According to data provided by the National Center for Health Statistics (NCHS), life expectancy in the US increased by six years between 1970 and 2000.1,2 Much of this increase can be attributed to reductions in mortality due to cardiovascular (CV) disease (CVD).2 Similarly, the incidence of CVD-related mortality and morbidity has been falling steadily across the EU since the early 1980s, in contrast to some countries in Eastern Europe.3 Despite this reduction in CVD-related mortality rates, CVD remains the main cause of death in Europe today, accounting for more than 4.3 million deaths each year.3 In total, 48% of all deaths reported in Europe (54% in women and 43% in men) are the result of CVD.3 In particular, coronary heart disease (CHD) and stroke are the most common causes of death in Europe, accounting for 1.92 and 1.24 million deaths per year, respectively.3
There are many reasons that rates of CVD-related mortality remain high despite rapid developments in therapeutic strategies. First, diagnosis of CVD is often suboptimal. For example, up to one-third of myocardial infarctions go unrecognised but are associated with a prognosis almost as poor as for individuals surviving symptomatic infarctions.4 In addition, many seemingly healthy individuals may have a risk of CVD as high as those who already have clinically manifest disease, but are not enrolled in primary prevention programmes.5 A second likely reason for high mortality rates is that patients identified with established CVD, or patients identified as being at risk of developing CVD, are not receiving ideal medical management.6 This is illustrated by a recent survey of risk factor management in 3,876 patients with CHD in Europe – European Action on Secondary Prevention through Intervention to Reduce Events (EUROASPIRE III) – which demonstrated that 54% of patients had not achieved target blood pressure (BP) levels and 41% of patients had not achieved target low-density lipoprotein cholesterol (LDL-C) levels.7 The reasons underlying these disappointing data are not fully understood but underprescription of efficacious cardioprotective pharmacological interventions (e.g. statins)8 is likely to be an important factor.
Although current guidelines take into account all major risk factors and strive to target patients who would benefit from therapeutic intervention, considerable discrepancies exist between guideline recommendations and clinical practice, often leading to under-treatment of patients.9 A third important reason for high CVD-related mortality is medication non-adherence, defined as a patient’s failure to follow a prescribed drug regimen. Non-adherence rates to pharmacological therapy is a major issue across various chronic diseases.10 Concerning CVD specifically, the World Health Organization (WHO) describes poor adherence as the single most important cause of uncontrolled, treated hypertension (HTN).11 Indeed, the rate of adherence to antihypertensive medication is estimated to be just over 50% in the US.12 Similarly, a recent meta-analysis of studies reporting patient adherence or persistence with CV medications revealed that 49% were non-adherent to lipid-lowering therapies.13
Clearly, there is a need to understand why patients may not adhere to prescribed treatment interventions and to develop strategies to increase adherence rates, thus improving overall clinical outcomes. In this article, we review the clinical and economic implications of non-adherence focusing on antihypertensive and lipid-lowering medications. We also consider possible strategies that may improve the adherence of patients to treatment.
Reasons for Non-adherence to Medication in Cardiovascular Disease
Reasons for non-adherence to pharmacological therapy can roughly be grouped under three main headings: patient-, physician- and medication-related (see Table 1).
There are a number of direct patient-related reasons for non-adherence to antihypertensive and lipid-lowering medicines.14–22 These include: forgetfulness (which is the primary reason cited by patients for non-adherence);15 a negative attitude towards medication; frustration with poor therapeutic responses; pre-conceived beliefs regarding health and medication; cultural beliefs; lack of education/inadequate literacy skills; and poor language proficiency. In addition, a poor understanding of the cost–benefit of a prescribed drug, including a lack of understanding of the benefit of the medication and a fear of drug-related adverse events, may also contribute to patient non-adherence.
