The issue of whether or not to legalize physician-assisted suicide has been front and center as a public policy issue around the world. Many proponents and opponents are largely operating on assumptions as to why people participate in physician-assisted suicide with only a limited amount of support for their attitudes. Not only do those that support physician-assisted suicide often assume that people participate in it for primarily rational and medical reasons from usually physical illnesses. Opponents of physician-assisted suicide often assume that the participants are motivated primarily by typical suicidal behavior, and push for suicide prevention intervention. Clearly one’s outlook toward this topic depends a great deal on the lenses through which one views it.
The present essay has two purposes. First, we will summarize some of the data emerging from the PAS sample we have been studying in Michigan. These data are striking in a number of important ways and are important in themselves in an attempt to characterize motivations of people who seek PAS. These data have been presented in detail in several recent articles published both by our research team (Kaplan, Lachenmeier et. al., 2000; Kaplan, O’Dell et. al., 2000) and others (Canetto and Hollenshead, 2000).
Here we summarize these data around gender in a particular way to meet the second purpose of our paper: to place these particular data set in a more general model of PAS world-wide, focusing on differences in gender-ratios across these samples. This model will introduce the conception of the degree of physician control as an ordering principle and will examine its relationship to the gender ratio of the PAS participants. In other words, degree of physician control will be treated as an independent variable, ranging across various data sets from unassisted suicide (no doctor involvement) on the one-hand to full euthanasia (full doctor control) on the other. The proportion of women versus men participating in hastened death will be treated as a dependent variable.
Generally, researchers agree that the relationship between physical illness and psychiatric symptamotology is complicated (Fawcett, 1972; Murphy, 1977; Conwell et. al., 1990).
The present research report presents data to determine the relative roles of psychosocial versus biomedical factors in the PAS cases in Michigan performed by Dr. Kevorkian and his team. Specifically, we focus on the question of gender differences in this regard.
Sample. The Michigan data derives from our own IRB approved research on the PAS conducted by Dr. Kevorkian and his team. Our research team, in conjunction with the Detroit Free Press, has identified 93 of these cases, and there is evidence that even with Kevorkian presently in prison, other members of his team have continued to perform additional PAS both in Michigan and in other states (Detroit Free Press, May 13, 1999).
From these 93 cases, we have administered a psychological autopsy to friends and relatives of the first 47 of these decedents who were assisted in their death in the period June 4, 1990 to February 2, 1997. This psychological autopsy technique has been widely used in the study of suicide and allows the reconstruction of the psychological profile of the decedent in a manner parallel to a physical autopsy (Barraclough et. al., 1974; Hagnell and Rorsman, 1979; Fowler et. al., 1986; Clark and Horton-Deutsch, 1992).
This will be discussed in more detail below.
We have also collected more cursory data gleaned from the death certificates and from the judgments of the medical examiners of the remaining 46 cases (Dr. Dragovic in Oakland County, Drs. Kahnluen and Somerset in Wayne County, and Dr. Spitz in Macomb County, all in Michigan).
The psychological autopsy involved collection of interview data with regard to demographic, biomedical, and psychosocial information about the decedent from families and friends in addition to medical personnel familiar with the case. The non-medical informants were of great value in obtaining information as to psychosocial issues regarding a decedent. To ensure accuracy, we followed standard psychological autopsy methodology in requiring the agreement of two or three informants who knew the decedent well. (Clark and HortonDeutsch, 1992; Kaplan and Maldaver, 1993).
Specifically, we only scored a symptom or piece of information as positive if it was indicated as present by at least two informants (Clark and Horton-Deutsch, 1992).
Guidelines were provided for the interviewers. For example, terminality was defined at autopsy as a projection of six months or less to live as determined by the respective medical examiner. Category of disease, anatomical basis for disease, and method of death were also defined by the medical examiner at autopsy. Reported pain, depression, and history of suicide attempts were assessed from the reports of friends and relatives regarding complaints of the decedent while anatomical basis for pain was defined by the medical examiner at autopsy as the physical specification of an anatomical basis for reported pain on the part of the decedent. This is admittedly a very difficult judgment to make by a medical examiner at autopsy, but we felt it was important to attempt to distinguish reported pain that had a specific anatomical referent from that which did not (Detroit Free Press Staff, 1997; Kaplan, Lachenmeier et. al., 2000; Kaplan, O’Dell et. al., 2000).
First of all, disability (87%) and fear of dependency (90%) seem more a factor than terminality (31.1%) for the Michigan sample of people dying by PAS. This is especially true among women. This raises the question of what can be done by society to better enable people with physical incapacitation and disabilities to live more full lives.
A second finding indicates that almost twice as many patients reported pain (73.6 %) than revealed an anatomical basis for that pain at autopsy (42.6 %).
This suggests that a considerable number of these decedents may have experienced psychological-based rather than anatomically-based pain. This trend seems to be higher among women than men, with 75% of both genders reporting pain, but men described as having almost twice the rate of anatomical basis for pain (60 %) as women (34.4%).
This pattern is very troubling. On the one hand, it may simply reflect the traditional inability of male physicians to adequately recognize the physical basis for women’s pain and thus to over-psychologize it. However, it may also indicate that there may be different bases for the pain behind the decision of men and women to seek death, and that differential techniques may be appropriate to control that pain.
The third finding is related. Psychosocial factors seem as important as biomedical factors in the Michigan PAS sample, especially among women. For example, a typical man choosing PAS may be suffering from terminal lung cancer and be in a great deal of physical pain. A typical woman choosing physicianassisted suicide may be somewhat disabled from chronic multiple sclerosis and as a result, see her marriage break up and her economic situation deteriorate. Her pain may be just as intense as that of the man described above but emerging from a partially psychosocial source.
Finally, the degree of physician control of the death situation seems to interact with patient gender. The very same factors which lead women to utilize social support in health care settings more than men may also affect their preferences for more paternalistic physicians (Emanuel and Emanuel, 1992) and more highly structured death situations where they may be more passive (Canetto and Hollenshead, 2000).
The danger is that such structure will create a sense of obligation on the part of a woman, especially one who subscribes to stereotypic sex roles (Flaherty and Richman, 1989; Vaux and Harrison, 1983; Kaplan, Schneiderhan, and Harrow, 2001) to complete a physician-assisted death towards which she may be initially ambivalent. The role of other patient factors such as disability, age, religiosity, ethnicity, and socioeconomic status must also be examined in the context of physician-assisted death. Research with regard to the effects of the physician-patient relation (Emanuel and Emanuel, 1992; Kaplan, Schneiderhan, and Harrow, 2001) is clearly needed to provide a foundation for intelligent and informed legislation in this very important area.