Jump to ContentJump to Main Navigation
Old and Sick in AmericaThe Journey through the Health Care System$

Muriel R. Gillick

Print publication date: 2017

Print ISBN-13: 9781469635248

Published to North Carolina Scholarship Online: May 2018

DOI: 10.5149/northcarolina/9781469635248.001.0001

Show Summary Details
Page of

PRINTED FROM UNIVERSITY OF NORTH CAROLINA PRESS SCHOLARSHIP ONLINE (www.northcarolina.universitypressscholarship.com). (c) Copyright University of North Carolina Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in NCSO for personal use. Subscriber: null; date: 21 September 2021

The Varieties of Hospital Experience

The Varieties of Hospital Experience

(p.97) Chapter Six The Varieties of Hospital Experience
Old and Sick in America

Muriel R. Gillick M.D.

University of North Carolina Press

Abstract and Keywords

American hospitals come in a variety of flavors: teaching and non-teaching, for profit and not-for-profit, large and small, government and private, urban and rural. While the patient’s experience varies slightly depending on the type of hospital, all hospitals could be improved to better serve the needs of older patients if they implemented basic geriatric principles.

Keywords:   Academic teaching hospital, For-profit hospital, Veterans Hospitals, Hospital systems

My first “rotation,” as the intense, one-to-three month hospital stints of medical students are known, was at the Beth Israel Hospital in Boston, which one merger later became the Beth Israel Deaconess Medical Center—the same hospital where Barbara Ellis would one day be a patient. The new version of this venerable institution (Beth Israel first opened its doors in 1916 as a forty-five-bed hospital serving the Yiddish-speaking Jews of Boston) spans two campuses, boasts 672 beds, incorporates an equally venerable Methodist hospital, and is a major teaching hospital of Harvard Medical School. But already in the 1970s, the BI, as we called it, was a sophisticated, modern academic hospital. Medical care was delivered by teams made up of three interns—newly minted physicians in the first year of a three-year residency—and a couple of junior residents, slightly more seasoned physicians in the second year of residency. The teams were supervised by a community-based doctor or sometimes by a specialist on the hospital staff who agreed to spend a month “attending” on one of the medical floors as part of his or her—mainly his, at the time—responsibilities. I remember that the senior physician during one month of my medical clerkship (another name for a rotation) was a primary care doctor who practiced at Harvard Community Health Plan, a newly formed multispecialty group of physicians, and the other was a gastroenterologist on staff at the BI who spent most of his time peering through endoscopes and colonoscopes. And then there were the medical students, one for each of the three interns.

We medical students spent our time shadowing our intern mentors. We were charged with replicating much of what they did, which involved performing a comprehensive admission history and physical examination and writing up our findings—by hand—in the medical record. The medical students also did the scut work, the lowest status busy work: we chased down lab results, drew blood samples, and looked at urine specimens under a microscope. What was striking about the Beth Israel Hospital, as I would come to appreciate when I rotated to one of the other hospitals in the Harvard Medical School orbit, was that it had its own culture.

Some hospitals, such as the BI, are known for being academic, which means the medical students are expected to look up journal articles pertaining (p.98) to the diseases for which their patients were admitted and report the principal findings. But within the designation “academic” was considerable variability. At Massachusetts General Hospital, another of the major Harvard teaching institutions, with its moniker “Man’s Greatest Hospital,” students were rumored to be required to make their presentations of journal articles without relying on notes. The BI, by contrast, was reputed to be a genial hospital, and its friendliness took many forms in addition to gentleness with students. Long before the term “patient-centered” entered into common usage, the BI had invented “primary nursing,” a model of care in which patients are assigned the same nurse every day of their stay, at least during the day shift. And the BI was generally felt to be to the Peter Bent Brigham Hospital, another major Harvard teaching hospital just a few blocks away, much like Boston was to New York: smaller, more intimate, more manageable, but still a first-class destination. For the BI, that meant the members of its medical staff were every bit as smart and its technology as cutting edge as the competition’s.

