The Current and Potential Capacity for Cardiac Rehabilitation Utilization in the United States

Transcription

1 CARDIAC REHABILITATION The Current and Potential Capacity for Cardiac Rehabilitation Utilization in the United States Quinn R. Pack, MD ; Ray W. Squires, PhD ; Francisco Lopez-Jimenez, MD, MSc ; Steven W. Lichtman, EdD ; Juan P. Rodriguez-Escudero, MD ; Victoria N. Zysek, DO, MBA ; Randal J. Thomas, MD, MS PURPOSE: Prior studies suggest that program capacity restraints may be an important reason for outpatient cardiac rehabilitation (CR) underutilization. We sought to measure current CR capacity and growth potential. METHODS: We surveyed all CR program directors listed in the American Association of Cardiovascular and Pulmonary Rehabilitation (AACVPR) database in November Respondents reported current enrollment levels, program capacity, expansion potential, and obstacles to growth. RESULTS: Of the 812 program directors in the AACVPR database, 290 (36%) completed the full survey. Respondents represented somewhat larger programs than nonrespondents but were otherwise representative of all registered AACVPR programs. Current enrollment, estimated capacity, and estimated expansion capacity were reported at a median (interquartile range) of 140 (75, 232), 192 (100, 300), and 240 (141, 380) patients annually, respectively. Using these data, we estimated that, in the year 2012, national CR utilization was 28% (min, max: 20, 38) of eligible patients. Even with modest expansion of all existing programs operating at capacity, a maximum of 47% (min, max: 32, 67) of qualifying patients in the United States could be serviced by existing CR programs. Obstacles to increasing patient participation were primarily controllable system-related problems such as facility restraints and staffing needs. CONCLUSIONS: Even with substantial expansion of all existing CR programs, there is currently insufficient capacity to meet national service needs. This limit probably contributes to CR underutilization and has important policy implications. Solutions to this problem will likely include the creation of new CR programs, improved CR reimbursement strategies, and new models of CR delivery. K E Y W O R D S capacity cardiac rehabilitation enrollment utilization Author Affiliations: Division of Cardiovascular Diseases, Department of Internal Medicine, Mayo Clinic, Rochester, Minnesota (Drs Squires, Lopez-Jimenez, Thomas, and Zysek); Department of Cardiovascular Medicine, Baystate Medical Group, Springfield, Massachusetts (Dr Pack); Tufts University School of Medicine, Boston, Massachusetts (Dr Pack); Department of Cardiology, Helen Hayes Hospital, West Haverstraw, New York (Dr Lichtman); and Department of Internal Medicine, Mount Sinai Medical Center, Miami, Florida (Dr Rodriguez-Escudero). Drs Randal J. Thomas and Steven W. Lichtman are past presidents of the AACVPR but neither received (nor currently receive) any compensation for their time and efforts. All other authors report no conflicts of interest relevant to the subject of this manuscript. Supplemental digital content is available for this article. Direct URL citation appears in the printed text and is provided in the HTML and PDF versions of this article on the journal s Web site ( ). Correspondence: Randal J. Thomas, MD, Division of Cardiovascular Disease and Internal Medicine, Mayo Clinic, 200 First St SW, Rochester, MN ( Thomas. Randal@mayo.edu ). DOI: /HCR / Journal of Cardiopulmonary Rehabilitation and Prevention 2014;34:

2 Cardiac rehabilitation (CR) is an effective secondary prevention intervention for eligible patients with known cardiovascular disease. Specifically, in patients who have suffered a myocardial infarction (MI), CR is known to improve quality of life, reduce mortality rate by 20% to 30%, and be cost-effective. 1-3 Unfortunately, the vast majority of patients do not participate in CR, 4-6 despite having the highest level of recommendation and evidence in national practice guidelines. 7 Decreasing this large quality gap is a national priority and has been the focus of recent American Heart Association presidential and scientific advisories. 8, 9 As part of this effort, referral to CR is soon to become a quality performance measure audited by Medicare and Medicaid. 10 However, there are many known obstacles to patient participation in CR. Commonly reported patient-centered modifiable factors include lack of transportation, limited social support, and competing family and employment responsibilities. 8, 11 Systemcentered obstacles include lack of referral, lack of physician endorsement, limited facilitation of referral, and limited program availability. 