Home > RNJ > 2007 > July/August > Predictors of Cardiac Rehabilitation Initiation

Predictors of Cardiac Rehabilitation Initiation
Linda Creadon Shanks, PhD DNP RN Shirley M. Moore, PhD RN FAAN Richard A. Zeller, PhD

This study determines the effects of 15 potential predictors on cardiac rehabilitation (CR) initiation: demographic information, measures of perceived severity, perceived susceptibility, perceived cardiac threat, social support, depression, comorbid conditions, left ventricular ejection fraction, strength of physician recommendation, and benefits and barriers. Results showed that greater strength of physician recommendation and less disease severity were significant predictors of higher levels of CR initiation; female gender was a marginally significant predictor of less CR initiation. The strength of the associations for these predictors varied. Strength of physician recommendation was the strongest predictor. This information can be used to increase the number of patients starting CR through programs designed to increase physician awareness of the importance of their recommendation, the continuing need to refer women to CR, and the need to design programs that meet women’s needs.

The goals of cardiac rehabilitation (CR) are to decrease morbidity and mortality and to improve clinical and behavioral outcomes through medically supervised exercise, lifestyle interventions, and health education (American Association of Cardiovascular and Pulmonary Rehabilitation [AACVPR], 2004). This is done through exercise training, exercise counseling, and lifestyle changes such as smoking cessation, blood lipid control, blood pressure control, weight control, diabetes management, and psychosocial management (Balady et al., 2000; Squires, Gau, Miller, Allison, & Lavie, 1990). The benefits of CR include improvement of exercise capacity (Conraads et al., 2004), improvement of skeletal muscle strength (Adams et al., 1999), decreased symptoms of heart failure (Conraads et al.; Mager, Reinhardt, Kleine, Rost, & Hopp, 2000), increased heart rate recovery after exercise (Tsai, Lin, & Wu, 2005), and decreased mortality (Lamm, Denolin, Dorossiev, & Pisa, 1982; Taylor et al., 2004). Despite these benefits, only about 11%–38% of potential candidates are referred to CR and only 11%–20% actually start (AACVPR; Wenger et al., 1995). Although some factors have been associated with CR participation, initiation remains a problem.

Factors associated with CR initiation and adherence include demographic variables (age, gender, socioeconomic status, education) (Ades, Waldmann, McCann, & Weaver, 1992; Ades, Waldmann, Polk, & Coflesky, 1992; Al-Ali & Haddad, 2004; Cooper, Jackson, Weinman, & Horne, 2002; Harlan, Sandler, Lee, Lam, & Mark, 1995; Husak et al., 2004), sociopsychological variables (social support, depression) (Ades, Waldmann, McCann, et al.; Hiatt, Hoenshell-Nelson, & Zimmerman, 1990), and physiologic variables (comorbidity, disease severity) (Harlan et al.; Oldridge et al., 1983). Strength of physician recommendation also is associated with initiation (Ades, Waldmann, Polk, et al.; Al-Ali & Haddad).

Patient perceptions, such as severity of disease or illness and susceptibility to disease or illness, have been shown to be positively related to various health behaviors such as compliance with antihypertensive medications and smoking cessation (Janz, 1988; Janz & Becker, 1984). However, these patient perceptions have not been consistently associated with initiation of or adherence to CR (Ades, Waldmann, McCann, et al., 1992; Oldridge & Streiner, 1990; Al-Ali & Haddad, 2004). Perceived high net benefits have been associated with increased CR initiation (Ades, Waldmann, McCann, et al.; Oldridge & Streiner; Al-Ali & Haddad) and cardiac exercise adherence (Muench, 1987; Robertson & Keller, 1992; Tirrell & Hart, 1980). Perceived threat of harm or loss related to cardiac disease (Bennett, 1992; Thompson, Webster, Cordle, & Sutton, 1987) is another patient perception that may influence CR initiation but has been little studied.

