More Than “Beating the Benchmark”: 5 Medicare ACOs, 2015-2019

ABSTRACT

Objectives: This retrospective study of a cohort of accountable care organizations (ACOs) that achieved success relative to their programmatic benchmarks in the Medicare Shared Savings Program (MSSP) sought to evaluate the impact of those ACOs relative to local trends on cost of care, inpatient utilization, the use of primary care services, and quality of care.

Study Design: This is a retrospective cohort study of the cost and utilization trends over time for beneficiaries attributed to 5 MSSP ACOs launched in 2016 compared with those of other traditional Medicare beneficiaries in the regions in which those ACOs operate.

Methods: The CMS Virtual Research Data Center was used to analyze Medicare claims data reflecting preventive and acute service utilization. Difference-in-differences analysis was done to compare performance on various end points vs regional trends.

Results: Over a 4-year period, the cohort of ACOs is estimated to have prevented 10,917 hospitalizations, 19,338 emergency department visits, and 8859 skilled nursing facility visits, compared with the region. This is believed to be largely driven by improvements in care transitions and preventive care, such as annual wellness visits, which the cohort of ACOs performed at 265% above the regional average in 2019.

Conclusions: Although it takes time to achieve results, value-based care has the potential to meaningfully reduce costs and improve the quality of care delivered by increasing preventive care and reducing utilization of acute services.

Am J Accountable Care. 2021;9(3):4-9. https://doi.org/10.37765/ajac.2021.88747

For most of modern American history, the fundamental economic incentives of the health care payment system have been to make more money by providing more services and demanding higher prices by consolidating market power. As a result, national health expenditures have grown considerably more than the economy for decades, and the Congressional Budget Office projects that Medicare’s federal Hospital Trust Fund will be insolvent by 2024.^1,2 Today, as new leadership takes the helm at HHS and CMS, the Biden administration has the opportunity to build on and scale what has been learned over the past decade about alternative payment models.

Policy makers have launched many initiatives to direct financial incentives toward improved outcomes for patients and bending the curve in total cost of care.^3,4 This movement toward “value-based care” has now spanned 3 administrations, garnered the support of both parties, and reached into health care markets in every corner of the country. The largest such initiative is the Medicare Shared Savings Program (MSSP). Since it launched in 2012, the MSSP has also proved to be one of the most successful value-based care programs.⁵ According to data from CMS, MSSP accountable care organizations (ACOs) have collectively improved the quality of health care for more than 11 million Medicare beneficiaries, while lowering costs to Medicare by an estimated $2.9 billion in the first 8 years of the program.⁶ Physician-led ACOs in particular generated 7 times as much savings as hospital-led ACOs in the 2018 performance year and generated more savings the longer they were in the program.⁷

Beyond the MSSP, private payers, including those administering commercial, Medicare Advantage, and Medicaid plans, have also moved to adopt value-based care over the past decade, with 35.8% of US health care payments flowing through advanced payment models as of 2019 (30.1% of commercial payments, 53.6% of Medicare Advantage payments, and 23.3% of Medicaid payments).⁸ Although multipayer adoption is important context for the shift to alternative payment models, this analysis focuses on a cohort of Medicare ACOs because of the robust data that CMS makes available for the MSSP program and the nationwide scale of the model.

The scale of the MSSP is clear, but the true value of its impact on our health care system is more complex to measure. The performance numbers mentioned earlier are derived from the benchmarks set by the regulations of the MSSP, a common (but not the only) way to measure the value created by alternative payment models. Evaluations of several models have found attenuated or absent cost savings compared with more rigorous counterfactuals.⁹ To more comprehensively evaluate whether ACOs in the MSSP provide value, it is also important to examine whether cost savings are associated with better care as intended or through “stinting” on care and worsening patient outcomes.¹⁰

For this case study, we evaluated the cohort of 5 physician-led ACOs launched in 2016 by a physician enablement organization, Aledade, each of which achieved success relative to their cost benchmarks. Using comprehensive claims data from the CMS Virtual Research Data Center (VRDC) for the 4-year period from 2016 through 2019, we sought to assess the cohort’s performance compared with local trends in 4 domains: cost of care; acute care utilization, such as visits to the emergency department (ED) and hospitalizations; service delivery, particularly primary care services; and quality of care.

