Association between patient beliefs and medication adherence following hospitalization for acute coronary syndrome

Association between patient beliefs and medication adherence following hospitalization for acute coronary syndrome

Association between patient beliefs and medication adherence following hospitalization for acute coronary syndrome Nancy M. Allen LaPointe, PharmD, a ...

277KB Sizes 2 Downloads 23 Views

Association between patient beliefs and medication adherence following hospitalization for acute coronary syndrome Nancy M. Allen LaPointe, PharmD, a Fang-Shu Ou, MS, b Sara B. Calvert, PharmD, b Chiara Melloni, MD, MHS, c Judith A. Stafford, MS, b Tina Harding, BSN, b Eric D. Peterson, MD, MPH, c and Karen P. Alexander, MD c Durham, NC

Background Patient adherence to medications is crucial for reducing risks following acute coronary syndrome (ACS). We assessed the degree to which medication beliefs were associated with patient adherence to β-blockers, angiotensinconverting enzyme inhibitor (ACEI)/angiotensin receptor blocker (ARB), and lipid-lowering medications (LL) 3 months following ACS hospitalization. Methods

We enrolled eligible ACS patients from 41 hospitals to participate in a telephone survey. The Beliefs in Medication Questionnaire-Specific was administered to assess perceived necessity for and concerns about heart medications. Three cohorts were identified for analysis: β-blockers, ACEI/ARBs, and LL. Patients discharged on or starting the medication class after discharge were included in the cohort. The primary outcome was self-reported nonadherence to the medication class 3 months following hospitalization. Factors associated with nonadherence to each medication class were determined using logistic regression analysis.

Results Overall, 973 patients were surveyed. Of these, 882 were in the β-blocker cohort, 702 in the ACEI/ARB cohort, and 873 in the LL cohort. Nonadherence rates at 3 months were 23%, 26%, and 23%, respectively. In adjusted analyses, greater perceived necessity for heart medications was significantly associated with lower likelihood of nonadherence in all cohorts (β-blocker: odds ratio 0.94, 95% CI 0.91-0.98; ACEI/ARB: OR 0.94, 95% CI 0.90-0.98; LL: OR 0.96, 95% CI 0.921.00). A greater perceived concern was significantly associated with a higher likelihood of nonadherence in all cohorts (β-blocker: OR 1.08, 95% CI 1.04-1.13; ACEI/ARB: OR 1.07, 95% CI 1.02-1.11; LL: OR 1.09, 95% CI 1.05-1.14). Conclusions Patients' perceived necessity for and concerns about heart medication were independently associated with adherence to 3 medication classes. Assessment of patient beliefs may be useful in clinical practice to identify those at greatest risk for nonadherence and to stimulate development of individualized interventions to change beliefs and improve adherence. (Am Heart J 2011;161:855-63.)

Medication nonadherence has been consistently associated with worse clinical outcomes and higher downstream rehospitalization and resource use.1-6 Therefore, understanding the factors associated with nonadherence and developing methods to improve medication adherence are high priorities. To date, studies on nonadher-

From the aDivision of Clinical Pharmacology and Duke Clinical Research Institute, Duke University Medical Center, Durham, NC, bDuke Clinical Research Institute, Duke University Medical Center, Durham, NC, and cDivision of Cardiology, Duke Clinical Research Institute, Duke University Medical Center, Durham, NC. Submitted November 9, 2010; accepted February 11, 2011. Reprint requests: Nancy M. Allen LaPointe, PharmD, P.O. Box 17969, Durham, NC 27715. E-mail: [email protected] 0002-8703/$ - see front matter © 2011, Mosby, Inc. All rights reserved. doi:10.1016/j.ahj.2011.02.009

ence have been conflicting and difficult to generalize, in part because many tools have looked only at demographic, clinical, practice setting, or socioeconomic factors associated with adherence rather than attempting to measure patient beliefs.7-10 The purpose of this study was to determine if there was an association between perceived necessity and concerns and self-reported adherence to β-blockers, angiotensin-converting enzyme inhibitor/angiotensin receptor blocker (ACEI/ARB), and lipid-lowering medications 3 months after hospital discharge for acute coronary syndrome (ACS).

Methods From January 2006 through September 2007, Englishspeaking patients at 41 hospitals who participated in Can

American Heart Journal May 2011

856 Allen LaPointe et al

Figure 1

Patient survey flow diagram for each cohort. MD, Physician.