The purpose of the present study was to determine factors that predict CR initiation. Nurses often coordinate patient care in interdisciplinary CR programs and have contact with patients immediately after their cardiac event. Therefore, nurses can provide interventions that influence CR initiation. Knowledge about the relationships between variables found to influence CR initiation will enable nurses to develop interventions designed to promote CR initiation. Nurses can develop interventions to influence patient perceptions or design CR programs to meet their needs.


Study Setting and Sample

The setting for this study was a 537-bed midwestern urban hospital that has an outpatient CR program. Myocardial infarction (MI) and coronary artery bypass graft (CABG) surgery patients were recruited from two cardiac telemetry units at this hospital. At the study hospital, the CR nurses see all potential CR patients and refer them to the closest CR facility. The CR nurse then contacts the physician, and the physician writes the order. The length of the CR program is approximately 12 weeks but can vary based on patient acuity.

Institutional review board approval was obtained before data collection began. Data were collected by the investigator and research assistants. Research assistants were trained by observing the investigator approach the patients and collect data. Research assistants were also given written instructions. The research assistants were then observed by the investigator before collecting data independently. A convenience sample of 116 hospitalized adult patients (older than 21 years) who had an MI or CABG completed the initial questionnaire in the hospital. Characteristics of these patients are listed in Table 1. Of these 116 subjects, 97 were able to be reached for a follow-up phone call to see whether they had started CR.

Independent Variables

Demographic data were collected by a self-report questionnaire during the initial patient contact.

Perceived severity and perceived susceptibility were measured using subscales from the Health Beliefs Questionnaire (Mirotznik, Feldman, & Stein, 1995). Mirotznik and Manfre (J. Mirotznik, personal communication, July 17, 2000) have tested the psychometric properties of the perceived severity scale and found that internal consistency was good, with an alpha of .92 for coronary heart disease (CHD) rehabilitation participants and .89 for college faculty and staff. Test–retest reliability was .40 for the CHD rehabilitation participants. Discriminant validity was assessed through a comparison of health beliefs of 120 CHD rehabilitation program participants with 85 college faculty and staff. The rehabilitation participants perceived CHD to be more severe and themselves to be more susceptible to it than the college faculty and staff, as expected. Cronbach’s alpha for this study with all 10 questions included was only .55. To increase Cronbach’s alpha and improve the reliability of the instrument, only questions 1, 2, 4, and 6 were included in the total score for the perceived severity questionnaire, which increased the alpha to .66.

Internal consistency of the perceived susceptibility subscale was .86 for Veteran’s Administration Hospital CHD rehabilitation participants and .62 for noncardiac college faculty and staff (J. Mirotznik, personal communication, July 17, 2000). Test–retest reliability of the Veteran’s Administration Hospital CHD rehabilitation participants was .68. Cronbach’s alpha for this study was .69.

Perceived threat was measured by the Cardiac Event Threat Questionnaire (Bennett, Puntenney, Walker, & Ashley, 1996). Bennett found an internal consistency of the entire Cardiac Event Threat Questionnaire of .94. Construct validity was examined and supported using the Profile of Mood States (McNair et al., 1992). In this study Cronbach’s alpha was .95.

Social support was measured using the Multidimensional Scale of Perceived Social Support (MSPSS), which assesses perceived social support from family, friends, and significant others (Zimet et al., 1988). The internal consistency for this study was .98. Previous studies reported Cronbach’s alphas of .88 with adolescent psychiatric patients and .87 with university students (Kazarian & McCabe, 1991). Zimet and colleagues reported a Cronbach’s alpha of .88 when the scale was administered to university undergraduates. When the scale was initially administered to university undergraduates, the total scale test–retest reliability was .85 (Zimet et al.). Construct validity was supported in a study by Kazarian and McCabe when it was correlated with the Beck Depression Inventory (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) and the Social Desirability Subscale of Jackson’s Personality Research Form (Jackson, 1984). Kazarian and McCabe also found concurrent validity of the MSPSS when it was correlated with the Social Support Behaviors Scale (Vaux, Riedel, & Stewart, 1987).

