The four officials lobbed ideas back and forth about how Moab might keep its financially beleaguered hospital, which was not only a source of acute health care but also a critical cog in the local economy, and at the same time accommodate the elderly who needed long-term care. Eventually, someone asked: Why not provide nursing home care in empty hospital beds?

At that time, Walter was Utah’s director of Medicare services. When he got back to Salt Lake City, he wrote to the regional Medicare office in Denver, laying out the problem. Rural communities faced a shortage of nursing home beds in Medicare-certified skilled nursing facilities, he noted. That reflected the “industry’s unwillingness or inability to meet the Medicare certification standards,” as well as a paucity of the private pay patients needed to support a skilled nursing facility, Walter said.¹ The shortage of skilled nursing beds—three per thousand people over sixty-five in South Dakota, Texas, and Iowa in the mid-1970s, compared to fifteen per thousand nationally—reflected business pragmatism. In other words, there were too few patients in rural areas to justify expansion. As a result, those facilities that did exist typically had a high occupancy rate and a long waiting list, leaving rural residents, especially elderly ones, with little access to long-term care.

Yet in many of the communities suffering a shortage of nursing home beds, the local hospitals had an oversupply of (or at least were underusing) acute care hospital beds. This reflected a declining population, a physician shortage, and the exodus of patients through referrals to an urban hospital. It might have seemed obvious to provide skilled nursing care in those empty beds, but federal regulations did not allow that. If a hospital wanted to provide intermediate-level care or skilled nursing care, Medicare and Medicaid regulations required doing so in a facility that was physically distinct from acute-care beds. The facility for long-term care could be used for no other purpose and had to provide social services, physical therapy, and social and recreational activities that would help patients remain ambulatory, or become ambulatory and be able to carry out activities of daily life.

Meeting this standard would have meant significant additional cost. Yet the Medicare reimbursement for the skilled nursing care long-term care patients would require was less than that for acute care. Many rural hospitals, especially smaller ones, struggled to meet these requirements; the paperwork threatened to swamp their accounting capability and skilled staff was often insufficient. The “physically distinct” requirement for long-term care facilities thus ended a tradition in community hospitals of caring for both acute and long-term patients. Unable to offer a physically distinct facility for such care, a rural hospital would be unable to offer it at all—just at a time when America’s elderly population was increasing and demands for such care were rising.

So two trends—excess acute bed capacity and shortage of extended-care beds in rural areas—crossed, and out of that juncture emerged Bruce Walter’s idea. Why not address both, he suggested, by letting rural hospitals fill their empty acute-care beds with long-term care patients? The regional Medicare office agreed, and encouraged Medicare headquarters back East to consider some kind of experimental program. After negotiations that were “more than a little tense,” and multiple presentations by Walter to U.S. Senate committees and meetings with Medicare officials, he finally got the answer he was looking for. “We’ll see what we can do,” Tom Tierney, Medicare’s first director, told Walter after an April 1972 meeting.²

What Tierney could do, it turned out, was fund a three-year experiment named the Utah Cost Improvement Project, starting in January 1973. It involved twenty-five small rural hospitals in Utah. These hospitals became eligible for Medicare and Medicaid reimbursement for long-term care that they provided in acute-care beds—and the first swing-bed program was under way.

It is worth pausing here to clarify some terminology. From its inception, the swing-bed program has been described as providing “long-term care.” There are three types of such care: skilled nursing, intermediate nursing, and custodial. Custodial care includes administering single medications and assisting with such activities of daily living as bathing and getting dressed. Medicare does not cover this. Nor does it cover intermediate-level care—what Marjorie Eddinger of the Centers for Medicare & Medicaid Services (formerly the Health Care Financing Administration) in the Department of Health and Human Services calls “caring for the walking wounded, like changing dressings.” What Medicare does cover is “skilled nursing care.” This includes injections, intravenous feedings, placing catheters, rehabilitation (speech, physical, and occupational), and the like. It is what participants in the swing-bed program meant by long-term care.

