Use of Voluntary Alignment in the Next Generation ACO Model

ABSTRACT

Objectives: To describe adoption and use of voluntary alignment throughout the Next Generation Accountable Care Organization (NGACO) model from 2016 through 2021. Voluntary alignment allows Medicare beneficiaries to self-attribute to a particular medical provider or organization by signing a form or making a selection in the MyMedicare.gov portal.

Study Design: We performed mixed-methods analyses of cross-sectional survey, enrollment, and claims data and coding transcripts of interviews with NGACO leadership.

Methods: We statistically compared characteristics of NGACOs and beneficiaries that engaged in voluntary alignment compared to claims alignment. Additionally, we grouped qualitative interview responses into 2 overarching themes that emerged around NGACO leaders’ perceptions, use, and experiences with voluntary alignment.

Results: Few NGACOs engaged in widespread use of voluntary alignment. NGACOs that adopted voluntary alignment were similar to those that did not in most respects, although beneficiaries aligned through voluntary alignment were sicker and more expensive than those aligned through claims only. Many NGACO leaders reported they were content with claims-based alignment and did not think implementing initiatives to increase voluntary alignment would be worthwhile.

Conclusions: The analysis suggests possible lessons for using voluntary alignment in future ACO models. NGACO leaders perceived that the use of voluntary alignment was limited by a high implementation effort, a need for patient education, and tight administrative time frames. Perceived benefits of voluntary alignment included attribution flexibility and creating opportunities for beneficiary engagement. Some leaders suggested allowing year-round voluntary alignment sign-up to better integrate voluntary alignment into their regular office workflows.

Am J Manag Care. 2025;31(6):In Press

Takeaway Points

Use of voluntary alignment attribution by Next Generation Accountable Care Organizations (NGACOs) was limited.

• Twenty-nine percent of NGACOs attributed 1% or more of their population through voluntary alignment.
• Leaders of NGACOs described low use of voluntary alignment as being due to implementation effort, short administrative time frames, and failing to meaningfully increase the total size of their aligned population.
• Reasons for using voluntary alignment included attribution flexibility and opportunities for enhanced patient engagement.
• Characteristics of NGACOs adopting voluntary alignment were similar to those that did not, but beneficiaries using voluntary alignment were sicker, had higher costs, and used more services.

Holding health care organizations accountable for the care and outcomes of their patients requires a method for assigning patients to health care providers. This method is called attribution or alignment. Attribution in some Medicare sectors is straightforward. For example, Medicare Advantage (MA) beneficiaries choose to participate in an MA plan, which facilitates attribution of beneficiaries to MA organizations for the purposes of capitation and quality payments. Attributing beneficiaries to health care providers requires a specific methodology and can be more ambiguous in traditional Medicare fee-for-service (FFS) than in MA. These beneficiaries are not required to make active provider choices during enrollment. Instead, they can choose to obtain care from any Medicare participating provider.

In CMS’ accountable care organization (ACO) models, the vast majority of beneficiaries are passively aligned to organizations and providers using the beneficiary’s medical claims that occurred during the performance year (retrospective alignment) or prior to the performance year (prospective alignment). Claims alignment has been criticized as complex and ambiguous, particularly for beneficiaries using few health services.1,2 Claims alignment is also sensitive to changes in a patient’s care patterns or an organization’s practitioners over time and may exhibit turnover in an organization’s aligned beneficiary population. An additional issue with retrospective claims alignment is that the exact makeup of an ACO’s attributed population is not known to the ACO until the end of a performance year.

To incorporate beneficiary preferences in ACO models with prospective alignment and stabilize year-to-year beneficiary alignment to ACOs, CMS allowed voluntary alignment.3 In voluntary alignment, beneficiaries, including those new to Medicare, actively self-attribute to a participating medical provider in an ACO. The medical provider may be a nonphysician or have a non–primary care specialty, but they qualify as long as they are listed in the ACO’s provider list.4 Allowing voluntary alignment in ACO models may remove some of the arbitrary fluctuations in claims alignment, promote greater year-to-year consistency in beneficiary alignment, and hold the ACO accountable for the risk-adjusted medical expenses of a prospectively attributed population, wherein some beneficiaries self-attribute.4 Voluntary alignment may also formalize and cement beneficiary engagement, promoting CMS’s goals of enhancing quality and reducing expenditures.5

Beneficiaries may also benefit from voluntary alignment. From a beneficiary perspective, voluntary alignment can assure alignment to a particular ACO even if their choice contradicts the claims alignment algorithm.5 This alignment may provide sicker beneficiaries with the assurance of better care over time, consistency of the care team providing services to the beneficiary, and access to additional services if elected by the ACO.4

Voluntary alignment may benefit ACOs by allowing them more control over the attribution process, reducing year-to-year churn in beneficiary alignment, providing a mechanism for ACO membership growth, and strengthening the link between a beneficiary and their designated primary care provider, which could potentially lead beneficiaries to seek less care from providers outside the ACO.