The ACO Cohort

In 2015, Aledade began working with a cohort of independent practices in 5 states: Louisiana, Kansas, West Virginia, Mississippi, and Florida.^11-15 Despite the diverse geographies, each state was represented by an ACO that began its performance in the MSSP in the same year—2016—and received similar technology, resources, and enablement support from Aledade. With some minor changes, this cohort of ACOs has remained surprisingly consistent in both its makeup and performance. Aledade works with many ACOs; we chose to study the ACOs whose first performance year was 2016 to ensure that sufficient time had elapsed to determine whether the effects are sustained.

In 2016, the cohort included 313 health care providers with 48,156 assigned Medicare beneficiaries; in the next 3 years the numbers of providers were 601, 752, and 670, respectively, representing 72,413, 68,130, and 67,733 assigned Medicare beneficiaries. The cohort’s corresponding spend under management was $451,504,894 in 2016, $698,979,707 in 2017, $696,369,458 in 2018, and $696,516,729 in 2019. The large change in Medicare beneficiaries from 2016 to 2017 was driven largely by growth in Louisiana. The large change in health care providers from 2017 to 2018 was largely due to the inclusion of community health centers with large numbers of providers but proportionally smaller Medicare attribution.

METHODS

Analysis of utilization and cost data for these ACOs over time was performed using the SAS analytic software (SAS Institute) within the CMS VRDC with ability to run aggregate analysis using all Medicare Part A and Part B claims paid between the years of 2015 and 2019. This analysis does not include Medicare Advantage or Part D. All references to beneficiaries are limited to those who have Part A and Part B coverage for the relevant performance year.

Data were aggregated at the ACO level for each of Aledade’s 5 ACOs using the CMS MSSP attribution methodology executed within the VRDC.¹⁶ Costs were truncated at the 99th percentile for each CMS-determined eligibility category as per MSSP program methodology. To be included in the analysis, a county must have had at least 10 ACO assigned beneficiaries residing in the county. For each ACO, a regional comparator was established by weighting the counties associated with the ACO by the proportion of ACO beneficiaries residing in the county. ACO beneficiaries were excluded from the region.

Every beneficiary was assigned a risk score using the CMS Hierarchical Condition Categories risk-adjustment methodology for the associated performance year except for 2015 (2015 risk scores were calculated using 2016 methodology to limit noise introduced by the substantial methodological changes implemented between 2015 and 2016). Cost, utilization, and outcome measures were adjusted by comparing the average risk score of the studied population with the average risk score of the reference population.¹⁷ Performance data for each year were calculated using that respective year’s ACO participant composition to appropriately report data for only those providers actively participating in ACO work in that year. Utilization and outcome measures were normalized against 2015 performance by dividing each year’s performance by 2015 performance for the same ACO participant composition. The single year of 2015 was chosen to isolate change associated with the multiple performance years in the ACO.

ACO demographic and quality data were derived from CMS ACO reconciliation files. Program summary data were calculated using CMS Public Use Files.

RESULTS

The proportion of ACOs exceeding their minimum savings rate and earning shared savings increased over time, starting with 1 of 5 in 2016, increasing to 3 in 2017, and growing to all 5 achieving savings in 2018. In 2019, all these ACOs earned shared savings once more, extending their average savings against benchmark from 4.3% to 6.6% under the historical benchmarks used in 1-sided risk ACO models. Under the Pathways to Success program launched in July 2019, each ACO transitioned to more aggressive downside risk models, and because of regional efficiency earned over the first 3 years, performance against their reset program benchmarks increased to 9.2%.

Over 4 years, the cohort saved Medicare $116.5 million against benchmark and received shared savings payments of $62.2 million (53%). On net, this means that this cohort returned to Medicare $54.3 million against benchmark, representing 2.1% of its spend under management over the 4-year period, and 3.1% in 2019.

Cost of Care

By the end of 2019, the cohort’s risk-adjusted costs were 13% lower than the region’s compared with a 2015 baseline. We note that savings attributed to this cohort were measured at 9% compared with official MSSP program benchmarks. In other words, these ACOs achieved even more savings to the taxpayer than performance against the benchmark suggests.

