Intended for healthcare professionals

Clinical Review State of the Art Review

# Medication adherence in cardiovascular medicine

BMJ 2021; 374 (Published 11 August 2021) Cite this as: BMJ 2021;374:n1493
1. Steven T Simon, assistant professor of medicine-cardiology1,
2. Vinay Kini, assistant professor of medicine-cardiology1,
3. Andrew E Levy, assistant professor of medicine-cardiology1 2,
4. P Michael Ho, professor of medicine-cardiology3 4
1. 1Division of Cardiology, University of Colorado Anschutz Medical Campus, Aurora, CO, USA
2. 2Division of Cardiology, Denver Health Medical Center, Denver, CO, USA
3. 3Division of Cardiology, University of Colorado Anschutz Medical Campus, Aurora, CO, USA
4. 4Cardiology Section, VA Eastern Colorado Health Care System, Aurora, CO, USA
1. Correspondence to S T Simon Steven.Simon{at}cuanschutz.edu

## Abstract

Cardiovascular disease is the leading cause of death globally. While pharmacological advancements have improved the morbidity and mortality associated with cardiovascular disease, non-adherence to prescribed treatment remains a significant barrier to improved patient outcomes. A variety of strategies to improve medication adherence have been tested in clinical trials, and include the following categories: improving patient education, implementing medication reminders, testing cognitive behavioral interventions, reducing medication costs, utilizing healthcare team members, and streamlining medication dosing regimens. In this review, we describe specific trials within each of these categories and highlight the impact of each on medication adherence. We also examine ongoing trials and future lines of inquiry for improving medication adherence in patients with cardiovascular diseases.

## Introduction

Cardiovascular disease (CVD) remains the leading cause of death globally.1 Pharmacological treatments can substantially reduce CVD morbidity and mortality; however, the effectiveness of these interventions is limited in cases of medication non-adherence and early discontinuation. While significant resources have been dedicated to the development of novel pharmaceutical agents and devices aimed to improve cardiovascular care, comparatively minimal resources have been aimed at improving patient adherence to these therapies.

Medication adherence has been defined as the extent to which a patient takes treatments as prescribed by their healthcare providers, and is often quantified as the percentage of prescribed treatments that are taken by the patient.234 The definitions and measurements of medication adherence are non-uniform across trials, which often makes it difficult to compare results. The term “medication adherence,” is preferred to “medication compliance,” because of the judgmental implications when referring to a patient as non-compliant.56 Non-adherence can be from “non-initiation,” meaning failure to start a new prescription, “non-persistence,” which is premature discontinuation of a medication, or the patient might not adhere to the daily routine of taking medication(s) as scheduled.37 Measurements of adherence can be direct, including observed administration or measuring the blood concentration of a metabolite; or they can be indirect, which includes patient self-report, pill counting, pharmacy refill rates, and electronic monitoring systems.8 Direct methods are more accurate, but require substantial time and expense to administer. Indirect methods are easier to evaluate but are often less reliable, with self-reported adherence especially prone to bias.9101112 Ultimately, the choice of measurement technique should be suited to the behavior change targeted by the intervention.7

This review presents data from recent clinical trial interventions aimed at improving medication adherence in patients with cardiovascular diseases. Of the 61 trials included in this review, the most common methods of adherence measurement were pharmacy refills (24 trials), self-report with clinical biomarker (21 trials), and electronic drug monitors (10 trials). We discuss the prevalence of non-adherence, common barriers to medication adherence, recent clinical trial data, and finally, discuss ongoing National Institutes of Health (NIH) clinical trials and the major society guidelines in this area.

