Strengthening Multipayer Collaboration: Lessons From the Comprehensive Primary Care Initiative

Abstract

Policy Points:

• Collaboration across payers to align financial incentives, quality measurement, and data feedback to support practice transformation is critical, but challenging due to competitive market dynamics and competing institutional priorities.

• The Centers for Medicare & Medicaid Services or other entities convening multipayer initiatives can build trust with other participants by clearly outlining each participant's role and the parameters of collaboration at the outset of the initiative.

• Multipayer collaboration can be improved if participating payers employ neutral, proactive meeting facilitators; develop formal decision‐making processes; seek input on decisions from practice representatives; and champion the initiative within their organizations.

Context

With increasing frequency, public and private payers are joining forces to align goals and resources for primary care transformation. However, sustaining engagement and achieving coordination among payers can be challenging. The Comprehensive Primary Care (CPC) initiative is one of the largest multipayer initiatives ever tested. Drawing on the experience of the CPC initiative, this paper examines the factors that influence the effectiveness of multipayer collaboration.

Methods

This paper draws largely on semistructured interviews with CPC‐participating payers and payer conveners that facilitated CPC discussions and on observation of payer meetings. We coded and analyzed these qualitative data to describe collaborative dynamics and outcomes and assess the factors influencing them.

Findings

We found that several factors appeared to increase the likelihood of successful payer collaboration: contracting with effective, neutral payer conveners; leveraging the support of payer champions, and seeking input on decisions from practice representatives. The presence of these factors helped some CPC regions overcome significant initial barriers to achieve common goals. We also found that leadership from the Centers for Medicare & Medicaid Services (CMS) was key to achieving broad payer engagement in CPC, but CMS's dual role as initiative convener and participating payer at times made collaboration challenging. CMS was able to build trust with other payers by clarifying which parts of CPC could be adapted to regional contexts, deferring to other payers for these decisions, and increasing opportunities for payers to meet with CMS representatives.

Conclusions

CPC demonstrates that when certain facilitating factors are present, payers can overcome competitive market dynamics and competing institutional priorities to align financial incentives, quality measurement, and data feedback to support practice transformation. Lessons from this large‐scale, multipayer initiative may be helpful for other multipayer efforts getting under way.

Over the past decade, public and private payers have increasingly focused on improving primary care, and multipayer initiatives—in which payers come together to align goals and resources to transform primary care—have sprung up across the country.1, 2, 3, 4, 5 Individually, payers often have relatively little leverage with which to influence provider behavior and encourage improvements in care; nonetheless, many offer incentives for providers to meet quality and cost goals. As a result, any given provider—which typically contracts with several different payers—often faces a large, disparate, and ultimately muddied set of financial incentives and quality metrics. Payers are increasingly recognizing that collective action can offer a more powerful, streamlined set of expectations and incentives for providers and potentially result in greater improvement in outcomes. Through such multipayer efforts, payers aim to amplify their individual efforts and drive health care delivery transformation on a broader scale.6 Collective action by a critical mass of payers—most critically, in the form of significant additional funding to practices—offers the promise of providing both the necessary motivation and the financial resources for practices to improve the way they deliver care.

A challenge for initiatives bringing together payers is learning how to foster relationships among competing—and sometimes adversarial—organizations in ways that maximize an initiative's impact. Several recent studies have examined factors that may influence the implementation of multipayer initiatives and other efforts bringing together payers, practices, and other stakeholders.6, 7, 8, 9, 10, 11, 12, 13, 14, 15 These studies point to market competitiveness and competing demands facing private payers as barriers to collaboration, and the presence of a neutral payer convener, strong leaders from stakeholder organizations, and prior collaborative experience as facilitators. For instance, studies of the Aligning Forces for Quality initiative found that neutral group facilitation was critical for keeping disparate stakeholders with competing priorities (and sometimes combative relationships) engaged and working to improve quality of care.5, 15 As another example, private payer engagement in the Multi‐Payer Advanced Primary Care Practice (MAPCP) initiative varied depending on local market contexts.2, 9 One market, Rhode Island, achieved strong multipayer engagement in MAPCP due to, in part, its relatively concentrated insurance market and its insurance commissioner's aggressive support for the initiative.

Multipayer efforts can also face important legal constraints—most notably antitrust concerns—which sometimes impede their work. Federal antitrust law prohibits payers working together to set prices and payments, which is viewed as collusion. Multipayer efforts with state governments acting as initiative conveners have been able to secure protection from federal antitrust law under the legal principle known as the state‐action immunity doctrine.3, 6 When initiatives have been convened by entities other than state governments, they generally have avoided antitrust scrutiny by coordinating and agreeing on key program design features but leaving each individual payer to negotiate directly with participating practices on payment amounts.6

Although state Medicaid agencies have some history of working with private payers, Centers for Medicare & Medicaid Services (CMS) has only more recently started to bring Medicare fee‐for‐service (FFS) dollars to the table. In large part because of its role as the single largest payer in the United States, Medicare has the potential to play an extremely influential role in multipayer initiatives. Medicare is viewed as a thought leader in the health care industry and, at times, state Medicaid agencies and private payers look to follow Medicare's approaches to improving care and reducing costs. Moreover, Medicare can bring substantial resources to practice transformation efforts. For example, once states had established MAPCP initiatives, Medicare FFS joined them and started providing enhanced payments to participating practices. Payers indicated these payments significantly increased resources available to support the primary care transformation efforts, bolstering the initiative's impact.2, 9

Although Medicare has the potential to strengthen the impact of multipayer initiatives, there is a paucity of research on how Medicare participation influences multipayer collaboration.16, 17 Understanding how CMS approaches multipayer initiatives—as the initiative convener, a large participating payer, or both—is important given its increasing involvement in multipayer initiatives and unique status among payers. As a government agency with the ability to set policies that profoundly affect other payers and stakeholders, CMS, acting as convener of a multipayer initiative, can be expected to exercise greater influence in shaping an initiative than a large private payer might have as an initiative convener. In some respects, this greater influence might confer advantages on the initiative: for example, the imprimatur of CMS as initiative convener might lead more private payers and practices to commit to an initiative than if a private payer or other party had acted as initiative convener. In other important respects, however, CMS's involvement in an initiative may pose challenges to collaboration. For example, as a federal agency, it is subject to a set of government contracting and procurement regulations that constrain the agreements it can forge with other participants and with vendors. In addition, there are key ways in which Medicare's interests as a payer may not be perfectly aligned with those of other payers. Most notably, the characteristics and needs of its patient population—Medicare FFS patients—differ from those of other payers, such as the commercial payers that focus on working‐age populations and Medicaid managed care plans that often focus on low‐income and pediatric populations. Moreover, since Medicare generally covers these patients later in life—when serious health conditions become more prevalent—and is responsible for covering them for the remainder of their lives, it may be more interested in affecting longer‐term outcomes than other insurers. In discussing issues such as quality metrics, these differences in covered populations might lead to a divergence of interests and impediments to collaboration.