Inadequate follow-up by physicians may compound the problem by sending out the message to patients that adherence to medication is unimportant. Patient-related reasons may be grouped into two simple categories: passive and active non-adherence. Passive non-adherence may include merely forgetting to take medication, perhaps because of a busy work or social life. Active non-adherence, on the other hand, is a conscious decision by the patient to ‘self-medicate’; that is, decide for themselves the dose and frequency of their treatment regimen. In both cases, patients require education regarding the need to follow prescribed treatment regimens as directed.
Currently, conflicting data from different studies have resulted in uncertainty regarding how some demographic factors influence adherence to antihypertensives and lipid-lowering therapies. For example, many studies suggest that adherence rates may be higher in older patients than younger patients,19,23–29 while other data suggest that the reverse may be true.30,31 Similarly, many but not all studies suggest that adherence is lower in women than men.25,29,32–37 However, there is consensus that certain demographic factors are associated with poor adherence to medication. Adherence tends to be lower in patients of non-Caucasian origin (e.g. African-American, Hispanic and Asian- Pacific patients) than in Caucasian patients, resulting in higher rates of uncontrolled, treated HTN or hyperlipidaemia in these patient subgroups.19,25,26,38–45 The presence of one or more non-CV co-morbidities (e.g. depression, respiratory disease, gastrointestinal disorders or osteoarthritis) also appears to negatively influence treatment adherence, particularly in elderly patients.26,37,40,46–48
Poor communication between physician and patient is a significant problem that may influence the adherence of patients to medication.14,25,49,50 As outlined above, one barrier to adherence is the lack of knowledge regarding why a particular treatment regimen has been prescribed and the consequences of not adhering to the intervention. This problem is compounded by the fact that physicians tend to overestimate public understanding of the impact of CV risk factor management on overall health. For example, in a survey that investigated perceptions of CHD from the viewpoint of both physicians and the general public (Reassessing European Attitudes about Cardiovascular Treatment [REACT] study), 92% of physicians believed that their patients were aware that cholesterol is associated with CVD,5 whereas only 51% of the general public knew that high cholesterol was associated with CHD.51
Moreover, only 45% of general public surveyed were aware that CHD was the leading cause of death in their country. These findings suggest that physicians’ perceptions of risk factor management may not reflect clinical reality in their patients. Indeed, in another recent survey in Europe, physicians estimated that over 60% of their hypertensive patients had controlled BP.52 In reality, less than 30% of hypertensive patients in Europe have controlled BP (BP<140/90mmHg for non-diabetic patients; BP<130/80mmHg for diabetic patients) according to a recent study.53
This gap in patient–physician communication may, in part, be due to the increasingly limited time physicians have available to spend with each patient. In the REACT study,5 time-constraints were identified as the largest single barrier to preventative cardiology. This was particularly true in the UK, where physicians reported the highest monthly consultation rates of all the countries surveyed.5
In addition to the communication gaps between patients and physicians, poor communication between primary and secondary care physicians may also be an issue. It has been suggested that cardiologists may mistakenly presume that the primary care physician will manage the follow-up care of the patient and monitoring of CV risk factors. As a consequence, the primary care physician may perceive this lack of treatment plan to mean that the cardiologist does not believe a risk-lowering strategy is necessary.54 Ultimately, this breakdown in communication means that the outpatient is not properly informed about their condition and is poorly cared for.
