Out-of-Pocket Spending for Cardiac Rehabilitation and Adherence Among US Adults

ABSTRACT

Objectives: Although cardiac rehabilitation (CR) improves cardiovascular outcomes, adherence remains low. Higher patient-incurred out-of-pocket (OOP) spending may be a barrier to CR adherence. We evaluated the association between OOP spending for the first CR session and adherence.

Study Design: Retrospective analysis.

Methods: Commercial and Medicare supplemental beneficiaries with a CR-qualifying event between 2016 and 2020 who attended at least 1 CR session within 6 months of discharge were identified in the MarketScan Commercial Database. OOP spending for the first session was categorized as zero or into 1 of 3 increasing tertiles of OOP spending. Poisson regression was used to determine the association between OOP-spending tertile and CR adherence, defined as the number of CR sessions attended within 6 months of discharge.

Results: A total of 43,992 beneficiaries attended at least 1 CR session. Of these, 35,883 (81.6%) paid $0, 2702 (6.1%) paid $0.01 to $25.39, 2704 (6.1%) paid $25.40 to $82.41, and 2703 (6.1%) paid at least $82.42 for the first session, constituting the first, second, and third OOP-spending tertiles, respectively. Compared with the zero-OOP cohort, the first-tertile cohort attended 13.5% (95% CI, 1.4%-27.1%; P = .028) more CR sessions and the second- and third-tertile cohorts attended 11.9% (95% CI, –16.4% to –7.1%; P < .001) and 30.9% (95% CI, –40.8% to –19.4%; P < .001) fewer CR sessions on average, respectively. For every additional $10 spent OOP on the first CR session, patients attended 0.41 fewer sessions on average (95% CI, –0.65 to –0.17; P < .001).

Conclusion: Among patients with OOP spending, higher spending was associated with lower CR adherence, dose dependently. Reducing OOP costs for CR may improve adherence for beneficiaries with cost sharing.

Am J Manag Care. 2024;30(12):651-657. https://doi.org/10.37765/ajmc.2024.89637

Takeaway Points

Although cardiac rehabilitation (CR) improves cardiovascular outcomes, CR adherence remains poor. Patient-incurred out-of-pocket (OOP) spending may be an important barrier to CR adherence.

• Although most beneficiaries had zero OOP costs for the first CR session, among those with cost sharing, higher OOP spending was associated with attending fewer CR sessions and lower odds of completing more than 24 sessions, dose dependently, after multivariable adjustment.
• Reducing OOP costs may improve CR adherence and clinical outcomes among those with cost sharing for CR.
• Because CR is a cost-effective intervention, insurance companies should reduce OOP costs for CR to improve the value of cardiovascular care.

Cardiovascular disease (CVD) remains the leading cause of morbidity and mortality in the US.1,2 Although great advancements have been made in treating CVD, there is a growing emphasis on promoting long-term health and preventing readmission following a cardiac event through cardiac rehabilitation (CR).3 CR is a comprehensive, multidisciplinary program designed to optimize cardiovascular-related health for those with CVD and typically entails 36 1-hour sessions of supervised exercise-based therapy and healthy lifestyle education.4 Recent meta-analyses have demonstrated CR’s proven role in decreasing cardiovascular-related mortality by 25% and hospital readmissions by 18%, driving its endorsement as a class 1a recommendation as standard of care following an acute myocardial infarction (AMI) or coronary artery bypass grafting (CABG).5,6 Additionally, a dose-dependent relationship exists between the number of CR sessions attended and improved cardiovascular outcomes, underscoring the importance of completing the program to fully realize its health benefits.7-9

Despite CR’s effectiveness, it remains underutilized, with only approximately one-third of qualified participants attending CR and evidence of poor adherence once enrolled.10-12 Although there are several well-studied barriers to CR adherence, out-of-pocket (OOP) costs may serve as an important barrier, particularly for commercial beneficiaries who may face higher co-pays per CR session compared with the average $23 paid by traditional Medicare fee-for-service participants.13-18 Because CR is billed per session, patient spending may increase with more sessions attended, potentially deterring those with OOP costs from completing the program. Therefore, it is imperative that we better understand how patients’ OOP spending is related to CR attendance to inform payment reform, reduce CR costs, and promote adherence. In this context, we used a Merative (formerly IBM) MarketScan database to obtain claims data of commercially insured beneficiaries and evaluated the association between the OOP spending for patients’ first CR session and CR adherence.

