Darien S. Fenn
Department of Psychiatry, Oregon Health Sciences University
Linda Ganzini
Department of Psychiatry, Oregon Health Sciences University
Geriatric Psychiatry, Portland Veterans Affairs Medical Center
Professional Psychology: Research and Practice
June 1999 Vol. 30, No. 3, 235Ð244
Abstract
Since the passage of the Oregon Death With Dignity Act (ODDA), psychologists have been grappling with how to fulfill their legally specified role in the process of physician,assisted suicide. We surveyed Oregon psychologists to elicit their views on assisted suicide and the process of assessing patients who request such assistance. There was a high degree of support for assisted suicide and the ODDA, but also a minority who were highly opposed. Most survey respondents raised ethical or practical concerns with their role as assessors. Many important questions about how evaluations in this complex area should be conducted remain unanswered.
The Oregon Death With Dignity Act (ODDA), passed by referendum in 1994, specifies that a physician may prescribe lethal medication that is to be used to hasten death for competent, terminally ill persons who voluntarily request it. Among the safeguards contained in this legislation is a provision that bars suicide assistance to a person who the attending physician feels may be “suffering from a psychiatric or psychological disorder, or depression causing impaired judgement” (ODDA, 1997). If such a disorder is suspected, the act specifies that a mental health assessment be provided by either a state-licensed psychologist or psychiatrist. If that assessment reveals a mental disorder that impairs judgment, no lethal medication may be prescribed until the mental health professional determines that the problem has been resolved.
The controversy over the issue of physician-assisted suicide has been heated in large part because individual views rest on a foundation of deeply held values. Proponents cite (in part) concerns with patient autonomy, the need to relieve suffering (Quill, Cassel, & Meier, 1992), and fears that life can now be technologically extended beyond the point where it is meaningful (Latimer, 1991* Weir, 1992). Opponents cite the ethic of “doing no harm” (Jennings, 1991), concerns about a “slippery slope” (Fenigsen, 1990) to abuse or misuse with the disadvantaged (Hendin & Klerman, 1993), and religious beliefs that such life-or-death decisions are solely the province of the divine (O’Rourke, 1991). Opponents also question whether a “rational suicide” is possible (Callahan, 1994) and emphasize the potential for personal growth in the last stage of life (Byock, 1993).
With the passage of the ODDA, the debate on assisted suicide moves to also address the pragmatic aspects of implementation. We surveyed psychologists in Oregon to ascertain their views and concerns with regard to their role in the process. As the first group of psychologists in this country to address the practical aspects of the assessment problems, their views should help guide discussion on the issue. At the same time, because proponents of assisted suicide have indicated their intent to introduce legislation in other states (as was recently the case in Michigan), what happens in Oregon may also impact psychologists in other states.
The Oregon Psychologist Assisted Suicide Survey
All questions in this anonymous survey were drawn from issues mentioned in the scientific literature or in the popular press and from discussions of physician-assisted suicide that have occurred among colleagues as the issue has unfolded in Oregon. The questionnaire is a companion survey to one conducted with Oregon psychiatrists (Ganzini, Fenn, Lee, Heintz, & Bloom, 1996) and assessed attitudes on a variety of ways in which a physician might be involved in hastening death, from withdrawal of life-sustaining treatment to the writing of a prescription for medication whose sole purpose was to end life. We also asked whether the respondent would consider physician-assisted suicide for himself or herself and under what conditions such assistance might be desired. We asked general questions about physician-assisted suicide and specific ones regarding the ODDA, its safeguards, and the impact that participation would have on psychology as a profession. We also asked several questions regarding willingness to participate in and confidence in the evaluation process as specified by the ODDA. Finally, we asked 10 questions describing factors that might be important in deciding when assisted suicide should be allowed for others, and an additional 21 items that described factors that influenced their personal views. Altogether, the questionnaire consisted of 8 demographic items, 51 closed-choice items, and 6 open-ended items. Copies of the questionnaire are available from Darien S. Fenn on request.