## Prevalence

Medication non-adherence is pervasive globally. In developed countries, studies have found non-adherence rates of 50% for chronic conditions and this is likely higher in developing nations that have fewer healthcare resources.41314 The overall impact of medication non-adherence is large, estimated to account for approximately 125 000 deaths, 10% of hospitalizations, and $100-500 billion of healthcare services in the US annually.41516171819202122 Among patients with cardiovascular disease, medication non-adherence is extensive and the clinical consequences are significant, especially in the period immediately following an acute cardiovascular event. A cohort study in Ontario found that 27% of patients did not fill discharge prescriptions one week after hospitalization for acute coronary syndrome (ACS), and these patients had increased 1 year mortality.23 At one month following ACS, 34% of patients discharged from hospitals in the US had stopped taking at least one of aspirin, β-blocker, or statin therapy.24 When looking at 1-2 years following ACS, non-persistence reaches 55-60%, with further reductions seen over 10 year follow-up.252627 Improved medication adherence following acute myocardial infarction (AMI) has been associated with a reduction in major vascular events and revascularization at one year as well as lower medical costs, major adverse cardiovascular events (MACE), and mortality.242829 Non-adherence to dual antiplatelet therapy following AMI has been shown to increase the risk of stent thrombosis30 and mortality at one year.31323334 High rates of non-adherence persist despite the growing research and resource efforts directed toward this problem. ### Sources and selection criteria We performed a comprehensive literature search for studies evaluating interventions to improve medication adherence in patients with cardiovascular disease. Using PubMed and Embase, we identified and reviewed randomized controlled trials (RCTs), controlled clinical trials, and meta-analyses from 1 January 2010 through 5 August 2020. MESH search terms used were “medication adherence”, “medication compliance” , and “cardiovascular disease”. We selected additional studies from references of key papers and review articles. We limited the results to those in the English language, performed in adults, with at least 50 participants, and follow-up period of greater than six months. We did not include studies that used self-report as the only measure of medication adherence, without an associated clinical biomarker. Trials selected for inclusion were mutually agreed on by the four authors, and disagreements were resolved through discussion. At least two authors reviewed each manuscript included in the review. We also reviewed the most recent guidelines and consensus statements from the American College of Cardiology/American Heart Association (ACC/AHA) and the European Society of Cardiology (ESC). ## Barriers to medication adherence Understanding the factors that affect adherence to medications is critical in the evaluation of interventional strategies. These factors have been broadly grouped into five categories by the World Health Organization: patient related factors, socioeconomic factors, health team and system factors, therapy factors, and condition related factors.14 Patient factors—A patient’s personal beliefs regarding their illness, such as the belief that medications are unnecessary to control their condition, or that medications will cause side effects, influence their adherence to prescribed treatment plans.353637 An individual’s health literacy, cognitive ability, and degree of forgetfulness, have each separately correlated with medication adherence, with forgetfulness having previously been attributed to more than one third of all cases of non-adherence.383940414243 These patient-level factors are clinically under-recognized but are important causes of non-adherence. Socioeconomic influences—Individuals categorized as having a lower socioeconomic status (SES) have been associated with lower medication adherence, as seen in a systematic review which finds this relationship among patients with a first AMI.4445 This is partly a result of medication cost, with one in eight patients with CVD reporting cost related non-adherence.46 Beyond income, low SES has been associated with decreased access to providers and pharmacies, as well as less coverage for prescription medications.474849 Socioeconomic factors have a profound influence on health outcomes, and poor medication adherence is another mechanism by which lower SES contributes to adverse health outcomes.50 Structure of the healthcare system—This can negatively influence a patient’s adherence to medications. Poor communication between inpatient and outpatient care teams, between specialists in different health systems, and between providers and pharmacists, often leads to poor communication with patients, ultimately leading to non-persistence owing to therapy confusion. In the US, most adult patients rely on private health insurance and are required to pay prescription copayments, the cost of which has been shown to contribute to medication non-persistence.5152 Prescribed therapies and the conditions for which they are prescribed—Many medications have unwanted side effects, require additional monitoring, and serve as a consistent reminder of the patient’s illness, all factors that might reduce patient persistence.53 Further, complicated dosing regimens can lead to inconvenient administration times and contribute to forgetting to take medications. Individuals with multiple medical conditions or conditions that require a large pill burden must adhere to complex regimens and may experience medication interactions and polypharmacy leading to non-adherence. The individual causes of a patient not adhering to prescribed medical therapy are not uniform, and have multiple contributing and intertwined factors.43 Grouping them into patient related factors, socioeconomic factors, health team and system factors, and therapy and condition related factors helps to organize approaches to improve adherence. Each