METHODS

Data Source and Variables

We used the MarketScan Commercial Database, which contains deidentified patient-level health data linked to inpatient and outpatient claims of employer-based commercially insured and Medicare supplemental beneficiaries.19 This database also discloses OOP charges for each CR claim, patient age, sex, date of admission/discharge for the qualifying event and CR sessions, and geographic region.19 This study was approved by the University of Michigan Institutional Review Board (HUM00211102) and adhered to the Strengthening of the Reporting of Observational Studies in Epidemiology guidelines.

Study Cohort

Using claims from January 1, 2016, to December 31, 2020, we identified commercial and Medicare supplemental beneficiaries aged 18 to 75 years who experienced at least 1 CR-qualifying event: AMI, CABG, percutaneous coronary intervention, or heart valve replacement or repair. Qualifying events were identified via International Statistical Classification of Diseases, Tenth Revision (ICD-10) diagnostic or procedure codes and Current Procedural Terminology codes as described elsewhere (eAppendix Methods [eAppendix available at ajmc.com]).12 CR sessions attended within 6 months of discharge were identified within outpatient facility and professional claims using Healthcare Common Procedure Coding System code 93797, 93798, G0422, or G0423 and revenue center code 0943.20 We included those who were continuously enrolled in their insurance policies 6 months after the event and who attended at least 1 CR session within 6 months of discharge from the qualifying event. Patients admitted for more than 30 days were excluded to ensure beneficiaries were reasonable CR candidates (eAppendix Figure 1).7

OOP Spending

For the first CR session, we obtained the total beneficiary OOP spending by summing the co-pay, coinsurance, and deductible amounts for that session. We categorized OOP spending for the first CR session into those with zero OOP spending and tertiles above zero to obtain 4 OOP-spending cohorts that comprised the covariate of interest.

CR Sessions

The total number of CR sessions completed by each beneficiary was compared across OOP-spending cohorts to assess the relationship between OOP spending for the first CR session and CR adherence. The primary outcome of interest was the total number of CR sessions attended within 6 months of discharge. The secondary outcome of interest was the odds of attending more than 24 sessions, given a potential plateau in the health benefits after attending more than 24 sessions.9

Covariates

To account for comorbidities that may confound CR adherence, we used secondary ICD-10 diagnosis codes from the index admission claim and calculated each patient’s Elixhauser Comorbidity Index (ECI) score.21,22 Our regression models also adjusted for age, sex, qualifying event, days elapsed between the qualifying event discharge date and first session, insurance type (ie, Medicare supplemental vs commercial), geographic region, and calendar year and quarter of the qualifying event. The yearly quarter was included to account for seasonal variations in CR attendance and patients’ remaining deductibles when approaching the end of the year.

Statistical Analysis

Differences between baseline characteristics and total CR sessions attended were compared using χ2 test and 1-way analysis of variance for categorical and continuous variables, respectively. The Kruskal-Wallis test was used to compare the medians of the total CR sessions attended and days elapsed from the date of discharge to the first session between OOP-spending cohorts. For the primary outcome, Poisson regression was used to estimate the mean number of CR sessions attended as a function of the OOP cohort, controlling for our covariates. The incident rate ratio for the OOP-spending cohort covariate was converted to relative rate change to intuitively represent the association between OOP spending and CR adherence.

Several sensitivity analyses were performed. Multivariable logistic regression was used for the secondary outcome to compare the probability of completing more than 24 sessions across OOP-spending cohorts and of attending a second CR session to account for the impact delays in CR enrollment may have on the effective time left to complete further sessions. We additionally performed a subgroup analysis restricted to patients who attended the first session in January and February to assess the impact of deductibles resetting at the beginning of the calendar year. The zero-OOP cohort served as the reference group for all regression models.