8, One potentially important system-centered obstacle to increased patient participation may be program capacity. This was first suggested by Gurewich et al, 15 who found that some CR centers pursued patient referral and enrollment only when CR census was low. These centers apparently preferred to match patient demand with program capacity rather than altering program capacity or offerings to meet the service needs of all eligible patients in their medical system. However, this preliminary study surveyed only 22 program directors and had limited generalizability. Consequently, we aimed to more formally assess individual program capacity and perceived obstacles to program growth, and to estimate national service utilization and potential capacity for CR. We hypothesized that limitations in program capacity would be an important restraint, and that most reported obstacles to program growth would be system-related rather than patient-centered barriers. METHODS In November 2012, we distributed a survey to all registered CR program directors in the American Association of Cardiovascular and Pulmonary Rehabilitation (AACVPR) database. The survey was anonymous and voluntary. A single reminder followed the original mailing after 2 weeks. Survey responses were collected using an online survey engine, SurveyMonkey ( ). A free health care oriented book (worth approximately $20) was distributed to 15 randomly selected participants as an incentive. This survey was reviewed and approved by the Mayo Clinic Institutional Review Board. Questions were developed by the primary author (Q.P.). They were then closely reviewed by all coauthors for clarity, face validity, and content, but no formal survey validation was undertaken. The final survey was reviewed and approved by the AACVPR Research Committee and the AACVPR Board of Directors. There were 28 content questions, which assessed program characteristics, enrollment practices, quality improvement projects, and program participation rates. The full 28-question survey will be presented elsewhere. For this article, we present data related to the current number of annually enrolled patients, estimated capacity, and estimated capacity assuming reasonable growth. We also assessed hours of program operation and obstacles to growth. The specific survey questions analyzed are found in the Appendix (see Supplemental Digital Content 1, available at ). The second survey question required participants to enter the address from which they had received the survey invitation. This allowed linkage to the AACVPR database and subsequently served as the basis for the nonresponse bias analysis. This database contains basic program information and is updated every 3 years at the time of program certification or recertification through the AACVPR. In addition to contact information, the AACVPR database contains details regarding characteristics of the individual CR programs such as program offerings, the number of program full-time equivalent (FTE) employees, referrals per year, and the estimated number of eligible patients per year in the program s service area. Some of these program data were used to validate directly comparable survey content items (such as the number of FTE staff members.) The final survey question asked CR program directors to describe their greatest obstacle(s) to improving CR participation. This question was designed for free text response, with self-responses being qualitatively grouped by theme into categories and tallied. Statistical Analysis Initial response rate and completed survey response rate were determined by dividing the total number of program directors who received a survey by the number of eligible program directors who completed the entire survey. Response bias analysis compared program characteristics between those who answered at least 1 survey content question compared with all nonresponding programs found in the AACVPR database. When Cardiac Rehabilitation Capacity / 319

3 data were missing, they were generally excluded. However, we could not distinguish missing information from a 0/null in the case of profit status, and consequently the reported percentage is likely an underestimate. Descriptive statistics were used for all variables. Chi-square test, t test, and Wilcoxon rank-sum test were used as appropriate to assess differences between survey respondents and nonrespondents. All analyses were carried out using JMP version (SAS institute, Cary, NC). National Estimates National estimates of CR program capacity were calculated by 2 separate methods. The first method used survey-derived factors to estimate the projected absolute numbers of patients attending CR. This projection was then divided by the absolute number of eligible patients in the United States on the basis of national estimates. 16 The second method used published national participation rates as the baseline 4, 6 and then used survey-derived proportions to estimate national capacity for provision of CR services. No single source of national statistics could be found, which gave a precise estimate of patients who qualify for CR each year in the United States. Consequently, we estimated this number from national statistics. 