of the interventions discussed in this review targets at least one of these domains, and often tries to address multiple factors to achieve sustained improvements in adherence (fig 1). Fig 1 Barriers to medication adherence and targeted interventions ## Results ### Interventions for medication non-adherence Considering the morbidity and mortality as well as the financial costs associated with medication non-adherence, strategies to address it have received increasing attention. Details of the search criteria for studies included in this review are in the box “Sources and selection criteria.” We identified 440 manuscripts by the search criteria, and found an additional five through additional resources.5455565758 Of these 445 records, 61 met the final criteria for inclusion (fig 2). The clinical trials included in this review (table 1) were classified into six distinct categories of intervention: patient education, technological reminders, cognitive behavioral and motivational interventions, medication cost interventions, the use of healthcare team members, and using fixed dose combination therapy regimens. Studies that use at least three of these categories without one dominant strategy are discussed separately as multifaceted interventions. This categorization provides a useful framework for discussion but is not absolute, as many of these interventions fall into multiple categories or do not fall exactly into any one of the categories. Fig 2 Flow diagram of manuscript selection Table 1 Key trials by category of intervention View this table: ## Educational interventions Improving a patient’s health education targets the barriers to adherence discussed above. Interventions focused on this aspect posit that if patients fully understand the extent of their illness and the benefits of the prescribed therapy, they would be more engaged in the therapy and therefore more likely to be adherent. The format and frequency with which educational material is delivered may have an impact on its effectiveness. Providing educational mailings did not improve persistence to secondary prevention medications at 1 year in an RCT of 852 patients who experienced an ST-elevation myocardial infarction (58.4% intervention versus 58.9% control; odds ratio (OR) 1.03; 95% confidence interval (CI) 0.77 to 1.36).120 Similarly, in the ISLAND trial, 2632 patients with obstructive coronary artery disease were randomized to receive usual care, five educational mail-outs, or the mail-outs plus interactive phone calls aimed at increasing adherence to cardiac rehabilitation and medications. Improvements were seen in completion of cardiac rehabilitation (27% in the usual care group, 32% in the mail only group, and 37% in the mail plus phone call group; adjusted OR 1.55, 95% CI 1.18 to 2.03); however, no significant difference was seen in adherence to secondary prevention medications (mail only versus usual care, OR 0.98, 95% CI 0.81 to 1.19; full intervention versus usual care, 0.99, 0.82 to 1.20).121 In another RCT providing written educational material, 1162 patients with non-valvular atrial fibrillation qualifying for oral anticoagulation showed no significant differences in medication adherence at 24 weeks (91.6% ±17.1 for standard of care and 91.9% ±16.1 for intervention; P>0.7) when provided with medication education in the form of information booklets, reminder tools, and access to a virtual clinic.54 Providing tailored educational information in a more interactive manner with repeated reinforcement may be more beneficial for improving adherence than single episode, universal information. Cardiac rehabilitation visits, which provide intensive in-person education, have been associated with improved medication adherence.122 Earlier and more frequent visits with cardiologists following percutaneous coronary intervention and acute myocardial infarction have been associated with higher adherence to cardiovascular medications.49123124 A trial of 694 patients with ACS found intensive follow-up care of telephone calls and regular consultations with cardiologists improved medication adherence (58% in the intervention group versus 40% in the control group (P<0.001)) and reduced major adverse cardiovascular events (19% in the intervention group versus 29% in the control group (P<0.001)) at 36 months follow-up.125 An RCT examining whether education delivered in person by nurses could improve statin adherence in 201 patients found significantly higher adherence to statin therapy in the intervention arm (P<0.01) with also significantly lower LDL cholesterol levels in primary prevention patients (2.66 versus 3.00 mmol/L, P=0.024).126 A randomized trial of 385 moderate to high risk patients with coronary heart disease showed improved adherence to medication when education was delivered in person by a counselor (21% improvement in medication adherence) or delivered in a web based format (18% improvement in medication adherence).127 These frequent healthcare contacts may not need to be in person. Both text messaging and structured educational phone calls from a nurse improved medication adherence (78.9% versus 81.4% versus 69.5%, P=0.011) and 180 day all-cause mortality or readmission (50.4% versus 41.3% versus 36.5%, both P<0.05) in a randomized trial of 767 patients recently admitted for acute heart failure.128 Similarly, an RCT of 1372 patients taking medications for hypertension found improved 12 month medication adherence among patients randomized to information-only text messages and interactive text messages compared with usual care (62.8% information-only, 59.7% interactive message group, and 49.4% usual care group; informative messages versus usual care, P<0.001; interactive messages versus usual care, P=0.002).129 An RCT of 300 patients with recent drug eluting stent placement showed that a series of four educational telephone calls made by nurses over nine months could improve persistence of taking dual antiplatelet therapy at 12 months (87.2% intervention compared with 43.1% control (P<0.001)).130 Improvements in understanding a condition may not always translate to improved