The Comprehensive Primary Care (CPC) initiative, a 4‐year initiative (October 2012‐December 2016) launched by CMS, is one of the largest multipayer initiatives ever tested. In the initiative's first year, 39 payers supported nearly 500 primary care practices’ efforts to provide more comprehensive care to their more than 2.8 million patients.18, 19 CMS considered the multipayer aspect of CPC a cornerstone of its design because aligning goals, financial incentives, quality measurement, and data feedback across payers was critical to create an “environment to support comprehensive primary care” for participating practices.18 Together, CMS and other participating payers provided care management fees for 43% of all patients served by participating practices in 2015.19 For the combined first 3 program years of the 4‐year initiative, the median care management payments to practices totaled $658,000 ($186,000 per clinician).

CPC provides a unique opportunity to further understand the factors that influence multipayer collaboration. CPC's unique combination of characteristics—the presence of CMS as both overall initiative convener and largest payer, its broad‐scale deployment in 7 geographically and economically diverse regions throughout the United States, and the substantial aggregate funding from payers to support practice transformation—sets it apart from other multipayer initiatives. In this study, we address 2 main research questions: (1) how did CMS's role as the initiative convener for CPC at the national level influence multipayer collaboration?, and (2) what local factors influenced collaboration within CPC regions? We also describe the outcomes CPC payer collaborations achieved and identify lessons learned from CPC experiences that may prove useful in the design and implementation of other multipayer initiatives.

Background on CPC

CPC was a CMS‐sponsored and funded initiative; other payers joined with CMS to support practices in the initiative. For CPC, CMS selected 7 geographically diverse regions in which a preponderance of payers expressed interest in participating in the initiative, including 4 states (Arkansas, Colorado, New Jersey, and Oregon) and portions of 3 other states (New York's Capital District Hudson Valley region, Ohio/Kentucky's Cincinnati‐Dayton region, and Oklahoma's Greater Tulsa region). Across all regions, CMS signed memoranda of understanding with 39 payers.20 (When payers in multiple regions are counted once for each region in which they participate, there were 31 distinct payers at the start of CPC.) All participating payers agreed to provide practices with care management fees, the opportunity to share in savings, and data feedback. In addition, CMS agreed to financially support CPC learning activities for practices.

With the exception of payers in Oregon, participating payers indicated that all major payers in each market were participating in CPC.21 In Oregon, payers noted that several local and national plans decided to participate in other single or multipayer initiatives instead of CPC. Oregon payers expressed some concern about these nonparticipating competitors benefiting from the work of CPC payers. In a few other regions, payers indicated that several national plans were not participating in CPC; however, this did not raise concerns among participating payers because of the relatively small market shares held by the nonparticipating national plans.

Private payers participated in all 7 regions and, in 5 regions, state Medicaid agencies also participated (CMS paid the CPC care management fees for Medicaid beneficiaries). Private payers included different lines of business in CPC: commercial (26 payers), Medicare Advantage (19 payers), and Medicaid managed care (11 payers).21 All 26 payers with commercial lines of business had self‐insured clients as well as fully insured business. The self‐insured population represents more than half of all commercially insured individuals in the United States and, thus, is an important group to include in multipayer initiatives.22 However, gaining self‐insured clients’ participation has historically been challenging due, in part, to self‐insured entities’ lack of knowledge about or confusion regarding the initiatives and their reluctance to join new initiatives without clear evidence of a return on investment. CPC payers worked to overcome these challenges and self‐insured participation increased over the course of the initiative. The number of payers reporting at least some self‐insured participation in CPC increased from 14 of 26 payers with self‐insured clients in 2013 to 20 out of 26 in 2016. Similarly, the number reporting full (or near full) self‐insured participation in CPC increased from 7 to 14. Most payers with full self‐insured participation either required all self‐insured clients to participate in CPC or automatically enrolled all self‐insured clients in CPC unless these clients expressly opted out.19

To achieve CPC's aims of strengthening primary care in the nearly 500 practices participating in the initiative, CPC provided enhanced payment, data feedback, and learning activities to support a core set of primary care functions, including access and continuity, planned chronic and preventive care, risk‐stratified care management, patient and caregiver engagement, and coordination of care across the medical neighborhood. By partnering with CMS and other payers in their CPC regions, participating payers aligned incentives and substantially increased the financial support for primary care transformation—thereby providing considerable resources to participating practices while sending them a clear and uniform signal about practice transformation activities and goals. Together, CMS and other participating payers provided care management fees for 43% of patients served by participating practices. Although Medicare FFS accounted for the bulk of care management payments to participating practices, non‐Medicare FFS payers contributed about 45% of all CPC payments to practices in 2015.19 For the combined first 3 program years of the 4‐year initiative, the median care management payments to practices totaled $658,000 ($186,000 per clinician).

In addition to providing enhanced payments and data feedback to practices, payers also agreed to work to further align their supports to and expectations of CPC practices over time. During the first year of CPC, payers focused on aligning quality metrics, selecting a common approach to data feedback, and developing an approach to engage other stakeholders in these discussions.19, 21 As the initiative progressed, payers focused increasing attention on implementing their selected approach to data feedback and developing plans to sustain support to practices after CPC ended. Payers also focused their efforts on other areas, such as coordinating learning activities between CPC's learning faculty and participating payers (2 regions), encouraging additional self‐insured clients to sign on to the CPC initiative (2 regions), improving health literacy and patient education (1 region), increasing information sharing between hospitals and CPC practices to support transitional care (1 region), or aligning their messaging on shared savings approaches (1 region).

