Professional Documents
Culture Documents
Circulationaha 108 768986
Circulationaha 108 768986
Medication Adherence
Its Importance in Cardiovascular Outcomes
P. Michael Ho, MD, PhD; Chris L. Bryson, MD, MS; John S. Rumsfeld, MD, PhD
Abstract—Medication adherence usually refers to whether patients take their medications as prescribed (eg, twice daily),
as well as whether they continue to take a prescribed medication. Medication nonadherence is a growing concern to
clinicians, healthcare systems, and other stakeholders (eg, payers) because of mounting evidence that it is prevalent and
associated with adverse outcomes and higher costs of care. To date, measurement of patient medication adherence and
use of interventions to improve adherence are rare in routine clinical practice. The goals of the present report are to
address (1) different methods of measuring adherence, (2) the prevalence of medication nonadherence, (3) the
association between nonadherence and outcomes, (4) the reasons for nonadherence, and finally, (5) interventions to
improve medication adherence. (Circulation. 2009;119:3028-3035.)
Key Words: cardiovascular diseases 䡲 healthcare quality assessment 䡲 medication adherence
䡲 outcomes research
“Drugs don’t work in patients who don’t take them.” control of care providers and healthcare systems (eg, pre-
—C. Everett Koop, MD scribing medications at discharge), the achievement of
longer-term therapeutic and outcome goals requires a part-
Adherence has been defined as the “active, voluntary, and
collaborative involvement of the patient in a mutually accept- nership with patients.
able course of behavior to produce a therapeutic result.”1,2 To date, measurement of patient medication adherence and
This definition implies that the patient has a choice and that use of interventions to improve adherence are rare in routine
both patients and providers mutually establish treatment goals clinical practice. For this reason, medication adherence has
been called the “next frontier in quality improvement” and is
Downloaded from http://ahajournals.org by on April 15, 2022
The views expressed in this article are those of the authors and do not necessarily represent the views of the US Department of Veterans Affairs.
From the Denver VA Medical Center (P.M.H., J.S.R.), Denver, Colo; University of Colorado Denver (P.M.H., J.S.R.), Denver, Color; Institute for
Health Research (P.M.H., J.S.R.), Kaiser Permanente of Colorado, Aurora, Colo; Puget Sound VA Medical Center (C.L.B.), Seattle, Wash; and University
of Washington (C.L.B.), Seattle, Wash.
Correspondence to P. Michael Ho, MD, PhD, 1055 Clermont St (111B), Denver, CO 80220. E-mail michael.ho@va.gov
© 2009 American Heart Association, Inc.
Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.108.768986
3028
Ho et al Medication Adherence 3029
Indirect methods of adherence assessment include patient each measure has its strengths and limitations, there is
questionnaires, self-reports, pill counts, rate of prescription currently no general consensus as to the best measure to use
refills, assessment of the patient’s clinical response, elec- to define adherence or persistence. These studies highlight the
tronic medication monitors, measurement of physiological challenges of measuring medication adherence in routine
markers, and patient diaries.5 The most commonly used clinical practice and in research studies given the lack of a
indirect methods include patient self-report, pill counts, and “gold standard” criterion.
pharmacy refills. The Morisky scale is a commonly used, On the basis of pharmacy refill data, patients with medi-
validated, 4-item self-reported adherence measure that has cations available 80% of the time have generally been
been shown to be predictive of adherence to cardiovascular categorized as adherent in the literature. This dichotomous
medications and blood pressure control.8,9 In addition, Gehi et cutoff is somewhat arbitrary; however, it has been used for a
al10 found that patient self-report of medication nonadherence majority of the studies in the literature on medication adher-
was strongly associated with adverse cardiac events, includ- ence, with data from both observational and randomized,
ing coronary heart disease death, myocardial infarction, and controlled clinical trials. In addition, adherence based on this
stroke, on the basis of a single screening question (“In the cut point has been associated with both intermediate and hard
past month, how often did you take your medications as the outcomes. However, a more recent analysis suggests that
doctor prescribed?”) among patients with known coronary there continues to be reductions in LDL cholesterol and blood
artery disease. However, self-report measures can be biased pressure with adherence levels beyond 80% (eg, 80% to
by inaccurate patient recall or by social desirability, whereby 100%), which suggests that the optimal level of adherence
patients report an overly optimistic estimation of adherence to may be higher than current cutoffs.20 Certainly for conditions
their healthcare providers. Pill counts are easy to perform, such as human immunodeficiency virus or medications such
have been correlated with electronic medication monitors, as oral contraceptives, the 80% cutoff may be too low.
