided by pharmacies

Medication problems for adherence baseline: 31% Outcomes met baseline: 58.0% 10 mo: 80.0%

NA


medication adherence, there was little direct acknowledgment of the concepts underlying the national standards on CLAS.8 Furthermore, since many of the CLAS standards are mandated for accredited agencies and organizations receiving US federal funds, it would be anticipated that the principles and activities underlying the standards should comprise “usual care” given to control groups. Yet, few details were provided about the nature and content of usual care initiatives given to people of CALD backgrounds. In the majority of studies, there was homogenization of adherence results for people of CALD and non-CALD backgrounds. This homogenization can lead to possible dilution of intervention effectiveness and lack of clarity about how interventions could be refined and improved in future work. Demographic characteristics were presented of various cultural groups, but aside from a few exceptions,24,55,57 there was a lack of subgroup analysis of adherence in participants of different CALD backgrounds. Similarly, only a few studies involved recruitment of participants solely from CALD backgrounds.12,16,36,38,41,44,45,51,52,54,56 Another concern was the highly varied terminology used to describe different population groups. For instance, the term “Latino” can include a person who has Mexican, Central American, South American, Puerto Rican, or Cuban ancestry. Individuals of a Hispanic origin are also interpreted in diverse ways, as they have been perceived to be individuals who come from Spain or Portugal or construed to be the same as individuals with a Latino background.63 Such delineations have limitations as they comprise people from highly diverse cultures. In addition, while the majority of studies tended to focus on people of Hispanic, Latino, or African American backgrounds, little interventional work was carried out on indigenous groups, such as Native Americans; people from European, Asian or Middle Eastern, and Pacific countries; and people at either end of the age spectrum. Future interventional work should not only target people from a specific CALD group, but also provide tailored strategies that take into account various differences within a particular CALD group, such as attitudes toward disease and treatment, educational attainment, religion, family support networks, and socioeconomic background. The process of developing interventions from patient perspectives through preliminary interviews and surveys can assist in identifying strategies that would normally not have been conceptualized if they were developed from health professionals’ perspectives.64 In most studies, measurement of medication adherence relied heavily on the use of self-reports. Reports from patients, caregivers, and physicians on adherence have been shown to underestimate the extent of adherence.65 Conversely, MEMS caps, blood drug concentrations, and pharmacy refill records provide more objective and accurate The Annals of Pharmacotherapy

n


n

977


measures.66 The most commonly used self-report measure was the Morisky Adherence Scale, which comprises 4 items. In the studies reviewed, people were deemed to be adherent if none of the 4 items evoked a “no” response, which suggested no missed doses. With repeated use of this scale for testing at various time points, it is possible that people may have responded in the same way, therefore leading to response bias with subsequent loss in sensitivity. The lack of a gold standard in adherence measurement means that researchers should be encouraged to use a com-

bination of approaches, such as MEMS, pill counts, prescription refills, and validated self-reported measures. Nine studies involved the use of combinations of different medication adherence measures and 41 studies involved the measurement of clinical or health outcome variables, such as blood pressure and number of days of asthma symptoms. To provide meaningful results to health professionals and individuals of CALD backgrounds, it is helpful to include clinical or health outcome measurements in future assessment of adherence.

Figure 2. Meta-analysis of studies using medication adherence as a dichotomous variable (n = 21).

Figure 3. Meta-analysis of studies using medication adherence as a continuous variable (n = 12).

978

n


n


theannals.com


Table 3. Quality Assessment of Studies on Interventions Provided to People of Culturally and Linguistically Diverse Backgrounds

Reference

Randomized Design

Selection Criteria Clearly Marked

Drug Starters/ Chronic Users/Both

Outcome Assessor Loss to Blinded to Follow-Up Intervention (%) Allocation Reported

