Even though most adult primary care physicians may not realize it, they each can be seen as a chief executive officer (CEO) in charge of approximately $10 million of annual revenue. Consider that a typical primary care physician has approximately 2000 patients, each of whom annually accounts for about $5000 for health care spending. Although primary care today accounts for only 5% of that spending, the decisions made in the primary care setting have important implications for downstream medical care, such as subspecialty referrals, imaging and other medical testing, invasive procedures, and hospitalizations. A group of 100 adult primary care physicians could potentially influence almost $1 billion in health care spending.
Yet for most physicians, practicing today certainly does not feel like being a CEO. Physicians see opportunities every day to improve quality and lower costs, but in a recent survey, the vast majority of physicians reported that they should not be expected to play a central role in controlling costs.1 They expressed this view even though most also reported that the best ways to control health care costs are through promoting continuity of care, using cost-effective treatments, chronic disease care coordination, prevention and adhering to clinical guidelines—all of which are controlled or influenced by physicians.
Primary care physicians understand that the current fee-for-service model does not provide adequate support for these interventions, and oppose cost-cutting approaches within that model, such as across-the-board payment cuts, limits on coverage, or increased administrative barriers such as prior approvals and contested payments that will further reduce their income. Despite this, only 7% of physicians enthusiastically supported moving away from the familiarity afforded by fee-for-service.1
This disconnect and the resulting frustration indicate that primary care physicians have been underused in leading health care reform. Typically, value-based purchasing schemes and pay-for-reporting measures have made only modest adjustments in physician reimbursement as incentives to improve quality and efficiency.
Primary care medical home payments have supported the implementation of structural capabilities believed to improve quality of care. But without direct incentives to control costs, implementing medical home capabilities alone may not be sufficient to reduce emergency department use, influence specialty and hospital care, or affect other major determinants of total costs.2
Moving beyond the typical medical home framework, physician-led accountable care organizations (ACOs) in Medicare, Medicaid, and private insurance plans could make primary care physicians more like CEOs, with accountability for the overall quality and cost results of their patients. In the Medicare Shared Savings Program, primary care services are the basis for assigning patients to the ACO, and 75% of the governance board seats must be held by ACO physicians. If ACOs demonstrate savings in the total cost of care while maintaining or improving quality and patient experience measures, those ACOs will be able to receive up to half of the savings without taking on downside financial risk.3 In many private insurance plans and multipayer collaboratives, primary care physicians receive some of their payments on a case basis (like a medical home) and also receive a share of the savings for reducing overall spending growth.
A key difference between physician-led ACOs compared with other ACOs, such as those organized by hospitals, is that physician-led ACOs have clearer financial benefits from reducing health care costs outside the physician group, which are much larger than physician costs. In contrast, hospital-based ACOs also receive shared savings for avoiding hospitalizations or shifting care to a less costly ambulatory setting, but those cost reductions are lost revenue for the hospital. The interests and incentives of physicians in physician-led ACOs are not similarly conflicted, and the benefits are more concentrated.
Because almost 40% of emergency department visits and roughly 10% to 17% of inpatient hospitalization costs are estimated to be preventable,4 primary care has opportunities to help reduce spending. Physician-led ACOs could also create efficient networks through their referral patterns, by partnering more closely with specialists, hospitals, diagnostic, and postacute services that provide evidence-based high-value care and that communicate and coordinate effectively. These models of care could also be more satisfying professionally because they require more intense patient engagement, align finances with a quality improvement mindset, and provide for greater autonomy.
A total of 5.3 million Medicare beneficiaries are now in Medicare ACOs, and physician-led ACOs have accounted for most of the recent growth in the Medicare ACO program. With an additional 57 organizations added in the fourth quarter of 2013, these entities have become the most common type of ACO in the United States (in numbers if not population), with 260 physician-led ACOs compared with 238 hospital-sponsored entities.5
One example is the Palm Beach ACO, which is self-funded by physicians in independent practices and has a service base of 30 000 Medicare beneficiaries. The physicians who established this ACO in 2012 have focused on understanding and addressing the key drivers of avoidable costs for their patient population. They have implemented systematic improvements in the care transitions for their patients, seeking notifications about emergency department visits, admissions, and discharges, in some cases directly to the smartphone of the physician. They have worked with nursing homes and home health agencies to set new expectations for improved communications and coordination, invested in increased patient outreach to ensure increased uptake of wellness and care transition encounters, and provided regular feedback of use patterns to physicians (K. Conroy and H. Zucker, Palm Beach Accountable Care Organization, written communication, January 31, 2014). When the early result of the first cohort of 114 Medicare Shared Savings Program ACOs were recently released, the Palm Beach ACO was 1 of 29 that exceeded the savings threshold; the ACO saved $22 million in projected costs in the first year, receiving nearly $11 million, or approximately $63 000 per member physician.6
Although these results are encouraging given the early stage of activities, the fact remains that in their first year, only 29% of the physician-led ACOs (and only 20% of the hospital-sponsored ACOs) demonstrated savings over their projected (albeit historically low) baseline rate of growth of Medicare costs. What needs to be done to support broader physician success in leading ACOs?
First, sizable start-up investments are needed for information technology and meaningful practice transformation support services. Many private insurers have provided up-front support similar to medical home payments. In addition, 35 Medicare Shared Savings Program participants have received advanced payments to support such investments. Other practices have relied on partnerships with health plans, a variety of other practice management and support organizations, and voluntary upfront contributions from the ACO physicians.
Second, lessons from successful physician-led ACOs must be widely and rapidly distributed. This new practice style requires new operating models and new skills, which take time and careful planning to implement. The core competencies successful practices should develop to improve quality and lower costs include the ability to analyze claims and clinical data to craft efficient referral networks, identify and manage individuals at highest risk of complications, track quality metrics for their patient population to identify gaps, implement protocols and decision support tools to improve compliance with guidelines, reach out to patients to establish and implement shared care plans, work with hospitals and skilled nursing facilities to improve care transitions, and increase access and communications with enrolled patients.
Third, transforming practices is difficult without practice-wide support and patient engagement. Physician-led ACOs need similar risk-sharing contracts from multiple payors—using consistent and reliable measures of accountability and ideally in conjunction with enhanced management fees or advanced payments. Payors must also provide timely and consistent access to claims data to assist with identifying patients with gaps in quality of care. Patient engagement is critical; some private insurers are now giving patients both quality information and lower co-pays (patient shared savings) when they use ACOs that reduce costs. Physicians can help develop the evidence and systems needed to speed these needed reforms.
Many primary care physicians are trying to continue their own practices, despite a sense of being overwhelmed by growing regulatory burdens alongside the frustrating history and uncertain future of physician reimbursement. Others are joining larger systems that seem better insulated from rate pressures and increasing practice costs. However, with growing pressures on fee-for-service rates and continuing gaps in the quality of care, changes in care delivery are inevitable, and physicians risk being overtaken by reforms in care that will necessarily shift leadership and control elsewhere.
Instead, an increasing number of primary care physicians see physician-led ACOs as a powerful opportunity to retain their autonomy and make a positive difference for their patients—as well as their practices’ bottom lines.
Corresponding Author: Mark McClellan, MD, PhD, The Brookings Institution, 1775 Massachusetts Ave NW, Washington, DC 20036 (email@example.com).
Published Online: April 10, 2014. doi:10.1001/jama.2014.4086.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Mostashari F, Sanghavi D, McClellan M. Health Reform and Physician-Led Accountable CareThe Paradox of Primary Care Physician Leadership. JAMA. 2014;311(18):1855-1856. doi:10.1001/jama.2014.4086