Depression was measured using the Beck Depression Inventory-II (BDI-II) (Beck et al., 1996). In this study Cronbach’s alpha was .87. In a study by Beck and colleagues internal consistency for this instrument was good, with a coefficient alpha of .92 for psychiatric outpatients and .93 for college students. Test–retest reliability was checked on 26 Philadelphia outpatients who were administered the BDI-II twice, about 1 week apart. A significant test–retest correlation of .93 (p < .001) was found. The BDI-II was also examined for convergent and discriminant validity using correlations with scores on a number of other psychological tests. Evidence of both types of validity was provided by Beck and colleagues.

Comorbidity was measured using the Charlson Comorbidity Index (Charlson et al., 1987). The Charlson Comorbidity Index is a method to classify comorbidity to estimate risk of death from comorbid disease. The index consists of a checklist of conditions weighted according to seriousness. Validity was assessed by comparing the Charlson Comorbidity Index with the Kaplan and Feinstein (1974) method for classifying comorbidity. Both methods were significant predictors of death, and survival curves were similar for both methods (Charlson et al., 1987; Charlson, Szatrowski, Peterson, & Gold, 1994).

Disease severity was measured using the patient’s most recent left ventricular ejection fraction reported in the medical record. The left ventricular ejection fraction must have been done within the previous 3 months.

Strength of the physician’s recommendation was measured using a scale similar to that developed by Ades, Waldmann, Polk, et al. (1992). Ades and colleagues developed a visual analog scale ranging from 1 (the physician did not recommend participation) to 5 (strong recommendation). A rating of 3 would be a patient perception of a moderate recommendation. In this study, patients were questioned over the telephone 4 months after initial consent and asked about how strongly their physician recommended participation using this scale.

Perceived benefits and barriers were measured using the Exercise Benefits/Barriers Scale (Sechrist, Walker, & Pender, 1987). In this study, Cronbach’s alpha was .86. This was slightly lower than the psychometric evaluation performed by Sechrist and colleagues, when the instrument was administered to healthy adults from the community. Their internal consistency for the Exercise Benefits/Barriers Scale was .95.

Dependent Variable

CR initiation consisted of patient self-report of attendance of at least four of the first six CR sessions. This was determined through a call to the patient 4 months after the initial consent. If a patient had initiated and attended four or more of the first six sessions, the patient was categorized as initiating CR.


Characteristics of the 116 research subjects can be seen in Table 1. A logistic regression equation was calculated predicting CR initiation from the 15 potential predictors (Table 2). Two of these variables, strength of physician recommendation and disease severity, were significant predictors (p < .05); gender was a marginally significant predictor (p = .06). A logistic regression equation was then run using the three predictors of strength of physician recommendation, disease severity, and gender. In this analysis, all three of these predictors were significant (p < .05). The logistic regression equation for this analysis is presented in Table 3. The odds of initiating CR were increased with stronger physician recommendation, for those with less severe disease, and for men. The strength of association for these predictors varied. Specifically, strength of physician recommendation was the strongest predictor. Differences in strength of physician recommendation accounted for 23.3% of the variance between those who initiated CR and those who did not, left ventricular ejection fraction accounted for 4.0% of the variance, and gender accounted for 2.4% of the variance. Age, years of school completed, employment status, income, race, marital status, social support, depression, comorbidity, and perceived exercise benefits and barriers were not found to be associated with CR initiation.


