Listen to this manuscript’s

audio summary by

JACC Editor-in-Chief

Dr. Valentin Fuster.

J O U R N A L O F T H E AM E R I C A N C O L L E G E O F C A R D I O L O G Y VO L . 6 9 , N O . 4 , 2 0 1 7

ª 2 0 1 7 B Y T H E AM E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N

P U B L I S H E D B Y E L S E V I E R

I S S N 0 7 3 5 - 1 0 9 7 / $ 3 6 . 0 0

h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j a c c . 2 0 1 6 . 1 1 . 0 3 4

THE PRESENT AND FUTURE

STATE-OF-THE-ART REVIEW

Improving Medication Adherence inCardiometabolic DiseasePractical and Regulatory Implications

Keith C. Ferdinand, MD,a Fortunato Fred Senatore, MD, PHD,b Helene Clayton-Jeter, OD,b Dennis R. Cryer, MD,c

John C. Lewin, MD,d Samar A. Nasser, PHD, MPH, PA-C,e Mona Fiuzat, PHARMD,b Robert M. Califf, MDb

ABSTRACT

Fro

Lo

Fo

Sch

ser

Ca

ap

Ins

My

He

He

rel

Ma

Medication nonadherence, a major problem in cardiovascular disease (CVD), contributes yearly to approximately 125,000

preventable deaths, which is partly attributable to only about one-half of CVD patients consistently taking prescribed

life-saving medications. Current interest has focused on how labeling and education influence adherence. This paper

summarizes the scope of CVD nonadherence, describes key U.S. Food and Drug Administration initiatives, and identifies

potential targets for improvement. We describe key adherence factors, methods, and technological applications for

simplifying regimens and enhancing adherence, and 4 areas where additional collaborative research and implementation

involving the regulatory system and clinical community could substantially reduce nonadherence: 1) identifying moni-

toring methods; 2) improving the evidence base to better understand adherence; 3) developing patient/health provider

team-based engagement strategies; and 4) alleviating health disparities. Alignment of U.S. Food and Drug Administration

approaches to dissemination of information about appropriate use with clinical practice could improve adherence,

and thereby reduce CVD death and disability. (J Am Coll Cardiol 2017;69:437–51) © 2017 by the American College of

Cardiology Foundation.

P atients who do not adhere to prescribedcourses of medication are at greater risk forpoor outcomes. Although the prevalence of

medication nonadherence is difficult to gauge (1),it remains an undermanaged problem (2). Approxi-mately 50% of patients with cardiovascular disease(CVD) had poor adherence to prescribed medica-tions (3). A study of 1,015 patients with stable coro-nary artery disease showed a 4.4-fold increase in

m the aDepartment of Medicine, Tulane Heart and Vascular Institute, T

uisiana; bU.S. Food and Drug Administration, Silver Spring, Maryland; cCry

undation, New York, New York; and the eDepartment of Clinical Resea

ool of Medicine and Health Sciences, Washington, DC. Dr. Ferdinand ha

ves as consultant for Amgen, Sanofi, Boehringer Ingelheim, and Eli Lilly.

liff currently holds the post of Commissioner of Food and Drugs, U.S

pointment to the FDA, he received research grant funding from the Patie

titutes of Health, FDA, Amylin, and Eli Lilly and Company; received res

ers Squibb, Janssen Research and Development, Merck, and Novartis;

althcare, BMEB Services, Genentech, GlaxoSmithKline, Heart.org–Daiichi

art.org, Regado, and Roche; and held equity in N30 Pharma and Portola

ationships relevant to the contents of this paper to disclose.

nuscript received June 13, 2016; revised manuscript received October 27,

the risk of stroke and a 3.8-fold increase in therisk of death among patients who self-reported asnonadherent (4). Full adherence to guideline-recommended therapies for the treatment of athero-sclerotic disease and management post-myocardialinfarction (MI), with a threshold >80% in thepost-MI population, was associated with a lowerrate of major adverse cardiac events, and withcost savings (5).

ulane University School of Medicine, New Orleans,

erHealth, Washington, DC; dCardiovascular Research

rch and Leadership, George Washington University

s received a grant from Boehringer Ingelheim; and

Dr. Cryer has served as a consultant for Amgen. Dr.

. Food and Drug Administration (FDA); before his

nt-Centered Outcomes Research Institute, National

earch grants and consulting payments from Bristol-

received consulting payments from Amgen, Bayer

Sankyo, Kowa, Les Laboratoires Servier, Medscape/

. All other authors have reported that they have no

2016, accepted November 7, 2016.

ABBR EV I A T I ON S

AND ACRONYMS

CVD = cardiovascular disease

EASi = Enhanced Adherence

Strategic Initiative

FDA = U.S. Food and Drug

Administration

HF = heart failure

OHCA = Office of Health and

Constituent Affairs

Ferdinand et al. J A C C V O L . 6 9 , N O . 4 , 2 0 1 7

Improving Medication Adherence in CVD J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1

438

As part of the Million Hearts Campaign (6),the U.S. Food and Drug Administration (FDA)convened a symposium addressing medica-tion adherence and its effect on the clinicaleffectiveness of approved drugs, patientoutcomes, and health disparities.

In this paper, we summarize key back-ground material from this symposium; pro-vide an overview of the status of currentwork aimed at improving medication adher-ence, particularly in the arena of cardiovas-

cular medicine; and discuss possible future directionsfor this field.

METHODOLOGY

We developed a collaborative initiative with thefollowing objectives: 1) identify key factors that affectmedication adherence; 2) describe an innovativecollaborative approach to address the medicationadherence problem; 3) detail specific features of FDAactivities that are applicable to taking innovativesteps toward attenuating the problem of medicationadherence; 4) describe specific measures to enhancemedication adherence; and 5) describe effective in-terventions for the practicing physician to improvemedication adherence. The authors independentlysearched their respective databases to achieve thecollaborative objective. Published studies wereincluded if they were considered to be significant andrelevant to the objective set forth in this initiative.

KEY FACTORS AFFECTING ADHERENCE

Patient nonadherence to prescribed medicationspresents a multifactorial challenge for physicians andother health care providers. In 2011, the AmericanCollege of Preventive Medicine documented 5 keyfactors that affect adherence and recommended thatresearch and monitoring efforts be focused on them(7). These include: 1) socioeconomic factors; 2) healthcare system–related factors; 3) medical condition–related factors; 4) therapy-related factors; and 5)patient-related factors (Table 1). In the followingsections, we address each factor in detail.

SOCIOECONOMIC FACTORS. Studies evaluating re-lationships between social support and medicationadherence have shown that weaker social support isassociated with poorer adherence. Socioeconomicfactors that adversely affect adherence includepoverty, illiteracy, unemployment, lack of socialsupport networks, unstable living conditions, greaterdistance from treatment centers, higher out-of-pocket cost of medications and care, lack of

transportation, cultural beliefs reflecting mistrust inthe health care system, and family dysfunction (8,9).In some cases, patients attempt to offset the cost ofprescription drugs by reducing the dosage and/orfrequency of a recommended therapy or by purchas-ing medications abroad, reinforcing the importance ofcost as a factor in adherence (10).

Patient demographic characteristics are also asso-ciated with differential rates of medication adher-ence. African-American race, Hispanic ethnicity,female sex, older age, and greater disability havebeen associated with lower rates of adherence.Women are less likely than men to adhere to pre-scribed long-term medications for diabetes mellitusand cardiovascular disease (11). Race and lower so-cioeconomic status are coexistent factors for non-adherence (12). A retrospective exploratory study of51,772 hypertensive patients found that adherencerates were lower among African-American patientscompared with white patients (13,14). Similarly, across-sectional analysis of data from the MedicationEvent Monitoring System over a 30-day intervalshowed that African-American race, female sex, andreceipt of care from a publicly funded versus a privateclinic were all associated with lower rates of adher-ence (15). In a Veterans Administration study of hy-pertensive patients, 63% of African-American menwere found to have inadequate blood pressure con-trol versus 50% of white men, and African-Americanmen were 81% more likely to be nonadherent (16).Of note, access to medications in the VeteransAdministration system is generally equal, regardlessof race, ethnicity, or income (17). In a study ofadherence in post-MI patients (18), 1-year rates ofadherence to beta-blockers were 68%, 66%, 61%,58%, and 57% for white, Asian, Hispanic, NativeAmerican, and African-American patients, respec-tively; among patients with disabilities, 1-yearadherence rates by respective race/ethnicity cate-gories were worse: 59%, 54%, 52%, 47%, and 43%,respectively. Finally, a number of studies havedemonstrated higher rates of nonadherence tomedication regimens among elderly patients. Barriersto adherence identified in these studies includedlower levels of health literacy, reduced cognitivefunction, polypharmacy, and various logistical bar-riers to obtaining medications (19).