Although it is permitted for ACOs in Shared Savings Program (SSP) and Next Generation Accountable Care Organization (NGACO) models, voluntary alignment remains a rarely used method of attribution.6 In 2018, fewer than 0.05% of SSP ACO beneficiaries were attributed to their ACO through voluntary alignment.6 Voluntary alignment was more common, although still infrequent, in the NGACO model. Just over 1% of NGACO-attributed beneficiaries in 2017 were voluntarily aligned.

The NGACO model’s evaluation provided an ideal opportunity to learn how voluntary alignment worked in practice and to better understand its strengths and limitations through a mixed-methods research design. Despite very limited uptake in the SSP and NGACO models, voluntary alignment was made an important feature in the ACO Realizing Health Equity, Access, and Community Health (REACH) model because beneficiaries attributed to organizations that have not traditionally served Medicare FFS patients (new entrant ACOs) may have to rely primarily on voluntary alignment during the first few performance years.7

Alignment in the NGACO Model

NGACO was an ACO model run by the CMS Innovation Center from 2016 to 2021. NGACOs were subject to high levels (80% or 100%) of 2-sided risk for shared savings and were rewarded or penalized based on the incurred parts A and B Medicare spending of their aligned population relative to their benchmark. Alignment in the NGACO model is performed prospectively so that NGACO providers know which beneficiaries will be included in the financial benchmark comparison before the start of the performance year.4,8

CMS offers 2 mechanisms for alignment in the NGACO model: claims alignment and voluntary alignment. Both claims-aligned and voluntarily aligned beneficiaries retain the right to seek care from any provider participating in Medicare FFS.5,8 Beneficiaries cannot be aligned to an NGACO if they are enrolled in an MA plan, lose Part A or Part B coverage, or have Medicare as a secondary payer.8,9

Rules for claims alignment in a given year require an NGACO to have provided the plurality of a beneficiary’s qualifying evaluation and management (QE&M) services during a 2-year alignment period preceding the year in question.8-10 Voluntary alignment provides Medicare beneficiaries with an opportunity to actively self-designate an NGACO-participating provider as their main source of care; this process supersedes the claims alignment algorithm.4,8 For a beneficiary to become voluntarily aligned to an NGACO, the NGACO must obtain a signed form from the beneficiary identifying an NGACO-participating provider as “the beneficiary’s main doctor, main provider, and/or the main place the beneficiary receives care” and the NGACO must notify CMS during the 8-month voluntary alignment period that precedes the start of a performance year.4,8,9 Additionally, all voluntarily aligned individuals must have obtained at least 1 QE&M service performed by the NGACO during the 2-year claims alignment period.8,9 Beneficiaries are not automatically realigned to the same NGACO in subsequent performance years; instead, they must remeet claims alignment or voluntary alignment requirements each year.8,9 During the fifth and sixth years of the NGACO demonstration, beneficiaries also were allowed to voluntarily align through the MyMedicare.gov portal.9,11

The Figure9 depicts how the timeline of the claims alignment and voluntary alignment periods was determined for an illustrative performance year, in this case, 2019 or calendar year 4. Specific dates coinciding with the alignment periods of all 6 of the NGACO performance years are included in eAppendix A (eAppendices available at ajmc.com).

A significant portion of voluntarily aligned beneficiaries were also claims-aligned to the same NGACO during the performance year. We refer to these beneficiaries as having dual alignment. A small proportion of voluntarily aligned beneficiaries would not have been aligned to the NGACO had it not been for these beneficiaries’ voluntary alignment forms. We refer to these beneficiaries as voluntarily aligned only because they were not also claims-aligned to the NGACO. Enrollment totals, the proportion of enrollees attributed by alignment type, and the proportion of NGACOs using voluntary alignment during all 6 years of the NGACO model are reported in Table 1.

