Together, hospital and physician services account for more than half of national health spending. In its 2014 National Health Expenditures estimates, the Centers for Medicare and Medicaid Services’ actuaries make the hospital (nearly $1 trillion) and physician practice (nearly $600 billion) sectors appear to be independent and non-overlapping. This is an optical illusion. Hospitals and physicians are, in day-to-day practice, hopelessly intertwined.

And while power appears to be shifting from physicians to hospitals with the increasing salaried employment of physicians, appearances can be deceiving. This post discusses the economic power balance between hospitals and physician communities, and the policy levers that influence this complex relationship — a relationship that is evolving in a way that could increase financial pressures on both hospitals and the American health system.

Physicians and hospitals must intimately collaborate or care does not get delivered. At the same time, hospitals and physicians directly compete in surgery, imaging, and other ambulatory services. In this relationship of simultaneous competition and interdependency, the borderline between hospitals and physicians is fraught both with economic conflict and moral/legal risk.

Conflict with physicians over contracts, practice prerogatives, and scope of professional practice poses one of the single most significant career threats to hospital administrators. Hospital executive colleagues have commented to us that half or more of their job is “political” — managing the diverse economic interests of their medical staffs. One confessed that there is nothing more dispiriting in his job than fighting with physicians over money.

Hospitals absolutely cannot function without close physician collaboration. Hospitals rely on physicians to admit patients to their facilities, refer to their specialists, and to use their lucrative diagnostic services. These high technology outpatient services represent a significant fraction, some believe well more than half, of hospital profit. Hospitals also cannot grow their market shares or execute any of the new alternative payment methodologies without physician participation.

In turn, hospitals also provide income and security for physicians. This is particularly the case in the frosty northern tier of the US where the economic foundation of private medical practice is increasingly not load bearing. Seeking income security appears to be the principal motivation for many physicians seeking hospital employment. In a recent Jackson survey, more than two-thirds of hospital-employed physicians reported that they initiated the discussions that led to employment.

Historical Divisions between the Hospital and Physician Spheres

Traditionally, there was a separation between the hospital and physician spheres — a “treaty” imposed by political fiat based on the social and political power of physicians. As Paul Starr documented in his magisterial history, The Social Transformation of American Medicine, physicians expended significant political capital to avoid being captured by the hospital, to maintain both professional autonomy and control over their incomes.

In some states, the physician franchise was explicitly protected by corporate practice of medicine laws which forbade non-physician controlled enterprises from employing physicians. The hospital was intended to function as a “doctors’ workshop,” or a “physicians’ co-operative” in the words of a famous treatise in health services research literature by Mark Pauly and Michael Redisch: an economic entity used by an independent medical community to support its incomes, where physicians could practice their profession free of managerial interference. Physicians set clinical standards and controlled access to the hospital’s services through membership on the medical staff, clinical ordering, and, frequently, vendor selection. Professional ethical standards were intended to moderate physician professional conduct and efforts to maximize income.

Power Begins Shifting in the 1990s

In the early 1990s, this physician-enforced independence began to break down. In their health reform plan, the Clinton administration proposed sweeping delivery system reforms designed to merge hospital and physician spheres into integrated delivery networks, known then as “accountable health plans.” These integrated entities would employ physicians, own hospitals, and offer fixed price health services through state-wide purchasing co-operatives to patients on a per capita basis. These new integrated enterprises were modeled on the Kaiser Foundation Health Plans which employed physicians in large medical groups and also owned hospitals, and which marketed themselves exclusively through their captive insurance product.

Even though the Clinton reform initiative collapsed in 1994, the merger and acquisition boom it catalyzed actually accelerated in the following years, resulting in a wave of regional hospital mergers as well as physician practice acquisitions by hospitals. These acquisitions were encouraged in many markets by competitive pressure from for-profit physician practice management firms—such as PhyCor and MedPartners—that were acquiring large multi-specialty physician groups with the intention of contracting directly with health plans. These firms threatened to turn the hospitals into commoditized subcontractors, squeezing both hospital volumes and payment levels.

By 1998, around 70,000 full-time equivalent physicians (and dentists) were employed by hospitals, in addition to roughly 108,000 interns and residents. After the large physician practice management firms collapsed in the late 1990s, cash losses generated by physician employment forced many hospitals to pare back. Practice acquisitions resumed in earnest in 2004 and continued for the next nine years.

By 2013, hospital full-time equivalent employment had grown to more than 122,000, a 74 percent increase since 1998, according to the American Hospital Association — roughly 18 percent of US practicing physicians. This number does not count physicians in non-profit “foundations,” some of which are hospital-controlled, which may include an additional 6.5 percent of practicing physicians. In 2014, there was almost no further growth in hospital-employed physicians, suggesting that this cycle of expansion may have come to an end.

