https://www.ajmc.com/journals/issue/2016/2016-vol22-n6/risk-contracting-and-operational-capabilities-in-large-medical-groups-during-national-healthcare-reform
Risk Contracting and Operational Capabilities in Large Medical Groups During National Healthcare Reform

Robert. E. Mechanic, MBA, and Darren Zinner, PhD

Under the Affordable Care Act (ACA), the federal government has implemented new initiatives to hold medical groups and health systems accountable for the cost and quality of care. As of January 2016, 434 organizations, covering 7.7 million beneficiaries, were participating in the Medicare Shared Savings Program (MSSP).1 Forty-three groups—about 1.2 million beneficiaries—were in the Pioneer Accountable Care Organization (ACO), Next Generation ACO, and Comprehensive End Stage Renal Disease Care Model, where participants bear financial risk if their spending exceeds a defined budget target. The federal government is accelerating its efforts and has announced a goal of moving 50% of Medicare payments into alternative payment models by 2018.
 
Medicare payment reforms have grown rapidly, but their ability to create systematic change in healthcare delivery that leads to better quality and lower cost will be limited without complementary efforts by private payers.2 Thus far, private payment reforms have evolved more slowly. According to one estimate, private-sector ACO contracts now cover 17.2 million individuals,3 which represents about 8.5% of the population covered by private insurance. Relatively little is known about financial arrangements in private accountable care programs, although a recent survey found that about half of ACOs had contracts with private health plans and about half of those contracts included some financial risk.4
 
In 2013, we set out to collect information about the full complement of public and private contracting arrangements for a subset of large, well-organized medical practices with employed physicians. Such organizations have demonstrated an ability to control spending and deliver high-quality care relative to other providers in their markets.5-7 We sent a detailed questionnaire to the medical groups affiliated with the Council of Accountable Physician Practices (CAPP). The groups reported the proportion of their total patient revenue paid through 7 distinct payment models (fee-for-service [FFS], pay-for-performance, episode payment, partial capitation, shared savings, shared risk, and global capitation). They also provided information about physician compensation arrangements, information management capabilities, and programs to improve quality and lower spending.
 
Although the groups in this study are larger and better integrated than physician organizations nationally, this study makes several unique contributions. First, it reports on payment methods across all public and private payers as a share of each organization’s total patient revenue—information that is not available elsewhere in the literature. One recent study reported on the characteristics of the largest private health plan contracts in a national sample of ACOs4; however, organizational decisions and investments depend on financial incentives across the full portfolio of medical groups’ revenue sources. Second, as health systems enter risk-based arrangements, their success will be influenced by how new financial incentives are conveyed to physicians. This study profiles how physicians were paid in large employed medical groups and how compensation models changed between 2011 and 2013. Finally, we report on changes in payment arrangements and internal performance management programs before and after the start of MSSP and the Pioneer ACO program. Despite not being nationally representative, this analysis of “advanced” medical groups provides one barometer of how the nation’s payment reform efforts are proceeding.
 
METHODS
We received 22 completed questionnaires from the CAPP-affiliated groups (82% response rate). Eleven additional organizations affiliated with The Group Practice Improvement Network expressed interest in participating in the study. Thus, our analysis profiled 33 medical groups. An advisory committee of medical directors from CAPP groups reviewed our survey instrument, analysis plan, and study findings.
 
We divided the groups into 3 categories of approximately equal size: FFS (0%-9% risk contract revenue), mixed payment (10%-35% risk contract revenue), and risk-based (45%-100% risk contract revenue). All of the statistics presented here are simple averages, because Kaiser Permanente is the size of all the other groups combined and weighed averages would bias results.
 
We defined risk-based contract revenue as all payments made to groups based on full- or partial-capitation, shared-risk, episode-payment, and shared-savings arrangements. FFS revenue includes straight FFS and FFS with pay-for-performance incentives. Shared savings revenue could reasonably be included in the FFS category, but we included it as risk-based revenue because it creates financial incentives for groups to manage spending growth as opposed to only improving quality of care.
 
About 45% of the groups serve patients enrolled in a health plan that they or a parent health system own. The executives we interviewed viewed the arrangement as equivalent to capitation because the overall organization was at risk; consequently, we classified payments from affiliated health plans as risk-based payments.
 