and are frequently used in randomized, controlled clinical Although the current 80% cutoff appears reasonable for
trials to assess medication adherence,11–13 Although simple to cardiovascular medications, future studies focused on medi-
measure, pill counts do not accurately capture the exact cation adherence using pharmacy refill data should report
timing of medication taking, and the data can be manipulated
both continuous measures of adherence and the distribution
by patients (eg, pill dumping). Each of these methods has
of adherent patients based on different dichotomous cutoffs.
advantages and disadvantages, and the use of a specific
The appropriate cutoff will depend on the specific medica-
method to measure adherence will depend on the clinical
tion, its formulation (eg, once daily versus twice daily), and
scenario and availability of the data.
the specific disease condition.
Downloaded from http://ahajournals.org by on April 15, 2022
For other cardiovascular conditions, the prevalence of an adverse outcome, including rehospitalization and mortal-
medication nonadherence varies tremendously depending on ity, in the subsequent 11 months.31 In addition, poor adher-
the population studied and the specific medications assessed. ence to heart failure drugs was associated with an increased
For example, Vrijens et al,25 using medication event monitor number of cardiovascular-related emergency department vis-
(MEMS) data, found that about half of all patients prescribed its.32 These effectiveness studies reinforce the benefits of
antihypertensive medications stopped taking them within 1 cardiovascular medications in routine clinical practice and
year of the initial prescription. They also found that on any 1 highlight the importance of taking these medication as pre-
day, patients omitted ⬇10% of the scheduled doses of scribed to optimize patient outcomes.
medications.25 In contrast, Bramley et al26 found that ⬇75%
of patients on monotherapy for hypertension were highly Healthy Adherer Effect
adherent, defined as a medication possession ratio of 80% to Although the association between medication nonadherence
100%. Among heart failure patients, studies of medication and adverse outcomes has been demonstrated in many obser-
adherence have also found widely differing rates of nonad- vational studies, some concern has been raised that this
herence. For example, 1 study reported persistence rates of association may be, at least in part, related to a “healthy
79% for renin-angiotensin inhibitors, 65% for -blockers, adherer” effect.33 The healthy adherer effect implies that the
56% for spironolactone, and 83% for statins 5 years after an lower risk of adverse outcomes associated with adherence
index heart failure hospitalization.27 In contrast, the rate of may be a surrogate marker for overall healthy behavior. This
nonadherence based on pill counts was much lower in the is supported by post hoc analyses of randomized, controlled
Candesartan in Heart Failure: Assessment of Reduction in clinical trials in which even adherence to placebo is associ-
Mortality and Morbidity (CHARM) randomized, controlled ated with better outcomes than for patients who are nonad-
trial of patients with heart failure, with 11% of patients taking herent to active treatment. These findings were first noted
fewer than 80% of the prescribed pills.28 Although nonadher- over 25 years ago in the Coronary Drug Project, which
ence to medications is prevalent in cardiovascular popula- involved clofibrate.34 In the -blocker Heart Attack Trial,
tions, the variability of the methods for assessment of poor adherence was associated with higher mortality risk
medication use (eg, self-report or pharmacy refill data) makes regardless of whether patients were randomized to propran-
comparisons across studies and across cardiovascular condi- olol or placebo.35 Similar findings were demonstrated in the
tions difficult. Canadian Amiodarone Myocardial Infarction Arrhythmia
Trial (CAMIAT) study for sudden death, total cardiac mor-
Association Between Medication Adherence tality, and all-cause mortality and in the CHARM program for
Downloaded from http://ahajournals.org by on April 15, 2022
and Outcomes all-cause mortality.28,36 It appears that patients who take their
Many observational studies have evaluated the association medication regularly are also more likely to perform other
between medication adherence and outcomes. In general, healthy behaviors, such as eating properly and exercising
these studies have focused on medications that have already regularly.37,38 These behaviors are often not measured di-
been demonstrated in clinical trials to be efficacious and rectly in prospective or retrospective studies.