Mount Auburn Hospital, where I had taken my introductory course in physical diagnosis during my second year of medical school—before entering the dizzying world of clinical rotations—had a very different culture. It was fundamentally a neighborhood hospital that catered to the surrounding community, but unlike most such institutions, it boasted a small residency training program and hosted occasional medical students. With some notable exceptions, Mount Auburn’s physicians did not perform the most elaborate operations or utilize the most sophisticated technology; the complex patients who stood to benefit from those modalities were transferred to a “tertiary” referral hospital such as Mass General or the Brigham or, for some procedures, the BI. What the Mount (as we called it) was best at were the more commonplace conditions—pneumonia, angina, hip fractures. Its staff physicians were held to a high standard and were expected to know about the latest treatments and the newest guidelines. But interns and residents cited Harrison’s textbook, Principles of Internal Medicine, on rounds, not last week’s issue of the New England Journal of Medicine. The hospital had only a couple of medical floors, a small ICU, and an equally small coronary care unit (CCU). The nurses knew all the doctors and the doctors knew all the nurses, as well as many of the other support staff. Many of the staff members had worked at Mount Auburn for years—they’d been born there, they lived in the vicinity, and they were employed there. The hospital looked like a diminutive version of Beth Israel: it had the same two-bed patient rooms flanked by long corridors, the same central nursing station where doctors wrote their notes and their orders, and the same x-ray machines tucked away in the basement. But it (p.99) felt different. If a hospital, with its polished floors and institutional décor, could feel homey, then Mount Auburn was homey.

When it was time for me to embark on my residency in internal medicine, I went to a very different kind of institution. I spent most of my waking hours and some sleeping hours at Boston City Hospital, a moderately large municipal hospital. Although it was a public hospital, it was also an academic teaching facility originally affiliated with all three Boston medical schools, but in my day, exclusively associated with Boston University School of Medicine. Boston City had its own culture too, featuring a pioneering, do-it-yourself spirit that arose initially out of necessity—public hospitals are not known for their amenities—and later by self-selection. There were teaching attending physicians, but it was the residents who ran the show. The residents, along with their interns and medical students, evaluated the patients, collected test results, read the medical literature, synthesized the information, made a diagnosis, and instituted a plan. The next day, after we’d already ordered the tests and procedures that comprised our plan, we informed the attending physician what we’d been up to. There were phlebotomists to draw blood, but more often than not they couldn’t get anything out of the scarred veins of drug addicts or didn’t dare approach the alcoholic patients in the throes of delirium tremens, leaving the job for the intern. There were a few IV nurses, nurses whose job was to put in an IV and check on its status every day or two, replacing it if needed, but I don’t remember seeing them around very often. If the team wanted to treat a patient with intravenous medication, the intern started the IV. At Boston City Hospital, we didn’t expect interns to quote chapter and verse of standard texts or to have read the latest journal article; what we valued was getting things done.

American hospitals come in a variety of flavors: teaching and non-teaching, for-profit and not-for-profit, large and small, government and private, urban and rural. Their different cultures translated into varying experiences for me as a medical student and then as a physician; surely those cultural differences affect what happens to patients as well. But how much the differences matter in terms of clinical outcomes, personal satisfaction, and actual care is not so clear. Would Barbara Ellis’s experiences—the unfamiliar and alien surroundings, a new doctor, delirium, a hospital-associated infection—have been any different at another hospital?

(p.100) The Academic Teaching Hospital

Of the roughly 5,000 general hospitals in the United States, only 400 call themselves academic medical centers. To qualify, a hospital has to be affiliated with a medical school and engage in teaching and research along with patient care. They are considered tertiary referral centers and are usually in urban locations. Not all major cities have an academic medical center—and some, such as Boston, have several.1

What the teaching mission means for inpatients at the hospital is that they will have an entire team participating in their care, typically comprised of medical students, residents, and another category of physician-in-training: fellows. Fellows have already completed a residency, whether in medicine, surgery, or something else—there are also separate residency programs in psychiatry and neurology in addition to fields that are irrelevant for older patients such as pediatrics—and are specializing further. In the case of internal medicine, that entails developing additional expertise in oncology, nephrology, cardiology, or any of a host of other subspecialties. In surgery, it involves acquiring special skills in a particular region of the body—thoracic surgeons operate on the heart and lungs, and otolaryngologists operate on the head and neck. Being cared for by a team creates built-in redundancy since each member of the team takes a history, performs a physical examination, and reviews the laboratory data as it trickles in. The system is supposed to provide a series of checks and balances—any time one member of the team finds something interesting, such as a new heart murmur or a new test result, all the others verify the finding and debate its significance. But it also stimulates extensive test-ordering since all those trainees don’t want to leave any stone unturned. They haven’t yet developed the finely honed clinical judgment that will allow them to focus on the most likely diagnoses and only seriously consider the more unlikely ones if the more plausible ones don’t pan out. Being a patient in a teaching hospital also means an increased chance of treatment with a new or experimental therapy that may not yet be available anywhere else.

