Estimating End of Life for Hospitalized Patients

Vetted prediction tools can help estimate end-of-life, but clear and empathetic discussions with dying patients and their families are more important.

by Susan Kreimer

End-of-life conversations are common in hospital medicine, and Caitlin Foxley, MD, FHM, is no stranger to their nuance. She offers patients and loved ones as much factual information as she can. And regardless of their preference—aggressive treatment, comfort care, something in between—it’s ultimately their choice, not hers. But no matter what, she will ensure the patient’s pain remains under control.

“The way I practice is to allow my patients to make the end-of-life decision that is in accordance with their wishes, and not simply push the least expensive one on them,” says Dr. Foxley, medical director of IMI Hospitalists and hospital service chief of internal medicine at Nebraska Medical Center in Omaha. However, she adds, “most people, given accurate information in a compassionate manner, would choose to die at home, and not in an ICU on a ventilator, with chemo and pressers going through a central line.”

Although hospitalists differ in their approaches to end-of-life discussions, most agree that the majority of critically ill patients want to know their prognosis. Tested end-of-life prediction tools can help physicians provide realistic ranges for patients and families (see “Helpful End-of-Life Prediction Tools,” p. 39). Armed with this insight, they can hope to deliver better and more cost-effective end-of-life care.

Nonetheless, “we cannot rely solely on a tool to make decisions,” says Alvin H. Moss, MD, FAAHPM, professor of medicine in the nephrology section and director of the Center for Health Ethics and Law at West Virginia University School of Medicine in Morgantown. “The tool is a decision aid.”

Clinicians still need to help patients and families identify their treatment goals while determining which life-sustaining options they would or wouldn’t want to pursue, Dr. Moss says. That conversation would include an estimated prognosis of survival.

“If you try to prognosticate a specific length of time, you will be wrong,” says Steven Z. Pantilat, MD, FACP, SFHM, professor of clinical medicine and director of the palliative care program at the University of California San Francisco Medical Center. “You can give patients a lot of useful information by speaking in ranges.”

But it’s important to also convey the inherent uncertainty of any prognosis, considering that a very sick patient might suffer a sudden decline. For this reason, even the best prognostic indicators aren’t exact, Dr. Pantilat cautions. A prediction tool could forecast a 20% chance of six-month survival on the day before a patient’s death in the ICU.

“Predictions really apply to groups of people, not individuals,” says J. Randall Curtis, MD, MPH, professor and section head in the division of pulmonary and critical care medicine at the University of Washington’s Harborview Medical Center in Seattle. Physicians can’t possibly know whether someone will fall into the 95% of patients who die or the 5% of patients who beat the odds.

“It’s never certain that a patient is not going to survive,” says Dr. Curtis, who is director of the Harborview/University of Washington End-of-Life Care Research Program. While patients are less likely to request aggressive care in light of a poor prognosis, some will elect intensive treatment in hopes of defying even the grimmest statistics.

More Medical Tests and Procedures

In the U.S., it’s much more common for patients to receive life-saving treatments than in other countries. The expectation is that expensive medical technology can always prolong life.1

“A lot of patients have that mentality,” says Ann Sheehy, MD, MS, director of the hospitalist service at University of Wisconsin Medical Center in Madison. “That makes it harder to have the discussion with patients that there isn’t something else we can do.” Patients feel “that you’re almost abandoning them, that there’s something you’re withholding.”

The widespread assumption that more medical tests and procedures lead to better outcomes goes hand in hand with the misperception that sufficiently controlling pain and other symptoms draws death closer. As a result, many patients end up dying with distressing symptoms in the hospital instead of peacefully at home.1

As physicians, Dr. Sheehy points out, “We don’t do a good job of saying, ‘This care probably is not going to help you or that it will leave you with a very bad quality of life in the end.’” But projections are far from perfect.

“Telling patients there is a 20% chance that they might die in the next year isn’t usually enough to change their thinking. Nor is it enough to justify withholding treatment,” says David Casarett, MD, MA, associate professor of medicine at the University of Pennsylvania and chief medical officer of Penn-Wissahickon Hospice in Philadelphia.

What prognostic information can do is play an important part in guiding appropriate screening and preventive health measures. For example, if a male patient has a 50% chance of dying within four years, it doesn’t make sense to screen for prostate cancer, a slow-growing malignancy that often takes years to develop. This protocol may also apply to cancer screenings, as well as treatments for diabetes, high blood pressure, or high cholesterol, Dr. Casarett says.