A swing bed does not swing physically. Rather, it swings in the way hospital accountants and medical staff treat the patient occupying it. In a swing-bed program, a patient being treated for an acute condition could remain in the hospital for follow-up long-term care rather than be discharged to a nursing home. She or he would usually stay in the same bed, but the kind of care would be different, and the accountants would bill for it differently. Medicare’s reimbursement for services such as x-rays and lab work was based on the hospital’s cost for these services. It reimbursed hospitals for skilled nursing care at the Medicaid per-diem rate, covering room, board, and nursing. A hospital received less money from the government for long-term care than for acute-level care, but it was income that would not be received if the patient were discharged to a nursing facility. Moreover, the hospital didn’t have to meet the Medicare standards for rehabilitation services, social services, and space for patient activities.

These two provisions—waiver of Medicare standards for long-term care and a novel reimbursement scheme—were crucial to the swing-bed experiment. All of the participating hospitals must have had fewer than fifty beds and less than 60 percent occupancy in the three previous years.

Yet even in such underused hospitals, the idea of the swing bed was not accepted without reservation. “The administrators were leery from the beginning,” says Peter Shaughnessy, of the University of Colorado Health Sciences Center, who led the team that evaluated the swing-bed program. They were suspicious of federal intervention, and most knew next to nothing about long-term care. “At the start, the reaction of the nursing staff was often, ‘This kind of care is beneath me; you don’t need my special skills,’” Shaughnessy says. The nursing home industry was none too happy, either, figuring that swing beds would steal their patients. Many health planners doubted the wisdom of trying to prop up rural hospitals: if the facilities were operating at such low capacity, the reasoning went, they should be closed, not bailed out with this or any other program. But the greatest concern of all was that the care in swing beds might be substandard; as the hospitals themselves admitted, what did they know about long-term care?

That’s what Shaughnessy and his team at the University of Colorado set out to answer. In an evaluation sponsored by the federal government, the Colorado team found, first of all, that long-term care patients loved swing beds: being in a hospital made them feel better cared for, and since they weren’t physically moved to a nursing home, the program was much less disruptive than a move to a nursing home would have been. Families loved keeping grandma in her home town and not having to travel to a distant nursing home to see her. Physicians liked swing beds because they could visit their patients more frequently than if they were sent to a nursing home. Administrators and nurses came to like swing beds for satisfying the community’s long-term care needs. “That’s what I’ve always liked about the staff at these small rural hospitals,” Shaughnessy says. “They were responsive to patients’ needs, and they were also very aware of the important place the hospital had in the community, both as an employer and as a symbol. Even though neither the acute-care staff nor the nurses’ aides were trained in long-term care, they quickly realized that these patients had different needs, and learned to meet those needs. What turned them around was seeing that their patients really benefited from swing-bed care.”

In 1976, the federal government expanded the trial program in hospital swing beds to 39 hospitals in Texas and 22 in South Dakota and western Iowa, and in 1977 to 22 in central Iowa, until the program encompassed 108 rural hospitals in those four states. As in the original Utah experiment, small rural hospitals (with fewer than fifty beds) would use their beds interchangeably, swinging them between acute care and nursing-home-type care, depending on demand.

Who were the early swing-bed patients? Most were and are female, typically a white widow, seventy-five or older, covered by Medicare. This typical patient was living at home, either alone or with family. A fracture (in 20 percent of cases) or stroke (12 percent) sent her to acute care. In long-term care, she sat in a chair for meals (53 percent of patients), received physical therapy (56 percent), participated in patient activities twice a week or more (41 percent), and received social services weekly (49 percent). After discharge, she returned home to live alone or with relatives, though one in four went to a nursing home. One in ten died while in the swing bed; one in eleven returned to an acute-care bed.³ They were patients like a seventy-eight-year-old woman at Sierra Vista Hospital in Truth or Consequences, New Mexico, who had chronic obstructive pulmonary disease. She was moved to long-term care for two weeks and then back to acute care (never physically changing beds, of course) when she contracted pneumonia. She went back to long-term care for a month and then back to acute care when she needed increased oxygen therapy.