adherence as was found in a trial of 794 hypertension patients randomized to usual care, electronic health record (EHR) based medication tools aimed at proper medication reconciliation, or EHR tools plus nurse led medication education. Following 12 months of intervention, understanding of hypertension medication instructions and dosing was greater in the EHR plus education group than the usual care group (OR, 2.3; 95% CI 1.1 to 4.8); however, this did not lead to improvements in adherence to antihypertensive medications (OR, 0.9; 95% CI 0.6-1.4).131 Improving a patient’s understanding of their medical condition and treatment would seem to be an effective strategy for improving medication adherence. However, recent trials have shown that improved education may not lead to improved adherence and that the mode of educational delivery may influence its impact. Use of the mobile phone is an easier method to enable frequent contact with patients, and this reinforcement may be more effective in improving medication adherence. ## Technology reminder interventions Even if patients are fully educated, motivated, and supported by the health system, remembering to take medications every day or multiple times a day can be difficult. Therefore, many trials have focused on interventions that remind patients to take their medications on schedule. Traditional pill boxes are useful in improving adherence, but more recent trials have relied on technology to expand the reminder toolkit. Additional technologies have been studied for their ability to improve adherence, including watches with alarms, pill box timers, phone apps, and “smart pill” containers. In a trial of 53 480 non-adhering patients, participants were randomized to receive a pill bottle strip with toggles, a pill bottle with a digital timer cap, a standard pill box, or usual care. Optimal medication adherence at 12 months was not significantly different between the control and any of the devices (standard pill box OR 1.03, 95% CI 0.95 to 1.13; digital timer cap OR 1.00, 95% CI 0.92 to 1.09; pill bottle strip with toggles OR 0.94, 95% CI 0.85 to 1.04).55 The near universal ownership of mobile phones has made them a natural choice of reminder tool. Telephone reminder calls, text message reminders, and mobile phone applications (apps) have all been examined for their capacity to improve medication adherence. A trial randomized 5216 patients who did not fill a prescription for a new statin to receive automated telephone calls followed by letters for continued non-adherence, versus no reminder. Ultimately, those who received outreach had significantly increased rates of filling their medication (42.3% intervention compared with 26.0% control; absolute difference 16.3%, P<0.001; relative risk (RR) 1.63, 95% CI 1.50 to 1.76).56 Similar improvements in medication adherence have been shown for automated phone call reminders aimed at patients with diabetes and atherosclerosis.132 The advent of text messaging provided a new way to communicate quickly with patients, and recent meta-analyses have found trials utilizing text message reminders to improve medication adherence both in chronic disease and cardiovascular disease specifically.133134 Mobile phone applications used as reminder tools have also been successful. In the MediSAFE-BP trial, 411 patients with uncontrolled hypertension were randomized to usual care or to download the Medisafe app, which provides reminder alerts, adherence reports, and optional peer support. Medication adherence, as measured by the self-reported Morisky medication adherence scale (MMAS) was significantly improved in the application group (between-group difference 0.4; 95% CI 0.1 to 0.7; P=0.01). This difference in adherence did not translate to a significant difference in blood pressure control between the groups (between-group difference: -0.5; 95% CI -3.7 to 2.7; P=0.78).135 An RCT exploring whether an advanced interactive app could have a greater impact on adherence than a basic app randomized 163 patients with coronary heart disease to these interventions. At 3 months, patients using either app had higher self-reported adherence compared with the usual care group (mean MMAS-8 score 7.11 app versus 6.63 usual care; mean difference 0.47, 95% CI 0.12 to 0.82, P=0.008) with no significant difference in adherence between the basic and advanced app (mean difference -0.16, 95% CI -0.56 to 0.24, P=0.428), and no significant difference in secondary clinical outcome measures.136 Forgetting to take medications is a significant cause of medication non-adherence and recent technologies have provided a cheap and relatively simple way to provide reminders. Electronic pill boxes alone may have limited utility; but the use of mobile phone calls, text messages, and applications can improve adherence, albeit without demonstrable clinical outcome improvement. Use of these techniques should be tailored to those patients who have difficulty remembering to take medications and are comfortable navigating the technology. ## Cognitive behavioral and motivational interventions Some patients are less motivated to take medications regularly, and changing this behavior can be challenging. Motivational interviewing, behavioral feedback, and social support have been explored as methods to collaboratively change behaviors surrounding medication adherence. Motivational interviewing is a counseling style that aims to evoke an individual’s own motivation for changing a behavior.137 A meta-analysis of studies examining the impact of motivational interviewing showed a small increase in medication adherence (pooled risk ratio for adherence 1.13, 95% CI 1.01 to 1.28).138 Within cardiovascular disease, four sessions of motivational interviewing were evaluated as a method to improve clinical outcomes and medication adherence in 386 patients with recent stroke. At the end of the 12 months, no significant change was seen in blood pressure (mean difference in change, -0.2.35, 95% confidence interval -6.16 to 1.47) or cholesterol (mean difference in change -0.0.12, 95% confidence interval -0.30 to 0.06); although self-reported