barriers to participation in and adherence to cardiac rehabilitation programs: a critical literature...

Progress in Cardiovascular NURSING Winter 20028

Despite the documented evidence of the benefits ofcardiac rehabilitation (CR) in enhancing recovery andreducing mortality following a myocardial infarction,only about one third of patients participate in such pro-grams. Adherence to these programs is an even big-ger problem, with only about one third maintainingattendance in these programs after 6 months. Thisreview summarizes research that has investigated bar-riers to participation and adherence to CR programs.Some consistent factors found to be associated withparticipation in CR programs include lack of referral byphysicians, associated illness, specific cardiac diag-noses, reimbursement, self-efficacy, perceived benefitsof CR, distance and transportation, self-concept, self-motivation, family composition, social support, self-esteem, and occupation. Factors associated withnon-adherence include being older, female gender,having fewer years of formal education, perceiving thebenefits of CR, having angina, and being less physi-cally active during leisure time. However, many of thestudies have methodologic flaws, with very few con-trolled, randomized studies, making the findings tenta-tive. Problems in objectively measuring adherence tounstructured, non-hospital-based programs, which arean increasingly popular alternative to traditional pro-grams, are discussed. Suggestions for reducing barri-ers to participation and adherence to CR programs, aswell as for future research aimed at clearly identifyingthese barriers, are discussed. (Prog Cardiovasc Nurs.2002;17:8–17) ©2002 CHF, Inc.

Heart disease is the leading cause of death in thedeveloped world, accounting for approximately

one quarter of deaths.1 Most deaths occur within thefirst 2 days after the onset, while those who survivethis period progress well if they have not developedheart failure or serious arrhythmias. Participation in a cardiac rehabilitation (CR) program can promoterecovery, enable patients to achieve and maintainbetter health, and reduce the risk of death in peoplewho have heart disease.2 It has been shown thatexercise-based CR can reduce fatal events by 25%in the first year of recovery and can significantlyreduce overall mortality from cardiac illness.3,4 Morerecent studies of psychosocial and education-basedCR have shown even more impressive benefits interms of mortality, morbidity, and quality of life.5,6

CR may be defined as “services that are com-prehensive, long-term programs involving medicalevaluation, prescribed exercise, cardiac risk-factormodification, education and counselling.”7 CR pro-grams are now regarded as an essential compo-nent of the overall care of patients following amyocardial infarction (MI).8 The overall aims are toimprove function, relieve symptoms, and enhancethe patient’s quality of life.9 Cardiac programs aredesigned to improve both the physiologic and psy-chological status of cardiac patients. Rather thanthe traditional approach of discouraging physicalactivity, in the past 40 years there has been a shiftto encouraging as much activity as a patient’ssymptoms and medical status permit.10 A CR mul-tidisciplinary team typically offers exercise training,education, and counselling of both client and fami-ly about risk factors, lifestyle changes, and copingwith the disease process.11 Not only do the pro-grams focus on enhancing recovery, they also focus onteaching ways of modifying risk factors (e.g., weightloss, cessation of smoking, stress management). Par-ticipation in these programs has been found to result inimproved exercise capacity and habits, improvement inblood lipid and lipoprotein levels, body weight, blood

From the University of Western Sydney, Penrith, Australia;1Concord Hospital, Sydney;2 University of York, York, UnitedKingdom;3 and St. George Hospital, Sydney, Australia4

Address for correspondence:John Daly, RN, PhD, Head, School of Nursing, Family and Community Health, College of Social andHealth Sciences, University of Western Sydney, Parra-matta Campus, Locked Bag 1797, Penrith South DCNSW 1797, AustraliaE-mail: [email protected] received January 26, 2001;accepted May 22, 2001

Barriers to Participation in and Adherence to Cardiac Rehabilitation

Programs: A Critical Literature Review

John Daly, RN, PhD;1 Andrew P. Sindone, BMed (Hons), MD, FRACP;1,2

David R. Thompson, RN, PhD, FRCN;3 Karen Hancock, BSc (Hons), PhD;1Esther Chang, RN, PhD;1 Patricia Davidson, RN, MEd1,4

Winter 2002 Progress in Cardiovascular NURSING 9

glucose, and blood pressure levels, and cessation ofsmoking,12 and to have psychological benefits, such asreductions in anxiety and depression.13,14 Another ben-efit of CR is that it increases the functional indepen-dence of patients, as measured by a return toappropriate and satisfactory work.8,12 In terms of costeffectiveness, an additional benefit is reduced costs forsubsequent hospital treatment.15

Despite the multifaceted benefits of CR programs,US figures suggest that only 15%–30% of patientswho have had an MI participate in such programs.8,16

Similar rates have also been reported in Australia.17 Anadditional problem is that those who do participate tendto have poor adherence to these programs. A patternof increasing attrition has been found in the first 3–6months of a CR program, and by 6 months, only30%–60% of individuals who attend CR programs par-ticipate in an exercise program.18–20 After 6 months, thedropout rate slows.19 Clinical trials have demonstratedthat the benefits of CR are greatest in those whoadhere to the program.12,19 While the problem ofadherence to CR has been extensively studied, theattributes which characterize adhererents and nonad-herents have been less conclusively established. How-ever, research conducted since 1978 reveals someconsistent factors to be associated with nonadherenceto CR programs: being older, being female, havingfewer years of formal education, having angina, andbeing less physically active during leisure time.19,21–23

Other studies have found lack of referral by physicians,associated illness, specific cardiac diagnoses, reim-bursement, self-efficacy, distance and transportation,self-concept, self-motivation, family composition, socialsupport, self-esteem, and occupation to influence par-ticipation in CR programs.25–28 Rather than singlefactors predicting adherence to CR programs, moststudies have found combinations of variables to berelated to participation and adherence to CR.

This paper critically reviews studies conducted overthe last 15 years that have investigated barriers to par-ticipation and adherence to CR programs, and offerssuggestions for improving participation and adherencerates to CR.