The Oregon Psychologist Assisted Suicide Survey
All questions in this anonymous survey were drawn from issues mentioned in the scientific literature or in the popular press and from discussions of physician-assisted suicide that have occurred among colleagues as the issue has unfolded in Oregon. The questionnaire is a companion survey to one conducted with Oregon psychiatrists (Ganzini, Fenn, Lee, Heintz, & Bloom, 1996) and assessed attitudes on a variety of ways in which a physician might be involved in hastening death, from withdrawal of life-sustaining treatment to the writing of a prescription for medication whose sole purpose was to end life. We also asked whether the respondent would consider physician-assisted suicide for himself or herself and under what conditions such assistance might be desired. We asked general questions about physician-assisted suicide and specific ones regarding the ODDA, its safeguards, and the impact that participation would have on psychology as a profession. We also asked several questions regarding willingness to participate in and confidence in the evaluation process as specified by the ODDA. Finally, we asked 10 questions describing factors that might be important in deciding when assisted suicide should be allowed for others, and an additional 21 items that described factors that influenced their personal views. Altogether, the questionnaire consisted of 8 demographic items, 51 closed-choice items, and 6 open-ended items. Copies of the questionnaire are available from Darien S. Fenn on request.
We selected at random a sample of 625 psychologists on the basis of a list of the 846 licensed psychologists in the state of Oregon obtained from the Oregon Board of Psychologist Examiners. We received 461 responses before the cutoff date of May 1996, a return rate of 74%. For the final analyses, we excluded 38 psychologists who indicated they were not in clinical practice. Of the final sample of 423, 95% had a doctoral-level education (PhD, PsyD, or EdD), and 57% were male. Practice settings included independent practice (80%), community mental health (16%), university (15%), and other (25%), with 70% of the respondents indicating they work in more than one setting. Age in the sample ranged from 31 to 76 years (average 48.8 + 8.1), with 1 to 45 years of practice as a psychologist (average = 13.8 + 8.5). Slightly more than half the participants (248) listed an additional 5 to 62 years of experience in other mental health professions (average = 10.9 + 6.9).
Attitudes Toward Suicide and Assisted Suicide
A majority of respondents felt that, for a competent individual, both suicide and assisted suicide were acceptable under some circumstances or that the decision should be left to the individual involved (see Figure 1). With regard to the physician’s role in hastening death, few opposed such measures as withdrawal of life support. Similarly, few opposed the use of high doses of analgesics, even when hastened death might occur as an unintended effect. However, if the specific purpose of providing medication was to assist in the patient’s suicide, opposition rose considerably (see Figure 2). As is clear in Figures 1 and 2, the majority of respondents supported assisted suicide only under “some circumstances.” Figure 3 shows how that support varied with a number of factors that are often mentioned in discussion of assisted suicide. Responses ranged from the amount of expected pain and physical discomfort, felt by 95% of respondents to be “important” or “very important,” to the wishes of the family, ranked as very important by only 10% and important by an additional 36%.
Eighty-two percent of psychologists (n = 345) indicated that they would consider obtaining a physician’s assistance to end their own lives under some circumstances. Of these respondents, 328 also provided from 1 to 6 (median = 3) open-ended responses describing those circumstances (see Table 1). For seven of these categories, there were corresponding items describing the factors respondents felt were important in deciding when assisted suicide should be allowed for other people, as shown in Table 1. For all but one of these (expected length of life) there was a strong association between the personal importance of the item from the open-ended responses and the importance of the corresponding factor for deciding when assisted suicide should be allowed for others. For all of the comparisons that were significantly related, the association was strongest between inclusion of the open-ended item and endorsement of the very important level of the support item.
