Science Progress | Where science, technology, and progressive policy meet

Harnessing Complex Hospital Care

Hospital Practices Must Match the Moral Ends of Treatment

A doctor goes over paper records in an examine room at the Indiana Heart Hospital in Indianapolis. SOURCE: AP/MICHAEL CONROY A doctor goes over paper records in an examine room at the Indiana Heart Hospital in Indianapolis. Doctors in the "paperless" hospital still face the challenge of working with patients referred from other hospitals and doctors that do not used electronic files.

Hospitals evolve. As the Pulitzer prize-winning, Princeton sociologist Paul Starr observed, “few institutions have undergone as radical a metamorphosis as have hospitals in their modern history.” This metamorphosis, “from places of dreaded impurity and exiled human wreckage into awesome citadels of science,” gave hospitals “a new moral identity.”[1]

Now hospitals must evolve again so that their moral identity is based on more than a commitment to scientific rigor and technical proficiency. Science and technology are necessary, of course, but not sufficient. Hospitals also must attend to markers of excellent patient care in a broader sense, using a wider-lens approach that focuses on change within organizational systems complexity.[2]

Now, it is time to move beyond the provision of merely technically feasible medical intervention. It is time to create mechanisms to forestall the use of alluring technology for its own sake, when it will not provide meaningful functional benefit. It is time to shift from the ineffective cost-cutting strategies that have reduced the practice of hospital medicine to running procedures as quickly as possible, so as to stabilize a patient for rapid discharge.

These outdated health care delivery approaches have resulted in hospital practice that is harmful to patients, families, and clinicians. For patients it often results in overtreatment. For families it results in feelings of dissatisfaction with hospital staff. For clinicians, it is crushing to morale and results in burn out. “Health-care costs are being treated as if they were largely an economic problem, but they are not,” observed the noted economist Lester Thurow. “To be solved, they will have to be treated as an ethical problem.”[3]

In short, it is time to reimagine hospital care.

Hospitals must reorganize their systems of care delivery if they expect to successfully meet the moral ends of medicine—that is, to treat pain and suffering, cure if possible, and honestly acknowledge when the patient’s disease cannot be cured. We need to find ways to enable hospitals to address systems complexity directly as they work to refocus care on what patients need to improve their health and functional status—or in the case of dying patients, what they need to die well. By doing so, morally grounded cost savings can accrue and hospital care in the United States can return to a leadership position in global health.[4]

These will not be easy tasks. Hospital care is complicated, expensive, and often emotionally charged. And, with our population aging, we will see sharp increases in the number of patients with chronic conditions who can be expected to be in and out of the acute care setting over the course of their disease. In the United States, the hospital is often the last place for medical care—20 percent of all U.S. deaths now occur in an intensive care unit or immediately thereafter.[5] At the same time, satisfaction with hospital care continues to be low.[6]

Numerous variables—including the continual introduction of new technologies, the burdens patients and families experience when they are expected to control medical decision-making, rampant fears of liability, dying as a chronic disease, and radical changes in physician training and health care cost coverage structures—all converge in the hospital setting to create systems of daunting complexity. One response to this complexity is to seek to overcome it through personal effort. Hard-working clinicians, administrators, and hospital support personnel focus on their own domains of work, trying to improve the outcome of their assigned tasks and thus take better care of patients.

Yet the generally accepted goals of quality improvement in health care—that care should be safe, effective, patient-centered, timely, efficient, and equitable—cannot be achieved without engaging system complexity in hospitals. As was made clear in “Crossing the Quality Chasm,” a landmark 2001 report from the Institute of Medicine, “Trying harder will not work. Changing systems of care will.”[7]

In the pages that follow, we will present two case studies that illustrate some of the systems complexities of modern hospital care. We follow these cases with solutions that require changes at all levels of systems complexity, working to paint a picture of wheels and gears within wheels and gears.

Case One: Ms. K’s Repetitive Admissions

Ms. K, 56 years old and living with her 59-year-old sister, is brought by ambulance to the emergency department in a diabetic coma. Ms. K is transferred to the intensive care unit where, during the next 24 to 36 hours, she is stabilized. Although she is still weak, at this point she is medically stable enough to be transferred out of the Intensive Care Unit and onto a general medicine service. The ICU is in need of beds, with critically unstable patients stacking up in the emergency department.