REFERRAL BY HEALTH CARE PRACTITIONERSParticipation in CR is dependent on a referral from aphysician.29 Hlatky et al.30 reported that 80%–85% ofphysicians recommend that their patients follow anexercise program after an uncomplicated MI. It isargued by many researchers that CR is beneficial fora wide scope of cardiac patients, such as those withcoronary artery bypass grafts (CABGs) and compli-cated MI,12,31–33 though training may be more gradualand supervised than for uncomplicated MI. However,

Bittner et al.34 found that only 7% of cardiac patientswho are eligible are actually referred to CR. Burns etal.29 also reported that only a small percentage ofpatients are referred for outpatient CR after hospitaldischarge. They noted that fewer MIs, fewer CABGs,less self-reported leisure physical activity, and previ-ous participation in CR were related to an increasedlikelihood of referral to CR.29 It may be that patientswith a history of MI or CABG were viewed as havingless potential for secondary prevention.

Pathman et al.35 reported that many physicians werenot following correct guidelines for referral, with lack ofawareness that guidelines for referral exist, or disagree-ment with the content of the guidelines, nonadoption ofthe guidelines in provision of care, and irregular adher-ence to the guidelines. Other factors argued to be relat-ed to lack of referral are the provision of risk-factormodification by the treating physician (making formalrehabilitation appear unnecessary) and professionalskepticism about the efficacy of CR or the ability or moti-vation of the patient to make lifestyle changes.16,32,36

Limited access and funding for available programs areother reasons for nonreferral.37 Indeed, the New SouthWales Division of the National Heart Foundation of Aus-tralia stated, “Rising health costs mean that it is nolonger reasonable to expect all people with heart dis-ease, irrespective of severity, to participate in medicallysupervised exercise programs.”38

The strength of the primary physician’s recom-mendation for participation has been found to be predictive of patient participation in conventional CRprograms.24 Ades et al.24 found that of all the predic-tors (demographic, medical, psychosocial, physician’sreferral), by far the most powerful predictor of partici-pation was the strength of the primary physician’s rec-ommendation. The stronger the recommendation, thegreater the participation. However, this study wasconducted in older patients, who represent one half ofcardiac patients, and the findings of the study cannotbe extrapolated to younger patients.

Although further controlled research is necessary,the above research suggests that the referring doctorhas a strong influence on whether CR patients partici-pate in or adhere to CR programs.

PSYCHOSOCIAL FACTORSSelf-EfficacySelf-efficacy reflects the individual’s perceptions orbeliefs about how capable he or she is in performinga specified activity.39 Oldridge40 suggested that self-efficacy enhances the self-regulating processes thatmaintain behaviors needed to reduce the risk of coro-nary artery disease. Several studies have found per-ceived self-efficacy to be predictive of CR exerciseadherence,41–43 while Vidmar and Robinson44 found

that self-efficacy was a significant predictor of compli-ance with a CR program. That is, those subjects whohad high self-efficacious beliefs were more likely to adhere to the CR exercise program. One implica-tion of these findings is that since patients with lowself-efficacy for risk reduction behaviors are less likelyto maintain the behaviors needed to lower the risk of coronary artery disease, these patients should be referred to a formal CR program rather than unsu-pervised home-based programs. However, furtherresearch is necessary to confirm findings, due to thelack of data in this area.

Self-MotivationIt has been argued that self-motivated people with highself-efficacy will be positively influenced by their atten-dance at CR.45,46 Self-motivation is the “want-to” com-ponent of individuals’ actions,47 making this conceptuseful in developing an understanding of the voluntarynature of an activity such as choosing to attend andadhere to CR. Although patients with low self-motiva-tion appear to be at risk for nonattendance and poorattendance at CR programs, there is a lack of empiricevidence demonstrating this relationship. However,Rhodes et al.48 did find that subjects who attendedand completed CR programs scored high on mea-sures of self-motivation. Comparisons with a matchedcontrol group who did not participate in CR are need-ed to establish the link between self-motivation factorsand participation in CR.

Self-EsteemA concept related to self-efficacy is self-esteem. Self-esteem indicates the extent to which an individualbelieves he or she is significant, worthy, capable, andsuccessful.49 Research has found that persons whohave experienced an MI are more likely to have loweredself-esteem due to the experience of redefining self-concept and changing roles and lifestyles.50–52 Connet al.53 found a positive relationship between self-esteem and MI patients’ participation in CR, thoughthe retrospective and crude measure of outcomemakes the findings tentative. Radtke20 found thatindividuals who complied with their home exerciseprogram also demonstrated a high degree of self-motivation. Thus, it appears ironic that MI is associ-ated with lowered self-esteem, yet those patientswith low self-esteem are less likely to participate inthe very program that would benefit them throughits treatment of self-esteem problems.

PersonalityThe findings concerning the relationship between par-ticipation/adherence with CR and personality aremixed. In a literature review, Emery54 concluded that

personality did not appear to be associated with adher-ence to CR exercise. Other reviews of CR adherencealso do not emphasise the role of personality.24,40 How-ever, Hershberger et al.55 found that personality vari-ables, such as sense of well-being, socialization(acceptance of rules and regulations), and communali-ty (view of self as similar to others), were associatedwith appointment-keeping, using the California Psycho-logical Inventory as a measure of personality. Oneimplication of these findings is that CR programs needto be flexible to suit the personalities of all patients.However, some methodologic problems with this study(cross-sectional design, small sample size, homoge-neous sample of older male veterans) suggest cautionbe used in interpreting the findings.

DepressionApproximately 10%–30% of patients are clinically de-pressed after an MI.56 Carney et al.57 found depressionto be related to adherence to CR programs. WhileAdes et al.24 did not find that depression during hospi-talization predicted participation in CR, subjects with ahistory of depression prior to the MI were less likely toparticipate in CR. Again, it seems ironic that CR pro-grams are designed to improve the psychological sta-tus of participants, yet those who are most likely tobenefit (i.e., those with depression) are less likely to par-ticipate. Some gender differences have been found inthe role of depression as a barrier to CR, as will be dis-cussed below under “gender.”