Reason for Loss to Follow-Up Reported

Power Calculation Provided

Mean % Quality Score

Armour (2004)50

No

Yes

Both

No

21

Yes

Yes

64.1

Barbamoto (2009)12

Yes

Yes

Starters

No

41

Yes

No

67.6

Berrien (2004)13

Yes, at pt. level

No

Both

No

8

Yes

No

67.6

Bonner (2002)14

Yes

No

Both

Yes

16

No

No

54.6

Burrelle (1987)15

Yes

Yes

Both

No

11

Yes

No

41.5

Canino (2008)16

Yes

Yes

Chronic users

No

1

No

No

62.1

Cooper (2009)17

Yes

Yes

Both

Yes

51

Yes

Yes

77.4

Dilorio (2003)18

Yes

Yes

Both

Yes

23

Yes

No

41.5

Fernandez (2008)51

No

Yes

Both

Yes

14

Yes

No

42.1

Gerin (2007)19

Yes

Yes

Both

Yes

38

Yes

Yes

77.8

Hirsch (2009)48

Yes

Yes

Both

Yes

NR

No

No

54.6

Jones (2003)20

Yes

Yes

Chronic users

Yes

NR

No

No

49.0

Krier (1999)21

Yes

Yes

Both

No

21

No

No

33.7

Lai (2007)52

NA

Yes

Both

No

51

Yes

No

48.2

Lam (2008)49

NA

Yes

Both

NA

NR

No

No

57.1

Leal (2008)53

NR

No

Both

No

NR

No

No

52.4

Lee (2006)22

Yes

Yes

Chronic users

Yes

27

Yes

No

68.2

Morisky (1985)23

Yes

Yes

Chronic users

NR

28

Yes

No

52.6

Morisky (2001)24

Yes

No

Starters

NR

20

No

No

44.8

Morisky (2002)25

Yes

No

Both

Yes

47

No

No

62.1

Murphy (2002)26

Yes

Yes

Chronic users

Yes

13

Yes

No

49.0

Ogedegbe (2007)27

Yes

Yes

Chronic users

Yes

15

Yes

Yes

75.9

Piette (2000)28

Yes

Yes

Both

Yes

11

Yes

No

67.6

Piette (2001)29

Yes

Yes

Chronic users

Yes

7

Yes

No

70.4

Rathbun (2005)30

Yes

Yes

Starters

Yes

23

Yes

No

77.2

Rawlings (2003)31

Yes

Yes

Both

Yes

28

Yes

No

85.5

Rehder (1980)32

Yes

Yes

Chronic users

NR

28

Yes

No

67.0

Resnick (2009)54

NA

Yes

Chronic users

No

9

Yes

No

31.0

Rich (1995)33

Yes

Yes

Chronic users

Yes

11

Yes (death only)

No

59.3

Robbins (2004)55

NA

Yes

Both

No

12

Yes

No

43.5

Safren (2001)34

Yes

Yes

Both

NR

5

No

No

35.9

Samet (2005)35

Yes

Yes

Both

Yes

24

No

Yes

90.5

Saunders (1991)36

Yes

Yes

Both

Yes

21

No

Yes

71.8

Schaffer (2004)37

Yes

Yes

Chronic users

Yes

4

Yes

No

71.2

Schwartz-Lookinland (1989)38 Yes

Yes

Starters

NR

6

Yes

No

53.8

Smith (2003)39

Yes

Yes

Both

NR

60

No

No

62.9

Solomon (1998)40

Yes

Yes

Chronic users

NR

NR

No

No

67.1

Telles (1995)41

Yes

Yes

Both

Yes

10

Yes

No

69.6

van Servellen (2003)56

NA

Yes

Both

NR

5

Yes

No

53.0

Velligan (2007)42

Yes

Yes

Chronic users

Yes

12

No

No

78.3

Vivian (2002)43

Yes

Yes

Chronic users

NR

5

Yes

No

72.4

Walker (2000)44

Yes

No

Both

Yes

17

Yes

No

47.8

Webb (1980)45

Yes

Yes

Chronic users

No

18

No

Yes

47.6

Westberg (2005)57

NA

No

Both

NA

NR

NR

No

30.4

Wyatt (2004)46

Yes

Yes

Both

NR

10

Yes

No

50.4

Yin (2008)47

Yes

Yes

Starters

Yes

7

Yes

Yes

69.6

NA = not applicable; NR = not reported.

theannals.com


n


n

979


Meta-analyses results showed statistically significant benefit in improving medication adherence when considering all reviewed studies. However, when subsets of studies involving only people of CALD backgrounds were examined, meta-analyses demonstrated no statistically significant benefits. The main reason for lack of benefit in these studies was the relatively low sample size where, on average, about 48 people participated in the control or intervention group. In addition, results of the tests for heterogeneity were significant, indicating that several factors need to be considered when evaluating adherence to medications. Such factors may include the use of community or hospital settings as study sites, the types of CALD groups involved, the use of culturally designed materials for interventions, and types of interventions tested.67 LIMITATIONS OF THE REVIEW

