On the brink

Tags: news, texas family physician, baylor college of medicine, kelsey-seybold, budget, legislature, family medicine residency program, elliott

The Baylor College of Medicine Kelsey-Seybold Family Medicine Residency Program is shutting its doors this year because it can’t secure stable funding from year to year. At a time when Texas desperately needs to produce more primary care physicians, many family medicine residency programs are

on the brink

By Jonathan Nelson

“After 10 years of preparing physicians-in-training for the practice of family medicine, we regretfully announce the closure of the Baylor College of Medicine Kelsey-Seybold Clinic Family Medicine Residency Program effective July 1, 2010. The residency program is unable to obtain the ongoing financial support necessary for the program to continue.”

So reads the first paragraph of an announcement released in the fall of 2009 by the residency program housed at Kelsey-Seybold Clinic in Houston. The irony of the program’s fate is that by all accounts, it could have served as the model for training the family physician of the future.

“Kelsey-Seybold Clinic is a great teaching environment,” says Patrick Carter, M.D., chair of the clinic’s department of family medicine. With more than 350 physicians, the multi-specialty ambulatory care center offers residents access to a broad range of resources including full laboratory and radiology support, specialty and subspecialty rotations all under one roof, plus rotations at several Houston hospitals. “Our family medicine center is right there on the main campus; the residents go right upstairs to do their specialty rotations. They’re the only residents in the whole place, so they get great teaching from all the specialists.”

The award-winning faculty design their curriculum to focus training on a collaborative model of high-quality, well-coordinated care that epitomizes the medical home concept. The residency even offers a dual-track M.B.A. program, one of only a handful of such programs in the nation.

“This is the model of care for the future that we are training our residents to do,” says Tricia Elliott, M.D., F.A.A.F.P., the residency’s program director.

In an era when family medicine residencies only manage to fill 45 percent of available residency positions with U.S. medical school graduates, 97 percent of the recruitment classes at the Kelsey-Seybold program over the last three years graduated from U.S. medical schools. More than 600 physicians applied for the four open positions at the residency in 2009, and of the four chosen, two are from out of state. “I’m constantly bombarded with people that would just love to come to our program,” Elliott says.

The quality of the program was never in question; the Accredita­tion Council for Graduate Medical Education recently granted the program a five-year accreditation, the longest duration possible. The funding, however, was always another matter.

“Our program was unique in many ways,” Carter says. “I was the first director of the program; it was sort of my baby, and it just breaks my heart to see it wind up like this.”

The first nail in the coffin came in 2006, when the program’s primary teaching hospital, St. Luke’s Episcopal Hospital, slashed support for the program in half. The move sent Baylor and the Kelsey-Seybold FMRP scrambling to find new sources of funding, none of which were stable from year to year.

By the fall of 2009, the parties agreed that the program was no longer financially sustainable. Carter says Kelsey-Seybold Clinic needed a subsidy from Baylor College of Medicine of between $400,000 and $450,000 to keep the program viable. But BCM, which has operated at a substantial deficit for the past several years, couldn’t save the program.

“We just were not in a position to write them a check to do this,” says Steve Spann, M.D., senior vice president and dean of clinical affairs at Baylor College of Medicine. “We felt that we could train those residents for much less cost in our own system.”


Residency training is expensive, and financial struggles are especially common to family medicine residency programs. The director of AAFP’s medical education division, Perry Pugno, M.D., says the situation at the Kelsey-Seybold FMRP is indicative of the challenges facing family medicine residency programs across the country. “In the last five years, we’ve lost over 30 programs for this exact reason.”

From 2000 to 2009, 67 family medicine residency programs closed and 19 opened for a net loss of 48 programs in the country. During the 2009-2010 academic year, there were 561 fewer family medicine residency positions filled than in the 2000-2001 academic year. Texas has experienced a similar decline. In 2000, 247 first-year family medicine residency positions were offered in Texas; there were only 201 by 2009, which is almost a 19-percent reduction in the state’s capacity to produce family physicians.

Texas already faces a shortage of primary care physicians to care for its growing and aging population. The national average for primary care physicians for every 100,000 people is 81, a number considered to be too few by AAFP and other health care policy institutions. Texas has 68 primary care physicians for every 100,000 people. Forty-five percent of physicians in Texas are 50 years of age or older.

“The significant problem we face is that at a time of health reform that says we need more family physicians, the present environment is so adverse to our training programs that they’re actually closing,” Pugno says.

After a rash of program closures between 2000 and 2002, he and some colleagues published a study in the journal Family Medicine that showed the primary reasons for the closures were financial and political, and that the majority of closures were community-based programs and those operating in public or non-profit hospitals.

Since then, he says, nothing has changed. Pugno calls the closure of the Kelsey-Seybold FMRP a “poster child” of what’s wrong with graduate medical education funding. The problem is not necessarily that there’s too little money, but that the funding streams are so varied, complex, and fraught with unintended incentives and disincentives.

“So few people understand how GME funding works,” Pugno says. “It’s so convoluted that most people just throw their hands up and don’t bother.”