medication adherence was improved at 9 months (4.295; 95% CI 1.56 to 11.84; P=0.005).139 Another RCT found that having physicians count pills during visits and use motivational interviewing improved rates of uncontrolled systolic blood pressure (OR 0.62; 95% CI 0.50 to 0.78) and increased adherence (OR 1.91; 95% CI 1.19 to 3.05) when compared with control group patients.140 Other counseling techniques explored include positive-affect induction and self-affirmation, which focus on having patients identify moments that make them happy and proud. In an RCT of 256 African American patients with hypertension, use of these techniques slightly improved medication adherence at 12 months (42% versus 36%; P=0.049) without a significant change in blood pressure.141 Providing patient feedback regarding their medication adherence has shown short term improvements in medication adherence without a sustained effect.142 In the PROMOTE randomized controlled trial, 201 adults with diabetes and a history of low adherence to statins were randomized to receive messages comparing their own statin adherence with that of other participants in the study, messages concerning only their own adherence, or no messages. No significant improvement was seen in adherence from the social support compared with control.143 Improvements in medication adherence have been seen when family members are involved to provide social support.144 Utilizing supportive methods to influence patient motivation and behaviors has not consistently shown significant improvements in adherence outcomes. These interventions should be targeted at those patients whose motivation is the primary barrier to medication adherence. ## Cost of medication and financial incentive interventions The cost to obtain medications can act as a major barrier to adherence and persistence, especially in healthcare systems that have a higher patient cost burden.4145 Interventions that reduce the cost of medications or provide financial incentives for taking medications have been explored as methods to improve adherence. The MI FREEE trial randomized 5855 patients hospitalized for AMI to full prescription coverage versus usual coverage for β blockers, statins, and angiotensin converting enzyme inhibitors (ACE-Is) or angiotensin receptor blockers (ARBs). Rates of adherence were increased in the full coverage group compared with the usual coverage group by 5.6 percentage points for ACE-I or ARBs (95% CI 3.4 to 7.7), by 4.4 percentage points for β blockers (95% CI 2.3 to 6.5), by 6.2 percentage points for statins (95% CI 3.9 to 8.5), and by 5.4 percentage points for all three medication groups (95% CI 3.6 to 7.2). No significant differences were seen in the primary clinical outcome, defined as first major vascular event or revascularization (17.6 per 100 person years in the full coverage group versus 18.8 in the usual coverage group; hazard ratio (HR) 0.93; 95% CI 0.82 to 1.04; P=0.21), but a significant reduction was seen in total major vascular events or revascularization in the full coverage group (21.5 versus 23.3; HR 0.89; 95% CI 0.90 to 0.99; P=0.03). Importantly, the elimination of copayments did not increase total spending ($66 008 for the full coverage group and $71 778 for the usual coverage group; relative spending 0.89; 95% CI 0.50 to 1.56; P=0.68).28 A similar result was found in the ARTEMIS trial, which followed 10 102 patients with AMI for one year following hospitalization, and randomized participants to an intervention of copayment vouchers for P2Y12 inhibitors for one year, or no vouchers. At 1 year, patient reported persistence was higher in the intervention group than in the control group (unadjusted rates, 5340/6135 (87.0%) versus 3324/3967 (83.8%), respectively; P<0.001; adjusted odds ratio 1.19, 95% CI 1.02 to 1.40); however, no significant difference was seen in major adverse cardiovascular events (10.2% versus 10.6%; P=0.65; adjusted HR 1.07, 95% CI 0.93 to 1.25).146 The evidence suggests that easing out-of-pocket costs for medications can improve patient adherence, although these improvements do not clearly translate to better clinical outcomes. These trials were performed in the US and may be less applicable to healthcare systems that have fewer medication cost burdens for patients. Rather than reducing upfront medication costs, financial incentives can be utilized to encourage adherence. The CHORD trial147 randomized 337 patients with poorly controlled hypertension to be paid$8 per medication per month for filling antihypertensive prescriptions, a computerized behavioral intervention (CBI), both payment and CBI, or usual care. At 12 months, no significant changes were seen in medication adherence or systolic blood pressure (SBP) improvement between the groups (SBP difference of -3.7 mm Hg; 95% CI -9.0 to 1.6; P=0.17). Recognizing that patients are not the only parties that can be influenced by financial incentives, a four group, cluster RCT148 was conducted to study the impact of financial incentives for patients, providers, or both, on medication adherence. A total of 340 primary care physicians and 1503 patients with elevated cardiovascular disease risk were randomized to receive financial incentives for taking their prescribed statin medications. Physicians were awarded monetary incentives per enrolled patient meeting goals for low density lipoprotein cholesterol (LDL-C) and patient incentives were distributed through daily lotteries tied to medication adherence. At 12 months, patients in the shared physician-patient incentives group achieved a significantly lower reduction in LDL-C than the control group (33.6 mg/dL improvement incentive group versus 25.1 mg/dL in the control group; difference 8.5 mg/dL; 95% CI 3.8 to 13.3; P=0.002). No significant differences were seen in LDL-C improvement between the physician-only incentive group and the patient-only incentive group compared with the control group. Only patients in the shared physician-patient incentives group had significantly higher mean medication adherence compared with control (39% CI 32% to 44%, versus 27% CI 23% to 31%; P<0.001).