Views on the Oregon Death With Dignity Act
Seventy-eight percent of respondents favored the enactment of the ODDA (see Figure 4). Conversely, a slight majority of respondents preferred that official representative organizations take no position on the measure, with those most in favor of the act also in favor of taking a supportive position (r = .563, p < .001). Even among the 9 who strongly opposed the ODDA, respondents were almost equally split between preferring organizational neutrality (4%) and opposition (5%). Although most respondents felt that the safeguards in the ODDA were adequate, there was also an association between positive views of the safeguards and general support for the enactment of the ODDA (r = .711 p < .001). Ninety-three respondents offered suggestions for improving safeguards (see Table 2). Calls for a mandatory mental health evaluation were associated with favoring the ODDA, c2(3, N = 93) = 11.86, p = .008, whereas concerns about the physicians' ability to adequately screen for mental health problems tended to be strongly opposed to the ODDA, c2(3, N = 93) = 8.72, p = .033. Only one fifth of respondents (20%) felt that psychologists’ participation in the assisted suicide process would constitute a threat to the profession. This view was associated with opposition to both the ODDA, c2(3, N = 389) = 82.1, p < .001, and assisted suicide, c2(2, N = 391) = 81.2, p < .001. Specific concerns mentioned included tarnishing psychology's public image (36%), politicization of professional issues (23%), and role conflicts (20%), with an additional 15% concerned about stepping outside the role of the neutral party. Respondents also mentioned concerns about the scientific validity of assessments (15%), assessor qualifications (9%), and the risk of "hired gun" specialists (7%). Concerns with the psychologist's role as a neutral party were associated with strong opposition to the ODDA, c2(3, N = 78) = 9.3, p = .025, and opposition to assisted suicide, c2(2, N = 80) = 8.1, p = .017.
The Psychologist’s Role in the Assessment Process
One third of the respondents indicated that performing an evaluation under the ODDA would be outside their practice area. Of the remaining 275 psychologists, 20 (7%) indicated that they would refuse to perform the evaluation and take no further action, 91 (33%) indicated they would refuse to perform the evaluation, but make a referral to a colleague, and 164 (60%) indicated they would perform the evaluation. Those who reported that they would refuse to perform the evaluation and take no further action were more likely than the other respondents to have had no experience with terminally ill family, friends, or patients, c2(2, N = 267) = 9.5, p = .009.
Of the people who reported that they would perform the evaluation, 151 also responded to the question on how they would follow up on the evaluation. The majority (62%) indicated that they would provide the assessment to the referring physician and take no further action. Twenty-eight percent indicated they would support the patient in obtaining a lethal medication from his or her primary physician. The remaining 10% of respondents reported that they would work with the patient or others involved in the case to try to prevent the suicide. This last group of respondents was more likely than the others to oppose both the ODDA, c2(6, N = 145) = 35.4, p < .001, and assisted suicide, c2(4, N = 147) = 31.1, p < .001. When psychologists were asked about the level of confidence in performing an assessment to determine whether a psychiatric disorder was impairing judgment, responses varied with the nature of the evaluation (see Table 3). When placed in the context of a single evaluation, half of the respondents replied that they were “not at all confident,” with only 7 reporting that they were “very confident.” Conversely, in the context of a long-term relationship with the patient, only 3 felt “not at all confident,” whereas 64 reported they would feel “very confident.” When we restricted the respondents to only those who stated they would perform the evaluation, the number who were “not at all confident” in the context of a single evaluation reduced to 28. Also, within this subgroup and in the context of a single evaluation, men were more likely to report being very confident and women being not confident, c2(2, N = 162) = 8.7, p = .013. We asked respondents if they felt that a request for assisted suicide from a terminally ill patient was prima facie evidence of a mental disorder. Only 13 people (3%) agreed or strongly agreed with this assertion, whereas 21% disagreed and 76% strongly disagreed. This item was strongly associated with most of the other major attitudinal measures, including the moral acceptability of suicide, c2(3, N = 384) = 57.6, p < .001, and assisted suicide, c2(3, N = 382) = 59.8, p < .001, whether assisted suicide should be permitted, c2(6, N = 382) = 69.8, p < .001, consideration of assisted suicide for one's self, c2(3, N = 381) = 34.8, p < .001, and support for the ODDA, c2(9, N = 377) = 105.4, p < .001. For all these comparisons, it was the views of those who strongly disagreed with the prima facie item that contrasted with the views of those who disagreed or agreed with the item.