Strength of physician recommendation, gender, and disease severity were found to be associated with CR initiation in this study. These findings are similar to those of Ades, Waldmann, McCann, et al. (1992), who found strength of physician recommendation to be the strongest predictor of CR participation. They found that only 1.8% of patients entered CR when the patient perceived the physician recommendation to be “not mentioned to moderately supportive” (p. 1034). When patients perceived a strong physician recommendation, the initiation rate jumped to 66%. Al-Ali and Haddad (2004) found that patients who received a physician recommendation to exercise were more concerned with health and more likely to engage in special health practices, although there were no significant differences in exercise participation between those who received a recommendation to exercise and those who were not recommended for exercise.

Gender also was a predictor of CR initiation in the present study. This is consistent with findings of numerous other studies (Ades, Waldmann, Polk, et al., 1992; Cooper et al., 2002; Harlan et al., 1995; Oldridge, Ragowski, & Gottlieb, 1992). Men’s greater participation in CR may result from gender differences in views about exercise or household and family responsibilities. Women traditionally are thought of as the caretakers of the family. Cannistra et al. (1992) found that women’s role responsibilities and their concerns for others led to less participation. Plach (2002) found that barriers to CR adherence for women were factors related to health, child care, transportation, work or time constraints, and personal preferences such as not liking exercise. Also, the literature indicates that men are more likely to get referred to CR (Ades, Waldmann, Polk, et al.). Moore and Kramer (1996) found that women may have a lower rate of participation than men because programs are designed to meet the needs of men. Features that women disliked included such things as lack of choice of exercises, lack of emotional support from staff, not enough opportunities to socialize, a “men’s club” feel to CR, feeling rushed and crowded, and being weighed. Therefore, women may not be as inclined to start CR if they feel it does not meet their needs. Further research and interventions may help to narrow the discrepancies between the participation of women and men in CR.

Disease severity (ejection fraction) was also found to predict CR initiation in this study, with lower disease severity leading to a greater likelihood of CR initiation. Although it seems reasonable that people who have better heart muscle function would feel better, perceive more energy, and get out and exercise, multiple studies have shown that left ventricular ejection fraction was not related to CR initiation (Ades, Waldmann, McCann, et al., 1992; Harlan et al., 1995). This may indicate that disease symptoms are not always related to the actual severity of disease. Patients with very severe disease may not always experience the most severe symptoms.

The findings of this study are significant for nursing science in that they confirm our understanding of the importance of the strength of physician recommendation as an influence on CR initiation (Ades, Waldmann, McCann, et al., 1992). This study also confirmed previous research findings about the influence of gender on CR initiation (Ades, Waldmann, Polk, et al., 1992; Al-Ali & Haddad, 2004; Husak et al., 2004). Men are still more likely to start CR than women, highlighting the continuing need to refer women to CR programs and design CR programs that are perceived by women as meeting their needs.

Several variables that were expected to be associated with CR initiation (age, years of school completed, employment status, income, race, marital status, social support, depression, comorbidity, and perceived exercise benefits and barriers) were not found to be associated with CR initiation in this study. It is possible that these variables do not play as great a role in CR initiation today as in the past. People may be more aware of the benefits of CR because of the increase in mass media campaigns about the importance of physical activity (Cavill & Bauman, 2004). Therefore, variables previously found to influence initiation do not have the same influence as in the past. According to the American Association of Cardiovascular and Pulmonary Rehabilitation (2004), a number of societal changes have occurred recently, such as “tobacco legislation, healthy food choices in restaurants, employee wellness programs, and local and national emphasis on successful aging programs” (p. 3). All of these societal changes raise awareness of the benefits of lifestyle changes, including CR after a cardiac event.

Findings from this study suggest several recommendations for practice. Because strength of physician recommendation was associated with CR initiation, and it has been shown previously to be one of the most significant predictors of CR initiation (Ades, Waldmann, McCann, et al., 1992), physicians need to understand the importance of their recommendation on whether a patient starts CR. Primary care physicians, cardiologists, and cardiac surgeons need to understand the numerous benefits of CR and strongly recommend CR to their patients. Educational programs about CR, its benefits, and the importance of physician recommendation can be implemented, such as a standard letter signed by the physician to recommend CR and flyers sent to physicians’ offices.