Only articles published in English were examined for this review. Potentially, there may have been studies published in non-English journals relating to research involving interventions for improving medication adherence in people of CALD backgrounds. In addition, research disseminated through gray literature, such as conference papers and unpublished reports, was not considered. Only 3 studies24,55,57 involving combinations of CALD and nonCALD groups had separate medication adherence results for the distinct groups. Due to the homogenization of medication adherence results for CALD and non-CALD groups, it was not possible to fully determine the impact of interventions. Nevertheless, meta-analyses results on 7 studies in which 100% of the participants were from a CALD background indicated that interventions did not produce a significant impact on medication adherence. Recommendations Generally, the reviewed studies had interventions that sometimes lacked articulation or reproducibility, missing controls, insufficient explanation of usual care initiatives, inadequate focus of minority populations in recruitment and analysis, missing power calculations, and inappropriate or missing statistical analyses. In developing well-designed intervention studies, greater attention needs to be given to examining the needs of specific CALD population groups. A predominant focus on bilingual peer-support people to deliver the intervention or the use of translated materials is insufficient to address the complex needs of CALD people. Instead, future researchers should consider gathering data from patient perspectives that can be used in interventions to take into account the enormous diversity that exists within any particular CALD group, such as values, beliefs, communication style, self-esteem, lifestyle activities, social support systems, language, religion, and socioeconomic status. It is therefore 980

n


n

important to minimize generalizations even within one CALD group. Well-developed interventions that are rigorously tested are more likely to succeed in improving medication adherence. Elizabeth Manias MPharm RN PhD, Professor, Melbourne School of Health Sciences, Faculty of Medicine Dentistry and Health Sciences, The University of Melbourne, Melbourne, Australia Allison Williams RN PhD, Senior Research Fellow, Melbourne School of Health Sciences, Faculty of Medicine Dentistry and Health Sciences, The University of Melbourne Reprints: Professor Manias, Melbourne School of Health Sciences, Faculty of Medicine Dentistry and Health Sciences, The University of Melbourne, Level 5, 234 Queensberry St., Carlton, Victoria 3053, Australia, fax 61 3 9317 4375, [email protected]

Conflict of interest: Authors reported none

References 1. World Health Organization. Adherence to long-term therapies. Evidence for action. Geneva: World Health Organization, 2003. 2. Ho PM, Rumsfeld JS, Masoudi FA, et al. Effect of medication nonadherence on hospitalization and mortality among patients with diabetes mellitus. Arch Intern Med 2006;166:1836- 41. 3. O’Connor PJ. Improving medication adherence: challenges for physicians, payers, and policy makers. Arch Intern Med 2006;166:1802-4. 4. Agency for Healthcare Research and Quality. National healthcare disparities report. Rockville, MD: US Department of Health and Human Services, 2008. 5. Poon I, Lal LS, Ford ME, Braun UK. Racial/ethnic disparities in medication use among veterans with hypertension and dementia: a national cohort study. Ann Pharmacother 2009;43:185-93. DOI 10.1345/aph.1L368 6. Diaz E, Woods SW, Rosenheck RA. Effects of ethnicity on psychotropic medications adherence. Community Mental Health J 2005;41:521-37. DOI 10.1007/s10597-005-6359-x 7. Duru OK, Gerzoff RB, Selby JV, et al. Identifying risk factors for racial disparities in diabetes outcomes: the Translating Research into Action for Diabetes study. MedCare 2009;47:700-6. 8. Office of Minority Health. National standards on culturally and linguistically appropriate services. Rockville, MD: US Department of Health and Human Services, 2007. 9. Haynes RB, Ackloo E, Sahota N, et al. Interventions for enhancing medication adherence. Cochrane Database Syst Rev 2008;2:CD000011. DOI 10.1002/14651858.CD000011.pub3 10. National Library of Medicine. Medical subject headings (MeSH) 2009. http://www.nlm.nih.gov/mesh/2009/mesh_browser/MBrowser.html (accessed 2009 Nov 9). 11. Nichol MB, Venturini F, Sung JCY. Compliance: a critical evaluation of the methodology of the literature on medication compliance. Ann Pharmacother 1999;33:531-40. DOI 10.1345/aph.18233 12. Babamoto KS, Sey KA, Camilleri AJ, at al. Improving diabetes care and health measures among Hispanics using community health workers: results from a randomized controlled trial. Health Educ Behav 2009;36: 113-26. 13. Berrien VM, Salazar JC, Reynolds E, McKay K. Adherence to antiretroviral therapy in HIV-infected pediatric patients improves with homebased intensive nursing intervention. AIDS Patient Care STDs 2004;18: 355-63. DOI 10.1089/1087291041444078 14. Bonner S, Zimmerman BJ, Evans D, Irigoyen M, Resnick D, Mellins RB. An individualized intervention to improve asthma management among urban Latino and African-American families. J Asthma 2002;39: 167-79. DOI 10.1081/JAS-120002198 15. Burrelle TN. Evaluation of an interdisciplinary compliance service for elderly hypertensives. J Geriatr Drug Ther 1986;1:23-49. 16. Canino G, Vila D, Normand S-LT, et al. Reducing asthma health disparities in poor Puerto Rican children: the effectiveness of a culturally tailored family intervention. J Allergy Clin Immunol 2008;121:665-70. DOI 10.1016/j.jaci.2007.10.022