At the same time that patients in a teaching hospital are surrounded by people who are eager to develop their clinical skills, who want to take care of patients, the more complicated the better, they are also surrounded by research physicians who may be more interested in what happens in their research lab than at the bedside. Physicians on the staff of teaching hospitals, at least of the major teaching hospitals, which are generally the largest academic institutions, spend a good deal of their time designing studies, writing grant (p.101) applications, and carrying out their research. The Beth Israel Deaconess Medical Center (BIDMC) for example, where Barbara Ellis was a patient, is regularly among the top five hospitals in the country as measured by the amount of grant money its researchers are awarded by the National Institutes of Health: in 2015, BIDMC received a total of $116 million, though this was overshadowed by two other Harvard affiliates, the Brigham and Women’s Hospital, with $334 million, and Massachusetts General Hospital, with $353 million.2

The crucial question is whether any of this affects what happens to patients. It certainly affects the day-to-day experience of hospitalization in small ways: perhaps patients will have their blood drawn daily rather than every other day, maybe when they go to the radiology suite for an x-ray, the test they have will be a PET scan rather than an MRI. But do teaching hospitals have any larger or more durable effects? Will patients go home with different medications when they are discharged from a teaching hospital rather than from a non-teaching hospital, or will they have more—or fewer—hospital-associated complications? Are they any more likely to improve clinically?

Most physicians who practice at teaching hospitals are convinced that the quality of care they provide is superior to what is available anywhere else. For some technically complex procedures—quintuple coronary artery bypass surgery or a Whipple’s procedure, in which the pancreas is removed and new connections established among the intestinal organs—it’s true that teaching hospitals usually perform better. Many small community hospitals won’t even undertake these challenging operations. But for the bread-and-butter problems such as appendectomies, pneumonia, and gastrointestinal bleeding, the outcomes at community hospitals are at least as good as at their academic counterparts.3 And in some of the areas that matter most to older people, teaching hospitals aren’t quite as good as they are cracked up to be.

For patients with advanced illness who may be in their final months of life, the nation’s teaching hospitals, even the most prestigious of them, offer varying approaches to care. That’s the conclusion drawn by researchers who examined the performance of those hospitals that U.S. News and World Report ranked among the best hospitals in the country in cardiology, pulmonary medicine, or geriatrics. Evidently somebody thought those seventy-seven hospitals were top-notch places for people with the kinds of medical problems that lead to serious illness. Next, the researchers identified a group of patients who were “loyal” to this group of hospitals, patients who got their care at those institutions on a regular basis. Finally, they asked what sorts of health care resources these loyal patients of the country’s best hospitals used (p.102) in their last six months of life and how intensively they used them. What they found was tremendous variability across the seventy-seven hospitals, variability that couldn’t plausibly be explained by differences in how sick the patients were. And since the hospitals couldn’t all be providing ideal medical care, given that what they provided varied so much, they couldn’t all be first-rate.4

The study’s authors were at great pains to say relatively little about what they thought constituted good care. But they made an exception for referral to hospice, which most authorities believe constitutes optimal care for dying patients. A low rate of hospice enrollment among patients who died, according to this standard, would imply that many patients didn’t get the best possible care—and in fact, hospice enrollment varied among the seventy-seven hospitals from 11 percent to 44 percent. The researchers also recognized that most people say they would prefer to die at home rather than in the hospital, and accordingly ranked care for terminal patients who died in hospitals as inferior to that for patients who stayed at home. It turned out that, in some of the hospitals, only 16 percent of patients died at home whereas in others, 56 percent did. Lastly, the researchers thought it was at least possible, and maybe even likely, that patients who were dying would prefer not only to die at home, but also to spend the months before their death at home rather than in the hospital. And perhaps not surprisingly, the amount of time that patients who ultimately died typically spent in the hospital during the six months prior to death could be as low as nine days or as high as twenty-seven days.5