CMS and other payers met regularly to discuss collaborative priorities. Although multipayer collaboration was a key CPC activity in all 7 regions, payers’ approaches to working together differed across regions and over time. The initial frequency of payer meetings varied from weekly in Arkansas to every 2 months in New Jersey. Early in the initiative, payers in Arkansas, New York, and Ohio/Kentucky also formed multistakeholder groups to gather feedback on CPC progress from CPC practices and other stakeholders, such as consumer representatives, employers, hospital associations, or health foundations. Other regions were slower to involve these groups. However, following encouragement by CMS, payers in all regions formed multistakeholder groups by September 2014. Over time, multistakeholder meetings largely replaced payer‐only meetings. CMS, which initially funded the facilitation of both meeting types, transitioned at the end of 2014 to only funding multistakeholder group facilitation due to budget constraints.19, 21

Methods

Data

This paper draws primarily on semistructured interviews conducted annually between 2013 and 2016 with the following respondents:

• CMS staff working directly with the regions (20 interviews; 1 interview per region per year; some respondents covered multiple regions);

• CPC‐participating payers (142 interviews; all payers were interviewed in 2013 and 2014; 35 of 36 payers participating in 2015 were interviewed [1 payer declined]; 31 of 36 participating in 2016 were interviewed [5 payers declined]); and

• Payer conveners funded by CMS to facilitate CPC regional meetings (30 interviews; at least 1 interview per region per year). These organizations, which CMS referred to as “multistakeholder faculty” for CPC, focused on promoting collaboration with CPC participants by regional health care organizations, such as policy centers at state universities and health information exchanges.

The interviews covered the implementation of all aspects of the CPC model, including payers’ level of engagement, relationship dynamics among participants, factors influencing collaboration, and outcomes of collaborative efforts.

We also regularly observed CPC multipayer or multistakeholder meetings in each region over the course of the initiative (November 2012 to December 2016). Multipayer meetings included all payers participating in CPC, and multistakeholder meetings typically also included representatives of selected practices and other selected stakeholders, such as community organizations. The frequency of these meetings varied by region. We tracked payer engagement in these meetings and monitored their participation in CPC, examining reasons for withdrawal among the few payers that decided to leave the initiative.

Analysis

To understand CPC implementation, we coded interview transcripts using 2 codebooks. The first codebook included codes that corresponded to components of the intervention such as payment, data feedback, learning, and multipayer collaboration. The second codebook, used to identify barriers and facilitators associated with implementation, included codes derived from the Consolidated Framework for Implementation Research.23 We reviewed the coded data to qualitatively assess factors facilitating or hindering collaboration. Specifically, we first reviewed notes to identify salient factors that payers raised by region and by year of implementation. We then used observation notes from multipayer and multistakeholder meetings to refine those findings. Next, we analyzed the data for patterns to assess whether some factors were more (or less) common in regions that experienced more (or less) success in achieving various collaborative outcomes across payers, such as aligning quality measures, adopting a coordinated approach to providing data feedback to participating practices, and developing a unified plan for sustaining support for practice transformation beyond the life span of the initiative. We also examined whether and how these factors changed over the course of CPC, as the initiative matured.

In this paper, we blind regions in our results section, labeling them A through G, to help preserve the confidentiality of our interview respondents.

What Collaborative Outcomes Did CPC Payers Achieve?

Our paper focuses on the major collaborative outcomes that CMS outlined in its original solicitation for CPC payers—such as aligning their quality measures and adopting a coordinated approach to providing data feedback to participating practices—as well as other goals payers identified as major collaborative outcomes they sought to achieve in CPC—such as coordinating technical assistance and developing a unified plan for sustaining support for practice transformation beyond the life span of the initiative.

These collaborative outcomes are different from the initiative's ultimate desired outcomes (improved care, smarter spending, and better health), but were viewed by payers as important prerequisites for making those ultimate outcomes more attainable. As noted above, the motivation for payers to reach agreement on key collaborative outcomes was to align practice expectations and incentives and to help reduce the barriers and challenges practices face in transforming care delivery.

The timetables on which CPC payers worked to achieve these collaborative outcomes varied considerably. On one key outcome—aligning financial incentives—payers had to reach agreement with CMS prior to the initiative, as a condition for participation. (This pre‐initiative alignment of financial incentives involved almost all payers agreeing to provide participating practices with per‐member, per‐month care management fees and the opportunity to earn shared savings. Payers made final financial arrangements through contracts with each participating practice; given antitrust concerns, CMS was not involved in these financial negotiations.) Working on other outcomes—such as deciding on a common set of quality measures—consumed substantial payer time during CPC's first year, while one of the most challenging outcomes—providing aggregated data feedback to practices—became a multiyear effort for the regions that managed to achieve this goal.

For some collaborative outcomes, such as aligning quality measures, payers in all 7 CPC regions reached alignment. (Table 1 lists outcomes achieved by each region. Regions are blinded to protect respondent confidentiality and sorted by the number of collaborative outcomes on which payers achieved alignment. Region A has the most collaborative outcomes on which payers achieved alignment, and Region G had the fewest.) However, in the case of the quality measures, the impact on practices was limited; only about half of payers across all 7 regions ultimately used the common set of quality measures to determine practices’ eligibility to participate in shared savings (use did not differ substantially by region). In 5 regions, payers developed a common approach to data feedback. Regions D and F agreed to provide aligned individual reports (covering a common set of cost and service use measures in individual reports). Regions A, B, and C achieved data aggregation (producing a single report that aggregates data across payers)—a much more costly and time‐consuming effort, but one widely viewed as more useful to practices than separate reports from each payer. Although not an explicit goal of CPC, payers in Region A also collaborated with the CMS‐funded learning contractor to provide coordinated and individualized technical assistance to CPC practices.

Table 1.

CPC Collaborative Outcomes, as of December 2016a

	A	B	C	D	E	F	G	Total Regions
Aligned goals and financial incentivesb	X	X	X	X	X	X	X	7
Agreed on a set of quality measuresc	X	X	X	X	X	X	X	7
Coordinated approach to data feedback
Single, aggregated data reportd	X	X	X					3
Separate, aligned data reportse				X		X		2
Coordinated technical assistancef	X							1
Coordinated plan for aligning work with other regional initiatives	X	X	X	X	X			5
Sustained payer participation in CPC and CPC+	X	X	X	X	X	Xg	Xg	7

Abbreviations: CPC, Comprehensive Primary Care; CMS, Centers for Medicare & Medicaid Services.