Once again, the evaluators found that hospital administrators, physicians, nurses, patients, and patients’ families were enthusiastic about swing beds.⁴ Even nursing home administrators in the community or nearby signed on, with the majority concluding that the swing-bed program should be not only continued but also expanded.

The biggest trouble spot was in quality of care. Some hospital staff members had difficulty switching gears to care for a long-term patient. Partly as a result, “the quality of long-term care provided in swing-bed hospitals was lower than . . . in comparison nursing homes.” In particular, the hospitals rated worse not only on such measures as addressing psychosocial problems, which fall outside their usual bailiwick, but even in such chronic-care needs as incontinence, skin condition, depression, loneliness, and isolation. Nursing homes were also better at “social-recreation, therapeutic-mental health, physical and occupational therapies, professional nursing,” the evaluators found, as well as in caring for patients requiring basic maintenance and support of their “functional, cognitive, emotional, and social needs.” Services were uneven: 53 percent of the swing-bed hospitals offered physical therapy and 36 percent offered social services, while speech therapy, occupational therapy, and patient activities were offered less frequently.

The extent to which the hospitals embraced swing beds also varied wildly, from 8 days of long-term care in one hospital to 3,667 in another. Not surprisingly, perhaps, the evaluators found that “hospitals with lower occupancy rates tended to provide more long-term care than those with higher occupancy rates.” The occupancy rates of nursing homes in rural areas hardly budged, however, indicating that swing beds were not substituting for rural nursing home beds but adding to them.

Swing beds were cost-effective, though they were not the revenue gusher that some administrators initially hoped for. “Swing beds were not big profit makers, but they did provide revenue and enable hospitals to keep open and retain staff,” Shaughnessy said. Compared to nursing homes, hospital swing beds were better able to provide the near-acute care (rather than chronic care) that most of their patients required. (Someone recovering from hip surgery or requiring an IV would be a near-acute patient.) Physicians visited swing-bed patients more frequently than they did nursing home patients. “Long-term care for the types of patients typically treated in hospital swing beds was adequate and possibly even above average, but . . . chronic care for patients with maintenance and even palliative care needs potentially required improvement,” the Colorado evaluators found.⁵ The experiment “showed that a hospital’s ability to ‘swing’ beds between acute and long-term care services could satisfy the need in rural communities for both services while maintaining sufficient reimbursement for the hospital,” the American Hospital Association concluded in 1982, referring to the 108-hospital program.⁶

The evaluators recommended that the swing-bed program go national. Many rural communities faced an unmet need for long-term care, they argued; swing beds could be more cost-effective than alternatives such as building more nursing homes; patients and their families benefited from staying near home; and swing beds could shore up rural hospitals financially. A swing-bed program “would be of benefit to rural communities in terms of meeting both long-term care and acute care needs,” the evaluation team wrote in 1980. Moreover, swing beds were “a cost-effective means of providing long-term care.” The quality of long-term care they offered was considered adequate, and although nursing homes scored better “the discrepancy was not substantial and . . . is likely to disappear over time as the staffs of swing-bed hospitals become familiar with the special problems of the long-term care patient.”⁷

The eight-year experiment was deemed enough of a success that in 1980, when Congress passed the Omnibus Budget Reconciliation Act, it included a provision allowing Medicare and Medicaid to pay for swing-bed care in rural hospitals that had fewer than fifty beds. In all, more than twenty-two hundred hospitals were eligible. The Health Care Financing Administration issued implementing regulations in July 1982: it imposed straightforward quality standards (requiring the specialized rehabilitation, patient activities, discharge planning, and other services of skilled nursing facilities) and reimbursement policies (as with the pilot program, routine room, board, and nursing were reimbursed according to the state average Medicaid day rate for skilled nursing facilities, and reimbursement for ancillary services were based on cost).