therefore are trying to assess the effectiveness of these However, there is also evidence against the healthy adherer
medications in routine clinical practice. Pharmacy refill data effect being a major factor in observed associations between
and patient self-report are the mostly commonly used adher- medication adherence and outcomes. For example, in a study
ence assessment methods in these studies. High adherence of post–myocardial infarction patients by Rasmussen et al,30
(defined as medication possession ratio of 80% to 100%) to the association between adherence to cardiovascular medica-
antihypertensive medications was associated with higher tions and outcomes in which neither clinical evidence nor
odds (odds ratio 1.45, 95% confidence interval 1.04 to 2.02) biological plausibility existed was assessed. They found
of blood pressure control compared with those with medium that adherence to cardiovascular medications was not
or low levels of adherence.26 Similarly, each incremental 25% associated with hospitalization for lung, prostate, or breast
increase in proportion of days covered for statin medications cancer and concluded that the benefit of adherence was
was associated with an ⬇3.8-mg/dL reduction in LDL cho- directly related to the cardiovascular medication rather
lesterol.29 Furthermore, nonadherence to cardiovascular med- than an epiphenomenon of a healthy adherer effect. Al-
ications has been associated with increased risk of morbidity though the debate will continue, the medications under
and mortality. For example, nonadherence to statins in the study have often been demonstrated in randomized, con-
year after hospitalization for myocardial infarction was asso- trolled clinical trials to be efficacious, and therefore, the
ciated with an ⬇12% to 25% increased relative hazard for importance of taking these medications as prescribed
mortality.30 In the chronic coronary artery disease setting, should be reinforced.
nonadherence to cardioprotective medications (-blockers,
statins, and/or angiotensin-converting enzyme inhibitors) was Association Between Medication Adherence
associated with a 10% to 40% relative increase in risk of and Costs
cardiovascular hospitalizations and a 50% to 80% relative Surprisingly little is known about the association between
increase in risk of mortality.14 Another study demonstrated medication adherence and healthcare costs in cardiovascular
that patients discontinuing clopidogrel within 1 month after populations. Sokol et al39 reported that greater adherence to
hospital discharge for acute myocardial infarction and drug- medications for chronic conditions such as hypertension,
eluting stent placement were significantly more likely to have diabetes mellitus, hypercholesterolemia, and heart failure was
Ho et al Medication Adherence 3031
associated with higher medication costs but lower nonmedi- Table. Reasons for Medication Nonadherence
cation medical costs, yielding a net overall reduction in Categories of
healthcare costs. A recent evaluation of statin medication Nonadherence Examples
adherence and healthcare costs among patients with coronary
Health system Poor quality of provider-patient relationship; poor
artery disease in a managed care organization demonstrated communication; lack of access to healthcare;
that higher medication adherence was associated with higher lack of continuity of care
pharmacy costs and lower medical costs; however, overall Condition Asymptomatic chronic disease (lack of physical
healthcare costs were similar between nonadherent and ad- cues); mental health disorders (eg, depression)
herent patients over several years of follow-up.40 On the basis Patient Physical impairments (eg, vision problems or
of available studies to date, it thus remains unclear whether impaired dexterity); cognitive impairment;
medication adherence is associated with lower overall health- psychological/behavioral; younger age;
care costs or is “cost neutral” to the healthcare system. Either nonwhite race
way, medication nonadherence is associated with worse Therapy Complexity of regimen; side effects
patient outcomes, which supports the need for interventions Socioeconomic Low literacy; higher medication costs; poor
to improve medication adherence. Yet, there is a clear need social support
for more research to better understand the association be-
tween adherence and healthcare costs and for formal cost-
occasional single day’s dose, and some timing inconsisten-
effectiveness evaluations to be embedded in studies of inter-
cies; (4) take drug holidays 3 to 4 times per year; (5) take drug
ventions to improve medication adherence.