theannals.com

Medication Adherence in People of Culturally and Linguistically Diverse Backgrounds 17. Cooper LA, Roter DL, Bone LR, et al. A randomized controlled trial of interventions to enhance patient-physician partnership, patient adherence and high blood pressure control among ethnic minorities and poor persons: study protocol NCT00123045. Implement Sci 2009;4:7. DOI 10.1186/1748-5908- 4-7 18. Dilorio C, Resnicow K, McDonnell M, et al. Using motivational interviewing to promote adherence to antiretroviral medications: a pilot study. J Assoc Nurses AIDS Care 2003;14:52-62. DOI 10.1177/1055329002250996 19. Gerin W, Tobin JN, Schwartz JE, et al. The medication Adherence and Blood Pressure Control (ABC) trial: a multi-site randomized controlled trial in a hypertensive, multi-cultural, economically disadvantaged population. Contemp Clin Tri 2007;28:459-71. DOI 10.1016j.cct.2007.01.003 20. Jones DL, Ishii M, LaPerriere A, et al. Influencing medication adherence among women with AIDS. AIDS Care 2003;15:463-74. 21. Krier BP, Parker RD, Grayson D, et al. Effect of diabetes education on glucose control. J La State Med Soc 1999;151:86-92. 22. Lee JK, Grace KA, Taylor AJ. Effect of a pharmacy care program on medication adherence and persistence, blood pressure, and low-density lipoprotein cholesterol: a randomized controlled trial. JAMA 2006;296:2563-71. 23. Morisky DE, DeMuth NM, Field-Fass M, et al. Evaluation of family health education to build social support for long-term control of high blood pressure. Health Educ Q 1985;12:35-50. 24. Morisky DE, Malotte CK, Ebin V, et al. Behavioral interventions for the control of tuberculosis among adolescents. Public Health Rep 2001;116: 568-74. 25. Morisky DE, Lees NB, Sharif BA, Liu KY, Ward HJ. Reducing disparities in hypertension control: a community-based hypertension control project (CHIP) for an ethnically diverse population. Health Promotion Pract 2002;3:264-75. DOI 10.1177/152483990200300221 26. Murphy DA, Lu MC, Martin D, et al. Results of a pilot intervention trial to improve antiretroviral adherence among HIV-positive patients. J Assoc Nurses AIDS Care 2002;13:57-69. DOI 10.1177/1055329002238026 27. Ogedegbe G, Schoenthaler A, Richardson T, et al. An RCT of the effect of motivational interviewing on medication adherence in hypertensive African Americans: rationale and design. Contemp Clin Trials 2007;28:169-81. DOI 10.1016/j.cct.2006.04.002 28. Piette JD, Weinberger M, McPhee SJ, Mah CA, Kraemer FB, Crapo LM. Do automated calls with nurse follow-up improve self-care and glycemic control among vulnerable patients with diabetes? Am J Med 2000;108: 207. DOI 10.1016/S0002-9343(99)00298-3 29. Piette JD, Kraemer FB, Weinberger M, et al. Impact of automated calls with nurse follow-up on diabetes treatment outcomes in a Department of Veteran Affairs health care system. Diabetes Care 2001;24:202-8. 30. Rathbun RC, Farmer KC, Stephens JR, et al. Impact of an adherence clinic on behavioral outcomes and virologic response in the treatment of HIV infection: a prospective, randomized, controlled pilot study. Clin Ther 2005;27:199-209. DOI 10.1016/j.clinthera.2005.02.010 31. Rawlings MK, Thompson MA, Farthing CF, et al. Impact of an educational program on efficacy and adherence with a twice-daily lamivudine/zidovudine/abacavir regimen in underrepresented HIV-infected patients. J Acquir Immune Defic Syndr 2003;34:174-83. 32. Rehder TL, McCoy LK, Blackwell B, et al. Improving medication compliance by counseling and special prescription container. Am J Hosp Pharm 1980;37:379-85. 33. Rich MW, Beckham V, Wittenberg C, et al. A multidisciplinary intervention to prevent the readmission of elderly patients with congestive heart failure. N Engl J Med 1995;333:1190-5. 34. Safren SA, Otto MW, Worth JL, et al. Two strategies to increase adherence to HIV antiretroviral medication: Life-Steps and medication monitoring. Behav Res Ther 2001;39:1151-62. 35. Samet JH, Horton NJ, Meli S, et al. A randomized controlled trial to enhance antiretroviral therapy adherence in patients with a history of alcohol problems. Antiviral Ther 2005;10:83-93. 36. Saunders LD, Irwig LM, Gear JSS, et al. A randomized controlled trial of compliance improving strategies in Soweto hypertensives. Med Care 1991;29:669-78. 37. Schaffer SD, Tian L. Promoting adherence: effects of theory-based asthma education. Clin Nurs Res 2004;13:69-89. DOI 10.1177/1054773803259300