Lastly, OOP spending was modeled as a continuous variable to estimate the change in the mean number of CR sessions attended and likelihood of attending more than 24 sessions for every $10 increase in OOP spending for the first session. We performed a robustness check with squared and cubic terms of the OOP-spending variable to allow for a nonlinear relationship between OOP spending and predicted number of CR sessions attended.

All analyses were conducted using Stata 15 MP (StataCorp LLC). A 2-sided P value of less than .05 was considered statistically significant.

RESULTS

Baseline Characteristics

Among the 43,992 beneficiaries who attended at least 1 CR session within 6 months of their qualifying event discharge date (33.6% of the CR-eligible cohort), the mean (SD) participant age was 57.4 (8.4) years, 32,797 (74.6%) were men, and 37,373 (85.0%) were commercially insured (eAppendix Table 1 and eAppendix Figure 1). Compared with nonparticipants, CR participants were more likely to be men and had a higher mean ECI score (1.12 vs 1.52; P < .001) (eAppendix Table 1).

OOP Spending and CR Adherence

Of the beneficiaries who attended CR, 35,883 (81.6%) had zero OOP spending for the first CR session. For those with OOP spending, 2702 (6.1%) paid between $0.01 and $25.39, 2704 (6.1%) paid between $25.40 and $82.41, and 2703 (6.1%) paid $82.42 or more for the first CR session, constituting the first, second, and third OOP-spending tertiles, respectively (Table 1). The mean (SD) OOP spending for the first session for commercial and Medicare supplemental beneficiaries with cost sharing was $81.97 ($77.75) (range, $0.01-$562.15) and $43.94 ($39.88) (range, $0.01-$238.12), respectively (P < .001). The zero-OOP and third-tertile cohorts had the highest mean ECI score, and the third-tertile cohort had the greatest delay in attending the first CR session (Table 1 and Table 2). The unadjusted mean and median number of CR sessions attended were highest in the first-tertile cohort and progressively decreased across the second and third OOP tertiles (Table 2).

For the primary outcome of total number of CR sessions attended within 6 months of discharge, compared with the zero-OOP cohort, the second- and third-tertile cohorts attended 11.9%(95% CI, –16.4% to –7.1%; P < .001) and 30.9% (95% CI, –40.8% to –19.4%; P < .001) fewer CR sessions on average, respectively, and the first-tertile cohort attended 13.5% (95% CI, 1.4%-27.1%; P = .028) more CR sessions after multivariable adjustment (Figure 1). Spending in the third OOP tertile was the strongest negative predictor of CR adherence in our model (eAppendix Table 2). Compared with the zero-OOP cohort, the first-tertile cohort was also more likely to attend a second CR session (adjusted OR [aOR], 1.21; 95% CI, 1.09-1.33; P = .007), whereas the third-tertile cohort was significantly less likely to do so (aOR, 0.72; 95% CI, 0.60-0.87; P < .001) (eAppendix Figure 2). For every additional $10 spent OOP on the first CR session, the mean number of total sessions attended decreased by 0.41 (95% CI, –0.65 to –0.17; P < .001) (Figure 2), with this inverse relationship persisting after transforming the OOP-spending variable into squared and cubic terms (eAppendix Figures 3 and 4).

For the secondary outcome, the second-tertile (aOR, 0.70; 95% CI, 0.60-0.82; P < .001) and third-tertile (aOR, 0.44; 95% CI, 0.36-0.53; P < .001) cohorts were significantly less likely to attend more than 24 sessions compared with the zero-OOP cohort after multivariable adjustment; there was no significant difference in the likelihood of attending more than 24 sessions between the first-tertile and zero-OOP cohorts (aOR, 1.21; 95% CI, 0.91-1.61; P = .190) (Figure 3). For every additional $10 spent OOP on the first CR session, the probability of attending more than 24 sessions decreased by 0.95% (95% CI, –1.37% to –0.54%; P < .001).