Rapid Risk Stratification of Unstable Angina Patients Suppress Adverse Outcomes with Early Implementation of the American College of Cardiology/American Heart Association Guidelines (CRUSADE), a national quality improvement initiative, were consented for participation in a longitudinal followup survey study called Medications Applied and Sustained Over Time (MAINTAIN).11 Entry criteria for CRUSADE were previously published.12 Follow-up surveys were conducted via telephone at 3 months after the ACS hospitalization using a standardized script and list of survey questions. During the telephone survey, patients were asked questions about their current medication regimen, out-of-pocket medication expenses, rehospitalizations, angina symptoms, symptoms of depression, communication with health care provider(s) regarding medications, and beliefs in heart medications.11 Baseline patient characteristics and in-hospital events and procedures were obtained from the CRUSADE registry and linked to survey responses. The institutional review board of each participating hospital approved its organization's participation in CRUSADE. In addition, the institutional review board of each participating hospital reviewed and approved the longitudinal survey study. Informed consent was obtained from each participant at the time of hospitalization by study staff at the participating hospital. Follow-up survey telephone calls were centrally conducted by the Duke Clinical Research Institute.

Study population Because not all participating patients were eligible for and/or discharged on all 3 of the medication classes of interest (βblockers, ACEI/ARB, and lipid-lowering medications), we identified 3 subpopulations to analyze adherence to each medication class separately. Each subpopulation included only those patients who were discharged from the hospital on the medication class of interest and those patients who started the medication class of interest after hospital discharge but before the 3-month follow-up survey. Patients who reported that their physician discontinued the medication(s) in the medication class of interest before the 3-month follow-up survey were excluded from the subpopulation for that medication class (Figure 1).

Beliefs in medications To assess beliefs in heart medications, the Beliefs in Medication Questionnaire-Specific (BMQ-Specific) was administered.13 The BMQ-Specific is a validated test instrument to assess personal beliefs about necessity and concerns related to diseasespecific medications. For our study, the questionnaire was specific for heart medications. The questionnaire consists of 2 scales—one assessing necessity of prescribed medications and one assessing concerns about potential adverse consequences of taking the medications. Each scale consists of 5 statements.

American Heart Journal Volume 161, Number 5

Respondents were asked to provide their level of agreement with each statement using a 5-point Likert scale (ranging from 1 = strongly disagree to 5 = strongly agree). Therefore, scores on each scale range from 5 to 25, with higher values indicating greater necessity or concern, respectively. In addition, to assess the relative importance of necessity and concerns for a patient, a necessity-concerns differential is calculated by subtracting the concerns score from the necessity score. The range of this score is −20 to 20, with a positive score indicating the perception that benefits outweigh risks and a negative score indicating the perception that risks outweigh benefits. Permission for use of this instrument was obtained from Dr Horne.

Medication adherence During the 3-month follow-up phone call, subjects were asked to provide a list of their current medications. Medications were categorized into predetermined medication classes. For these analyses, the medication classes of interest were β-blockers, ACEI/ARB, and lipid-lowering therapies. All branded and generic names of medications within these classes were obtained and updated throughout the course of the study. Patients who were discharged on a medication class of interest but who did not include a medication in that class on their self-reported list were specifically asked about why they did not include the medication. As described above, patients who reported that their physician discontinued the medication were excluded from the study cohort (Figure 1). Patients who reported that they self-discontinued the medication were included in the study cohort and considered to be nonadherent to that medication class. For all self-reported medications, patients were asked how often they missed a dose in the past month using a 4-point scale (never missed, rarely missed [missed 1 dose per week], sometimes missed [missed 2-3 doses per week], or often missed [missed N 3 doses per week]). Given the expected tendency for patients to underreport missed doses, for these analyses, only patients reporting that they never missed a dose were considered adherent.