theannals.com

38. Schwartz-Lookinland S, McKeever LC, Saputo M. Compliance with antibiotic regimens in Hispanic mothers. Patient Educ Couns 1989;13:171-82. 39. Smith SR, Rublein JC, Marcus C, et al. A medication self-management program to improve adherence to HIV therapy regimens. Patient Educ Couns 2003;50:187-99. 40. Solomon DK, Portner TS, Bass GE, et al. Part 2. Clinical and economic outcomes in the hypertension and COPD arms of a multicenter outcomes study. J Am Pharm Assoc 1998;38:574-85. 41. Telles C, Karno M, Mintz J, et al. Immigrant families coping with schizophrenia. Behavioral family intervention v. case management with a low-income Spanish-speaking population. Br J Psychiatry 1995;167:473-9. 42. Velligan DI, Diamond PM, Mintz J, et al. The use of individually tailored environmental supports to improve medication adherence and outcomes in schizophrenia. Schizophr Bull 2008;34:483-93. DOI 10.1093/schbul/sbm111 43. Vivian EM. Improving blood pressure control in a pharmacist-managed hypertension clinic. Pharmacotherapy 2002;22:1533- 40. 44. Walker CC. An educational intervention for hypertension management in older African Americans. Ethnicity Dis 2000;10:165-74. 45. Webb PA. Effectiveness of patient education and psychosocial counseling in promoting compliance and control among hypertensive patients. J Fam Pract 1980;10:1047-55. 46. Wyatt GE, Longshore D, Chin D, et al. The efficacy of an integrated risk reduction intervention for HIV-positive women with child sexual abuse histories. AIDS Behavior 2004;8:453-62. DOI 10.1007/s10461-004-7329-y 47. Yin HS, Dreyer BP, van Schaick L, Foltin GL, Dinglas C, Meldelsohn AL. Randomized controlled trial of a pictogram-based intervention to reduce liquid dosing errors and improve adherence among caregivers of young children. Arch Pediatr Adolesc 2008;162:814-22. 48. Hirsch JD, Rosenquist A, Brookie M, et al. Evaluation of the first year of a pilot program in community pharmacy: HIV/AIDS medication therapy management for Medi-Cal beneficiaries. J Manage Care Pharm 2009; 15:32-41. 49. Lam AY. Assessing medication consultations, hypertension control, awareness, and treatment among elderly Asian community dwellers. Consult Pharm 2008;23:795-803. 50. Amour CL, Taylor SJ, Hourihan F, et al. Implementation and evaluation of Australian pharmacists’ diabetes care services. J Am Pharm Assoc 2004;44:455-66. 51. Fernandez S, Scales KL, Pineiro JM, Schoenthaler AM, Ogedegbe G. A senior center-based pilot trial of the effect of lifestyle intervention on blood pressure in minority elderly people with hypertension. J Am Geriatr Soc 2008;56:1860-6. DOI 10.1111/j.1532-5415.2008.01863.x 52. Lai LL. Community pharmacy-based hypertension disease-management program in a Latino/Hispanic-American population. Consult Pharm 2007;22:411-6. 53. Leal S, Soto M. Chronic kidney disease risk reduction in a Hispanic population through pharmacist-based disease-state management. Adv Chronic Kidney Dis 2008;15:162-7. DOI 10.1053/j.ackd.2008.01.007 54. Resnick B, Shaughnessy M, Galik E, et al. Pilot testing of the PRAISEDD intervention among African American and low-income older adults. J Cardiovasc Nurs 2009;24:352-61. 55. Robbins B, Rausch KJ, Garcia RI, Prestwood KM. Multicultural medication adherence: a comparative study. J Gerontol Nurs 2004;30:25-32. 56. van Servellen G, Carpio F, Lopez M, et al. Program to enhance health literacy and treatment adherence in low-income HIV-infected Latino men and women. AIDS Patient Care STDs 2003;17:581-94. DOI 10.1089/108729103322555971 57. Westberg SM, Sorensen TD. Pharmacy-related health disparities experienced by non-English-speaking patients: impact of pharmaceutical care. J Am Pharm Assoc 2005;45:48-54. DOI 10.1331/1544345052843066 58. Miller NH. The use of the telephone in cardiac and pulmonary rehabilitation. J Cardiopulm Rehab 1996;16:349-52. 59. Kim MT, Hill MN, Bone LR, Levine DM. Development and testing of the Hill-Bone compliance to High Blood Pressure Therapy Scale. Prog Cardiovasc Nurs 2000;15:90-6. DOI 10.1111/j.1751-7117.2000.tb00211.x 60. Chin MH, Walters AE, Cook SC, et al. Interventions to reduce racial and ethnic disparities in health care. Med Care Res Rev 2007;64:7S-28S. DOI 10.1177/1077558707305413


