Is it appropriate for physicians to accept assisted-death requests at face value, or should they be interpreted as clinical indications of suffering? Should physicians act on patient requests to die, or should they address patient needs through other measures? Such are the difficult questions facing most physicians today.
Is it appropriate for physicians to accept assisted-death requests at face value, or should they be interpreted as clinical indications of suffering? Should physicians act on patient requests to die, or should they address patient needs through other measures? How can such decisions be made when patients themselves cannot voice their wishes? What factors other than patient suffering influence requests for assisted death?
Such are the difficult questions facing most physicians today, particularly the more than 102,000 medical residents in this country who are just learning to identify and respond therapeutically to the devastating effects of serious illness, says Laura Weiss Roberts, M.D., assistant professor of psychiatry at University of New Mexico (UNM) School of Medicine. These are the questions that Roberts and her colleagues are working to answer.
"Recent controversy surrounding physician-assisted suicide and euthanasia practices has created even greater challenges for residency education," Roberts said. "The well-being of vast numbers of seriously ill and dying patients, now and in the future, is in the hands of today's physicians-in-training. Yet there is an important and disturbing gap in the literature regarding how residents view end-of-life care."
Roberts said that to her knowledge only one recent study (Caralis and Hammond, 1992) looked at views of residents regarding assisted-death activities. In that study, views of 100 internal medicine residents and 17 surgery residents were compared with views of 127 faculty physicians and 116 medical students.
"Residents placed emphasis on perceived quality of life considerations, whereas faculty physicians assigned greater valence to disease-based information in making decisions, such as the natural history of diagnosed illness or the patient's prognosis with various treatments," Roberts said.
"In this article, responses given by surgery and internal medicine residents were combined; we found no other studies comparing attitudes of residents from different specialties."
To further investigate residents' attitudes toward assisted death, Roberts and her team of researchers surveyed 96 residents in emergency medicine, internal medicine and psychiatry at UNM. The study, which was published last year in the Archives of Internal Medicine (Roberts et al., 1997b), employed a survey instrument that was created by the Therapeutic Care Committee of the Group for the Advancement of Psychiatry.
"We felt that they had done very valuable work and wanted to build upon that," said Roberts. "Their survey was modified for this study to include questions about training background, ethnicity and experience in providing end-of-life care to patients."
The Roberts et al. study presented six patient vignettes to participants. Four of the vignettes were descriptions of patients who had asked to die, including a 30-year-old man with rapidly progressing amyotrophic lateral sclerosis; a 50-year-old woman with severe pelvic pain who had refused antidepressant or ECT (electroconvulsive therapy) treatment for what had been diagnosed as psychosomatic pain; the same 50-year-old woman who agreed to both psychiatric treatments, both of which then failed to relieve her perceived pain; and a 37-year-old man serving life in prison for the raping and killing of women, behavior that was repeated despite intensive psychotherapy and treatment with medroxyprogesterone acetate to diminish sexual drive.
For each of these vignettes, residents were asked whether they would "directly assist this patient to die," "refer this patient to someone to assist [this patient] to die," "find it acceptable for other physicians to assist similar patients to die," and "find it acceptable for nonphysicians to assist similar patients to die."
The fifth vignette described a 27-year-old woman with AIDS and severe pneumonia, who was unresponsive to antibiotics and had been in a coma for the past two months. She was receiving intravenous morphine and was on a respirator. In this case, it was the family who hoped death would occur soon, and questions relating to the vignette were similar to those for the previous vignettes, except that they referred specifically to physicians or nonphysicians, including family, turning off the patient's respirator.
The final vignette described the same AIDS patient, but in this case her respirator had been turned off, and three weeks later she was still alive. The questions here related to increasing morphine to hasten death.
"We hypothesized that resident attitudes would be affected by clinical aspects of patient cases and by who might be involved in carrying out assisted suicide or euthanasia [the 'agents']," said Roberts. "We also wondered whether residents from the three different specialties would have differing views and whether variables such as gender, ethnicity, religious beliefs and personal philosophy might predict attitudes expressed by these residents."
As the authors predicted, results of statistical analyses revealed that residents responded differently to the six vignettes, based on clinical aspects.
"They clearly opposed hastening death of the woman with severe pelvic pain, although they were somewhat less opposed if she had experienced failed psychiatric treatments," Roberts said. "They were only slightly opposed to death assistance for the rapist/killer with failed psychological and medical treatments. They were more uncertain or neutral about hastening death in the cases of the man with amyotrophic lateral sclerosis and the comatose woman with AIDS who had survived despite discontinuation of respiratory support. Residents were in favor of withdrawal of life support for the woman with AIDS."
Regarding the effect of which agent should hasten death, "Residents did not favor assisted-suicide practices conducted by nonphysicians and were disinclined to hasten patient death personally," Roberts said. "They were neutral or uncertain about assisted suicide conducted by referral physicians or by physicians in general."
In addition, there appeared to be a relationship between agent and patients' circumstances. "That is, assisted death was perceived as more or less acceptable for a certain patient depending on which agent was involved, or conversely, assisted death performed by a certain agent was perceived as more or less acceptable depending on which patient was involved," said Roberts. Specifically, for the first four vignettes, residents were less likely to approve of assisted suicide carried out by themselves or nonphysicians in general.
