Adherence to Cardiovascular Treatment Regimens
Lora E. Burke
Kyeongra Yang
Sushama D. Acharya
Adherence or compliance has been studied extensively in recent decades.1, 2, 3, 4 The terms adherence and compliance have been used interchangeably in the literature; however, more recently the use of adherence has superseded compliance.5 Compliance is viewed by many as having a negative connotation that implies an authoritarian relationship between the provider and the patient with the provider issuing instructions that the patient is expected to follow. Adherence is similar but is seen as recognizing the rights of the patient to chose and thus removes the concept of blame.6 Concordance, a term used mainly in the United Kingdom, has been more broadly defined and today ranges from prescribing and communicating to supporting the patient in medication taking, and includes consideration for the preferences and beliefs of the patient.6 Today, it seems to be more widely recognized that the patient is only one part of the equation when adherence is considered. Numerous factors may play a part in adherence, which involves the health care professional, the system or organization in which care is delivered, and the patient, for example, the provider’s suboptimal use of evidence-based treatment guidelines or the health organization’s practices that present barriers to the patient’s attempts to being adherent.7
Other terms that sometimes are considered synonymous or related include self-management and disease management.5,7 Heart failure disease management focuses on educating patients about adherence, monitoring symptoms that may warn of decompensation, factors that may precipitate an exacerbation, and being seen in close follow-up by nurses specialized in heart failure care. Broader than adherence or compliance, self-management includes general strategies and behaviors that contribute to disease management, improved health, and prevention or reduction of complications rather than mainly focusing on following specific regimen components.8,9
This chapter reviews adherence and the significance of nonadherence in the management of the cardiac patient. Methods used to assess adherence across the behaviors of medication taking, dietary self-management, following an exercise program, and smoking cessation are reviewed. Factors that influence adherence and strategies to enhance adherence are discussed and guidelines for implementing educational and behavioral strategies are provided.
SIGNIFICANCE OF NONADHERENCEA number of pharmacologic therapies are used in the prevention, as well as the acute and chronic management of cardiovascular disease (CVD). However, the extent to which these therapies can be demonstrated to be efficacious in clinical trials and later effective when prescribed by practitioners can be influenced by the patient’s adherence to the treatment regimen,10 which is less than ideal in both clinical trial and clinical practice settings. The survival benefits of several drugs have been demonstrated in large-scale clinical trials. However, it has been shown repeatedly that 50% of individuals prescribed statins will discontinue the therapy within 6 months10 or stop taking the drug for an extended period.11 A quantitative review of 50 years of research in patient adherence revealed that the average nonadherence rate is 24.8%; the highest adherence rates are among patients with HIV disease, arthritis, gastrointestinal disorders, and cancer, the lowest are among patients with pulmonary disease, diabetes, and sleep disorders.12 Approximately 31% of hypertensive participants in a Veterans Affairs study reported unintentional nonadherence mainly due to carelessness or forgetfulness and 9% reported intentional nonadherence.13 The rates of nonadherence to treatment recommendations are found to be 20% to 40% for acute illness, 30% to 60% for chronic illness, and 80% for prevention.14 The most common preventable cause of rehospitalization in the heart failure population is nonadherence to the regimen.15 In the United States, 33% to 69% of medication-related hospital admissions are the result of poor medication adherence, resulting in an annual cost of $100 billion.10 These statistics illustrate how medication nonadherence is a major health problem; however, it is not limited to the United States.
Nonadherence is a ubiquitous problem that spans across continents and treatment regimens. Indeed, the magnitude of this problem was underscored by the World Health Organization (WHO) convening a panel of experts to examine its prevalence and develop an evidence-based report on treatment strategies. The panel reported that adherence to long-term therapies in developed countries is approximately 50% and is much lower in developing countries. Poluzzi et al.16 reported 69% adherence during the second year and 60% during the third year of antihypertensive therapy among Italian adults. A study among Asian Pacific Americans (Japanese, Filipino, Chinese, Korean, and part-Hawaiian; N = 28,395) showed that Japanese living in Hawaii were 21% more likely to adhere to antihypertensive medications than white population while individuals of Korean, Hawaiian, and Filipino descent were less likely to adhere than white population, after controlling for patient’s education and physician characteristics such as specialty, gender, and race.17 Among all ethnic groups in this sample, overall adherence rates were less than 60%.
