Chapter 6 – Delivery System Reform


  • As of January 2016, 838 active accountable care organizations (ACOs) had formed in all 50 states and Washington, D.C., caring for 28.3 million individuals (source).
  • Nearly 7,000 post-acute care providers, hospitals, and physician organizations have signed up to participate in bundled-payment demonstrations (source).
  • As of July 2015, the National Committee for Quality Assurance had certified more than 53,000 clinicians and 11,300 sites as patient-centered medical homes (PCMHs) (source).

The coverage expansion elements of the Affordable Care Act (ACA) have received far more media and political attention than delivery system reforms. Nevertheless, major payment and delivery system reforms are underway in both the public and private sectors, with the aim of fundamentally moving the health care system in the United States from one that pays per service to one that encourages and rewards coordinated, efficient care. Realigning incentives is intended to improve quality while slowing the growth of health care costs.

Even if Congress and the Trump administration repeal or make substantial changes to the ACA, the imperative to find ways to keep health care cost growth under control will remain, and both private and federal delivery system reforms are likely to continue, though they may be modified from current efforts.

In 2015, the federal government spent more on health care ($936 billion) than on Social Security ($882 billion) for the first time in U.S. history (source). These health care expenditures include spending on Medicare, Medicaid, the Children’s Health Insurance Program (CHIP) and tax subsidies that help people pay for private health plans under the ACA. After a period when costs have grown more slowly than historical averages, growth rates are projected to increase to an average of 5.7 percent from 2017 through 2019 and an average of 6 percent from 2020 through 2025 (source).

Quality concerns also have been a driving factor for delivery system reform. The Institute of Medicine created a stir in 1999 when it released its landmark study, “To Err is Human,” which documented the many instances of errors by medical providers, leading to tens of thousands of unnecessary deaths (source). The second volume, released in 2001, “Crossing the Quality Chasm,” laid out a broader strategy for improving quality, going beyond errors and safety (source). The focus of the reports was on problems of unnecessary infections and other complications that can lead to death or impairment, but the reports also covered the missed opportunities to provide preventive and other known high quality care.

Despite its high spending, the U.S. health care delivery system still fares poorly in certain quality measures. High rates of medical error continue to exist. These include never events, or events that should never occur, including failure to remove surgical foreign bodies at the end of procedures; and adverse events, such as post-operative sepsis, which cannot be avoided in all cases but should be minimized (source).

Another quality indicator is hospital readmission rates. Medicare 30-day readmissions dropped from 18.4 percent in 2007 to 16.9 percent in 2013 as financial penalties for hospitals with high readmissions rates were implemented. However, significant decreases in readmissions were recorded at only 8 percent of participating hospitals (source). This suggests that the improvement is not widespread, indicating inadequate transitional care.

To address issues of cost and quality, the ACA included significant funding for demonstration projects to test various delivery system reforms as part of Medicare, Medicaid and CHIP (source). This chapter describes three of the most broadly adopted projects: Accountable Care Organizations (ACOs), patient-centered medical homes, and bundled payments. At the same time, insurers, hospitals, physicians and other health care stakeholders in the private sector have been moving on a parallel track. Many of the new approaches are geared to acute care, but others address the long-term care system. Ultimately, delivery system reform involves shifting payments from traditional fee-for-service (FFS) (which reimburses based on the number of services provided) to payment systems that incorporate some “value” of care as part of their metric, including patient outcomes and clinical guideline adherence.

Projections vary for how much savings there will be as models such as ACOs, patient-centered medical homes, and bundled payments are instituted. Some early research concluded that the cost-saving potential of these changes is significant. According to the RAND Corporation, bundling payments to cover all of the services for certain medical conditions could reduce costs by 5.4 percent between 2010 and 2019, under optimistic scenarios (source). Researchers at Dartmouth have estimated that using global payments for medical providers under such payment systems as ACOs could slow annual spending growth by a full percentage point (source).

In a March 2013 Alliance for Health Reform video, Princeton University economics professor Uwe Reinhardt, however, expressed his concern that some of the new models could have a negative effect on costs. “The big fear is that the ACO’s … could create local monopolies that could dictate to you what that bundled price would be, and some of us fear that bundled prices might be even more than what the fee-for-service for that bundle would be today,” he said (source).


The Centers for Medicare & Medicaid Services (CMS), mostly through its entity, the Center for Medicare and Medicaid Innovation (CMMI), is undertaking a large number of payment and delivery reform initiatives called for in the ACA (source).