Ms. K’s transfer orders call for her to receive diabetic education before she is discharged from the hospital. By the time Ms. K is transferred out of the ICU, however, it is late in the evening. The next day, the floor nurse calls the diabetic nurse educator, who comes and thoroughly explains the necessary medications and the importance of adhering to the prescribed dietary regimen. Thereafter, Ms. K is discharged home. About three months later, Ms. K comes back into the ICU, once again through the emergency department, for the same problem.

During rounds in the ICU, the first-year resident physician presenting the case summarizes the physiological evidence showing no co-morbidities, that is, no other serious medical problems. The first-year resident also points out that the patient has not been recently ill, nor is she a drug addict or alcoholic—all possible explanations for the patient’s present problem. Therefore, when the first-year resident ties up the patient’s symptom profile, clinical examination findings, laboratory and other diagnostic test results—concluding that the patient’s present problem is from medication non-compliance for her diabetes—the resident’s assessment goes unchallenged by the rest of the clinical team.

By happenstance the third-year resident physician who took care of Ms. K on the floor three months earlier is now rotating through the ICU. She remembers the patient from the floor, and her medical ethics training has made her alert to the issue that merely technically competent care of the patient, at the time the patient is in front of the treating clinician, often is not sufficient for excellence in the art of medicine. Thus, the conclusion of the first-year resident leaves the third-year resident uneasy because “intentional non-compliance” (the term “non-compliance” casts the patient in a pejorative light) somehow does not jibe with the resident’s recollection of Ms. K.

The third-year resident manages to steal time to read through archived chart entries and discovers that the patient had been admitted six times in the past three years to various area hospitals in a diabetic coma. Previous chart notes confirm that the sister with whom Ms. K lived was caring and attentive, although the resident had formed an impression of the sister as not really fully appreciating what medication Ms. K would need to take when she returned home.

The resident, after continuing to read through pages of archived charts, finds mention of below-normal intelligence in the patient, which suggests that when the medication and dietary regimens had been explained, the patient might have been unable to retain the information. This likely gap in the patient’s understanding, coupled with the probability that Ms. K’s sister did not fully grasp the gravity of adhering to tight dosing requirements, could account for the recurring problem.

With this information, the third-year resident realizes that both the patient and her surrogate need additional assistance. By the time the third-year resident figures all this out, however, and realizes that social work needs to be called to arrange for a visiting nurse or nurse’s aide to come into the patient’s home frequently enough to check on medication management, the patient is ready for discharge. Realistically, it will probably take the social worker a day or two to get the needed home health care supports set up.

Applying a Systems Complexity Approach to Ms. K’s Care

That Ms. K returns again and again to her local hospitals for the same problem reflects bad patient care. It is dangerous for Ms. K to repeatedly experience a diabetic coma. Viewed at the broadest level, it is expensive for the national health care system. From the perspective of those rendering the care, it is dispiriting, given that diabetic coma is, under normal conditions, an avoidable event.

Because of systems complexity, however, the repetitiveness is even difficult to spot. At one level, her care was flawless. The ICU team (that is, residents, attending physician, nurses, unit clerks) took medically appropriate actions while the patient was in the ICU; transfer orders to the floor were written properly and followed; and discharge planning and education were carried out. Nor is this a case of a patient facing socioeconomic barriers. Ms. K has insurance, so payment for care and pharmaceuticals is not the problem. She lives with her sister, with whom she has a loving relationship, so this is not a matter of having an isolated patient without adequate social support.

Rather, implementing solutions that will work for patients like Ms. K, not merely tinkering with old ones that did not, will require reinventing how we think about hospital operations. Given that information crucial to proper care can easily become invisible in a complex system, hospitals need to invest in the time, space, and staffing that would allow problems like Ms. K’s to be spotted early enough for effective intervention. This investment in problem case identification is needed for outpatients and inpatients alike, because coordinated outpatient care, or the lack of it, directly affects hospitals.

Although time, physical space and professional staff cost money, they are integral to all systems and subsystems influencing and embedded within health care delivery. Taking the short-sighted approach of looking for yet more ways to cut costs is precisely the worst strategy for managing costs, systems complexity and ethically justifiable medical care.