holidays monthly or more often and have frequent omissions;
A larger number of studies have evaluated the impact of
and (6) take few or no doses.5,47 Most deviations in medica-
changing costs of medications on individual patient adher-
tion taking are due to omissions of doses or delays in taking
ence. Among Medicare⫹Choice beneficiaries, patients who
doses. In addition, it is common for patients to improve their
had drug benefit caps were more likely to be nonadherent to
medications for hypertension, hyperlipidemia, and diabetes.41 medication-taking behavior shortly before and after an ap-
In addition, patients with caps on drug benefits had worse pointment with a healthcare provider, which has been termed
intermediate outcomes (eg, LDL levels and blood pressure) “white-coat adherence.”5
and higher rates of emergency department visits and nonelec- The World Health Organization has categorized potential
tive hospitalizations. In separate studies, changes to out-of- reasons for medication nonadherence into 5 broad groupings
pocket spending doubled the risk of stopping statin therapy, that include patient, condition, therapy, socioeconomic, and
Downloaded from http://ahajournals.org by on April 15, 2022
and higher copayments were associated with lower adherence health system–related factors.48 – 62 Examples for each of
to statins.42,43 Taira et al44 also showed a graded relationship these categories are detailed in the Table. Patient factors
between the level of copayment and medication adherence, associated with medication nonadherence include younger
with patients more likely to refill medications for antihyper- age, nonwhite race, and depression. Conditions that are
tensive medications that had a lower copayment. Finally, asymptomatic and chronic in nature that require long-term
Cole et al45 demonstrated that higher drug copayments for therapy have also been associated with nonadherence. For
angiotensin-converting enzyme inhibitors and -blockers therapy-related factors, the complexity of the regimen and the
were associated with a small decrease in the medication perceived or experienced side effects can impact adherence.
possession ratio among patients with heart failure. These Socioeconomic factors such as lower education level and low
studies suggest that medication costs can have a significant health literacy have been correlated with nonadherence.
impact on nonadherence, and future studies are needed to Although the cost of medications is another important socio-
assess whether lowering medication costs can improve med- economic factor, medication nonadherence remains common
ication adherence and clinical outcomes. even when cost is less of a factor, such as in the Canadian
healthcare system or the US Veterans Health Administration.
Patterns and Reasons for Finally, it remains unclear whether these individual factors
Medication Nonadherence can adequately discriminate between patients who are adher-
The reasons for poor medication adherence are often multi- ent or are not adherent, and this suggests that evaluations of
factorial. Nonadherence to medications can be intentional or nonadherence cannot be targeted to specific patient
nonintentional. Intentional nonadherence is an active process populations/characteristics.
whereby the patient chooses to deviate from the treatment Next, although patients are often implicated as the cause
regimen.46 This may be a rational decision process in which for medication nonadherence, healthcare system factors can
the individual weighs the risk and benefits of treatment also have a significant impact on a patient’s nonadherence to
against any adverse effects. Unintentional nonadherence is a medications. Makaryus et al63 found that ⬍50% of patients
passive process in which the patient may be careless or were able to list all of their medications, and even fewer could
forgetful about adhering to the treatment regimen. This is also recount the purpose of their medications at hospital discharge,
referred to by Vrijens et al25 as the execution of the prescribed which suggests that system factors like the educational
regimen, or how well patients adhere to the dosing regimen. process at hospital discharge can impact medication adher-
On the basis of electronic monitoring data, there are 6 general ence after discharge. Supporting this idea is a prior study that
patterns of execution: (1) Close to perfect adherence; (2) take found that discharge counseling was associated with im-
nearly all doses with some timing irregularity; (3) miss an proved adherence after hospital discharge for myocardial
3032 Circulation June 16, 2009
infarction.21 Next, Coleman et al64 found that medication and hospital admissions, as well as lower healthcare costs. In
discrepancies, defined as lack of agreement between the contrast, Lee et al77 randomized patients to an intervention
prehospital, discharge, and posthospital medication regimens, composed of patient education, medication reminder packag-
were common after hospital discharge, occurring ⬇15% of ing, and frequent clinic visits (every 2 months) versus usual
the time. The causes for these discrepancies were just as care and showed improvements in medication adherence
likely to be related to system factors (eg, conflicting infor- (⬇30%) and systolic blood pressure but not LDL cholesterol.