METHODS

We conducted both quantitative and qualitative analyses to understand the adoption and use of voluntary alignment by NGACO participants. We used quantitative analysis of cross-sectional data to understand the characteristics of NGACOs associated with the adoption of voluntary alignment and the characteristics of beneficiaries attributed through voluntary alignment compared with claims-only alignment. This analysis was conducted using χ2 and regression analysis of enrollment and FFS claims data as well as NGACO characteristics collected and coded by the evaluator throughout the model from 2016 through 2021. These analyses include all NGACOs and all their attributed Medicare enrollees and were completed using SAS Enterprise Guide 7.1 (SAS Institute Inc) and Stata 16 (StataCorp LLC).

As part of the evaluation of the NGACO model, the evaluator conducted 44 interviews with leaders of NGACOs that entered the model in 2016 and 2017. One interview question asked leaders to describe actions they took to promote voluntary alignment among beneficiaries and their effectiveness or to explain why no actions were taken.12 This was the only interview question that directly asked about voluntary alignment activities during the 60- to 90-minute, semistructured interview and was the source of qualitative data for this study.

RESULTS

Characteristics of ACOs Adopting Voluntary Alignment

To understand organizational characteristics associated with NGACO use of voluntary alignment, we classified NGACOs into 2 categories: voluntary alignment adopters and nonadopters. NGACOs were classified as adopters if at least 1% of their attributed population was voluntarily aligned during at least 1 performance year. In total, there were 18 adopters and 44 nonadopters. Only 2 NGACOs had voluntary alignment rates that exceeded 10%: UniPhy ACO, based in Florida (16.4%), and Accountable Care Coalition of Southeast Texas (17.1%).

Table 2 compares market and provider characteristics of NGACO voluntary alignment adopters and nonadopters. Because some of these characteristics change over time, values associated with an NGACO’s highest annual rate of voluntary alignment were used for comparison. Data from the most recent performance year were used for NGACOs with 0 voluntarily aligned beneficiaries.

Only 1 comparison between adopter and nonadopter characteristics reached statistical significance at P < .1. Adopters were more likely to be part of the first NGACO cohort, having entered the model in 2016.

Characteristics of Voluntarily Aligned Beneficiaries

Table 3 compares the characteristics of claims-only aligned and voluntarily aligned beneficiaries in 2017 (performance year [PY] 2). The primary goal of this analysis was to identify the types of beneficiaries who select voluntary alignment. We chose PY 2 because it contained the highest number of voluntarily aligned beneficiaries. In the table, the column labeled voluntarily aligned includes beneficiaries who were dual-aligned as well as those who were voluntarily aligned only. The large sample sizes result in many statistically significant differences of modest magnitude.

One of the most striking comparisons is that the prevalence of chronic illness was substantially higher among voluntarily aligned beneficiaries than those who were claims-only aligned. Voluntarily aligned beneficiaries were also more likely to utilize Medicare annual wellness visits and home health visits in 2017. eAppendices B and C show additional comparisons across chronic-condition diagnoses and service use.

Mean annual Medicare parts A and B expenses for voluntarily aligned beneficiaries were $5068 higher than for claims-only aligned beneficiaries in 2017 ($16,187 vs $11,119). This coincides with the fact that voluntarily aligned beneficiaries were, on average, older and had more chronic conditions. In addition to these raw cost differences, Table 4 reports cost differences after adjusting for age, sex, 28 chronic conditions, and NGACO fixed effects using multiple linear regression. These adjustments reduce but do not eliminate the higher spending rates of voluntarily aligned vs claims-only aligned beneficiaries. After adjustment, voluntarily aligned beneficiaries spent $1490 more than claims-only aligned beneficiaries, and the difference was statistically significant.

Perspectives on Voluntary Alignment From NGACO Leaders

We grouped qualitative interview responses into 2 overarching themes that emerged around NGACO leaders’ perceptions, use, and experiences with voluntary alignment in the first year of their participation in the model: those expressing perceived barriers to and challenges of voluntary alignment and those expressing perceived benefits.

Barriers to and Challenges of Voluntary Alignment Perceived by NGACO Leaders in the First Year of Model Participation

Implementation effort. Leaders used phrases such as “operationally intensive” and “an extreme amount of work” to describe the process of crafting and administering a voluntary alignment initiative. These phrases may reflect the level of effort of administering these initiatives relative to the increase in aligned beneficiaries or other returns NGACOs realized. Leaders identified logistical challenges requiring intensive work, such as getting providers on board, determining which beneficiaries should be included in voluntary alignment outreach, obtaining CMS’s approval of outreach letters, mailing letters, training staff to handle beneficiary calls in response to letters, collecting beneficiary signatures, and reporting the signatures to CMS.