The For-Profit Hospital

If academic affiliation does not predict what happens to patients in the hospital, maybe ownership status does. The majority of American hospitals are nonprofit institutions, but a substantial and growing number are for-profit, with the remainder publicly owned. These proportions vary tremendously by geography, with far more hospitals in the South owned and operated by for-profit corporations than in the Northeast. Nevada and Florida are essentially tied for having the highest rate of for-profit hospitals, with just over half of their institutions in this category. Four states, Hawaii, Minnesota, New York, and Vermont, don’t have any for-profit hospitals at all, and another three, Connecticut, Maine, and Maryland, have just a handful.6 While ownership might differ, in actuality all these hospitals seek to make a profit; the difference between them lies in what they do with the profit. For-profits use the income they earn to distribute to shareholders, while not-for-profits plow it (p.103) back into the facility in the form of new or enhanced technology and services. But even that distinction is not quite accurate since a growing number of for-profits are owned by private equity firms. In 2010, Nashville, Tennessee-based Vanguard Health Systems, which is owned by the private equity firm Blackstone Group, purchased Detroit Medical Center. A year later, St. Louis, Missouri-based Ascension Health took investment from Oak Hill Capital Partners, another private equity firm. In such cases, both the hospital’s earnings and also capital from the parent company can be invested in upgrading the facility, at least over the short run.7

Not-for-profit hospitals are supposed to provide a public good in exchange for which they don’t have to pay taxes. For-profits don’t operate under the same constraints, but you’d never know it from their mission statements. Memorial Hospital in Jacksonville, Florida, is a 418-bed community hospital that is part of the giant, for-profit hospital system Hospital Corporation of America (HCA). Its patients are drawn from the residents of Jacksonville, the largest city in Florida, a city where 11 percent of the population is over age sixty-five. All in all, it’s an average hospital in an average American city. Its mission is to “deliver high quality compassionate healthcare to all in our community.” Travel west for 2,000 miles and you will reach the Banner Estrella Medical Center in Phoenix, Arizona, a 305-bed community hospital that is part of the large, not-for-profit hospital system Banner Health. Its patients are residents of Phoenix, the largest city in Arizona, where 8.4 percent of the population is over age sixty-five. Banner Estrella is another typical hospital in a typical American city. Its mission statement is “to make a difference in people’s lives through excellent patient care.”8

The real question is whether what happens to patients at Memorial Hospital is substantively different at Banner Estrella Medical Center. In general, for-profit hospitals are more likely to offer relatively lucrative medical services such as open-heart surgery, and are less likely to offer money-losing services such as emergency psychiatric care than are nonprofits.9 But, while some studies have found that for-profit hospitals have higher mortality rates and more complications than not-for-profit hospitals, other studies fail to bear this out. Whether there’s a difference depends on geography and on what diseases the patients were admitted for, as well as on the era in which the study was conducted.10

To try to tease out whether ownership matters for clinical outcomes, researchers at the Harvard School of Public Health looked at what happened when hospitals were acquired and changed from a not-for-profit status to a for-profit status. The buyouts regularly improved the hospital’s financials but (p.104) had no impact on either mortality or clinical outcomes of Medicare patients.11 If ownership matters, independent of hospital size, location, and the type of patients, it’s awfully hard to prove that it does.

The Veterans Administration Hospitals

Maybe private hospitals are all alike, whether they are for-profit or not-for-profit, but perhaps public hospitals are different. A sizable number of older people, primarily older men, receive health care through the Veterans Administration (VA) system, and some of these veterans, though not all, use VA hospitals for inpatient care.12 The VA is a fascinating anomaly in the United States health care system: it’s totally run by the federal government, unlike Medicare, which is just a government-operated health insurance program. And despite the 2014 scandal over prolonged wait times for appointments, the VA system overall provides high-quality medical care. It has offered first-rate care since its transformation in the 1990s, when it made managers more accountable, improved the coordination of care, introduced quality improvement measures, and upgraded its information systems.13

The VA has a long history of innovating in the arena of geriatric care. It has a network of Geriatric Research and Education Clinical Centers that conduct cutting edge research in how to care for older patients. The earliest study of an inpatient geriatric service was carried out at the Sepulveda VA Medical Center in Los Angeles, California; it showed dramatically better outcomes for those frail older people cared for in this special environment compared to those getting usual care. Patients went home with better function and, after one year, they had experienced less mortality and fewer admissions to nursing homes.14 The improvement in survival was not borne out in subsequent studies, but another examination of the VA’s “Geriatric Evaluation and Management” units, this one focusing on frail older patients and conduced at eleven VA centers, confirmed that patients didn’t decline nearly as much during their hospital stay if they had the benefit of a specialized geriatric unit.15

The VA also has a model home visit program and an Extended Care Program, its name for a service that provides care in long-stay nursing facilities. Both options offer an alternative to hospitalization for selected patients. These special programs and services translate into a unique hospital experience for VA patients, and the VA regularly outperforms Medicare on quality indicators.16 But not many older patients get medical care in the VA system. Only those with a “service-related condition” qualify, and since older (p.105) people who qualify for VA benefits are also eligible for Medicare, many of them use the private system of health care. Even among patients who make use of various outpatient VA facilities, a substantial proportion seek care in hospitals outside the VA system.