mation, incomplete discharge instructions) as to patient Across these studies, one of the consistent features of suc-
factors. Furthermore, system-level factors such as the bureau- cessful interventions has been regular follow-up with the
cratic processes associated with insurance claims can also healthcare system. Although multimodal interventions are
influence adherence. In a qualitative study, Bokhour et al65 more likely to be successful than unimodal interventions,
found that in approximately one third of hypertension-related some of the study findings remain mixed with regard to
visits in which blood pressure was not controlled, the care outcomes, and the complexity of the interventions makes it
provider did not even ask about medication taking, and difficult to implement them in routine clinical care.
closed-ended questions during the visit inhibited discussions Although observational studies have highlighted oppor-
on medications or medication-taking behaviors. Finally, tunities to improve medication adherence, the implemen-
Svensson et al66 found that adherence to antihypertensive tation of these interventions in routine clinical practice poses
medication was related to faith in the physician, which many hurdles. First, the successful interventions to date have
suggests that the quality of the provider-patient relationship included multiple components, and the components have
was also an important determinant. Although patients are often been heterogeneous, which makes implementation into
central to taking medications as prescribed, there are many routine practice difficult. These multiple components often
non–patient-related factors that can also affect adherence to require clinical personnel to manage and/or coordinate them,
medications. which increases the cost of implementing the intervention.
Furthermore, because of the current fractured healthcare
Interventions to Improve system, there are logistical challenges to the coordination of
Medication Adherence any intervention between various care providers and across
To date, interventions targeting medication adherence have different healthcare systems. Although integrated health sys-
produced only modest success. In general, unimodal inter- tems like Kaiser or the Veterans Administration may be able
ventions have been less successful than multimodal interven- to overcome these barriers, it remains a significant problem
tions, because the reasons for nonadherence are often multi- for the rest of the healthcare system. Finally, current financial
Downloaded from http://ahajournals.org by on April 15, 2022
factorial.67–72 Unimodal interventions that have demonstrated incentives are not aligned with the promotion of the impor-
some success include those that reduced the number of daily tance of medication adherence. Until some of these barriers
doses of medications, used motivational strategies, packaged are lowered, improving medication adherence broadly in
medications into special containers (eg, pill boxes or blister routine clinical practice will remain challenging.
packs), provided more convenient care, educated patients, or
involved monitoring and feedback. For example, Smith et al73 Research and Clinical Implications
recently published findings of a cluster randomized trial in 4 There have been great strides made toward a better under-
geographically dispersed health maintenance organizations standing of medication adherence and its impact in clinical
that sent informational mailings on -blocker therapy to practice; however, more research is needed to address critical
patients recently discharged for acute myocardial infarction issues in the field. One of these issues is coming to a
and prescribed -blockers. The intervention resulted in an consensus on how to uniformly report measures of medica-
absolute increase of 4.3% of days covered per month for tion adherence and persistence. Because of the variety of data
-blocker medications compared with usual-care patients and sources and adherence measures, it is often difficult to
a 17% relative increase in the likelihood of being adherent to compare adherence rates across studies and conditions. A
-blockers. Although the study findings are positive, it is consensus on the method(s) of measurement will provide
unclear whether the modest improvement in adherence will more comparability across studies. Second, prospective stud-
translate into differences in clinical outcomes. ies focusing on adherence should measure adherence using
Multimodal interventions have shown the most promise different methods to help inform the specific type of nonad-
and have improved both adherence and outcomes. For exam- herence behavior, such as primary nonadherence or problems
ple, Piette et al74,75 randomized veterans with diabetes melli- with execution of the regimen, because each of these behav-
tus to an intervention that consisted of telemonitoring with iors may necessitate a different intervention. Third, new
interactive voice response technology and weekly nurse strategies to improve medication adherence need to be tested
feedback and demonstrated improvements in medication to add to the current knowledge base on how to improve
adherence and diabetes-related symptoms, as well as a trend medication adherence and persistence. These interventions
for improvement in hemoglobin A1C. For heart failure, should focus on improving medication adherence, in addition
Murray et al76 randomized clinically stable outpatients with a to intermediate and hard outcomes. Finally, more research is
diagnosis of heart failure to an intensive pharmacist-led needed to better understand the association between adher-
intervention versus usual care and found a 10.9% improve- ence and healthcare costs, and there is a need for formal
ment in adherence to cardiovascular medications. The inter- cost-effectiveness evaluations to be embedded in studies of
vention patients also had fewer emergency department visits interventions to improve medication adherence. Many new
Ho et al Medication Adherence 3033
especially important in the post– hospital discharge setting. records, and pill counts with electronic monitoring of adherence to anti-
hypertensive therapy. Med Care. 1999;37:846 – 857.