Of the patients who had their first CR session in January or February, 3732 (53.6%) had zero OOP costs for the first session. The second-tertile and third-tertile cohorts attended 14.6% (95% CI, –23.0% to –5.2%) and 32.2% (95% CI, –43.0% to –19.4%) fewer mean sessions, respectively, compared with the zero-OOP cohort, but there was no significant difference in the adjusted mean number of sessions attended between the first-tertile and zero-OOP cohorts (6.26%; 95% CI, –0.18% to 13.12%) (eAppendix Figure 5).

DISCUSSION

In this retrospective study involving a large cohort of commercial and Medicare supplemental beneficiaries who attended at least 1 CR session, we observed 3 key findings. First, more than 80% of our cohort had no OOP spending for the first CR session. Second, those in the first OOP-spending tertile attended more CR sessions than the zero-OOP cohort. Third, among the remaining CR participants with cost sharing, greater OOP spending for the first session was associated with lower CR adherence in a dose-dependent manner, suggesting that lower OOP costs are associated with increased CR adherence among beneficiaries with cost sharing for CR.

Our findings are consistent with those of a smaller study of 603 patients by Farah et al, which demonstrated that greater cost sharing was associated with lower CR attendance and found a greater effect of 1.5 fewer sessions attended for every $10 increase in total OOP spending for CR.23 However, our study is the first to showcase similar findings in a nationally representative commercially insured population that may be subject to more variable OOP costs for CR compared with traditional Medicare patients.24 Indeed, of those with cost sharing for CR, we found commercial beneficiaries spent significantly more, on average, and had more variable OOP spending for the first session than their Medicare counterparts, which is concordant with findings of prior studies comparing cost-sharing patterns between commercial and Medicare beneficiaries for outpatient services.13-15,24 Overall, our study provides important and timely insight on the relationship between cost sharing and CR adherence, a topic of interest among investigators seeking to promote CR participation and equitable delivery of CR services.17,25,26

Importantly, we found that most patients had zero cost sharing for the first CR session, suggesting that OOP costs may not be a barrier to CR adherence for many beneficiaries. Nonmonetary factors (eg, time, travel distance) may be more important barriers. However, our findings may be explained by differences in benefit design and health care spending between the zero- and nonzero-OOP groups. The zero-OOP cohort may have been less healthy and utilized more health care, consequently meeting their deductible prior to CR enrollment. Similarly, the 63% of patients in the Farah et al study with zero cost sharing were older with lower functional capacity.23 In our study, most beneficiaries’ OOP spending may have been clustered around the CR-qualifying hospitalization, increasing the likelihood of meeting their deductible before enrollment, as demonstrated in a previous study evaluating clustering in OOP spending for hospitalized commercial beneficiaries.27 Indeed, the proportion of beneficiaries in our study with zero cost sharing decreased to 53% for those who attended the first session after deductibles reset, suggesting our results were driven in part by the deductible being met prior to enrollment. However, because most patients still had zero cost sharing, employer-based policies may cover CR as an essential health benefit, with a minority of policies having deductibles that exceed the qualifying-event cost or that charge coinsurance for CR.28 Without further information on patients’ benefit design, we cannot make conclusions on what drove this finding. Nevertheless, our study findings demonstrated a consistent inverse relationship between adherence and OOP spending for the first CR session among those with cost sharing. Therefore, reducing OOP costs for CR may be an effective strategy to promote adherence in a subpopulation with cost sharing for CR.

Notably, patients in the first OOP-spending tertile attended more sessions than the zero-OOP group. Although we cannot definitively determine why this relationship exists, we suspect that because the zero-OOP cohort comprised the majority of our sample, it is likely a more heterogeneous population with clinical and demographic confounders that we could not account for. These patients had a higher comorbidity index score, which may have prevented them from attending further CR sessions.18,29 Although we controlled for ECI score, we could not assess functional capacity or disease severity. Moreover, our ECI score was derived from the index hospitalization claim rather than all claims from the previous year, limiting its utility. The zero-OOP cohort also may have had less reliable transportation, lived further from the CR facility, or been composed of a higher proportion of patients of lower socioeconomic status or minority groups with historically lower CR adherence rates.11,23,29-33 Because we used claims data, we could not assess granular patient-level variables that may account for this discrepancy. OOP costs are one of many factors associated with CR adherence and likely impact CR participation in subpopulations differently. Future research should evaluate socioeconomic and demographic factors that impact CR adherence in commercial beneficiaries.