n


n

981

E Manias and A Williams 61. Shumaker SA, Ockene JK, Riekert KA. The handbook of health behavior change. 3rd ed. New York: Springer, 2009. 62. Schouten BC, Meeuwesen L. Cultural differences in medical communication: a review of the literature. Patient Educ Counseling 2006;64:21-34. DOI 10.1016/j.pec.2005.11.014 63. Lanouette NM, Folsom DP, Sciolla A, Jeste DV. Psychotropic medication nonadherence among United States Latinos: a comprehensive literature review. Psychiatr Serv 2009;60:157-74. DOI 10.1176/appi.ps.60.2.157 64. Williams A, Manias E, Walker R. Adherence to multiple, prescribed medications in diabetic kidney disease: a qualitative study of consumers’ and health professionals’ perspectives. Int J Nurs Stud 2008;45:1742-56. 65. Velligan DI, Wang M, Diamond P, et al. Relationships among subjective and objective measures of adherence to oral antipsychotic medications. Psychiatr Serv 2007;58:1187-92. DOI 10.1176/appi.ps.58.9.1187 66. Grymonpre RE, Didur CD, Montgomery PR, Sitar DS. Pill count, selfreport, and pharmacy claims data to measure medication adherence in the elderly. Ann Pharmacother 1998;32:749-54. DOI 10.1345/aph.17423 67. Robiner WN. Enhancing adherence in clinical research. Contemp Clin Trial 2005;26:59-77. DOI 10.1016/j.cct.2004.11.015