While we await additional evidence from studies, these 13. Gossec L, Tubach F, Dougados M, Ravaud P. Reporting of adherence to
strategies can be easily implemented in routine clinical medication in recent randomized controlled trials of 6 chronic diseases: a
practice without much additional time or resources to help systematic literature review. Am J Med Sci. 2007;334:248 –254.
14. Ho PM, Magid DJ, Shetterly SM, Olson KL, Maddox TM, Peterson PN,
patients take their medications as prescribed. Masoudi FA, Rumsfeld JS. Medication nonadherence is associated with a
broad range of adverse outcomes in patients with coronary artery disease.
Conclusions Am Heart J. 2008;155:772–779.
15. Steiner JF, Koepsell TD, Fihn SD, Inui TS. A general method of com-
Building on recent improvements in the prescription of pliance assessment using centralized pharmacy records: description and
indicated cardiovascular medication to patients, the next step validation. Med Care. 1988;26:814 – 823.
is to improve adherence to prescribed medications.78 Nonad- 16. Hess LM, Raebel MA, Conner DA, Malone DC. Measurement of
adherence in pharmacy administrative databases: a proposal for standard
herence to medications is common and is associated with
definitions and preferred measures. Ann Pharmacother. 2006;40:
adverse outcomes. Nonadherence is not solely a patient 1280 –1288.
problem but is impacted by both care providers and the 17. Steiner JF, Prochazka AV. The assessment of refill compliance using
pharmacy records: methods, validity, and applications. J Clin Epidemiol.
healthcare system. As the first step toward improving adher-
1997;50:105–116.
ence, there needs to be a broader recognition of the problem 18. Halpern MT, Khan ZM, Schmier JK, Burnier M, Caro JJ, Cramer J, Daley
of nonadherence, and once identified, simple strategies WL, Gurwitz J, Hollenberg NK. Recommendations for evaluating com-
should be implemented in daily practice to improve adher- pliance and persistence with hypertension therapy using retrospective
data. Hypertension. 2006;47:1039 –1048.
ence. Certainly, there are still many challenges in further 19. Andrade SE, Kahler KH, Frech F, Chan KA. Methods for evaluation
understanding the reasons for nonadherence and designing of medication adherence and persistence using automated databases.
better interventions to improve adherence; however, getting Pharmacoepidemiol Drug Saf. 2006;15:565–574.
20. Bryson CL, Au DH, Young B, McDonell MB, Fihn SD. A refill
patients to take their medication as prescribed is a worthy adherence algorithm for multiple short intervals to estimate refill com-
goal in order for patients to derive the maximal benefit of pliance (ReComp). Med Care. 2007;45:497–504.
prescribed therapies and is also highly consistent with one 21. Jackevicius CA, Li P, Tu JV. Prevalence, predictors, and outcomes of
primary nonadherence after acute myocardial infarction. Circulation.
of the Institute of Medicine’s goals of care of patient 2008;117:1028 –1036.
centeredness.79 22. Ho PM, Spertus JA, Masoudi FA, Reid KJ, Peterson ED, Magid DJ,
Krumholz HM, Rumsfeld JS. Impact of medication therapy discontin-
uation on mortality after myocardial infarction. Arch Intern Med. 2006;
Sources of Funding 166:1842–1847.