Statistical analyses Medians (interquartile ranges) were reported for continuous variables, and frequencies (percentages) were reported for categorical variables. Median and interquartile ranges for the BMQ-Specific necessity scale, concern scale, and differential (necessity score minus concern score) were calculated for each study cohort. Differences in baseline patient characteristics, BMQ-Specific scores, and other survey responses between patients who were adherent and nonadherent were compared in each of the 3 study cohorts. Continuous variables were compared using Wilcoxon rank sum tests, and categorical variables were compared using χ2 tests. To investigate the association between adherence and necessity and concern scores, patient characteristics, and other survey responses, a multivariable logistic regression method was used. The selection of variables entered into each of the 3 models was based upon those with statistically significant differences in the univariate comparisons for each cohort (β-blockers at 3 months, ACEI/ARBs at 3 months, and lipid-lowering therapy at 3 months). In addition, because patients within the same hospital are more likely to be treated in a similar way, generalized estimating equation

Allen LaPointe et al 857

models with exchangeable working correlation structure were used to adjust for correlations among clustered responses (within-hospital correlations).14 The specific variables included in each model are presented with the results. In addition, because of the possibility that patients who self-discontinued a medication might differ from patients who missed doses of a medication, a sensitivity analysis was done in which patients who reported self-discontinuation of the medication in the medication class of interest for each model were excluded from the analyses. A P value b .05 was considered statistically significant for all tests, and all tests were 2-tailed. No adjustments were made for multiple comparisons because this was an observational study and exploratory in nature. All analyses were performed using SAS software (version 9.2; SAS Institute, Cary, NC). CRUSADE is funded by the Schering-Plough Corporation. Bristol-Myers Squibb/Sanofi Pharmaceuticals Partnership provides additional funding support. MAINTAIN is funded by Bristol-Myers Squibb/Sanofi-Aventis Pharmaceutical Partnership and by Merck Schering-Plough Pharmaceutical. The authors are solely responsible for the design and conduct of this study, all study analyses, the drafting and editing of the paper, and its final content.

Results A total of 1,195 patients at 41 hospitals were enrolled in MAINTAIN; however, only 1,182 patients were alive at the time of hospital discharge. Of these, 973 patients completed the 3-month follow-up survey. The 3 study cohorts for analysis of 3-month adherence to β-blockers, ACEI/ARB, and lipid-lowering therapy were obtained from these 973 patients. The number of patients within each cohort and the reasons for exclusion for each cohort are presented in Figure 1.

Study cohorts and nonadherence rates at 3 months A total of 882 patients were included in the β-blocker cohort. Of these, 843 (95.6%) patients were discharged from the hospital with a β-blocker; and 39 (4.4%) patients started the β-blocker after hospital discharge. A total of 204 (23.1%) patients reported nonadherence to the βblocker. These included 26 (2.9%) patients who reported self-discontinuation of the β-blocker. A total of 702 patients were included in the ACEI/ARB cohort. Of these, 623 (88.7%) patients were discharged from the hospital with an ACEI/ARB; and 79 (11.3%) patients started the ACEI/ARB after hospital discharge. Most patients (552 or 78.6%) were taking an ACEI instead of an ARB. A total of 182 (25.9%) patients reported nonadherence to the ACEI/ARB. This included 55 (7.8%) patients who reported self-discontinuation of the ACEI/ARB. A total of 873 patients were included in the lipidlowering medication cohort. Of these, 829 (95.0%) patients were discharged from the hospital with a lipidlowering medication; and 44 (5.0%) patients started the lipid-lowering medication after hospital discharge. At

American Heart Journal May 2011

858 Allen LaPointe et al

Table I. Characteristics and survey responses of adherent versus nonadherent patients at 3 months following ACS hospitalization β-Blocker

Characteristic Age, median (25th, 75th percentile) Male sex BMI, median (25th, 75th percentile)

Adherent Nonadherent (n = 678) (n = 204) 61 (53, 72) 69% 28.8 (25.5, 33.4)

56 (49, 64) 71% 29.8 (27.4, 34.0)

Race White 81% 80% Black 13% 15% Other/missing 6% 5% Insurance at discharge HMO/private 55% 57% Medicare 28% 21% Military/VA 4% 3% Medicaid 3% 5% Self/none 8% 12% Missing 1% b1% History of Hypertension 65% 64% Diabetes 28% 34% PAD 8% 9% Dyslipidemia 55% 59% Heart failure 8% 11% Stroke 7% 8% MI 23% 23% PCI 23% 27% CABG 15% 15% Any revascularization 32% 35% In-hospital PCI 62% 50% Medication before admission Lipid medication 43% 45% Aspirin 44% 47% β-Blocker 39% 43% ACEI/ARB 36% 38% Total no. of discharge medications b7 41% 36% ≥7 59% 64% Missing b1% b1% Nonstatin prescribed NA NA at discharge ACEI prescribed NA NA at discharge Patient reported receipt 83% 80% of medication list and written instructions at discharge Patient reported receipt of 79% 76% verbal medication instructions at discharge No. of evidence-based medications⁎ at discharge 0 1% 1% 1 1% 1% 2 6% 6% 3 33% 30% 4 59% 62%