Some of this renewed hospital activism was driven by explicit strategy in preparing for new generation of integrated delivery models—accountable care organizations (ACOs). In other cases, it was an effort to grow fee-for-service (FFS) market share by acquiring and “realigning” physicians who admitted patients to competing facilities. Hospitals were subjected in particular to intense pressure to employ their local private practicing cardiologists; failure to employ them frequently meant losing their catheterization/imaging volumes and open heart surgery referrals to competing facilities. Private practice cardiology was upended by the 2005 federal Deficit Reduction Act, which markedly reduced Medicare’s office-based imaging payments on which cardiologists had increasingly depended for incremental income.

Lots of Other Ways Hospitals Supply Physicians Income

Direct employment does not exhaust the means by which hospitals provide physicians with income. Many physicians who are not employed by the hospital contract to provide hospital-based services on a 24/7 basis, e.g. to support hospitals’ emergency and intensive care services. Hospitals are required by the federal Emergency Medical Treatment and Active Labor Act of 1986 (EMTALA) to provide emergency services to patients regardless of their ability to pay, which in turn requires 24/7 physician coverage.

Traditionally, physicians voluntarily provided hospital call coverage for evenings and weekends, in exchange for the hospital granting them admitting privileges. Because an increasing percentage of physicians no longer hospitalized patients or even visited the hospital, this practice no longer met the demands of a 24/7 enterprise. And because community physicians no longer were willing to supervise in-hospital care for their patients, hospitals also needed to hire or contract with physicians to provide in-house medical supervision. Hospital medicine, or “hospitalist” services, has been the most rapidly growing medical specialty in the past 15 years, now numbering more than 44,000 practitioners — roughly equivalent to the number of general internists.

While hospitals often contract with local physician groups for coverage in “hospital-based” disciplines—notably emergency physicians, hospitalists, anesthesiologists, and neonatal and adult intensivists—an increasing percentage of coverage in these disciplines is being provided by billion dollar-plus national firms such as MedNax and Team Health. However, the highly paid general medical and surgical subspecialties that demand call pay tended to be locally controlled. The “pinch point” sub specialties are: general surgery, cardiology, orthopedics, and neurosurgery.

According to Sullivan Cotter’s 2014 survey, the average daily call pay stipend for orthopedics is around $1,200; the amount for neurosurgery and interventional cardiology exceeds $2,000. These subspecialties have simultaneously experienced a wave of consolidation into large single specialty medical groups. It is not unusual today for all the orthopedic surgeons or neurosurgeons in large communities to practice in a single group.

While the initial impulse for this consolidation may have been to spread fixed practice costs over a larger patient base and to avoid being excluded from health insurers’ specialty panels, a fortuitous byproduct of consolidation is to increase specialists’ bargaining leverage over the local hospital regarding call pay and directorships. These dominant single specialty groups have also shown an unsurprising disinclination to participate in their hospitals’ ACOs, given the fact that reducing specialty use is an inevitable future ACO agenda item. Hospitals have often cited unaffordable call pay demands by private practitioners as a reason why they chose to employ the relevant subspecialists directly, triggering explosive political reactions from what remains of their private staffs.

Hospitals’ Fastest Rising Expense is Physician Related

By our best estimates, perhaps 10 percent of a typical community hospital budget involves physician subsidies of various kinds. For smaller institutions in economically troubled rural areas, these percentages may be much higher because the hospital has no choice but to employ local physicians to keep them in the community.

Not all physicians focus mainly on income maximization. Many or most are simply too busy seeing patients to maximize the economic returns on their practices. Further, in our experience, medical communities differ markedly in their “mercantile” character. However, with physicians’ incomes squeezed by consumer financial problems, the growth in high deductible health plans, and the relentless press of prior authorization by health insurers, pressuring the local hospital for additional income has become the easiest way for physicians to improve their finances.

With hospitals’ unit price growth at historic lows, physician compensation in all its varieties has become the fastest growing expense in many hospital budgets, forcing managements to cut expenses in other areas. This increased cost of physician subsidies of various kinds comes without compensating revenue. A recent report by Moody’s on hospital financial trends found that while physician employment was an effective strategy for growing hospital and health system revenues, it comes at a price of damaging their operating margins and, by implication, their creditworthiness

Regulatory Constraints on the Hospital Physician Relationships

Two key elements of federal law constrain the ability of physicians and hospitals to exploit one another at patients’ and Medicare’s expense. Under Medicare’s “anti-kickback” statutes, it is illegal for hospitals to offer physicians financial inducements to admit patients to their institutions or to order the hospitals’ services. This prohibition covers such inducements as directorships and practice or medical office rent subsidies and loans. But these prohibitions also extend to the hospital’s paying physicians more than market value for their own services as employees.