Adherencia a Medicamentos en Personas con Trasfondos Culturales y Lingüísticos Diversos: una Revisión Sistemática y un Meta-Análisis E Manias y A Williams Ann Pharmacother 2010;44:964-82. EXTRACTO TRASFONDO: La adherencia a los medicamentos es de particular importancia para personas de trasfondos culturales y lingüísticos diversos (TCLD) debido a la dificultad con el lenguaje, falta de apoyo social y organizacional, falta de acceso a los recursos de cuidado de la salud y desencaje con el sistema de cuidado de salud. OBJETIVO: Evaluar el impacto de intervenciones para mejorar la adherencia a medicamentos en personas con TCLD a través de una revisión sistemática y meta-análisis. MÉTODO: Se realizfi una búsqueda utilizando las siguientes fuentes de información: The Cochrane Database of Systematic Reviews, Cumulative Index to Nursing & Allied Health Literature, EMBASE, Journals@OVID, PsychInfo, PubMed, Science Direct, Scopus, y Web of Science. La búsqueda cubrió desde enero 1978 hasta octubre 2009. RESULTADOS: Cuarenta y seis artículos revisados se evaluaron como relevantes, los cuales incluyeron 36 estudios aleatorios y controlados, 2 estudios observacionales de cohorte, y 8 estudios cuasi experimentales. El método más común para evaluar adherencia a medicamentos fue métodos de auto-informe, tales como la escala Morisky y sus modificaciones. Algunos estudios utilizaron combinaciones de medidas de adherencia, y la adherencia incluyendo sistemas de monitoreo de eventos de medicación (SMEM) se utilizfi en solamente 6 estudios. Individuos de TCLD fueron reclutados conjuntamente con personas con trasfondos no-culturalmente o lingüísticamente diversos y los análisis subsiguientes tendieron a cubrir toda la muestra. Veinte estudios demostraron mejoría estadísticamente significativa en la adherencia a medicamentos, 15 de los cuales fueron estudios aleatorios y controlados. Seis de las intervenciones exitosas incluyeron la entrega por una persona bilingüe o el uso de material traducido, y 4 incluyeron el uso de un modelo conceptual. Los meta-análisis demostraron mejoría moderada en la adherencia a medicamentos.

982

n


n

Se han realizado relativamente pocos estudios de alta calidad sobre intervenciones para aumentar la adherencia a medicamentos en personas de TCLD. Se necesita que se dé más atención a examinar las necesidades específicas de grupos poblacionales de TCLD. Los futuros investigadores deben considerar evaluar rigurosamente las intervenciones que toman en cuenta la enorme diversidad y diferencias que existen en cualquier grupo particular de TCLD.

CONCLUSIONES:

Traducido por Giselle Rivera Miranda

Adhésion aux Médicaments chez une Population avec Diversité Culturelle et Linguistique: Une Revue Systématique et une MétaAnalyse. E Manias et A Williams Ann Pharmacother 2010;44:964-82. RÉSUMÉ OBJECTIF: Évaluer l’impact des interventions pour améliorer l’adhésion aux médicaments chez une population avec diversité culturelle et linguistique via une revue systématique et une méta-analyse. MÉTHODES: Une recherche a été effectuée dans les banques de données suivantes: Cochrane Database of Systematic Reviews, Cumulative Index to Nursing & Allied Health Literature, EMBASE, Journals@Ovid, PsychInfo, PubMed, Science Direct, Scopus, et Web of Science. Les banques de données ont été explorées de janvier 1978 à octobre 2009. RÉSULTATS: Un nombre de 46 articles de revue ont été évalué comme étant pertinent; ceci incluait 36 articles randomisés, 2 études observationnelles de cohorte et 8 études avec devis quasi expérimentales. La méthode la plus fréquemment utilisée pour évaluer l’adhésion aux médicaments était un questionnaire d’auto-estimation soit l’échelle Morisky modifiée. Peu d’études ont utilisé d’autres méthodes pour évaluer l’adhésion au traitement et un système de surveillance des événements de médication a été utilisé dans 6 études seulement. Les personnes avec diversité culturelle et linguistique ont été recrutées avec les personnes sans diversité culturelle et linguistique et les analyses postérieures avaient tendance à inclure l’échantillon au complet. Vingt études ont démontré une amélioration statistiquement différente dans l’adhésion aux médicaments dont 15 étant des etudes cliniques aléatoires. Six des interventions les plus réussies ont impliqué la présence d’une personne bilingue ou l’utilisation des documents traduits et dans 4 des interventions l’utilisation d’un modèle conceptuel. Les méta-analyses ont démontré des améliorations modestes dans l’adhésion aux médicaments. CONCLUSIONS: Il existe peu d’études de bonne qualité pour évaluer l’adhésion aux médicaments chez une population avec diversité culturelle et linguistique. Une attention particulière doit être portée aux populations avec diversité culturelle et linguistique. Les études à venir doivent valider des interventions qui prennent en considération les différences qui existent dans les différents groupes avec diversité culturelle et linguistique.


Traduit par Louise Mallet

theannals.com