ACEI/ARB P value b.0001

Adherent Nonadherent (n = 520) (n = 182)

Lipid-lowering drug P value

Adherent Nonadherent (n = 674) (n = 199)

P value

60 (53, 71) 69% 28.8 (25.5, 33.3)

56 (48, 64) 69% 29.4 (26.7, 33.3)

b.0001

.45

82% 12% 6%

82% 13% 5%

.96

54% 21% 4% 4% 14% 1%

.07

56% 28% 4% 3% 8% 1%

58% 20% 4% 4% 15% ≤1%

.03

68% 32% 7% 55% 9% 7% 22% 23% 16% 33% 63%

65% 34% 9% 60% 9% 8% 24% 30% 12% 35% 51%

.49 .74 .38 .30 .86 .66 .58 .09 .13 .62 .005

64% 30% 8% 56% 8% 7% 22% 23% 15% 32% 61%

62% 31% 8% 55% 10% 7% 23% 25% 16% 33% 52%

.68 .71 .83 .82 .38 .77 .81 .53 .80 .69 .05

.56 .42 .43 .62

44% 42% 38% 44%

47% 49% 45% 41%

.64 .10 .14 .53

46% 43% 36% 36%

43% 56% 42% 35%

.38 .79 .16 .81

.19

35% 64% 1% NA

.51

NA

41% 58% 1% 18%

36% 63% 1% 26%

.21

NA

38% 62% b1% NA

.02

NA

76%

85%

.01

NA

NA

NA

.57

83%

79%

.29

84%

78%

.82

.35

78%

74%

.51

78%

76%

.77

.93

b1% b1% 5% 23% 71%

0% 1% 3% 19% 77%

.45

1% 1% 6% 33% 59%

0% 1% 5% 32% 63%

.61

61 (54, 72) 66% 29.0 (25.7, 33.7)

57 (51, 65) 74% 29.6 (26.9, 33.7)

.0005

.82

81% 14% 5%

77% 15% 8%

.08

52% 29% 5% 4% 8% 2%

.70 .13 .64 .34 .30 .61 .80 .16 .77 .37 .004

.71 .0009

.11 .10

.89 .02

American Heart Journal Volume 161, Number 5

Allen LaPointe et al 859

Table I (continued ) β-Blocker

Characteristic

Adherent Nonadherent (n = 678) (n = 204)

No. of evidence-based medications⁎ at 3 months 0 0% 0% 1 1% 4% 2 6% 9% 3 36% 35% 4 57% 52% Rehospitalization within 25% 29% 3 months Depression (score ≥ 10) 15% 22% at 3 months Angina in past 2 weeks 20% 30% Participation in diet/ 66% 55% exercise program since discharge Participation in formal 35% 24% cardiac rehabilitation program since discharge Change in work status 15% 18% attributed to heart disease Insurance/program that 90% 83% assists with medication cost Monthly out-of-pocket 26% 27% medication expenses ≥$150 at 3 months Patient strongly agrees 91% 86% or agrees that provider listens Patient strongly agrees 95% 92% or agrees that provider communicates in understandable language Patient strongly agrees 80% 77% or agrees that provider involves patient in treatment decisions Appointment with 96% 94% physician since discharge Patient reported method 70% 63% to assist with daily medication regimen BMQ necessity score, 19 19 median (25th, (17, 21) (16, 20) 75th percentile) BMQ concern score, 12 14 median (25th, (10, 15) (11, 16) 75th percentile) BMQ necessity score − 6 4 BMQ concern score (3, 9) (1, 8) (25th, 75th percentile)

ACEI/ARB P value

Adherent Nonadherent (n = 520) (n = 182)

Lipid-lowering drug P value

1% 4% 6% 35% 54% 31%

b.0001

.18

0% 0% 3% 22% 74% 26%

.02

15%

.003 .006

.03

Adherent Nonadherent (n = 674) (n = 199)