Another phenomenon, known as clinical inertia, is rapidly becoming recognised as a barrier to optimal clinical outcomes. Clinical inertia can be thought of as the failure of a physician to intensify medication in patients who are not achieving goals that they are set. This is intrinsically related to non-adherence as the physician has to judge whether the set treatment goals are not being achieved due to an insufficient therapeutic dose or through failure on behalf of the patient to adhere to the treatment regimen as prescribed. In fact, both factors need to be considered. This was demonstrated in a retrospective cohort study of patients with coronary disease (n=10,447) evaluating the impact of medication non-adherence and therapy intensification on reaching target BP goals. In this study, medication non-adherence (odds ratio [OR] 1.73; 95% confidence interval [CI] 1.34–2.24) and therapy intensification (OR 1.31; 95% CI 1.01–1.70) were associated with having uncontrolled BP compared with having high systolic BP that became controlled over time.55
While the concept of considering medication intensification and non-adherence together might seem obvious, the paucity of clinical evidence to support the notion means that this is not often applied in clinical practice. For example, in a retrospective cohort study of 38,327 hypertensive patients who had filled prescriptions for one or more BP medications at Veterans’ Affairs (VA) healthcare facilities in a midwestern VA administrative region, providers only intensified medications in response to 30% of the 68,610 elevated BP events. In this study, there was almost no variation in intensification regardless of whether patients had good or poor BP medication adherence.56
Complicated drug regimens, including polypharmacy14,17,46,57 and high dosing frequency,58–63 are frequently reported as a reason for non-adherence. Polypharmacy is common in patients at risk of CVD due to the multifactorial aetiology of the disease. Furthermore, patients may have one or more co-morbidities that necessitate the use of additional, non-CV medications. In one study, around one-third of patients receiving both antihypertensive and lipid-lowering therapies were adherent to their medication after six months,46 although adherence was improved if both drugs were initiated at the same time.46 It has been suggested that polypharmacy may have an adverse effect on adherence because patients do not understand their complex dosing regimens and/or experience problems in organising their daily schedules to accommodate these regimens.30
Adherence may also be affected by the choice of drug prescribed. For example, adherence rates tend to be higher with the newer classes of antihypertensives, such as angiotensin II blockers, newer calcium channel blockers and angiotensin-converting enzyme inhibitors, than with older drugs, such as thiazide diuretics and non-selective beta-blockers.36,64–69 Similarly, adherence rates tend be higher with statins than with older lipid-lowering therapies.70,71 Increased adherence with newer agents may be related to the improved tolerability profiles of these drugs.
Economic factors also contribute, as high drug costs and the inability of patients to afford them are also occasionally reported as reasons for non-adherence, particularly in developing countries.14,29,72–76
Clinical Implications of Non-adherence
Several studies have identified non-adherence to therapy as an important factor that underlies both poor BP control14,26,50,55,57,72,77 and poor lipid control.29,33,78–81 Clearly, such inadequate control of CV risk factors can lead to serious clinical consequences (see Figure 1).82 Non-adherence has been linked with an increased risk of CV events and mortality. A recent meta-analysis of 21 observational studies of pharmacological therapy in chronic diseases, including predominantly antihypertensive and lipid-lowering therapies, showed that ‘good’ adherence was associated with a significant reduction in the risk of overall mortality compared with ‘poor’ adherence.83
In terms of antihypertensives, in the second Australian National Blood Pressure Study (ANBP2), patients (elderly patients with no prior history of CVD) who adhered to their antihypertensive medication regimen were less likely to experience CV morbidity or death (first CV event or first nonfatal CV event; a fatal other CV event; or a first occurrence of heart failure) than those who were non-adherent.20 Similarly, non-adherence to antihypertensive treatment had a substantial effect on mortality in women (enrolled in the National Heart, Lung and Blood Institute beta- Blocker Heart Attack Trial) who had previously experienced a myocardial infarction (MI).84
In terms of lipid-lowering therapies, patients with hypercholesterolaemia and no history of MI who took ≥75% of their prescribed dose of pravastatin in the West Of Scotland COronary Prevention Study (WOSCOPS), had significantly lower rates of non-fatal MI, revascularisation procedures, death from any cause and CV death compared with those who took <75% of their prescribed dose.85
In another primary prevention study (a nested case-control study) ≥90% adherence to statin therapy was associated with a significant reduction in non-fatal coronary artery disease events compared with <90% adherence.86 In addition, negative effects of non-adherence to lipid-lowering therapy on CV morbidity, CV mortality and all-cause mortality have been reported in large, observational, secondary prevention studies.87–90 Non-adherence to statins has also been shown to increase the risk of all-cause mortality in patients with diabetes in a large observational study.91
In addition to directly affecting CV outcomes, non-adherence also appears to predict treatment discontinuation in patients receiving initial antihypertensive treatment.92 This may be important, as early discontinuation increases the risk of acute MI and stroke.93
Economic Implications of Non-adherence
As well as affecting clinical outcomes in CVD, non-adherence to medication in chronic disease places a considerable burden on healthcare resources. Non-adherence is believed to be responsible for 10% of all hospitalisations and 23% of all nursing home admissions.94 In the US, the annual cost of non-adherence has been estimated at approximately US$300 million (2000 prices).95
Although improved adherence to medication may increase drug acquisition costs, these costs are likely to be balanced by reductions in secondary care expenditure. For example, in patients with HTN and hyperlipidaemia, meta-analyses and observational studies have shown that while ‘good’ adherence is associated with increased drug costs, ‘poor’ adherence leads to more CV events, and hence more physician visits/hospital admissions and longer hospital stays.96–98
Ultimately, it is clear that non-adherence to medication has serious negative implications in terms of both clinical and economic outcomes. Therefore, strategies to improve adherence are essential.