AnxietyThere has been little recent research during the last 15years on the role of anxiety as a barrier to participationor adherence to CR. Conn et al.53 did not find a sig-nificant association between CR participation and anx-iety. However, they found that the anxiety levels of the197 subjects were relatively low prior to participation.More recent research is needed to confirm olderresearch that has shown that CR patients who areanxious are less likely to adhere to CR programs.58

Social SupportPatients who have their family’s support for attend-ing a CR program are more likely to adhere to theprogram than those who do not.28 Social supporthas been identified as influential in determiningattendance at a CR program and as being moreinfluential for women than for men.59,60 Patients’perceptions of support (including the strength ofthe recommendation for CR attendance) fromhealth care providers and from family membershave also been related to attendance at CR pro-grams.24,59,61 There is also evidence that involve-ment of partners in the rehabilitation process is a



critical factor in its effectiveness, even though fewprograms include them.62

PATIENTS’ PERCEPTIONS OF THE BENEFITS OF CRSome research has examined whether patients’ per-ceptions of the barriers to and benefits of CR are relat-ed to adherence to CR. Australian research by Johnsonand Heller,63 in post-MI and angina patients found thatamong cardiac patients recently discharged followingan MI, perceptions of the benefits of and the physicalenvironment and time barriers to home exercise while inthe hospital were predictive of adherence with regularexercise 6 months after discharge. Only a few otherstudies have been conducted to determine factorsassociated with CR adherence to unsupervised exer-cise. These studies also support the findings of John-son and Heller.43,54 However, the findings of thesestudies are limited due to the inherent problems ofmeasuring compliance in unsupervised programs (see“barriers to CR exercise programs”).

Hellman41 found that the perceived benefit ofexercise was a predictor of adherence to exercise.Pender et al.42 also found that the perceived bene-fits of exercise, perceived barriers to exercise, andperceived self-efficacy for exercise were predictiveof CR exercise adherence and were significantly dif-ferent between participants initiating and thosemaintaining a CR exercise program. Kelly et al.65

found that the strongest predictors of lifestylechanges following a cardiac event were perceivedbenefits associated with CR, as well as self-efficacy.

Although further longitudinal research using prospec-tive data is needed to confirm these findings, the aboveresearch suggests that psychological variables of per-ceived benefits of exercise/CR programs and perceivedbarriers to exercise/CR programs predict adherence.

GENDER AS A BARRIER TO CRStudies on rehabilitation outcome reveal poorer pro-gram uptake, poorer adherence, and significantly high-er drop-out rates for women than for men.66 Afterhospitalization, only 20% of all patients entering struc-tured rehabilitation programs are women, despite thefact that 40% of coronary events occur in women.67–69

It is of concern that women have lower participationrates, given that women seem to be at higher risk ofpsychosocial impairment than men following an MI,including such problems as psychosomatic and func-tional complaints, more sleep disturbances, greateranxiety and depression, and a generally lower quality oflife than men.27,46,70,71

Moore et al.72 found that of the 40 women whoparticipated in an exercise program following an acutecardiac event, although 83% started exercising during

the first month, after 1 month, one third of the partici-pants had stopped exercising, while only 50% werestill exercising 3 months later. Although the findingsare limited by the convenience nature of the sampleand relatively small sample size, this study is to becommended for its objective measures of exercise toprevent bias in self-reported exercise levels. Moore etal.72 found that most of the women were exercisingwell below the recommended guidelines for exerciseafter an acute cardiac event. One possible explana-tion for this finding is that cardiac events generallyoccur later in women than in men, and they tend to have more severe pathology with poorer prog-noses. Participation in CR decreases with age andwith patients with more severe pathology.69 Anotherreason is that women are known to have more diffi-culty traveling to and consistently attending CR pro-grams.59 An additional explanation is that women tendto be referred to CR less often than men, despite asimilar clinical profile and improvement in functionalcapacity compared to men.69

A further possible reason for the lack of participa-tion of women in CR programs is that perhaps manywomen do not find the CR programs relevant to their needs. For instance, psychosocial factors suchas depressive symptoms and perceived social sup-port have been found to be related to the severity ofpain symptoms in women following an MI.73 The impli-cation is that women may find a CR program moreappealing if there is a strong psychological emphasis,rather than exercise being the main focus, as is cur-rently the case.36,74 Female patients have been identi-fied as less efficacious and less able to toleratephysical activity than men.68,75,76 This may partly explainwhy women do not participate in CR; that is, the physi-cal components of CR and women’s perception of theirphysical abilities may hinder their participation in CR.

More research is needed to investigate reasons fornot participating, since most of the research intoadherence to CR has been conducted in men.

BARRIERS TO CR EXERCISE PROGRAMSSeveral studies have examined only the exercise com-ponent of CR programs to determine factors associat-ed with participation or adherence. Some of thesestudies have been discussed above (see “patients’ per-ceptions of the benefits of CR”). Noncompliance withthe exercise component of CR is a particular problembecause of the extended period of exercise mainte-nance required before benefits are achieved.77 TheSTAGES model (Stages of Change in Exercise Adher-ence behavior), a process-oriented model, has beenapplied to many studies on adherence to CR exerciseto better understand exercise adherence behavior.41

This model postulates that individuals progress through

distinct stages as they initiate and maintain exercise. Ateach stage, the predictors of exercise adherence areused in varying degrees.41 As discussed above (under“self-efficacy”), Hellman found perceived self-efficacypredictive of exercise adherence. Exercise time signifi-cantly increased with each subsequent stage of exer-cise adherence, from the precontemplation stage to themaintenance stage.41 While this study is to be com-mended for its random selection of participants, theissue of self-reporting bias limits the findings (seebelow in “limitations of studies”).