For Ms. K, smart systems complexity management could start with reinvention of hospital staffing patterns. Smart systems reinvention would include attaching a clinical social worker to Ms. K at her second admission. This did not happen in the case because hospitals have generally reduced the role of social workers to discharge facilitators. To achieve ethical and effective care of Ms. K, her social worker would serve as the social issues coordinator.

This reimagined social worker is the fulcrum for smartly coordinated care, linking all relevant medical, ethical, and social information so that it is not lost in a system of ever-changing medical teams. Ms. K’s social worker would assist in the coordination of all meetings about patient needs, decision-making, and transfers to various floors and units. To reduce the post-discharge risk of another bout of diabetic coma caused by non-adherence, Ms. K’s social worker would be expected to coordinate with Ms. K, her inpatient and outpatient clinicians, home health support personnel, and her sister, tracking Ms. K’s post-hospitalization progress long enough to avoid, if at all possible, rehospitalization.

What is being called for here is not simply more social workers doing real, clinical social work. Rather, the addition of more clinical social workers is one instance of the broader and more fundamental change that comes with recognizing, and responding to, the systems complexity of health care. For example, the preventive care coordination for which an augmented clinical social work staff would be responsible implies other changes, not the least of which is that information be accessible through widely adopted and compatible electronic systems of health records.

Case Two: To Transfer or Not To Transfer Mr. J.

Mr. J., an active 72 year old, lives in a rural area with his wife. On a cold day in January, Mr. J has a massive brain bleed and a cardiac arrest. His wife calls 911, and the local emergency personnel arrive within a couple of minutes, immediately beginning cardio-pulmonary resuscitation. On their way to the nearest community hospital, they regain Mr. J.’s pulse, which is sustained in the hospital’s emergency room.

Within minutes, Mr. J. is transferred to the hospital’s ICU. The ICU’s attending physician immediately recognizes that Mr. J’s brain bleed is still evolving. The attending physician has the hospital’s on-call neurologist contacted; they confer and agree that the patient may be a surgical candidate. The surgery, however, cannot be performed at the community hospital, as there are no local neurosurgeons. The ICU attending physician has the closest regional hospital contacted that can accommodate a patient in Mr. J’s condition. The accepting neurosurgeon at the regional hospital communicates with the rural neurologist, obtaining as much information as possible while the patient is being airlifted by medical helicopter to the accepting regional hospital.

Mr. J arrives at the distant, regional hospital, where the accepting neurosurgical staff immediately evaluates him. The team determines that Mr. J is not, after all, a surgical candidate. The patient is transferred to the medical ICU, where he languishes for weeks and eventually dies, after a rancorous series of encounters between the medical team and the family about establishing and sustaining realistic expectations for meaningful functional recovery in the patient. Clinically, the patient arrived unconscious and never improved during the many weeks of his ICU stay.

Applying a Systems Complexity Approach to Mr. J’s Care

As in the first case, on the face of it Mr. J’s care was technically adept. The emergency medical team ran a successful resuscitation in the field. At the rural hospital, Mr. J was evaluated quickly and competently by the appropriate complement of nurses, emergency room physician, ICU attending physician, and on-call neurologist. The determination that the patient might be a surgical candidate was made expeditiously, as was the coordination with the closest hospital that could perform the surgery and the airlift to get the patient to that next hospital. Unfortunately, all of that activity only ended up with a patient who was not, after all, a surgical candidate, in a hospital far from the patient’s home, and with family who had to traverse long distances while experiencing a roller coaster of emotions and eventually dashed expectations.

If, however, we take proper account of systems complexity throughout our hospital systems and subsystems smartly, we can reduce the misery of these situations. We can provide families and friends the environment and resources they need to process calamitous medical events and, where such turns out to be the need, to begin grieving. One of the solutions is to reinvent our hospital spaces to more currently match the severity, acuity, and social complexities 21st century patients present.[8]

Many hospitals have emergency rooms. Many have ICUs. But most have nothing in between other than standard lower-level acuity care units. So in Mr. J’s case, once the regional hospital’s neurosurgical team had decided that, for whatever panoply of conditions specific to Mr. J’s clinical status, he was not a surgical candidate, the only safe place in the hospital for Mr. J was the medical ICU. He was on a ventilator (a breathing machine), and there are few places in a hospital other than an ICU that have the staffing capability to manage patients on ventilators.