P value

.15

0% 1% 6% 36% 58% 24%

0% 4% 8% 37% 51% 32%

.01

.03

23%

.006

13%

26%

b.0001

19% 65%

30% 57%

.004 .09

20% 68%

28% 51%

.009 b.0001

.005

33%

31%

.53

36%

26%

.01

.27

17%

20%

.31

15%

17%

.35

.007

90%

82%

.01

90%

82%

.007

.69

24%

26%

.50

26%

25%

.73

.05

91%

88%

.57

92%

86%

.02

.22

96%

92%

.09

97%

94%

.07

.46

81%

75%

.11

80%

75%

.21

.37

97%

94%

.17

97%

92%

.04

.09

71%

69%

.73

71%

63%

.07

.06

19 (17, 21)

19 (16, 21)

.17

19 (17, 21)

19 (16, 21)

.31

b.0001

12 (10, 15)

14 (11, 16)

.0003

12 (10, 15)

14 (11, 17)

b.0001

b.0001

6 (3, 10)

5 (1, 8)

b.0001

6 (3, 10)

4 (1, 8)

b.0001

BMI, Body mass index; HMO, health maintenance organization; VA, Veterans Administration; PAD, peripheral artery disease; MI, myocardial infarction; PCI, percutaneous coronary intervention; CABG, coronary artery bypass grafting. Bolded variables were selected for inclusion in the logistic regression models. ⁎ Evidence-based medications were considered to be aspirin, beta-blocker, ACEI/ARB, and any lipid-lowering medication.

American Heart Journal May 2011

860 Allen LaPointe et al

Table II. Factors associated with nonadherence to cardiovascular drugs 3 months following ACS hospitalization Variable BMQ necessity score (per 1-point increase) BMQ concern score (per 1-point increase) Age (per 10-year increase) BMI (per 5-unit increase) In-hospital PCI procedure Participation in cardiac rehabilitation program Participation in diet/exercise program Insurance/program that assists with medication cost No. of evidence-based medications⁎ at 3 months (per 1 increase) Depression at 3 months Angina in past 2 weeks Rehospitalization Doctor appointment since hospital discharge Patient strongly agrees or agrees that provider listens Hospital discharge with nonstatin Hospital discharge with ACEI

Nonadherence to β-blocker (adjusted OR [95% CI])

Nonadherence to ACEI/ARB (adjusted OR [95% CI])

Nonadherence to lipid-lowering therapy (adjusted OR [95% CI])

0.94 (0.91-0.98)

0.94 (0.90-0.98)

0.96 (0.92-1.00)

1.08 (1.04-1.13)

1.07 (1.02-1.11)

1.09 (1.05-1.14)

0.79 (0.71-0.88) 1.13 (1.00-1.27) 0.83 (0.59-1.17) 0.61 (0.47-0.79)

0.76 (0.65-0.89) NA 0.77 (0.61-0.99) NA

0.77 (0.66-0.90) 1.10 (0.98-1.23) NA 0.70 (0.49-0.99)

1.30 (0.59-2.88)

NA

0.95 (0.34-2.65)

0.64 (0.41-0.98)

0.67 (0.45-1.00)

0.66 (0.39-1.11)

0.85 (0.69-1.04)

0.46 (0.38-0.55)

0.73 (0.59-0.89)

1.28 (0.95-1.73) 1.32 (0.98-1.77) NA NA

1.33 (1.01-1.76) 1.68 (1.17-2.42) NA NA

1.74 (1.36-2.23) 1.11 (0.79-1.58) 1.42 (1.01-1.98) 0.46 (0.21-1.03)

0.76 (0.50-1.16)

NA

0.65 (0.38-1.10)

NA NA

NA 1.77 (1.09-2.85)

1.74 (1.31-2.32) NA

Bolded text indicates statistically significant values. NA, Variable was not selected for the model. ⁎ Evidence-based medications were considered to be aspirin, beta-blocker, ACEI/ARB, and any lipid-lowering medication.

3 months following hospital discharge, the majority of patients were taking a statin (92.1%); and 149 (17.1%) patients received both a statin and nonstatin drug. A total of 199 (22.8%) patients reported nonadherence to the lipid-lowering medication(s). These included 36 (4.1%) patients who reported self-discontinuation of the lipidlowering medication.