Several risk factors associated with CVD are related to lifestyle; however, adherence to public health recommendations for dietary and physical activity habits is also lacking. Generally, Americans exceed the dietary fat limit by 2% to 5%, depending on ethnic group and exceed the 2,400 mg of sodium per day guideline by
1,000 mg. In contrast, the reported intake of dietary fiber is at approximately 50% to 75% of the recommended intake.18 Results from the behavioral risk factor surveillance system revealed that 50.01% of adults engaged in regular activities and that 13.5% were physically inactive.19 In general, white population reported being more physically active than black population and Hispanics. The dietary excess and physical activity deficiency rates are reflected in the prevalence of overweight and obesity in the United States and other westernized countries; the most recent reports indicate a combined prevalence of 66% for the United States.20
1,000 mg. In contrast, the reported intake of dietary fiber is at approximately 50% to 75% of the recommended intake.18 Results from the behavioral risk factor surveillance system revealed that 50.01% of adults engaged in regular activities and that 13.5% were physically inactive.19 In general, white population reported being more physically active than black population and Hispanics. The dietary excess and physical activity deficiency rates are reflected in the prevalence of overweight and obesity in the United States and other westernized countries; the most recent reports indicate a combined prevalence of 66% for the United States.20
The poor rates of smoking has been relatively static with an estimated 23.9% men and 18.1% women current smokers in the United States as indicated by National Health Interview Survey. The highest proportion of smokers were American Indian or Alaska Native adults (25%) followed by black (21%), white (21%), and Asian populations (13%).21 Worldwide more than one third of the population is estimated to be smokers.22 Studies suggest that smoking cessation rates remain low with the majority of self-quitters relapsing within the first 8 days of initial cessation, and that only about 3% to 5% of self-quitters are able to successfully achieve abstinence for 6 to 12 months after initial cessation.23 Among those who participate in a formal treatment program, relapse during or after treatment is usual and might require treatment several times.24
The duration of treatment is usually a factor influencing compliance, with an initial decline in adherence observed in the first year followed by a gradual decline over time. This pattern is observed repeatedly among those participating in long-term programs, for example, weight loss programs.25 The prevention and treatment of CVD requires ongoing management of lifestyle habits and, increasingly, inclusion of pharmacologic therapy, such as aspirin, hyperlipidemic agents, β-blockers, or calcium channel blockers. In the absence of sustained adherence, the benefits of prevention or treatment cannot be realized. This may be critical among patients who have had solid organ transplantation. A meta-analysis of 147 studies revealed that average nonadherence rates to immunosuppressants, diet, exercise, and other health care requirements ranged from 19 to 25 cases per 100 patients per year; failure to exercise was highest among heart recipients.26
In the clinical arena, nonadherence at any point in the treatment continuum poses a threat to satisfactory outcomes. Medication nonadherence has been associated with increased risk of coronary heart disease, precipitated episodes of heart failure, late organ rejection among heart transplant recipients, and mortality. The literature emphasizes the mediating effects of adherence on clinical outcomes, and the impact nonadherence can have on morbidity and mortality associated with CVD regardless of when it occurs in the treatment continuum.7,12,26
In the research arena, nonadherence affects therapy evaluation before its introduction into the clinical setting. Incomplete adherence to the treatment under study underestimates its efficacy, and the diminished effect reduces the study’s power to detect a difference between treatment groups, thus preventing the study from meeting the assumptions of the projected sample size. In this situation, when nonadherence to the study protocol results in diminished effect, additional subjects are required. Furthermore, nonadherence to the treatment protocol may mask side effects or result in an overestimation of optimal dosage.1,27, 28, 29 Finally, intermittent or varying adherence to the study protocol may reflect varying adherence to concomitantly prescribed therapeutic modalities, which may affect study outcomes.30
METHODS OF MEASUREMENTAssessment of adherence needs to be incorporated into each clinical encounter. It is important that the clinician separate adherence from therapeutic or clinical outcome, which can be affected by a myriad of variables besides adherence. For example, inadequate control of serum cholesterol may be due to inadequate drug dosage, individual variation in pharmacokinetic factors of different drugs, daytime or seasonal variations in measurement values, or personal factors. Conversely, the absence of symptoms or achievement of goal does not confirm adherence. Clinical outcomes are indirect measures of adherence, whereas patient behaviors (e.g., weight loss, exercise, taking the medication) are direct measures of adherence. Both direct and indirect measures have inherent advantages and disadvantages.31,32 Unfortunately, it is difficult to measure behavior directly, and thus there is a great reliance on self-reported behavior. Table 40-1 summarizes the numerous measurement methods and the advantages and disadvantages of their use.