Coping with Medicare regulations was often difficult for small rural hospitals, and probably for others as well. To make things even more confusing, in March 1983 Medicare began reimbursing hospitals under a “prospective payment system.” Rather than reimbursing hospitals on the basis of the number of days they cared for patients and what they did for them, Medicare now reimbursed hospitals according to a patient’s diagnostic-related group, or DRG. A hospital that kept, say, a hip-replacement patient for ten days would get the same reimbursement as one that kept him for nine. This was an incentive to discharge patients as soon as possible—and, according to critics, in some cases sooner than was reasonable. Such patients seemed more likely to receive the postacute care they needed in a swing bed, which had a stronger medical orientation, than in a nursing home bed.

The change in reimbursement also increased the controversy swirling around swing beds. The nursing home industry charged that hospitals were using swing beds to do an end run around the prospective payment system. (Medicare will pay for only this number of days for my hip fracture patient? No problem, from that day on I’ll have Accounting say she’s in a swing bed.) There were even concerns that swing beds might allow double dipping, with hospitals being reimbursed twice for services that the prospective payment system was intended to cover. According to this argument, the services that swing patients received—caring for postsurgical wounds, for instance—were included in the hospital’s DRG.

Although the new Medicare prospective payment system slowed adoption of swing-bed programs as small rural hospitals sorted out the incentives and the disincentives, by 1984 it was clear that the prospective payment system offered incentives for rural hospitals to provide swing-bed care.⁸ The number of participating hospitals grew steadily: 149 at the end of 1983 to 771 (out of an estimated 1,600 eligible rural hospitals in thirty-nine states) at the end of 1985 and 1,207 by August 1989.

Starting a swing-bed program requires meeting the care and administrative requirements that Congress and the Health Care Financing Agency imposed. The complexity of the reimbursement procedures created management problems for many small rural hospitals. Moreover, directors and staff members of many of those hospitals resisted the idea that an acute-care hospital should become a nursing home. As a result of these and other barriers, by the end of the four-state federal demonstration a third of the eligible hospitals had elected not to participate. Although Congress had changed federal reimbursement mechanisms, it had done little or nothing to address such issues.

The Robert Wood Johnson Foundation already had a vibrant program in rural health care, recalls senior program officer Nancy Barrand: “This seemed like the next logical step in our efforts to help small rural hospitals and the communities they served.” So in April 1981, the Foundation launched a national demonstration program called the Rural Hospital Program of Extended-Care Services. The $6.5 million program provided funding for five state hospital associations and twenty-six small rural hospitals in those states to convert acute-care beds to swing beds. The goals, explained Thomas Gregg, a Robert Wood Johnson Foundation program officer, and Robert Blendon, a Foundation vice president, were to improve patient access to long-term care and help financially precarious rural hospitals, and do it all in a cost-effective way. To accomplish all this, the Foundation would promote swing beds by establishing models of how a small rural hospital with excess acute care beds in an area that also had a nursing home shortage could provide high-quality long-term care.⁹

How? By raising awareness of the program, showing hospitals how to adopt it successfully, developing technical expertise in the hospitals and state associations, and having participants share what they learned.¹⁰ Hospitals could use the grants from The Robert Wood Johnson Foundation for training, to pay for replacements of those being trained, to pay for those providing specialized services like physical therapy, and to obtain equipment. They could organize training sessions on billing Medicare and Medicaid for swing-bed stays, since negotiating the shoals of Medicare reimbursement was no easy matter. Grantees could also offer educational programs in gerontology, to help nurses deal with what, to many, was the more frustrating and less interesting work (compared to acute care) of caring for the chronically ill. Training was crucial, since, as the Foundation officers gently noted, “hospital staff may be reluctant to change their methods of operation.”