What Shapes the Hospital Experience?

Big or Small

Perhaps what primarily determines the patient’s hospital experience is simply the size of the hospital. Large hospitals are inevitably more confusing for older people to navigate; they’re by definition less intimate. They feel different, too: the employees don’t all know each other and they certainly don’t know all the patients. But size is intimately intertwined with geography and with academic affiliation. A hospital in the rural Northeast is categorized as small if it has fewer than fifty beds, but in the rural West it is deemed small if it has less than twenty-five beds. Similarly, an urban teaching hospital in the Midwest is designated large if it has more than 375 beds, but it’s considered large only if it has more than 175 beds if it’s an urban, Midwestern non-teaching hospital. And in the South, urban teaching hospitals are large if they have more than 450 beds but urban non-teaching hospitals are categorized as large if they have over 200 beds.17

Intuitively, size ought to affect the patient experience. Large hospitals do more of everything, including surgical procedures, and it’s well established that mortality rates among Medicare patients are inversely associated with surgical volume for all kinds of procedures, including colectomy, coronary artery bypass surgery, and prostate surgery.18 Larger hospitals can afford to purchase the latest, most expensive high-tech equipment, confident that their investment will pay off because they have the necessary patient volume to support its use. Some hospitals operate their scanners around the clock, and not just for inpatients—outpatients will have appointments scheduled at nine in the evening or six in the morning to keep the machines whirring. But technology is like housing—as the saying goes, if you build it, they will come. When a patient like Barbara Ellis developed delirium, the temptation was great to send her for a CT scan just to make absolutely sure that the cause of her confusion wasn’t a stroke or bleeding in the brain, even though the doctors had a perfectly good explanation for her condition—the new medication they had prescribed for her. If they had had to arrange to send her across town to another facility for the scan, suddenly the test wouldn’t have (p.106) seemed so important, but at the large urban hospital, all it took to order a scan was a click of the mouse.

A single extra scan might not dramatically alter the patient’s experience—unless, of course, the sedative the patient gets in order to stay still for the scan leads to other problems, such as trouble swallowing, resulting in further complications, such as pneumonia. But when the tests add up, when they include very invasive techniques, and when they are carried out in the last months of life, at which point they are very unlikely to be beneficial, they do affect quality of life in important ways. Certain procedures in particular, including use of a ventilator for breathing, or more generally ICU care, in which testing is a way of life, are generally acknowledged to be burdensome for patients. If they have a reasonable chance of conferring benefit—of curing disease or prolonging life—many patients are willing to accept the extreme discomfort of these procedures. But if they don’t have much likelihood of benefit, then they’re best avoided. And it turns out that in hospitals all over the country, patients like Barbara Ellis with advanced heart disease, as well as patients with advanced dementia, advanced cancer, and other similarly dire conditions, are getting just this kind of care, whatever the facility’s size.

Among older patients with very advanced cancer, by way of example, 29 percent are admitted to the ICU in what will turn out to be the last month of their lives. A smaller, but still significant, percentage gets some kind of life-sustaining treatment in their last month, whether it’s attempted CPR, a respirator, or a feeding tube. Six percent of them even get chemotherapy just days before death.19

Extensive use of high-tech treatment isn’t limited to cancer patients in their final weeks. Medicare patients who died in 2012 spent an average of 3.6 days in the ICU at some point during their last six months of life, with the amount of time ranging from one day in some areas to nine days in cities such as Miami and Los Angeles. And patients with dementia who died continued to have feeding tubes inserted to supply nutrition during the last six months, even though the procedure is widely held to confer no advantage to such patients. The risk of having a needless surgical procedure to insert a tube in the stomach varied considerably from 2 percent in Portland, Oregon, to 12.8 percent in Los Angeles, California. The exact reason for the variability is uncertain, but hospital size was largely irrelevant.20

Geography as Destiny

Maybe what determines both patients’ degree of satisfaction and their clinical outcomes in the hospital is whether they happen to be in a rural area, (p.107) where only 38 percent of U.S. hospitals are located, or an urban area, where the majority are found.21 Figuring out whether the urban/rural distinction matters proves to be rather tricky since rural hospitals differ from urban hospitals in two important ways: they are rarely teaching institutions and they are almost always small. Once these factors are taken into account, there’s not much that distinguishes rural from urban hospitals.