End-of-Life Conversations

By not taking a patient’s prognosis into account, many clinical decisions are not fully informed. In physicians’ clinical practice and training, there tends to be less emphasis on estimating prognosis than on diagnosing and treating illness. This is particularly significant in older adults with competing chronic conditions and diminished life expectancy.2

“Many physicians have not been trained in how to have these conversations, which is something we’re trying to change,” says Dr. Curtis, the pulmonary and critical care specialist at the University of Washington. “This is very emotionally difficult for patients and families, and therefore, it can also be emotionally difficult for physicians.”

Starting this summer, medical residents at Cooper University Hospital in Camden, N.J., will have mandated exposure to palliative care. The rotation, lasting from two to four weeks, will occur in their second year of training, says Mark Angelo, MD, FACP, director of palliative medicine.

Residents will accompany the palliative-care team for an intensive period of time to learn about different techniques for symptom control and to observe end-of-life conversations, which often elicit patients’ angst, depression, and physiologic and spiritual unrest.

“We already have residents rotating with us now in the palliative-care program, and everybody is very struck by how honest patients want you to be,” Dr. Angelo says. “It certainly is devastating, and we understand that. But it does give patients the opportunity to reorganize their lives and to prioritize a number of issues, including medical treatment.”

Some patients opt to spend more time with their grandchildren rather than stay in the hospital. Others prefer to eat and drink whatever they want. Many terminally ill Hispanic patients in the Camden area travel to Latin America, where they were born, or they invite relatives to visit them here, Dr. Angelo says.

While it’s difficult to accept finality, “there’s a certain amount of freedom that comes with that,” he says. “If someone has a prognosis of two months, they may make very different decisions than someone who has a prognosis of two years.”

Physicians tend to be overly optimistic, notes JoAnn Wood, MD, MSEd, MHA, a hospitalist and division director of general internal medicine at the University of Arkansas for Medical Sciences in Little Rock. “The data suggest that, if you ask a physician to provide you with his or her impression of a patient progression, we generally tend to overestimate how well they’re doing,” particularly with cancer patients, Dr. Wood says. “Physicians don’t choose this line of work to facilitate people’s dying.”

Growing evidence indicates that treating a patient’s discomfort is linked to improvement in physical status and might even increase survival.

When Doing Less Is More

End-of-life prediction tools enable clinicians to keep their expectations of a patient’s survival grounded in reality. And in many hospitals, palliative-care providers are available to lend their expertise. “The field of palliative medicine has taken a lot of strides,” Dr. Wood says, “in helping us to seeing that dying is something that can be done well, just like living can be done well.”

In fact, growing evidence indicates that treating a patient’s discomfort is linked to improvement in physical status and might even increase survival.3 What this means is that, at the end of life, sometimes doing less is actually more. And if patients request less medical care, physicians should honor those desires.4,5

Dr. Pantilat, the UCSF palliative-care expert, suggests asking open-ended questions to get at the heart of a patient’s wishes. For instance: “‘When you look to the future, what do you hope will happen?’ Or ‘When you think of life ahead, what worries you the most?’” The responses guide physicians in devising a plan of care that is consistent with a patient’s values. Having these discussions sooner rather than later is best for everyone involved, Dr. Pantilat says. Unfortunately, that’s not what usually happens.

About half of the more than 40% of Americans who die annually under hospice care do so within two weeks after being admitted. In such a short timeframe, even the most skilled experts are limited in what they can do, according to the National Hospice and Palliative Care Organization. To reverse this trend, Dr. Pantilat recommends that physicians consider making referrals to palliative care much earlier, whenever they sense that a patient may have a year or less to live.

Providing palliative care can be done in conjunction with life-prolonging therapies or as the central focus. The goal is to prevent and relieve suffering and to ensure the best possible quality of life for patients and their families, regardless of disease stage or the need for other treatments. Such care is suitable for patients with cancer, heart conditions, liver or renal failure, Alzheimer’s disease, spinal cord injuries, and a number of other illnesses, according to the National Consensus Project for Quality Palliative Care.

We have a lot of work to do to educate the public about the dying process. We are all going to die someday, and we all need to think about how and where we want to be when it happens, if we have any choice in it.

—Caitlin Foxley, MD, FHM, medical director, IMI Hospitalists, hospital service chief of internal medicine, Nebraska Medical Center, Omaha

Planning Ahead

In Arkansas, Dr. Wood typically turns to the hospital’s palliative-care team in end-of-life situations, asking its members to join in a conference with the patient and family. The team typically consists of a nurse, social worker and chaplain who can address various concerns. An employee from the medical billing department participates if necessary.

“Physicians should never assume that they understand the healthcare system, because it’s incredibly complicated,” Dr. Wood says, adding that she doesn’t pretend to be an expert in Medicare, Medicaid, or private insurance reimbursement issues.