Figure 1 shows the reductions in total cost of care, risk adjusted, for the 2016 cohort ACOs (in dark blue) compared with the regional average where these ACOs were located (in light blue). We found a difference between the ACOs and the region that grew more pronounced over time.

Utilization

Figure 2 shows a reduction in risk-adjusted utilization across acute and postacute care services across the board for the ACO cohort compared with the region (normalized against a 2015 historical baseline). This includes a reduction in ED visits (17% below the cohort’s historical baseline, and 14% better than the region’s 4-year trend), inpatient visits (20% below the cohort’s historical baseline, and 15% better than the region’s 4-year trend), and skilled nursing facility (SNF) visits (27% below the cohort’s historical baseline, and 18% better than the region’s 4-year trend) per 1000 beneficiary-months.

We again find an early difference that becomes more pronounced over time. Over 4 years, we equated this gap with an effective prevention of 10,917 hospitalizations, 19,338 ED visits, and 8859 SNF visits compared with the region.

Service Delivery

Each ACO in the cohort focused on 3 population health activities: (1) preventive care through annual wellness visits (AWVs), (2) increased access to primary care, and (3) transitional care management, which reduces hospital readmissions.

The Medicare AWV, a crucial tool for value-based care, helps a physician understand the full spectrum of health care needs for each patient and provides comprehensive screening and preventive counseling before these needs lead to hospitalizations or ED visits.¹⁸ In Figure 3, we see a consistent increase in the percentage of patients receiving AWVs, up to 70% of all attributed beneficiaries, performance that was 265% above the regional average in 2019.

Transitional care visits provide an opportunity to update care plans, reconcile or update prescriptions, and help patients avoid costly readmissions.19 The ACO cohort showed a significant increase in the number of patients who received a transitional care visit compared with total inpatient visits for the ACO population—ending 2019 at 194% higher than the regional average.

Overall, although utilization decreased for acute and postacute services, delivery of services like AWVs and transitional care visits, as well as ACO initiatives to increase primary care access (such as open scheduling, patient education, and after-hours call management) translated into an overall increase in primary care services accessed by beneficiaries. The ACO cohort demonstrated an increase in primary care visits compared with Medicare beneficiaries in their respective regions, ending 2019 at a rate 34% higher than the regional average.

Quality

For this cohort of ACOs, overall quality scores in the MSSP measurement years of 2017 through 2019 were consistently high (97%, 95%, and 96%, respectively). The MSSP averages for those years were 92%, 93%, and 94%.

Aledade focused ACO efforts on certain clinical quality measures that were believed to have the greatest impact on health and safety: glycated hemoglobin A_1C (HbA_1C) control for patients with diabetes (National Quality Forum [NQF] 0059) and blood pressure control for patients with hypertension (NQF 0018).^20-22 Our programmatic focus on these measures resulted in active improvement in these measures for the ACO cohort over the period. HbA_1C control for patients with diabetes improved 20% from 2017 to 2019. Blood pressure control for patients with hypertension improved 9%.

Patient-reported outcomes also remained very high over the years. Between 93% and 94% of patients rated their provider highly (MSSP average was 92%). Patient-reported health and functional status improved slightly from 71.7 in 2017 (MSSP average, 71.8) to 73.4 (MSSP average, 73.6) in 2019.

DISCUSSION

Although it takes time to achieve results, value-based care has the potential to meaningfully reduce costs and improve the quality of care delivered by increasing preventive care and reducing utilization of acute services.

According to public use data, in 2016, there were 100 new ACOs launched. Of that cohort, 42 have since dropped out of the MSSP program, and in the most recent performance year, 25 achieved savings. The experiences of this cohort of 5 physician-led ACOs provide potential lessons for practitioners and policy makers.

For practitioners: First, the results in value-based care are not simply an inevitable result of better financial incentives. There is significant new work required to empower providers, change workflows, and leverage data and software, which reveals itself in changes in the actual care that practices deliver. Second, value-based care, as these results show, takes time. The consistent results did not happen often within the first year, or sometimes even the second. Within 3 years, however, data show consistent results in both outcomes and savings. As providers hone and practice the delivery of care in these models, they achieve better results. Third, these changes can be reinforced by taking risk on additional patient segments, from Medicare Advantage to commercial and Medicaid.