Adherence assessment can be conducted through numerous methods. However, a weakness common to all forms of measurement is a bias toward overestimation of adherence.31 One of the reasons for this measurement error is that the period being measured is usually not representative of the patient’s usual behavior. Research has shown that patients’ adherence varies in relation to the clinical appointment, with adherence increasing immediately prior to and after the visit.33 An example of this would be the patient taking medicines very closely to how they were prescribed, or closely following a low-cholesterol eating plan for the 7 days prior to the clinic appointment. Thus, when the patient is asked to report on his or her behavior, the report may be influenced by the individual’s recollection of the most recent behavior and thus overestimates adherence for the longer period.31 Cramer’s research also showed that the patient was more adherent in the 7 days following the appointment, and then adherence again tapered off until a week prior to the next appointment.34 A variety of methods are available to measure adherence in the clinical setting. These include self-report, biologic and electronic measures, pill counts, and records such as pharmacy refills.
Self-Report Measures
Self-report measures consist of interviews, structured questionnaires, and diaries, which can be in either paper-and-pencil or electronic formats. This form of adherence assessment is used most frequently, which is probably explained by its ease of administration and low cost.
Interviews
Interviews, often used in the research setting to assess adherence behavior at each contact, can easily be conducted in the clinical setting. A brief interview scale was developed to assess global medication compliance among hypertensive patients. The four-item scale developed by Morisky and colleagues pertains to areas of omission, such as forgetting, being careless, and stopping the medication when feeling better or when feeling worse. The literature would suggest that the scale is used more often as a questionnaire.
Adherence can also be ascertained through a 7-day recall interview by asking the patient to report the number of pills and the times at which these were taken for each day of the week before
the visit. However, these tend to provide an overestimation of adherence.35 When comparing self-reported interview adherence to electronic measured adherence, Dunbar-Jacob et al.35 found 97% adherence reported in the interview compared with 84% adherence measured by an unobtrusive electronic event monitor.
the visit. However, these tend to provide an overestimation of adherence.35 When comparing self-reported interview adherence to electronic measured adherence, Dunbar-Jacob et al.35 found 97% adherence reported in the interview compared with 84% adherence measured by an unobtrusive electronic event monitor.
Table 40-1 ▪ METHODS OF ADHERENCE MEASUREMENT AND FEATURES OF THEIR USE | |||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| |||||||||||||||||||||||||||||||||||||||||
Assessing dietary adherence requires a determination of what the person eats and the degree to which the food intake approximates the recommended diet.36 The most widely used and rigorous measure of dietary adherence in population studies is the 24-hour dietary recall, where individuals are asked to recall their food and beverage intake in the previous day.37 The recall is conducted unannounced so that individuals cannot change eating habits in anticipation of the recall. This method allows more exact description of foods (e.g., brands, degree of fat modification) but also requires interviewer skill at prompting recall and eliciting detail. Benefits of the 24-hour recall are increased accuracy because of the shortened recall period and reduced patient burden compared to recording in a food diary, but a disadvantage is that there may be increased bias if the recall is conducted for days on which the eating pattern is atypical.37 To compensate for this weakness, some studies have multiple 24-hour dietary recalls (from 3 to 7) performed on nonconsecutive days to account for daily variations in food intake; however, three is most typical.