Short time frame. Some leaders expressed frustration with the relatively short time frame in which NGACOs had to collect beneficiary signatures. They would have preferred a year-round voluntary alignment period to better integrate voluntary alignment into their regular office workflow or extending the 8-month deadline to allow more time to collect and report beneficiary signatures.

Lack of increase in aligned population. Leaders expressed satisfaction with traditional claims-based alignment and skepticism that voluntary alignment would meaningfully increase the size of their attributed population. Leaders at a few NGACOs reported scaling back voluntary alignment initiatives after previous attempts failed to enlist as many beneficiaries as anticipated.

Need to educate patients about voluntary alignment. Leaders reported concerns that patients may not know what voluntary alignment means. They spoke of having to alleviate concerns that voluntary alignment might increase a patient’s out-of-pocket costs or that the voluntary alignment letter sent by the provider was fraudulent.

Benefits of Voluntary Alignment Perceived by NGACO Leaders

Attribution flexibility. Some appreciated the flexibility allowed by voluntary alignment, citing its potential to stabilize a beneficiary’s year-to-year alignment with the NGACO and using it to retain a retiring or leaving clinician’s patient population.

Beneficiary engagement. Voluntary alignment was sometimes viewed as an opportunity to encourage ongoing engagement between physician practices and their patients, standing alongside other engagement opportunities like Medicare’s annual wellness visit.

DISCUSSION

Few NGACOs engaged in widespread use of voluntary alignment. Adopter analysis showed that the organizational characteristics of NGACOs that did engage in voluntary alignment were not very different from those that did not, although there was a slight overrepresentation of NGACO adopters from the initial 2016 cohort. One potential explanation for the cohort differences may be that NGACOs in the later cohorts may have had fewer performance years in which to implement voluntary alignment.

Data from 2017 revealed that beneficiaries who were attributed through voluntary alignment (including dual-aligned and voluntarily aligned only) were older and had more chronic conditions than enrollees attributed by claims alignment alone. The voluntarily aligned enrollees may have selected voluntary alignment due to their heightened health needs and familiarity with the health care system or may have had it recommended to them by their physician or NGACO staff who felt it would benefit the patient.

Looking at the interviews, many NGACO leaders appeared content with prospective claims-based alignment and did not think that implementing initiatives to increase voluntary alignment would be worthwhile. Other NGACOs acknowledged voluntary alignment might be of some use but were not ready to begin organization-wide recruiting efforts due to competing priorities. However, some NGACOs did identify specific use cases for voluntary alignment. For example, Partners HealthCare Accountable Care Organization, an integrated delivery system located in Massachusetts, targeted voluntary alignment information to patients enrolled in its intensive care management program.

Limitations

Several limitations of our study should be noted. First, our statistical comparisons of organizational and beneficiary characteristics should be viewed as descriptive rather than causal because there is potential for many unobserved confounding factors. Second, our finding that voluntarily aligned beneficiaries were sicker and had higher spending than those who were claims-only aligned was based on a single year of data and masked heterogeneity across different NGACOs. Third, our qualitative data and analysis assessed the perspectives of health care administrators early in their NGACO participation. Although health care administrators are a crucial factor in the adoption of voluntary alignment, the views and opinions of beneficiaries and clinicians are also important.

CONCLUSIONS

Voluntary alignment may improve year-to-year stability in beneficiary alignment to ACOs, but it was rarely used in the NGACO model. NGACO leaders proposed year-round voluntary alignment sign-up to facilitate its adoption among provider organizations. This modification was made in the CMS Innovation Center’s ACO REACH model.13 In ACO REACH, beneficiaries can voluntarily align to participant ACO REACH providers prospectively, year-round, to ACOs that elect prospective-plus alignment.13

The use of voluntary alignment in CMS models could be encouraged by increasing beneficiary demand for attribution, for example, by increasing the number of special services associated with model membership. Additionally, creating more awareness of the ability of FFS beneficiaries to self-attribute to providers through voluntary alignment using the MyMedicare.gov portal might increase provider attribution, even in cases where beneficiaries infrequently seek care from medical providers.

Future research should examine how voluntary alignment is used in other ACO models and also should investigate the effects of voluntary alignment on patient care and health care utilization. For example, it is important to understand how formalizing the patient-provider relationship through voluntary alignment could affect the stability of beneficiary alignment to an ACO over time and leakage of utilization to providers outside an ACO’s provider network. Additional research can also explore patients’ perceptions and experiences with voluntary alignment and identify ways to reduce the need for ACOs to conduct additional beneficiary education on voluntary alignment.