Levels of education and literacy are also importantfactors in medication adherence. In a study of 61 pa-tients with heart failure (HF) (20), inability to readprescription labels and lack of understanding wereassociated with increased cardiovascular-relatedemergency department visits, as well as emergencyvisits specifically related to HF. Furthermore,

J A C C V O L . 6 9 , N O . 4 , 2 0 1 7 Ferdinand et al.J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1 Improving Medication Adherence in CVD

439

medication education at the time of hospitaldischarge has shown substantial success in reducinghospital readmission rates (21).

There is significant interaction among these fac-tors, suggesting that efforts to alter nonadherence onan individual or population level must take intoaccount community dynamics that affect the envi-ronment in which individuals make decisions aboutadherence. Efforts that intervene on single factors, orthat occur only in a clinical setting removed from thepatients’ daily environment, seem unlikely to havethe desired level of effect.

HEALTH CARE SYSTEMS–RELATED FACTORS. Theorganization of the health care delivery system isbelieved to play an important role with regard tomedication adherence, but empirical data to supportthis view are scant. Within a given delivery system,physician and caregiver skills are thought to beimportant in developing supportive patient relation-ships. Studies have shown that team-based care, inwhich physician extenders, pharmacists, and nursingteam members are engaged in patient education andmonitoring of adherence, can be effective inimproving adherence and health outcomes (22,23).The role of the pharmacist is critical in the post-hospital setting. Hospital discharge is often associ-ated with multiple changes in medication regimensand discontinuity of care, which is often compoundedby inadequate patient education.

Pharmacist intervention focusing on clarifyingmedication regimens, directions for use, reviewingside effects, and providing patient counseling resul-ted in a significant reduction in adverse drug events30 days after hospital discharge (24). In a study oflow-income, inner-city patients 50 years of age orolder with primary care physician-confirmed HF, a 9-month pharmacist intervention protocol, includingverbal instructions, patient monitoring of medicationmanagement, and other health care encounters andtechnical support, yielded an 11% increase in medi-cation adherence (95% confidence interval: 5% to17%) over standard of care. However, medicationadherence declined to that of the standard of care inthe 3-month post-intervention follow-up period (25).Other studies have shown that early follow-up visitsafter hospital discharge, as well as physician conti-nuity, had a significant effect on outcomes (26,27).This was due to various factors, including assessmentof adherence to medications.

Substantial evidence shows that nonadherence isassociated with increases in hospitalization and useof other medical resources, and these differences inresource use have been demonstrated to translate

into substantially higher costs for nonadherent pa-tients and populations (28–30). However, there arecurrently few studies documenting that specific in-terventions result in significant cost savings at thesocietal or health system levels. There is a criticalneed to understand cost-effectiveness, and the asso-ciation between medication adherence and healthcare costs for health system improvements andimproved access to care within these systems.

MEDICAL CONDITION–RELATED FACTORS. Concomi-tant conditions, particularly mental health diseasesand cognitive impairment, have been associated withpoorer medication adherence in patients with car-diometabolic disease. Depression has been associatedwith nonadherence after adjustment for potentialconfounders, such as age, ethnicity, education, socialsupport, and measures of CVD severity (4). In recentlyhospitalized cardiac patients, improvement in symp-toms of depression was consistently and indepen-dently associated with superior self-reportedadherence to medications and secondary preventionbehaviors (i.e., diet, exercise, and stress reduction)across a 6-month span, whereas improvement inanxiety symptoms was not (31).

Furthermore, comorbidities contribute to poly-pharmacy issues; for example, patients with HF aretypically prescribed 6 to 8 daily medications (32).Collaborative disease management programs, inwhich medications are reviewed with the goal ofeliminating unneeded therapies, may be helpful inthis regard, but little evidence has been generated toguide recommendations for the most effective in-terventions and implementation schemes to improveadherence in the setting of complex conditions.

THERAPY-RELATED FACTORS. Therapy-related bar-riers to adherence appear to be substantial. Sideeffects, greater complexity of regimen, and longerduration of regimen are all associated with lowerrates of adherence, as are frequent changes to theregimen itself (33). In 1 study, nuisance bleeding,which occurred in up to 60% of patients receivingdual antiplatelet therapy, was correlated with non-adherence (34).

In a retrospective cohort study of 255,500 hyper-tensive patients followed for 2 years, adherence wasmonitored across the spectrum of antihypertensivemedications given as monotherapy or in a fixed-dosecombination (35). A total of 56% of the cohort takingmonotherapy and 50% of the cohort taking the fixed-dose combination were defined as nonadherent(Table 2). The percentage of nonadherent patientswas highest for diuretic therapy, either as mono-therapy or in a fixed-dose combination. There was a

TABLE 1 Factors Reported to Affect Adherence

Social and economic dimension

Limited English proficiency

Low health literacy

Lack of family or social support network

Unstable living conditions/homelessness

Burdensome schedule

Limited access to health care facilities

Lack of health care insurance

Inability/difficulty accessing pharmacy

Medication cost

Cultural and lay beliefs about illness and treatment

Health care system dimension

Provider-patient relationship

Provider communication skills

Disparity between health beliefs of patient and provider

Lack of positive reinforcement from provider

Weak system capacity for patient education and follow-up

Lack of knowledge about adherence and interventions for improving it

Patient information materials written at too high of a literacy level

Changes/restrictions affecting formulary

High drug costs/copayments

Poor access/missed appointments

Long wait times

Lack of continuity of care

Condition-related dimension

Chronic conditions

Lack of symptoms

Severity of symptoms

Depression

Psychotic disorders

Mental retardation/developmental disability

Therapy-related dimension

Complexity of medication regimen (number of doses, number of concurrent medications)

Treatment requires mastery of certain techniques (injection, inhaler)

Duration of therapy

Frequent changes in medication regimen

Lack of immediate therapeutic benefit

Medications with associated social stigma

Actual/perceived unpleasant side effects

Treatment interferes with lifestyle/requires significant behavioral change

Continued on the next page

Ferdinand et al. J A C C V O L . 6 9 , N O . 4 , 2 0 1 7

Improving Medication Adherence in CVD J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1

440

trend toward an attenuation of nonadherence forfixed-dose combination therapy. This suggested thatfixed-dose combination therapy might prove benefi-cial in promoting medication adherence.

A nonrandomized study of synchronized prescrip-tion refills was conducted to evaluate whether fillingprescriptions at the same time for common mainte-nance medications would improve medicationadherence (36). This study, although limited (i.e., nota randomized trial, low participation relative toeligible participants, and mail-order pharmacy ser-vice only for maintenance prescriptions), suggestedthe possibility that synchronized refills is a therapy-related streamline mechanism that may improvemedication adherence.

Increased number of medications, prescriptionrefill patterns across different time points, not usingreminder tools (e.g., pillbox), no insurance, lowerlevel of education, and the need for dialysis were allfactors associated with medication discontinuation.These findings suggest a need for common-senseapproaches to simplifying regimens and clear logis-tics as critical factors in better adherence.

PATIENT-RELATED FACTORS. Medication adherenceis adversely affected by patient-related factors thatare closely linked to the other factors previouslydiscussed. Visual, hearing, cognitive, mobility, andswallowing impairments are obvious barriers, andindividual patients with fewer resources or lowerliteracy levels have more difficulty filling pre-scriptions and adhering to complex regimens. Addi-tional factors include lack of knowledge orunderstanding about the disease or need for theprescribed medication, expectations about andperceived benefits of treatment, ability and motiva-tion to follow a medical regimen, frustration, anxiety,and substance or alcohol abuse (37).

COLLABORATIVE APPROACH TO ADDRESS

THE MEDICATION ADHERENCE PROBLEM

The multifactorial challenges of medication adher-ence require a multifactorial solution, and have thusprompted a collaborative effort between the FDA andother agencies or campaigns (i.e., Million HeartsCampaign, National Forum for Heart Disease andStroke Prevention, National Consumer League [NCL],Duke Adherence Alliance, and the Enhanced Adher-ence Strategic Initiative [EASi] Consortium) (CentralIllustration). This collaborative effort has linkedtogether several independent programs in an effort toformulate an overall strategy to improve medicationadherence: knowledge dissemination, patient advo-cacy, improvement of the evidence base to optimizeadherence monitoring and engagement strategies,and alleviation of health disparities that have anadverse effect on medication adherence.

MILLION HEARTS. The Million Hearts Campaign (6) isa national initiative of the Department of Health andHuman Services (HHS), whose goal is to prevent 1million heart attacks and strokes by 2017. Thiscollaborative effort involves multiple governmentagencies and nongovernmental collaborators. One ofits main areas of focus is improving medicationadherence through knowledge dissemination, stake-holder activation, creation of incentives, measuringand reporting, improving population health, andresearch. The initiative seeks to leverage the research

TABLE 1 Continued

Patient-related dimension

Physical factors

Visual impairment

Hearing impairment

Cognitive impairment

Impaired mobility/dexterity

Swallowing problems

Psychological/behavioral factors

Perceived risk/susceptibility to disease

Understanding reason medication is needed

Expectations/attitudes toward treatment

Perceived benefit of treatment

Confidence in ability to follow treatment regimen

Motivation

Fear of possible adverse effects

Fear of dependence

Feeling stigmatized by disease

Frustration with health care providers

Psychosocial stress/anxiety/anger

Alcohol/substance abuse

Adapted with permission from the American College of Preventive Medicine (7).