Although we evaluated OOP spending for the first CR session, we cannot assume a temporal relationship between the cost for the first session and subsequent adherence because patients may only be billed after multiple sessions. However, because the per-session OOP fee is based on a reimbursement amount set by the insurance payer and we were evaluating the same outpatient service, the OOP cost may vary minimally between CR sessions.15 Higher OOP spending for the first session may therefore translate to proportionally higher spending for multiple sessions, which may disincentivize participation and account for our findings. A bundled payment model in which patients are charged a onetime fee for CR may improve adherence by mitigating higher anticipated costs when billing for sessions recurrently.34,35 Nevertheless, there is likely substantial variability in billing practices, and we did not know the frequency or timing of beneficiaries’ CR charges. The impact of specific billing practices on CR adherence should be explored in future studies.

In this study, we demonstrate the paradox of cost sharing and its association with CR adherence and timing of enrollment. Although cost sharing is meant to disincentivize unnecessary health care consumption, it also results in the underutilization of cost-effective outpatient interventions.36,37 Increased co-pays for ambulatory care have been shown to result in 19.8 fewer appointments attended per year, which may pertain to CR as it entails repetitive outpatient sessions.37 Those in the highest OOP-spending tertile also had the greatest delay in enrollment and were less likely to attend a second session, suggesting their lower adherence was unlikely due to having less time to complete more sessions. Because delays in enrollment are related to lower CR participation and adherence, reducing cost sharing may improve these metrics by mitigating delays in enrollment itself.38,39 However, we do not know when patients were informed of their projected costs and how that impacted their decision to attend CR.

Our study supports quality improvement initiatives outlined in the Million Hearts Cardiac Rehabilitation Change Package, a collaboration between the CDC and the American Association of Cardiovascular and Pulmonary Rehabilitation, targeting cost-sharing barriers for CR services to promote CR adherence.40 For example, by developing philanthropic financial assistance programs, hospitals can help offset OOP expenses for CR for those who are underinsured, which may improve CR participation for patients with less comprehensive health plans and reduce disparities in cardiovascular care.40,41 Institutions may also proactively connect patients with their billing department to discuss CR costs and formulate affordable payment plans for those unable to pay their co-pays in full.40 Although high-value care is being incentivized through value-based physician reimbursement models, future payment-reform policy should focus on developing payment models that reduce patient-incurred costs for cost-effective interventions such as CR.41-45

Limitations

Our study has multiple limitations. First, we did not include patients who did not attend CR and could not identify their OOP spending because we could not access their benefit design. We were therefore unable to assess the effect of cost sharing or the expected cost of attending CR among those who never attended. Additionally, we did not have information on patients’ remaining deductible, coverage benefits, frequency of billing, or timing of when they were informed of their projected OOP costs, which may have influenced our findings. We also used claims data, so we were unable to evaluate potential socioeconomic and demographic confounders. Lastly, because the MarketScan Commercial Database only contains beneficiaries with employer-based policies, we cannot generalize these findings to non–employer-based beneficiaries with potentially different OOP liabilities for CR.

CONCLUSIONS

Among commercial and Medicare supplemental beneficiaries insured by an employer, most who attended CR had zero OOP spending for their first session, suggesting that OOP costs are not a barrier to CR adherence for most employer-based beneficiaries. However, for the minority of beneficiaries with cost sharing, higher spending for the first session was associated with lower CR adherence in a dose-dependent manner. Because CR is a cost-effective intervention that significantly improves cardiovascular-related morbidity and mortality and the value of cardiovascular care, insurance companies should consider policies that reduce patient-incurred OOP spending for those with cost sharing for CR to ensure CR adherence.37,45

Author Affiliations: Johns Hopkins Hospital (AIM), Baltimore, MD; University of Michigan Institute for Healthcare Policy and Innovation (UN), Ann Arbor, MI; Michigan Medicine (MPT, DS), Ann Arbor, MI; Henry Ford Health (SK), Detroit, MI; Michigan Medicine Department of Cardiology (DS), Ann Arbor, MI.

