| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||
Correspondence: Address correspondence to Howard Tuch, MD, CMD, Director of Palliative Care Resources, Associate Director of the Hospice of the Florida Suncoast, 3324 Westmoreland Drive, Tampa, FL 33618. E-mail: hstuch{at}msn.com
Nearly 20 years ago, when I began a residency in internal medicine and geriatrics, our hospitals and intensive care units were filled with dying nursing home patients. These patients were uniformly dreaded by the interns and residents. They were usually admitted to the hospital with a new catastrophic clinical event superimposed on a background of multiple chronic illnesses. They came with bucket loads of medications to sort through and suffered severe cognitive and functional impairments. Do not resuscitate orders were still unusual; orders to withhold or remove a feeding tube or ventilator support were scandalous. Senior physicians would proclaim that failing to offer aggressive support was tantamount to euthanasia. Although patients were dying all around us, physicians managed never to talk about death, except in the context of a failure to respond to medical intervention. Indeed, failing to respond to medical intervention was the only way anyone could die in physicians' hands.
Well, dying in health care has changed. Physicians now routinely make decisions to limit care. They are required to ask about advance directives and respect them when present. Older patients, especially those aged older than 85 years are typically managed less aggressively (Hogan, Lunney, Gabel, & Lynn, 2001; Lubitz, Beebe, & Baker, 1995), and physicians are much more likely to focus on issues of quality of life when making decisions. Patients from nursing homes are less commonly transferred to the hospital when death appears imminent. Yet, despite much progress, dying in the U.S. health care system remains a difficult grace. Unrelieved pain, failure to follow advance directives, invasive interventions, isolated deaths, and inadequate bereavement support remain far too common (Teno et al., 1997).
When asked, Americans say that they want to die at home, in comfort, and surrounded by their loved ones. Americans also say that they would rather die than enter a nursing home. Chances are good that they will be disappointed on both counts. Nearly 20% (and growing) of Americans now die in nursing homes. Already, in some states and in older age groups, closer to 40% of deaths now occur in a nursing home (Brown Atlas on Dying, 2000). Health care professionals know very little of what it is like to die in this setting. They know even less about how or why treatment decisions are made in the nursing home. They do know that pain management, advance care planning, bereavement support, and involvement of Hospice are not used adequately (Miller, Gozalo, & Mor, 2000).
It is within this context that health care professionals must appreciate the article "Lower Respiratory Infections in Nursing Home Residents With Dementia: A Tale of Two Countries," by Mehr and colleagues (2003). The authors study prospective, nonrandomized cohorts of nursing home patients in the Netherlands (706 patients living in 61 dementia-specific nursing homes with clinical diagnoses of pneumonia) and in the United States (701 dementia patients residing in 36 nursing homes in central Missouri with evidence of lower respiratory tract infections). The results suggest two very different approaches to patients with end-stage dementia and may provide some insight into the cultural determinants of medical decisions at the end of life.
In the Netherlands, patients with severe dementia were rarely tube fed or sent to the hospital. Withholding of antibiotics in dementia patients with pneumonia was more common than in the U.S. homes studied (23.4% vs. 12.3%). When they were treated, 90% of the patients in the Netherlands received oral antibiotics. The 23.4% of the residents who were not treated had more severe dementia (80% of the severely demented patients were not treated with antibiotics), had more functional impairment, and seemed more acutely ill on clinical exam (a larger number had signs and symptoms of pneumonia, presence of dehydration, or likelihood of aspiration as a cause of the pneumonia). Mortality within 30 days among those not treated with antibiotics was 90.2% compared to 29.8% of those treated.
In the United States, by contrast, tube feeding (12.3%) and hospitalization (26%) were significantly higher. Treatment with antibiotics was both more common and more invasive. Residents not treated with antibiotics (14.7%) were more severely demented (53% of severely demented patients were not treated). Those with the clearest evidence of severe pneumonia (high temperature, high white blood cell count, positive chest x-ray) were more likely to be treated in the United States, but not in the Netherlands. Remarkably, there was no difference in mortality between those treated with antibiotics and those not treated in the U.S. cohort (16.7% and 17.5%, respectively).
There are important limitations to the study that make comparing the two populations difficult. In the Netherlands, patients were residents of psychogeriatric units spread throughout the country and were probably more severely demented. Diagnosis of pneumonia in the Netherlands was based on clinical evaluations by the primary physicians whereas in the United States diagnosis of lower respiratory tract infection relied on defined criteria (and usually chest x-rays and laboratory values). The Dutch cohort may also have been more acutely ill than in the United States. The differences seen in mortality as well as nontreatment decisions were likely magnified because of the population differences.
Mehr and colleagues (2003) suggest that the study provides insight into the difficulty we have in transitioning to a palliative care approach in the United States. Others have suggested that the real problem with end-of-life care in this country is the difficulty health professionals have in identifying who is really dying. When in the course of illness does a gravely ill patient turn into a dying patient (Finucane, 1999)? In the United States, physicians are more likely to offer treatment when the evidence for an acute infection is strong even in the setting of severe dementia. Physicians seem willing to concede that although dying with severe dementia is acceptable, dying from something potentially treatable is not. I wonder what kind of a death these patients will experience if physicians insist that death from pneumonia will not occur.
I think that many patients and families go through multiple transitions at the end of life, accommodating to ever-increasing severity of illness. Hospice patients are frequently treated with antibiotics as families first come to terms with the idea that aggressive treatment is not beneficial, then maybe that hospitalization, hydration, or nutritional intervention may not be indicated. Withholding antibiotics may fall late in the course of such decision making. Maybe in a more ideal system of care, health professionals would not see such a discrete transition to palliation but rather a mixture of curative and palliative approaches all along the disease trajectory (Lynn, 2000). The relative proportion of palliative care would increase gradually as death approaches. The questions raised by this study are important and timely. As those dying with dementia fill America's nursing homes, health professionals need to ask what kind of care these patients and their families want and deserve. Health professionals should know how and why decisions to limit care are made. Perhaps even more important, however, they need to have a long overdue discussion of what kind of place their parents and themselves will likely go to live, and to die over the next several decades.
 |
Footnotes |
|---|
Decision Editor: Laurence G. Branch,, PhD
Received for publication October 28, 2002. Accepted for publication October 28, 2002.
|
References |
|---|
 TOP References |
|---|
| ||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