J A C C V O L . 6 9 , N O . 4 , 2 0 1 7 Ferdinand et al.J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1 Improving Medication Adherence in CVD

441

capabilities of the National Institutes of Health andthe Agency for Healthcare Research and Quality, thepayment and quality measures of the Centers forMedicare and Medicaid Services, and the FDA’sexpertise in the assessment and regulation oftherapeutics.

NATIONAL FORUM. The National Forum for HeartDisease and Stroke Prevention is a consortium of ac-ademic and community health care advocacy groups.The FDA’s Office of Health and Constituent Affairshas signed a 5-year memorandum of understandingwith the National Forum to promote and increase theuse of health knowledge, skills, and practices by thepublic in their daily lives that are designed to reducethe burden of heart disease and stroke.

NATIONAL CONSUMER LEAGUE. The FDA has part-nered with the NCL, a private, nonprofit advocacygroup representing consumers in the marketplaceand workplace. The NCL’s “Script Your Future” effort(38), launched in 2011, is a national campaign focusedon raising awareness of the importance of takingmedications as directed. This campaign is theculmination of years of research, collaboration, andcontributions from more than 135 committed stake-holders, comprising governmental agencies(including the FDA), health professionals, patient andconsumer organizations, industry, and academia.

DUKE ADHERENCE ALLIANCE. The FDA has alsojoined the Duke Adherence Alliance, a think tank thatevaluates the state of medication adherence andmakes recommendations to help ensure that Ameri-cans get the greatest benefit from advances in phar-macology (39).

ENHANCED ADHERENCE STRATEGIC INITIATIVE.

The EASi consortium, which developed alongside theMillion Hearts Campaign, has committed its mem-bers’ expertise to contribute to this campaign.Collectively, the team plans to fulfill the objectivesdeveloped linked to the Million Hearts Campaign andoutlined in this paper: 1) identifying methods formonitoring adherence; 2) improving the evidencebase to better understand adherence; 3) developingpatient/health provider team-based engagementstrategies; and 4) alleviating health disparities.

SPECIFIC FDA ACTIVITIES

During a recent meeting involving the FDA, the Na-tional Institutes of Health, academicians, and repre-sentatives from the medical products industry, anevaluation of medical adherence programs revealedthat for long-term treatments, 36 of 83 interventions

reported in 70 randomized clinical trials were asso-ciated with improvements in adherence, but only 25interventions led to improvements in at least 1treatment outcome (40). No single intervention wasidentified as having a significant effect on adherence.Slight improvements (4% to 11%) were observed inunimodal interventions that included reduction indaily doses, use of motivational strategies, andoptimal packaging strategies, with education, moni-toring, and feedback from the patients. As notedearlier, multimodal interventions have generallyachieved greater success than unimodal ones (41).Interventions that were most effective includedcombinations of more convenient care, information,reminders, self-monitoring, reinforcement, coun-seling, family therapy, psychological therapy, crisisintervention, manual telephone follow-up, and sup-portive care.

In its capacity as both a regulatory and publichealth agency, the FDA is systematically examiningactions and activities that can improve appropriateuse of approved medications. However, the FDA ispart of a much larger ecosystem whose members mustact in concert if adherence is to be substantiallyimproved. In particular, the primary issues involvemedical practice, patient commitment, communityhealth services organization and delivery, andinvolvement in and solutions to access and dispar-ities in health care. Accordingly, the FDA’s efforts arefocused on ways in which regulatory strategies, la-beling, and patient information can interdigitate with

TABLE 2 Nonadherent Patients Taking Antihypertensive Medications*

ACE-i ARB BB CCB Diuretic Agents Total

Monotherapy 31,867 (55.8) 8,462 (57.6) 54,559 (55.2) 18,772 (81.5) 4,774 (66.3) 113,136 (56.3)

Combination 12,268 (50.3) 6,030 (45.9) 4,808 (49.3) 43 (33.9) 4,438 (61.0) 25,587 (50.4)

Values are n (%). *Includes 200,787 patients (78.6%) with monotherapy and 54,714 patients (21.4%) with fixed-dose combination therapy. Reprinted with permission fromSchulz et al. (35).

ACE-i ¼ angiotensin-converting enzyme inhibitors; ARB ¼ angiotensin II receptor antagonists; BB ¼ beta-blocker; CCB ¼ calcium-channel blocker.

CENTRAL IL

Key factors th

Socioe

Health

Conco

Thera

Patien

Ferdinand, K.C. et

Patient nonadherenc

Preventive Medicine

include: 1) socioecon

The multifactorial ch

Administration and o

Duke Adherence Alli

an effort to formulat

adherence monitorin

Ferdinand et al. J A C C V O L . 6 9 , N O . 4 , 2 0 1 7

Improving Medication Adherence in CVD J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1

442

medical practice and health care delivery to improveadherence, as summarized in Table 3.

ROLE OF GENERIC MEDICATIONS. One regulatorystrategy being used to address socioeconomic bar-riers to adherence is to advance access to genericmedications, which facilitates greater penetrationof medication availability across socioeconomicstrata through lower prices. Under the Drug Price

LUSTRATION Multifactorial Approach to Influence

at affect medication adherence

conomic factors

care system-related factors

mitant illness

py-related factors

t-related factors

Collaborative

Key U.S.Food and DrugAdministration

(FDA)partnerships:

al. J Am Coll Cardiol. 2017;69(4):437–51.

e to prescribed medications presents a multifactorial challenge for physician

documented 5 key factors that affect adherence (7), and recommended that

omic factors; 2) health care system–related factors; 3) medical condition–relat

allenges of medication adherence require a multifactorial solution, and have

ther agencies or campaigns (i.e., Million Hearts Campaign, National Forum for

ance, and the Enhanced Adherence Strategic Initiative Consortium). This collab

e an overall strategy to improve medication adherence: knowledge disseminat

g and engagement strategies, and alleviation of health disparities that adve

Competition and Patent Term Restoration Act (PublicLaw 98-417), also known as the Hatch-Waxman Act,generic pharmaceutical manufacturers may obtainFDA approval to market a copy of an innovator drugwithout duplicating the clinical and nonclinicalstudies required for approval of the innovator. Underthe Generic Drug User Fee Act, the FDA hascommitted to rigorous timelines for review of genericdrug applications, and prioritizes applications that

Medication Adherence

Million Hearts Campaign

National Forum

National Consumer League

Duke Adherence Alliance

Enhanced AdherenceStrategic Initiative

approach for enhancing medication adherence

s and other health care providers. In 2011, the American College of

research and monitoring efforts be focused on them. These factors

ed factors; 4) therapy-related factors; and 5) patient-related factors.

thus prompted a collaborative effort between the Food and Drug

Heart Disease and Stroke Prevention, National Consumer League,

orative effort has linked together several independent programs in

ion, patient advocacy, improvement of the evidence base to optimize

rsely affect medication adherence.

TABLE 3 Specific FDA Activities

Generic medications (42)

Critical Path and Sentinel System (43–46)

Health literacy (47–54)

Medication guides (55–57)

Essential medical information (58,59)

Cardiovascular and Endocrine Liaison Program (60)

FDA Advisory committees-Polypill (61–64)

FDA ¼ U.S. Food and Drug Administration.

J A C C V O L . 6 9 , N O . 4 , 2 0 1 7 Ferdinand et al.J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1 Improving Medication Adherence in CVD

443

deal with therapeutic options lacking in competition(42). The FDA continues to develop strategies tooptimize accessibility of medications within thisparadigm. Currently, 88% of prescriptions are forgeneric drugs, including common guideline-recommended cardiovascular therapies, such as sta-tins, aspirin, beta-blockers, angiotensin-convertingenzyme inhibitors, angiotensin receptor blockers,warfarin, and clopidogrel, suggesting that althoughsuch a strategy may be useful, it will not be sufficientin itself to overcome barriers to adherence withoutbeing combined with other approaches.