RESULTS
Medical Group Characteristics
The groups vary in size, geographic location, and degree of integration with other healthcare facilities (see eAppendix, available at www.ajmc.com). The median group employed 406 full-time employed physicians, and about 40% of employed physicians were in primary care. Eighteen groups (55%) belonged to delivery systems that included hospitals. Fifteen (45%) had affiliated health plans, and 11 (33%) were not formally affiliated with hospitals or health plans.
 
2013 Risk Contracting and Operating Characteristics
Risk contracting. The 33 groups received about two-thirds of their total patient revenue from FFS payments (unweighted average) and 16% from global capitation
(Table 1). Risk-based groups reported 25% FFS revenue and 45% global capitation, whereas FFS groups reported 97% FFS revenue and 1% global capitation. Mixed-payment groups had 71% FFS revenue and 4% global capitation. The study predated Medicare’s bundled payment initiatives, and only 3 groups reported any episode-based payments. Just under half of the groups joined the Pioneer ACO program or MSSP, including three-fourths of the risk-based groups and about one-third of the others.
 
Physician compensation. The compensation models reported by risk-based groups differed substantially from the other groups. On average, 52% of risk-based group primary care compensation was based on salary and panel size and 33% was tied to production (ie, relative value units [RVUs] generated). The mixed-payment and FFS groups paid primary care physicians (PCPs) 81% and 88%, respectively, based on production (Table 2). Risk-based groups tied a greater proportion of primary care pay to performance on quality, efficiency, or patient satisfaction metrics (13% vs about 6% for the other groups).
 
Specialist compensation was 92% production-based in FFS groups, 74% in the mixed-payment groups, and 57% in risk-based groups. About 4.5% of specialist pay was tied to performance metrics versus 8.5% for PCPs.
 
Information management. Groups reported their stage of implementation for selected information management tools on the following scale: 1) none, 2) planning underway, 3) partially implemented, and 4) fully implemented. All of the groups reported having electronic medical records, and about 80% reported full implementation of a common system across their organization. Risk-based groups were more likely than other groups to report full implementation of decision-support tools, including systemwide data warehouses (80%), disease registries (50%), practice variation analysis (50%), and clinical guideline reminders (70%) (Table 2).
 
Quality and cost management. The groups reported progress implementing 14 types of quality and cost management programs based on the following scale: 1) no progress, 2) planning underway, 3) just getting started, 4) far along, and 5) advanced. Table 2 shows the proportion of each group category reporting that they were far along or advanced for each of the 14 programs.
 
Risk-based groups reported more advanced implementation than other groups. The greatest differences were in programs to reduce unnecessary hospital admissions, develop preferred relationships with efficient specialists and hospitals, manage care for high-risk patients, link physician compensation to performance, and manage postacute care. The risk-based groups were also further along on programs to reduce treatment variation, engage patients, and provide electronic consultations.
 
Mixed payment groups reported implementation levels similar to risk-based groups in care process redesign and prescription drug management. They were further along than the FFS groups in readmission reduction initiatives, high-risk care management, and programs to reduce “leakage” of their primary care patients to nonaffiliated hospitals and specialists.
 
Relatively few groups reported “advanced” implementation of the programs. The risk-based groups reported “advanced” implementation for about one-third of their responses (across all 14 programs) compared with 8% for the mixed-payment groups and 3% for FFS groups (data not shown).
 
Financial and Operational Changes (2011-2013)
Thirty-one groups reported payment model information for 2011, 2012, and 2013. We also analyzed changes in compensation, information management, and cost and quality management programs for 20 CAPP groups that completed questionnaires in both 2011 and 2013.8
 
Changes in risk contracting. We divided the 31 groups into “movers” that increased risk-based contract revenue by at least 5 percentage points and “stable” groups that increased by less than 5 percentage points. The 9 “mover” groups had an average increase of 15.4 percentage points (range = 7%-31%). The 22 “stable” groups reported average growth of only 0.2 percentage points (see eAppendix).
 
Four of the 9 “movers” already had substantial risk in 2011, and the bulk of their new risk revenue came from joining the MSSP or Pioneer ACO program. The other 5 began primarily as FFS groups in 2011; most of their new payment arrangements were commercial shared savings contracts with limited downside risk.
 