Even if, over the days and weeks of Mr. J’s ICU stay, he had been weaned off the ventilator and able to breathe on his own, his blood pressure was unstable, which meant that he still needed medications to artificially elevate his blood pressure. His immune system was also impaired, and he continued to need antibiotics. Unable to swallow, he needed these medications administered through intravenous drips, none of which can usually be managed anywhere in the hospital other than the ICU.

If, however, high acuity hospitals—the kind where patients like Mr. J are taken—were to create intensive care step-down units, then it is possible that even such hi-tech care as is required for critically ill, stable patients could be provided more humanely and with less expense. While the systems change solutions suggested for our first case study almost completely revolved around altered staffing patterns, here the focus is on altered staffing patterns and redesign of physical space.

An intensive care step-down unit, a unit some might call a hospice ICU or by some other comparable title, would be quieter. For many of its patients, the goal would be reducing aggressive, non-beneficial care. The primary oversight would be through the nurses, with expensive physicians on call as needed rather than as primary staffing for the unit. Wake/sleep lighting would be normalized. Family and friends would be welcome at any time. Such a unit would build time and space back into the dying process.

An even more fundamental reinvention solution would prevent Mr. J from ever having had to be moved in the first place. What’s more, the technology already exists.

The reason Mr. J had to be airlifted into the higher acuity hospital was because neurosurgeons make “go/no go” surgical decisions, and most community rural hospitals do not have neurosurgeons available. The neurologist at Mr. J’s rural hospital, however, would be expected to be able to conduct, under the guidance and supervision of a neurosurgeon, an examination sufficient to allow a watching neurosurgeon make the surgery decision. That is, if Mr. J’s rural hospital were equipped with the requisite telemedicine technology, the regional hospital’s neurosurgeon could direct and “watch” the examination needed to determine whether or not Mr. J should be transferred.

It also is possible, given Mr. J’s particular clinical problem—a brain bleed—that although he might have been judged not to be a surgical candidate at first examination, he might have become one over the next several days. With the electronic technology to do the original examination remotely, the geographically distant neurosurgeon could continue to “watch” Mr. J and, if the evolving clinical situation justified doing so, make a different decision.

Applying a Systems Complexity Approach

A paradigm shift towards thinking about hospital processes from the perspective of systems complexity and systems change will be difficult to accomplish.[9] But this new approach to excellence in hospital care is necessary. Achieving this shift will take the combined efforts of many disciplines working together and well-conceived policy incentives. It will also take a willingness to reimagine physician-patient relationships in new ways.

Relevant here is the reality that hospital medicine is delivered by teams. Hospital clinicians appreciate this, even if others do not. Although the traditional model of the doctor/patient relationship is still central to physicians’ professional identity, appreciation of fiduciary responsibilities and developing the character of the virtuous physician, the model is no longer the lived experience of most physicians or patients. For the hospital patient, the traditional model has been replaced by a dizzying array of doctor/patient encounters. Especially in teaching hospitals, patients and families interact with so many doctors, in such staccato exchanges, that it is virtually impossible for the traditional concept of the doctor/patient relationship to remain meaningful. This reality of contemporary hospital medicine has important implications for health care delivery organization.

In Ms. K’s case, her doctors all provided technically competent and, we will assume, empathetic care. But each time Ms. K came into the hospital, a different attending physician was on-call. The residents responsible for her care in the ICU on previous admissions had long shifted their unit rotations, and the one or two residents who might have crossed Ms. K’s path, as well as her nurses, were on different shifts during her brief stays. The fact that Ms K had health insurance does not imply that she had a primary care physician who would be noted in a hospital chart. That the third-year resident remembered Ms. K from the floor and made the time to go back through old charts was lucky indeed.