Source of Funding: None.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (AIM, MPT, SK, DS); acquisition of data (DS); analysis and interpretation of data (AIM, UN, MPT, DS); drafting of the manuscript (AIM, SK, DS); critical revision of the manuscript for important intellectual content (AIM, UN, MPT, SK, DS); statistical analysis (UN); administrative, technical, or logistic support (AIM); and supervision (DS).

Address Correspondence to: Alexandra I. Mansour, MD, Johns Hopkins Hospital, 1305 Dock St, Apt 310, Baltimore, MD 21231. Email: amansou6@jhu.edu.

REFERENCES

1. Heart disease facts. CDC. Accessed April 18, 2022. https://www.cdc.gov/heart-disease/data-research/facts-stats

2. HCUP fast stats: most common diagnoses for inpatient stays, national. Agency for Healthcare Research and Quality. Accessed November 5, 2024. https://datatools.ahrq.gov/hcup-fast-stats/?tab=national-hospital-utilization-costs&dash=75

3. Bachmann JM, Shah AS, Duncan MS, et al. Cardiac rehabilitation and readmissions after heart transplantation. J Heart Lung Transplant. 2018;37(4):467-476. doi:10.1016/j.healun.2017.05.017

4. Servey JT, Stephens M. Cardiac rehabilitation: improving function and reducing risk. Am Fam Physician. 2016;94(1):37-43.

5. Anderson L, Oldridge N, Thompson DR, et al. Exercise-based cardiac rehabilitation for coronary heart disease: Cochrane systematic review and meta-analysis. J Am Coll Cardiol. 2016;67(1):1-12. doi:10.1016/j.jacc.2015.10.044

6. Lawton JS, Tamis-Holland JE, Bangalore S, et al; Writing Committee Members. 2021 ACC/AHA/SCAI guideline for coronary artery revascularization: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. J Am Coll Cardiol. 2022;79(2):e21-e129. doi:10.1016/j.jacc.2021.09.006

7. Medina-Inojosa JR, Grace SL, Supervia M, et al. Dose of cardiac rehabilitation to reduce mortality and morbidity: a population-based study. J Am Heart Assoc. 2021;10(20):e021356. doi:10.1161/JAHA.120.021356

8. Suaya JA, Stason WB, Ades PA, Normand SL, Shepard DS. Cardiac rehabilitation and survival in older coronary patients. J Am Coll Cardiol. 2009;54(1):25-33. doi:10.1016/j.jacc.2009.01.078

9. Hammill BG, Curtis LH, Schulman KA, Whellan DJ. Relationship between cardiac rehabilitation and long-term risks of death and myocardial infarction among elderly Medicare beneficiaries. Circulation. 2010;121(1):63-70. doi:10.1161/CIRCULATIONAHA.109.876383

10. Cardiac rehabilitation. Million Hearts. Updated May 30, 2024. Accessed November 6, 2022. https://millionhearts.hhs.gov/about-million-hearts/optimizing-care/cardiac-rehabilitation.html

11. 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. 2007;116(15):1653-1662. doi:10.1161/CIRCULATIONAHA.107.701466

12. Keteyian SJ, Jackson SL, Chang A, et al. Tracking cardiac rehabilitation utilization in Medicare beneficiaries: 2017 update. J Cardiopulm Rehabil Prev. 2022;42(4):235-245. doi:10.1097/HCR.0000000000000675

13. Aragam KG, Dai D, Neely ML, et al. Gaps in referral to cardiac rehabilitation of patients undergoing percutaneous coronary intervention in the United States. J Am Coll Cardiol. 2015;65(19):2079-2088. doi:10.1016/j.jacc.2015.02.063