16 We first included patients with qualifying procedures such as percutaneous coronary intervention (PCI) ( ; page e229), coronary artery bypass graft surgery (CABG) ( ; page e230), heart valve surgery (85 000, pages e206 and e232), and heart transplantation (2300; page e321) for a total of patients. We then added patients with MI and incident stable angina but had to make assumptions about the likely overlap of patients in > 1 of the diagnostic categories (ie, a patient with MI who also underwent PCI or CABG). As a result, we estimated that 40% ( of ; page 186) of MIs and 40% ( of ; page e186) of recurrent MIs 17 qualified for CR, although they were not revascularized with either PCI or CABG. We then estimated that 60% of incident angina cases ( of ; p e194) were not revascularized and qualified for CR. Finally, we subtracted all patients who died ( ; p e9) for an estimated total of 1.25 million eligible patients for CR services each year in the United States. 16 Proportional increases from current CR enrollment levels were calculated by comparing a CR program s current/baseline annual enrollment to (1) their estimated current capacity; and (2) their capacity assuming reasonable growth and increased resource allocation. We then projected these proportional increases to (1) all AACVPR programs (adjusting for the number of FTEs reported in the survey); and (2) all CR programs in the United States, regardless of AACVPR registration. 18 We further used published national participation rates 4, 6 and the estimated annual number of patients treated in CR compared with the current annual number of eligible patients from national statistics. 16 For the national projection, we assumed that most non-aacvpr registered programs were considerably smaller and had approximately 60% of the program size when compared with registered AACVPR programs. For the total number of nonregistered CR programs, we used a previously published ratio of AACVPR registered to nonregistered programs of 27% (709/2621) 18 to estimate the current national total of CR programs. To test our level of precision under alternative scenarios (ie, sensitivity analysis), we varied our assumptions and assessed the impact of these assumptions on the main outcome. We varied the percentage of nonrevascularized angina from 40% to 80%, nonrevascularized recurrent and incident MI from 35% to 45%, the ratio of AACVPR registered programs to nonregistered CR program from 24% to 30%, and the relative nonregistered CR program size from 40% to 80%. For each variable, we combined the ranges that would give the lowest and highest possible results and reported those as the minimum and maximum around our best estimate. RESULTS We found and surveyed 823 unique CR program directors who were associated with 884 AACVPR registered programs. There were 11 invalid addresses; 334 program directors (41%) answered at least 1 content question. In total, we received completed surveys from 290 programs for a response rate of 36% (290/812, see Figure 1 ). As a whole, the group was composed of 40% CR program directors, 54% CR program managers, and 6% CR program staff members. Of the 334 survey participants, 57 addresses could not be matched to a recognized AACVPR program. We presumed that some program directors forwarded this survey to program managers or knowledgeable program staff members, thus accounting for the 57 unmatched addresses. No surveys were returned from programs in the states of Alaska, Hawaii, Kentucky, New Mexico, or Nevada. Table 1 compares the characteristics of respondents and nonrespondents as found in the AACVPR survey database. Survey respondents were generally representative of the entire AACVPR program membership. Importantly, respondents represented somewhat larger programs (4.60 ± 3.2 vs 4.0 ± 2.8 FTE, respectively; 320 / Journal of Cardiopulmonary Rehabilitation and Prevention 2014;34:

4 Figure 1. Survey response flow diagram and response bias comparison groups. The nonresponse bias analysis was based on program characteristics found in the American Association of Cardiovascular and Pulmonary Rehabilitation (AACVPR) database. Unknown participants were people for whom the entered did not match the AACVPR database for program directors. P =.01) In addition, although respondents represented programs with more eligible patients per year, other measures of program size, such as the number of hospital beds and referrals per year, were not statistically different. We compared the answers of survey respondents to their equivalent answers as found in the AACVPR database. Survey respondents reported 4.64 ± 3.3 FTE and a geographic regional breakdown of 18%, 39%, 33%, and 10% for Northeast, Midwest, South, and West regions, respectively. These measures were similar to the answers found for the same survey respondents in the AACVPR database ( P =.86 for FTE and P =.50 for geographic regions). However, program availability was 60.0%, 99.7%, and 82.6% for phase I, II, and III CR, respectively, which was significantly different from the AACVPR database ( P <.001 for all 3 comparisons, see Table 1 ). It