Generally, research has found that subjects who donot adhere to CR exercise programs include thosewho are older60,61; those who lack spousal or familysupport41,78; those with lower income and educa-tion61; and those of lower socioeconomic and poorpsychological status.79,80 The findings concerninginactive lifestyle are mixed, with some studies findingthat adherence to CR exercise programs was relatedto an inactive lifestyle (often characterized by a job withlow energy demand and sedentary leisure pursuits)and poor previous exercise habits,19,79 and others notfinding a relationship.20,63,64

Mitchell et al.81 found that individuals who consis-tently participated in a CR exercise program did not usestrategies to overcome barriers; rather, in their quest tosurvive, they used strategies to ensure their participa-tion in the program. Their quest to survive became thedriving force in their lifestyle modification. However, thishaving been a qualitative study with only six subjects,the findings cannot be generalized.

One area of research into adherence to CR exer-cise programs that particularly needs investigatingis whether adherence to supervised exercise pro-grams differs from unsupervised, home-based pro-grams. As discussed by Johnson and Heller,63

budget restraints38 have led to an increase in thenumber of unsupervised programs. However, apartfrom their lack of empiric evaluation, another prob-lem in measuring adherence to these programs isthat it is difficult to measure compliance with theseprograms due to the reliance on self-reports, whichmay be biased and inaccurate. More creative waysof solving this problem are needed in order to eval-uate adherence to these programs.

BARRIERS TO MAINTENANCE OF DIET RECOMMENDATIONS OF CR PROGRAMSKoikkalainen et al.82 found that the main reasonswhy cardiac patients did not follow nutritionists’advice during CR were related to certain situations,such as eating in company or having too much foodavailable, and to the perception that healthy foodsdo not taste good. Barnes and Terry83 found thatthe majority of men who were hospitalized for an MI

did not like the taste of the cardiac diet, and thatthey found it difficult to choose appropriate foods inthe supermarket. They also had difficulty complyingwith the diet away from home. Wright84 found thatthe majority of patients at a follow-up clinic consid-ered the recommended cardiac diet boring and stilldesired foods not prescribed in their diet. Lloyd etal.85 found that inferior quality of food taste was oneof the largest barriers to adherence to diet recom-mendations, while the greater cost and lack of fam-ily support also hindered adherence. More recentstudies are also needed in this area to determinewhether such barriers as taste, cost, and availabilityof low-fat foods in supermarkets are as significant aproblem as these studies indicated, given the surgein the last few years in the range and quality of low-fat products available to consumers.

While many of the above studies have methodolog-ic problems in terms of limited sample sizes, togetherthese findings suggest that CR programs may need tofocus on teaching social skills necessary to manageeating in social situations, and on encouraging familymembers to actively support the CR participant inadherence to the diet. However, an Australian study86

found that most nutrition education programs focusedon knowledge-based objectives rather than those relat-ed to behavior change. Reported barriers to the deliv-ery of nutrition education included the venue, variabilityof participants' needs, lack of funding, and time limita-tions. This study suggests that further resources areneeded to adequately address dietary issues in manyCR programs.

OTHER SOCIODEMOGRAPHIC AND MEDICAL VARIABLESSmokingAlthough evidence is accumulating that smoking cessa-tion may be the most important predictor of mortalityamong MI survivors,87 there is little evidence that smok-ing cessation is associated with participation in CR.53

Further research is needed to investigate this issue,especially as the study of Conn et al.53 was retrospec-tive and used only percentage attendance as a measureof adherence.

Marital and Socioeconomic StatusHiatt et al.88 investigated factors influencing patiententrance into a CR program and found that demo-graphic variables, such as marital status and income,predicted participation in CR, with wealthier and mar-ried patients more likely to participate. Marital statusmay be linked to social support, with encouragementfrom spouses possibly a factor in CR participation.These findings suggest that unmarried MI patientsmay need additional professional support in terms of



encouragement to participate and adhere to CR pro-grams. Although cost effectiveness of CR in terms ofimproving health outcomes and preventing morbidityhas been demonstrated,15,89 the financial cost to thepatient, particularly in the United States, where lack ofinsurance reimbursement has been reported,90 maybe another factor related to poor participation in CRprograms. Alternatives such as home-based or cost-modified protocols that have been developed mayhelp reduce this problem.

ConvenienceVarious studies25,59,72,91,92 have shown that conve-nience factors, such as distance and availability oftransportation, influence patients’ participation in CR.Women and elderly people are known to have difficultytraveling to and consistently attending CR programs.11

Ades et al.24 found that commuting time was predictiveof participation in CR. Participants were more likely tolive closer to the facility and to own and drive a car thannonparticipants. Johnson and Heller63 found that par-ticipants who believed that there were few practical(physical environment and time) barriers to participationwere more likely to adhere to home exercise 6 monthsafter discharge. De Busk16 reported that many patientsprefer home-based to group-based exercise trainingbecause of the greater convenience of the former. Healso noted that patients undergoing group-based train-ing spent more time traveling to and from the programthan patients in the home training program spent ontheir exercise bikes. Home-based, unsupervised exer-cise programs may overcome some of these problems.

Functional CapacityHarlan et al.93 found that patients with a lesser degreeof functional capacity were less likely to enroll in CR. Thismay be related to the above discussion about reasonsfor lack of referral by a physician, with physicians lesslikely to refer patients with severe impairment followingan MI. Burns et al.29 reported that individuals with poorfunctional capacity and poor exercise habits before thecardiac event were less likely to adhere to CR. Howev-er, Ades et al.24 did not find markers of the severity ofcardiac disease to be predictive of participation in CR.Further research is needed to clarify the role of function-al capacity in CR participation.

Treatment Advances Reduce the Need for CRAnother possible reason for the decline in CR partici-pation may relate to more effective therapies than pre-viously existed, precluding the need for CR, particularlythe exercise component.16 Thrombolytic therapy andmechanical revascularization at the onset of an MI havereduced the mortality rate to the extent that demon-strating the value of exercise training would require

studies involving many patients.16 It is argued that thenewer therapies diminish the potential of exercise train-ing to lower mortality still further, and that exercisetraining designed to facilitate weight management andenhance psychological status and physical endurancecan be provided in a home setting to a number ofpatients recovering from MI.

