Identifying Factors for Low-Risk Participation in Alternative Cardiac Rehabilitation Models for Patients with Coronary Heart Disease Using MI'S SCOREPAD

Introduction Although a recent joint society scientific statement (the American Association of Cardiovascular Pulmonary Rehabilitation, the American Heart Association, and the American College of Cardiology) suggests home-based cardiac rehab (CR) is appropriate for low- and moderate-risk patients, there are no paradigms to define such individuals with coronary heart disease. Methods We reviewed a decade of data from all patients with coronary heart disease enrolled in a single CR center (University of Michigan) to identify the prevalence of low-risk factors, which may inform on consideration for participation in alternative models of CR. Low-risk factors included not having any of the following: metabolic syndrome, presence of implantable cardioverter defibrillator or permanent pacemaker, active smoking, prior stroke, congestive heart failure, obesity, advanced renal disease, poor exercise capacity, peripheral arterial disease, angina, or clinical depression (MI'S SCOREPAD). We report on the proportion of participants with these risk factors and the proportion with all of these low-risk factors. Results The mean age of CR participants (n = 1984) was 63 years; 25% were women, and 82% were non-Hispanic White. The mean number of low-risk factors was 8.5, which was similar in the 2011-2012 and 2018-2019 cohorts (8.5 vs. 8.3, respectively, P = 0.08). Additionally, 9.3% of the 2011-2012 cohort and 7.6% of the 2018-2019 cohort had all 11 of the low-risk factors. Conclusion In this observational study, we provide a first paradigm of identifying factors among coronary heart disease patients that may be considered low-risk and likely high-gain for participation in alternative models of CR. Further work is needed to track clinical outcomes in patients with these factors to determine thresholds for enrolling participants in alternative forms of CR.


Introduction
Cardiac rehabilitation (CR) is effective for secondary prevention following a major adverse cardiovascular event [1,2].CR has a class I indication with evidence that comprehensive medically supervised risk reduction strategies lead to improved compliance with healthy lifestyles and evidencebased therapies, quality of life, cardiorespiratory fitness, and exercise tolerance, quicker return to work, and decreases in angina, subsequent hospitalizations, and mortality [3].CR is entering an era wherein alternative forms of CR, including abbreviated, home-based, or hybrid CR, are being increas-ingly considered, especially for some patients with coronary heart disease (CHD).
A recent joint Scientific Statement from the American Association of Cardiovascular Pulmonary Rehabilitation (AACVPR), the American Heart Association (AHA), and the American College of Cardiology (ACC) states that home-based CR "may be a reasonable option for selected clinically stable low-to moderate-risk patients" [4] However, no definition for low or moderate risk is provided in this scientific statement.There are definitions in the AACVPR Guidelines for CR programs (6 th Edition) for low or moderate risk factors for participation in CR's exercise component.
However, there remain no paradigms to define individuals with CHD who may be low enough risk based on exercise and nonexercise factors and possibly more suitable for alternative nontraditional CR models (i.e., abbreviated, homebased, or hybrid CR).
Thus, we reviewed a decade of data from all patients with CHD enrolled in a single CR center to identify the prevalence of low-risk factors that may be considered for suitable participation in alternative models of CR.We hypothesized that commonly recognized low-risk factors would be common in the CR population.Furthermore, since comorbidities have shifted over time in the general population, we hypothesized that the prevalence of these risk factors would differ in the first and last two years of the cohort.

Methods
2.1.Data.This prospective observational study followed the Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines.Our study consisted of patients enrolled in the University of Michigan CR program from 1/1/2011-2/29/2020 whose indication for CR was CHD-related (n = 1984).This study was approved by the University of Michigan IRB.Consent to participate in CR and for data to be used for research purposes were obtained from all participants at the time of CR enrollment (IRB#: HUM00045929).
Self-reported patient characteristics were collected from standardized health questionnaires and at intake via exercise physiologists, including age, sex, race/ethnicity, physical activity level, smoking status, and history of peripheral arterial disease.Stroke, heart failure, and presence of cardiac pacemaker or implantable cardioverter defibrillator history were derived from past diagnostic codes.Angina was based on indication for referral as documented by the exercise physiologist.Psychological distress was evaluated using the Brief Symptom Inventory (BSI), consisting of 53 items and covering nine symptom dimensions [5].Adult nonpatients were used as the reference group to convert raw scores to T-scores.Depression was defined as a T-score of ≥63 (90 th percentile) [6].Ejection fraction was obtained from the ventriculogram or, if not available, the most recent echocardiogram or nuclear study.Definitions for major CHD risk factors and the criteria for the metabolic syndrome were based upon those of the American Heart Association [7].For determining the metabolic syndrome, we used laboratory data (fasting glucose and lipid profiles) from the date closest to baseline CR evaluation.Clinical exercise physiologists measured blood pressure, body weight, and waist circumference at CR entry.Cardiorespiratory fitness was measured at CR entry by peak oxygen consumption (VO 2peak ) and evaluated using electronic/motorized treadmill test (Ultima CPX metabolic stress test system, MCG Diagnostics).Patients were tested until requesting to stop, general/leg fatigue, clinical decision to terminate, or maximal effort.Patients were not asked to discontinue medications before measurement.

