The Current and Potential Capacity for Cardiac Rehab in the U.S.
Limited program capacity may contribute to low rates of cardiac rehabilitation (CR) utilization in the U.S. However, whether the collective expansion capacity of existing CR programs is sufficient to support high participation rates is not known.
This was a survey study conducted in 2012 and aimed at directors of all CR programs registered with the American Association of Cardiovascular and Pulmonary Rehabilitation (AACVPR) and engaged in reporting to the AACVPR database. Potential respondents were contacted by email and asked to complete an anonymous, voluntary, online survey created and fielded using Survey Monkey® software. The 28 questions, written by the authors and approved by the AACVPR research committee and board of directors, were designed to assess current CR utilization, current program capacity, expansion capacity, and obstacles to growth. Emails collected during the survey were matched to those stored in the AACVPR database and used to validate responses to questions regarding program demographics and average enrollment. A 'bias analysis' was conducted to examine differences in program characteristics between survey respondents and AACVPR-registered non-respondents. Between group differences were assessed by the Chi-square test, t test and Wilcoxon rank-sum test for each descriptive variable.
Of 812 AACVPR-registered program directors, 290 (36%) responded fully to the survey. Median annual current enrollment, current capacity and estimated expansion capacity of the programs (and interquartile ranges), were 140 patients (range 75, 232), 192 patients (range 100, 300), and 240 patients (range 141, 380), respectively. After extrapolating from national statistics to compute the annual number of CR-eligible patients in the U.S. (~1.25 million), and from survey and other AACVPR and non-AACVPR data to compute the annual number of CR enrollees (~350,000), the authors calculated a utilization rate for CR (for the year 2012) of 28%. If all existing programs maximized their current capacity, 37% of potential enrollees could be accommodated. If all programs expanded capacity 'modestly,' the authors calculated that existing programs could serve only 47% of CR-eligible patients in the U.S. This mismatch persisted even when the authors used different assumptions to calculate expansion 'supply' and 'demand.' Regarding obstacles to growth, 25-30% of respondents cited systems-related constraints (lack of facilities and/or staffing) as the principal barriers to expansion, while almost 25% cited insurance copays. In the 'bias analysis,' the only significant difference between respondents and non-responders was that the former group directed larger programs with more staff.
The authors concluded that even if all CR programs were able to achieve "substantial" expansion, there would be insufficient capacity to accommodate all potential annual CR enrollees, and this gap will continue to drive underutilization of CR. They recommended the creation of more CR programs, alternative delivery models (e.g., home-based and web-based CR), and higher reimbursement/lower copays to increase CR capacity and utilization in the U.S.
The primary aim of this study was to predict whether the collective expansion capacity of existing CR programs could support a large increase in CR utilization in the U.S. However, as the authors admit, their data have limitations. First, they surveyed only AACVPR-accredited programs (which account for only 27% of current CR programs), and, thus, their 'numerator' was based largely on data extrapolated from AACVPR-programs to non-AACVPR programs. Second, the percentage of AACVPR-registered respondents was only 36%, and the larger sizes of their programs and other unaccounted factors may have led to selection bias and further miscalculations within the numerator. Third, because national statistics on the number of CR-eligible patients are lacking, the authors were forced to compute this number from a variety of sources, which may have led to over- or underestimation of their 'denominator.' Fourth, the authors did not indicate that they provided a formula for respondents to calculate their expansion capacity, implying that in most cases this figure was a loose estimate. This fact, and the lack of control over which program personnel actually answered the survey, may have impacted significantly the accuracy of the authors' final estimates of expansion capacity. Nonetheless, the data suggest that the number of existing site-based programs in the U.S. is probably insufficient to support large increases in CR enrollment. These findings may be of particular interest to payers and policy makers. They also may be important for programs that operate within coordinated health care systems transitioning to accountable care (ACOs), in which high participation in some type of CR or secondary prevention program is necessary to optimize population management of cardiovascular disease. However, until all CR programs are operating within ACOs, and not within thin 'profit' margins, maximizing individual (and collective) current capacity, through provider education and quality initiatives, is the logical first step toward expansion and wider CR utilization.
Keywords: Cardiovascular Diseases, Chi-Square Distribution, Secondary Prevention, Selection Bias, Haplorhini, Rehabilitation, Exercise Therapy
< Back to Listings