is unclear whether this difference reflects a true difference between survey respondents and nonrespondents or whether the difference is simply a reflection of missing data in the AACVPR database. Hours and days of CR program operation are shown in Figure 2. All programs offered CR sessions at least 3 times per week. Only 2 of 282 ( < 1%) programs offered CR sessions on Saturday, and no CR programs offered CR sessions on Sundays. Hours of operation ranged from 12 to 82 hours per week, with 61.2% of programs offering sessions starting prior to 8 AM and 16.5% offering sessions starting after 5 PM. Estimates of current utilization, capacity, and expansion capacity for CR services in the United States are shown in Table 2. For this analysis, only 252 respondents answered all relevant questions. Survey respondents reported enrolling a total of patients in their CR programs per year. Programs estimated that they could increase services by 70 ± 82 patients per program if they ran at current maximal capacity, an increase of 33% (95% CI, 27.6%-40.0%). Similarly, programs estimated that, if they were given reasonable resources to expand, they could increase services by 112 ± 118 patients per program, an increase in CR patient enrollment of 68% (95% CI, 59%-77%). Higher hours of operation were marginally predictive of higher estimated expansion capacity ( r = 0.14; P =.03), but no relationship was seen when accounting for baseline program size ( r = 0.05; P =.45). Program directors reported, on average, 1.3 ± 0.9 obstacles to enrolling patients in their CR programs. Cardiac Rehabilitation Capacity / 321

5 Table 1 Survey Nonresponse Bias Analysis From the AACVPR Database a Nonrespondents (n = 602) Respondents (n = 282) P Values b Program age, y a 23 ± 9 23 ± 8.98 Referrals per year, n a 246 ± ± Full-time equivalent staff a 4.0 ± ± For profit status, % Geographic region.59 Northeast, % 106 (17.6) 51 (18.1) Midwest, % 260 (43.2) 110 (39.0) South, % 174 (28.9) 93 (33.0) West, % 62 (10.3) 28 (9.9) n = 368 n = 194 Program characteristics Clinic-based, n (%) 15 (4.1) 12 (6.2).23 Hospital-based, n (%) 346 (94.0) 181 (93.3) Other, n (%) 7 (1.9) 1 (0.5) Number of hospital beds.42 < 25, n (%) 57 (15.5) 33 (17.0) 26-49, n (%) 13 (3.5) 8 (4.1) , n (%) 128 (34.8) 54 (27.8) > 200, n (%) 170 (46.2) 99 (51.0) Eligible patients/year c.04 < 200, n (%) 188 (51.1) 77 (39.7) , n (%) 155 (42.1) 100 (51.6) > 500, n (%) 25 (6.8) 17 (8.8) Program offerings Phase I, n (%) 153 (41.6) 83 (42.8).76 Phase II d, n (%) 356 (96.8) 180 (92.8).03 Phase III, n (%) 243 (66.0) 119 (61.3).27 Abbreviation: AACVPR, American Association of Cardiovascular and Pulmonary Rehabilitation. a Continuous values expressed as mean ± SD. b Statistical significance was set at P <.05. c Based on estimated number of patients eligible for CR in the program s service area as reported in the AACVPR database. Data was not used in national projection calculations. d When respondents were asked about program offerings in the survey, 99.7% (281/282) reported offering phase II cardiac rehabilitation. See text for further details. Obstacles to program growth are shown in Figure 3. As shown, the majority (88%) of reported obstacles were structural, administrative, or financial barriers. Staffing needs, facility restraints, and insurance issues/co-pays were the most commonly reported obstacles to program growth. Finally, when asked to rate how difficult a hypothetical 20% increase in program size would be, 47% felt this would be very easy or somewhat easy, 36% felt this would be possible but with some problems, and 17% felt this would be difficult, or very difficult/impossible. National Projections On the basis of our sample of patients in the CR programs we surveyed, we estimated national CR participation to be approximately patients annually 322 / Journal of Cardiopulmonary Rehabilitation and Prevention 2014;34:

6 Figure 2. Hours of operation for current cardiac rehabilitation programs. Hours per week were determined as 10 to 19 hours in the 10 category, 20 to 29 hours in the 20 category, and so forth. These findings suggest that national capacity could be increased simply by expanding service hours. However, no such correlation was seen. Although higher hours of operation were marginally predictive of higher estimated expansion capacity ( r = 0.14; P =.03), no such relationship was seen when baseline program size was accounted for ( r = 0.05; P =.45). (see Table 2 ). We further estimated the national percentage of eligible patients who participated in CR for the year 2012 to be 28% ( /1.25 million) with estimates for maximal current CR program capacity and for the expanded CR program capacity to be 37% and 47%, respectively. Other estimates are additionally reported in Table 3 but generally show significant underutilization and important capacity limits. The sensitivity analyses suggested a substantial range in the potential utilization rates. Nevertheless, even in the most optimistic models, we found