Strategies for Improving Adherence
There are various strategies that can be employed to improve adherence to medication in patients with HTN and/or hyperlipidaemia.
The results of several meta-analyses have suggested that behavioural/ motivational, educational and combined/simplified interventions may improve medication adherence to a small extent.58,59,61,62,99–101 While there is little evidence to support the superiority of any single strategy over another, a personalised, intensified, patient-focused programme that involves either frequent contact with healthcare professionals or a combination of interventions appears to be the most effective approach.61,100
This theory was recently evaluated in the Risk Evaluation And Communication Health Outcomes and Utilization Trial (REACH-OUT), which was a randomised study designed to evaluate if a physician-delivered CHD risk evaluation and communication programme (intervention programme) could lower the predicted Framingham 10-year risk of MI or death due to CHD by 10% versus usual care within six months.102 The intervention programme included informing patients of their 10-year risk of MI or death due to CHD, educating patients about modifiable risk factors and risk-factor control and three follow-up phone calls by a physician or study nurse. After six months, mean predicted CHD risk was 12.5% with the intervention programme, and 13.7% with usual care (OR 0.896; p=0.001, adjusted for risk at baseline and other covariates). The proportion of patients achieving both BP and LDL-C targets was significantly higher with the intervention programme (25.4%) than usual care (14.1%; p<0.001). The reasons underlying these improvements are not fully understood. However, the impact of the intervention programme on the knowledge, attitudes and behaviour of patients were assessed by means of a questionnaire completed at baseline and at six months.103 The results of this survey indicated that the programme-mediated improvements in the attitudes and awareness of patients of CHD risk, leading to changes in behaviour with respect to adherence to medication and lifestyle modifications.103
Other studies that have investigated the benefits of assessing and communicating the predicted CVD/CHD risk of patients (with the aim of improving risk factor management) have shown limited benefits.104–106 Ultimately, patient–physician communication intervention programmes need to constantly evolve in order to reflect changes in social attitudes and medical practice.