In order to improve accuracy in reporting dietary intake, various techniques have been employed to help individuals estimate their intake accurately. One such example includes the United States Department of Agriculture automated multiple-pass method (AMPM), a five-step multiple-pass 24-hour dietary recall method.38 It is a computer assisted 24-hour dietary recall designed to provide better cues for respondents’ cognitive processes. The AMPM has been shown to provide valid measures of total energy and nutrient intake among healthy normal weight women38 and obese women.39 Based on the AMPM approach, the Nutrition Data System for Research is a comprehensive software program available for research purposes for dietary data collection and analysis through 24-hour dietary recalls, food records, menus, and recipes. It also features optional dietary supplement data that may be included with 24-hour dietary recalls or food records. This software was developed by the Nutrition Coordinating Center at the University of Minnesota and is updated annually to reflect marketplace changes and new analytic data. It contains values for 155 nutrients, nutrient ratios, and food components and includes over 18,000 foods, including ethnic foods and over 8,000 brand products (NDSR, 2006 to 2007, University of Minnesota, Minneapolis). Additional software programs are available to collect dietary data, for example, the United States Department of Agriculture Nutrient Database for Standard Reference (Washington, DC), ProNESSy (Princeton, NJ), and Food Processor (Salem, OR). All these programs provide summarization of dietary data and detailed reports of macronutrients (carbohydrate, fat) and micronutrients (vitamins, minerals).
Adherence to exercise regimens may also be assessed through 7-day physical activity recall interviews. One study reported a very weak association between the 7-day physical activity recall and the energy expenditure as assessed by doubly labeled water.40 However, on balance, self-report measures provide the most practical and cost-effective method for assessing adherence. Interviews may be guided by established questionnaires, for example, the Paffenbarger, the Physical Activity Recall, and the Modified Activity Questionnaire.
Nicotine dependence was assessed through the nicotine dependence module of the Composite International Diagnostic Interview.41 The Composite International Diagnostic Interview is a comprehensive and standardized instrument to assess the presence of nicotine dependence in the past year and a lifetime history of nicotine dependence. It is reported to provide reliable and valid psychiatric diagnosis of nicotine dependence based on the International Classification of Diseases (ICD-10) and the Diagnostic and Statistical Manual of Mental Disorder (DSM-IV).42
Questionnaires
Questionnaires are available to assess adherence across multiple behaviors. Although there are numerous scales available for assessment of eating and exercise behaviors, few exist for medication-taking behavior. The Morisky scale, first published in 1986, has been adapted by several investigators and used as a paper-and-pencil questionnaire in several populations, including those being treated for hypercholesterolemia, rheumatoid arthritis, and HIV. This scale, for which adequate psychometric properties have been reported, was used recently in a study of medication adherence among older Chinese immigrants43 and patients taking cardiovascular medications.44,45 Shalansky et al.45 recommended rewording the questions, increasing the number of items, and the use of graded response options to improve the scale’s consistency. Rottlaender et al.46 used the scale and found that 83% of the patients reported they were absolutely compliant to their medication regimen but the Morisky score indicated high adherence in only 52% of the sample. Similarly, the Morisky questionnaire revealed lower adherence rates than that measured by pill count.47
Dietary adherence can be measured by several established questionnaires including the Connor Diet Habit Survey, the Eating Pattern Questionnaire, and food frequency questionnaires (FFQ). The first two questionnaires focus on fat intake and have reported psychometric properties when used in cardiac and general populations.48,49 However, the FFQ is now the most commonly used dietary measure to provide estimates of usual dietary intake over time (typically 6 months to a year) in large epidemiological studies.50 FFQs include a list of foods with a frequency response section to report how often and how much each food item was consumed. Examples of FFQ are the Harvard/Willett FFQ, National Cancer Institute’s Diet History Questionnaire (DHQ), and the Fred Hutchinson FFQ. The Fred Hutchinson FFQ was updated in 2001 and has a separate questionnaire for men and women. The DHQ was recently updated in 2007 to reflect changes in food availability. Both Fred Hutchinson FFQ and the DHQ are available in English and Spanish. Although FFQs provide a relatively inexpensive and standardized way of collecting dietary information, their major limitation is the number and types of items listed thereby reducing its utility among ethnic groups.37 It becomes very important for the FFQ to be culture-specific to capture dietary intake of specific racial/ethnic groups. FFQs have been adapted and validated to assess the diet of diverse populations,51 US Chinese women,52 South Asians in the UK,53 and elderly populations of low socioeconomic status.54 A regionally specific FFQ has been developed for white and black adults residing in the southern region of the United States.55
