Hospitals and physicians increasingly are asking about the future of alternative payment programs, especially the comprehensive primary care initiative (Section 3021), accountable care organizations (Section 3022) and bundled payment programs (Section 3023) included in the Affordable Care Act. Per the Centers for Medicare & Medicaid Services, participation by hospitals and physicians has been widespread but results have been mixed.


The use of alternative payment programs to rein in health spending by improving care coordination is not new. Two decades ago they were called “pay for performance” programs, or P4P, prompted by annual health cost increases of 7 percent and mixed results from managed care efforts in the ’90s. Most P4P programs targeted hospitals; a few targeted physicians; and all resulted in a flurry of activity to create physician-hospital organizations (or PHOs), independent practitioner associations (or IPAs) and other weird arrangements so providers could collaborate in risk-sharing arrangements legally. The results were mixed: Savings for payers were modest, but important lessons were learned:

  1. It’s possible to improve quality and reduce costs simultaneously – the two are complimentary, not contradictory, aims. Data from venerable organizations like RAND and Dartmouth showed more care is not necessarily better care.
  2. There is widespread variability in the performance of providers around outcomes, costs, charges and profitability, regardless of their status as nonprofit, investor-owned or public entities. Thus, there was ample opportunity for payers to drive direct patient care to higher-performing organizations with attractive incentives that rewarded efficiency and effectiveness.
  3. Provider organizations pushed back from P4P programs if their costs for participating were too high. Most underestimated their costs for credentialing; operational changes; data collection and results reporting; information systems; and professional services provided by their lawyers, accountants and consultants. And participation by providers was especially problematic in markets where competing payers sponsored P4P programs with different specifications.
  4. Access to valid and reliable data about costs, quality, outcomes, referrals and savings was a major sticking point. Physicians were skeptical about data they believed invalid or unreliable, and suspected that payers distorted what they reported. Even measuring costs and savings was contentious: Claims data was not thought to be a complete picture of all costs, a debate that continues today.
  5. Physician performance was key to savings. P4P efforts that resulted in cost reduction centered around physician referrals and their clinical judgment.

The same, but different

That was yesterday. Today we’re calling the programs “alternative payment models,” but the aim is the same: Reduce costs by improving how care is managed. And the lessons from P4Ps are very much part of current APM pedagogy. But this time there are two key differences.

First, participation in APMs today is being driven by Medicare, not private insurers, to bend the federal government’s cost curve. Medicare is arguably the most important payer in the U.S. health system, covering 58 million seniors and disabled adults and accounting for 20 percent of total health care spending. It is not by chance that APMs were included in the Affordable Care Act to shift accountability for costs to hospitals and physicians. And it is consequential that participation in APMs is included in Medicare’s new physician pay formula, the Medicare Access and CHIP Reauthorization Act of 2015, as an option for physicians to optimize their reimbursement from Medicare.

Second, the measures used by Medicare to score the effectiveness of these programs are more sophisticated, more granular, more clinically relevant, more verifiable, more transparent and constantly improving. In the days of P4P, access to and control of data was the source of conflict: Providers were at a disadvantage because they lacked access to claims data and distrusted private insurers. Data today is more widely accessible, more sensitive to variations in the acuity of clinical populations and balanced in weighting improved quality and reduced costs. And CMS has demonstrated its willingness to modify its specifications and measures to make it easier for providers to participate successfully. That’s why every year new measures and methodologies are announced in each program.

So, will APMs will be around in coming years? Yes, but with these caveats:

  1. Incentives for participation in APMs must be increased by Medicare. Many organizations spent more to participate than they saved. For example, after three years, the costs of participation in the Medicare Shared Savings Program (ACO) exceeded savings for three out of four participants. Nonetheless, most considered their investments worthwhile as their care coordination processes improved. And in its report last month, CMS observed that savings in APMs seem to stagnate after three years – logical since early results focus on low-hanging fruit. So, as Congress debates tax reform and the federal budget, continued funding for the Center for Medicare and Medicaid Innovation, the primary overseer of APMs on behalf of CMS, and increased upside for providers in the savings they produce for Medicare are keys to the viability of APMs long-term.
  2. Consumers must be engaged. Improved quality and reduced costs are largely dependent on patient engagement, but most APMs do not include them as active participants. APMs must clearly and definitely communicate the results of their cost savings and quality improvements to patients. Sharing financial risks and savings with consumers might be game-changing for Medicare.
  3. Additional payers – private insurers, Medicaid, large employers – must piggyback on Medicare’s APM efforts. Medicaid covers 74 million people and the federal government funds 57 percent of these costs. CMS could incentivize states to standardize APM programs and share data and best practices consistent with its programs. Ditto efforts by private insurers and large employers where APM programs for the under-65 population hold great potential. Simplifying the ways provider performance is evaluated, standardizing payment methods agnostic to the sponsor, coordinating data sharing and engaging consumers as active participants in shared savings are levers the federal government has at its disposal. More consistent and widespread use of APMs to organize and evaluate costs and quality will result in lower costs and better care for everyone.

I think APMs are here to stay, but they’ll continue to morph. With that, organizational capabilities in care coordination, informatics, consumer behavior modification, medical management and cost controls will be put to the test.

Paul H. Keckley, Ph.D. (, does independent health research and policy analysis and is managing editor of The Keckley Report []. He is also a member of Speakers Express. For speaking opportunities, please contact Laura Woodburn

The opinions expressed by the author do not necessarily reflect the policy of the American Hospital Association.