THE CRITICAL PATH AND SENTINEL SYSTEM. Inresponse to societal expectations about drug safetyand efficacy, the FDA introduced the Critical PathInitiative, with the intent of modernizing drugdevelopment (43). A key facet of this effort is theSentinel Initiative (44,45), a national electronic safetysurveillance system launched in 2008, with the pur-pose of greatly strengthening the FDA’s ability toconduct post-market safety monitoring of medicalproducts. Prior to Sentinel, the FDA relied on report-ing systems such as the FDA Adverse Event ReportingSystem, Manufacturer and User Facility Device Expe-rience Database, and Vaccine Adverse Event Report-ing System, all of which depend on manufacturers,health care providers, and patients to report adverseevents related to health care products. Sentinel wasdesigned to complement the existing FDA systemsthrough active monitoring of regulated products. In2015, Sentinel was used in 17 active protocol assess-ments. The system’s rapid query capabilities ran closeto 2,000 different scenarios. The current Sentineldatabase includes 200 million individuals and morethan 350 million person-years of observation time.The database includes electronic health records pro-vided by hospitals, insurers, and universities (46). AsSentinel evolves, it is anticipated that it will expedi-tiously produce more knowledge about safety and therisk/benefit balance of medical products. This isanticipated to translate into more appropriate use,and better instructions to prescribing physicians and

patients. These features are key parameters in medi-cation adherence, and therefore, it is anticipated thatthe Sentinel program will contribute to improvedadherence. The effect of the Sentinel program onmedication adherence will require evaluation throughan adequately designed monitoring program.HEALTH LITERACY. To increase the utility of the la-bel, the FDA revised the label design by adding a“highlights” section to emphasize the drug’s in-dications and warnings (47). In addition, the FDAadvanced a proposal calling for labels to be moretransparent by reflecting information the reviewingmedical officer used when evaluating the New DrugApplication and by displaying any uncertainties in therisk-benefit analysis (43). However, despite the meritof this proposal, incorporating information used bythe FDA reviewer may not lead to a better under-standing of the drug’s safety and efficacy profileamong health care providers and, in turn, patients whoare prescribed such drugs. A key policy challenge foradvancing high-quality health care is health literacy—the degree towhich people have the capacity to obtain,process, and understand basic health information andservices needed to make appropriate health decisions(48). Health literacy requires both the patient’s abilityto read and understand, as well as the healthcare provider’s ability to adequately communicateimportant medical information. A patient-centeredapproach to designing consumer medication infor-mation could improve the comprehensibility of exist-ing warnings in package inserts (49). The use of precisewording on prescription drug label instructions hasbeen shown to improve patient comprehension, butpatients with limited literacy were more likely tomisinterpret instructions, despite the use of moreexplicit language (50). In addition, specific imple-mentation strategies designed to provide useful in-formation for patients with limited health literacy,such as a 3-part intervention consisting of automatedtelephone reminders, picture prescription cards, andpharmacist communication skills training, did notsignificantly improve refill adherence among inner-city patients. Further exploration of strategies to in-crease health literacy, and thus improve adherence, isclearly warranted (51), and the FDA has an importantrole to play in fostering improvements in communi-cation that lead to improved adherence.

In 2010, the HHS initiated a national action plan toimprove health literacy (52) on the basis of 2 coreprinciples: 1) all people have the right to health infor-mation that helps them make informed decisions; and2) health services should be delivered in ways thatare easy to understand and that improve health,longevity, and quality of life. As part of this national

Ferdinand et al. J A C C V O L . 6 9 , N O . 4 , 2 0 1 7

Improving Medication Adherence in CVD J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1

444

plan, the FDA is working to develop language forproduct labeling that patients can easily understand.The Labeling Development Team, located in the Officeof New Drugs in the Center for Drug Evaluation andResearch at the FDA, has been tasked with ensuringthat prescribing information: 1) is a useful communi-cation tool for health care providers; 2) is on the basis ofregulations and guidance; and 3) conveys the essentialscientific information needed to prescribe and usedrugs and biological products safely and effectively(53). A 2014 FDA guidance described Section 17 inprescription drug product labeling (Patient CounselingInformation), which summarizes information that ahealth care provider should convey to a patient orcaregiver during a counseling discussion (54). ThisHHS health literacy initiative is a work in progress, andits effects on adherence have yet to be assessed.

MEDICATION GUIDES. As part of a strategy toimprove adherence by improving medication knowl-edge, the FDA has required medication guidesdesigned to be issued to consumers along with certainprescribed medications (55). Medication guides arehandouts that address issues specific to particulardrugs and drug classes, and are written in nontech-nical and understandable language, on the basis ofapproved professional labeling for the drug. The FDArequires medication guides to be issued when theAgency determines that 1 or more of the followingcircumstances exists: 1) the drug product is one forwhich patient labeling could help prevent seriousadverse effects; 2) the drug product is one that hasserious risk(s) (relative to benefits), of which patientsshould be made aware because information concern-ing the risk(s) could affect patients’ decisions to use,or continue to use, the product; or 3) the drug productis important to health and has specific directions forits use that require patient adherence (such as keepaway from light, store in refrigerator, do not chew,take only with food, and so on). A key feature of themedication guide is the heading, “What is the mostimportant information I should know about (name ofdrug)?” followed by a statement describing the healthconcern that necessitated the medication guide. Thecommunication features associated with medicationguides are postulated to favorably affect medicationadherence, specifically in a manner that enhances thebenefit-risk ratio of such therapy.

An evaluation of the medication guides program(56) revealed that none of the 40 guides available atthat time (written from 2001 to 2006) met federalrecommendations for readability and use of plainlanguage. All exceeded the recommended seventh- toeighth-grade level for reading comprehensibility;

only 10% provided a summary of content; and 76%failed to limit the scope of information; furthermore,the guides ranged in length from 667 to 4,900 words.Only 23% of patients reported having looked at theguides or accompanying patient information mate-rial. Patients with low literacy were less likely to lookat the medication guides, and they are likely inef-fective in these patients.

Between March 2008 and January 2011, the FDAapproved more than 150 medication guides for prod-ucts approved under a New Drug Application andBiologics License Application as part of a risk evalu-ation and mitigation strategy; currently, 14 cardio-vascular drugs are accompanied by such guides. Asample guide for prasugrel shows that the language isdirected toward a general U.S. population (57), butit has not been ascertained whether adherence,adverse events, or efficacy have been affected by themedication guide. Efforts should be focused ondetermining whether there has been improvement inthis decade on adherence attributable to the use ofmedication guides.

ESSENTIAL MEDICAL INFORMATION. The FDA isdeveloping a framework to provide patients withhigh-quality, up-to-date essential prescription drugproduct information (58) by convening public work-shops on the basis of a cooperative agreement be-tween the Agency and the Brookings Institution’sEngelberg Center for Health Care Reform. Currentlyavailable forms of prescription drug informationintended for patients include the Patient PackageInsert, the Medication Guide, and Consumer Medi-cation Information. From these workshops, aconsensus emerged in favor of a new framework forproviding an easy-to-read, FDA-driven Patient Medi-cation Information document.

A study on the best way to display information thataccompanies prescription medication was conductedby interviewing patients with select immune disor-ders (n ¼ 30), other chronic diseases (n ¼ 30), and thegeneral public (n ¼ 30) (59). Respondent preferencesvaried according to age, education, and health status.The FDA is committed to continued efforts to create afocused, single standardized document that: 1) comesin an easily understood format; 2) is consumer-tested;3) is provided when a patient receives a prescriptionmedication; 4) is intended to be used at home, butdoes not replace or affect professional labeling, in-structions for use, or patient counseling; and 5) isfreely available and easily accessible to health careproviders and patients.

Adherence is increasingly likely to improve to theextent that this publicly vetted information from the

J A C C V O L . 6 9 , N O . 4 , 2 0 1 7 Ferdinand et al.J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1 Improving Medication Adherence in CVD

445

FDA is: 1) consistent with the messages from clini-cians and educators; and 2) can be adapted to beappropriate for local circumstances and the needs ofindividual patients.

FDA OFFICE OF HEALTH AND CONSTITUENT

AFFAIRS. The FDA’s Office of Health and ConstituentAffairs (OHCA) serves as the liaison between the FDAand stakeholder organizations. The OHCA providescoordination for activities related to patient advo-cacy, health professional organizations, consumergroups, state organizations, and industry groups onhigh-priority topics, such as serious and life-threatening diseases, imminent public health needs,and other special health issues, such as medicationadherence. By working with these organizations andpatient advocacy groups, the OHCA may better un-derstand patient-specific concerns and work directlywith these groups on initiatives for implementation.

CARDIOVASCULAR AND ENDOCRINE LIAISON

PROGRAM. The Cardiovascular and EndocrineLiaison Program (CELP) has been under developmentsince November 2011 (60). The CELP was designed toserve as a liaison among the FDA, cardiovascular andendocrine health professionals, and patient commu-nities. CELP encourages and supports the commun-ity’s active participation in developing FDA policiesthat promote healthy dietary and nutritional prac-tices and advance the safety and effectiveness ofmedical products that treat diabetes, hypertension,heart disease, and obesity.

Forums of participation include FDA open publicmeetings, stakeholder calls, listening sessions, andworkshops and conferences, as well as providingwritten or electronic comments to the FDA’s pub-lished documents via open public dockets (GetIllness/Condition Information). FDA-mediated pa-tient engagement across the health care continuumdirectly addresses a key issue (i.e., patient engage-ment and education) that was shown to improvemedication adherence (22,23). Feedback from thepatient’s perspective may help FDA policies to evolveregarding drug development by addressing patientconcerns and thus increasing the likelihood of medi-cation adherence.