Regions are blinded to protect respondent confidentiality and sorted by the number of collaborative outcomes on which payers achieved alignment. Region A had the most collaborative outcomes on which payers achieved alignment, and Region G had the fewest.

^aData derived from agendas and notes from payer and multistakeholder meetings and information provided by payer conveners and CMS staff.

^bCMS and other payer alignment of goals and financial incentives was a direct outcome of payers joining CPC, as opposed to an outcome from ongoing collaborative discussions.

^cPayers agreed on a common set of quality measures. However, only around half of payers ultimately used them to determine practices’ eligibility to participate in shared savings.

^dPayers achieved data aggregation (producing a single report that aggregates data across payers).

^ePayers aligned individual reports (covering a common set of cost and service utilization measures in individual reports). However, several payers in Region F stopped producing these reports after the second year of the initiative.

^fThe table indicates only 1 region collaborated with the CMS‐funded learning contractor to provide coordinated, individualized technical assistance to CPC practices, which was not an explicit goal of CPC. Payers in other regions were involved in CPC learning in less intensive ways, such as participating in learning sessions for practices.

^gPayers continued to participate in the initiative, but engagement in CPC meetings waned over the course of the initiative.

Additionally, payers in 5 regions (A, B, C, D, and E) coordinated CPC with other regional delivery system reform efforts, most notably with the State Innovation Models (SIM) Initiative, which is funded by CMS and led by the state's Medicaid program.1 In these regions, SIM awards were the largest initiative to support primary care transformation other than CPC and were viewed by many payers as a way to expand and sustain practice transformation started under CPC. As described by one payer convener, “One of the things that has happened as a result [of CPC] is payers have been able to move from representing their organization to each other to where they are representing the collaborative to the community.… I think that ultimately impacts the way they have approached their [SIM] and their commitment to [it].”

Another marker of collaborative success, CPC payer participation, remained notably stable in all 7 regions. Over the course of the initiative, only 3 payers across the 7 regions withdrew from the initiative. (One participating payer also acquired another participating payer.) Two of these 3 payers had fewer than 4,000 patients attributed to CPC practices; the third payer withdrew early in the initiative and the number of lives it had attributed to CPC practices is unknown. None of the 3 payers (all of which were private payers) withdrew because of dissatisfaction with CPC. Rather, 2 payers withdrew because of significant declines in their market share and 1 because its self‐insured clients would not contribute enhanced CPC payments. In addition, the majority of payers in all 7 CPC regions agreed to participate in the Comprehensive Primary Care Plus (CPC+) initiative, which CMS launched in January 2017 and builds on lessons learned from CPC.24 Specifically, as of April 2017, 27 of the 36 payers remaining in CPC planned to participate in CPC+.

In 5 regions (A, B, C, D, and E), payers not only remained in the initiative but also continued to actively engage in efforts to align supports for CPC practices throughout the initiative. In Regions F and G, payers became less actively engaged in collaborative discussions over the course of CPC, though engagement in these regions was somewhat reinvigorated after CMS announced the CPC+ initiative in April 2016.

To help structure our analysis of which facilitators and barriers most influenced CPC collaboration, we grouped regions into 3 clusters based on payers’ success achieving collaborative outcomes. Payers in Regions A, B, and C experienced the highest degree of success, followed by Regions D and E. Payers in Regions F and G achieved the fewest collaborative outcomes.

How Did CMS's Role as the National Convener for CPC Influence Multipayer Collaboration?

Factors Facilitating Collaboration

Across all 7 CPC regions, most payers identified CMS's leadership and financial contribution to CPC as critical for achieving broad payer participation and active engagement in CPC (Table 2). Many payers joined CPC because CMS participation brought substantial resources to their region, potentially increasing the impact of their ongoing initiatives.21 These payers frequently indicated that the aggregate care‐management fees for and data feedback on Medicare FFS beneficiaries that CMS contributed to CPC set this initiative apart from prior regional multipayer efforts. Because Medicare FFS accounts for a substantial share of a typical primary care practice's patient panel, prior multipayer efforts without CMS participation could not approach the critical mass that CPC attained in terms of both total enhanced payments and coordinated data feedback to practices. Additionally, in all 7 regions, payers reported that CMS funding for CPC practice learning activities and meeting facilitation, viewed as essential elements of CPC, encouraged sustained payer participation.

Table 2.

Factors Facilitating and Hindering Collaboration Between CMS and Other Payersa

	A	B	C	D	E	F	G
Factors facilitating collaboration
CMS's leadership and financial contribution to CPC encouraged sustained, active participation.	X	X	X	X	X	X	X
Factors hindering collaboration
CMS was initially viewed as an outsider by non‐CMS payers, especially local payers or those located in tight‐knit communities.	X	X	X	X	X	X	X
CMS's dual role as an initiative convener and participating payer resulted in confusion or frustration about the extent to which other payers could drive the initiative's direction.	X	X	X	X	X	X	X
CPC's effort to maintain national consistency frustrated payers desiring adaptation to local contexts.	X			X		X

Abbreviations: CPC, Comprehensive Primary Care; CMS, Centers for Medicare & Medicaid Services.

Regions received an “X” for a given factor if we identified it as facilitating or hindering collaboration through our analysis of interviews conducted with CMS, non‐CMS payers, and payer conveners. If a factor was present in a region but was not identified by respondents as influencing collaboration, that region did not receive an “X” for that factor. Regions are blinded to protect respondent confidentiality, and sorted by the number of collaborative outcomes on which payers achieved alignment. Region A had the most collaborative outcomes on which payers achieved alignment, and Region G had the fewest.

^aData derived from interviews with CPC payers, payer conveners, and other stakeholders and observations of CPC meetings.

In addition to committing substantial Medicare FFS resources, CMS also negotiated with participating payers prior to the start of CPC to help ensure that other payers were contributing adequate resources to support CPC practices, potentially reducing payers’ concerns that other payers might free ride. Of the 38 payers interviewed in the first year of the initiative, 7 reported agreeing to raise the level of their care management fees during these negotiations, 2 reduced the number of attributed lives a practice needed in order to sign a CPC agreement with the payer, and 2 national payers agreed to participate in additional CPC regions. A few payers suggested that CMS was able to garner payer participation and navigate these negotiations because, unlike other private payers in the region, it was not competing with payers for business.