Drs Ho and Bryson are supported by Veterans Administration Health 23. Newby LK, LaPointe NMA, Chen AY, Kramer JM, Hammill BG,
Service Research and Career Development. DeLong ER, Muhlbaier LH, Califf RM. Long-term adherence to
3034 Circulation June 16, 2009
evidence-based secondary prevention therapies in coronary artery disease. 44. Taira DA, Wong KS, Frech-Tamas F, Chung RS. Copayment level and
Circulation. 2006;113:203–212. compliance with antihypertensive medication: analysis and policy impli-
24. Jackevicius CA, Mamdani M, Tu JV. Adherence with statin therapy in cations for managed care. Am J Manag Care. 2006;12:678 – 683.
elderly patients with and without acute coronary syndromes. JAMA. 45. Cole JA, Norman H, Weatherby LB, Walker AM. Drug copayment
2002;288:462– 467. and adherence in chronic heart failure: effect on cost and outcomes.
25. Vrijens B, Vincze G, Kristanto P, Urquhart J, Burnier M. Adherence to Pharmacotherapy. 2006;26:1157–1164.
prescribed antihypertensive drug treatments: longitudinal study of elec- 46. Lowry KP, Dudley TK, Oddone EZ, Bosworth HB. Intentional and
tronically compiled dosing histories. BMJ. 2008;336:1114 –1117. unintentional nonadherence to antihypertensive medication. Ann
26. Bramley TJ, Gerbino PP, Nightengale BS, Frech-Tamas F. Relationship Pharmacother. 2005;39:1198 –1203.
of blood pressure control to adherence with antihypertensive mono- 47. Urquhart J. The electronic medication event monitor: lessons for phar-
therapy in 13 managed care organizations. J Manag Care Pharm. 2006; macotherapy. Clin Pharmacokinet. 1997;32:345–356.
12:239 –245. 48. World Health Organization. 2003. Adherence to Long-Term Therapy:
27. Gislason GH, Rasmussen JN, Abildstrom SZ, Schramm TK, Hansen Evidence for Action. Available at: http://www.who.int/chp/knowledge/
ML, Buch P, Sørensen R, Folke F, Gadsbøll N, Rasmussen S, Køber publications/adherence_introduction.pdf. Accessed May 27, 2009.
L, Madsen M, Torp-Pedersen C. Persistent use of evidence-based 49. National Council on Patient Information and Education. Enhancing Pre-
pharmacotherapy in heart failure is associated with improved scription Medicine Adherence: A National Action Plan. Available at:
outcomes. Circulation. 2007;116:737–744. http://www.talkaboutrx.org/. Accessed September 3, 2008.
28. Granger BB, Swedberg K, Ekman I, Granger CB, Olofsson B, McMurray 50. Bardel A, Wallander MA, Svärdsudd K. Factors associated with
JJ, Yusuf S, Michelson EL, Pfeffer MA; CHARM Investigators. adherence to drug therapy: a population-based study. Eur J Clin
Adherence to candesartan and placebo and outcomes in chronic heart Pharmacol. 2007;63:307–314.
failure in the CHARM programme: double-blind, randomised, controlled 51. Hyre AD, Krousel-Wood MA, Muntner P, Kawasaki L, DeSalvo KB.
clinical trial. Lancet. 2005;366:2005–2011. Prevalence and predictors of poor antihypertensive medication adherence
29. Ho PM, Rumsfeld JS, Masoudi FA, McClure DL, Plomondon ME, in an urban health clinic setting. J Clin Hypertens (Greenwich). 2007;9:
Steiner JF, Magid DJ. Effect of medication nonadherence on hospital- 179 –186.
ization and mortality among patients with diabetes mellitus. Arch Intern 52. Mochari H, Ferris A, Adigopula S, Henry G, Mosca L. Cardiovascular
Med. 2006;166:1836 –1841. disease knowledge, medication adherence, and barriers to preventive
30. Rasmussen JN, Chong A, Alter DA. Relationship between adherence to action in a minority population. Prev Cardiol. 2007;10:190 –195.
evidence-based pharmacotherapy and long-term mortality after acute 53. Charles H, Good CB, Hanusa BH, Chang CC, Whittle J. Racial dif-
myocardial infarction. JAMA. 2007;297:177–186. ferences in adherence to cardiac medications. J Natl Med Assoc. 2003;
31. Spertus JA, Kettelkamp R, Vance C, Decker C, Jones PG, Rumsfeld JS, 95:17–27.
Messenger JC, Khanal S, Peterson ED, Bach RG, Krumholz HM, Cohen 54. Gazmararian JA, Kripalani S, Miller MJ, Echt KV, Ren J, Rask K.