FDA ADVISORY COMMITTEES. Advisory Committeesare used to provide advice on individual productapprovals and also to consider major policyissues, including health information and useful ap-proaches to product development to meet needs,such as the need for greater adherence. A recent FDAAdvisory Committee (61) addressed the cardiovascu-lar polypill, a single-dose regimen that typicallycombines a low-dose statin, a beta-blocker, an

angiotensin-converting enzyme inhibitor/angio-tensin receptor blocker, and low-dose aspirin, in aneffort to improve adherence by reducing patient pillburdens. The basic question at issue was whetherpeople are better off receiving some reasonable doseof drugs for secondary prevention of CVD rather thannone, even if they are not getting what is believed tobe optimal dosing. The advisory committee agreedthat there was a significant group of patients who arenot seen frequently for medical care, and for whomtaking the polypill might offer a real advantage. Therewas concern, however, that discontinuing the polypilldue to a side effect associated with just 1 component(e.g., statin intolerance) would thereby result in dis-continuing all other components. The committee’sconsensus was that many in the general populationmight be suitable candidates for the polypill, whichcould improve adherence. Clinical studies of poly-pills, including the UMPIRE (Use of a Multidrug Pill InReducing Cardiovascular Events) (62) and FOCUS(Fixed-Dose Combination Drug for Secondary Car-diovascular Prevention) (63) trials, have shownimproved medication adherence, but the in-vestigators suggested that further research withappropriate trial designs were required to validatethis approach. This sentiment was echoed in theconclusion of the advisory committee.

The HOPE-3 (Heart Outcomes PreventionEvaluation-3) trial demonstrated a significantly lowerrate of cardiovascular events in intermediate-risksubjects without cardiovascular disease for the com-bination of rosuvastatin and candesartan/hydrochlo-rothiazide compared with dual placebo (64). In 12,705racially/ethnically diverse subjects who were takingcombination therapy, adherence was 83.6% at 2 yearsand 74.6% at the end of the trial (median follow-up5.6 years), which was similar to dual placebo,despite a higher incidence of muscle weakness anddizziness. Although the study was not designed toevaluate medical adherence, the data suggested thatcombination therapy would not attenuate adherence,despite additive side effects.

SPECIFIC MEASURES TO ENHANCE

MEDICATION ADHERENCE

Four broad areas where additional focus couldsignificantly contribute to reducing the scope of themedication adherence problem are summarized inTable 4 and are described in the following sections.

METHODS FOR MEASURING AND MONITORING

ADHERENCE. A fundamental issue in developingeffective approaches to improve adherence is thedifficulty of measuring it. As discussed previously,

TABLE 4 Specific Measures to Enhance Medication Adherence

Methods for measuring and monitoring adherence (65–68)

Improvement of evidence base (69–71)

Patient/health care provider team-basedengagement strategies

(72–78)

Alleviation of health disparities (79–83)

Ferdinand et al. J A C C V O L . 6 9 , N O . 4 , 2 0 1 7

Improving Medication Adherence in CVD J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1

446

the series of behaviors required to fill a prescriptionand take the right dose of the right medicine at theright time are complex; furthermore, self-reporting issubject to numerous, well-documented biases andinaccuracies. Studies reporting on approaches thatused substantial focus and concentration of sup-portive resources have shown improved estimates,but the generalizability of such methods to practicewhere additional resources are not available is ques-tionable. New methods, including novel technologiescombined with patient-reported outcome measures,are increasingly being used to monitor and measureadherence. These new methods offer the prospect ofmuch more accurate measurement, with fidelity atthe level of the individual for health care, and forpopulation assessment in research designed todevelop new interventions. Validated methods foranalyzing, monitoring, and ranking identified pre-dictors of nonadherence are critical to these efforts.

A plethora of new biomonitoring and Internetreporting systems have emerged to target adherence(e.g., Telehomecare). Tools, such as special watcheswith alarms, pillbox timers, “smart pill” containers,and automated pill dispensers, are readily available(65). Electronic monitoring has been favored over pillcounts by many authorities as producing the mostaccurate data on adherence (66). Although traditionalpillboxes are useful in improving adherence, ad-vances in digital pillboxes can improve medicationadherence by engaging patients in innovative ways.In 1 study, a wireless electronic pill bottle, whichreminded hypertensive patients to take their medi-cation by emitting light and sound, increased medi-cation adherence by 27% (67). The FDA’s approach toregulating wireless technology in medical devices isevolving in an effort to balance the need for innova-tion with the responsibility for assuring the publicthat devices are safe and effective (68).

IMPROVING THE EVIDENCE BASE FOR

EFFECTIVE THERAPIES, RISK FACTORS

FOR POOR ADHERENCE, AND

INTERVENTIONS TO IMPROVE ADHERENCE

Simple and clear directions for medications areenhanced by consensus in the professional com-munity about the body of evidence supporting

therapeutics. The development of the evidence-basedmedicine paradigm, defined as the systematic colla-tion, synthesis, and application of high-qualityempirical evidence, has been adversely affected byan unmanageably high volume of evidence; statisti-cally significant benefits that have marginal effect inclinical practice; rules and practice prompts thatproduce management-driven, rather than patient-centered care; and poor mapping of evidence-basedguidelines to complex multimorbidity (69).

The developing strategies to improve adherencehave likely been influenced by the challenges of thedeveloping evidence-based medicine paradigm. Ad-vances in information technology will facilitate newapproaches to monitoring and improving medicationadherence, through combinations of structured andunstructured data. As these interventions advance,attention to rigorous science, transparency, andcompleteness of reporting will be crucial, becausemuch of the adherence intervention research in-volves behavioral interventions that are not subjectto federal requirements for reporting in theClinicalTrials.gov registry.

A poor-quality evidence base about interventionscan lead to unsettled science and to differences ofopinion among clinicians about the appropriate use orduration of therapy; this can often affect publicopinion, which can pose challenges to adherence.A recent Danish study (70) showed that when nega-tive news appeared about statins appeared in themedia, patients discontinued their statins and hadworse outcomes.

The Agency for Healthcare Research and Qualityrecently articulated its research priorities forimproving the database by tracking: 1) the dayscovered and prescriptions filled; 2) medicationpossession versus medication prescription; 3) adher-ence by class related to multiple chronic conditions;and 4) patient-reported self-adherence evaluationsystems, and associating prescription data withpharmacy/payer data to track adherence (71). Thecohesiveness of the database and the quality of evi-dence that populates it are central to an effectiveunderstanding of adherence and how to improve it.

PATIENT/HEALTH CARE PROVIDER

TEAM-BASED ENGAGEMENT STRATEGIES

When we consider challenges posed in this arena, theevidence points to a complex problem involving in-dividual, social, and health care–related causes. Forthis reason, it is likely that the entire health careteam, including the patient and family, must beengaged in efforts to improve adherence.

J A C C V O L . 6 9 , N O . 4 , 2 0 1 7 Ferdinand et al.J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1 Improving Medication Adherence in CVD

447

Cardiac rehabilitation represents an engagementstrategy with the potential to improve medicationadherence. This health care resource usually followshospitalization for acute coronary syndrome. Post-hospitalization referral to cardiac rehabilitation is apivotal moment in the patient’s management, and itis likely to have a significant effect on medicationadherence via patient engagement. A survey of car-diac rehabilitation programs listed in the AmericanAssociation of Cardiovascular and Pulmonary Reha-bilitation database showed that only 28% of eligiblepatients used cardiac rehabilitation services. If all 812of these programs functioned at maximum capacity,they could service a maximum of 47% of qualifyingU.S. patients (72). A study of cardiac rehabilitationprogram adherence among women evaluated 3rehabilitation models: supervised mixed-sex, super-vised women-only, or home-based cardiac rehabili-tation. Adherence was defined as the recordednumber of on-site or telephone sessions (in the caseof home-based programs). The overall adherence was64%. The results did not clearly favor any model (73).This study suggested that offering women alternativemodels for cardiac rehabilitation services did notpromote adherence. In a study of 822 consecutivepatients referred to cardiac rehabilitation at theMontefiore Medical Center in the Bronx, New York,nonwhite patients initiated cardiac rehabilitation lessoften than white patients (54% vs. 65%; p ¼ 0.003).Copayment did not influence initiation. White pa-tients were 61% more likely to complete the pre-scribed 36 sessions. Cardiac rehabilitation initiationwas associated with cardiologist involvement. Initi-ation was associated with a survival benefit, andthere were no differences between white andnonwhite patients in survival benefit for those pa-tients initiating rehabilitation (74). Even in a com-munity of a predominantly minority population,racial disparity exists among cardiac rehabilitationparticipants. This study suggested that greaterinvolvement by the cardiologist was pivotal forrehabilitation initiation and adherence, with conse-quent survival benefit.