Factors Hindering Collaboration

During interviews, CMS, other payers, and payer conveners identified 3 factors that made collaboration between CMS and other payers challenging (see Table 2). CMS adjusted its collaborative strategy to address these challenges and, by the third year of CPC, had improved its relationship with most payers. The factors that hindered collaboration and CMS's response to each are described below.

Establishing Trust With Other Payers

Most payers indicated that they had pursued prior single‐payer or multipayer practice transformation work and that CMS was a new partner in these efforts. As such, CMS and other payers needed to build a trusting, collaborative relationship for CPC. While the need to build trust was raised by payers in all regions, it was identified as particularly important in 3 of the regions characterized by strong prior collaborative efforts: A, D, and F. Payers in these regions, including some state Medicaid agencies, had formed strong, long‐standing relationships among themselves and viewed national CMS representatives as outsiders at the start of CPC. In these tight‐knit communities, the tendency to view both CMS and other new payers as outsiders initially resulted in an “us versus them” dynamic, which impeded collaboration at the outset.

To build trust with payers, CMS representatives worked to enhance communication with non‐CMS payers. For example, CMS increased the number of individual calls it had with payer representatives to better understand their perspectives on CPC. In response to concerns expressed by other payers on these calls, CMS made several changes to the initiative. For example, in the initiative's second year, CMS started requiring practices to report to CMS on their progress in transforming care delivery on a quarterly instead of an annual basis. CMS's decision to implement this change was reinforced by feedback from payers that annual reporting was insufficient to monitor practice change.

Additionally, CMS decided to change the staff it assigned to participate in CPC meetings. Initially, staff from CMS national headquarters served as CPC representatives in all regions; eventually, nearly all were replaced by staff from CMS regional offices, who could more easily attend meetings in person (which many payers indicated was important to building trust) and were better positioned to understand and work within the regional context. Some payers reported that this staffing change helped improve communication between CMS and other payers.

Balancing CMS's Role as Initiative Convener and Participating Payer

CMS's dual role as both the initiative's convener and a participating payer initially created tensions in all CPC regions. CMS provided a higher proportion of CPC care‐management payments for practices than other payers (accounting for 55% of total payments from all payers in CPC's third year) and funded CPC meeting facilitation and learning support to practices.19 Given its substantial role, CMS wielded considerable influence over the direction of the initiative. As a result, many participating payers reported confusion and frustration about the extent to which they could drive the initiative's direction. For example, some payers felt that CMS dictated the terms of participation prior to the start of CPC instead of negotiating on equal footing with other payers.

Further compounding other payers’ frustrations, CMS faced unforeseen challenges implementing CPC. The most serious of these challenges involved legal and bureaucratic hurdles related to government contracting and procurement guidelines, which required CMS to modify its initial approach to joining regional data aggregation efforts several times during the initiative. CMS's modifications required other payers to restart or rework processes already set in motion for selecting and contracting with a regional data aggregation vendor. In all 6 regions that initially pursued data aggregation (A, B, C, D, E, and F), payers indicated that these changes delayed data aggregation work and strained relationships between CMS and non‐CMS payers. In part due to these frustrations, payers in Regions D, E, and F ultimately decided not to pursue data aggregation.

Recognizing this issue, CMS strived to clarify its role and the limitations it faced. Specifically, CMS more clearly communicated to regional payers when it was acting as the initiative convener and when it was serving only as another payer collaborator. CMS also more clearly communicated its organizational constraints, such as federal government contracting requirements. Payers in the regions that achieved data aggregation (A, B, and C) reported that this improved communication from CMS helped enhance trust and remove roadblocks to their joint data aggregation efforts.

Balancing National and Regional Priorities

In 3 regions (A, D, and F), payers expressed frustration that CMS's need to create a single national program meant that most components of CPC had to be standardized across the 7 regions rather than tailored for local contexts. Payers in these regions had strong prior multipayer initiatives, sought to align CPC with those efforts, and were disappointed by their limited ability to do so. Frustrated with the lack of regional alignment, one payer said, “We are talking only about CPC in these meetings, but there is just so much overlap with work we are doing on a broader basis [in the region].… We want to tie this to broader conversations in other forums, where the same kind of issues are being talked about.” In Region F, these frustrations resulted in non‐CMS payers’ engagement with CPC waning and efforts to align supports for CPC stalling (while payers moved ahead with alignment work on other initiatives).

In the third year of CPC, CMS improved its relationships with payers in Regions A and D, in part by intentionally taking a back seat in some regional collaborative decisions so as not to let its own bureaucratic constraints slow the momentum achieved by regional stakeholders. For example, CMS decided that non‐CMS payers could move ahead with selecting and contracting with data aggregation vendors, with CMS later joining those efforts once they were already under way.

What Local Factors Influenced Collaboration Among Non‐CMS Payers Within CPC Regions?

From our analysis of interviews conducted with non‐CMS payers and payer conveners, we identified factors that influenced collaboration among non‐CMS payers within CPC regions. We identified 5 factors that facilitated collaboration and 3 that hindered collaboration (Table 3).

Table 3.

Factors Facilitating and Hindering Collaboration Within CPC Regionsa

	A	B	C	D	E	F	G
Factors facilitating collaboration
Payers were excited about and committed to primary care transformation.	X	X	X	X	X	X	X
Non‐CMS payer representatives with decision‐making power served as regional CPC champions.	X	X	X	X	X
Non‐CMS payers engaged practices during meetings in a meaningful way.	X	X	X	X	X
Payer convener was neutral and effective at driving payers to establish collaborative goals and take concrete steps to address them.	X	X	X		X
Non‐CMS payers developed good working relationships and/or a strong sense of community during prior multipayer collaborations.	X			X	X	X
Factors hindering collaboration
Non‐CMS payers had competing initiatives or differing priorities.	X	X	X		X	X
Region is characterized by competitive market dynamics, which inhibited payer collaboration.		X	X
One non‐CMS payer dominated the market, making it more difficult to engage others.							X

Abbreviations: CPC, Comprehensive Primary Care; CMS, Centers for Medicare & Medicaid Services.

