|
|
| |||||||||||||||||||||||||||||||||||||||||||||
Background and Methods The characteristics and frequency of clinical problems with the performance of euthanasia and physician-assisted suicide are uncertain. We analyzed data from two studies of euthanasia and physician-assisted suicide in the Netherlands (one conducted in 1990 and 1991 and the other in 1995 and 1996), with a total of 649 cases. We categorized clinical problems as technical problems, such as difficulty inserting an intravenous line; complications, such as myoclonus or vomiting; or problems with completion, such as a longer-than-expected interval between the administration of medications and death.
Results In 114 cases, the physician's intention was to provide assistance with suicide, and in 535, the intention was to perform euthanasia. Problems of any type were more frequent in cases of assisted suicide than in cases of euthanasia. Complications occurred in 7 percent of cases of assisted suicide, and problems with completion (a longer-than-expected time to death, failure to induce coma, or induction of coma followed by awakening of the patient) occurred in 16 percent of the cases; complications and problems with completion occurred in 3 percent and 6 percent of cases of euthanasia, respectively. The physician decided to administer a lethal medication in 21 of the cases of assisted suicide (18 percent), which thus became cases of euthanasia. The reasons for this decision included problems with completion (in 12 cases) and the inability of the patient to take all the medications (in 5).
Conclusions There may be clinical problems with the performance of euthanasia and physician-assisted suicide. In the Netherlands, physicians who intend to provide assistance with suicide sometimes end up administering a lethal medication themselves because of the patient's inability to take the medication or because of problems with the completion of physician-assisted suicide.
The incidence of physician-assisted suicide and euthanasia and attitudes toward these practices have been studied extensively, but the few reports on the clinical aspects of these practices are based on limited data or small numbers of cases.14,15,16,17,18 We performed a study to determine whether there are problems with the clinical aspects of euthanasia and physician-assisted suicide as reported by the physicians involved, including complications and problems with completion, such as a prolonged interval between the administration of medications and the patient's death.
Methods
Study Design
In 1990 and 1991 and in 1995 and 1996, we performed two studies of euthanasia, physician-assisted suicide, and other medical practices involving the end of life in the Netherlands. Detailed information about the design of these studies has been reported elsewhere.1,19,20 In three parts of the studies, detailed information on the clinical aspects of euthanasia and physician-assisted suicide was collected.
In 1990 and 1991, we interviewed a stratified random sample of 405 physicians that included 152 general practitioners, 50 nursing home physicians, and 203 physicians in the specialties of cardiology, surgery, internal medicine, pulmonology, and neurology. Nine percent of eligible physicians declined to take part in the study. The interviews were based on an extensive questionnaire and lasted an average of 2.5 hours. Respondents were asked whether they had ever performed euthanasia or provided assistance with suicide. If the answer was yes, more detailed questions were asked about the most recent case. Some of the questions were about technical problems, complications, and the length of the interval between the administration of medications and death.
In 1995 and 1996, we replicated the earlier interview study. We interviewed a randomized stratified sample of 405 physicians, including 124 general practitioners, 74 nursing home physicians, and 207 physicians in the specialties noted above. Eleven percent of eligible physicians declined to take part in the study. The questionnaire was virtually identical to the one used in the 1990–1991 study.
In 1995 and 1996, we also interviewed a random sample of 147 physicians who had reported a case of physician-assisted suicide or euthanasia for judicial review by the public prosecutor between August 1, 1994, and February 1, 1995. Ten percent of eligible physicians declined to be interviewed. We asked the physicians questions about the reported case and the most recent unreported case (if any). In addition, we interviewed 63 of 64 eligible physicians who had been involved in cases of physician-assisted suicide or euthanasia discussed in the Assembly of Prosecutors General between January 1, 1991, and July 1, 1995. One physician declined to be interviewed. The questionnaire was similar to the one used in the 1990–1991 and 1995–1996 interview studies.
The protocol for the studies was reviewed and approved by a special committee of the Dutch Ministry of Health, Welfare, and Sports and the Dutch Ministry of Justice. It was also approved by the Royal Dutch Medical Association.
To obtain data from a sufficiently large number of cases, we combined all cases of euthanasia and physician-assisted suicide for which information on clinical problems was available. There were 649 such cases: 185 from the 1990–1991 interview study, 194 from the 1995–1996 interview study, and 270 from the 1995–1996 reported-cases study. These cases were divided into two groups according to whether the physician's intention had been to provide assistance with suicide or to perform euthanasia. Cases in which all the lethal drugs were administered by the patient were assigned to the assisted-suicide group, as were cases in which the last drug was administered by another person, although the intention had been that the patient would administer all the drugs. Cases in which someone other than the patient administered at least one of the lethal drugs at the patient's explicit request were assigned to the euthanasia group.