Most residency programs cobble together funding from various sources to cover expenses, the largest of which are salaries for faculty and the residents in training. The bulk of GME funding comes from the Centers for Medicare and Medicaid Services through the Medicare program in the form of direct GME payments (DGME)—intended to reimburse teaching hospitals for the direct cost of training residents including resident stipends and faculty salaries—and indirect medical education payments (IME)—intended to pay for the higher cost of patient care at teaching hospitals. Funding amounts for DGME and IME are calculated through two different methodologies, both of which are based on the number of residents in training at a hospital.

Two fundamental problems with Medicare GME funding combine to put family medicine residency training in jeopardy. First, Congress capped the number of residency positions paid for by Medicare in the Balanced Budget Act of 1997. With few exceptions, a teaching hospital can only receive Medicare GME funding for the number of residents it trained in 1996, and while many teaching hospitals exceed that cap, they do so at their own expense.

The other problem is that Medicare will only reimburse teaching hospitals for the time residents spend in the hospital, which is fine for most specialties, but detrimental to primary care. For family medicine residents, the most important classroom is the outpatient clinic, and unless that clinic is attached to the teaching hospital, residency administrators have to fight tooth and nail to secure resident stipends.

When St. Luke’s Episcopal Hospital reduced the stipends they paid to the Kelsey-Seybold FMRP, the program had no recourse to secure Medicare GME funds other than to find additional teaching hospitals in the area that would agree to host a few residency rotations. According to Spann, St. Luke’s cut the stipends without consulting Kelsey-Seybold Clinic or Baylor. “They did that unilaterally and despite some pretty strong protest from us, but they felt it was more to their benefit to put those stipends into neurosurgery.”

St. Luke’s did not agree to an interview for this article despite multiple requests.

Since Medicare GME funds go to teaching hospitals instead of residency programs, hospitals determine which residencies get funded and which don’t, and building a robust primary care physician workforce to care for patients in ambulatory clinics is not a priority for the hospitals.

The Christus St. Elizabeth Family Practice Residency Program in Beaumont was forced to shut down in 2002 after its teaching hospital withdrew support for the program, citing financial reasons. Researchers later found that of the 74 graduates from the program practicing medicine in 2005, 88 percent practiced in health professional shortage areas.

In 2008, the Texas Tech University Rural Program in Abilene closed for the same reason. Mike Ragain, M.D., chair of the Department of Family and Community Medicine at Texas Tech University Health Science Center and the program’s first director, says the purpose of the family medicine residency was to place family physicians in rural communities around Abilene, but their training hospital didn’t share that goal.

“Down in Abilene, we saw a bunch of the indigent population … and tried to provide excellent primary care and keep those patients from having to use the ER as their source of care,” he says, adding that he and the program administration presented financial analysis of these savings to the hospital to prove the program’s worth. “They didn’t buy it.”

In a May 5, 2009 letter to Kathleen Sebelius, secretary of the U.S. Department of Health and Human Services, the Council on Graduate Medical Education wrote:

“Congress, as part of health care reform, should modernize GME funding under Medicare and Medicaid to align financial and educational incentives to produce more primary care physicians capable of practicing in patient-centered medical homes in order to serve the growing need of Americans. This would help to satisfy a growing need for first-line and coordinated health and would begin to remedy the changes of the last 10 years where nearly all GME expansion in teaching hospitals has been in subspecialty medicine, often to the detriment of primary care.”

When COGME Vice Chair Robert Phillips, M.D., M.P.H., considers how the decline in the number of primary care residency positions offered nationwide correlates with specialty income, he draws a troubling conclusion: “Academic hospitals are expanding GME to staff their more lucrative services, not to produce the workforce America needs.”

Ragain, Carter, and Elliott all say that family medicine residency training would be much more stable if federal GME funds went directly to residency programs. AAFP has been pushing for this change for years with some recent success. The health reform bill currently under consideration in Congress contains provisions for teaching health centers that AAFP says would modernize training for primary care residents in non-hospital settings. In recent letters to congressional leadership, the Academy continues to recommend that CMS be given authority to establish pilot projects to test the use of Medicare GME funds for direct support of primary care residencies, a proposal that has been entertained throughout the reform deliberations.


Unlike many states, Texas provides some funding for graduate medical education, and family medicine residencies depend heavily on those funds. In fact, the only governmental funding that goes directly to family medicine residencies in Texas comes from the Texas Higher Education Coordinating Board. In 1977, TAFP played a seminal role in the establishment of a state program under the purview of the board to invest in family medicine training, the Family Practice Residency Program, which has since supported well over 7,000 family medicine residents. That program is one of several under the THECB dedicated to graduate medical education, and like the rest, it suffered a blow following the actions of the 77th Texas Legislature.

The state faced a projected $10-billion shortfall as the Legislature convened in 2003, and medical education was to be only one of many health-care-related sectors that would feel the axe. In total, state funding for GME allocated by the THECB was cut by 46 percent. A $4-million-per-year program to compensate teaching hospitals for their educational costs was eliminated, along with a line item for pilot projects that enabled innovation in the care of indigent patients at family medicine residencies.