Given the relatively simple medical problem Ms K presented, the speed of her transfers from emergency department to the ICU to the floor and then her discharge, the system of care did not encourage the clinicians to discover the high number of previous admissions or the underlying causes that would account for her in-and-out hospitalization pattern. The time pressure on physicians that force them to focus almost exclusively on the scientific and technical aspects of care; the lack of other clinical professionals, such as clinical social workers who could devote time with the patient and her sister to gain a broader perspective on care issues; the inability to easily access previous hospital admission information—all of these factors conspired so that Mr. K’s technically flawless care was actually poor.

For Mr. J, ill fortune dictated his death, and rescue medicine was not available to save him. What could have been averted were weeks of nonbeneficial care and much family and clinician distress, had hospital space been organized to provide critically ill, dying patients and their families care that meets their special medical and moral needs in units other than standard ICUs. Indeed, if already existing technology were disseminated properly, critically ill, dying patients like Mr. J. can remain well cared for in rural hospitals.

We need to take a fresh look at hospital staffing. Hospital staffing patterns must be reoriented to provide patients with the caring attention they have always needed and will always need within a health care system that will only become more technologically complex. Determining exactly how to achieve this goal is a problem fraught with complexity. Hospital clinical care is, today, so heavily technological that to expect physicians, especially physicians still in training, to be as skilled at the nontechnical parts of the job as at the technical parts may be unrealistic.

The solution is not to simply try to reduce the technological knowledge required of clinicians. In time, the requirements for physicians and nurses to understand more and more technologically advanced, scientifically justified procedures will likely become more, not less, integral to hospital medicine. Instead of adding to physician and nurse burdens, we have to become more skilled at managing continuous flux and change within complex systems, becoming creative about arranging who does what in the care of hospital patients.

Another one of these system-wide, systems-redesign solutions is in the communications domain. One piece of smart complexity management here is for hospital professionals, other than physicians, to take over some of the tasks previously thought of as being only under the purview of a physician, such as discussions about end-of-life care planning, freeing up physicians to do the technical parts of medical care that only physicians are trained and licensed to do. Instituting such a change, however, requires overcoming entrenched patterns and attitudes. Many physicians believe that talking to patients about medical care planning is solely a physician prerogative, even though physicians may never have been adequately trained in effective communications skills and, especially as the prospect of successful rescue medicine grows bleaker, often shy away from exercising their assumed authority.

The result is that important planning conversations never occur. This problem is magnified by the related systems problem of the misguided cost-cutting strategies that have decimated hospital’s clinical social work staff and continue to act as barriers to building clinical ethics, palliative care, and hospice programs into daily hospital operations. The challenge of rethinking traditional hierarchy and professional obligations long embedded in the practice of medicine is difficult indeed. But these traditional practices, along with all other hospital practices and systems, must be open to review and they must be changed wherever necessary.

Reform of health care must include steps to encourage a paradigm shift from hospital medicine that is focused too narrowly on procedural skill to the realities of 21st century medicine, which is characterized by complex systems surrounding diagnostics and procedures and is delivered by teams of clinicians and support personnel. This is a task that we must undertake and complete successfully for the sake of everyone who may someday be a patient in a hospital—that is, all of us.

About the Authors

Evan G. DeRenzo, Ph.D., is Senior Bioethicist, Center for Ethics, Washington Hospital Center, Washington, D.C. Adjunct faculty at the Johns Hopkins University, Graduate Program in Biotechnology, she received her Ph.D. from the University of Maryland in Human Development with a concentration in gerontology. She serves or has served as consulting bioethicist to a wide variety of private organizations and Federal and Maryland state agencies, on multiple Institutional Review Boards and Data Safety Monitoring Boards, and on numerous hospital and professional association ethics committees, including the global ethics committee of the Association for Clinical Research Professionals and the World Federation of Societies for Biological Psychiatry. She has published widely in the areas of clinical ethics and human participant research ethics, and is the Editor-in-Chief of the Journal of Hospital Ethics.

Jack Schwartz is Visiting Professor of Law and Senior Health Law and Policy Fellow at the University of Maryland School of Law. He formerly served as an Assistant Attorney General and Director of Health Policy in the Maryland Attorney General’s Office. His undergraduate work was at the University of Maryland Baltimore County, and he then went to Yale Law School. Prior to joining the Maryland Attorney General’s Office in 1982, he held a series of senior staff positions at the Federal Trade Commission in Washington. Currently, he is a member of an institutional review board at the National Cancer Institute, a data and safety monitoring board at the National Institute on Aging, the Bioethics Committee of the Washington Hospital Center, a State of Maryland advisory committee on infections in health care facilities, and the advisory board of the Maryland Healthcare Ethics Committee Network. In recent years, he served as a member of the American Bar Association’s Commission on Law and Aging, a senior consultant to the National Bioethics Advisory Commission, a member of a national advisory committee for the Robert Wood Johnson Foundation, the State Advisory Council on Quality Care at the End of Life, the Maryland Stem Cell Research Commission, and the Maryland Task Force to Study Electronic Health Records. He writes and lectures frequently on legal, policy, and ethical issues in health care.

Stephen R. Selinger, M.D., FCCP, is President of Pulmonary & Critical Care Associates of Baltimore, currently the largest pulmonary and critical care group in the United States; staffing six hospitals and nine critical care units, and Director of Critical Care Services at Franklin Square Hospital. Dr. Selinger, a graduate from Johns Hopkins University School of Medicine 1979 and Assistant Professor of Medicine Johns Hopkins Hospital from 2003-2004, is currently Clinical Assistant Professor of Medicine, Education Office University of Maryland School of Medicine 2003-present. He is also Associate Chairman, Department of Internal Medicine, Franklin Square Hospital Center, and the recipient of multiple Excellence in Teaching awards from Good Samaritan Hospital; and Distinguished Attending of Year awards from Franklin Square Hospital. Dr Selinger is a member on multiple hospital committees and MedStar TICU that creates and implements protocols and tools to improve quality of care in critical care settings.

Disclaimer: This report contains the views of the authors, only, and does not represent the policies or procedures of any institution or organization to which any of the authors are affiliated.

Note: Portions of this article will appear in E.G. DeRenzo, “Seeking Excellence in Hospital Care: Evolving toward a Systems Approach,” in The Journal of Clinical Ethics 20, no. 1 (Spring 2009): in press.


[1] P. Starr, The Social Transformation of American Medicine (New York: Basic Books, 1982), 145.

[2] R.D. Stacey, Strategic Management and Organizational Dynamics: The Challenge of Complexity (Harlow, UK: Prentice Hall, 2003).

[3] Thurow LC., “Learning to say no,” NEJM. 311(24)(1984):1569-72, p. 1572

[4] S. Brownlee and E. Emanuel, “5 Myths on Our Sick Health Care System,” The Washington Post, November 23, 2008.

[5] R.D. Truog et al., “Recommendations for end-of-life care in the intensive care unit: A consensus statement by the American Academy of Critical Care Medicine,” Critical Care Medicine 36 (3) (2008): 953-63.

[6] S.M. Dy, H.R. Rubin, and H.P. Lehmann. 2005. Why do patients and families request transfer to tertiary care?: a qualitative study.” Soc. Sci. Med. (61):1846-1856.

[7] Institute of Medicine, Crossing the Quality Chasm: A New Health System for the 21st Century (Washington, D.C.: National Academy Press, 2001).

[8] See Mokwunye NO., “A case study of cost, mortality, and quality of life outcomes of inappropriate and appropriate admissions to a medical intensive care unit” Dissertation Abstract International, 69 (4A) (2008); Smith TJ, Coyne P, Penberthy L, Hager A, Hopson MA, “A high-volume specialist palliative care unit and team may reduce in-hospital end-of-life care costs,” Journal of Pallliative Medicine, 2003: 699-705; Cohen SR, “Changes in quality of life following admission to palliative care units,” Palliative Medicine, 2001: 363-371; Goldney R, Bowes J, Spence N, Czechowicz A, Hurley R, “The psychiatric intensive care unit,” Br. J. Psychiatry 1985: 146:5054; and Winkler D, Scharfetter J, Kasper S, Frey R. “Ethical Issues and Intensive Care in Psychiatry” Journal of Hospital Ethics, in press.

[9] T.S. Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago Press, 1962).

Tags: ,

Comments on this article

By clicking and submitting a comment I acknowledge the Science Progress Privacy Policy and agree to the Science Progress Terms of Use. I understand that my comments are also being governed by Facebook's Terms of Use and Privacy Policy.