Regions received an “X” for a given factor if we identified it as facilitating or hindering collaboration through our analysis of interviews conducted with CMS, non‐CMS payers, and payer conveners. If a factor was present in a region but was not identified by respondents as influencing collaboration, that region did not receive an “X” for that factor. Regions are blinded to protect respondent confidentiality, and sorted by the number of collaborative outcomes on which payers achieved alignment. Region A had the most collaborative outcomes on which payers achieved alignment, and Region G had the fewest.

^aData derived from interviews with CPC payers, payer conveners, and other stakeholders and observations of CPC meetings.

Factors Facilitating Collaboration

Commitment to Primary Care Transformation

In all 7 regions, payers were excited about CPC and committed to supporting practice transformation. While insufficient to overcome challenges to aligning CPC supports, payer enthusiasm helped sustain payers’ participation in CPC despite competing priorities for payers’ time.

Payer Champions

In 5 regions (A, B, C, D, and E), the presence of payer champions also helped propel collaborative efforts forward. For example, in Regions A and C, a payer champion spearheaded data aggregation efforts both by assuming a lead role on key tasks (such as reviewing vendor qualifications) and by encouraging other payers to commit the needed time and resources. Senior staff at participating payers served as particularly effective champions, often marshaling resources within their own organizations and energizing their counterparts within other payer organizations. The payer convener in Region D indicated one senior leader was “really working hard to make sure that everybody's able to move forward as a cohesive group and not being held up by corporate bureaucracy.” In contrast, a lack of strong payer leadership in Region F contributed to waning engagement of several CPC payers.

Practice Engagement

In all 7 regions, payers incorporated perspectives from representatives of CPC practices, such as clinicians and office managers. In 5 of these regions, practice engagement was identified as a key to successful collaboration (A, B, C, D, and E). In many regions, payer commitment to CPC was reinvigorated after hearing about practices’ successes in CPC as well as their challenges. Payers also indicated that practice perspectives were critical to designing and obtaining participation in data aggregation efforts and to achieving alignment between CPC and other initiatives.

Meaningful practice participation was predicated on a thoughtful engagement strategy. Payers received more useful feedback from practices when they, along with the payer convener, delineated clear roles for practices’ participation and purposely selected participants with time to commit to meetings and with skills and experience that matched the goals of the group. (In Region B, for example, practices with experience in data analysis were recruited by the payer convener to play key roles in data aggregation discussions and decisions.) Payers also found that early practice engagement helped build trust, break down silos, and create a cohesive multistakeholder group. Finally, many payers indicated that maintaining a venue for payers to discuss CPC without practices present allowed them to discuss either technical issues not requiring other stakeholders’ input or sensitive issues that payers were reluctant to raise in front of other stakeholders. For example, in multistakeholder meetings in Region D, payers learned that patient‐level data would be useful in practices’ transformation efforts and then used payer‐only meetings to finalize a plan for providing that information to practices.

Effective Payer Conveners

All 7 regions had a payer convener; in 4 regions (A, B, C, and E), payers identified their payer conveners as the most critical factor facilitating CPC collaboration. In these 4 regions, payer conveners played roles far surpassing logistical and administrative support (eg, scheduling meetings and taking notes). Instead, effective payer conveners gained participant trust; fostered positive working relationships among payers; broke down broad initiative goals into more concrete, achievable goals; and identified constructive steps to overcome barriers and make progress toward those goals. Specifically, Region B's payer convener built group cohesion by holding individual calls with payers to discuss issues they were reticent to raise in a group setting, identifying areas of common interest and concern across payers, and then focusing on this common ground in subsequent group meetings. As another example, Region C's payer convener spearheaded the development of a formal charter that outlined the goals for payer collaboration and the responsibilities of payer partners. As one payer explained, “[The payer convener] knows what's going on. They are helping us understand what we were doing. They are actually engaged, not just facilitating.”

Payer conveners were able to drive progress, in part because payers viewed them as trusted, neutral partners. In 3 regions with strong payer conveners (A, B, and E), those organizations had deep roots in the local community and built on their established rapport with payers. In the fourth region (C), an out‐of‐state organization filled this role because payers did not identify a local, neutral organization. At the outset of the initiative, this payer convener built trust and rapport with payers in the region, in part by taking the time to get to know payer representatives personally prior to diving into CPC discussions.

Prior Multipayer Collaborations

In most regions, payers had worked together previously in the areas of primary care redesign, payment reform, or health information technology. Payers in 4 regions (A, D, E, and F) reported that they had established good working relationships and/or a strong sense of community during those collaborations. Payers in these regions believed that this collaborative foundation provided them a head start at the beginning of CPC. For example, payers in Region E, disappointed that changes effected by prior initiatives had not been sustained beyond the life span of those initiatives, made it a priority to develop a strong vision for sustainability from CPC's start. In Region A, payers’ previous work together to develop a state health information exchange served as a starting point for data aggregation discussions.

However, although prior collaborative experience initially facilitated progress toward payers’ goals, by the end of the initiative's first year, there was no noticeable difference in collaborative dynamics in regions with and without prior collaborative experience.

Factors Hindering Collaboration

Competing Initiatives or Priorities

In 5 regions (A, B, C, E, and F), competing initiatives or differing priorities within and among payer organizations sometimes hindered collaboration. Most notably, in Region F, payers expressed disappointment at the lack of alignment between CPC and other multipayer initiatives in the region and at what they perceived as CPC's lack of flexibility in adapting program requirements to region‐specific contexts and needs. As a result, payer engagement in CPC waned over time, as payers and other stakeholders committed more time and resources to other key regional efforts not aligned with CPC.

In other regions (A, B, C, and E), national payers participating in multiple CPC regions were initially hesitant to join regional data aggregation efforts or participate in efforts to align CPC with other regional initiatives. Instead, these national payers expressed the desire to maintain data feedback and quality assessment approaches that were standardized across regions within their own organizations. One payer convener described this dynamic: “As a collaborative, [payers] are hitting the point where their organizational interests are bumping up against [CPC's] greater plan for the region.… That is a challenge that any collaborative faces.” Ultimately, most payers were pursuaded to join the region's collaborative efforts on data feedback and other program dimensions, largely through the efforts of skillful, persuasive payer conveners and/or payer champions.