14. Appleby J. Cardiac rehab improves health, but cost and access issues complicate success. KFF Health News. August 31, 2016. Accessed April 21, 2023. https://kffhealthnews.org/news/cardiac-rehab-improves-health-but-cost-and-access-issues-complicate-success/

15. Johnson B, Kennedy K, Kurowski D, et al. Comparing commercial and Medicare professional service prices. Health Care Cost Institute. August 13, 2020. Accessed April 21, 2023. https://healthcostinstitute.org/all-hcci-reports/comparing-commercial-and-medicare-professional-service-prices

16. Schopfer DW, Priano S, Allsup K, et al. Factors associated with utilization of cardiac rehabilitation among patients with ischemic heart disease in the Veterans Health Administration: a qualitative study. J Cardiopulm Rehabil Prev. 2016;36(3):167-173. doi:10.1097/HCR.0000000000000166

17. Ades PA, Keteyian SJ, Wright JS, et al. Increasing cardiac rehabilitation participation from 20% to 70%: a road map from the Million Hearts cardiac rehabilitation collaborative. Mayo Clin Proc. 2017;92(2):234-242. doi:10.1016/j.mayocp.2016.10.014

18. Ritchey MD, Maresh S, McNeely J, et al. Tracking cardiac rehabilitation participation and completion among Medicare beneficiaries to inform the efforts of a national initiative. Circ Cardiovasc Qual Outcomes. 2020;13(1):e005902. doi:10.1161/CIRCOUTCOMES.119.005902

19.Merative MarketScan research databases. Merative. 2022.Accessed November 5, 2024. https://www.merative.com/content/dam/merative/documents/brief/marketscan-explainer-general.pdf

20. Billing and coding: frequency and duration for cardiac rehabilitation and intensive cardiac rehabilitation. CMS. October 1, 2015. Updated August 27, 2020. Accessed November 5, 2024. https://localcoverage.cms.gov/mcd_archive/view/article.aspx?articleInfo=53775:24

21. User Guide: Elixhauser Comorbidity Software Refined for ICD-10-CM Diagnoses, v2022.1. Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project. October 2021. Accessed June 25, 2022.
https://www.hcup-us.ahrq.gov/toolssoftware/comorbidityicd10/CMR-User-Guide-v2022-1.pdf

22. Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8-27. doi:10.1097/00005650-199801000-00004

23. Farah M, Abdallah M, Szalai H, et al. Association between patient cost sharing and cardiac rehabilitation adherence. Mayo Clin Proc. 2019;94(12):2390-2398. doi:10.1016/j.mayocp.2019.07.018

24. Lopez E, Neuman T, Jacobson G, Levitt L. How much more than Medicare do private insurers pay? a review of the literature. KFF. April 15, 2020. Accessed April 21, 2023. https://www.kff.org/medicare/issue-brief/how-much-more-than-medicare-do-private-insurers-pay-a-review-of-the-literature/

25. Beatty AL, Beckie TM, Dodson J, et al. A new era in cardiac rehabilitation delivery: research gaps, questions, strategies, and priorities. Circulation. 2023;147(3):254-266. doi:10.1161/circulationaha.122.061046

26. Thomas RJ. Cardiac rehabilitation — challenges, advances, and the road ahead. N Engl J Med. 2024;390(9):830-841. doi:10.1056/nejmra2302291

27. Chen S, Shafer PR, Dusetzina SB, Horný M. Annual out-of-pocket spending clusters within short time intervals: implications for health care affordability. Health Aff (Millwood). 2021;40(2):274-280. doi:10.1377/hlthaff.2020.00714

28. What Marketplace health insurance plans cover. HealthCare.gov. Accessed April 21, 2023. https://www.healthcare.gov/coverage/what-marketplace-plans-cover/

29. Borg S, Öberg B, Leosdottir M, Lindolm D, Nilsson L, Bäck M. Factors associated with non-attendance at exercise-based cardiac rehabilitation. BMC Sports Sci Med Rehabil. 2019;11:13. doi:10.1186/s13102-019-0125-9