For policy makers: First, the MSSP model design is working. The MSSP not only is the most scalable and successful value-based payment model in the United States, but it is also creating greater benefits than indicated by program benchmarks and providing a sustainable balance between rewarding historical improvement and regional attainment. Second, we know that health care providers can, in fact, make money in value-based care without stinting on services. In fact, this ACO cohort delivered more services—more preventive and coordinated care—and achieved improved outcomes and quality measures, as well as lower costs. Third, the future is not defined exclusively by large, integrated health systems. A network of independent primary care practices, like those in this ACO cohort, was able to achieve significant cost savings and quality improvements by sharing knowledge and best practices—which was demonstrated by steady improvement in key metrics. Finally, we know that value-based models are not some precious gems, unable to scale. The practices in this cohort of ACOs stretched across a wide variety of regions and patient populations, yet they achieved remarkably similar, significant, and consistent results.

As more models of value-based care are developed, it is important to continue to build on the established results of physician-led ACOs. The new administration and private insurance companies alike should cultivate a strong ACO model that (1) keeps administrative requirements to a minimum to encourage more physician-led ACOs with primary care at the center of the model; (2) encourages greater participation by building ACO models that reward both improvement over one’s own historical costs and recognizes lower costs compared with other health care providers; and (3) incentivizes quality, particularly reductions in health disparities, as a source of value along with lower costs.

CONCLUSIONS

Value-based care programs like the MSSP have achieved measurable, consistent, and accumulating results.10,23,24 Our nation’s health care leaders should take note of them and learn from them. Our health care spiral of higher spending and more services finally has an off-ramp, backed by years of data, and these 5 ACOs illuminate the path forward.

Acknowledgments

The authors would like to thank Laurel Trantham for providing access to some of the data used for the analyses in this article; Britainy Barnes, Ryann Miguel, and Corinne Sullivan for editorial support; and Evelyn Cayson, Andrew Chang, Erin Patrick, Nadine Robin, and Colleen Wagner for their work as the executive directors of the ACOs profiled in this article.

Author Affiliations: Aledade Inc (DAS, TB, BC, FM), Bethesda, MD.

Source of Funding: None.

Author Disclosures: Mr Streat, Mr Broome, Mr Chiglinsky, and Dr Mostashari are full-time employees at Aledade and hold stock options in Aledade as part of the employee benefits package. Dr Mostashari also serves on the board of directors of Aledade. Aledade is a partner organization that shares in the risk and reward of the accountable care organizations profiled in this article. Mr Broome is a board member of the National Association of Accountable Care Organizations.

Authorship Information: Concept and design (DAS, TB, BC, FM); acquisition of data (DAS, TB); analysis and interpretation of data (DAS, TB, BC, FM); drafting of the manuscript (DAS, TB, BC); critical revision of the manuscript for important intellectual content (DAS, TB, BC, FM); statistical analysis (DAS, TB); administrative, technical, or logistic support (DAS, TB, BC); and supervision (TB, FM).

Send Correspondence to: Travis Broome, MPH, MBA, Aledade Inc, 4550 Montgomery Ave, Ste 950N, Bethesda, MD 20814. Email: travis@aledade.com.

REFERENCES

1. Kamal R, McDermott D, Ramirez G, Cox C. How has U.S. spending on healthcare changed over time? Peterson-KFF Health System Tracker. December 23, 2020. Accessed November 15, 2020. https://www.healthsystemtracker.org/chart-collection/u-s-spending-healthcare-changed-time/

2. The outlook for major federal trust funds: 2020 to 2030. Congressional Budget Office. September 2, 2020. Accessed November 15, 2020. https://www.cbo.gov/publication/56523

3. Value-based programs. CMS. Accessed November 15, 2020. https://www.cms.gov/Medicare/Quality-Initiatives-
Patient-Assessment-Instruments/Value-Based-Programs/Value-Based-Programs