Changes in physician compensation. Medical groups that enter risk contracts may try to align physician compensation with the organization’s overall financial objectives. Eight of 18 groups that reported compensation data in both years changed the basis for PCP pay by at least 10 percentage points. The most common change was reducing the share of compensation based on production (eg, billed RVUs). Two groups reduced PCP production-based compensation by more than 40 percentage points: one group shifted 30 percentage points from production to panel size and doubled physician performance incentives from 10% to 20%, while another group shifted 40 percentage points from production to a mix of quality and patient-access metrics (eg, wait times, evening hours). Only 3 of the 18 groups made large changes in specialist compensation.
 
Changes in information management capability. The proportion of groups reporting that they had fully implemented clinical guideline reminders, an enterprisewide data warehouse, and a disease registry increased measurably from 2011 to 2013; two-thirds of the progress was reported by “mixed payment” groups.
 
Changes in quality and cost management programs. The number of groups reporting that their program implementation was “far along” increased substantially across the 14 program categories between 2011 and 2013 (see eAppendix). Of groups reporting that they were considering or getting started with these programs in 2011, 40% to 70% (depending on the program) reported progress in 2013. The programs where the greatest number of groups advanced—care management and reducing treatment variation—are important for successful performance under risk contracts. Notably, by 2013, 16 of 19 groups reported that their high-risk care management program was far along or advanced.
 
Factors Affecting Medical Group Risk Contract Adoption
We asked the groups to rate the importance of selected factors on their ability or willingness to adopt risk contracts. In 2013, 55% of groups said that the need to improve care management capability was “very important” and 45% said that the need to improve data systems was “very important.” Twenty-two percent of the groups said that limited payer willingness or ability to offer risk contracts was “very important” and 48% said it was “somewhat important.”9 More than 60% said that local market dominance of preferred provider organization insurance products that were less suited to risk contracting was either “somewhat important” or “very important.” Only 7% of the groups said physician resistance was “very important,” whereas 24% said it was “somewhat important” (data not shown).

DISCUSSION
The federal government’s goal of moving half of Medicare spending into alternative payment models by 2018 is ambitious. The pace of change will depend on the nature and scope of public and private payment arrangements currently in place, future models or model changes that make these programs more (or less) attractive, and provider system readiness. Very little systematic information about providers’ payment arrangements or capacity to manage risk-based contracts is currently available. This study provides more detailed information for a subset of advanced delivery systems as one view of the progress of payment reform.

Commercial insurers have announced plans to expand their value-based contracts to 50% or more of their total payments to providers.10 However, “value-based” is used to describe programs ranging from FFS with small quality incentive payments, to patient-centered medical home “care management” fees, to full risk contracts. Industry sources suggest that contracts with downside risk are rare in most areas of the country. Pay-for-performance, the most common form of value-based payment today, has been widely studied with mixed results.11
 
New global budget programs, in which providers bear substantial financial risk, have had positive initial results,12,13 but such programs are a minority of value-based pay initiatives and it is unknown whether they could be successful at a larger scale. CMS’ decision to continue the upside-only MSSP option through 2018 reflects a continuing reluctance of providers to take on risk.
These study results suggest the need for realistic expectations about the pace of payment system change. Most participating groups would be considered prototypical ACOs that are well positioned to succeed under risk-based arrangements. Nonetheless, two-thirds of their 2013 patient revenue came from FFS payments and only 16% from global capitation; moreover, although one-third of the groups received half or more of their revenue from risk-based contracts, another one-third were paid mostly by FFS. Additionally, only one-third added risk contracts between 2011 and 2013 despite new Medicare and private plan options.
 
CONCLUSIONS
Although many of these advanced groups have embraced payment reforms, those newer to risk-contracting appear more comfortable building infrastructure before taking on additional risk. The ACA has made payment delivery system reform top-of-mind for provider systems and health plans. Medicare’s ability to drive delivery system transformation will depend on how aggressively commercial health plans push risk contracts and whether providers will accept them. This is beginning to happen—but relatively slowly—at least through 2013. Ultimately, the experience of the advanced groups in this study suggests that expanding risk-based arrangements across the health system will likely be slower and more challenging than many people assume. 
Print | AJMC Printing...