LIMITATIONS OF STUDIESMany of the above studies were cross-sectional innature, so conclusions as to the direction of theobserved associations are not possible. While cross-sectional research allows factors associated with par-ticipation or adherence to be identified, it is importantfor clinicians to know what factors predict participa-tion or adherence, so that these patients can be tar-geted to emphasize the benefits of CR, or to preventdrop-outs from occurring. Furthermore, in retrospec-tive studies, problems with the validity of measuresarise (especially with self-reporting), as well as diffi-culties in predicting variables associated with out-comes. Ideally, studies designed to predict factorsassociated with participation or adherence to CRshould be prospective and longitudinal.

Another problem with most of the studies reviewedis that the selection of participants was not random-ized. Many of the subjects were selected for thesestudies on the basis that they volunteered to partici-pate in the CR program, and they were not comparedwith a control group matched for variables such asage and gender. Thus, the subjects may not be rep-resentative of the population of potentially eligiblepatients. It has been argued that volunteers are morelikely to comply with exercise programs because oftheir expressed desire to participate.79 Systematicresearch is needed to look at not only patients whoinconsistently attended or dropped out of programs,but also patients who were never referred to or chosenot to enroll in programs. If it is not possible to ran-domize the selection of participants, it is important toalso study a matched cohort to compare differencesbetween groups. Another factor related to sampleselection in these studies is the homogeneous natureof many of the samples. For example, some studiesassessed males only, others used patients over 75years of age or excluded those over 65, and someassessed male war veterans only. Furthermore, sam-ple sizes were often not adequate to provide sufficientpower to detect significant findings.

One of the problems with relying on participants’evaluations of adherence (as is the case for unsuper-vised programs) is that the actual process of record-ing frequency and duration of exercise can in itselfincrease adherence (Tooth et al.64). This is a particu-lar problem in evaluating unsupervised programs,

where compliance is more difficult to measure. Anoth-er problem with relying on participants’ responses isthat there may be a self-reporting bias, with partici-pants wishing to respond in a socially desirable man-ner. There is also controversy regarding the extent towhich individuals inflate their amount of exercise whenself-reporting.94

Another liability in comparing studies in this area con-cerns differences in measures of adherence. Somestudies use only the percentage of sessions attendedas a measure, while others take into account suchissues as whether the participant actually followed theguidelines and participated effectively in the programwhen attending formal sessions. The problem with rely-ing on measures of attendance at organized sessions ordiscontinuation of formal program attendance is thatsome of the dropouts may continue to exercise on theirown, but because of the lack of follow-up, this informa-tion is not available. Some investigators attempt to over-come subjective self-reports of exercise by objectivelymeasuring physiologic exercise indices (Moore et al.72).

In summary, prospective, longitudinal studies, withadequate sample sizes and randomly selected sub-jects with matched controls, are needed to provideempiric evidence for factors predictive of participationin or adherence to CR programs.

CR is a proven, effective intervention that should beoffered routinely to all those who are likely to benefit.Unfortunately, the focus (in practice and in research) hasbeen primarily on men who have had a first and uncom-plicated MI. Certain groups are neglected by either notbeing offered or not taking up rehabilitation. Theseinclude women, elderly people, ethnic minority groups,and individuals who live in rural areas. Also, people whohave heart failure and angina are neglected.

There is a need to make CR more widely availableand to integrate it into the secondary prevention of com-prehensive cardiac care. A flexible approach to rehabil-itation is warranted in order to enhance adherence andparticipation. Consideration should be given to modesof delivery (home vs. hospital-based) that are based onan individual assessment of need. There should be rou-tine monitoring and evaluation of participation.

FACILITATING PARTICIPATION AND ADHERENCEAlthough there are methodologic problems withresearch into participation and adherence to CRprograms, some recommendations arise from thereview of the above research and views of cliniciansworking in CR.

• It has been suggested that a history of minimalexertion is likely to cause an initial reluctance to participate in exercise. However, someresearchers have found that if initial hesitancy

can be overcome, it may not be a large barrierto exercise maintenance.95 Thus, health careprofessionals have a role in assisting thepatient to overcome hesitancy, through educa-tion and reassurance.

• Related to the above point is that participation inCR programs may be enhanced by marketing ofCR programs by health professionals and organiza-tions. In particular, emphasis should be placed onadvertising CR programs as multiple risk reductionprograms so that all lifestyle factors are addressed,rather than emphasizing only exercise or diet.

• To improve the problem of lack of referral byphysicians, it is recommended that mechanismsto enhance doctors’ adherence to practice guide-lines continue to be investigated. King and Teo11

suggest that particular areas for investigationinclude issues regarding consensus building; crit-ical path development, implementation, and evalu-ation; and the streamlining or automating ofreferral mechanisms for CR programs. CR staffshould ensure that referring physicians do nothave an inherent bias against referring certainpatients to CR (e.g., older patients and those ofpoor functional status).

• CR patients’ perceptions of the barriers to andbenefits of CR may be modifiable sources of non-adherence, through education and encourage-ment by health professionals.

• The finding that psychosocial variables play dif-ferent roles in the recovery paths of men andwomen73 suggests that CR programs would bemore effective with gender-specific tailoring.

• One of the recent trends in CR is the developmentof home rehabilitation programs. Advantages ofhome CR are that it addresses the problem ofaccessibility and reduces hospital costs. This canbe particularly beneficial for low-risk patients.

• Greater resources need to be allocated to diet inCR programs in order to facilitate behaviorchange in patients.

• Patients with known risk factors for nonadher-ence or nonparticipation (e.g., older individuals,women, and those with low self-efficacy, motiva-tion, and social support) should be targeted earlyafter MI and educated about the benefits of CR.These patients should be referred for formal CR,since the likelihood of maintenance is greaterthan in unsupervised programs.