Legislators also cut all GME funding through the Medicaid program, a $48-million state expenditure for the 2002-2003 biennium that drew down an estimated $72 million in federal matching funds. According to Roland Goertz, M.D., M.B.A., AAFP president-elect and chair of the Family Practice Residency Advisory Committee of the THECB, Medicaid GME funding was more important for some family medicine residencies than others. Still it represented a monumental drawback of state support for GME. Before the 77th Legislature, Texas spent $181.3 million on GME including those federal matching dollars. In the next biennium, Texas spent $35.9 million on GME.

As the CEO of the three foundations behind the Waco Family Practice Center, which houses the McLennan County Family Medicine Residency Program, Goertz was keenly aware of the reduction in state support. The cuts left a $150,000 hole in the program’s budget for the biennium. “We had to go back and look at almost every cost center the residency had to see if there was some way to make them operate with less funding.” They found few.

Four years earlier, the Waco program had fulfilled the requirements to become a federally qualified health center, which allowed it to fund some of the direct clinical costs. If not for that, Goertz says the program would not have been able to balance the budget. “That’s a rare situation,” he says of their timely maneuver. “What literally happened was that the state made a decision and then the programs had to either appeal to their partners or their sponsors, and say ‘please bail us out of this, or we’ll have to reduce numbers of residents, numbers of faculty, or ability to see patients.’” All of those things happened at many of the residency programs in Texas, he says.

Since then, the state has increased its investment in graduate medical education, adding a budget line item to the appropriations for each medical school entitled, “GME formula funding,” beginning in the 2006-2007 biennium. The amount each school gets is based on the number of residents in the programs affiliated with the university, and it has increased significantly since its inception, from $25 million to $79 million for the 2010-2011 biennium. This funding, however, goes to medical schools and not residency programs, so as with Medicare GME funding, family medicine residencies must lobby their institutions in competition with all other residencies to receive that support. Baylor College of Medicine’s appropriation for GME formula funding for 2010-2011 is $15.3 million, up from $12.8 million in the previous biennium, and yet the Kelsey-Seybold Family Medicine Residency Program is still closing.

Including the loss of Medicaid GME and its federal match, and the establishment of GME formula funding to medical schools, the state will spend $65.6 million less on GME in 2010-2011 than it did 2002-2003. Even after a $3-million raise from the last Legislature, the THECB funding for GME is only 52 percent of what it was before the 77th Legislature.

All the while, residency costs have been climbing. Texas Tech’s Ragain and colleagues are in the process of publishing a study in which they examined financial data from 2005 to 2008 reported to the THECB by the state’s 26 family medicine residencies. Preliminary data shows that while revenue for the programs increased by 4.1 percent on average, total expenses jumped by 6.7 percent, leading to an overall increase in cost per resident of 13.8 percent.

Relief from the state may be a long shot at best. No official predictions have been cast, but word around the Capitol is that as the 82nd Legislature convenes in January 2011, Texas could face between a $12-billion and a $17-billion shortfall. The method the state selected for funding public education has left it with what some economists call a structural deficit, meaning unless unanticipated funds—like federal stimulus dollars—are available, the budget process starts off in a significant hole.

Dick Lavine, senior fiscal analyst for the Center for Public Policy Priorities, says the last Legislature used a lot of one-time money to fund the current budget. “Most of those expenditures are ongoing costs, so [the state is] at least $12 billion short in revenue sources that aren’t going to appear again, unless the federal government has a second stimulus package.”

Sales tax accounts for more than half of all state revenue, but in the first four months of the current fiscal year, state sales tax revenue is down 12.9 percent from last year. “Who knows what the national economy or the state economy is going to do,” Lavine says, “but it means that right now, this is the worst performance by the sales tax since it was started in 1961.” Both the governor and lieutenant governor have already hinted at budget restrictions for the current biennium.

Texas has socked away money in a rainy day fund that was estimated to be $9 billion, but recent estimates have been somewhat lower. Lavine says that’s because it’s mostly funded by natural gas taxes, and prices for natural gas are falling. New discoveries like the Barnett Shale have increased supply while the recession has caused demand to drop. “So we can’t be sure about the rainy day fund, and in any case, it’s unlikely that state leadership would want to spend it all.”

The coming biennium is beginning to look a lot like what the Legislature faced in 2003; comparisons are in order. “That’s what we’re going to be replaying, or worse,” Lavine says.

Goertz worries that if in the face of a severe pending budget deficit, state leaders slash funding for GME as they did in 2003, the state’s infrastructure for producing increasingly valuable primary care physicians will be seriously damaged. “There is a point where there is no more flexibility left and I am very convinced that’s not far off,” he says. “The organizations won’t be able to cover the shortfalls any longer, no matter how much they want to. Kelsey-Seybold is an example of that. Beaumont before it, Abilene since then. If the reports to the coordinating board are any example, a large number of the programs are right at that cliff, so to speak.”