As hospitals search for strategic clarity in the wake of health reform, they face major uncertainty in how to manage the looming generational transition in their medical communities. The hospital and physician practice sectors are the core of our health system; together they represent a little more than half of U.S. medical spending. These two sectors are inextricably intertwined — and getting more so.

Hospitals today employ about one-fifth of practicing physicians (not counting about 116,000 residents and interns) and contract with another fifth or more to support their 24/7 services such as the emergency department and intensive care units. Physician subsidies of various types — employment contracts, call pay, directorships, etc. — are the fastest growing and seemingly least well-controlled hospital expense.

There is significant geographic variation in physician employment by hospitals. Across the hard-pressed northern swath of the United States, from Maine to Montana, independent physician practice has pretty much collapsed, except for scattered, large multispecialty physician practices. For better or worse, the future of physician practice in this tier of the country lies firmly, if uneasily, in hospitals' hands.

However, in most sunbelt communities and many large urban areas, private practice is still very much alive, though hospital employment is significant. In these communities, the boundary between the hospital and medical practice is both blurry and volatile. Chronic problems include: granting of hospital privileges to perform certain procedures for competing specialties, covering the hospitals' 24/7 service obligations, managing care of the indigent, joint ownership of ambulatory services and contracting for physician services generally.

Shifting Boundaries

Conflicts between hospitals and physicians markedly worsened during the late 1980s and 1990s as ambulatory technologies like imaging, chemotherapy and less-invasive surgery enabled physicians to peel off these services and perform them in their own offices or freestanding facilities. Urgent care also posed a challenge, since hospitals' urgent care locations often competed with the private practices of nearby primary care physician offices.

In the last decade, the boundaries separating the hospital from physician practice have shifted yet again, posing a major strategic challenge in a time of slowing hospital topline growth. Today, the U.S. physician practice sector has two basic components: a tenacious small-scale practice framework controlled by physicians, and a growing trend of very large practices — not only hospital-sponsored megagroups, but also successful regional multispecialty physician groups like Kaiser, Mayo, Geisinger and Lahey, and corporations like MEDNAX, HealthCare Partners (now part of DaVita) and Sheridan Healthcare. In 2012, more than 110,000 physicians practiced in groups larger than 100, according to the American Medical Association's Physician Characteristics and Distribution in the U.S. (2014). By contrast, nearly 41 percent of physicians practiced in groups smaller than four.

In the next five years, I expect this small physician practice sector to shrink as the boomer physicians who form its core retire. However, the large practice sector is economically fragile, and the independent practices within it are often capital-starved. There is scant evidence of economies of scale in physician practice, and as practices grow, they become more dependent on so-called ancillary income.

Boomer docs — the most entrepreneurial generation of physicians in history — are gearing down or retiring. They are being replaced by frightened Generation Y docs looking for practice stability and a way to pay down their debts. In the wake of the 2008 recession, many private practices stopped hiring, leaving the hospital the employer of last resort. As a result, in the decade that overlapped the recession (2003–2013), the hospital's employment "footprint" in physician practice increased by somewhat more than half.

For many Gen Y docs, hospital employment is like an extension of the residency, only with 35-hour shifts instead of 80-hour ones, but with significantly higher per hour compensation. As many hospitals are discovering, this "extension of the residency" continues a cost-insensitive, consultant-happy style of medical practice profoundly at odds with the needs of a disinflationary, "accountable care" environment.

The Cost of Employment

Hospitals' physician employment surge has caused a serious case of financial indigestion. According to the Medical Group Management Association in 2013, the average hospital-employed physician generated over $206,000 in losses (the difference between total practice revenue and total costs of operating the practice and paying its providers). These burgeoning losses have occurred at the same time hospitals' top lines have stopped growing. As can be seen from the following figure, physician "investment" costs are devouring hospital earnings.



Green acres health system financials

Source: Nathan Kaufman, Kaufman Strategic Advisors, 2014. Used with permission.

Of course, it has been argued that if you acquire practices of physicians who don't use the hospital, the losses are offset by additional outpatient utilization and hospital admissions. This argument does not apply to new physicians, however, nor to physicians who were loyal users of the hospital, where the major income offset is marking up their fees to the higher Medicare "provider-based" rates. "Beggar thy neighbor" physician employment strategies (hiring away a competitor's physicians) appear to be an efficient way to destroy both hospitals' bottom lines.

It is my perception that the physician employment boom is reaching a tipping point. Most hospitals are now rationalizing their physician spending, renegotiating excessive income guarantees, strengthening productivity requirements and firing physicians who just wanted shelter from economic risk or who did not work effectively with colleagues.

In right-sizing their physician employment expenses, hospital executives face the following challenge: They cannot afford their primary care physician base not to renew itself. If primary care private practitioners are not doing well enough to take in a junior partner who can later take over the practice or to sell the practice to younger physicians, eventually the hospital will wither. Some form of hospital participation in this renewal seems inevitable, even in some prosperous communities.

However, here is the ultimate question: Will the hospital need to be the permanent final employer of all its primary care physicians (let alone its highly paid procedure-oriented physicians)? Many hospitals or systems are trying to create management services organizations to strengthen their physicians' practices as an alternative to employment, and are encouraging consolidation of the healthy private practices as older physicians contemplate retirement.

Some Successes

It is unlikely that we will return to the time when most primary care physicians admitted their own patients to the hospital and managed their hospital care subsequently. Some element of the recent expansion of the hospital into the physician world is likely permanent; hospitalists and intensivists (pediatric and adult) are almost certainly a part of most hospitals' futures.

For these emerging specialties, as well as for the more traditional hospital-based disciplines of emergency medicine, radiology, anesthesiology and pathology, there is a new question: Do I contract for hospital-based physician coverage, or employ them directly?

Many hospital leaders believe they can manage these practices better than local or national groups, e.g., that they can recruit, equip, motivate, renew and systematize the practices of these physician specialties. But is the hospital truly able to manage the logistics and politics of direct employment well enough to realize savings? The fact that the large national and regional groups that dominate these specialties continue to grow suggests the answer to this question may be "no."

It will not be financially possible for hospitals to absorb all of Gen Y physicians' economic risk. Hospital leaders will need pluralism, flexibility, transparency and fairness to manage this tricky generational transition.

Several regional health systems seem to be walking the tightrope effectively. One is Memorial Hermann in metropolitan Houston. It has created a 3,500-physician clinically integrated network that offers Houston physicians a variety of ways to participate in an aligned clinical enterprise, only one of which is direct employment. It is possible to drive toward clinical protocols, a single electronic record and single signature risk contracting without employing all the physicians.

Another hospital that seems to be striking the right balance is Rady Children's Hospital in San Diego. It has created Rady Children's Physician Management Services, which supports both a large primary care medical group and a hospital-sponsored foundation that employs the hospitals' subspecialists without creating an economic burden on the hospital. Both Texas and California have strong "corporate practice" statutes that limit hospital employment options.

In the last 10 years, hospitals have become unwitting midwives to an extremely expensive generational transition in medicine. The illusion of control that physician employment represents may be just that — an expensive experiment in the limits of management capacity. Managing employed physicians is not a core competency for most hospital leaders. If hospitals can add value to physician practices by marking up their rates to Medicare and commercial payers, rather than improving the practices — lowering their overhead, decreasing the "hassle factor" of medical practice, improving professional satisfaction and, most importantly, improving service to patients — the present trend toward physician employment will reverse yet again.

Jeff Goldsmith Ph.D., is the president of Health Futures Inc., and associate professor of public health sciences at the University of Virginia, Charlottesville.