Helping patients with advanced care planning can minimize difficulties later. Most patients who die in hospitals are admitted with end-stage disease, and most spend time in the ICU with mechanical ventilation. Physicians often are unaware of patients’ preferences, and this could lead to misunderstandings, especially in the ICU setting, where prognoses can shift quickly. One study showed that, in order for some of these patients to die, clinicians and families generally had to decide explicitly to strive toward less-than-completely-aggressive care.1

“We’re trying to make sure that patients and families have an opportunity to do advanced-care planning,” Dr. Curtis says. Talking with them about their values and goals is essential for clinicians to understand their preferences.

Part of this involves crafting advanced-care directives. One such directive would be a power of attorney for healthcare, in which a patient states who is authorized to make decisions if he or she becomes unable to do so. This is particularly important for patients who select someone other than whom their state’s law would normally designate.

Another document—the living will—allows patients to specify their own preferences for end-of-life care. Yet “it’s often very difficult to know exactly what decisions will need to be made,” Dr. Curtis says. “Those documents are rarely determinative.” Even when a patient stipulates his or her wishes against “extraordinary life-sustaining measures, it still leaves a lot for interpretation.”

Some patients may spell out more clearly whether they wouldn’t want tube-feeding, CPR, or ventilation. This can be prescribed in written and signed Physician Orders for Life-Sustaining Treatment (POLST).

If you try to prognosticate a specific length of time, you will be wrong. You can give patients a lot of useful information by speaking in ranges.

—Steven Z. Pantilat, MD, FACP, professor of clinical medicine, director of the palliative-care program, University of California San Francisco Medical Center

Educating the Public

End-of-life discussions also pose a threat of litigation. “It takes a significant amount of time, often during a very busy day, to sit down with a patient and family members to bring up an issue that will undoubtedly raise many questions, some of which are impossible to answer,” says Dr. Foxley, the hospitalist service chief in Omaha. “I’m sure many physicians are uncomfortable with the tears that are shed.”

When Dr. Foxley recently advised a patient’s family that aggressive care would be futile, they directed their anger toward her. Their loved one died, despite the intensive treatment. It’s just one example of many in which Dr. Foxley has witnessed how high-tech medical treatments can incur astronomical hospital bills after just a few days while doing little—if anything—helpful for the patient.

“We have a lot of work to do to educate the public about the dying process,” she says, adding that the entire burden shouldn’t fall on physicians, and that patients should inform family members of their end-of-life wishes. “We are all going to die someday, and we all need to think about how and where we want to be when it happens, if we have any choice in it.”

Susan Kreimer is a freelance medical writer based in New York.


APACHE II (Acute Physiology and Chronic Health Evaluation II)

A severity-of-disease classification system and one of several ICU scoring systems, it is applied within 24 hours of patient admission. An integer score from 0 to 71 is computed based on several measurements; higher scores correspond to more severe disease and a higher risk of death.

SOFA (Sequential Organ Failure Assessment)

SOFA tracks a patient’s status during an ICU stay. This scoring system determines the extent of a person’s organ function or rate of failure. The overall score is based on scores of the respiratory, cardiovascular, hepatic, coagulation, renal, and neurological systems.

The Karnofsky Performance Scale Index

This tool allows patients to be classified by functional impairment. It compares effectiveness of different therapies and assesses the prognosis in individual patients. The lower the Karnofsky score, the worse the survival outlook for most serious illnesses.

PPS (Palliative Performance Scale) 

First introduced in 1996 as a tool for performance status in palliative care, the PPS uses five observer-rated domains correlated to the Karnofsky index (100-0). PPS is used in many countries and is translated into many languages.

The BODE Index

BODE helps predict mortality from chronic obstructive pulmonary disease (COPD) after diagnosis. Presumably, a higher BODE score correlates with an increased risk of death.

The Charlson Comorbidity Index

This index predicts the 10-year mortality for a patient who might have a range of comorbid conditions, such as heart disease, AIDS, or cancer (a total of 22 conditions). Each condition is assigned a score, depending on the risk of dying associated with the condition.

Hemodialysis Mortality Predictor

This online calculator estimates prognosis in end-stage renal disease patients using an integrated model that incorporates the patient age, serum albumin, comorbidities, and clinician assessment of the patient’s likelihood of being dead within a year.

—Susan Kreimer

Source: The Hospitalist, August 2012


Subscribe To Our Newsletter

Sign up for our newsletter and receive a FREE emergency guide for your pets.

You have Successfully Subscribed!

Submit a Comment

Your email address will not be published. Required fields are marked *

Pin It on Pinterest

Share This

Share this post with your friends!

Share This

Share this post with your friends!

Password Reset
Please enter your e-mail address. You will receive a new password via e-mail.