Cost reduction strategies using either copayment reduction or financial incentives have shown only modest changes in medication adherence when applied only to the patient. Whether short term incentives can form habits that persist when the incentive is removed needs to be further examined to determine the sustainability of these interventions.

## Healthcare team members

Expanding the healthcare team role in reinforcing medication adherence can also help to target the difficulty patients experience in navigating the healthcare system. For the purposes of this review, we defined interventions using healthcare team members as those where the primary intervention involved expanded roles for non-prescribing healthcare team members. Pharmacists often interact regularly with patients and provide access, education, and support to patients taking medications. Naturally, enhanced pharmacist involvement has been explored as a means to improve medication adherence. An intervention trial for patients with hypertension, combining pharmacist intervention with motivational interviewing and telephone follow-ups, found 1 year reductions in the percentage of patients who did not adhere (20.3% of the patients in the intervention group compared with 30.2% in the control group, risk difference -9.8; 95% CI -17.3 to -2.4, P=0.02), but with no significant differences in clinical endpoints of blood pressure control or hospital admission.149 Other studies evaluating the use of pharmacists to engage in additional medication support have found modest or no improvements in adherence.150151

Similarly, community health workers have been employed to act as community connections to the healthcare system. An RCT of 806 patients with recent ACS found that an intervention of phone calls and visits by community health workers which focused on healthy lifestyle education led to higher adherence at 1 year to all secondary prevention treatments among the intervention group (97% versus 92%, odds ratio 2.62; P=0.006), although adherence rates in this trial for both groups exceeded most reported rates.152

Healthcare team members can be utilized to increase interaction with patients in the community, but these are generally resource and time intensive interventions that require substantial investments from health systems. Trials have generally shown modest improvements in medication adherence and clinical outcomes with these interventions.

## Fixed dose combination therapy

Polypharmacy and complicated dosing regimens contribute to poor medication adherence, and polypills have been employed to address this barrier. Retrospective analysis has shown that initiating antihypertensive agents as a fixed dose combination (FDC) of multiple classes, rather than initiating multidrug therapy, is associated with improved adherence and reduced cardiovascular outcomes.153 Blood pressure control was also improved in an RCT of 700 hypertensive patients initiating or escalating medications as an FDC versus usual care (risk difference 12.7%, 95% CI 3.2% to 22.0%; P<0.001).154 Hypertension treatment clearly benefits from the use of FDC therapy and has become standard initial therapy for patients with uncontrolled hypertension.

Other RCTs have combined therapies across multiple pharmaceutical categories to target CVD prevention. These have consistently shown improved medication adherence with FDC therapy, although improvements in clinical endpoints are less consistent.58155 The UMPIRE trial was an RCT among 2004 patients with CVD or at risk of CVD randomized to FDC therapy of aspirin, simvastatin, lisinopril, and atenolol or hydrochlorothiazide compared with usual care. At a median follow-up of 15 months, the FDC group had improved adherence versus usual care (86% versus 65%; relative risk (RR) of being adherent 1.33; 95% CI 1.26 to 1.41; P<0.001) and experienced improvements in clinical outcomes including reductions in SBP (-2.6 mm Hg; 95% CI -4.0 to -1.1 mm Hg; P<0.001) and LDL-C (-4.2 mg/dL; 95% CI -6.6 to -1.9 mg/dL; P<0.001).156 An RCT involving 303 adults with intermediate atherosclerotic cardiovascular disease risk and uncontrolled hypertension found that a four drug polypill (atorvastatin 10 mg, amlodipine 2.5 mg, losartan 25 mg, and hydrochlorothiazide 12.5 mg), when compared with usual therapy, led to a reduction in SBP (difference -7 mm Hg; 95% CI -12 to -2; P=0.003) and mean LDL-C (difference -11 mg/dL; 95% CI -18 to -5; P<0.001) at 12 months. Adherence to the polypill regimen, as assessed on the basis of pill counts, was 86% at 12 months.157 The intervention phase of the FOCUS study enrolled 695 patients after AMI and randomized participants to receive a polypill containing aspirin, simvastatin, and an ACE inhibitor, compared with giving the three drugs separately. After nine months of follow-up, the group randomized to polypill showed improved medication adherence (intention-to-treat 50.8% versus 41%; P=0.019). No difference was seen in mean systolic blood pressure, low density lipoprotein cholesterol levels, serious adverse events, or death.158

As the number of recommended medications to treat cardiovascular conditions continues to expand, interventions like FDC therapy provide an opportunity to improve adherence and clinical outcomes. The evidence suggests a substantial improvement in clinical outcomes can be obtained with this strategy. While there are difficulties with matching patients to a given polypill and changing doses of individual elements of a polypill, this is an intervention that can be sustained long term without substantial cost or effort in the appropriately selected patient.

