Benefits of Ethics Consultation When Life-Sustaining Treatments are Unlikely to be Beneficial 

Abstract & Commentary

In this randomized control trial, Schneiderman and Associates evaluated the effect of an intervention, namely, an ethics consultation, on several variables associated with the care of patients eventually dying in the ICU. They hypothesized that in critically ill patients, when a conflict arises as to the nature and duration of life-saving therapy, an ethics consultation would reduce the number of days such patients spend in ICU and in the hospital and also the number of days such therapy is continued. They further hypothesized that such intervention would not increase the mortality of the group that received the ethics consultation.

To test their hypothesis, they enrolled 551 patients in 7 diverse ICUs across the country. Nurses routinely rounding in the ICU identified eligible patients. Included were adult patients in whom "value-laden treatment conflicts" were imminent or manifest that could lead to incompatible courses of action. The investigators had established that such conflicts could arise in 6 circumstances:

1. within the health care team as to whether to pursue aggressive care or comfort measures;

2. within the health care team as to whether to pursue aggressive care or comfort measures when a decision maker was unavailable;

3. within the health care team when 1 or more members of the team thought the care was futile;

4. between the health care team and the family as to whether to pursue aggressive care or comfort measures;

5. within the family as to who should be the surrogate decision maker; or,

6. between the family/friends and the health care team when 1 or more of the team members thought the care was futile.

When such patients were identified, an investigator at the local site enrolled the patient. A computer-generated block randomization scheme was used and patients were randomized to receive either an ethics consultation or usual care. The patient or the family members of the usual care group were not informed about the study and did not give informed consent; however, at most institutions, ethics consults were initiated by the health care team, so in general such consultations were not initiated by families. At all centers, any member of the health care team could request an ethics consult for the usual care group.

In the treatment group, the investigator contacted the treating physician to obtain a verbal consent to perform an ethics consultation. If the physician agreed, the ethics consultant followed a general process model of ethics consultation. After consent, an ethical diagnosis was made. Varieties of options were provided to the parties involved. A follow-up meeting was offered and performed if the family agreed. An ethics consultation note was entered in the chart. A follow-up interview was performed after several weeks with the health care team and the patient or surrogate decision maker to assess whether the ethics consult was helpful. The primary outcome measures were hospital days, ICU days, and days of life-saving treatments in those patients who did not survive to hospital discharge. For the purpose of analysis, intent-to-treat basis was used with assignment at the time of randomization as the basis of group allocation.

Out of 551 patients enrolled in the study, 278 were assigned to be offered ethics consultation, of whom 211 received an ethics consult. Of the 273 patients in the usual care group, 77 patients received an ethics consult. The groups were comparable in terms of age, sex, ethnicity, primary diagnosis, and surrogate decision makers. A total of 173 patients in the ethics consultation group (63%) died compared to 156 patients (57.8%) in the usual care group. The mortality rates of the 2 groups were not significantly different.

The ethics consultation group had fewer hospital days (-2.95 days, mean, 8.66 vs 11.62), fewer ICU days (-1.44 days, mean, 6.42 vs 7.86) and fewer days receiving ventilation (-1.7 days, mean, 6.52 vs 8.22), all statistically significant with P values < 0.05. The number of days receiving nutrition/hydration was not statistically different between the 2 groups (-1.03, 7.36 vs 8.38; P = .14). Health care team members and surrogate decision makers were satisfied with the ethics consultation and did not feel coerced to make a decision by the consultant. Schneiderman et al conclude that in those patients who did not survive to discharge from the hospital, ethics consultations were associated with a significant reduction in likely nonbeneficial treatments, without affecting the mortality (Schneiderman LJ, et al. JAMA. 2003;290:1166-1172).

Comment by Uday B. Nanavaty, MD

Addressing end-of-life issues and facilitating discussions regarding the aggressive care or "dying peacefully" during the hospitalization has become an important daily practice for the critical care provider. As the health care costs continue to climb, a variety of approaches are taken to control costs, especially in the areas consuming large amounts of resources. It has been clearly established that a large proportion of overall health care dollars are spent in the care of people during the last 6 months of their lives. A large share of this is in providing critical care to those who are not likely to benefit from it. However, in spite of all the research, it is hard to decide which patient is not likely to benefit from critical care.

It is not clear who should make the decision that critical care is no longer going to be beneficial to the patient. It is not clear what value system or health status should be used to define the group of patients who are not likely to benefit from critical care. Should cost be an issue? Who should decide where the health care dollars are spent? Who is to decide if the patient’s quality of life is such that life-saving treatments should be withheld or withdrawn? When so much of it is unknown yet the decision results in loss of life, ethical conflicts arise.

One way to resolve this dilemma has been to obtain the help of a bioethicist. The benefits of obtaining an ethics consult have largely been studied in case series. Hence the randomized control trial design used in this study is a good step toward starting to resolve the issues. The study also nicely delineates the framework of an ethics consult with all the steps involved in performing one.

This trial has several limitations. One of the foremost in my mind is that mortality was higher in the group of patients who were assigned to the intervention. Although statistically no different, it is possible that the study was not powered enough to detect a difference. A 5% difference may become statistically significant if a larger study group is used and may in fact make one argue that the ethics consultations were "mere subterfuge to pulling the plug"—the very fear of obtaining ethics consultation, as sited by Schneiderman et al. Also, although the study was powered to detect 3 days’ difference in ICU stay, the difference was substantially less than hypothesized. Schneiderman et al used intent-to-treat analysis, a rigorous method, to study outcome. However, there was a large crossover population, which may make it difficult to decide whether the effects observed were really due to the intervention.

In spite of these and perhaps other limitations, the study is in the right direction of starting a debate as to the needs to address the cost/resource consumption during seemingly "nonbeneficial" treatments toward the end of life. I believe that rather than counting and saving the ICU days at the very end of life, we need to change the focus toward months and years before that point is reached. As the mother of a dying young woman once said to the patient’s husband, "We know she is dying. What is the hurry to withdraw care?" I don’t have an answer for that myself. Maybe one of the ethics consultants will answer it one day.

Dr. Nanavaty, MD Pulmonary and Critical Care Medicine Rockville, Md.