In addition to these decades-old prohibitions, the so-called Stark Laws enacted in the early 1990’s (named after longtime Ways and Means health subcommittee chair Pete Stark [D-CA, retired]) prohibit physicians from having an ownership interest in services they order and use. This law was originally intended to curb the practice of physicians “self-referring” patients to imaging or surgical facilities in which they had an ownership interest. But as hospitals created hospital-physician joint ventures for surgical and imaging services during the 1990’s, the Stark laws also acted as a brake on economically abusive partnerships that rebated back to physicians some of the income they generated by using the joint-ventured services. The Stark laws required these transactions to be “commercially reasonable.”

However, the Stark laws are riddled with loopholes (euphemistically termed “safe harbors”), including exceptions for physicians that own an entire hospital (tightened in the ACA in 2010); for group practices which own their own imaging or surgical facilities; and for imaging services located in the same physical office location as a physician practice. Since the government does not provide guidance on fair market value or commercial reasonableness of these arrangements, management (or artful use) of these loopholes has become a vast and lucrative subspecialty in health care law.

Medicare also exerts tremendous influence over where physicians practice by how the program prices ambulatory services. In 2005, for example, in the budget reconciliation law known as the Deficit Reduction Act, Congress dramatically reduced what Medicare paid for freestanding or office-based imaging services compared to rates paid to hospitals for the same services. This triggered the aforementioned surge of cardiologists, whose incomes had become highly dependent on their own office-based imaging revenues from nuclear scans and CT, into hospital employment.

In addition due to Medicare Part B payment policy, hospitals can mark up their employed physicians’ services as “provider based” and can charge technical fees for their services. This in turn enables hospitals to offer some physicians salaries that significantly exceed what they can earn in private practice. These physicians refer patients to the higher-reimbursed hospital ancillaries, whose profits hospitals use to support physician compensation.

The recent (2015) MACRA legislation, which “fixed” the Medicare Part B Sustainable Growth Rate (SGR) problem, appeared to correct this payment anomaly, i.e., that physician services are worth more to Medicare in hospital employment than in private practice. In reality, however much protesting hospital representatives did during the negotiations, what MACRA actually did was grandfather in most of the existing payment differentials while reducing some payments for hospital ambulatory services provided more than 200 yards from the main hospital campus.

Policy Goals Do Not Align with Reality in the Hospital/Physician Nexus

The movement of formerly FFS-based private practitioners to hospital salaried employment is conceptually appealing to policymakers because it seemingly dilutes physicians’ incentive to render more services than patients need, and also offers the potential for better coordinated care. The reality, however, appears to be that salaried employment actually increases health costs.

Market pressures have pushed hospital compensation and practice expenses for some specialties far above what they actually generate in cash collections. Losses in excess of $200,000 per hospital employed physician are not unusual, according to the Medical Group Management Association. Hospitals make up the cash losses on the practices by placing their employed physicians on RVU-based incentive compensation formulae. (RVUs, or relative value units, are Medicare’s billing units for physician services.) These formulae encourage the employed physician to maximize consultative requests for other specialists, as well to use in-house imaging and laboratory services (which are highly profitable to the hospital). In effect, physician care has become a “loss leader” for the hospital’s profitable diagnostic and surgical services.

Whatever the theoretical benefits in improved efficiency or care coordination, a growing body of evidence suggests that hospital employment of physicians increases health costs. Some hospitals have amassed sufficient bargaining leverage with health plans to increase their physician groups’ rates, raising antitrust concerns about hospital monopolization of local physician markets. Physician employment apparently raises hospital prices and spending over institutions that do not employ physicians (testimony to the success of the RVU compensation models discussed above). Finally, physician employment also appears to raise per capita health spending relative to when hospitals and physicians are nominally separate.

Looking Ahead

Overall, despite the appearance of increasing hospital influence, the terms of trade between hospitals and physicians actually appear to be deteriorating for hospitals as baby boom physicians retire. Entrepreneurial baby boom physicians who did not wish to be employed by the hospital are being replaced by Millennial physicians heavily burdened by medical school debt, with a revealed preference for significantly shorter work weeks compared to their elders. Many younger physicians also appear disinclined to participate in the system development and medical policymaking required to participate in new payment models like the ACO and bundled payment.

Hospitals have a vital interest in the renewal of the primary care physician base in their communities, a particularly vital one if the community is struggling economically. Yet all over the US, hospitals have become midwives to an expensive intergenerational transition in medicine, perhaps permanently raising their expense base. Due to competitive pressures, hospitals are supplying an increasing percentage of physician income at a time their top line revenues are growing in the low single digits, if at all. This rise in physician expense challenges hospital managers and clinical leaders to improve clinician productivity as well as the quality of their work product.