Medicaid, Indian Health Service programs, andsome community health center programs providedevices for disease-monitoring purposes, e-mailmessaging, clinical reminders, and medication edu-cation. In addition to educating patients about theimportance of adherence, these tools can help pa-tients share concerns about side effects, cost barriers,or other factors that may contribute to nonadherence.Call centers for personal communications from a careteam member are used when patients fail to respondto e-mail or text messages.

Patient web portals have also been used, and theyhold promise to increase patient health under-standing and shared decision making. Nevertheless,although most users (60%) consider portals to be veryuseful (75), overall use remains relatively low, withonly 28% of patients being offered this tool as of 2013.Optimally, patient portals could be used to hoste-visits and telemedicine, with interoperability acrossmultiple providers (e.g., record sharing) and healthcoaching (76). In addition, mobile technology andassociated health applications (apps) have gainedpopularity, with a growing market in recent years.The most common method for engaging primaryusers appears to be self-monitoring (77). However, arecent study that validated blood pressure measure-ments from a highly popular mHealth app (>148,000units sold) found that the app’s measurementswere highly inaccurate. In particular, the appdemonstrated poor sensitivity for hypertensive mea-surements, with the result that approximately four-fifths (77.5%) of individuals with blood pressuremeasurements in the hypertensive range wouldreceive a false assurance that their blood pressurewas within the nonhypertensive range (78). Thesetypes of interventions need to be carefully evaluatedso as to optimize engagement strategies.

ALLEVIATION OF HEALTH DISPARITY

As previously noted, poor adherence is embedded inand exacerbates health care disparities. Someongoing projects to address this include an initiativestarted in September 2015 by the Morehouse School ofMedicine and Emory University. This initiative ex-plores the risk of heart disease and stroke amongminority patients by establishing the MorehouseSchool of Medicine/Emory Cardiovascular Center forHealth Equity, with funding from the American HeartAssociation’s (AHA’s) new Strategically FocusedResearch Network (79).

Another initiative is the African American HealthDisparity Project, which is designed to improve thehealth status of African Americans and to eliminateinstitutional racism in the health care system of SanFrancisco, California. Some of the project’s accom-plishments include no-cost treatment for AfricanAmericans diagnosed with breast cancer, establish-ment of an African American Health Initiative, com-munity grants to address health disparities indisproportionately affected neighborhoods, main-taining posting of pledge statements, providingeducational seminars, and adding a cultural diversitycompetency component to the employee evaluationtool. Some key obstacles included the availability of

TABLE 5 Intervention by the Practicing Clinician to Improve Adherence

Intervention Reference

Improve patient health literacy: label highlights and conveyance ofliteracy-sensitive information to patients. Engage the patient forgreater health care provider-patient interaction: follow-upcommunication with the patient (e.g., cardiac rehabilitation), clinicalreminders, phone calls, communication with the pharmacist,medical education, and use of patient web portals.

(43,47–54,76,77)

Use of generics. (42)

Use of the Sentinel System for reporting adverse events. (43–46)

Communicate with the FDA on key issues from your practice via theCardiovascular and Endocrine Liaison Program.

(60)

Develop an office-based strategy to monitor adherence. (65)

Be judicious about poor-quality evidence. (70)

Establish greater cognizance of inherent health care disparities:build community trust.

(80,81)

FDA ¼ U.S. Food and Drug Administration.

Ferdinand et al. J A C C V O L . 6 9 , N O . 4 , 2 0 1 7

Improving Medication Adherence in CVD J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1

448

funding, the immensity and complexity of the factorsaffecting health disparities, and mistrust in the healthsystem among the African-American community, aswell as lack of communication. Key lessons includedthe need for partnership with the external commu-nity, the importance of developing community trustby building relationships, and the critical need toleverage greater resources for a multidisciplinaryeffort, to educate staff, and to develop culturalsensitivity (80).

The BARBER-1 (Effectiveness of a Barbershop-Based Program to Improve High Blood PressureControl and Awareness in Black Men) study (81)demonstrated a neighborhood-based interventionstrategy involving black-owned barbershops with>95% black male clientele. The cluster-randomizedstudy assigned barbershops to receive either AHApamphlets only, with no other intervention, or toreceive interventions including blood pressurechecks and health messaging through the barber andpeers (but no AHA pamphlets). The results demon-strated a significant increase in hypertension controlin the intervention arm (p ¼ 0.035). These resultssuggest that neighborhood programs designed toimprove blood pressure outcome in African-Americanmen may have clinical utility. These results may beapplicable to other ethnic groups. Further researchinto this neighborhood-based strategy is warranted.

A review of interventions, and state and nationalpolicy initiatives, aimed at reducing racial and ethnicdisparities in blood pressure control showed mixedresults (82). The 39 interventions that were examinedmostly focused on patient education, communica-tion, behavioral intervention, multilevel settings,cultural competency training, community interven-tion, and/or social support. Of these, 27 resulted in

some improvement of blood pressure control; 7 didnot. Among 6 studies that focused on disparities,3 had interventions that led to reduced disparity,2 precipitated increased levels of disparity, and 1showed no effect on disparity. Although severaleffective interventions have been identified toimprove blood pressure in racial and ethnic minor-ities, data on whether these interventions will reducedisparity are limited. The Million Hearts Campaign isaddressing this issue, as is a joint series of projectsundertaken by the National Heart, Lung, and BloodInstitute and the Patient-Centered OutcomesResearch Institute (83).

Another critically important issue is to ensure thatresearch cohorts are diverse in terms of race/ethnicity, sex, and age, as well as socioeconomicstatus. Better information about adherence in thecontext of clinical trials from representative pop-ulations should enable the development of strategiestailored to the needs of disadvantaged populations.The collaborative initiative of the EASi consortiumwill address health disparities and its effect onmedication adherence as a priority item. The magni-tude of this problem requires more than interventionby the FDA; a concerted effort that engages the entirecommunity is needed to ensure success.

EFFECTIVE INTERVENTIONS FOR

THE PRACTICING CLINICIAN TO

IMPROVE ADHERENCE

The state-of-the-art paradigm for enhancing medica-tion adherence is complex and involves multiple hi-erarchies. Collaborative efforts at the national leveland FDA activities steered toward enhancing medi-cation adherence are tools for the practicing clinician.Practical guidelines for the clinician to improvemedication adherence are shown in Table 5. Given ourfindings, patient engagement via communicationwith the patient in a literacy-sensitive manner is asignificant parameter. Phone calls to the patient andproviding educational services have been shown toimprove adherence. FDA label highlights have facili-tated implementation of this guideline. Greater use ofgenerics to reduce cost is warranted. Use of theSentinel program will enhance knowledge about drugsafety and consequent risk/benefit, thereby facili-tating the physician knowledge base in optimizingcommunication with patients. The clinician shouldcommunicate with the FDA on key issues affectingadherence through the CELP.

The office practice should design a monitoringstrategy using current technologies. The physicianshould be judicious about poor-quality evidence that

J A C C V O L . 6 9 , N O . 4 , 2 0 1 7 Ferdinand et al.J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1 Improving Medication Adherence in CVD

449

can lead to unsettled science, resulting in inappro-priate medication discontinuation and worsening ofoutcomes. There should be heightened sensitivity tohealth care disparity, and measures should be takento attenuate this inherent issue.

SUMMARY AND CONCLUSIONS

Improving medication adherence could be one of themost effective and efficient ways of improving healthoutcomes, especially for cardiometabolic disease, theleading cause of mortality in the United States.Because less than one-half of patients with chroniccardiovascular conditions take their prescriptionsconsistently, it is critical that we develop an effectiveand generalizable adherence strategy for patients.Approaches for enhancing adherence are multifac-eted, requiring the participation of governmentalagencies, academia, and organizations devoted tooptimizing health care. The FDA and collaboratorshave taken substantial steps to enhance medicationadherence, and are working closely with stakeholderson a variety of activities. Nevertheless, because of themany social and economic causes of nonadherence,

much work remains before we can claim to have metthe challenge of medication adherence. Alignmentbetween therapeutic evaluation and labeling, and theefforts of clinicians and health systems to improveadherence, could ultimately have a major effect onhealth outcomes.

ACKNOWLEDGMENTS The authors are grateful toJonathan McCall, MS, of the Duke Clinical ResearchInstitute for his invaluable assistance in preparingthis paper. Mr. McCall received no compensation forhis work on this paper other than usual salary. Theauthors would also like to thank Dr. Norman Stock-bridge, Director of the Division of Cardiovascularand Renal Products at the FDA, for facilitating theactivities of the EASi Board. Finally, the authorswould like to thank Julie Harvill, National Forum,for her tireless coordination of collaboration meet-ings and planning.