Hierarchical Logistic Regression Modeling
With regard to self-reported attitudes that influenced a respondent’s views on the acceptability of assisted suicide, all but three of the items (experience with terminally ill family or friends, experience with terminally ill patients, and views regarding the consistency with the physician’s role in relieving pain and suffering) were significantly associated with views on assisted suicide (see Table 4). These items also had an average intercorrelation of .178, with one third of the correlations greater than .30. We therefore performed a stepwise logistic regression in order to determine which of these attitudes were independent predictors of a respondent’s position with regard to assisted suicide. We also included age, sex, years in practice, experience with terminally ill family, friends, or clients, number of clients who had died, attitudes toward suicide (not assisted suicide), and whether one would consider assisted suicide for himself or herself, as predictors. Bivariate analyses suggested that there were not only differences in the predictor items between those most opposed to assisted suicide and those who felt that the procedure should be allowed sometimes or always, but that some differences existed between these latter two groups as well. We therefore compared those who stated that assisted suicide should never be allowed to the remainder of the sample and then looked at those items that separated the “sometimes” and the “always” groups.
Strong opposition to assisted suicide was predicted by six factors: a) not considering assisted suicide as a possible personal option, b) a belief that suicide per se was not moral, c) the view that the physician’s role was one of preserving life, d) placing less importance on a person’s right to self-determination, e) concern that assisted suicide might be misused with the disadvantaged, and f) younger age. The model accounted for 76% of the variance (Nagelkerke’s R2* Nagelkerke, 1991) in the dependent measure and successfully categorized 93% of the respondents.
Comparing those who would support assisted suicide “under some circumstances” with those who reported that it should be allowed “always” yielded a different set of predictors. Those who felt assisted suicide should always be allowed were a) less confident in the degree to which their future quality of life could be determined, b) more supportive of a person’s right to self-determination, c) more confident that impairment in judgment due to mental condition could be assessed in a single assessment, and d) less concerned about “allowing the natural dying process to take its course.” This model, however, was far less predictive than the previous one, accounting for 20% of the variance in the dependent measure and successfully categorizing 70% of the respondents.
Other Concerns
We asked respondents if there were any other important considerations with regard to assisted suicide. A total of 170 respondents (40%) provided us with comments. We coded these responses into 1 or more of 10 categories (see Table 5).
Implications and Discussion
Although as of this writing, the ODDA has been in effect for 1 year, concerns over whether the procedure should be allowed and over details of its implementation are far from resolved. For psychologists outside of Oregon, there are two primary implications. First, although further legal challenges to the ODDA have been promised (Barnett, 1997* Green & Hoover, 1997), some commentators believe that the ODDA will not only survive these challenges (Green & Hoover, 1997), but that its mere passage in Oregon may stimulate further debate or similar legislation in other states (Hendin, 1997). Public support for some sort of control over the dying process is high, and the high profile created by the public discussion of the issue may result in physicians receiving more requests for assisted suicide. Such a result may further encourage popular referenda, as occurred in the 1998 ballot in Michigan. Our survey results suggest that the professional issues are more subtle and complex than what enters the public rhetoric. Indeed, many respondents commented that they felt they had not thought through the issues thoroughly enough and that just responding to the survey had stimulated much reflection on their own views. Because depression and competence will continue to be central issues, psychologists’ input should play an important part of the public debate.
Second, assisted suicide is legal in Oregon, and what happens in this state will likely set the standards by which psychology’s role in the process will initially be judged. Whether psychologists should be involved in assisted suicide evaluations, under what circumstances, whether specialized training and/or certification should be required, and the development of a research base and technology of assessment are issues that need development at a professionwide level. Addressing these concerns may be more difficult than it might at first appear. For example, some of our respondents felt that standards for evaluations should include assessments at multiple points in time, extended waiting periods, mandatory psychotherapy, or trials with antidepressants. Yet Oregon’s experience with the ODDA to date shows that a substantial fraction of the people who have made requests for assisted suicide under the law have not lived through the 15-day waiting period (Hoover, 1998), rendering the above suggestions impractical. At the same time, the sensitive nature of this time for patients and their families and concerns for privacy heightened by media attention create considerable obstacles to the research needed to inform a training and standards effort.