Healthcare providers have a higher level of control over recommendations to begin a CR program than over other factors, such as demographics or health beliefs. Nurses can promote CR to the patients they see in the hospital. Nurses spend more time with inpatients than the physicians and may have a large influence on a patient’s decision to begin CR. It is likely that if a physician’s recommendation is influential, recommendations by a patient’s nurse may also be significant. Multidisciplinary teams can also develop videos promoting CR that can be shown to patients before they are discharged from the hospital. Collaboration between the nurses and physicians to promote CR would be beneficial. Information that nurses identify through patient contact in the hospital regarding the patient’s willingness to start CR can be shared with the physician. This information can then be used during outpatient visits.

A recommendation for future research is to determine the effect of the recommendation of nurses on CR initiation and how this compares with the effect of physician recommendation. Also, this study should be replicated to determine whether the negative findings regarding many of the expected candidate predictors are spurious findings or are representative of the predictors of CR initiation of contemporary patients.


The purpose of this study was to examine the effects of 15 potential predictors on CR initiation. Predictors of CR initiation were strength of physician recommendation, gender, and disease severity. This information can be used to increase the number of patients starting CR through programs designed to increase physician awareness of the importance of their recommendations. Nurses and physicians also need to collaborate in their efforts to promote CR. The influence of gender on CR initiation highlights the continuing need to refer women to CR and to design programs that meet their needs.


Dissertation work was completed at Case Western Reserve University.

About the Authors

Linda Creadon Shanks, PhD DNP RN, is an assistant professor at the University of Akron. Address correspondence to her at University of Akron, 302 Buchtel Common, Akron, OH 44325 or at shanks@uakron.edu.

Shirley M. Moore, PhD RN FAAN, is a professor and associate dean for research at Case Western Reserve University.

Richard A. Zeller, PhD, is a visiting professor at Kent State University, Kent, OH.


Adams, K.J., Barnard, K.L., Swank, A.M., Mann, E., Kushnick, M. R., & Denny, D. M. (1999). Combined high-intensity strength and aerobic training in diverse phase II cardiac rehabilitation patients. Journal of Cardiopulmonary Rehabilitation, 12, 209–215.

Ades, P.A., Waldmann, M.L., McCann, W. ., & Weaver, S.O. (1992). Predictors of cardiac rehabilitation participation in older coronary patients. Archives of Internal Medicine, 152, 1033–1035.

Ades, P.A., Waldmann, M.L., Polk, D. M., & Coflesky, J.T. (1992). Referral patterns and exercise response in the rehabilitation of female coronary patients aged > 62 years. American Journal of Cardiology, 69, 1422–1425.

Al-Ali, N., & Haddad, L.G. (2004). The effect of the Health Belief Model in explaining exercise participation among Jordanian myocardial infarction patients. Journal of Transcultural Nursing, 15, 114–121.

American Association of Cardiovascular and Pulmonary Rehabilitation. (2004). Guidelines for cardiac rehabilitation and secondary prevention programs. Champaign, IL: Human Kinetics.

Balady, G.J., Ades, P. A., Comoss, P., Limacher, M., Pina, I. L., Southard, D., et al. (2000). Core components of cardiac rehabilitation/secondary prevention programs: A statement for healthcare professionals from the American Heart Association and the American Association of Cardiovascular and Pulmonary Rehabilitation writing group. Circulation, 102, 1069–1073.

Beck, A.T., Steer, R.A., & Brown, G.K. (1996). Beck Depression Inventory: Manual (2nd ed.). San Antonio: The Psychological Corporation, Harcourt Brace & Company.

Beck, A.T., Ward, C.H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561–571.

Bennett, S.J. (1992). Perceived threats of individuals recovering from myocardial infarction. Heart & Lung, 21, 322–326.

