December 1995 XVII/4
LESSONS IN CARE OF THE DYING
An article in the November 22nd issue of JAMA reports the results of a major, multi-center clinical study (1) undertaken to change the way care is rendered to persons with life-threatening illnesses. Phase I of the study documented the need for such change by showing that seriously ill patients were frequently over-treated, experienced moderate to severe unrelieved pain, and that expressed patient preferences about end-of-life treatment often were disregarded by physicians in the decision-making process. Phase II of the study proposed to alter physician behavior in order to 1) reduce the technologic assault on dying patients; 2) reduce the incidence of undesirable states such as coma; 3) insure adequate pain management; 4) shorten length of ICU days; and 5) promote appropriate resource use. The change was to come about by an intervention which enhanced physicians' understanding of patient preferences primarily through nurse-mediated communication and gave physicians more timely and accurate prognostic information upon which to base treatment decisions. Unfortunately, and perhaps predictably, the intervention had no impact on the manner in which dying patients were treated by physicians in the study. After the intervention, patients were still over-treated and continued to experience unrelieved pain and physicians persisted in ignoring stated patient preferences.
In an editorial in the same issue of JAMA, (2) physician Bernard Lo reflects on possible reasons that the SUPPORT study intervention did not change physician behavior. For example, he notes that although physicians in Phase I asserted that more accurate, computer-generated prognostic data would make a difference in the care of seriously ill patients, the availability of such data made no difference at all. Lo suggests the reason is that physicians' subjective prognostications are as reliable as those produced by computers. Lo proposes further that the study intervention may have been unsuccessful because of the inadequacies of communication between physician and patient. He suggests that the study may not have taken into account some of the significant barriers to improved communication. Perhaps, he muses, physicians do not really believe their patients. Perhaps physicians are convinced that they know best regardless of patient input. Perhaps physicians would prefer to communicate first-hand with patients rather than rely on a nurse go-between, especially when addressing issues of such import.
Finally, Dr. Lo suggests that patients' pain was treated no better after the intervention was made than before because "no component of the SUPPORT intervention directly addressed the problem of inadequate pain control.
Lo offers some suggestions on how to insure that future studies do promote more appropriate care for critically ill persons. Do not simply give physicians more information and expect a behavior change. Change the culture in which they practice. Offer conferences on end-of-life decision making and pain control, conduct case management meetings, give feedback to physicians and offer "recognition for clinicians who provide outstanding care at the end of life." When all else fails, reduce the number of ICU beds available. Fine physicians for over- treatment and provide incentives for those who can be persuaded to give appropriate care only by appealing to self-interest. Lo concludes: "it would be simple if computers and nurses alone could improve the care of seriously ill patients. Instead, physicians (emphasis added) will need to change hospital culture and practices and our own behavior.
Questions Raised
Dr. Lo's comments raise some interesting questions. First, physicians can predict at least as well as computers, the likelihood of meaningful survival for patients with life-threatening illnesses. Yet, even when physicians know that the chances of meaningful survival are minimal, many frequently fail to act on this knowledge. Wh~ Is it because that even with the best prognostic information available there remains some margin for error, however slight? Is it because, when faced with the possibility of death physicians demand absolute certitude regardless of what this means for patients and their families? Second, physicians may indeed harbor suspicions about the ability of patients to understand the complex issues involved in treatment decisions in serious illness. Physicians truly may wish to spend adequate time helping patients understand and participate more fully in the decision making process. But many physicians fail to do so. Why/! Is it because they simply do not have the time, as many claim? Or is it because they do not take the time because they fear honest discussion with patients and families about topics they find troubling? Third, physicians may want to attend more appropriately to the pain experienced by critically ill patients. However, while there is a wealth of information available in the medical literature and elsewhere on both the theory and mechanics of good pain management many physicians do not make use of this information. Physicians routinely up-date their knowledge and skills in other, often more complex areas of endeavor. Why do so many resist becoming adept at controlling pain in critically ill and dying patients? Could it be that beyond the pain is a suffering person who, when relieved of the physical pain, may raise questions of meaning that some physicians find disquieting? Fourth, appropriate physician behavior with regard to the treatment of persons with life-threatening illness does require an organizational culture supportive of such behavior. Many health care organizations around the country strive to create and sustain such a culture by doing just as Dr. Lo suggests. They offer programs on end-of-life issues. They support case conferences which are initiated routinely in acute care settings often and only by nursing personnel. However, many physicians refuse to participate in either and thus frustrate the creation of a culture in which appropriate care is possible. Why/! Is it, again as Dr. Lo suggests, that physicians interpret such programs and conferences as threats to their autonomy and self-esteem? Could it be that the kind of knowledge offered at such programs and conferences raises issues that some physicians are unwilling or unable to deal with?
Dr. Lo's final suggestions are quite troubling. Why would anyone who is a member of the medical profession propose as reasonable that appropriate care might be achieved only by means which would undermine the very meaning of the profession and the integrity of individual physicians? Could it be that some physicians will yield to the dying of their patients only when coerced to do so by financial harm or gain?
Conclusion
The most significant outcome of the SUPPORT study seems to be data that was known well before the study was undertaken. That is, many physicians do not deal well with the possibility or reality of death. Because of this, critically ill persons and their loved ones often suffer unnecessarily as death approaches. The increasing number of persons advocating the legalization of physician assisted suicide attests to this. Dr. Lo and many others seem to suggest that the way to address this persistent problem is by more and better constructed studies. However, future studies will be helpful only when some of the fundamental impediments to appropriate care of critically ill and dying persons are confronted. Here Dr. Lo is certainly correct. This is work that physicians themselves must do. It is simply not acceptable to continue to blame the environment, other health care professionals, the limitations of patients and families or the constraints of time or regulations for the inadequate care that so many critically ill and dying persons experience. Physicians must address their own beliefs, anxieties and fears about personal mortality before they can care appropriately for patients who face their own dying. Physicians must realize that dying is a personal, communal and spiritual process, not simply a medical event. They must realize that the dying process can be aided by good medicine but that it can be perverted by medicine which seeks only to master that which is ultimately beyond human control. In other words, physicians who seek to care well for the dying must learn to befriend death, to be as open to death as they are to life.
Kevin O'Rourke, OP
Footnotes 1. SUPPORT Principal Investigators. "A Controlled Trial To Improve Care for Seriously III Hospitalized Patients: The Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments." JAMA Nov. 22/29, 1995.
2. Bernard Lo, "improving Care Near the End of Life: Why Is It So Hard? JAMA Nov. 22/29, 1995.
| INDEX | © Kevin O'Rourke, O.P.