The data we used did not represent a random sample of cases, since we combined cases from different (stratified) samples and asked physicians about only the most recent case. We performed three analyses to assess the degree to which our data were representative of all cases of euthanasia or assisted suicide in the Netherlands. We compared the prevalence of technical problems, complications, and problems with completion between samples and according to the type of practice. We also analyzed the probability of a problem according to the number of cases of physician-assisted suicide or euthanasia that physicians had been involved in during the two years before the interview. Finally, we repeated all our analyses with only the data from the 1990–1991 interview study and the 1995–1996 interview study, which were both based on representative samples. The prevalence of technical problems, complications, and problems with completion was virtually identical for all samples and for all types of practice, and the likelihood of a problem was not related to the physician's prior experience with physician-assisted suicide or euthanasia. Therefore, we report the results for the overall samples.
We categorized clinical problems as technical problems, such as difficulty inserting an intravenous line; complications, such as spasm, myoclonus, nausea, or vomiting; and problems with completion, such as a longer-than-expected time to death or failure to induce coma. If more than one type of medication was administered, the combination of drugs was classified according to the drug with the greatest potential for causing death. Neuromuscular relaxants and potassium chloride were considered to have the greatest lethal potential, followed by barbiturates, and then opioids. Thus, for example, if a barbiturate and a neuromuscular relaxant had been administered, the medication was classified as a neuromuscular relaxant, and if a barbiturate and an opioid had been administered, the medication was classified as a barbiturate.
Statistical Analysis
The chi-square test was used for comparisons of categorical variables. Two-sided P values of less than 0.05 were considered to indicate statistical significance. All statistical analyses were performed with SPSS software for Windows, version 7.5 (SPSS, Chicago).
Results
The characteristics of the 649 cases of euthanasia and assisted suicide are shown in Table 1. Fifty-five percent of the patients were men, and the mean age was 63.9 years (range, 21 to 96). Seventy-five percent of the patients had cancer. In 535 cases (82 percent), the physician intended to perform euthanasia, and in 114 cases (18 percent), the intention was to provide assistance with suicide.
|
In 367 of the 535 cases in which euthanasia was intended (69 percent), a neuromuscular relaxant was given (Table 2), in most cases after the administration of a barbiturate to induce coma. In 10 percent of the euthanasia cases, a barbiturate was the only or the most potent lethal drug administered; in 13 percent, an opioid was the only or most potent lethal drug administered. Potassium chloride was administered in 2 percent of the euthanasia cases. In 485 cases of euthanasia (91 percent), a physician administered all or some of the medications. In 29 cases (5 percent), a physician did not administer any of the drugs: a nurse administered the drugs in 23 cases, and another person (generally a family member) did so in 6 cases. In 384 cases of euthanasia (72 percent), the physician who completed the interview was present continuously from the time the first drug was administered until the patient's death. In 10 cases (2 percent), the respondent was not present at all; instead, a colleague or a nurse administered the medication.
|
Clinical Problems
Table 3 shows the frequency of the clinical problems that occurred. Technical problems occurred in 35 cases (5 percent), complications in 24 cases (4 percent), and problems with completion in 44 cases (7 percent). In 10 cases, more than one type of problem occurred. Assisted suicide was more frequently associated with each type of problem than was euthanasia (P=0.03, P=0.03, and P=0.001, respectively). General practitioners and nursing home physicians were more likely than specialists to report technical problems (P<0.001) and problems with completion (P=0.04). Oral or rectal administration of medications was more likely than parenteral administration to be associated with technical problems (P=0.003) and problems with completion (P<0.001).
|
|
|
In 21 of the cases in which the intention had been to provide assistance with suicide, the physician administered the lethal drug. In 12 of these cases, a problem with completion was mentioned as the reason for the physician's intervention: the interval between the administration of medication and the occurrence of coma or death was too long (in 9 cases), the patient did not become comatose after taking the first drug (in 1), or the patient awoke from coma (in 2). In five other cases, the physician intervened because the patient had difficulty taking all the oral medication; for example, one patient vomited after taking the first medication, and one patient fell asleep before taking all the medication. In four cases, no specific reason for the physician's intervention was given.
Discussion
We found that there were sometimes clinical problems with the performance of physician-assisted suicide and euthanasia. The problems reported by physicians frequently concerned a longer-than-expected interval between the administration of the lethal drug (or the first drug, if more than one was administered) and the patient's death. In general, problems were more frequently reported in cases of assisted suicide than in cases of euthanasia. The responsible physician decided to administer the lethal medication in 21 of 114 cases in which the original intention had been only to provide assistance with suicide. In most of these cases, the patient did not die as soon as expected or awoke from coma, and the physician felt compelled to administer a lethal injection because of the anticipated failure of the assisted suicide. In some cases, the physician administered a lethal injection because the patient had difficulty swallowing the oral medication, vomited after swallowing it, or became unconscious before swallowing all of it.