REPRINT REQUESTS AND CORRESPONDENCE: Dr.Keith C. Ferdinand, Tulane University School ofMedicine, 1430 Tulane Avenue, SL-48, New Orleans,Louisiana 70112. E-mail: kcferdmd@aol.com.

RE F E RENCE S

1. Kolandaivelu K, Leiden BB, O’Gara PT, et al.Non-adherence to cardiovascular medications. EurHeart J 2014;35:3267–76.

2. Baroletti S, Dell’Orfano H. Medication adher-ence in cardiovascular disease. Circulation 2010;121:1455–8.

3. Kronish I, Ye S. Adherence to cardiovascularmedication: lessons learned and future direction.Prog Cardiovasc Dis 2013;55:590–600.

4. Gehi AK, Ali S, Na B, et al. Self-reported medi-cation adherence and cardiovascular events inpatients with stable coronary heart disease: theheart and soul study. Arch Intern Med 2007;167:1798–803.

5. Bansilal S, Castellano JM, Garrido E. Assessingthe impact of medication adherence on long-termcardiovascular outcomes. J Am Coll Cardiol 2016;68:789–801.

6. Centers for Disease Control and Prevention.Million Hearts. Available at: http://millionhearts.hhs.gov. Accessed November 23, 2016.

7. American College of Preventive Medicine.Medication adherence—improving health out-comes. A resource from the American College ofPreventive Medicine. 2011. Available at: http://www.acpm.org/?Adherence. Accessed November22, 2016.

8. Lewis LM, Ogedegbe C, Ogedegbe G. Enhancingadherence of antihypertensive regimens in hy-pertensive African-Americans: current and futureprospects. Expert Rev Cardiovasc Ther 2012;10:1375–80.

9. Braverman J, Dedier J. Predictors of medicationadherence for African American patients diag-nosed with hypertension. Ethn Dis 2009;19:396–400.

10. Cohen RA, Villarroel MA. Strategies Used byAdults to Reduce Their Prescription Drug Costs:United States, 2013. NCHS Data Brief No. 184.Hyattsville, MD: National Center for Health Sta-tistics, 2015.

11. Manteuffel M, Williams S, Chen W, et al. In-fluence of patient sex and gender on medicationuse, adherence, and prescribing alignment withguidelines. J Womens Health (Larchmt) 2014;23:112–9.

12. Rolnick SJ, Pawloski PA, Hedblom BD, et al.Patient characteristics associated with medicationadherence. Clin Med Res 2013;11:54–65.

13. Holmes HM, Luo R, Hanlon JT, et al. Ethnicdisparities in adherence to antihypertensive med-ications of Medicare Part D beneficiaries. J AmGeriatr Soc 2012;60:1298–303.

14. Ishisaka DY, Jukes T, Romanelli RJ, et al. Dis-parities in adherence to and persistence withantihypertensive regimens: an exploratory analysisfrom a community-based provider network. J AmSoc Hypertens 2012;6:201–9.

15. Grigoryan L, Pavlik VN, Hyman DJ. Predictorsof antihypertensive medication adherence in twourban health-care systems. Am J Hypertens 2012;25:735–8.

16. Bosworth HB, Dudley T, Olsen MK, et al. Racialdifferences in blood pressure control: potential

explanatory factors. Am J Med 2006;119:70.e9–e15.

17. Lewis LM, Schoenthaler AM, Ogedegbe G. Pa-tient factors, but not provider and health caresystem factors, predict medication adherence inhypertensive black men. J Clin Hypertens (Green-wich) 2012;14:250–5.

18. Zhang Y, Baik SH. Race/ethnicity, disability,and medication adherence among Medicare ben-eficiaries with heart failure. J Gen Intern Med2013;29:602–7.

19. Gellad WF, Grenard JL, Marcum ZA.A systematic review of barriers to medicationadherence in the elderly: looking beyond cost andregimen complexity. Am J Geriatr Pharmacother2011;9:11–23.

20. Hope CJ, Wu J, Tu W, et al. Association ofmedication adherence, knowledge, and skills withemergency department visits by adults 50 years orolder with congestive heart failure. Am J HealthSyst Pharm 2004;61:2043–9.

21. Bradley EH, Curry L, Horwitz LI, et al.Contemporary evidence about hospital strategiesfor reducing 30-day readmissions: a nationalstudy. J Am Coll Cardiol 2012;60:607–14.

22. Granger BB, Ekman I, Hernandez AF, et al.Results of the Chronic Heart Failure Interventionto Improve MEdication Adherence study: a ran-domized intervention in high-risk patients. AmHeart J 2015;169:539–48.

23. Dunn SP, Birtcher KK, Beavers CJ, et al. Therole of the clinical pharmacist in the care of

Ferdinand et al. J A C C V O L . 6 9 , N O . 4 , 2 0 1 7

Improving Medication Adherence in CVD J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1

450

patients with cardiovascular disease. J Am CollCardiol 2015;66:2129–39.

24. Schnipper JL, Kirwin JL, Cotugno MC, et al.Role of pharmacist counseling in preventingadverse drug events after hospitalization. ArchIntern Med 2006;166:565–71.

25. Murray MD, Young J, Hoke S, et al. Pharmacistintervention to improve medication adherence inheart failure. Ann Intern Med 2007;146:714–25.

26. Hernandez AF, Greiner MA, Fonarow GC, et al.Relationship between early physician follow-upand 30-day readmission among Medicare benefi-ciaries hospitalized for heart failure. JAMA 2010;303:1716–22.

27. Sidhu RS, Youngson E, McAlister FA. Physiciancontinuity improves outcomes for heart failurepatients treated and released from the emergencydepartment. J Am Coll Cardiol HF 2014;2:368–76.

28. Granger BB, Swedberg K, Ekman I, et al., forthe CHARM Investigators. Adherence to cande-sartan and placebo and outcomes in chronic heartfailure in the CHARM programme: double-blind,randomised, controlled clinical trial. Lancet2005;366:2005–11.

29. Hernandez AF, Hammill BG, Peterson ED, et al.Relationships between emerging measures ofheart failure processes of care and clinical out-comes. Am Heart J 2010;159:406–13.

30. Retrum JH, Boggs J, Hersh A, et al. Patient-identified factors related to heart failure read-missions. Circ Cardiovasc Qual Outcomes 2013;6:171–7.

31. Bauer LK, Caro MA, Beach SR, et al. Effects ofdepression and anxiety improvement on adher-ence to medication and health behaviors inrecently hospitalized cardiac patients. Am J Car-diol 2012;109:1266–71.

32. Wong CY, Chaudhry SI, Desai MM, et al. Trendsin comorbidity, disability, and polypharmacy inheart failure. Am J Med 2011;124:136–43.

33. Fried T, Tinetti ME, Towle V, et al. Effects ofbenefits and harms on older persons’ willingnessto take medications for primary cardiovascularprevention. Arch Intern Med 2011;171:923–8.

34. Serebruany V, Rao SV, Silva MA, et al. Corre-lation of inhibition of platelet aggregation afterclopidogrel with post discharge bleeding events;assessment by different bleeding classifications.Eur Heart J 2010;31:227–35.

35. Schulz M, Krueger K, et al. Medication adher-ence and persistence according to different anti-hypertensive drug classes: a retrospective cohortstudy of 255,500 patients. Int J Cardiol 2016;220:668–76.

36. Doshi J, Lim R, Li P, et al. A synchronizedprescription refill program improved medicationadherence. Health Aff (Millwood) 2016;35:1504–12.

37. Brown MT, Bussell JK. Medication adherence:WHO cares. Mayo Clin Proc 2011;86:304–14.

38. National Consumers League. Script YourFuture. 2016. Available at: http://www.scriptyourfuture.org. Accessed November 23,2016.

39. Mozaffarian D, Benjamin EJ, Go AS, et al., forthe American Heart Association Statistics Com-mittee and Stroke Statistics Subcommittee. Heartdisease and stroke statistics—2015 update: areport of the American Heart Association. Circu-lation 2015;131:e29–322.

40. Bosworth HB, Granger BB, Mendys P, et al.Medical adherence: a call for action. Am Heart J2011;162:412–24.

41. Ho PM, Bryson CL, Rumsfeld JS. Medicationadherence: its importance in cardiovascular out-comes. Circulation 2009;119:3028–35.

42. U.S. Food and Drug Administration. GenericDrug User Fee Act program performance goals andprocedures. Available at: http://www.fda.gov/downloads/ForIndustry/UserFees/GenericDrugUserFees/UCM282505.pdf. Accessed November22, 2016.

43. Woodcock J, Woosley R. The FDA Critical PathInitiative and its influence on new drug develop-ment. Ann Rev Med 2008;59:1–12.

44. Robb M, Racoosin JA, Sherman RE, et al. TheUS Food and Drug Administration’s SentinelInitiative: expanding the horizons of medicalproduct safety. Pharmacoepidemiol Drug Saf2012;21 Suppl 1:9–11.