"However, the pattern changed for the vignettes of the woman with AIDS," Roberts said. "For this comatose woman, residents were as willing to withdraw support or hasten death by increasing morphine themselves as they were to have a referral physician or physicians to do so."
Regarding effects of respondent characteristics, seven such factors were examined: gender, year of training, percentage of residents' patients dealing with end-of-life issues, religious influence, personal philosophy influence, medical training influence and ethnic group. Only religious influence and personal philosophy influence were found to reliably predict respondent perceptions.
"Twenty-two percent of respondents reported being influenced by religious beliefs, with those being religiously influenced less likely to support assisted death," said Roberts. However, "religious influence...did not have any effect for the woman with pelvic pain."
With respect to personal philosophy, 77% of respondents reported that their views had been strongly influenced by this factor. Those reporting this influence were less opposed to assisted-death practices.
"This analysis also revealed a personal philosophy influence by patient vignette interaction," said Roberts. "Interestingly, this interaction qualifies the main effect of personal philosophy influence in the same manner as with religious influence, i.e., for the woman with pelvic pain, personal philosophy did not predict responses, whereas with the other patients, the pattern was consistent with the main effect."
Overall, the residents indicated uncertainty or opposition toward assisted suicide. However, averaged across patient and agent, emergency medicine residents were more likely to support assisted suicide and euthanasia than were either internal medicine or psychiatry residents.
Roberts noted that several other recent surveys in this country have investigated clinicians' attitudes toward, and their direct experience with, assisted death and euthanasia.
"This early empirical work involving physicians and nurses suggests that patient requests for these practices are not uncommon," said Roberts, citing a survey conducted by Back et al. (1996). In that study of 828 Washington state physicians, 16% of respondents reported that patients had requested help in dying. In another study by Lee et al. that same year, 21% of 2,761 Oregon physicians reported such requests. Also, 17% of 852 critical care nurses surveyed by Asch, also in 1996, reported suicide requests.
Roberts said that several large studies recently documented that the majority of physicians are personally unwilling to participate in assisted-death practices, especially euthanasia (Caralis and Hammond, 1992; Cohen et al., 1994; Lee et al., 1996; Roberts et al., 1997b).
"Nevertheless, early empirical investigation has also shown that physicians from various specialties often view these practices as permissible under certain circumstances, and are relatively tolerant of other physicians' involvement in facilitating or directly causing patient death," Roberts said.
Using the same study instrument, Roberts and colleagues investigated attitudes of 184 consultation-liaison (C-L) psychiatrists (Roberts et al., 1997a).
"A comparison of the C-L psychiatrist study and the resident study indicates that resident physicians are somewhat more accepting of assisted death and euthanasia than C-L psychiatrists," Roberts reported. "In both studies, less opposition was expressed overall regarding assisted suicide for the man with amyotrophic lateral sclerosis and regarding euthanasia for the comatose woman with AIDS. Findings were also similar in that referral and other physician conduct of assisted death was seen as more acceptable than either direct personal involvement or participation of nonphysicians."
Roberts and colleagues have also completed a third study that looked at the attitudes of 166 medical students. Those results, not yet published, indicate that compared to respondents in Roberts' other studies, medical students are the most receptive to assisted-death practices.
"They didn't embrace assisted-death practices but they were more tolerant," Roberts said. "This makes sense because medical students are more similar to the lay public, whereas residents were becoming more conservative but still kind of open to it, and then when you get to the C-L psychiatrists, they're actually quite conservative and unreceptive regarding this [practice]. Part of what's interesting to us is that there might be some developmental process here, but we don't know what it is yet."
Roberts said that taken together, these studies raise important questions that need answers. "Why is assisted suicide viewed as permissible for certain patients and not others? Why do differences exist around who can acceptably hasten patient deaths? Why are physicians willing to accept assisted death performed by colleagues when they find it objectionable to do themselves? What informs these views?
"Further investigation is imperative so that care of severely ill patients is derived from explicit, clinically and ethically sound principles of medicine and not based on uncertain motives, incorrect information or prejudicial attitudes," she said.
Asch DA (1996), The role of critical care nurses in euthanasia and assisted suicide. N Engl J Med 334(21):1374-1379. See comments.
Back AL, Wallace JI, Starks HE, Pearlman PA (1996), Physician-assisted suicide and euthanasia in Washington State. Patient requests and physician responses. JAMA 275(12):919-925. See comments.
Caralis PV, Hammond JS (1992), Attitudes of medical students, housestaff, and faculty physicians toward euthanasia and termination of life-sustaining treatment. Crit Care Med 20(5):683-690.
Cohen JS, Fihn SD, Boyko EJ et al. (1994), Attitudes toward assisted suicide and euthanasia among physicians in Washington State. N Engl J Med 331(2):89-94. See comments.
Lee MA, Nelson HD, Tilden VP et al. (1996), Legalizing assisted suicide-views of physicians in Oregon. N Engl J Med 334(5):310-315. See comments.
Roberts LW, Muskin PR, Warner TD et al. (1997a), Attitudes of consultation-liason psychiatrists toward physician-assisted death practices. Psychosomatics 38(5):459-471.
Roberts LW, Roberts BB, Warner TD (1997b), Internal medicine, psychiatry, and emergency medicine residents' views of assisted death practices. Arch Intern Med 157(14):1603-1609.