Geography does turn out to matter, but sorting out its precise role is a good deal more complicated than just whether the hospital’s location is rural or urban. Since 1996, researchers at the Dartmouth Institute (which used to be called The Center for Evaluative Clinical Sciences) have been tracking how much and what kind of care Medicare enrollees receive in their last two years of life. What they find, no matter whether they are looking at the number of different doctors that patients see—they are particularly interested in patients who see ten or more physicians over the course of a year—or the number of days they spend in an ICU in their last year of life, or the proportion of patients who die in the hospital, is mind-boggling geographic variability. When they correct for regional differences among patients, such as differences in the average age or in racial makeup or in socioeconomic status, the distinctions persist. But the phenomenon isn’t purely regional: there is no one pattern that describes all Midwestern hospitals and another that describes Southern hospitals. There is variability between cities within the same state. There is variability between hospitals in the same city. As long as there’s no absolutely, unequivocally optimal treatment for a particular condition—say, surgery for a hip fracture—whenever the treatment is “preference-sensitive,” or affected by the values and attitudes of patients, physicians, or both, there is geographic variability. And different kinds of treatment translate into vastly different experiences for patients.22

Whether a chronically ill older patient will spend any time at all in a hospital depends as much on where he lives as on what he has wrong with him. Manhattanites as recently as 2007 spent an average of 20.6 days in the hospital in the six months before they died; residents of Ogden, Utah, spent 5.2 days. When patients did end up in the hospital, whether they were admitted to a regular hospital floor or the ICU, with its beeping alarms, perpetual daylight, and sophisticated technology, was also a function of where they lived. In Miami, Florida, they spent 10.7 days in the ICU; if they lived in Bismarck, North Dakota, they spent 1.1 days in the ICU.23

(p.108) Hospital Systems

Perhaps the patient’s hospital experience is shaped by whether or not the hospital is part of a health care system. Over the past twenty-five years, a growing percentage of hospitals have joined some kind of larger umbrella organization, fueled in part by hospital executives’ efforts to improve efficiency through mergers and acquisitions. The first wave of hospital consolidation occurred in the 1990s; the merger fever subsequently subsided, but since the passage of the Affordable Care Act (ACA) in 2010, with its new incentives promoting efficiency, mergers and acquisitions have again been on the rise. Those hospitals that weren’t already part of a system hastily sought to join one or create a new system by developing a business relationship with other solo hospitals. Existing hospital systems went on a buying spree, snapping up small community hospitals, and in some cases converting not-for-profit hospitals into for-profit hospitals.24

Hospital CEOs, anxious about what they regard as inadequate reimbursement by Medicaid, promote geographic expansion for another reason—to gain access to better-insured patients. Sometimes this approach involves building a new, full-service hospital outside their usual jurisdiction; in other cases, hospitals buy or merge with existing community institutions in areas with a high proportion of the privately insured—patients who don’t have government insurance such as Medicare or Medicaid. At the same time, hospitals buy up or occasionally form physician practices to serve as referral sources for their hospitals.25

Hospital systems are not confined to the for-profit world. One system that was born during the first wave of mergers and that underwent a second growth spurt during the post-ACA era is the non-profit Massachusetts heavyweight, Partners HealthCare. Partners burst onto the Boston scene in 1994, when the two largest and most prestigious Harvard Medical School teaching hospitals, Massachusetts General Hospital (MGH) and the Brigham and Women’s Hospital (BWH), announced they were forming an alliance. In their first decade, they acquired five community hospitals extending from Martha’s Vineyard, an island off the coast, to the Pioneer Valley in the western part of the state. Included were two small neighborhood hospitals in their own backyard, one in the affluent western suburbs and another just one mile from BWH. As a result, Partners could shunt patients with routine medical problems such as pneumonia or appendicitis to one of its smaller facilities, whether the patients liked it or not. The idea was that the medical issue would be dealt with relatively inexpensively in the small community hospital, thus (p.109) freeing up beds in the tertiary care center for more complex patients. In addition, the network facilitated the transfer of patients from a low-tech to a high-tech facility if their condition warranted it—cardiac catheterization and open-heart surgery were done at BWH and MGH, but not at the lesser community hospitals. For patients, the net effect of the expansion is that they often start out at one hospital near home and end up downtown; the result for Partners was that by 2012, the chain collected nearly one-third of the money spent by Massachusetts commercial insurers on acute care.26