Bennett, S.J., Puntenney, P.J., Walker, N.L., & Ashley, N.D. (1996). Development of an instrument to measure threat related to cardiac events. Nursing Research, 45, 266–270.

Cannistra, L.B., Balady, G.J., O’Malley, C.J., Weiner, D.A., & Ryan, T.J. (1992). Comparison of the clinical profile and outcome of women and men in cardiac rehabilitation. The American Journal of Cardiology, 69, 1274–1279.

Cavill, N., & Bauman, A. (2004). Changing the way people think about health-enhancing physical activity: Do mass media campaigns have a role? Journal of Sports Sciences, 22, 771–790.

Charlson, M.E., Pompei, P., Ales, K.L., & MacKenzie, C. R. (1987). A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. Journal of Chronic Disease. 40, 373–383.

Charlson, M., Szatrowski, T. P., Peterson, J., & Gold, J. (1994). Validation of a combined comorbidity index. Journal of Clinical Epidemiology, 47, 1245–1251.

Conraads, V. M., Beckers, P., Vaes, J., Martin, M., Van Hoof, V., De Maeyer, C., et al. (2004). Combined endurance/resistance training reduces NT-proBNP levels in patients with chronic heart failure. European Heart Journal, 25, 1797–1805.

Cooper, A. F., Jackson, G., Weinman, J., & Horne, R. (2002). Factors associated with cardiac rehabilitation attendance: A systematic review of the literature. Clinical Rehabilitation, 16, 541–552.

Goodman, J. M., Pallandi, D. V., Reading, J. R., Plyley, M. J., Liu, P. P., & Kavanagh, T. (1999). Central and peripheral adaptations after 12 weeks of exercise training in post–coronary artery bypass surgery patients. Journal of Cardiopulmonary Rehabilitation, 19, 144–150.

Harlan, W. R. III, Sandler, S. A., Lee, K. L., Lam, L. C., & Mark, D. B. (1995). Importance of baseline functional and socioeconomic factors for participation in cardiac rehabilitation. American Journal of Cardiology, 76, 36–39.

Hiatt, A. M., Hoenshell-Nelson, N., & Zimmerman, L. (1990). Factors influencing patient entrance into a cardiac rehabilitation program. Cardiovascular Nursing, 26, 25–29.

Husak, L., Krumholz, H. M., Lin, Z. Q., Kasl, S. V., Mattera, J. A., Roumanis, S. A., et al. (2004). Social support as a predictor of participation in cardiac rehabilitation after coronary artery bypass graft surgery. Journal of Cardiopulmonary Rehabilitation, 24, 19–26.

Jackson, D. N. (1984). Personality research form. Port Huron, MI: Research Psychologists Press.

Janz, N. K. (1988). The Health Belief Model in understanding cardiovascular risk factor reduction behaviors. Cardiovascular Nursing, 24, 39–41.

Janz, N. K., & Becker, M. H. (1984). The Health Belief Model: A decade later. Health Education Quarterly, 11, 1–47.

Kaplan, M. H., & Feinstein, A. R. (1974). The importance of classifying initial comorbidity in evaluating the outcome of diabetes mellitus. Journal of Chronic Disease, 27, 387–404.

Kazarian, S. S., & McCabe, S. B. (1991). Dimensions of social support in the MSPSS: Factorial structure, reliability, and theoretical implications. Journal of Community Psychology, 19, 150–160.

Lamm, G., Denolin, H., Dorossiev, D., & Pisa, Z. (1982). Rehabilitation and secondary prevention of patients after acute myocardial infarction. Advanced Cardiology, 31, 107–111.

Mager, G., Reinhardt, C., Kleine, M., Rost, R., & Hopp, H. W. (2000). Patients with dilated cardiomyopathy and less than 20% ejection fraction increase exercise capacity and have less severe arrhythmia after controlled exercise training. Journal of Cardiopulmonary Rehabilitation, 20, 196–198.