CMMI, also called the Innovation Center, has a $10 billion budget over 10 years and has the authority to test innovative payment and service delivery models that have the potential to reduce program expenditures while also enhancing the quality of care (source). Evaluations must assess quality of care, including patient-level outcomes, patient-centered criteria and changes in spending. Upon completing the evaluation, the secretary of the Department of Health and Human Services (HHS) can use the rulemaking process to expand the model’s duration and scope, including implementation on a nationwide basis (source). From 2010 – 2015, the center’s initiatives reached more than 2.5 million patients and 60,000 clinicians across the 50 states (source).

In its first round of solicitations announced in November of 2011, CMMI received about 3,000 applications and awarded $900 million for 107 projects in both rural and urban areas in all 50 states over a three-year period. This first round was a broad solicitation, inviting proposals in all settings of care and including a focus on workforce training (source).

CMMI’s second round, launched in 2013, excludes models primarily focused on acute hospital in-patient care and emphasizes the testing of innovation in payment models. CMMI lists its four categorical goals as reducing program expenditures in outpatient and/or post-acute settings; improving care for populations with specialized needs; transforming the financial and clinical models for specific types of providers and suppliers; and improving the health of populations (source).

CMMI is overseeing the testing of a variety of models in government health care programs, including ACOs, bundled payments and patient-centered medical homes. In addition, CMMI offers grants to state governments experimenting with broader-based reforms. The State Innovations Models (SIM) initiative was designed to foster innovation in state-based, multi-payer health care delivery and payment systems. Since states have policy and regulatory authority over many aspects of the health care system, they are able to conduct reforms that involve both public and private stakeholders, and SIM grants are designed to encourage that (source).

The election of Donald Trump as president places the future of CMMI in doubt (source), and the Trump administration has already issued a proposed rule to cancel a CMMI demonstration of bundled payments under Medicare Part B (source).


In April 2015, President Barack Obama signed the Medicare Access and CHIP Reauthorization Act (MACRA) into law, which replaced the longstanding Medicare Sustainable Growth Rate (SGR) method for controlling spending by Medicare on physician services. Under MACRA, health care professionals who see Medicare patients are expected to participate in one of two separate Quality Payment Programs: the Merit-based Incentive Payment System (MIPS) or Advanced Alternative Payment Models (Advanced APMs) (source).

An APM is a payment system that provides health professionals with incentives to provide high-quality, cost-efficient care. Advanced APMs differ from other APMs by allowing providers to earn more by taking on some of the risk related to patient outcomes. MIPS, on the other hand, allows providers to remain in traditional, fee-for-service Medicare. Based on practice-specific quality data, providers in the MIPS program will see their Medicare reimbursements adjusted up or down by up to 4 percent in 2019 and up to 9 percent after 2021 (source)


Accountable Care Organizations

Accountable Care Organizations (ACOs) are designed to share financial risk and incentives across multiple health care providers, which collectively are accountable for a patient population. The ACO invests in infrastructure and redesigned care processes that provide for coordinated care, high quality and efficient service delivery (source).

The idea behind ACOs is that reforming payment for health care services will drive innovation and reform in delivery of those services. The long-term success of ACOs will depend on changes not only in how health care services are paid for, but also in how they are delivered. Organizations such as the Accountable Care Learning Collaborative and private ACO enablers are focusing on delivery issues and identifying strategies for succeeding in risk-bearing payment models (source).

As of January 2016, 477 active Medicare accountable care organizations (ACOs) had formed in 49 states, caring for more than 8.9 million individuals (source).

There are also ACO contracts being formed in the private sector. While some of these commercial ACO contracts are with Medicare, some providers are forming ACOs that only have contracts with private payers. Leavitt Partners has identified a total of 838 active ACO contracts covering 28.3 million people as of January 2016 (source). Without the legal and regulatory restrictions of Medicare ACOs, private ACO contracts don’t generally have the strict financial requirements and quality metrics of Medicare ACOs. As of May 2015, Cigna, Aetna and Anthem were the private insurers with the most ACO contracts, though other private payers were increasing their involvement (source).

Medicare offers several ACO programs, including the pre-ACA Medicare Shared Savings Program (MSSP) for fee-for-service beneficiaries, the Advance Payment ACO Model for certain eligible providers already in or interested in the Medicare Shared Savings Program, the Pioneer ACO Model for health care organizations and providers already experienced in coordinating care for patients across care settings, and, starting in 2016, the Next Generation ACO Model. These models set a standard for commercial ACOs and represent major shifts in delivery system reforms.

CMMI announced 32 participants in its Pioneer ACO Model demonstration in December 2011 (source). The performance period began on January 1, 2012. Overall, the Pioneer ACOs performed better on quality measures when compared to care in fee-for-service Medicare, when comparable data were available. Mean quality composite scores across Pioneer ACOs increased from 72 percent in the first year to 87 percent in the second year. Areas of improvement included medication reconciliation, depression screens with follow-up plans, and electronic health record qualification. Earlier results from 2013 found that Pioneer ACO beneficiaries had lower hospital admission rates for certain medical conditions and higher rates of physician follow-up after hospital admissions. As of June 2016, there were nine ACOs participating in the Pioneer ACO Model.