One of the program’s chief elements was to help grantee hospitals narrow the gap between the services they offered in swing beds and those that nursing homes offered in long-term care beds. Hospitals offered swing-bed patients fewer social and recreational services, therapeutic and mental health services, and physical and occupational therapy services than did high-quality nursing homes. There was genuine concern among Foundation officers that the hospital-based long-term care was inadequate in these areas and had to be improved. Finally, but not least, the Foundation’s program had the ambitious intent of overcoming reluctance to adopt the swing-bed approach by enhancing the medical, administrative, and financial capacity to deliver care.

“What the Foundation did was immensely clever,” evaluator Peter Shaughnessy said. “They took a program we knew something about, and as national implementation was going forward they said, ‘Let’s be sure it’s done well.’ Without these grants, swing beds would not have brought the widespread benefits for so many patients that they did. The education and training that resulted from the grants was exactly what was needed. They provided a template.”

To direct the program, the Foundation named Tony Kovner of New York University’s Robert F. Wagner Graduate School of Public Service and a former rural hospital administrator.

As the swing-bed program was launched, hopes were high. In a memo recommending the grant program, Gregg and Blendon noted that if every rural hospital participated in the swing-bed program, “between three-quarters of a million and two million days of long-term care would be provided in rural hospital ‘swing-beds,’” a 10 percent increase in the availability of such care in rural areas, and the program would offer rural residents “an instant alternative to some of the smaller and less adequate proprietary nursing homes.”¹¹

The Robert Wood Johnson Foundation program got off to a rocky start, however. Despite intensive training, many hospitals found that nurses resented the extra workload and, inappropriately and expensively, provided the same high level of care to swing-bed patients that they did to acute-care patients. Because of inexperience and lacking local resources such as physical therapists, many of the grantee hospitals were unable to offer enough high-quality patient activity programs. In one New Mexico hospital, two patients who were admitted to long-term care in swing beds stayed only a few days, since the nursing staff refused to take care of them and the director of nursing did not support the program. Even where hospitals did a good job, as in Kansas, there were glitches: hospitals were denied reimbursement because of confusion about the definition of skilled-level care.

Even with these startup problems, the number of patient days in the swing beds rose steadily, from 1,526 in 1983 to 8,522 in 1984 and 10,221 in 1985. Over the same period, patient-days in acute-care beds fell 20 percent, suggesting that swing beds had a stabilizing effect on the hospitals.¹²

Any assessment of the swing-bed program needs to be clear about whether it is addressing the program’s value to the patient or its value to the hospital. On the first count, both those running the program and those evaluating it found swing beds a success. Tony Kovner, the program director, and Hila Richardson, the deputy director, found that swing beds increased access to long-term care for rural residents, especially those with intense medical needs. More residents stayed in their community for such care than did those where swing beds were unavailable. Compared with nursing homes, swing beds excelled in several areas. They made possible a smooth continuum of care from initial hospital admission through discharge to home or nursing home. Swing-bed patients had easier access to such medical services as respiratory therapy and lab tests.¹³

Perhaps their care improved simply because they were allowed to recover near home, where family and friends could be involved in their care.¹⁴ Peter Shaughnessy was particularly impressed that “critical ingredients of long-term care that were often missing in other settings were beginning to occur in swing-bed settings.”¹⁵ In particular, physicians became more involved in long-term care. Certain kinds of rehabilitation as well as skilled nursing care were more available than in a nursing home. Care for patients segueing from acute care to long-term care was better coordinated. These differences from traditional long-term care accounted for the swing-bed program’s success.