Factors associated with medication refill adherence in cardiovascular-
DJ. Prevalence, predictors, and outcomes of premature discontinuation of
related diseases: a focus on health literacy. J Gen Intern Med. 2006;21:
thienopyridine therapy after drug-eluting stent placement: results from the
1215–1221.
PREMIER registry. Circulation. 2006;113:2803–2809.
55. Rieckmann N, Gerin W, Kronish IM, Burg MM, Chaplin WF, Kong G,
32. Hope CJ, Wu J, Tu W, Young J, Murray MD. Association of medication
Lespérance F, Davidson KW. Course of depressive symptoms and med-
adherence, knowledge, and skills with emergency department visits by
ication adherence after acute coronary syndromes: an electronic medi-
Downloaded from http://ahajournals.org by on April 15, 2022
68. Heneghan CJ, Glasziou P, Petra R. Reminder packaging for improving 74. Piette JD, Kraemer FB, Weinberger M, McPhee SJ. Impact of automated
adherence to self-administered long-term medications. Cochrane calls with nurse follow-up on diabetes treatment outcomes in a
Database Syst Rev. 2006 Jan 25:CD005025. Department of Veterans Affairs Health Care System. Diabetes Care.
69. Schroeder K, Fahey T, Ebrahim S. How can we improve adherence to 2001;24:202–208.
blood pressure-lowering medication in ambulatory care? Systematic 75. Piette JD. Interactive voice response systems in the diagnosis and man-
review of randomized controlled trials. Arch Intern Med. 2004;164: agement of chronic disease. Am J Manag Care. 2000;6:817– 827.
722–732. 76. Murray MD, Young J, Hoke S, Tu W, Weiner M, Morrow D, Stroupe KT,
70. McDonald HP, Garg AX, Haynes RB. Interventions to enhance patient Wu J, Clark D, Smith F, Gradus-Pizlo I, Weinberger M, Brater DC.
adherence to medication prescriptions: scientific review. JAMA. 2002; Pharmacist intervention to improve medication adherence in heart failure:
288:2868 –2879. a randomized trial. Ann Intern Med. 2007;146:714 –725.
71. Kripalani S, Yao X, Haynes RB. Interventions to enhance medication 77. Lee JK, Grace KA, Taylor AJ. Effect of a pharmacy care program on
adherence in chronic medical conditions: a systematic review. Arch Intern medication adherence and persistence, blood pressure, and low-density
Med. 2007;167:540 –550. lipoprotein cholesterol: a randomized controlled trial. JAMA. 2006;296:
72. Petrilla AA, Benner JS, Battleman DS, Tierce JC, Hazard EH. 2563–2571.
Evidence-based interventions to improve compliance with antihyper- 78. Setoguchi S, Glynn RJ, Avorn J, Mittleman MA, Levin R, Winkelmayer
tensive and lipid-lowering medications. Int J Clin Pract. 2005;59: WC. Improvements in long-term mortality after myocardial infarction and
1441–1451. increased use of cardiovascular drugs after discharge: a 10-year trend
73. Smith DH, Kramer JM, Perrin N, Platt R, Roblin DW, Lane K, Goodman analysis. J Am Coll Cardiol. 2008;51:1247–1254.
M, Nelson WW, Yang X, Soumerai SB. A randomized trial of direct-to- 79. Committee on Quality of Health Care in America, Institute of Medicine.
patient communication to enhance adherence to beta-blocker therapy Crossing the Quality Chasm: A New Health System for the 21st Century.
following myocardial infarction. Arch Intern Med. 2008;168:477– 483. Washington, DC: National Academy Press; 2001.
Downloaded from http://ahajournals.org by on April 15, 2022