It is well known that center-based cardiac rehabilitation (CBCR), despite being highly efficacious and cost-effective, is a significantly underutilized medical service.1,2 Of course there are numerous barriers that reduce participation in CBCR, including time and schedule limitations, lack of referral and support of the medical community, transportation issues, and insurance/financial constraints. Consequently, home-based cardiac rehabilitation programs (HBCR), first described by researchers at Stanford more than 20 years ago,3,4 have become recognized as a viable alternative for patients unable to participate in CBCR owing to the presence of the aforementioned barriers. Over the last two decades, a number of studies on HBCR (employing somewhat different methodologies) have provided adequate data to demonstrate that this approach is not only safe but also efficacious and cost-effective.5-10 Despite these findings, HBCR programs are not without their own inherent limitations, particularly a lack of availability (to my knowledge, few programs offer this option) and lack of financial support by third-party payers. Moreover, little is known about overall patient preference for this type of program and/or preference by specific subgroups for HBCR versus CBCR, as these issues have never been systematically examined.

Grace and colleagues11 are to be commended for examining the important issue of patient preference of HBCR versus CBCR. This investigation attempted to identify potential barriers to, and predictors of, participation in either HBCR or CBCR using a cross-sectional, observational-type design. The paper is very well written, and the statistical approaches employed to identify predictors are quite sophisticated. The findings of this investigation are intriguing as they suggest that Caucasian, time-constrained, working individuals would likely prefer a HBCR while certain subgroups (nonworking, ethnocultural background) appear to prefer a traditional CBCR program. While intriguing, the conclusions drawn from this investigation have to be interpreted carefully as the study has a number of limitations, many of which are appropriately acknowledged by the authors. First, the overall sample size is small and this becomes a significant issue when attempts are made to predict participation in HBCR versus CBCR based on ethnocultural background. Second, the study design was nonprospective and nonrandomized. While this type of research design is a useful initial approach, as suggested by the authors, it should be followed by a prospective, randomized, multisite clinical trial before definitive conclusions can be drawn. In addition, there are several issues inadequately addressed in the manuscript regarding how and when the two options (HBCR vs CBCR) were presented and/or delivered. For instance, in the "introductory" session-what type of information did patients receive about safety and monitoring provided in the HBCR versus CBCR? Were the two interventions presented in a "balanced" manner? Furthermore, the CBCR was offered in multiple languages; was the same true for the HBCR? Clearly, these types of issues are likely to affect how patients perceive and ultimately determine which program they prefer. Furthermore, some discussion of the Canadian model of reimbursement/support for cardiac rehabilitation services (including home-based components) would have been useful in the interpretation of these findings.

To my knowledge, HBCR program services are not commonly covered by most US private health insurance providers or Medicare. Consequently, most CBCR programs have very little incentive to develop and offer this approach to prospective patients. Even simple, albeit effective, interventions typically employed by HBCR research protocols that include phone calls, review of exercise logs, etc still require the use of costly resources-primarily staff time. In the current fiscal environment surrounding cardiac rehabilitation, at least in the United States, hospital administrators are not likely to be supportive of this non-revenue-generating activity. Consequently, programs attempting to offer HBCR would need to pass the costs directly onto the participant. Several years ago, after observing and reporting the significant benefits of a research-funded home-based "maintenance" cardiac rehabilitation program,12 we attempted to continue to provide this service to our participants as a "fee for service" program, as clearly there would be no third-party reimbursement for a "maintenance" exercise program. Unfortunately, we quickly found that most patients were unwilling to pay even $25.00 per month for an intervention that consisted of a one-monthly home visit by a nurse and exercise physiologist, analysis and feedback on monthly exercise logs, as well as a weekly 30-minute phone call with each patient to discuss progress and problems in the HBCR program. Given the financial disincentive of this approach, we terminated our HBCR maintenance program. There are a number of well-designed and tested HBCR models,9,13 including the Cardiovascular Health Activity Maintenance Program (CHAMP) developed at Wake Forest, that have been shown to be equal or more effective than CBCR.14,15 However, until there is financial support for these alternative approaches by the health insurance providers and/or the patient perception and willingness to pay for these services greatly improves, it is doubtful as to whether HBCR will ever become a truly viable "option" for patients who are unable/unwilling to participate in CBCR. A large prospective, randomized, multisite clinical trial could potentially provide the data necessary to convince third-party payers to support these alternative approaches of cardiac rehabilitation program delivery. Soon we will have the opportunity to see if the results of a large clinical exercise training trial (HF-ACTION) can impact the financial support for cardiac rehabilitation services, specifically for heart failure patients. The HF-ACTION trial, which was started early 2003, and will ultimately randomly assign 3000 systolic heart failure patients at 50 + centers across United States and Canada to a long-term exercise program (3 months center-based followed by home-based) or to a control group. Primary outcome is the effects of exercise (most of which is home-based) on morbidity and mortality in heart failure patients. Clearly, the results of HF-ACTION have the potential to change policy regarding third-party reimbursement for heart failure exercise programs-whether or not this will happen remains to be seen.

In closing, Grace and colleagues should be applauded for initiating a study and subsequent dialog on a complex, yet important, issue regarding the optimal delivery of cardiac rehabilitation services. Unfortunately, direct application of the information derived from this study to actual clinical practice will have to wait until these findings are confirmed through a more rigorous investigational process. Furthermore, until third-party payers recognize and financially support HBCR, this alternative method of delivery may never become a true option for potential cardiac rehabilitation program participants.

References