Comparisons of characteristics of adherent versus nonadherent patients Characteristics of those who were adherent versus those who were nonadherent in each cohort are presented in Table I. The underlying distribution of necessity scores was similar between those who were adherent versus nonadherent in all 3 cohorts. However, there were statistically significant higher concern scores and lower differential scores among those who were nonadherent as compared with those who were adherent in all 3 cohorts (Table I). Factors associated with nonadherence In the multivariable model, an increase in necessity score was statistically significantly associated with lower odds of nonadherence to β-blockers (adjusted odds ratio [OR] 0.94, 95% CI 0.91-0.98), ACEI/ARB (adjusted OR 0.94, 95% CI 0.90-0.98), and lipid-lowering medications (adjusted OR 0.96, 95% CI 0.92-1.00; P = .049). An increase

in concern score was statistically significantly associated with a higher odds of nonadherence to β-blockers (adjusted OR 1.08, 95% CI 1.04-1.13), ACEI/ARB (adjusted OR 1.07, 95% CI 1.02-1.11), and lipid-lowering medications (adjusted OR 1.09, 95% CI 1.05-1.14). Other factors found to be statistically significantly associated with nonadherence to β-blockers, ACEI/ARB, and lipid-lowering medications at 3 months are shown in Table II. In the sensitivity analysis, in which patients who reported self-discontinuation of the medication in the medication class of interest for each model were excluded, necessity and concern scores remained statistically significantly associated with nonadherence in all cohorts (Table III). Increasing age also remained significantly associated with lower likelihood of nonadherence in all cohorts; however, some of the other factors associated with nonadherence were different in this sensitivity analysis and are shown in Table III.

Discussion In this study, lower perceived necessity and higher perceived concern for heart medications were independently and significantly associated with patient-reported nonadherence to β-blocker, ACEI/ARBs, and lipid-lowering medications. Younger age was the only other characteristic found to be statistically significantly associated with a higher risk of self-reported nonadherence to

American Heart Journal Volume 161, Number 5

Allen LaPointe et al 861

Table III. Sensitivity analysis: factors associated with nonadherence to cardiovascular drugs 3 months following ACS hospitalization (patients with self-reported discontinuation excluded) Variable BMQ necessity score (per 1-point increase) BMQ concern score (per 1-point increase) Age (per 10-year increase) BMI (per 5-unit increase) In-hospital PCI procedure Participation in cardiac rehabilitation program Participation in diet/exercise program Insurance/program that assists with medication cost No. of evidence-based medications⁎ at 3 months (per 1 increase) Depression at 3 months Angina in past 2 weeks Rehospitalization Doctor appointment since hospital discharge Patient strongly agrees or agrees that provider listens Hospital discharge with nonstatin Hospital discharge with ACEI

Nonadherence to β-blocker (adjusted OR [95% CI])

Nonadherence to ACEI/ARB (adjusted OR [95% CI])

Nonadherence to lipid-lowering therapy (adjusted OR [95% CI])

0.94 (0.90-0.98)

0.93 (0.88-0.98)

0.95 (0.91-0.99)

1.10 (1.05-1.16)

1.06 (1.01-1.11)

1.09 (1.04-1.14)

0.76 (0.67-0.86) 1.12 (0.99-1.26) 0.89 (0.63-1.24) 0.61 (0.44-0.84)

0.73 (0.63-0.84) n/a 0.76 (0.53-1.09) n/a

0.74 (0.62-0.87) 1.07 (0.94-1.22) n/a 0.73 (0.49-1.07)

1.19 (0.62-2.30)

n/a

1.22 (0.42-3.55)

0.66 (0.43-1.01)

0.61 (0.37-0.98)

0.64 (0.39-1.06)

1.21 (0.98-1.50)

1.29 (0.83-2.00)

1.29 (0.99-1.68)

1.24 (0.92-1.68) 1.44 (1.04-2.00) n/a n/a

1.77 (1.28-2.44) 1.78 (1.12-2.85) n/a n/a

1.54 (1.19-2.00) 1.29 (0.90-1.85) 1.47 (0.94-2.29) 0.41 (0.18-0.92)

0.69 (0.49-0.97)

n/a

0.65 (0.41-1.02)

n/a n/a

n/a 1.19 (0.75-1.90)

1.75 (1.22-2.51) n/a

Bolded text indicates statistically significant values. ⁎ Evidence-based medications were considered to be aspirin, beta-blocker, ACEI/ARB, and any lipid-lowering medication.