a maximum current utilization rate of 38% and a maximal expansion capacity of 67%, leaving even in the best-case scenario of universal program expansion -approximately 33% of eligible patients unable to participate in CR due to program capacity restraints. DISCUSSION In this national survey, we found that, even using the most optimistic scenarios of significant program Table 2 Current Utilization, Capacity, and Estimated Expansion Capacity Among US Cardiac Rehabilitation Programs Current CR Utilization Current CR Capacity Expansion CR Capacity Survey response, n = 252 Median patients/program, n (IQ range) 140 (75, 232) 192 (100, 300) 240 (141, 380) Average patients/program, n, mean ± SD 165 ± ± ± 260 Total patients, n Estimated capacity AACVPR registered programs a, n = All other US CR programs b, n = Abbeviations: AACVPR, American Association of Cardiovascular and Pulmonary Rehabilitation; CR, cardiac rehabilitation; IQ, interquartile. a Adjusted for baseline differences in program size between survey respondents and nonrespondents using program full-time equivalents. b Estimated from Curnier et al. 18 See the Methods section. Cardiac Rehabilitation Capacity / 323

7 Figure 3. Obstacles to increasing patient participation in cardiac rehabilitation as reported by program directors. As shown, the majority of reported obstacles were structural, administrative, or financial barriers unrelated to patient behavior. CR, cardiac rehabilitation. expansion, estimated national CR capacity is insufficient to meet current CR needs. With more realistic estimates, on the basis of maximizing current CR program capacity, only about half of eligible patients could receive CR program services. This general finding persisted even when we used several different assumptions and methods in calculating current capacity. In addition, most program directors felt primarily limited by facilities, personnel, and co-pays/ insurance problems rather than patient behavior, suggesting that program capacity restraint is a significant contributor to the underutilization of CR services across the United States. Similar concerns about CR capacity have been previously raised in Europe. 19 This capacity limitation is critically important to know about for several reasons. First, this limitation suggests that alternative models of CR delivery need to be further developed and implemented if national CR participation rates are to improve significantly. As currently structured and staffed, center-based CR programs simply do not have the capacity, by themselves, to provide services to all eligible patients even in the setting of perfect referral and enrollment. Certainly, improving patient referral and enrollment will still be an essential part of improving CR participation. Furthermore, it will be very important to continue to work with providers, hospitals, and insurance companies to improve referral to traditional centerbased CR. However, our data suggest that alternative models of CR delivery will need to be explored and implemented to substantially increase national CR participation rates. Second, limited resources hinder the further expansion and improvement of CR in the United States. As Table 3 Cardiac Rehabilitation Utilization Rates and Estimates Among Eligible US Patients Population Sample Year Estimated Population or Source Current CR Utilization, % (Min, Max) Current CR Capacity, % (Min, Max) Expansion CR Capacity, % (Min, Max) US national statistics, 2012 Qualifying diagnoses for CR million patients a 28 (20, 38) 37 (26, 53) 47 (32, 67) National samples Prevalent CR participation rates 2005 CDC Medicare CR participation rates 1997 Suaya et al Abbreviations: CDC, Center for Disease Control; CR, cardiac rehabilitation. a Based on estimates provided by Go et al. 16 Qualifying diagnoses considered were percutaneous coronary intervention, coronary artery bypass graft surgery, valve surgery, heart transplant, myocardial infarction, or incident stable angina. 324 / Journal of Cardiopulmonary Rehabilitation and Prevention 2014;34:

8 demonstrated in our survey, the top 2 self-reported concerns voiced by program directors were facility limitations and insufficient personnel resources, rather than lack of patient interest or other patient-related obstacles. In addition, average program age was 23 ± 8 years, implying that the vast majority of programs are > 5 years old and new CR programs are uncommon. These findings suggest that financial concerns (ie, low reimbursement for current CR services provided) are a significant driver of inadequate capacity and may adversely affect utilization rates. Given the known cost-effectiveness of CR, 2, 3 improving reimbursement for CR services and eliminating excessive patient co-pays would likely improve patient outcomes in a very cost-effective manner. Furthermore, an expansion of reimbursement strategies would accelerate development and implementation of alternative CR delivery models. Third, our results will be important for policy makers, facing decisions about insurance coverage policies for other medical conditions where lifestyle-based therapy is effective. For example, there is significant evidence, showing that CR services are beneficial for patients with heart failure, 20 peripheral vascular disease, 21 and metabolic syndrome. 