In May 2016, CMS stated that net saving totaled $465 million for MSSP ACOs; $43 million for Pioneer ACOs; and $40 million for Advanced Payment ACOs (source). About one-quarter of the MSSP ACOs and about half of the Pioneer ACOs received shared savings payments. CMS also stated that ACO performance on quality measures is good as or better than the traditional Medicare program overall.

Starting in 2016 with 21 participating ACOs, the Next Generation ACO model will allow ACOs with at least 10,000 beneficiaries to assume higher levels of risk and reward compared to earlier ACO models. Additionally, beneficiaries may receive coordinated care incentives (payments made directly to beneficiaries by CMS) if at least a specified percentage of their patient encounters are with Next Generation ACO’s providers/suppliers (source).

In 2012, Oregon created 16 regional ACOs for its Medicaid program. Called “coordinated care organizations (CCOs), they have enrolled approximately 90 percent of the state’s 1.1 million Medicaid enrollees (source). Each regional CCO is allocated a fix amount of funds per patient and is given the authority to spend those dollars as they see fit, provided that they meet certain quality metrics (source). As of January 2016, the CCOs had succeeded in meeting their cost and quality goals (source).

Bundled Payments

A bundled payment is a single payment for a group of medical services, and it is intended to incentivize better coordination among health care providers treating a patient during an episode of care (source).

Under this model, the payer determines this payment for different courses of treatment for a given clinical condition over a defined period of time. If total expenses for an episode of care exceed the bundle, the affiliated providers may “lose” money on that episode; conversely, if their costs are lower, the providers may share in the savings. Offering these providers a single, bundled payment for an episode of care, rather than individual fees for services provided, makes them jointly accountable for the patient’s care. It also allows providers to achieve savings based on effectively managing resources as they provide treatment to the beneficiary throughout the episode.

Some of the highest-profile bundled payment initiatives are happening within Medicare. In January 2013, CMS announced a three-year bundled payment demonstration, called the Bundled Payments for Care Improvement (BPCI) Initiative. The bundling demonstration tests four payment models aimed at achieving higher quality and more coordinated care at a lower cost for Medicare. CMMI’s bundled payment demonstration offers four different models. The first includes an episode of care focused on acute care inpatient hospitalization. Two involve a retrospective bundled payment arrangement. The fourth model involves a prospective bundled payment arrangement (source).

Under the BCPI initiative, organizations enter into payment arrangements that include financial and performance accountability for episodes of care. Participants choose from 48 episodes, including, for example, pacemakers, stroke, diabetes and renal failure (source). In the private sector, commercial insurers and employers are also adopting bundled payment initiatives.

In November 2015, CMS passed its first mandatory bundled payments program, the Comprehensive Care for Joint Replacement (CJR) Final Rule. This mandate holds participating hospitals in certain markets financially accountable for generating savings for total hip and/or knee replacements (source).

Patient-Centered Medical Homes

A patient-centered medical home (PCMH) is not a place, but a model for delivering comprehensive primary care services. It is patient centered and focused on coordinated, quality care.

There are many ways a medical practice can adopt the PCMH model, but most of them incorporate a team-based approach to care that seeks to coordinate patients’ use of appropriate medical services, and increasingly connects them to community resources that can support their health (source). Many PCMH practices do not seek any sort of certification, but national accrediting bodies, some states, and some private insurers do offer accreditation. This accreditation can be necessary to participate in health plans that offer increased reimbursement for adopting the PCMH model (source).

CMMI is currently testing three PCMH models, including Multi-Payer Advanced Primary Care Practice (MAPCP), Comprehensive Primary Care (CPC), and Independence at Home (IAH). MAPCP and CPC both test multi-payer approaches where medical homes receive monthly care management fees for most of their patients across multiple insurers (including commercial plans, Medicare, and Medicaid). IAH providers are not paid monthly care management fees, but receive incentive payments for providing services to chronically ill beneficiaries at their own homes (source).

In 2012, more than 90 commercial insurance plans, multiple employers, about half of Medicaid programs, federal agencies, the Department of Defense, hundreds of safety net clinics and thousands of small and large clinical practices were using the PCMH model.

CMMI has required state governments to recruit private plans to cooperate with Medicaid and Medicare in setting up PCMH programs (the Comprehensive Primary Care Initiative and the Multi-payer Advanced Primary Care Practice Demonstration).