Author Affiliations: University of Minnesota School of Public Health (TB, BED, RDF), Minneapolis, MN; NORC at the University of Chicago (KR, DC, SC, SP), Bethesda, MD; Center for Medicare and Medicare Innovation, CMS (WL), Baltimore, MD.

Source of Funding: Research conducted for this manuscript was supported by the Center for Medicare and Medicaid Innovation under HHSM-500-2014-00035I, contract for “Evaluation of the Next Generation Accountable Care Organization (ACO) Model.”

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 (TB, BED, RDF, SC, SP); acquisition of data (KR, DC, SC, SP); analysis and interpretation of data (TB, BED, RDF, WL, SC); drafting of the manuscript (TB, BED); critical revision of the manuscript for important intellectual content (TB, BED, RDF, KR, WL, SC, SP); statistical analysis (TB, BED, DC); provision of patients or study materials (KR, DC); obtaining funding (RDF, SP); administrative, technical, or logistic support (WL, DC); and supervision (SP).

Address Correspondence to: Tyler Boese, MS, University of Minnesota School of Public Health, 420 Delaware St SE, Minneapolis, MN 55455. Email: boes0106@umn.edu.

REFERENCES

1. Heiser S, Conway PH, Rajkumar R. Primary care selection: a building block for value-based health care. JAMA. 2019;322(16):1551-1552. doi:10.1001/jama.2019.14190

2. Mehrotra A, Adams JL, Thomas JW, McGlynn EA. The effect of different attribution rules on individual physician cost profiles. Ann Intern Med. 2010;152(10):649-654. doi:10.7326/0003-4819-152-10-201005180-00005

3. Medicare Shared Savings Program: Shared Savings and Losses and Assignment Methodology. CMS. February 2021. Accessed July 14, 2023. https://www.cms.gov/files/document/medicare-shared-savings-program-shared-savings-and-losses-and-assignment-methodology-specifications.pdf-0

4. Next Generation ACO Model: Request for Applications. CMS. Accessed August 30, 2023. https://innovation.cms.gov/files/x/nextgenacorfa.pdf

5. CMS. Voluntary alignment ACO fact sheet. Rural Health Group, Inc. Accessed July 14, 2023. https://web.archive.org/web/20230128104101/https://rhgnc.org/wp-content/uploads/2020/01/Voluntary_Alignment_ACO_Fact_Sheet.pdf 

6. RE: File code CMS-1701-P. Medicare Payment Advisory Commission. October 15, 2018. Accessed July 14, 2023. https://www.medpac.gov/wp-content/uploads/import_data/scrape_files/docs/default-source/comment-letters/10152018_aco_2018_medpac_comment_v2_sec283312adfa9c665e80adff00009edf9c.pdf

7. ACO REACH. CMS. Accessed August 30, 2023. https://innovation.cms.gov/innovation-models/aco-reach

8. CMS. Next Generation ACO model: amended and restated participation agreement. Trinity Health. November 18, 2019. Accessed April 9, 2025. https://www.trinity-health.org/sites/default/files/hg_features/mercury_standard_layout/bf5a9b611eb37ae6de4c59aa218b22be.pdf

9. RTI International. Next Generation ACO Model: Calculation of the Performance Year Benchmark: Performance Years 2019 and 2020. CMS. September 2018. Accessed July 14, 2023. https://innovation.cms.gov/files/x/nextgenaco-benchmarkmethodology-py4.pdf

10. NORC at the University of Chicago. Fifth Evaluation Report: Next Generation Accountable Care Organization (NGACO) Model Evaluation. CMS. November 2022. Accessed February 26, 2023. https://www.cms.gov/priorities/innovation/data-and-reports/2022/nextgenaco-fifthevalrpt

11. Next Generation ACO Model: voluntary alignment frequently asked questions: beneficiaries and caregivers. CMS. April 2019. Accessed July 14, 2023. https://innovation.cms.gov/files/x/nextgenaco-volalignmentfaq.pdf

12. NORC at the University of Chicago. Third Evaluation Report: Technical Appendices: Next Generation Accountable Care Organization Model Evaluation. CMS. September 2020. Accessed July 14, 2023. https://innovation.cms.gov/data-and-reports/2022/nextgenaco-thirdevalrpt-appendices

13. ACO Realizing Equity, Access, and Community Health (REACH) Model: Request for Applications. CMS. February 24, 2022. Accessed August 30, 2023. https://innovation.cms.gov/media/document/aco-reach-rfa