In order to understand the debate over assisted suicide, it is important to note that proponents and opponents of assisted suicide view the issues quite differently. Our findings suggest that questions such as whether assisted suicide will divert resources from palliative care or result in distrust of medical professionals are most likely a reflection of deeply held values such as religious beliefs, the moral acceptability of suicide (not assisted suicide), or views that the role of the physician is foremost to preserve life. Whereas the debate on assisted suicide is multifaceted and complex on its face, differences at these core levels may underlie the voiced disagreements and thus contribute to the often intractable nature of the debate.
Although our survey found broad support for assisted suicide in general, levels of support were less for the ODDA, suggesting that there is disagreement over how the procedure should be implemented. Some of these disagreements were apparent in comparisons between those who were somewhat supportive and those who were strongly supportive of assisted suicide. Those who strongly supported assisted suicide were less confident in psychologists’ ability to assess the quality of life of patients than were those who endorsed a more moderate view, but more confident that one could assess whether judgment was impaired with regard to the request for assisted suicide. For the strong proponents of assisted suicide, this combination of beliefs would give greater weight to the patient’s self-report of quality of life and apply reduced concern to the question of how well one could guard against requests made in an impaired condition. More or less support for assisted suicide would logically flow from one’s position on these two items. Regardless of how one views these issues, concerns about the reliability and validity of assessments are more amenable to resolution than are questions of a moral nature. These findings suggest that increased attention to such concerns could help move the strong and moderate supporters toward a consensus but are unlikely to bridge the gap between supporters and opponents.
Several of our negative findings are notable. We asked respondents about their experiences with terminally ill patients, family, and friends, whether they had been in the role of caregiver for a dying family member, and whether significant pain or suffering had been a part of that experience. None of these experiences influenced reported views on assisted suicide or on the ODDA. Neither did we find a difference in views on the basis of gender or years of experience. Although they were unrelated to views on the issue, gender and experience both emerged as factors when we asked how people would respond to a request to perform an evaluation as specified by the ODDA, with women and those with less experience with the terminally ill opting for less participation in the process.
The ODDA specifies that the physician can make a referral for an evaluation to determine “whether the patient is suffering from a psychiatric or psychological disorder, or depression causing impaired judgement” (ODDA, 1997). The nature of that evaluation is not specified. As such, we deliberately asked our questions regarding respondents’ confidence that they could perform the assessment in the context of a single evaluation or within the context of a long-term relationship, but without specifying any details. Half of all respondents reported that they were “not at all confident” that they could do what was asked in a single evaluation. In contrast, the percentage of respondents who were “not at all confident” diminished to 3% when the evaluation was in the context of a long-term relationship with the patient. This finding highlights a dilemma that may confront evaluating psychologists. By definition, patients who need a mental health evaluation will be very near the end of life, so the time to establish a therapeutic relationship may simply not be available. Further, many will be severely physically debilitated or have periods of cognitive impairment, which will limit the quality and effectiveness of what can be done in the available time. Thus, psychologists may be asked to do extraordinarily important work in far from ideal conditions. To the extent that these conditions influence the quality of the evaluations, their reliability and resulting adequacy as a safeguard may be diminished.
To complicate this matter, support for assisted suicide was related to how psychologists viewed the evaluation process. Those who indicated that they would be willing to perform the evaluation were more likely to be supporters of assisted suicide, whereas opponents were more likely to refuse and take no further action. Supporters were also more likely to have a greater degree of confidence in both assessment conditions (long-term or short-term). Contrary to what one might expect, psychologists’ confidence was not associated with factors such as age, years in practice, or experience with terminally ill family or patients. What these findings do not make clear is the reason for the association between views on assisted suicide and the evaluation process. On one hand, greater confidence in the evaluation process might justify support for assisted suicide, or conversely, lack of confidence might justify opposition. Alternatively, it is important to remember that this survey was conducted in the midst of a very politicized debate on legalization of assisted suicide. Respondents may have consciously or unconsciously biased their reported confidence levels to be consistent with their position on assisted suicide.