Although the quality of care got mostly good marks, there were some stumbling blocks. The first was that caring for a patient in a swing bed is significantly different from caring for a patient in a traditional hospital bed. As a result, the medical staff had to change its focus from diagnosis-centered, doctor-dominated, acute care to the nursing-centered, multidimensional needs of the long-term-care patient. There was no question that the old mold of hospital care needed to be broken when it came to the swing-bed patient. At Cedar County Memorial Hospital in El Dorado Springs, Missouri, staff members learned not to do everything for the patient, as they did for those in acute care, but to assist (or simply stand by) while the swing-bed patients fed and dressed themselves. The staff involved the family in a patient’s care, encouraging relatives to come at mealtime and bath time to lend a hand, and to stay after visiting hours to help the patient get ready to sleep. At all the hospitals, the staff learned a new way of working: physicians did not necessarily visit swing-bed patients daily, as they typically do acute-care patients. Nurses carried out assessment and referral for swing-bed services and made recommendations for long-term care to the physician rather than the other way around.

Not surprisingly, perhaps, hospitals did better at things they were used to doing: meeting basic medical and nursing needs, and providing diagnostic and lab tests. Tasks such as making an activity available for patients and performing a functional assessment presented more problems, and hospital staff members generally did not carry them out as well as nursing home staff people did. In large part, this difference reflected the low number of swing-bed patients. It’s one thing to organize bridge games, music recitals, crafts activities, and outings for a score of patients in a nursing home, but quite another to do so for the three patients in your swing beds. The staff time per patient soars, and it’s tough to motivate staff members to shift gears when they typically have only three to five swing-bed patients. When there’s an acute-care patient in this bed, an aide may feed her. But the aide has to sit patiently while the swing-bed patient right next door feeds herself. Finally, some 40 percent of swing-bed patients stayed fewer than ten days, so the difficulty of long-term assessment left many staff members frustrated. Swing beds “lack the critical mass of patients to carry out successful patient activities or to justify hiring some specialized staff,” concluded Joshua Wiener, a Brookings Institution researcher.¹⁶

An unforeseen problem arose from the fact that swing beds are not physically distinct or separate. Tension sometimes arose: the patient in the acute-care bed received more nursing than the person in the swing bed beside him, causing resentment by the swing-bed patient (and her family). Why is Grandma getting less help and care than her roommate? The American Hospital Association, or AHA, which cosponsored the swing-bed program, concluded, “When mismanaged, swing beds can induce negative side effects . . . reducing the quality of both the acute and long-term care available to the community and having a detrimental effect on hospital finances and work relationships.”¹⁷

Although “work relationships” mattered, the AHA made it clear that the finances of swing beds concerned it most. The association exhorted members to use available beds as acute-care beds whenever possible; if a hospital has so many swing-bed patients (requiring skilled nursing care) that it has to turn away acute-care patients, the bottom line suffers (since reimbursement rates are less). Even hiring staff members for skilled nursing-care patients is financially problematic, the AHA warned.¹⁸ If such patients “are given the same level of nursing care as acute patients, the income statement is again threatened,” the association continued. It exhorted hospitals to get relatives to lend a hand feeding and dressing patients.

Unexpectedly, swing beds benefited hospital staff in some ways. The presence of specialized professionals and the chance to learn about the needs of the elderly made hospital staff members more aware of the rehabilitative and psychosocial needs of all patients, particularly elderly acute patients. Also, the need to provide swing-bed patients with physical, occupational, and speech therapy, as well as social services and discharge planning, made these processes available to acute care patients too. Acute patients often availed themselves of social programs. Also, swing beds let nurses take more responsibility for patient care, and physicians less, offering a better chance for professional development. Swing beds were even good for the bottom line, yielding some 8 percent of total inpatient revenue. This covered the cost of the program and, in the majority of hospitals, reduced the deficit or increased a surplus.¹⁹