Competitive Market Dynamics

In the absence of strong, prior working relationships among payers, competitive market dynamics in Regions B and C initially resulted in a lack of trust among payers. Early CPC discussions in these regions often became heated. However, by the second year of CPC, payers in these regions had established positive working relationships—a turnaround that they attributed largely to their payer conveners’ focused, strategic efforts to improve collaborative dynamics (see “Effective Payer Conveners” above).

Highly Consolidated Payer Market

In Region G, payers indicated that one non‐CMS payer with a dominant share of the region's commercial market tended to drive decision making. Smaller payers in the region generally indicated that they believed this dynamic to be not only inevitable but also fair, and they were willing to let the dominant private payer take the lead. However, over the course of the initiative, smaller payers’ engagement in the initiative decreased. Additionally, payers decided not to pursue a common approach to data feedback because the highly skewed market shares meant that payer investment in a common approach would add limited value for CPC practices beyond the feedback reports they were already receiving from the dominant payer.

How Do Factors Facilitating and Hindering Collaboration Interact?

Factors influencing collaboration interacted with one another in complex ways that determined each region's collaborative dynamics—a concept that includes the extent of payer participation and engagement in the initiative and the degree of cohesion and trust among payers. Some factors had the potential to reinforce or amplify the effects of other factors. For example, in Region E, the payer champion was able to leverage the relationships it had established with other payers in prior initiatives to help drive payer alignment within CPC. In other cases, facilitating factors played a countervailing or mitigating role to hindering factors. As an example, the involvement of a skilled, committed payer convener helped Regions B and C overcome barriers to collaboration such as a highly competitive payer market. (See Boxes 1 and 2 for detail on how collaborative dynamics evolved in Regions A and B.)

Box 1. A Closer Look: Overcoming Collaborative Challenges to Coordinate Technical Assistance in Region A^a .

1.

Outcome. CMS and all other payers in the region provided coordinated, enhanced technical assistance to practices. The region established a field service team made up of representatives from each of the regional payers and the CPC RLF funded through CMS to expand the capacity of technical assistance available to practices beyond what the RLF alone could provide. The team also worked collaboratively to make the clinical information provided to practices more useful.

Barriers. In the first year of the initiative, payers raised several concerns regarding the extent of learning activities offered to practices. CMS rebid the CPC learning contract in summer 2013; during the procurement process no region‐specific learning activities were held by CMS or its contractors. Additionally, payers believed that the RLF did not devote adequate staff resources to provide the level of technical assistance needed by practices. Payers also viewed the RLF as taking a directive, top‐down approach to practice coaching and other learning activities, as well as failing to provide enough concrete, actionable clinical information and individualized support to practices.

Facilitators. The following factors enabled payers to come together to enhance and coordinate technical assistance to practices:

• Prior collaborative experience and a strong sense of community. The region has a relatively concentrated payer market and, correspondingly, a small number of payers participated in CPC. Participating payers all had strong ties to the local community and had worked together on prior regional initiatives. During CPC, payers continued to build on this strong collaborative foundation by meeting frequently to discuss the initiative and other regional issues.

• Payer champions with decision‐making power. Each payer committed senior management and staff to work on CPC. These personnel were all strong advocates for CPC and practice transformation; they also had the authority to marshal resources within their organizations to solve problems encountered in CPC.

• Relevant expertise. Much of the expertise that the region had on care management prior to CPC was concentrated within one payer organization. This payer drew on its knowledge base and readily committed its care management staff to help enhance technical assistance to CPC practices.

Moreover, payers viewed the field service team as not only having the direct impact in providing more and better technical assistance to practices, but also an indirect beneficial impact for multipayer collaboration. The experience of working together on the team enhanced payers’ mutual trust and sense of shared purpose.

Abbreviations: CPC, Comprehensive Primary Care; CMS, Centers for Medicare & Medicaid Services; RLF, regional learning faculty.

^aData derived from interviews with CPC payers, payer convener, and other stakeholders and observations of CPC meetings.

Box 2. A Closer Look: Overcoming Collaborative Challenges to Achieve Data Aggregation in Region B^a .

1.

Outcome. Region achieved full CPC payer participation in data aggregation.

Barriers at the outset of CPC. Early in CPC, most payers were skeptical that they would achieve data aggregation given the following factors:

• The region's intensely competitive insurance market. Payers were reportedly so reluctant to divulge any information to rivals that might reveal either a competitive advantage or disadvantage that they discussed little of substantive value during the initiative's first year.

• Several payers prioritizing their corporate initiatives. Several payers with business in multiple regions, for example, were initially reluctant to join data aggregation efforts, instead preferring to standardize reporting across all their regions.

Facilitators. The following powerful facilitators allowed payers to overcome initial barriers:

• Strong, effective leadership from the region's payer convener.

  • The payer convener helped payers develop trust and a unified sense of purpose by expertly facilitating meetings, holding individual calls with payers to discuss issues they were reticent to raise in a group setting, and identifying common points of interest and concern across payers and emphasizing them during meetings.

  • The payer convener also thought creatively and strategically about how the region could achieve CPC goals. Most notably, the faculty suggested a novel way to achieve data aggregation not explored by other CPC regions—asking practices to partially fund it. Payers largely credit the payer convener with achieving this breakthrough by convincing all payers, and nearly all practices, to pay a share of the costs proportional to their number of CPC patients.

• Meaningful provider engagement. Payers invited practices to CPC meetings early in the initiative to help them identify measures for aggregation, design reports, and select the data aggregation vendor. Providers’ early involvement in shaping data aggregation efforts likely made providers more willing to fund part of this effort than they otherwise might have been.

Abbreviation: CPC, Comprehensive Primary Care.

^aData derived from interviews with CPC payers, payer convener, and other stakeholders and observations of CPC meetings.