Approximately 7,000 practices have been accredited or recognized as a PCMH by groups such as the Joint Commission and the National Committee for Quality Assurance, while additional provider groups operate as medical homes in collaboration with private insurers and state Medicaid programs (source).

As of July 2015, the National Committee for Quality Assurance had certified more than 53,000 clinicians and 11,300 sites as patient centered medical homes (PCMHs) (source).


Melinda Abrams, vice president, delivery system reform, The Commonwealth Fund, 212/606-3831,

Deborah Bachrach, partner, Manatt Health Solutions, 212-790-4594,

Michael Barr, executive vice president, National Committee for Quality Assurance, 202/955-3500,

Robert Berenson, institute fellow, Urban Institute, 202/261-5709,

David Blumenthal, president, The Commonwealth Fund, 212/606-3825,

Cristina Boccuti, associate director, Program on Medicare Policy, Kaiser Family Foundation, 202/347-5270

Amitabh Chandra, professor, social policy, Harvard University, 617/496-7356,

Michael Chernew, professor of health care policy, Harvard Medical School, Harvard University, 617/432-0174,

Sabrina Corlette, research professor and project director, Georgetown University Center on Health Insurance Reforms, 202-687-3003,

Suzanne Delbanco, executive director, Catalyst for Payment Reform, 510/328-3328,

Mark Ganz, president and CEO, Cambia Health Solutions, 503/225-5229,

Paul Ginsburg, director, Leonard D. Schaeffer Initiative for Innovation in Health Policy, University of Southern California, 202-797-6268,

Dana Goldman, Leonard D. Schaeffer director’s chair, University of Southern California Center for Health Policy and Economics, 213/821-7948,

Darin Gordon, founding partner, Speire Healthcare Strategies, 615-386-7061,

Kate Goodrich, Centers for Medicare & Medicaid Services, 410-786-6841,

Ann Greiner, president and CEO, The Patient-Centered Primary Care Collaborative, 202-417-2062,

Benjamin Isgur, PwC Health Research Institute, 214-754-5091,

Ashish Jha, director, Harvard Global Health Institute, 617-384-5367,

Timothy Jost, Robert L. Willett family professor of law, Washington and Lee School of Law, 540-564-2524,

Cindy Mann, partner, Manatt Health Solutions, 202-585-6572,

Douglas McCarthy, senior research director, The Commonwealth Fund, 970/259-7961,

John McCarthy, founding partner, Speire Healthcare Strategies, 615-386-7061,

Robert Mechanic, senior fellow, Heller School of Social Policy and Management, Brandeis University,

Dan Mendelson, president, Avalere Health, 202/207-1310,

Thomas Miller, resident fellow, American Enterprise Institute, 202/862-5886,

Elizabeth Mitchell, president and CEO, Network for Regional Healthcare Improvement, 202-747-5104,

MaryBeth Musumeci, associate director, Kaiser Commission on Medicaid and the Uninsured, Kaiser Family Foundation, 202/347-5270

Tricia Neuman, senior vice president, Kaiser Family Foundation, 202/347-5270,

Julia Paradise, associate director, Kaiser Commission on Medicaid and the Uninsured, Kaiser Family Foundation, 202/347-5270

Kavita Patel, nonresident fellow, economic studies, The Brookings Institution, 310/968-1627

Eleanor Perfetto, senior vice president, strategic initiatives, National Health Council, 202/785-3910,

Ed Pezalla, vice president, national medical director, pharmaceutical policy and strategy, Aetna, 415/502-1603,

Pamela Riley, assistant vice president, delivery system reform, The Commonwealth Fund, 212/606-3862,

Chas Roades, chief research officer, The Advisory Board Company, 202/266-5326,

John Rother, President and CEO, National Coalition on Health Care, 202/638-7151,

Robin Rudowitz, associate director, Kaiser Commission on Medicaid and the Uninsured, Kaiser Family Foundation, 202/347-5270,

Eric Schneider, senior vice president for policy and research, The Commonwealth Fund, 212/606-3864,

Tanya Shah, senior program officer, delivery system reform, The Commonwealth Fund, 212/606-3856,

David Share, senior vice president for value partnerships, Blue Cross Blue Shield of Michigan, 313/448-6142,

Sara Traigle van Geertruyden, executive director, Partnership to Improve Patient Care, 202/688-0226,

Joel Zinberg, visiting scholar, American Enterprise Institute,

Stephen Zuckerman, codirector, Urban Institute Health Policy Center, 202/261-5709,

This guide was made possible with the support of the National Institute for Health Care Management (NIHCM) Foundation. This edition of the Sourcebook also had initial support from the Robert Wood Johnson Foundation.