• Rehabilitation services should be flexible enoughto cater to individual personalities, varying func-tional capacities, and gender differences.

• Ice21 suggests that post-MI patients enter a CRprogram no later than 3 weeks after the event, ashe found this to improve compliance rates.



CONCLUSIONThe problem of poor participation and adherence toCR programs following an MI is well documented,and this literature review has resulted in some suggestions for reducing the barriers to participa-tion/adherence. While several studies have investi-gated factors associated with or predictive ofparticipation or adherence to CR programs, veryfew were randomized, controlled studies. Further-more, many measures of outcome are subject tobias. These factors, along with differing measuresof adherence/compliance, make it difficult to com-pare the study outcomes.

However, some consistent findings have emergedfrom these studies. One is that factors strongly relatedto poor participation or adherence include the strengthof the physician’s referral, being female, being older,having a lower education status, and having a poorfunctional capacity. Several other factors have alsobeen isolated; barriers to participation/adherence aregenerally multifactorial. Given the increasing trendtoward unsupervised CR exercise programs, the chal-lenge remains for researchers to conduct randomized,controlled studies that use valid and reliable measuresof adherence in order to determine what the barriersare to participation/adherence to CR programs.

REFERENCES

1 Julian D. Medical background to cardiac rehabilitation. In:Jones D, West R, eds. Cardiac Rehabilitation. London:BMJ; 1995:1–30.

2 Thompson D, Bowman G. An Audit of Cardiac RehabilitationServices in England and Wales. Leeds: NHS Executive; 1995.

3 O’Connor G, Buring J, Yusuf S, et al. An overview of ran-domized trials of rehabilitation with exercise after myocardialinfarction. Circulation. 1989;80(2):234–244.

4 Oldridge N, Guyatt G, Fischer M, et al. Cardiac rehabilita-tion after myocardial infarction. JAMA. 1988;260:945–950.

5 Linden W, Stossel C, Maurice J. Psychosocial interventionsfor patients with coronary artery disease. A meta-analysis.Arch Intern Med. 1996;156:745–752.

6 Dusseldorp E, van Elderen T, Maes S, et al. A meta-analysisof psychoeducational programs for coronary heart diseasepatients. Health Psychol. 1999;18:506–519.

7 Feigenbaum D, Carter E. Cardiac Rehabilitation Services.Health Technology Assessment Report. Washington, DC:National Center for Health Services Research and HealthCare Technology Assessment, Public Health Service, USDepartment of Health and Human Services; 1988.

8 Wenger N, Froelicher E, Smith L, et al. Cardiac Rehabilita-tion: Clinical Practice Guideline No. 17. Rockville, MD: USDepartment of Health and Human Services, Public HealthService, Agency for Health Care Policy and Research, andthe National Heart, Lung, and Blood Institute; 1995.AHCPR Publication No. 96–0672.

9 Cay E. Goals of rehabilitation. In: Jones D, West R, eds.Cardiac Rehabilitation. London: BMJ; 1995:31–53.

10 Pashkow F. Issues in contemporary cardiac rehabilitation: a his-torical perspective. J Am Coll Cardiol. 1993;21:822–834.

11 King K, Teo K. Cardiac rehabilitation referral and attendance: notone and the same. Rehabil Nurs. 1998;23(5):246–251.

12 Balady G, Fletcher B, Froelicher E, et al. Cardiac rehabilitationprograms: a statement for health care professionals from the

American Heart Association. Circulation. 1994;90:1602–1610.13 Dickson H, Fastier G, Oldenburg B, et al. Effectiveness of

cardiac rehabilitation: a controlled trial. In: Proceedings ofthe Seventh Annual Scientific Meeting: The Assessmentand Achievement of Fitness. Canberra, Australia, March26–28, 1987. Australian College of Rehabilitation Medi-cine, Ryde, NSW. Australian College of Rehabilitation Med-icine; 168–177.

14 Taylor C, Houston Miller N, Ahn D, et al. The effect of exer-cise training programs on psychological improvement inuncomplicated postmyocardial infarction patients. J Psy-chosom Res. 1986;30:581–587.

15 Levin L, Perk J, Hedback B. Cardiac rehabilitation—costanalysis. J Intern Med. 1991;230:427–434.

16 De Busk R. Why is cardiac rehabilitation not widely used?West J Med. 1992;156:206–208.

17 Bunker S, McBurney H, Cox H, et al. Identifying participationrates at outpatient cardiac rehabilitation programs in Victoria,Australia. J Cardiopulm Rehabil. 1999;19(6):334–338.

18 Oldridge N, Streiner D. The health belief model: predictingcompliance and dropout in cardiac rehabilitation. Med SciSports Exerc. 1990;22:678–683.

19 Oldridge N. Compliance with cardiac rehabilitation services. J Cardiopulm Rehabil. 1991;11:115–127.

20 Radtke K. Exercise compliance in cardiac rehabilitation.Rehabil Nurs. 1989;14(4):182–195.

21 Ice R. Long-term compliance. Phys Ther. 1985;65(12):1832–1839.22 Oldridge N. Compliance and dropout in cardiac exercise

rehabilitation. J Cardiopulm Rehabil. 1984;4:166–177.23 Oldridge N. Patient compliance. In: Pollock N, Schmidt D,

eds. Heart Disease and Rehabilitation. 3rd ed. Champaign,IL: Human Kinetics; 1995.

24 Ades P, Waldmann M, McCann W, et al. Predictors of car-diac rehabilitation in older coronary patients. Arch InternMed. 1992;152:1033–1035.

25 Ades P, Waldmann M, Guillespie C. A controlled trial ofexercise training in older coronary patients. J Gerontol ABiol Sci Med Sci. 1995;50A(1):M7–M11.

26 Cannistra L, Balady G, O’Malley C, et al. Comparison of theclinical profile and outcome of women and men in cardiacrehabilitation. Am J Cardiol. 1992;69:1274–1279.