## Multifaceted interventions

Many patients have multiple barriers to medication adherence, making a single focus of intervention less likely to be effective.43 A few trials have implemented multifaceted strategies, here defined as employing at least three different categories of intervention simultaneously without any one dominant strategy. A randomized control trial of 253 patients who were hospitalized for ACS tested a multifaceted strategy in which the intervention group received pharmacist led medication reconciliation and tailoring, patient education, collaborative care between pharmacist and a patient’s clinician, and educational and refill reminder telephone voice messages. This strategy resulted in a greater proportion of adherent patients to four secondary prevention medications compared with usual care (89.3% versus 73.9%; P=0.003) and a greater proportion of days covered (0.94 versus 0.87; P<0.001). However, no statistically significant differences in blood pressure and lipid goals were seen between the groups.57 A separate trial randomized 1509 patients with AMI to a combination of electronic pill bottles, lottery financial incentives, and social support. At 12 months, no statistically significant differences were seen in time to first re-hospitalization for a vascular event or death (HR 1.04; 95% CI 0.71 to 1.52; P=0.84), and no difference in mean medication adherence (difference 0.04; 95% CI -0.01 to 0.09; P=0.10).159 A cluster randomized clinical trial combining behavioral interviewing led by pharmacists, educational tools, and text message based reminders for 4078 non-adhering patients with hypertension, hyperlipidemia, or diabetes found the combined interventions improved adherence by 4.7% (95% CI 3.0% to 6.4%) with no improvement in the clinical endpoints.160 Strategies combining multiple interventions have generally managed to improve measured adherence; however, a detectable improvement in clinical outcomes has not been seen.

## Emerging treatments

Medication adherence in cardiovascular disease continues to be an active area of research and many clinical trials are ongoing within each of the intervention categories detailed above. We performed a search of clinicaltrials.gov for active, enrolling, recruiting, or recently completed interventional studies with an outcome of medication adherence utilizing MESH terms “cardiovascular diseases.” Studies were limited to those using an objective criterion for medication adherence, adult populations, and those with more than 60 participants. This yielded 22 studies, which are presented in table 2. These interventions target multiple domains including education,1 technology,6 behavioral,8 healthcare delivery system,5 and healthcare team members.2

Table 2

Ongoing clinical trials targeting medication adherence in cardiovascular disease

View this table:

Within behavioral interventions, trials are continuing to explore the concept of financial incentives. Novel methods include pairing these incentives with smartphones (NCT04114669, NCT04075045), and adding social support as well as daily monitoring of vital signs (NCT02708654). Additional interest focuses on the use of community health workers to help navigate the healthcare system (NCT02674464, NCT03077386). Several trials are investigating the use of smartphone applications to serve as reminder tools and educational sources, with some incorporating the monitoring of vital signs and the use of provider alerts. One trial will explore the use of a smartphone application which connects with a home blood pressure monitoring device to provide feedback and reminders regarding medications, weight, and blood pressure (NCT03288142). Another will pair a smartphone application with an Apple iWatch and a blood pressure cuff to track vitals and provide medication reminders (NCT03760796). Given the ubiquitous use of personal phones, they have become an obvious vehicle for interventions to improve adherence. Use of cellular phone applications will continue to receive research attention as they can serve as tools for many of the interventions described, including delivering educational material, reminding patients to take medications, providing behavioral feedback, and helping patients to navigate the healthcare system, all from their pockets.

## Guidelines

The importance of medication adherence has been incorporated into recent guidelines from ACC, AHA, and ESC. As the study of medication adherence has progressed, the guidelines have become more specific in their recommendations. The 2013 ACC/AHA guideline for the management of heart failure makes a class I recommendation for reinforcing the need for medication adherence throughout hospitalization, before hospital discharge, and at follow-up visits.161 Similarly, the 2014 ACC/AHA guidelines regarding stable ischemic heart disease and non-ST-elevation acute coronary syndrome make class I recommendations for patients with these conditions to have treatment plans promoting medication adherence and suggest follow-up to assess adherence.162163

## Conclusions

Medication non-adherence in cardiovascular medicine is a preventable cause of substantial morbidity and mortality. Improvements in medication adherence that translate to improved clinical outcomes could lead to significant reductions in healthcare costs.20 However, the relative resources dedicated to improving adherence are minimal compared with those dedicated to introducing new therapies. This may be in part a result of misaligned incentives within the healthcare system, which tend to reward new treatments over the encouraging adherence to existing ones.