Swing beds were still not stealing patients from nursing homes, but not for the reason their proponents expected. By 1985, it was clear that swing beds were serving primarily near-acute care patients. Their “long-term care” wasn’t very long; recall that the average length of stay for all swing-bed patients was twenty days. Compared to a nursing home patient, a swing-bed patient was more likely to be recovering from surgery, to have an IV catheter, and to have had a recent stroke or hip fracture or heart attack with congestive heart failure. “In rural communities, swing-bed hospitals had gravitated predominantly toward providing near-acute care and were not competing with nursing homes for more traditional chronic care patients who required maintenance and palliative care,” Shaughnessy concluded.²⁰ Hospital swing-bed programs, in other words, were playing to their strength—something that continues today. Swing beds, as Kovner and his colleagues noted, seem to “work best for patients who require short-stay, medically intensive services in small, rural hospitals that can recruit the necessary specialized staff. It is not a program for all post-acute patients, particularly long-stay patients.”²¹

That did not always placate nursing home operators, however. Throughout the 1980s, swing-bed programs still faced opposition from the industry, which in some states lobbied hard to require a hospital to be licensed as a nursing home before being allowed to operate swing beds.

Are swing beds cost-effective? As long as the hospital has surplus physical capacity and staff, the answer is yes: the marginal cost of providing long-term care in a swing bed runs some 45 percent less than it would have been to provide that care by building a nursing home, noted Joshua Wiener of Brookings.²² If an existing home has excess beds, however, or if it can easily add beds, then that option is cheaper than swing beds. “The swing-bed program cannot, on its own, save a failing hospital,” a 1985 conference sponsored by The Robert Wood Johnson Foundation program and the Texas Hospital Association concluded, “but it can provide revenue to help a hospital survive.”²³ By the mid-1980s, hospitals with swing-bed programs were gaining, on average, between eight and ten dollars a day per patient under the program. The ancillary care cost hospitals about twenty dollars a day per patient, but brought in some thirty-two dollars in daily revenue. Although some hospitals groused that they should be reimbursed at a higher rate, the financial set-up cannot have been too bad; by the mid-1980s, more than half the eligible hospitals were participating. Although revenues typically exceeded costs in hospital swing beds, the programs nevertheless saved money for Medicare and Medicaid. “It appears that the swing-bed program has paid for itself under Medicare by lowering costs through reductions in rehospitalizations and physician reimbursement after discharge,” Peter Shaughnessy observed in 1991.²⁴

By that year, swing-bed programs were “providing patients with demonstrably higher-quality care than they were likely to receive elsewhere, minimizing burdens on family members, and helping to insure the survival of rural health care institutions,” wrote Bruce Vladeck, then president of the United Hospital Fund of New York and former head of the federal Health Care Financing Administration. But if the outcome was noteworthy, so was the process. Swing beds evolved “from an initial though preliminary experimentation, to more systematic testing, and then through successive iterations of modest expansion accompanied by thorough evaluation,” Vladeck noted—the way innovative health care programs are supposed to arise, perhaps, but rarely do. He elaborated: “An innovative idea was tested on a small scale; the apparent success of that innovation was followed by somewhat larger tests of the idea with minor variants, and those tests were closely monitored; new policy, based largely on the results of those tests, was adopted.”²⁵

Swing beds have remained a rural program. Periodically throughout the 1990s, advocates of swing beds approached the Health Care Financing Administration to argue that they be expanded to urban hospitals. City hospitals, however, could not make the case that they were providing a service for which the elderly would otherwise have to travel to a distant facility (after all, most cities have lots of nursing homes). In rural areas, however, swing beds have taken hold. In 2001, “about 63 percent of rural hospitals, about one thousand, had swing bed agreements,” says Ira Moscovice, professor and director of the University of Minnesota Rural Health Research Center. “Swing beds accounted for about 20 percent of the patient-days in these hospitals, or eight hundred days per hospital. Just as in the beginning of the program, swing beds won’t make or break a rural hospital, but they do make a difference on the margin—the difference between a hospital being a little bit in the hole to balancing its books. Swing beds remain an important program for rural hospitals.”