McNair, L. D. N., Lorr, M., & Droppleman, L. F. (1992). Manual: Profile of mood states. San Diego, CA: Educational and Industrial Testing Service.

Mirotznik, J., Feldman, L., & Stein, R. (1995). The Health Belief Model and adherence with a community center-based, supervised coronary heart disease exercise program. Journal of Community Health, 20, 233–247.

Moore, S. M., & Kramer, F. M. (1996). Women’s and men’s preferences for cardiac rehabilitation program features. Journal of Cardiopulmonary Rehabilitation, 16, 163–168.

Muench, J. (1987). Health beliefs of patients with coronary heart disease enrolled in a cardiac exercise program. Journal of Cardiopulmonary Rehabilitation, 7, 130–135.

Oldridge, N. B., Donner, A., Buck, C., Andrew, G., Jones, N. L., Parker, J. O., et al. (1983). Predictors of dropout from the Ontario Exercise–Heart Collaborative Study. American Journal of Cardiology, 51, 70–74.

Oldridge, N. B., Ragowski, B., & Gottlieb, M. (1992). Use of outpatient cardiac rehabilitation services: Factors associated with attendance. Journal of Cardiopulmonary Rehabilitation, 12, 25–31.

Oldridge, N. B., & Streiner, D. L. (1990). The Health Belief Model: Predicting compliance and dropout in cardiac rehabilitation. Medicine and Science in Sports and Exercise, 22, 678–683.

Plach, S. K. (2002). Women and cardiac rehabilitation after heart surgery: Patterns of referral and adherence. Rehabilitation Nursing, 27, 104–109.

Robertson, D., & Keller, C. (1992). Relationships among health beliefs, self-efficacy, and exercise adherence in patients with coronary artery disease. Heart & Lung, 21, 56–63.

Sechrist, K. R., Walker, S. N., & Pender, N. J. (1987). Development and psychometric evaluation of the exercise benefits/barriers scale. Research in Nursing and Health, 10, 357–365.

Squires, R. W., Gau, G. T., Miller, T. D., Allison, T. G., & Lavie, C. J. (1990). Cardiovascular rehabilitation: Status, 1990. Mayo Clinic Proceedings, 65, 731–755.

Taylor, R. S., Brown, A., Ebrahim, D. M., Jolliffe, J., Noorani, H., Rees, K., et al. (2004). Exercise-based rehabilitation for patients with coronary heart disease: Systematic review and meta-analysis of randomized controlled trials. American Journal of Medicine, 116, 682–692.

Thompson, D. R., Webster, R. A., Cordle, C. J., & Sutton, T. W. (1987). Specific sources and patterns of anxiety in male patients with first myocardial infarction. British Journal of Medical Psychology, 60, 343–348.

Tirrell, B. E., & Hart, L. K. (1980). The relationship of health beliefs and knowledge to exercise compliance in patients after coronary bypass. Heart & Lung, 9, 487–493.

Tsai, S., Lin, Y., & Wu, S. (2005). The effect of cardiac rehabilitation on recovery of heart rate over one minute after exercise in patients with coronary artery bypass graft surgery. Clinical Rehabilitation, 19, 843–849.

Vaux, A., Riedel, S., & Stewart, D. (1987). Modes of social support: The Social Support Behaviors (SS-B) scale. American Journal of Community Psychology, 15, 209–237.

Wenger, N. K., Froelicher, E. S., Smith, L. K., Ades, P. A., Berra, K., Blumenthal, J. A., et al. (1995). Cardiac rehabilitation. Clinical Practice Guideline No. 17. Rockville, MD: U.S. Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research and the National Heart, Lung, and Blood Institute. AHCPR Publication No. 96-0672.

Zimet, G. D., Dahlem, N. W., Zimet, S. G., & Farley, G. K. (1988). The multidimensional scale of perceived social support. Journal of Personality Assessment, 52, 30–41.