The trials presented in this review find heterogeneous effects within each category of intervention. Often, an intervention shows an improvement in measured adherence without associated clinical outcome improvement. To help overcome this, future clinical trials should use a standardized approach for measuring adherence, avoid using self-report as the sole method for assessing adherence, and evaluate clinical outcomes alongside adherence measurements. Incongruity between a patient’s individual adherence barrier and the intervention applied also contributes to the varied results of these trials. Clinicians and investigators should assess for these common barriers to adherence, and interventions should then be tailored to overcome these barriers (fig 2).

Individual trials in each of the intervention categories outlined in this review have shown success; however, some categories appear more effective than others. Trials utilizing fixed dose combination therapy consistently showed improved measures of adherence and clinical outcomes. These interventions are limited to individuals who require multiple medications, but are likely sustainable at a relatively low cost for those patients.

Providing reminders via smartphones as either an application or text message appear to consistently improve adherence, with select trials additionally showing improvements in clinical outcomes. These tools are relatively easy to implement and sustain, but are only helpful if a patient’s non-adherence is the result of forgetfulness. Similarly, interventions that focused on patient education often improved adherence with occasional clinical improvement. The trials that reinforced the material with repeat contacts seemed to be the most effective, but are also more costly and time intensive. These interventions will be most effective for patients who have a lower comprehension of their medical condition or medications.

Reducing medication cost to patients through copayment reduction or financial incentives appears to improve measured adherence without a demonstrable clinical improvement. These can be expensive interventions requiring significant support from health insurance providers, and are best targeted at patients with cost barriers.

Finally, the cognitive and motivational behavioral interventions and interventions using pharmacists and community health workers to individually work with patients can be costly in time commitment and do not consistently show significant improvements in adherence and clinical measures. Individuals lacking motivation to adhere to therapy may benefit most from behavioral interventions, while those who have difficulty navigating the healthcare system are most likely to benefit from health team member interventions.

The heterogeneity of effect within each category may be a result of imbalanced matching of intervention and barrier to adherence. Providing medication reminders to an individual who cannot afford to fill a prescription, or alternatively, covering copayments for patients not willing to take a medication, are not likely to have significant impact. For a patient to be fully adherent to medications, they need access to prescribers and pharmacies as well as have the finances to obtain the prescription. Further, they need to want to take the medication and remember to take it. Additional research and clinical commitment is needed in personalizing adherence interventions to an individual’s adherence barriers.

### Questions for future research

• Does tailoring specific interventions to identified personal barriers improve adherence to medications?

• What has been the impact of the coronavirus pandemic on medication adherence?

• How do racial and ethnic disparities contribute to medication non-adherence and what interventions can specifically address these disparities?

• How can wearable technology further contribute to identifying and treating medication non-adherence?

• What is the role of depression in medication adherence and how can pharmacological and non-pharmacological (eg, peer group) treatments improve adherence?

• How can health systems be incentivized to implement interventions that have shown improvements in adherence?

• Do interventions that improve adherence reduce overall healthcare spending?

This manuscript was reviewed by a nurse who was the caregiver and medication manager for her husband who had cardiovascular disease. She provided suggestions regarding the content, specifically in the barriers to medication adherence, conclusion, and questions for future research sections.

## Acknowledgments

We would like to thank Johann Chanin for reviewing the manuscript and providing comments from the perspective of a patient.

## Footnotes

• Series explanation: State of the Art Reviews are commissioned on the basis of their relevance to academics and specialists in the US and internationally. For this reason they are written predominantly by US authors

• Competing interests: We have read and understood the BMJ policy on declaration of interests and declare the following interests: STS reports none, AEL reports none, VK reports unrelated research grants from the NIH and Greenwall Foundation, PMH reports research grants from Department of Veterans Affairs, NHLBI, and University of Colorado School of Medicine. His NHLBI grant is focused on medication adherence. He has a research agreement with Bristol Myers Squibb unrelated to medication adherence. He serves as a deputy editor for Circulation: Cardiovascular Quality and Outcomes.

• Contributors and guarantor: STS prepared the manuscript draft, performed literature review, and is the guarantor of the manuscript. AEL and VK contributed to the writing of the manuscript, selection of articles for inclusion, and reviewing the manuscript. PMH constructed the idea and framework of the manuscript, selection of articles for inclusion, contributed to writing of the manuscript, and reviewing the manuscript. STS is guarantor.

• Provenance and peer review: commissioned; externally peer reviewed.

View Abstract