22 Given the recent Medicare decision intending to reimburse CR services for stable heart failure, it seems even more likely that patient demand will exceed current capacity, and that policy and coverage changes that promote even greater expansion of CR capacity will be increasingly needed. Potential solutions to this problem may be to supplement traditional CR programs with alternative models of delivery, including group-based CR programs in community centers, home-based programs, and Web-based methods of delivering CR. Safety and effectiveness of each of these approaches will need to be shown more clearly, although studies support the potential benefits of both approaches. 23, 24 To make the use of these approaches more feasible, reimbursement strategies will need to be adjusted, as will CR certification processes, which both define and drive the appropriate, evidence-based models of CR delivery. Ideally, such programs would potentially be able to reach 19.3% of the US population that lives in rural areas 25 and would additionally be well suited to assist the currently underserved populations of the elderly, women, and minorities. Although it may be unreasonable to expect that 100% of patients will participate in CR, it is notable that some programs currently report approximately 70% participation rates in center-based programs These reports stand in stark contrast to reported national participation rates of 18% to 35% 4, 6 and suggest that, if given the proper resources, CR programs could improve low participation rates through quality improvement activities, an idea supported by our study. Furthermore, if additional alternative models of CR delivery suggested previously were developed as additional options for patients, it seems possible that participation rates of > 80% could become possible. Such participation rates would still far outstrip current available resources. The primary limitation of this study was the fact that we were unable to identify reliable information on CR programs that are not registered with AACVPR. As a result, we needed to make several assumptions in our estimates, including the assumption that AACVPR CR programs and non-aacvpr CR programs are similar in characteristics and CR delivery. However, very little is known about non-aacvpr registered CR programs including their size, age, geographical distribution, contact information, and their scope. The last time a total census of these programs was done was in 2003, 18 and substantial changes in CR program numbers could have occurred since that time. The primary strength of the study was our ability to perform a nonresponse bias analysis and demonstrate that our survey participants were generally representative of all AACVPR programs. This adds substantial confidence to our overall generalizability and projections. Although the survey response rate was not ideal, recent editorials have suggested that representativeness is more important than response rate. 29 Furthermore, our current national utilization estimate of 27% is in the range previously observed in the medical literature, and our sensitivity analysis suggested that substantially altering our basic assumptions did not change the main results of underutilization from capacity restraints. CONCLUSIONS Even with substantial expansion of all existing CR programs, there is currently insufficient capacity in current CR programs to meet national CR service delivery needs. Solutions to this problem will likely include the creation of new CR programs, improved CR reimbursement strategies, and new approaches to delivering CR services. Traditional center-based CR programs will continue to be an important part of future CR delivery, but our results suggest that current models of delivery are insufficient to meet the large national need for CR services. Acknowledgments We thank Jonah Gorski, BA, and the AACVPR leadership for allowing and assisting with this survey. The Cardiac Rehabilitation Capacity / 325