Simplifying Dosing Regimens
As polypharmacy has been shown to negatively influence patient adherence, it is unsurprising that there is evidence to suggest that simplifying dosing regimens may improve adherence. Several meta-analyses suggest that simplifying the dosing regimen, in particular reducing the dosing frequency, may improve adherence to both antihypertensive and lipid-lowering therapy.58–62,100 Once-daily antihypertensive regimens have been associated with adherence rates exceeding 75%.107
In the Caduet Adherence Research Program and Education (CARPE) study, a retrospective analysis of 4,703 patients at risk of CVD, pharmacy-claims data from a large US pharmacy benefit-management company were used to evaluate adherence to a fixed single-pill combination of amlodipine/atorvastatin (Caduet®; Pfizer Inc.) compared with two-drug regimens.108 Patients were newly initiated on the single-pill combination or concomitant calcium channel blocker and statin therapy (initiated within 30 days of each other regardless of initiation sequence) and were followed up for 180 days commencing at the prescription fill date of the second medication class. After six months, patients taking single-pill amlodipine/atorvastatin were almost twice as likely to achieve adherence compared with patients taking the two compounds separately (OR=1.95, 95% CI=1.80–2.13; p<0.0001).108 Furthermore, the percentage of patients achieving a proportion of days covered (PDC) of ≥80% was higher in the single-pill amlodipine/atorvastatin cohort (67.7%) versus other cohorts (≤49.9%).108 Similarly, in a subset analysis of 3,561 patients followed up for a year, the odds of achieving a PDC of ≥80% was higher in the amlodipine/atorvastatin single-pill cohort versus other cohorts.108
Furthermore, it has been shown that such fixed combinations effectively lower both BP and lipid levels. In the clinical utility of Caduet in simultaneously Achieving Blood Pressure And Lipid Endpoints (CAPABLE) trial (African-Americans with concomitant HTN and dyslipidaemia, and either no additional risk factors, one or more CV factors, or CHD or a CHD-risk equivalent; n=499) it was shown that: 48.3% of patients reached both their BP and LDL-C goals (versus 0.8% at baseline); 56.8% of patients reached BP goals (versus 1.4% at baseline); and 73.7% of patients reached LDL-C goals (versus 28.5% at baseline).109 Likewise, in the Avoiding Cardiovascular events through COMbination therapy in Patients LIving with Systolic Hypertension (ACCOMPLISH) outcome study investigating aggressive antihypertensive combination treatment (benazepril plus amlodipine), mean BP fell from 145.2/80.2mmHg to 132.7/74.7mmHg at 12 months of follow-up using the combination pill.110 The benefit of fixed-dose combinations in terms of adherence is supported by several retrospective studies39,47,111,112 and a meta-analysis of fixed-combination medications.113
Finally, because of the associated improvement in disease control and reduction in healthcare utilisation, improved adherence with fixed-dose combinations is associated with lower overall costs of care compared with the drug-free combinations.39,96,112
Although the problem of non-adherence may not be fully recognised by physicians, the clinical and economic consequences of this phenomenon are severe and multifaceted. The foundation for improving adherence will be to quantify the scale and impact of the problem accurately and disseminate this finding across the wide range of stakeholders. Educational approaches to encourage greater adherence should be aimed at healthcare providers, payers and patients and need to be co-ordinated, consistent and comprehensive. Such approaches should highlight potential economic advantages, which are of greatest interest to payers, clinical efficacy improvements to appeal to the prescribing physician and pharmacist, and for the patient, long-term benefits in terms of reduced CV events.
It is clear that more research, based on prospective clinical data, needs to be undertaken in order to identify effective approaches to improve adherence. Therefore, it is encouraging that the volume of such research is increasing. For example, numerous studies are currently under way assessing approaches such as community pharmacist-led interventions,114 online peer support115 and tools for shared decision-making.116 As reasons for non-adherence vary greatly between patients, the more weapons we have in our armamentarium the more likely we will be able to provide the right solution for each individual patient.
Ultimately, physicians need to understand the importance of improving adherence and have the tools to be able to measure it effectively in order to make decisions regarding medication intensification. Achieving optimal clinical outcomes in the future will be linked closely with the management of patient adherence to treatment, coupled with appropriate intensification/escalation of therapy.
It is clear that adherence to medication is highly important in terms of clinical and economic outcomes, and incredibly complex in terms of management. Human nature and the pressures of modern living mean that new strategies have to be developed and employed in order to improve adherence. Perhaps paramount among such strategies is patient education and motivation, with the rationale being that an educated and motivated patient is much more likely to follow prescribed treatment interventions. In addition, simplified-dosing regimens can play an important role in improving patient adherence and make the task of managing multiple CV risk factors that little bit easier. Ultimately, optimal clinical and economic outcomes are the result of excellent partnerships between patients and physicians.