Measurement of physical activity, which continues to receive high priority in the public health field, has relied primarily on the questionnaire.56,57 Exercise assessment questionnaires, which are subjective measures, have been validated by objective measures of physical activity, such as measures of total energy expenditure (doubly labeled water), estimates of physical fitness (heart rate), or measures of physical motion by accelerometers.57 A beneficial trait of the questionnaire is that it does not influence the behavior being measured, and although less precise than the objective measures, it estimates activity relative to others in the population. The questionnaire may range from one item to an array of questions covering a wide range of occupational and leisure activities, and may cover varying time intervals. A compilation of physical activity questionnaires and a review of their psychometric properties was published, providing an excellent resource for anyone wishing to measure exercise adherence.58 In selecting a questionnaire, the investigator must consider characteristics of the population, such as gender, age, culture, and the outcome of interest. Most of the activity questionnaires were developed with men’s activities in mind making them less sensitive to differences in physical activity levels in women.57 The Kaiser Physical Activity Survey, which includes questions on household/care giving activities, is available to measure physical activity in women.59 The 7-day Physical Activity Recall and the Paffenbarger Physical Activity Questionnaire are widely used for various groups including patients in cardiac rehabilitation programs,60 male veterans,61,62 individuals in the National Weight Control Registry,63 and adults with a body mass index greater than 27.64 The Community Healthy Activities Model Program is a valid and reliable questionnaire to estimate physical activity among middle- and older-aged adults.65
The most common measure of nicotine dependence among cigarette smokers is the Fagerstorm Tolerance Questionnaire (FTQ),66 which was designed to estimate the degree of nicotine dependence in smoking. The Fagerstrom Test for Nicotine Dependence (FTND) is a shortened version of FTQ that emphasizes cigarette consumption and time to first cigarette after awakening.67 Recently, the Cigarette Dependence Scale (CDS-12) has been proposed as a good alternative to FTND for measuring nicotine dependence with better validity and internal consistency than FTND.68
In summary, questionnaires with a shorter time interval are less vulnerable to recall bias and easier to validate with objective measures. However, using a shorter time frame reduces the likelihood of obtaining a picture of usual behavior, because eating and exercise patterns may vary by season. Reliability and validity are affected by the person’s ability to store and retrieve information, and by potential influence of the interviewer or respondent bias.69,70
Diaries
Daily diaries for food intake or exercise circumvent the bias of recall, but require training and cooperation of the patient or study participant, which limits its use to highly motivated, literate individuals. While diaries may be used as part of an intervention to achieve awareness of one’s behavior, the focus here is on assessment of adherence. Food and exercise diaries are often used periodically and cover a 3- or 7-day period, including one nonwork or leisure day. Recording for extended periods (i.e., over 3 days) may reduce accuracy, and the recording may begin to influence the recorder’s behavior. Several investigators71, 72, 73 have used diaries to measure exercise and dietary adherence. Wickel74 reported high convergent validity for the Bouchard activity diary with an accelerometry-based monitor. However, when comparing self-report (questionnaire) to doubly labeled water data, Walsh et al.75 demonstrated that sedentary overweight women overreported their exercise in comparison to normal weight control counterparts.
Issues of concern with self-report measures include response biases due to social desirability, deliberate and nondeliberate
errors in recall or reporting; for example, underreporting food consumption76,77 and overreporting energy expenditure75,78 are common phenomenon among obese subjects. Moreover, there are concerns specific to self-report of eating and exercise behaviors. It is difficult for individuals to accurately estimate the portion size and components of mixed foods. Because there are so many dimensions of exercise, it is a challenge for individuals to accurately characterize the type of exercise, its frequency, duration, and intensity.78 Staff need to be trained on how to teach participants to record the information, and potential problems with memory and social desirability need to be reduced. For example, recording the behavior immediately reduces forgetting and conveying an expectation of a full range of behaviors may help reduce less than truthful reports. Despite their limitations, self-report measures are common, easy to use, inexpensive, and provide information on the circumstances surrounding the good or poor adherence.
errors in recall or reporting; for example, underreporting food consumption76,77 and overreporting energy expenditure75,78 are common phenomenon among obese subjects. Moreover, there are concerns specific to self-report of eating and exercise behaviors. It is difficult for individuals to accurately estimate the portion size and components of mixed foods. Because there are so many dimensions of exercise, it is a challenge for individuals to accurately characterize the type of exercise, its frequency, duration, and intensity.78 Staff need to be trained on how to teach participants to record the information, and potential problems with memory and social desirability need to be reduced. For example, recording the behavior immediately reduces forgetting and conveying an expectation of a full range of behaviors may help reduce less than truthful reports. Despite their limitations, self-report measures are common, easy to use, inexpensive, and provide information on the circumstances surrounding the good or poor adherence.
Biologic Measures
Adherence is often reported in terms of biologic end points, such as serum cholesterol or glycosylated hemoglobin level. Other biologic assays frequently used include serum, urine, or saliva level of a drug or its metabolites. Examples include antihypertensive medication adherence measured by serum bromide level,79 dietary adherence measured by urine sodium,80 smoking cessation by serum or saliva thiocyanate or cotinine,81 and exercise by direct or indirect calorimetry and maximal oxygen uptake.82 Doubly labeled water (2H218O), a procedure that requires the subject to ingest water enriched with18O and2H isotopes, is the most accurate measurement of total energy expenditure available,83 but is too costly to be used on a widespread basis.84 A limitation of biologic assays is that daily variability in compliance cannot be detected. Instead, they indicate if the person has been adherent close to the time of assessment and may serve as a validation of the behavior. Moreover, biological measures are not available for many drugs and dietary factors; moreover, biologic assays may be influenced by many other factors such as ethnic differences in urinary sodium excretion.85
Electronic Monitoring
Technology has provided tools for ongoing and detailed assessment of adherence behavior. Electronic methods include unobtrusive electronic medication monitors, heart rate monitors, electronic motion detectors for exercise (pedometers, accelerometers, SenseWear Armband), and electronic diaries for self-report data.86 The Medication Event Monitor System (MEMS; APREX, a division of AARDEX Ltd., Union City, CA), consists of an electronic chip housed inside the medication bottle cap, provides date and time data on medication bottle openings and closures. This assessment is based on the assumption that bottle opening leads to pill removal and ingestion.87,88 An additional applications of an electronic monitor for medication use includes the IDAS II (Intelligent Drug Administration System, Bang and Olufsen Medicom, Denmark) that accommodates blister pill packs.88 This new device uses visual and audible reminders to the patients to enhance adherence. Similar to the MEMS, the IDAS II has been demonstrated to be acceptable to people with hypertension.88 Santschi et al.88 reported that there was a tendency for patients using the IDAS II to take their drug more regularly.
Several devices measure exercise adherence. The Polar® monitor can be worn during ambulatory exercises (walking, bicycling). It includes a microprocessor that measures and sequentially stores average heart rate values, which provide data on adherence to the exercise prescription (Polar Electro, Inc., Lake Success, NY).89,90 Pedometers are used to measure steps and miles during ambulatory activity and are the most inexpensive objective monitoring device for physical activity. However, pedometer accuracy is influenced by walking speed; there was significant improvement in pedometer accuracy in fast pace walking, compared to that in slow pace walking.91, 92, 93 Electronic accelerometers are motion sensors that register body accelerations and decelerations, and thus provide a direct and objective measure of movement intensity and frequency during physical activity. The Actigraph accelerometer (Manufacturing Technologies Inc., Fort Walton Beach, FL) and the Caltrac accelerometer (Muscle Dynamics Corp., Torrance, CA) were used to assess physical activity levels among adults with congenital heart disease,94 and among individuals of all ages from the 2003 to 2004 National Health and Nutritional Examination Survey.95 The SenseWear® Armband (BodyMedia, Inc., Pittsburgh, PA) is a relatively new product that monitors physical activity and energy expenditure. It is worn on the upper right arm. It includes a two-axis accelerometer, heat flux sensor, galvanic skin response sensor, skin temperature sensor, and a near-body ambient temperature sensor, which enables it to measure heat produced by the body as a result of basic metabolism and from all forms of physical activity. Data are stored up to 12 days and can be uploaded to a personal computer for analysis using the SenseWear® Software. This device has been shown to reliably determine energy expenditure in both the active and resting state, and thus measures activity adherence in an unsupervised setting.96, 97, 98, 99
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