4. Value-based care in America: state-by-state. Primary Care Collaborative. 2019. Accessed November 15, 2020. https://www.pcpcc.org/sites/default/files/resources/%7Ba7b8bcb8-
0b4c-4c46-b453-2fc58cefb9ba%7D_Change_Healthcare_Value-Based_Care_in_America_State-by-State_Report.pdf

5. Verma S. 2019 Medicare Shared Savings Program ACO performance: lower costs and promising results under ‘Pathways to Success.’ Health Affairs. September 14, 2020. Accessed November 15, 2020. https://www.healthaffairs.org/do/10.1377/hblog20200914.598838/full/

6. Verma S. Number of ACOs taking downside risk doubles under ‘Pathways to Success.’ Health Affairs. January 10, 2020. Accessed November 15, 2020. https://www.healthaffairs.org/do/10.1377/hblog20200110.9101/full/

7. Sullivan G, Feore J. Physician-led accountable care organizations outperform hospital-led counterparts. Avalere. October 15, 2019. Accessed November 15, 2020. https://avalere.com/press-releases/physician-led-accountable-care-organizations-outperform-hospital-led-counterparts

8. Measuring progress: adoption of alternative payment models in commercial, Medicaid, Medicare Advantage, and traditional Medicare programs. Health Care Payment Learning & Action Network. October 24, 2019. Accessed June 8, 2021. https://hcp-lan.org/apm-measurement-effort/2019-apm/

9. Smith B. CMS innovation center at 10 years—progress and lessons learned. N Engl J Med. 2021;384(8):759-764. doi:10.1056/NEJMsb2031138

10. McWilliams JM, Hatfield LA, Landon BE, Hamed P, Chernew ME. Medicare spending after 3 years of the Medicare Shared Savings Program. N Engl J Med. 2018;379(12):1139-1149. doi:10.1056/NEJMsa1803388

11. LA MSSP 2016. Aledade. Accessed November 15, 2020. https://www.aledade.com/la-mssp-2016

12. KS MSSP. Aledade. Accessed November 15, 2020. https://www.aledade.com/ks-mssp

13. WV MSSP 2016. Aledade. Accessed November 15, 2020. https://www.aledade.com/wv-mssp-2016

14. MS MSSP. Aledade. Accessed November 15, 2020. https://www.aledade.com/ms-mssp

15. FL MSSP. Aledade. Accessed November 15, 2020. https://www.aledade.com/fl-mssp

16. Medicare Shared Savings Program: shared savings and losses and assignment methodology: specifications. CMS. February 2019. Accessed November 15, 2020. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/sharedsavingsprogram/Downloads/Shared-Savings-Losses-Assignment-Spec-V7.pdf

17. Risk adjustment. CMS. Accessed November 15, 2020. https://www.cms.gov/Medicare/Health-Plans/MedicareAdvtgSpecRateStats/Risk-Adjustors

18. Beckman AL, Becerra AZ, Marcus A, et al. Medicare annual wellness visit association with healthcare quality and costs. Am J Manag Care. 2019;25(3):e76-e82.

19. Patel SK, Miller A, Chen S, Lindsay A, Gray M, Su YP. Transitional care management visits to improve coordination of care. Am J Manag Care. 2021;27(4):e130-e134. doi:10.37765/ajmc.2021.88622

20. Comprehensive diabetes care: hemoglobin A1c (HbA1c) poor control (>9.0%). National Quality Forum: Quality Positioning System. Accessed May 17, 2021. https://www.qualityforum.org/QPS/QPSTool.aspx

21. Controlling high blood pressure. National Quality Forum. Accessed May 17, 2021. https://www.qualityforum.org/QPS/QPSTool.aspx

22. Hospital-wide all-cause unplanned readmission measure (HWR). National Quality Forum. Accessed May 17, 2021. https://www.qualityforum.org/QPS/QPSTool.aspx

23. Medicare Shared Savings Program accountable care organizations have shown potential for reducing spending and improving quality. HHS Office of Inspector General. August 28, 2017. Accessed November 15, 2020. https://oig.hhs.gov/oei/reports/oei-02-15-00450.asp

24. Kyle MA, McWilliams JM, Landrum MB, et al. Spending variation among ACOs in the Medicare Shared Savings Program. Am J Manag Care. 2020;26(4):170-175. doi:10.37765/ajmc.2020.42834