27 Conn V, Taylor S, Wiman P. Anxiety, depression, quality oflife, and self-care among survivors of myocardial infarction.Issues Ment Health Nurs. 1991;12:321–333.

28 Yates B, Skaggs B, Parker J. Theoretical perspectives onthe nature of social support in cardiovascular illness. J Car-diovasc Nurs. 1994;2(1):1–15.

29 Burns K, Camaione D, Froman R, et al. Predictors of refer-ral to cardiac rehabilitation and cardiac exercise self-effica-cy. Clin Nurs Res. 1998;7(2):147–163.

30 Hlatky M, Cotugno H, Mark D, et al. Trends in physicianmanagement of uncomplicated myocardial infarction: 1970to 1987. Am J Cardiol. 1988;61:515–518.

31 Greenland P, Chu J. Efficacy of cardiac rehabilitation ser-vices. In: American College of Physicians. Clinical PracticeGuideline. Philadelphia, PA: American College of Physi-cians; 1995:II25–II39 [Reprinted from Ann Intern Med.1988;109:650–663].

32 Pashkow F. Cardiac rehabilitation: not just exercise any-more. Clev J Clin Med. 1996;63:116–123.

33 Smith S, Blair S, Criqui M, et al. AHA medical/scientificstatement: preventing heart attack and death in patientswith coronary artery disease. Circulation. 1995;92:2–4.

34 Bittner V, Sanderson B, Taylor H, et al. Referral patterns toa university-based cardiac rehabilitation program: how wellare we following AHCPR guidelines [abstract]? Can J Car-diol. 1997;139(suppl B):188B.

35 Pathman D, Konrad T, Freed G, et al. The awareness-to-adherence model of steps to clinical guideline compliance.Med Care. 1996;34:873–889.

36 Dafoe W, Huston P. Current trends in cardiac rehabilitation.Can Med Assoc J. 1997;156(4):527–587.

37 Wolfe L, Dafoe W, Henderson-Roberge E. Cardiovascular


rehabilitation in Ontario (Canada). J Cardiopulm Rehabil.1990;10:130–140.

38 National Heart Foundation NSW Division. Exercise for Peo-ple With Heart Disease: Guidelines for the Prescriptionand Conduct of Non-Medically Supervised, Community-Based Exercise Programs. New South Wales: The Founda-tion; 1995.

39 Bandura A. Social Foundations of Thought and Action: aSocial Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall; 1986.

40 Oldridge N. Cardiac rehabilitation exercise programme:compliance and compliance-enhancing strategies. SportsMed. 1988;6:42–55.

41 Hellman E. Use of the stages of change in exercise adher-ence model among adults with a cardiac diagnosis. J Car-diopulm Rehabil. 1997;17:145–155.

42 Pender N, Walker S, Sechrist K, et al. The Health Promo-tion Model: Refinement and Validation. Washington, DC:US Public Health Service; 1990.

43 Robertson D, Keller C. Relationships among health beliefs,self-efficacy, and exercise adherence in patients with coro-nary artery disease. Heart Lung. 1992;21:121–127.

44 Vidmar P, Robinson L. The relationship between self-effica-cy and exercise compliance in a cardiac population. J Car-diopulm Rehabil. 1994;14:246–254.

45 Ben-Sira Z. The structure of readjustment after heart attack.Soc Sci Med. 1990;30:523–536.

46 Lavie C, Milani R. Effects of cardiac rehabilitation and exer-cise training on exercise capacity, coronary risk factors,behavioral characteristics and quality of life in women. Am JCardiol. 1995;75:340–343.

47 Bandura A. Self-efficacy: Exercise of Control. New York,NY: W Freeman; 1977.

48 Rhodes R, Morrissey M, Ward A. Self-motivation: a drivingforce for elders in cardiac rehabilitation. Geriatr Nurs.1992;13(2):94–98.

49 Robinson J, Shaver P, Wrightsman, L, eds. Measures ofPersonality and Social Psychological Attitudes. New York,NY: Academy Press Inc; 1991.

50 Johnson J, Morse J. Regaining control: the process ofadjustment after myocardial infarction. Heart Lung.1990;19:126–135.

51 McGlashan R. Strategies for rebuilding self-esteem for thecardiac patient. Dimens Crit Care Nurs. 1988;7:28–38.

52 Runion J. A program for psychological and social enhance-ment during rehabilitation after myocardial infarction. HeartLung. 1985;14:117–125.

53 Conn A, Taylor S, Casey B. Cardiac rehabilitation participa-tion and outcomes after myocardial infarction. Rehabil Nurs.1992;17:58–62.

54 Emery C. Adherence in cardiac and pulmonary rehabilita-tion. J Cardiopulm Rehabil. 1995;15:420–423.

55 Hershberger P, Robertson K, Markert R. Personality andappointment-keeping adherence in cardiac rehabilitation. JCardiopulm Rehabil. 1999;19:106–111.

56 Carney R, Rich M, Tevelde A, et al. Major depressive disor-der in coronary artery disease. Am J Cardiol. 1987;60:1273–1275.

57 Carney R, Freedland K, Eisen S, et al. Major depression andmedication adherence in elderly patients with coronaryartery disease. Health Psychol. 1995;14:88–90.

58 Blumenthal J, Williams R, Wallace A, et al. Physiological andpsychological variables predict compliance to prescribedexercise therapy in patients recovering from myocardialinfarction. Psychosom Med. 1982;44(6):519–527.

59 Moore S. Women’s views of cardiac rehabilitation pro-grams. J Cardiopulm Rehabil. 1996;16:163–168.

60 Sallis J, Hovell M, Hofstetter C. Predictors of adoption andmaintenance of vigorous physical activity in men andwomen. Prev Med. 1992;21:237–251.

61 Dishman R, Sallis J, Ornstein D. The determinants of physicalactivity and exercise. Public Health Rep. 1985;100:158–171.

62 Thompson D, Meddis R. Wives’ responses to counseling

early after myocardial infarction men. J Psychosom Res.1990;34:249–258.