9 Mayo Clinic Cardiovascular Rehabilitation Program provided funding to cover the modest cost of the survey incentive. References 1. Heran BS, Chen JM, Ebrahim S, et al. Exercise-based cardiac rehabilitation for coronary heart disease. Cochrane Database Syst Rev ;( 7 ): CD Ades PA, Pashkow FJ, Nestor JR. Cost-effectiveness of cardiac rehabilitation after myocardial infarction. J Cardiopulm Rehabil ; 17 ( 4 ): Wong WP, Feng J, Pwee KH, Lim J. A systematic review of economic evaluations of cardiac rehabilitation. BMC Health Serv Res ; 12 : Suaya JA, Shepard DS, Normand SL, Ades PA, Prottas J, Stason WB. Use of cardiac rehabilitation by Medicare beneficiaries after myocardial infarction or coronary bypass surgery. Circulation ; 116 ( 15 ): Centers for Disease Control and Prevention. Receipt of cardiac rehabilitation services among heart attack survivors 19 states and the District of Columbia, MMWR Morb Mortal Wkly Rep ; 52 ( 44 ): Centers for Disease Control and Prevention. Receipt of outpatient cardiac rehabilitation among heart attack survivors United States, MMWR Morb Mortal Wkly Rep ; 57 ( 4 ): Smith SC, Jr, Benjamin EJ, Bonow RO, et al. AHA/ACCF secondary prevention and risk reduction therapy for patients with coronary and other atherosclerotic vascular disease: 2011 update. J Am Coll Cardiol ; 58 ( 23 ): Balady GJ, Ades PA, Bittner VA, et al. Referral, enrollment, and delivery of cardiac rehabilitation/secondary prevention programs at clinical centers and beyond: a presidential advisory from the American Heart Association. Circulation ; 124 : Arena R, Williams M, Forman DE, et al. Increasing referral and participation rates to outpatient cardiac rehabilitation: the valuable role of healthcare professionals in the inpatient and home health settings: a science advisory from the American Heart Association. Circulation ; 125 : Thomas RJ, King M, Lui K, Oldridge N, Pina IL, Spertus J. AACVPR/ACCF/AHA 2010 update: performance measures on cardiac rehabilitation for referral to cardiac rehabilitation/secondary prevention services: a report of the American Association of Cardiovascular and Pulmonary Rehabilitation and the American College of Cardiology Foundation/American Heart Association Task Force on Performance Measures (Writing Committee to Develop Clinical Performance Measures for Cardiac Rehabilitation). Circulation ; 122 ( 13 ): Dunlay SM, Witt BJ, Allison TG, et al. Barriers to participation in cardiac rehabilitation. Am Heart J ; 158 ( 5 ): Ades PA, Waldmann ML, McCann WJ, Weaver SO. Predictors of cardiac rehabilitation participation in older coronary patients. Arch Intern Med ; 152 ( 5 ): Brown TM, Hernandez AF, Bittner V, et al. Predictors of cardiac rehabilitation referral in coronary artery disease patients: find- ings from the American Heart Association s Get With The Guidelines Program. J Am Coll Cardiol ; 54 ( 6 ): Parashar S, Spertus JA, Tang F, et al. Predictors of early and late enrollment in cardiac rehabilitation, among those referred, after acute myocardial infarction. Circulation ; 126 ( 13 ): Gurewich D, Prottas J, Bhalotra S, Suaya JA, Shepard DS. System-level factors and use of cardiac rehabilitation. J Cardiopulm Rehabil Prev ; 28 ( 6 ): Go AS, Mozaffarian D, Roger VL, et al. Heart disease and stroke statistics 2013 update: a report from the American Heart Association. Circulation ; 127 ( 1 ): e6-e Roger VL, Weston SA, Gerber Y, et al. Trends in incidence, severity, and outcome of hospitalized myocardial infarction. Circulation ; 121 ( 7 ): Curnier DY, Savage PD, Ades PA. Geographic distribution of cardiac rehabilitation programs in the United States. J Cardiopulm Rehabil ; 25 ( 2 ): Bjarnason-Wehrens B, McGee H, Zwisler AD, et al. Cardiac rehabilitation in Europe: results from the European Cardiac Rehabilitation Inventory Survey. Eur J Prev Cardiol ; 17 ( 4 ): O Connor CM, Whellan DJ, Lee KL, et al. Efficacy and safety of exercise training in patients with chronic heart failure: HF-ACTION randomized controlled trial. JAMA ; 301 ( 14 ): Murphy TP, Cutlip DE, Regensteiner JG, et al. Supervised exercise versus primary stenting for claudication resulting from aortoiliac peripheral artery disease: six-month outcomes from the claudication: exercise versus endoluminal revascularization (CLEVER) study. Circulation ; 125 ( 1 ): Rubenfire M, Mollo L, Krishnan S, et al. The metabolic fitness program: lifestyle modification for the metabolic syndrome using the resources of cardiac rehabilitation. J Cardiopulm Rehabil Prev ; 31 ( 5 ): DeBusk RF, Miller NH, Superko HR, et al. A case-management system for coronary risk factor modification after acute myocardial infarction. Ann Intern Med ; 120 ( 9 ): Taylor RS, Dalal H, Jolly K, Moxham T, Zawada A. Home-based versus centre-based cardiac rehabilitation. Cochrane Database Syst Rev ( 1 ): CD United States Department of Health and Human Services. Defining the rural population. Accessed June 20, Harkness K, Smith KM, Taraba L, Mackenzie CL, Gunn E, Arthur HM. Effect of a postoperative telephone intervention on attendance at intake for cardiac rehabilitation after coronary artery bypass graft surgery. Heart Lung ; 34 ( 3 ): Pack QR, Goel K, Lahr BD, et al. Participation in cardiac rehabilitation and survival after coronary artery bypass graft surgery: a community-based study. Circulation ; 128 ( 6 ): Pack QR, Mansour M, Barboza JS, et al. An early appointment to outpatient cardiac rehabilitation at hospital discharge improves attendance at orientation: a randomized, single-blind, controlled trial. Circulation ; 127 ( 3 ): Johnson TP, Wislar JS. Response rates and nonresponse errors in surveys. JAMA ; 307 : / Journal of Cardiopulmonary Rehabilitation and Prevention 2014;34: