Poor Management of the End of Life

Not only are patients' rights violated in this last act of life, money is wasted. Indeed, at least a quarter of Medicare spending is for the last year of life(1). It is not necessarily a bad thing that so much money is spent in the last year of life, but how it is spent is another matter.

The manner of Ed Fitzsimmons's death, the fictitious patient from the section on the erosion of primary care, illustrates the point. Shortly after Ed Fitzsimmons suffered his heart attack and was taken from his long sojourn in the waiting room of the ER, his breathing became so rapid that a tube was put in his airway, a ventilator attached, and he was rushed to the cardiac intensive care unit (ICU). Elaine Fitzsimmons suspected then that her husband was dying. The doctors in the ICU took three-day rotations. Ed Fitzsimmons's first physician, Dr. Parsons, was a fastidious person, so precise in speech, appearance, and bearing that even his expressions of concern seemed studied. He attended, even fussed, over every detail of management.

Ed Fitzsimmons had specifically told Elaine after his first heart attack that he did not want intensive care if he could not come out of it with high odds of being able to walk out of doors and living without assistance. He had filled out a standard advance directive to be sure his wishes were met. It said, "If at any time I should have an incurable condition caused by injury, disease, or illness certified to be a terminal condition by two physicians, and if the applications of life-sustaining procedures would serve only to artificially postpone the moment of my death, I direct that those procedures be withheld or withdrawn and that I be permitted to die naturally." He had assumed that this said to doctors what he meant.

Unfortunately, advance directives are not specific documents and do not provide doctors with precise information. No reasonable person would wish to be kept alive if they were inevitably to die during a hospitalization, so the language in Ed Fitzsimmons's advance directive only told a physician what was already inherently understood. But left undefined by advance directives are the terms "prospect of recovery" and "hopeless". One person might want everything done, no matter the cost and consequences to his family, for a slight chance of high-functioning recovery, when the downside was a very high chance of living in dependence, severely disabled. Another might view those odds with more horror than a certain death. Much depends upon how a patient views his life before the hospitalization and what he is willing to accept after it.

Doctors mostly rely upon their own judgment and the wishes of the family in deciding the practice of life assistance or its withdrawal. And that judgment also depends upon the doctor's views of life and death, but also upon their habits of practice. Some doctors think about such matters and make careful individual judgments, but many simply apply technology by reflex. And Ed Fitzsimmons's physician, Dr. Parsons, would not let anyone die on his service if he could prevent it.

Under sedation and with tubes projecting from his mouth, Ed was mute. Nurses pushed his bed down to x-ray day after day for new studies; more tubes each day exiting his arms, mouth, penis, and chest. On the sixth day of Ed Fitzsimmons's hospitalization, his fingers and toes had taken a blue tinge and the family requested a meeting with Dr. Parsons. "Well," Dr. Parsons began, "He is not doing very well today. His heart failure is worse. We're using high doses of dobutamine for blood pressure support. He has some peripheral ischemia. This is why his fingers and toes are cyanotic."

"Dr. Parsons," Elaine interrupted, "Do you really think that my husband can survive this?"

"Well," Parsons responded, "He is very sick but we have to keep working to turn his heart around."

"But what are his chances, Doctor? We don't know if he would want this," interjected the Fitzsimmons's oldest grandchild.

Dr. Parsons's response was, "His chances are not very good, but there is always hope. He could turn around."

Just as Parsons applied his skills without much thought for the consequences, he spoke with families without the ability to imagine himself in their place. Doctors vary in their ability to make value judgments. Most doctors get better at such value judgments as they gain more life experience but Parsons was not among them. His statements were technically true, if sometimes undecipherable. Ed Fitzsimmons's condition at this time statistically yielded a probable survival rate of five percent. If he lived, his remaining cardiac function would permit only the exertion that walking around the house at a slow pace would afford. Parsons's view was that this constituted hope, not to mention an opportunity to use his considerable abilities. The family did not know to ask the explicit detail that would have allowed an informed decision, and Dr. Parsons did not volunteer it. But they suspected that Ed was dying, no matter what the doctors might do.

Elaine again took charge of the situation. "Let us talk about it among ourselves, Dr. Parsons."

After Parsons left, the family conferred with Ed's nurse who attended the session. She told Elaine confidentially that Dr. Parsons never took anyone off life-support and that she was correct in the judgment that her husband was dying. The nurse explained the process of withdrawal of care and told them that he would probably live only a short time without the ventilator and the drugs, but he would not be in pain or suffer. The family spent two days comparing their observations of Ed's physiologic state as they stared at the monitors and made mental notes during visiting hours, notes of his heart rate, his blood pressure, how he was responding to the medications. Each day brought worse numbers, and after a resigned family meeting and reassurances from his nurse that he was dying, the family simply asked the nurse to convey to Dr. Parsons their wish that the breathing tube be removed and the medications stopped.

The scene was common in the ICU's small pods with the family transfixed around the bed in suspense that each un-natural gasp would be the last. The older grandchildren were sobbing, having never imagined anything like this. The adults worked to bring dignity to Ed's death by retaining their composure. Finally Elaine whispered to him, "Ed, let go, it's all right," after which he died.

Ed Fitzsimmons, against his wishes, had expended over $50,000 in medical costs during the ten days of his ICU stay and had undergone a series of fruitless medical interventions that he did not want, but was powerless to deny.(2) From the moment his heart failure became so bad that he required a ventilator and his blood pressure so low that medications were required to support it, his outcome was certain.

Ed Fitzsimmons's stay in the ICU had consumed enough resources to provide health insurance for a year to thirty-three uninsured children. Had he been able to discuss his options with his physician in an office setting before his heart attack, the doctor would have told him that his heart disease was so severe that the likelihood of his dying if he had another heart attack or needed a ventilator was ninety-five percent and with severely limited independence if he should survive. Ed would have responded that he wanted everything done up to the point of being put on a ventilator, but no more. The family endured ten days of gnawing anxiety and powerlessness that robbed their grief of peace and left unresolved doubt in the family. Whether they had done the right thing was a topic of family discussion for the next year.

People who talk about this problem believe that an unreasonable public demands excessive care in the face of hopeless injury, and yet the medical literature indicates that this happens in only about four percent of cases, and my experience is the same.(3) I only remember three or four cases out of maybe a thousand that I managed in which a family insisted that we continue intensive care on someone who was hopelessly damaged. It was usual for a family to want a day or two to think about what they had been told or to wait for a relative to arrive. In most cases the family would make the decision to withdraw care within hours, particularly in older patients.

A study called SUPPORT dispelled any illusions that money could be saved and patient autonomy preserved by encouraging doctors in hospitals to use intensive care more judiciously. By the early 1990s researchers had made the observation that death rates for the same conditions varied widely among US hospitals, and this puzzled them. The corollary was that no one knew what the death rate for various conditions ought to be. The lack of measures that allowed prediction of terminal illnesses made it impossible to determine which hospitals were doing the better job managing patients. Permitting patients who were inevitably going to die to do so in the hospital rather than later in a nursing home after much care might be in accord with the patient's wishes and be better medicine, but the practice would increase a hospital's mortality rate. In many conditions a high mortality rate is bad; in the case of an inevitably terminal illness it is not.

For two years, eight investigators funded by the Robert Wood Johnson Foundation studied the outcomes and interviewed the families of 4,301 patients in advanced stages of nine diseases who were admitted to one of five large hospitals in different metropolitan areas. When someone's heart or breathing fails in a hospital, the nurses must instantly "call a code" unless the patient has "Do Not Resuscitate"(DNR) orders. When a code is called, a system is initiated that brings anesthesiologists, cardiologists, EKG technicians, radiology technicians, and any nearby medical personnel running to the patient to conduct cardiopulmonary resuscitation (CPR). It is a frightening scene; the bed is completely enclosed by a white-coated horde thumping the chest, drawing blood, and inserting tubes. The SUPPORT investigators found that doctors wrote "Do Not Resuscitate" orders in only half of the patients who did not want cardiopulmonary resuscitation if their heart or breathing stopped. Half of the patients' doctors did not know the patient's preference. SUPPORT found that thirty-eight percent of the patients who died spent at least ten days in ICU and that dying patients were in pain at least half of that time. (4)

The investigators then tested the effect of a program of educating doctors, patients, and families in order to improve management of life-sustaining treatments. Nearly five thousand patients in the same hospitals were divided into two groups--control and intervention. For the intervention group, the investigators provided the doctors with a statistical prediction of whether his or her patient would live in the event that intensive care was required. The investigators documented patients' and families' preferences on care and also educated families as to prognosis and treatment options. Nurses working for the study organized these conversations. The control group was managed in the usual way without the extra educational activities. Somewhat to everyone's amazement, the study found no difference at all between the groups in the number of patients that received unwanted cardiopulmonary resuscitation and intensive care and no difference from the original observational study. (5)

Of 479 patients in SUPPORT that lived through their first hospitalization and then died within six months, eighty-one percent wanted to die at home and only about half got their wish.(6) SUPPORT's extensive on-site and real-time physician education had no effect whatsoever on physician behavior towards patients at the end of life. Reading between the lines of the scientific papers, there was anguish. I think the investigators did not want to believe the results they reported but knew them to be accurate and thus had to report them.

When the investigators at the Dartmouth Institute for Health Policy and Clinical Practice examined regional differences in Medicare spending for patients who were hospitalized with the diagnoses of hip fracture, heart attack, or colorectal cancer, they found that in regions with high bed counts, fifty-four percent of patients died in the hospital. In regions with lower bed counts, twenty-three percent died in the hospital. Personal preference did not influence the place of death at all. (7)Similar variation in end-of-life care is found even among the nation's top hospitals.(8)(9)Patients' dignity would be respected and a lot of money saved if people's wishes at the end of life were communicated accurately and if the members of the health care system respected them.

1. Christopher Hogan, June Lunney, Jon Gabel, et al., "Medicare Beneficiaries' Costs of Care in the Last Year of Life,” Health Affairs 20, no. 4 (July/August 2001): 188-195.
2. Matthew R. Williams, Rachel B. Wellner, Elizabeth A. Hartnett, et al., "Long-Term Survival and Quality of Life in Cardiac Surgical Patients with Prolonged Intensive Care Unit Length of Stay,” Annals of Thoracic Surgery 73, no. 5 (May 2002): 1472-1478.
3. Thomas J. Prendergast and John M. Luce, "Increasing Incidence of Withholding and Withdrawal of Life Support from the Critically Ill,” American Journal of Respiratory and Critical Care Medicine 155, no. 1 January 1997: 15-20.
4. SUPPORT Principal Investigators, "A controlled trial to improve care for seriously ill hospitalized patients. The study to understand prognoses and preferences for outcomes and risks of treatments (SUPPORT),” JAMA 274, no. 20 (November 22, 1995): 1591-1598.
5. Ibid.
6. Robert S. Pritchard, Elliott S. Fisher ES, Joan M. Teno, et al., "Influence of Patient Preferences and Local Health System Characteristics on the Place of Death. SUPPORT Investigators. Study to Understand Prognoses and Preferences for Risks and Outcomes of Treatment,” Journal of American Geriatric Society 46, no. 10 (October 1998): 1242-1250.
7. Ibid.
8. Thomas J. Prendergast, Michael T. Claessens, and John M. Luce, "A National Survey of End of Life Care for Critically Ill Patients,” American Journal of Respiratory and Critical Care Medicine158, no. 4(October 1998): 1163-1167.
9. John E. Wennberg, Elliott S. Fisher, Therese A. Stukel, et al., "Use of Hospitals, Physician Visits, and Hospice Care During Last Six Months of Life Among Cohorts Loyal to Highly Respected Hospitals in the United States,” BMJ 328, no. 7440 (March 13, 2004): 607.