30. Sukul D, Seth M, Barnes GD, et al. Cardiac rehabilitation use after percutaneous coronary intervention. J Am Coll Cardiol. 2019;73(24):3148-3152. doi:10.1016/j.jacc.2019.03.515

31. Mathews L, Brewer LC. A review of disparities in cardiac rehabilitation: evidence, drivers, and solutions. J Cardiopulm Rehabil Prev. 2021;41(6):375-382. doi:10.1097/HCR.0000000000000659

32. Bachmann JM, Huang S, Gupta DK, et al. Association of neighborhood socioeconomic context with participation in cardiac rehabilitation. J Am Heart Assoc. 2017;6(10):e006260. doi:10.1161/jaha.117.006260

33. Alter DA, Iron K, Austin PC, Naylor CD; SESAMI Study Group. Socioeconomic status, service patterns, and perceptions of care among survivors of acute myocardial infarction in Canada. JAMA. 2004;291(9):1100-1107. doi:10.1001/jama.291.9.1100

34. Mead H, Grantham S, Siegel B. Improving cardiovascular care through outpatient cardiac rehabilitation. J Cardiovasc Nurs. 2014;29(2):158-164. doi:10.1097/JCN.0b013e31828568f7

35. Kuehn BM. Post-acute care takes center stage in CMS (Centers for Medicare and Medicaid Services) plan to expand use of bundled payments for heart attack. Circulation. 2016;134(19):1503-1504. doi:10.1161/CIRCULATIONAHA.116.025796

36. Reddy SR, Ross-Degnan D, Zaslavsky AM, Soumerai SB, Wharam JF. Impact of a high-deductible health plan on outpatient sessions and associated diagnostic tests. Med Care. 2014;52(1):86-92. doi:10.1097/MLR.0000000000000008

37. Trivedi AN, Moloo H, Mor V. Increased ambulatory care copayments and hospitalizations among the elderly. N Engl J Med. 2010;362(4):320-328. doi:10.1056/NEJMsa0904533

38. Marzolini S, Blanchard C, Alter DA, Grace SL, Oh PI. Delays in referral and enrolment are associated with mitigated benefits of cardiac rehabilitation after coronary artery bypass surgery. Circ Cardiovasc Qual Outcomes. 2015;8(6):608-620. doi:10.1161/CIRCOUTCOMES.115.001751

39. Russell KL, Holloway TM, Brum M, Caruso V, Chessex C, Grace SL. Cardiac rehabilitation wait times: effect on enrollment. J Cardiopulm Rehabil Prev. 2011;31(6):373-377. doi:10.1097/HCR.0b013e318228a32f

40. Cardiac Rehabilitation Change Package. 2nd ed. CDC; 2023. Accessed November 5, 2024. https://millionhearts.hhs.gov/files/Cardiac_Rehab_Change_Pkg.pdf

41. Remler DK, Greene J. Cost-sharing: a blunt instrument. Annu Rev Public Health. 2009:30:293-311. doi:10.1146/annurev.publhealth.29.020907.090804

42. Chee TT, Ryan AM, Wasfy JH, Borden WB. Current state of value-based purchasing programs. Circulation. 2016;133(22):2197-2205. doi:10.1161/CIRCULATIONAHA.115.010268

43. MACRA. CMS. Updated September 10, 2024. Accessed December 6, 2022. https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Value-Based-Programs/MACRA-MIPS-and-APMs/MACRA-MIPS-and-APMs

44. Shields GE, Wells A, Doherty P, Heagerty A, Buck D, Davies LM. Cost-effectiveness of cardiac rehabilitation: a systematic review. Heart. 2018;104(17):1403-1410. doi:10.1136/heartjnl-2017-312809

45. Fendrick AM. Value-based insurance design: enhancing access and affordability to essential cardiovascular services. University of Michigan Center for Value-Based Insured Design. 2021. Accessed November 5, 2024. https://vbidcenter.org/wp-content/uploads/2021/02/CDCMAVBIDintro.pdf