2.2.
Outcome.The primary outcome was the proportion of CR participants who were low-risk and therefore could likely qualify for abbreviated, home, or hybrid CR participation.We were able to identify 11 commonly measured factors in CR and defined low risk as having all 11 of these low-risk factors.
The factors included five factors that the AACVPR defines as low-risk factors for participating in the exercise component of CR, including peak aerobic capacity ≥7 METS (METS = VO 2peak /3:5), no heart failure (including ejection fraction ðEFÞ ≥ 50%), no angina, and no clinical depression [8,9].Presence of cardiac pacemaker or implantable cardioverter defibrillator was available and used as a substitute for risk for complex dysrhythmia.Notably, the AACVPR also identified complex dysrhythmia and abnormal hemodynamics during exercise testing and recovery as risk factors, but these data were not available for this study.
We expanded on the AACVPR risk factors by including six additional factors that do not increase the risk of participating in the exercise component of CR but require major lifestyle changes that may not be equally addressed from alternative CR approaches or are major cardiovascular risk factors.These additional low-risk factors included current nonsmoking status, class I obesity or less (BMI ≤ 35 kg/ m 2 ), no metabolic syndrome, no peripheral arterial disease, no prior stroke, no advanced renal disease (eGFR ≥ 45 ml/ min/m 2 ), and no depression.These risk factors can be recalled using the acronym MI'S SCOREPAD (see Table 1 for review of risk factor definitions).

Statistical
Analysis.Among the entire study cohort, the extent of missing data for the relevant factors ranged from 0% to less than 10%, except for exercise capacity measure (36%).Thus, we performed multiple imputations with 100 imputed datasets, which were combined for estimates of the individual factors and corresponding comparisons.Nonimputed results are reported in Supplementary Table 1.Comparisons were made using chi-squared tests, Fisher's exact tests, and Wilcoxon rank-sum tests.We compared the prevalence of factors in the earliest (2011-2012) and most contemporary (2018-2019) cohorts.All analyses were performed using SAS v9.4 (SAS Institute, Inc.).A 2-tailed P < 0:05 was used to indicate statistical significance.
Proportions of participants with factors associated with a low-risk for participation in abbreviated/home/hybrid CR are provided in Table 3.The mean number of low-risk factors was 8.5.Additionally, 7.3% of the entire cohort had all 11 of the low-risk factors.Mean risk factors and number with all 11 risk factors were similar in the 2011-2012 and 2018-2019 cohorts (8.5 vs. 8.3, P = 0:08, and 9.3% vs. 7.6%, P = 0:70, respectively).However, compared to the 2011-2012 cohort, the 2018-2019 cohort was more likely to have 2 Cardiovascular Therapeutics a history of congestive heart failure (P = 0:002) and advanced renal disease (P = 0:03).
Among the five AACVPR-driven guideline-based factors, the median number of low-risk factor was 4 and mean was 3.5 (Table 3).There were 12.5% of the cohort with all five of the AACVPR low-risk factors, which was similar in the 2011-2012 (13.0%) and 2018-2019 (14.0%) cohorts (P = 0:67).
Values without multiple imputation are reported in Supplementary Table 1.

Discussion
In this observational study, we provide a first paradigm of identifying factors among CHD patients that may be considered low-risk and likely high-gain for participation in alternative models of CR (i.e., abbreviated, home-based, or hybrid CR).These factors are overall common with most having at least 8 low-risk factors.Our findings are relevant to ongoing decisions for the future of CR and how CR can better serve the needs of current patients with CHD.Although paradigms such as home-based CR are being used clinically [10], there is no guideline-chosen paradigm to define level of risk and aid in deciding which patients are best qualified for different lengths and types of CR.As new types of CR evolve, there should be ongoing consideration of which factors inform the preferred modality.It is important to understand this because alternative forms of CR, such as home-based CR, may be just as, if not more, cost-effective than center-based options, but patients should be chosen wisely to ensure safety [11][12][13].
We have provided a first paradigm of factors based on currently recognized clinical risk factors to determine which patients may be low-risk for alternatives to inperson CR.Notably, 7.3% of participants had all 11 factors from the MI'S SCOREPAD and 12.5% of participants has all five AACVPR guideline-driven criteria.Our analysis does not determine an appropriate cut point for alternative CR models.This research is direly needed.What our work does do is lend a view into a potential population for these future CR strategies.
4.1.Limitations.First, the study utilized medical records to identify relevant data, some of which were not originally collected for research purposes.Second, there were a large number of missing data and repeat analysis after multiple imputation did not confirm all findings.Third, our patient sample was predominantly White at a major academic medical center, which limits generalizability.

Conclusion
This study introduces a novel framework of factors to consider for determining whether CHD patients are at lowrisk for alternative models of CR.Given the urgent need for enhanced approaches to secondary CHD prevention [14], additional study is warranted to optimize CR programs to increase accessibility and promote cardiometabolic health [15].Our findings can provide a basis for future studies that seek to clearly define low and moderate risk for participation in home-based CR and other CR models based on recent joint-society scientific statements.

Table 2 :
Characteristics of cardiac rehabilitation study population, 2011-2020, and at the beginning and end of the study period.

Table 3 :
Low-risk factors to consider for favorable participant in alternative forms of cardiac rehabilitation based on 100 imputed datasets.P value from χ 2 tests.†Defined as able to achieve ≥7 METS.‡ AACVPR factors included peak aerobic capacity ≥7 METS (METS = VO 2peak /3:5), no heart failure (including ejection fraction ðEFÞ ≥ 50%), no angina, no clinical depression, and presence of cardiac pacemaker or implantable cardioverter defibrillator.