45. U.S. Food and Drug Administration. FDA’sSentinel Initiative–News and Events. 2016.Available at: http://www.fda.gov/Safety/FDAsSentinelInitiative/ucm149341.htm. AccessedNovember 23, 2016.

46. MezherM.Woodcock: Drug Safety SurveillanceSystem Ready for Full Operation. RAPS. February 3,2016. Available at: http://raps.org/Regulatory-Focus/News/2016/02/03/24248/Woodcock-Drug-Safety-Surveillance-System-Ready-for-Full-Operation/. Accessed November 23, 2016.

47. Schwartz L, Woloshin S. Lost in transmission—FDA drug information that never reaches clini-cians. N Engl J Med 2009;361:1717–20.

48. Parker RM, Ratzan SC, Lurie N. Health literacy:a policy challenge for advancing high-qualityhealth care. Health Aff (Millwood) 2003;22:147–53.

49. Webb J, Davis TC, Bernadella P, et al. Patient-centered approach for improving prescription drugwarning labels. Patient Educ Couns 2008;72:443–9.

50. Davis TC, Federman AD, Bass PF III, et al.Improving patient understanding of prescriptiondrug label instructions. J Gen Intern Med 2009;24:57–62.

51. Gazmararian J, Jacobson KL, Pan Y, et al. Ef-fect of pharmacy-based health literacy interven-tion and patient characteristics on medication refilladherence in an urban health system. Ann Phar-macother 2010;44:80–7.

52. U.S. Department of Health and Human Ser-vices, Office of Disease Prevention and HealthPromotion. National Action Plan to ImproveHealth Literacy. Washington, DC: US Departmentof Health and Human Services, 2010.

53. U.S. Food and Drug Administration. LabelingDevelopment Team. 2015. Available at: http://www.fda.gov/AboutFDA/CentersOffices/Officeof

MedicalProductsandTobacco/CDER/ucm443026.htm. Accessed November 22, 2016.

54. U.S. Department of Health and Human Ser-vices, Food and Drug Administration, Center forDrug Evaluation and Research (CDER), Center forBiologics Evaluation and Research (CBER). Patientcounseling information section of labeling forhuman prescription drugs and biological products–content and format. Guidance for industry.December 2014. Available at: http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM368602.pdf.Accessed November 22, 2016.

55. Electronic Code of Federal Regulations. Title21. Chapter I, Subchapter C, Part 208. Medicationguides for prescription drug products. Available at:http://www.ecfr.gov/cgi-bin/text-idx?SID¼a4598f2153bbff076bd49ee19fbbe5eb&mc¼true&node¼pt21.4.208&rgn¼div5. Accessed December 4,2016.

56. Wolf MS, Davis TC, Osborn CY, et al. Literacy,self-efficacy, and HIV medication adherence. Pa-tient Educ Couns 2007;65:253–60.

57. US Food and Drug Administration. Medicationguide—Effient (prasugrel) tablets. Available at:http://www.fda.gov/downloads/Drugs/DrugSafety/UCM238428.pdf. Accessed March 4, 2016.

58. Pearsall BM, Araojo R, Hinton D. Essentialmedication information for patients: Ensuring ac-cess. Ther Innov Regul Sci 2014;48:162–4.

59. Kish-Doto J, Scales M, Eguino-Medina P, et al.Preferences for patient medication information:what do patients want? J Health Commun 2014;19Suppl 2:77–88.

60. U.S. Food and Drug Administration. Cardio-vascular and Endocrine Liaison Program (CELP).2015. Available at: http://www.fda.gov/forhealthprofessionals/liaisonactivities/ucm402222.htm. Accessed November 22, 2016.

61. Food and Drug Administration, Center for DrugEvaluation and Research. Summary minutes of thecardiovascular and renal drugs advisory committeemeeting, September 10, 2014. Available at: http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/CardiovascularandRenalDrugsAdvisoryCommittee/UCM456600.pdf.Accessed November 22, 2016.

62. Thom S, Poulter N, Field J, et al., for theUMPIRE Collaborative Group. Effects of a fixed-dose combination strategy on adherence and riskfactors in patients with or at high risk of CVD: theUMPIRE Randomized Clinical Trial. JAMA 2013;310:918–29.

63. Castellano JM, Sanz G, Peñalvo JL, et al.A polypill strategy to improve adherence: resultsfrom the FOCUS project. J Am Coll Cardiol 2014;64:2071–82.

64. Yusuf S, Lonn E, Pais P, et al., for the HOPE-3Investigators. Blood-pressure and cholesterollowering in persons without cardiovascular dis-ease. N Engl J Med 2016;374:2032–43.

65. Figge HL. Electronic tools to measure andenhance medication adherence. US Pharm 2010;36. 4 Compliance & Adherence Suppl:6–10.

J A C C V O L . 6 9 , N O . 4 , 2 0 1 7 Ferdinand et al.J A N U A R Y 3 1 , 2 0 1 7 : 4 3 7 – 5 1 Improving Medication Adherence in CVD

451

66. Urquhart J. The electronic medication eventmonitor. Lessons for pharmacotherapy. ClinPharmacokinet 1997;32:345–56.

67. Dolan B. Wireless Medication Adherence StudyConducted at the Partners Center for ConnectedHealth Shows Promising Initial Findings [pressrelease]. Boston, MA: Center for ConnectedHealth, 2010.

68. U.S. Food and Drug Administration. FDASIAHealth IT Report. Proposed strategy and recom-mendations for a risk-based framework. April2014. Available at: http://www.fda.gov/downloads/aboutfda/centersoffices/officeofmedicalproductsandtobacco/cdrh/cdrhreports/ucm391521.pdf. Accessed November 22, 2016.

69. Greenhalgh T, Howick J, Maskrey N, for theEvidence Based Medicine Renaissance Group. Evi-dence based medicine: a movement in crisis? BMJ2014;348:g3725.

70. Nielsen SF, Nordestgaard BG. Negative statin-related news stories decrease statin persistenceand increase myocardial infarction and cardiovas-cular mortality: a nationwide prospective cohortstudy. Eur Heart J 2016;37:908–16.

71. Agency for Healthcare Research and Quality.Medication adherence. Available at: http://healthit.ahrq.gov/ahrq-funded-projects/emerging-lessons/medication-adherence. Accessed November22, 2016.

72. Pack QR, Squires RW, Lopez-Jimenez F, et al.The current and potential capacity for cardiacrehabilitation utilization in the United States.J Cardiopulm Rehabil Prev 2014;34:318–26.

73. Grace S, Midence L, Brister S, et al. Cardiacrehabilitation program adherence and functionalcapacity among women: a randomized controlledtrial. Mayo Clinic Proc 2016;91:140–8.

74. Prince DZ, Sobolev M, Gao J, et al. Racialdisparities in cardiac rehabilitation initiation andthe effect on survival. PM R 2014;6:486–92.

75. Patel V, Barker W, Siminerio E. Individuals’Access and Use of Their Online Medical RecordNationwide. ONC Data Brief No 20. Washington,DC: Office of the National Coordinator for HealthInformation Technology, 2014.

76. Jacobs J. Best practices for patient portals.HIMSS.org. 2014. Available at: http://www.himss.org/News/NewsDetail.aspx?ItemNumber¼43127.Accessed November 22, 2016.

77. Sama PR, Eapen ZJ, Weinfurt KP, et al. Anevaluation of mobile health application tools.JMIR MHealth Uhealth 2014;2:e19.

78. Plante TB, Urrea B, MacFarlane ZT, et al.Validation of the Instant Blood Pressure smart-phone app. JAMA Intern Med 2016;176:700–2.

79. Morehouse School of Medicine and EmoryExplore Risk for Heart Disease and Stroke Among

Minorities [press release]. Atlanta, GA: MorehouseSchool of Medicine, September 1, 2015.

80. African American Health Disparity Project–SanFrancisco. Four-year project report. Available at:http://www.sfhip.org/javascript/htmleditor/uploads/AAHD_Four_Year_Report.pdf. AccessedNovember 22, 2016.

81. Victor RG, Ravenell JE, Freeman A, et al.Effectiveness of a barber-based intervention forimproving hypertension control in black men: theBARBER-1 study: a cluster randomized trial. ArchIntern Med 2011;171:342–50.

82. Mueller M, Purnell TS, Mensah GA, et al.Reducing racial and ethnic disparities in hyper-tension prevention and control: what will it take totranslate research into practice and policy? Am JHypertens 2015;28:699–716.

83. National Heart, Lung, and Blood Institute.Testing multi-level interventions to improveblood pressure control in minority racial/ethnic,low socioeconomic status, and/or rural pop-ulations FAQs. 2014. Available at: http://www.nhlbi.nih.gov/research/funding/opportunities/faq-foa-HL-15-021. Accessed November 22, 2016.

KEY WORDS cardiovascular disease,nonadherence, polypill, United States Foodand Drug Administration