A central issue in past discussions of assisted suicide has been concern that the deeply held personal views that underlie the debate over assisted suicide may influence the thinking and decision making of evaluators. If this is a fundamentally new role for psychology, how much does the absence of training, standards, an empirical database, and clinical experience threaten professional objectivity? Respondents’ comments raise concerns about objectivity for both proponents and opponents of assisted suicide. On one hand, some respondents stated that “the first thing I would do is try to talk the patient out of committing suicide,” whereas others, frankly admitting bias, noted that under the circumstances, they would set a very low threshold for what would constitute competence. If an evaluator judges the rationality of a patient by the degree to which the reasons underlying a patient’s request for assisted suicide match his or her own, it is possible that the evaluator’s personal views could sway the outcome of the evaluation (Sullivan, Ganzini, & Younger, 1998). We were able to examine this possibility because we asked respondents to indicate the factors that they thought should be considered in allowing assisted suicide for others and what factors would influence their own decisions to request assisted suicide. We found that the two sets of items were strongly related. For example, as a group, our respondents felt that financial impact on the patient’s family was among the least important factors that should be considered in deciding when assisted suicide should be allowed for others. However, individually, those who indicated that such financial concerns would influence a decision to request assisted suicide for themselves rated this item as highly important for evaluating the appropriateness of assisted suicide for others. Such results suggest that concerns raised about individual bias should be explored further. Our respondents were clearly aware of these potential issues. Consistent with this awareness, the most numerous of open-ended comments on the survey were related to problems with the evaluation process, the need for research, standards and training for the evaluations, and concerns that the evaluation process not come to be dominated by “hired guns” who would routinely find either for or against a patient’s psychological competence.
Evaluations of patients requesting assisted suicide will carry a level of import rarely found in psychology: Every evaluation will be a decision between life and death or, depending on the perspective of the individual, a decision between an unnatural intervention in the process of death or allowing unnecessary and prolonged suffering. At the same time, little has been done in the area of conducting evaluations for such decisions and many unanswered ethical and practical questions remain (Hadjistavropoulos, 1996). There may be some ambiguity as to whether the patient or the referring physician is the client. Whether the psychologist can ethically serve as an advocate for the patient, work to prevent the suicide, or should merely provide a report to the physician depends on how that question is resolved. Should a psychologist who feels strongly that assisted suicide is morally unacceptable recuse himself or herself from the evaluation process, or are such dissenting voices an important counterbalance? What should be included in an assessment? The ODDA specifies that an assessment must determine whether the patient has impaired judgment specifically due to a psychiatric or psychological disorder or depression. What should the psychologist do if he or she believes that the patient’s judgment is unaffected by such a mental disorder, but nonetheless believes that the patient’s judgment is questionable because of other factors such as family members’ attempts to sway the patient’s decision or the patient’s concerns that he or she is becoming a financial or emotional burden to the family? To what degree is the specified assessment equivalent to an evaluation of decision-making competence, and if it is, which standards for competence apply (Marson, 1995* Moye, 1996)? Problems with traditional methods of competency assessment in other contexts have already been documented (Silberfeld, Corber, & Checkland, 1995* Tancredi, 1987* Twining, 1983* Tymchuk, Ouslander, Rahbar, & Fitten, 1988). Can the capacity of a terminally ill patient to choose suicide be understood outside the context of a long-term system of beliefs and attitudes? If not, how does one reliably and validly assess those beliefs and incorporate them into a final conclusion? And of particular significance, how well do assessment or treatment methods derived from other populations generalize to the unique circumstances and physical condition of a terminally ill person?
Undoubtedly, mental health issues will continue to play a pivotal role in the debate over assisted suicide. Until very recently, mental health professionals have participated little in the discussion (Block & Billings, 1995* Peruzzi, Canapary, & Bongar, 1996), but concerns raised by Oregon psychologists in response to this survey suggest that there are many unresolved philosophical and practical issues, both with respect to the specific instance of the ODDA, and with regard to assisted suicide in general.
References
Barnett, J. (1997, November 10). Utah Sen. Hatch vows to fight Oregon’s assisted suicide law. The Oregonian, A1.
Block, S. D., Billings, J. A. (1995). Patient requests for euthanasia and assisted suicide in terminal illness. The role of the psychiatrist. Psychosomatics, 36, 445–457.
Byock, I. R. (1993). Consciously walking the fine line: Thoughts on a hospice response to assisted suicide and euthanasia. Journal of Palliative Care, 9, (3) 25–28.
Callahan, J. (1994). The e