63 Johnson N, Heller R. Prediction of patient nonadherencewith home-based exercise for cardiac rehabilitation: the role of perceived barriers and perceived benefits. Prev Med.1998;27:56–64.

64 Tooth L, McKenna K, Colquhoun D. Prediction of compli-ance with a post-myocardial infarction home-based walkingprogramme. Aust Occup Ther J. 1993;40:17–22.

65 Kelly R, Zyzanski S, Alemagno S. Prediction of motivationand behavior change following health promotion: role ofhealth beliefs, social support, and self-efficacy. Soc SciMed. 1991;32:311–320.

66 Brezinka V, Kittel F. Psychosocial factors of coronary heart disease in women: a review. Soc Sci Med. 1996;429(10):1351–1365.

67 Lerner D, Kannel W. Patterns of coronary heart diseasemortality in the sexes. A 26-year follow-up of the Framing-ham population. Am Heart J. 1986;111:383–390.

68 Schuster P, Waldron J. Gender differences in cardiac reha-bilitation patients. Rehabil Nurs. 1991;16:248–53.

69 Ades P, Waldmann M, Polk D, et al. Referral patterns and exer-cise response in the rehabilitation of female coronary patientsaged >62 years. Am J Cardiol. 1992;69:1422–1445.

70 Guiri E, Conroy R, Hickey N, et al. Psychological response to anacute coronary event and its effect on subsequent rehabilitationand lifestyle change. Clin Cardiol 1987;10:256–260.

71 Wilkund I, Herlitz J, Johansson S, et al. Subjective symptoms andwell-being differ in women and men after myocardial infarction.Eur Heart J. 1993;14:1315–1319.

72 Moore S, Ruland C, Pashkow F, et al. Women’s patterns of exercise following cardiac rehabilitation. Nurs Res.1998;47(6):318–324.

73 Con A, Linden W, Thompson J, et al. The psychology of menand women recovering from coronary artery bypass surgery.J Cardiopulm Rehabil. 1999;19(3):152–161.

74 Thompson DR. Cardiac rehabilitation: how can it be improved?J Psychosom Res. 1995;39(5):519–523.

75 Cochrane B. Acute myocardial infarction in women. CritCare Nurs Clin North Am. 1992;4:279–288.

76 Hamilton G. Recovery from acute myocardial infarction inwomen. Cardiology. 1990;77:58–70.

77 Andrew G, Oldridge N, Parker J, et al. Reasons for dropoutfrom exercise programs in post coronary patients. Med SciSports Exerc. 1981;13:164–168.

78 Knapp D, Blackwell B. Emotional and behavioral problemsin cardiac rehabilitation patients. J Cardiopulm Rehabil.1985;5:112–123.

79 Daltroy L. Improving cardiac patient adherence to exerciseregimes: a clinical trial of health education. J CardiopulmRehabil. 1985;5:40–49.

80 Holm K, Christman N, Ashley W. The cardiac patient and exer-cise: a sociobehavioral analysis. Heart Lung. 1985;14:586–593.

81 Mitchell R, Muggli M, Sato A. Cardiac rehabilitation: partici-pating in an exercise program in a quest to survive. RehabilNurs. 1999;24(6):236–239.

82 Koikkalainen M, Lappalainen R, Mykkanen H. Why cardiacpatients do not follow the nutritionist’s advice: barriers innutritional advice perceived during rehabilitation. DisabilRehabil. 1996;18(12):619–623.

83 Barnes M, Terry R. Adherence to the cardiac diet: attitudesof patients after myocardial infarction. J Am Diet Assoc.1991;91:1435–1437.

84 Wright C. Preliminary survey of attitudes to, and factors affecting,perceived success or failure among adults attending a hospitallipid clinic. J Hum Nutr Diet. 1994;7:153–159.

85 Lloyd H, Paisley C, Mela D. Barriers to the adoption ofreduced-fat diets in a UK population. J Am Diet Assoc.1995;95:316–322.

86 Paxton F, Ball M, Bunker S, et al. Nutrition education in out-patient cardiac rehabilitation programs. Aust J Nutr Diet.1999;56:76–80.

87 Stegman R, Miller P, Hageman R, et al. Myocardial infarction


survival: how important are patients’ attitudes and adher-ence? Am J Prev Med. 1987;3:147–151.

88 Hiatt A, Hoenshell-Nelson N, Zimmerman L. Factors influ-encing patient entrance into a cardiac rehabilitation pro-gram. Cardiovasc Nurs. 1990;26:25–29.

89 Ades P, Pashkow F, Nestor J. Cost-effectiveness of cardiacrehabilitation after myocardial infarction. J CardiopulmRehabil. 1997;17(4):222–231.

90 Carlson J, Johnson J, Franklin B, et al. Program participation,exercise adherence, cardiovascular outcomes, and programcost of traditional versus modified cardiac rehabilitation. AmJ Cardiol. 2000;86(1):17–23.

91 Gori P, Pivotti F, Mase N, et al. Compliance with cardiac rehabili-

tation in the elderly. Eur Heart J. 1984;5(suppl E):109–111.92 Moore S, Kramer F. Women’s and men’s preferences for

cardiac rehabilitation program features. J Cardiopulm Reha-bil. 1996;16:163–168.

93 Harlan W, Sandler S, Lee K, et al. Importance of baselinefunctional and socioeconomic factors for participation incardiac rehabilitation. Am J Cardiol. 1995;76:36–39.

94 Haskell W, Leon A, Casperson C, et al. Cardiovascular ben-efits and assessment of physical activity and physical fitnessin adults. Med Sci Sports Exerc. 1992;18(5):557–562.

95 Fontana A, Kerns R, Rosenberg R, et al. Exercise training forcardiac patients: adherence, fitness, and benefits. J Car-diopulm Rehabil. 1986;6:4–15.

barriers to participation in and adherence to cardiac rehabilitation programs: a critical literature...

Documents