Having made major inroads into the markets east and west of Boston, Partners was eager to have a presence north and south of the city. As a result, when the 378-bed South Shore Hospital showed signs of financial instability in 2012, Partners swiftly stepped in with a plan to buy it, hoping to extend its empire south of Boston. The following year, it set its sights on two small hospitals comprising the Hallmark Health System, which would facilitate expansion north of Boston. Both deals were challenged on the grounds that the cost of health care in Massachusetts, already one of the costliest states in the country, would have risen further if Partners gobbled up additional hospitals. The Health Policy Commission, appointed by the governor to help constrain costs after Massachusetts passed its universal health care law in 2006, agreed, estimating that the deal would increase health care spending by $23 to $26 million per year.27 Subsequently, the Commission revised its estimate upwards. Even with the settlement proposed by the Massachusetts attorney general, which would have limited price increases across the network until 2020 and allowed insurance companies to bargain with the constituent hospitals individually rather than with the Partners behemoth, prices were predicted to rise considerably. Partners, by virtue of its size, exercised enormous bargaining clout with third-party payers. It was in the enviable position, from the perspective of hospital CEOs, of being able to extract generous payments from insurers—or it would refuse to care for the insurer’s members. With strong political winds threatening to topple the agreement, Partners agreed to put its proposal on hold. It was a proposal that would have affected costs for every patient in the Commonwealth and that might have forced patients to change doctors or hospitals for insurance reasons.

At the same time that Partners was flexing its muscles, another hospital system arose to provide some competition, Steward Healthcare. Steward represented a new departure for Massachusetts, its first major foray into for-profit medicine. While states such as Tennessee, home to HCA, have long been dominated by for-profit hospitals, other states such as Massachusetts have a tradition of not-for-profit medicine. In a bold move, the private equity (p.110) firm Cerberus Capital Management bought the ailing Catholic hospital chain Caritas Christi Health Care, which owned two financially troubled Boston hospitals. Cerberus immediately converted Caritas to a for-profit company which it named the Steward Healthcare System. Within two years, Steward had transformed itself into a network of eleven hospitals with roughly 2,100 beds. Every year it operated at a loss, plowing millions of dollars into construction projects (including a new cardiac catheterization laboratory at one hospital and new emergency departments at two others), as well as into information technology. Steward gambled that it could provide lower cost care than Partners without requiring patients to travel to Boston for complex treatment.28 Ultimately, it would have to stop relying on infusions of funds from its private equity progenitor and become profitable—typically, private equity firms buy companies, restructure them, making them more efficient, and then sell them within eight years.29 Constrained to some extent from closing critical but money-losing services such as mental health by its agreement with the Massachusetts attorney general, Steward nonetheless engaged in other cost-cutting activities. It repeatedly cut nursing staff, leading the Massachusetts Nursing Association to file a complaint against Steward, since quality of care has been shown to be directly related to the ratio between nurses and patients. In fact, the nurses filed over 1,000 complaints against Steward, alleging that its practices impair quality. In addition to firing nurses at several of its hospitals, it eliminated the geriatric psychiatry unit at one of them, Quincy Medical Center, leaving older psychiatric patients without any comparable facility in the area. Then it closed down the hospital almost entirely, retaining only an urgent care center and a free-standing emergency room, leaving thousands of residents of the area without a community hospital.30

Steward also affected patients’ choice of doctors and hospitals. It aggressively recruited physician practices to join its network in order to keep the pipeline to its hospital chain full. Its tactics sometimes backfired, as when it signed on the 200-physician Whittier Independent Practices Association, luring them away from the third major hospital system in the Boston area, CareGroup, only to have the group turn around and rejoin CareGroup two years later. Whittier then sued Steward for withholding the incentives contractually owed them, a step taken by Steward, or so the physicians argued, to punish them for having withdrawn from the system.31 Patients, in the meantime, had a choice between switching doctors to keep their ties with the hospital they had been using, or keeping their doctor and switching hospitals.