all 3 medication classes. The BMQ is a validated instrument for assessing a patient's belief in the necessity for and concerns about medications.13 The BMQ scores have previously been shown to be associated with overall medication adherence, including among patients with heart disease.13,15 However, to our knowledge, this is the first evaluation of the association of BMQ scores and self-reported adherence to specific medication classes among patients with ischemic heart disease. Medication nonadherence is a known barrier to successful management of many acute and chronic diseases, resulting in underachievement of therapeutic goals and higher health care costs. Many studies have found numerous and variable factors associated with medication nonadherence; however, many challenges exist in attempting to fully understand and improve medication adherence.4,8,10 Lack of consistency in defining adherence and in methods used to measure adherence makes comparisons of study results extremely difficult. Factors associated with nonadherence appear to be highly specific to the individual and thus not readily generalizable to observable physical characteristics, specific medical conditions, or medical procedures. Therefore, tools that attempt to capture patient beliefs, abilities, and decision-making priorities may be more successful in identifying and/or predicting patients who may be at greater risk of medication nonadherence. The BMQ appears to be a good initial

attempt at quantifying these largely unobservable characteristics and factors; but additional research is needed, including a better understanding of the relative importance of perceived necessity versus concern. In addition, the application of this instrument within clinical practice is untested; and therefore, additional research is needed in translating this or similar instruments into clinical practice to assess feasibility and outcomes. At present, even when medication nonadherence is identified, the armamentarium of proven interventions and/or tools to improve medication adherence is extremely limited and may require individualization.7-9,16 Other studies have shown that the steepest decline in medication adherence occurs during the early months following hospital discharge.17,18 In this study, only patient age, necessity score, and concern score were found to be associated with nonadherence to all 3 medication classes. There were, however, several other characteristics—such as depression and lack of insurance for medications—that one might expect to be associated with nonadherence that were found to be independently and significantly associated with nonadherence to 2 of the medication classes. Overall, very few other characteristics were found to be significantly associated with nonadherence to individual medication classes when patient beliefs were included in the analytical model. Interestingly, participation in a cardiac rehabilitation program was associated with lower

American Heart Journal May 2011

862 Allen LaPointe et al

nonadherence to β-blockers and lipid-lowering therapy and may be indicative that the individualized attention, follow-up, and reinforcement of good medication-taking behavior that may be delivered during a regular physician appointment could be part of an effective intervention to improve medication adherence. Many studies have evaluated the association between patient age and medication adherence and found variable results.1,3,4,6,18-23 In this study, we found that increasing age was associated with a lower likelihood of nonadherence in all of the medication cohorts. The more widely accepted perception is that older patients are more likely to be nonadherent. Our results indicate that older patients may have developed more reliable methods for self-management of their medications or that they are less willing to report nonadherence. There was no statistically significant difference in the proportion of patients who reported a tool or person to assist with medication self-management between those who were adherent and those who were nonadherent to each of the medication groups, and we did not have the ability to assess accuracy of self-reported adherence. Therefore, we are unable to further speculate on the reason for the identified association between greater age and lower likelihood of nonadherence. There are several limitations to this study. First, patient self-report of medication adherence is not a very robust method for assessing adherence. Previous studies have indicated that patients tend to overrepresent their adherence; however, several qualitative rather than quantitative methods for assessing adherence are commonly used and are found to be somewhat associated with the more quantitative methods. In this study, we only considered patients who reported no missed doses in the preceding 4 weeks to be adherent because of the anticipated tendency for overrepresenting one's true adherence. Secondly, although the study included patients from multiple hospitals, the patients who agreed to participate at these hospitals may not be representative of all patients with an ACS hospitalization in the United States or the CRUSADE registry population. Because of the need to obtain informed consent from the subjects for the conduct of this study, there may have been a selection bias. Lastly, no adjustments were made for multiple comparisons because this was an observational study and exploratory in nature.

Conclusion Patient beliefs in the necessity for and concerns about their heart medications were found in this study to be independently associated with adherence to β-blockers, ACEI/ARBs, and lipid-lowering medications. This 10question assessment of patient beliefs may be useful in clinical practice to identify those at greatest risk for nonadherence and to stimulate the development and testing of individualized interventions to change beliefs and improve adherence.