Two limitations of our study should be considered. First, we included all cases of euthanasia and physician-assisted suicide for which we had clinical information. These cases may not be representative of all cases of euthanasia and assisted suicide in the Netherlands. We performed several analyses to check for differences between samples or strata, which did not yield any significant differences. Second, the information was provided by the responsible physician in each case, who may have underestimated the number or seriousness of problems.
In two other Dutch studies of the clinical aspects of euthanasia and physician-assisted suicide, general practitioners and nursing home physicians reported problems in 12 percent and 7 percent of cases, respectively.14,15 In a study of physician-assisted suicide in the United States, Emanuel et al. reported that assisted suicide failed in 3 of 20 cases described by American oncologists.16 In a study in Washington State, no unsuccessful attempts at physician-assisted suicide were reported.17 In 15 cases of physician-assisted suicide in Oregon, no complications, such as vomiting or seizures, occurred.18
Information on problems with the performance of euthanasia or physician-assisted suicide is relevant to the debate about whether to legalize these practices, for several reasons. First, such information may be helpful in cases in which the distinction between physician-assisted suicide and euthanasia is unclear. We found that in most cases in which assistance with suicide had been intended, the physician's role was confined to prescribing or supplying the drugs to be used. In 21 percent of the cases, however, the physician or another person helped the patient take one or more of the drugs. It is not clear at what point a physician's assistance with suicide becomes euthanasia. For example, how should cases be classified in which the physician pours an oral medication into the patient's mouth because the patient is unable to administer it? A judicial distinction between physician-assisted suicide and euthanasia may be complicated by such practical issues.21
Second, if physician-assisted suicide is legalized but euthanasia is not, some competent patients may not be able to end their own lives for purely physical reasons, as in the case of patients with neurologic illnesses who have problems with swallowing or using their hands and patients who are physically too weak to take all the oral medication themselves. One study found that general practitioners and nursing home physicians preferred euthanasia to assisted suicide because of limitations imposed by the patient's condition in 48 percent of 155 cases of euthanasia and in 78 percent of 50 cases, respectively.22
Third, a physician who decides to assist with a patient's death should have adequate knowledge of the technical performance of euthanasia and physician-assisted suicide. In the Netherlands, one of the requirements for prudent practice concerns care in choosing doses and administering medications.23 The Royal Dutch Association of Pharmacy recognized in the 1980s that physicians and pharmacists should have at their disposal reliable information on the technical performance of euthanasia, and the association regularly updates its guidelines on the use and preparation of relevant drugs.12,13 Studies in the United States and in the Netherlands have shown that many physicians lack knowledge about the use of lethal drugs and that recommendations about which drugs to use are not always followed.2,13,16,17,18,24,25
Finally, the responsible physician's presence must be considered. According to the Royal Dutch Medical Association, the physician should be present when euthanasia or physician-assisted suicide is carried out. A physician who decides not to be present (e.g., at the patient's request) should make arrangements to be available if the prescribed drugs fail to have the intended effect.26 About two thirds of both general practitioners and nursing home physicians in the Netherlands endorse the statement that a physician who provides assistance with suicide should be prepared to administer a lethal drug if the suicide attempt fails.27 To be available under such circumstances, the physician must have previously informed the patient and the patient's relatives about what to expect during the actual performance of physician-assisted suicide, including possible complications and the expected interval between the administration of the medication and death.
Decisions about physician-assisted suicide and euthanasia are difficult and emotional for all persons involved. Unexpected events can be traumatic. Professional training, empirical research, and an open debate on the practical aspects of physician-assisted suicide and euthanasia may contribute to the prevention of complications and other clinical problems.
Supported by a grant from the Dutch Ministry of Health, Welfare, and Sports and the Dutch Ministry of Justice.
We are indebted to Johannes J.M. van Delden and Loes Pijnenborg, for their contributions to the 1990–1991 study; to Ilinka Haverkate and Jacqueline M. Cuperus-Bosma, for their contributions to the 1995–1996 study; to Caspar W.N. Looman, for his statistical advice; to all the physicians who participated in the two studies; and to the interviewers.
Source Information
From the Department of Public Health, Erasmus University, Rotterdam (J.H.G., A.H., P.J.M.); and the Institute for Research in Extramural Medicine and the Department of Social Medicine, Vrije Universiteit, Amsterdam (B.D.O.-P., D.L.W., G.W.) — both in the Netherlands.
Address reprint requests to Dr. Groenewoud at the Department of Public Health, Erasmus University, Rotterdam, P.O. Box 1738, 3000 DR Rotterdam, the Netherlands, or at groenewoud{at}mgz.fgg.eur.nl.
References
| |||||||||||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