(p.111) More Alike Than Different

The older patient’s hospital experience varies depending on the type of hospital—a little bit. At an academic teaching hospital, patients are more likely to get lots of tests than at community hospitals, but they can count on having plenty of diagnostic tests and procedures wherever they land. They are more apt to receive the latest therapies at a teaching hospital, but they are also less likely to be cared for by a physician who knows them, sometimes leading to their getting a treatment they’ve had—and that failed—in the past. If patients are at a large hospital, they are on average more likely to receive technologically intensive medical care than at a smaller facility with fewer resources, but they may simply be transferred from one site to another if they could potentially benefit from more elaborate interventions.

Similarly, whether the hospital is for-profit or not-for-profit can affect the patient’s experience—sometimes. The for-profits are more likely to cut money-losing services such as mental health and to build up lucrative ones like invasive cardiology (that’s actually the name that cardiologists use for performing procedures such as cardiac catheterization). But if you have a common condition that just about any hospital is well-equipped to manage, such as pneumonia, you will probably fare just as well—or as poorly, if you develop a hospital-associated complication—at either type of institution.

Size matters to the extent that small hospitals seldom provide all possible services. But that doesn’t mean that for rarer or more complex conditions patients get poorer care at small facilities; it just means they are likely to be transferred elsewhere. And geography makes a difference for the patient’s experience, but not so much for clinical outcomes; the amount done to patients suffering from a heart attack or colon cancer or a hip fracture varies enormously from one part of the country to another, with some patients getting 60 percent more interventions than others, but mortality, patient satisfaction, and the ability to function independently are unaffected.32

The only type of hospital that consistently makes a difference in the lives of older patients is a VA hospital. The VA offers a more comprehensive package of services for frail older patients than other hospitals. It is a pioneer in geriatric assessment, home care, and long-term care for its oldest, most vulnerable patients.

All hospitals could be improved to better address the needs of older patients. The basic principles required—avoiding medications that often precipitate delirium, promoting a good night’s sleep, providing input from the primary care physician or, better yet, assuring that the primary care doctor is the (p.112) attending physician, and tailoring treatment to conform to the patient’s goals of care—are not esoteric knowledge, they are not the carefully guarded secrets of an arcane cult. They constitute the fundamentals of geriatric medicine. But hospitals only rarely implement these principles. Powerful forces—the now familiar litany of physicians, drug companies, device manufacturers, and Medicare itself—all have a vested interest in maintaining the status quo, an arrangement that functions reasonably well for younger or more vigorous patients. How they manage this feat requires looking at the role of each one in the modern hospital.


(1.) The terminology is confusing. The federal government defines a community hospital as any hospital that’s not a psychiatric hospital, not a chronic disease hospital, and not run by the federal government. I will instead use the term “general hospital” for these facilities and will distinguish between large, teaching institutions on the one hand and community hospitals on the other. Dunn and Becker, “50 Things to Know about the Hospital Industry.”

(2.) U.S. Department of Health and Human Services, “NIH Research Portfolio Online Reporting Tools (RePORT), NIH Awards.” Beth Israel Deaconess Medical Center was number 5 in the country with 237 awards; BWH was number 2 with 535, and MGH was number 1 with 772. https://report.nih.gov/award. Accessed February 22, 2016.

(6.) Kaiser Family Foundation, “Hospitals by Ownership Type.” This research found that 58.4 percent of hospitals are nonprofit and 21.4 percent are for-profit. The for-profit hospital proportion in Nevada is 51.7 percent, and in Florida it’s 51.4 percent.

(12.) Auerbach, Weeks, and Brantley, “Health Care Spending and Efficiency in the U.S. Department of Veterans Affairs.” It’s very difficult to be more precise about how many older vets get their care in the VA system because almost all veterans eligible for VA benefits (and only those with service-related injuries qualify) also have Medicare. Many patients go to VA facilities for some services and to private health care institutions for others. They also cycle in and out of the VA system.

(16.) Jha et al., “Effect of the Transformation of the Veterans Affairs Health Care System on the Quality of Care.” In 2000, the VA did better on 12 out of 13 quality indicators, compared to Medicare patients.

(19.) Goodman et al., “Trends in Cancer Care.” A total of 9.4 percent of patients got life-sustaining treatment in the last month of life.