References 1. Dragomir A, Cote R, Roy L. Impact of adherence to antihypertensive agents on clinical outcomes and hospitalization costs. Med Care 2010;48:418-25. 2. Rasmussen JN, Chong A, Alter DA. Relationship between adherence to evidence-based pharmacotherapy and long-term mortality after acute myocardial infarction. JAMA 2007;297: 177-86. 3. Fischer MA, Stedman MR, Lii J. Primary medication non-adherence: analysis of 195,930 electronic prescriptions. J Gen Intern Med 2010; 25:284-90. 4. Jackevicius CA, Li P, Tu JV. Prevalence, predictors, and outcomes of primary nonadherence after acute myocardial infarction. Circulation 2008;117:1028-36. 5. Simpson SH, Eurich DT, Majumdar SR. A meta-analysis of the association between adherence to drug therapy and mortality. BMJ 2006;333:15. 6. Allen LaPointe NM, Zhou Y, Stafford JA, et al. Association between mortality and persistent use of beta blockers and angiotensinconverting enzyme inhibitors in patients with left ventricular systolic dysfunction and coronary artery disease. Am J Cardiol 2009;103: 1518-24. 7. McDonald HP, Garg AX, Haynes RB. Interventions to enhance patient adherence to medication prescriptions: scientific review. JAMA 2002;288:2868-79. 8. Haynes RB, Ackloo E, Sahota N, et al. Interventions for enhancing medication adherence. Cochrane Database Syst Rev 2008(2): CD000011. 9. Kripalani S, Yao X, Haynes RB. Interventions to enhance medication adherence in chronic medical conditions: a systematic review. Arch Intern Med 2007;167:540-50. 10. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med 2005;353:487-97. 11. Melloni C, Alexander KP, Ou FS. Predictors of early discontinuation of evidence-based medicine after acute coronary syndrome. Am J Cardiol 2009;104:175-81. 12. Hoekstra JW, Pollack Jr CV, Roe MT. Improving the care of patients with non–ST-elevation acute coronary syndromes in the emergency department: the CRUSADE initiative. Acad Emerg Med 2002;9: 1146-55. 13. Horne R, Weinman J, Hankins M. The beliefs about medicines questionnaire: the development and evaluation of a new method for assessing the cognitive representation of medication. Psychol Health 1999;14:1-24. 14. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13-22. 15. Phatak HM, Thomas III J. Relationships between beliefs about medications and nonadherence to prescribed chronic medications. Ann Pharmacother 2006;40:1737-42. 16. Haynes RB, Yao X, Degani A, et al. Interventions to enhance medication adherence. Cochrane Database Syst Rev 2005(4): CD000011. 17. Kramer JM, Hammill B, Anstrom KJ. National evaluation of adherence to beta-blocker therapy for 1 year after acute myocardial infarction in patients with commercial health insurance. Am Heart J 2006;152:454.e1-8. 18. Butler J, Arbogast PG, BeLue R. Outpatient adherence to beta-blocker therapy after acute myocardial infarction. J Am Coll Cardiol 2002; 40:1589-95. 19. Shalansky SJ, Levy AR. Effect of number of medications on cardiovascular therapy adherence. Ann Pharmacother 2002;36: 1532-9.

American Heart Journal Volume 161, Number 5

Allen LaPointe et al 863

20. Newby LK, LaPointe NM, Chen AY. Long-term adherence to evidence-based secondary prevention therapies in coronary artery disease. Circulation 2006;113:203-12. 21. Murray MD, Morrow DG, Weiner M. A conceptual framework to study medication adherence in older adults. Am J Geriatr Pharmacother 2004;2:36-43.

22. Kulkarni SP, Alexander KP, Lytle B, et al. Long-term adherence with cardiovascular drug regimens. Am Heart J 2006;151: 185-91. 23. Dunbar-Jacob J, Bohachick P, Mortimer MK, et al. Medication adherence in persons with cardiovascular disease. J Cardiovasc Nurs 2003;18:209-18.

O N THE MOVE?

Send us your new address at least 6 weeks ahead

Don’t miss a single issue of the journal! To ensure prompt service when you change your address, please photocopy and complete the form below. Please send your change of address notification at least 6 weeks before your move to ensure continued service. We regret we cannot guarantee replacement of issues missed because of late notification. JOURNAL TITLE: Fill in the title of the journal here.

__________________________________________________________

OLD ADDRESS:

NEW ADDRESS:

Affix the address label from a recent issue of the journal here.

Clearly print your new address here. Name ____________________________________________________ Address __________________________________________________ City/State/ZIP _____________________________________________

COPY AND MAIL THIS FORM TO: Elsevier Health Sciences Division Subscription Customer Service 3251 Riverport Lane, Maryland Heights, MO 63043

OR FAX TO:

OR PHONE:

341-447-8029

1-800-654-2452 Outside the U.S., call 341-447-8871

OR E-MAIL:

[email protected]