While the factors that influenced collaboration between CMS and non‐CMS payers differed from those that influenced collaboration among non‐CMS payers, these 2 types of factors also interacted with one another in important ways to determine outcomes in each region. Region A, for example, began CPC with some key region‐level advantages: a robust track record of communitywide collaborations (including a health information exchange), which helped foster cohesion among regional payers and a strong shared vision for data aggregation. Initially, however, these facilitating factors were at least partially counterbalanced by regional payers’ mistrust of CMS as both an outsider and the dominant party in the initiative—one that the regional payers perceived to be unresponsive to region‐level needs. Payers became so frustrated with CMS's changing approach to data aggregation that those tensions threatened to undermine the substantial progress the region had already made toward that goal. However, in CPC's second and third years, CMS's decision to defer to regional stakeholders in regional CPC discussions helped reinvigorate other payers’ commitment to data aggregation, and ultimately, Region A succeeded in achieving aggregated data feedback.

As another example of the interaction between national and local factors, Region G began CPC with a strong collaborative foundation and a clear shared vision for how regional payers wanted to implement CPC. However, CMS's need to maintain a standardized national model limited the ability of the region's stakeholders to adapt CPC to fit their concurrent multipayer initiatives. As a result, payers shifted their attention and resources from CPC to the other key initiatives in the region.

In all regions, collaborative dynamics were subject to change over time, as key factors such as the degree of payer engagement and the level of collaborative effort could and did gain momentum either in a positive or negative direction. As an example of positive momentum, once the majority of payers in a region had made a strong commitment to funding their share of a data aggregation effort, remaining payers felt peer pressure to join the effort, according to interviews with payers and payer conveners. In contrast, in regions where discussions around data aggregation began fracturing and some payers made known their decisions not to contribute either data or funding to the effort, the remaining payers became concerned that they would be left to bear higher costs for a less comprehensive, less useful aggregated data tool. As a result, these regions saw disengagement growing among all payers and data aggregation efforts losing traction over time.

Discussion

Multipayer efforts are expected to continue gaining traction in coming years because payers see value in aligning incentives to improve care delivery. Among the most notable is CPC+, the largest multipayer initiative ever tested to improve primary care.24 Launched by CMS in January 2017 and building on lessons learned from CPC, the first round of CPC+ will support almost 2,900 practices in 14 regions (including the 7 CPC regions). Other important multipayer efforts are also under way; for example, some of the 21 states that received SIM design awards in 2014 will move forward with implementing multipayer models.1

Our findings from CPC build on previous research on other multipayer and multistakeholder initiatives in several important ways. Most importantly, this study built on the scant evidence on the role of Medicare FFS in multipayer initiatives by examining in depth the challenges and opportunities arising from CMS's unique role as a federal government agency serving as both the initiative convener and the dominant payer. As CMS continues to grow its role in multipayer initiatives, it should consider:

• Building trust with other payers early in an initiative. In several regions, CMS was viewed as an outsider joining long‐existing multipayer efforts. CMS built trust with payers by increasing the frequency and transparency of its communications with other payers and by taking payers’ suggestions for improving the initiative.

• Clarifying its role as the initiative convener and the limitations on its ability to collaborate. During CPC's first several years, CMS found its ability to collaborate with payers on regional objectives—such as data aggregation—was limited by its dual role as initiative convener and participating payer. Two strategies improved CMS's collaborative relationships with payers: (1) clearly delineating what aspects of the initiative are flexible and subject to change and which aspects must be standardized across all regions, and (2) deferring to non‐CMS payers on region‐specific collaborative discussions.

• Coordinating with other regional initiatives when possible. CMS's limited ability to coordinate with regional initiatives due to CPC's scope and design hindered collaboration in some regions. Identifying opportunities for increased alignment between future CMS efforts and regional initiatives may result in more progress toward collaborative objectives.

Given the scope of CPC and the diversity of CPC regions, our evaluation also provided an opportunity to thoroughly assess the factors influencing collaboration among non‐CMS payers in a given region. Our findings build on prior studies, providing a nuanced view of how payers can work together to achieve aligned aims. Our findings point to several strategies that multipayer initiatives should consider:

• Supporting neutral, skilled payer conveners. We found that payer conveners not only needed to be neutral (as noted in many previous studies) but also, to be most effective, needed to play an active and strategic leadership role for the initiative. In regions that achieved the most collaborative outcomes, payers indicated their strong, effective payer conveners went far beyond the simple logistics of convening meetings to take on proactive, substantive roles in fostering collaboration. For example, in the 3 CPC regions that accomplished data aggregation, payer conveners worked to build trust among payers by holding in‐person meetings, establishing formal charters or decision‐making processes, and/or holding offline discussions with payers to identify and build on areas of common interest.

• Undertaking thoughtful engagement of stakeholders beyond payers. Payer conveners and payers in most regions noted that engaging practices and other stakeholders (such as employers) helped promote CPC collaboration. Several strategies helped facilitate meaningful stakeholder engagement: delineating clear goals for engagement, selecting stakeholders with the time and skills needed to contribute to meetings, building trust among payers and other stakeholders, and maintaining the option for payers to meet about CPC without other stakeholders present if needed.

• Encouraging leadership of payer champions. Payer champions played an important role in keeping individual payer organizations and regions engaged in CPC despite competing organizational priorities. Typically, at least one payer champion per region emerged and pushed its payer peers to make decisions and keep moving forward on the region's goals.

Limitations

There are a few limitations to our findings. First, some of the factors influencing collaboration identified in this study apply specifically to CMS, including its dual role as initiative convener and participating payer and the bureaucratic constraints it faced as a federal government agency; however, state Medicaid agencies and other initiative conveners can learn from these lessons. Second, the experiences discussed here reflect only 7 regions—some of which had prior multipayer efforts—and their broader applicability to other regions is uncertain. For example, payers in some CPC regions had successfully collaborated on prior practice transformation efforts or had come together before CPC to coordinate their applications to the initiative. CPC findings may be most generalizable to multipayer initiatives with similar goals, including transforming primary care and developing a common approach to data feedback.

Conclusion

Multipayer initiatives can help align incentives for primary care transformation and ensure practices receive substantial support to improve care delivery by bringing together a consortium of payers that collectively represent a substantial market share. But given that payers compete for business and have different priorities, building trust and a common sense of purpose among payers can be challenging. The experience of CPC demonstrates that payers can overcome collaborative challenges to promote common goals and align financial incentives and quality measures, implement a common approach to data feedback, and develop a coordinated approach to support practice transformation after the initiative ends. Conveners of future multipayer initiatives can amplify their outcomes by building on lessons from the CPC initiative.