Prevalence and Content Analysis of Guidelines on Handling Requests for Euthanasia or Assisted Suicide in Dutch Nursing Homes

Background  The growing number of requests for euthanasia or assisted suicide (EAS) makes it imperative for health care institutions, such as nursing homes, to have written guidelines on how to handle requests for EAS. The objective of this study was to determine the prevalence of EAS guidelines in Dutch nursing homes and to analyze the content.

Methods  Directors of patient care in 324 Dutch nursing homes were asked, by means of a mailed short list of questions, if they had an institutional guideline on EAS and, if so, to provide a copy. Guidelines were analyzed according to a structured list of items based on current jurisprudence, model documents, and opinions of experts.

Results  Of the 324 directors, 313 (97%) responded. In 58% of the nursing homes that responded, there existed written guidelines for EAS. Of those guidelines, 74% concerned EAS; in 26%, EAS was integrated in a guideline on terminal care. Of the guidelines, 165 (90%) were based on the policy that EAS is acceptable under specific conditions, and 18 (10%) banned EAS completely. Of the first-mentioned guidelines, 81% described one or more procedures for in-principle objections. In 65% of these guidelines, all official requirements for prudent practice were described.

Conclusions  Despite the rapidly growing number of nursing-home guidelines on EAS and the existence of model documents, there is still considerable variation in the guidelines, and they can be improved in many aspects. A basic prerequisite is that the guidelines include all the official requirements for prudent practice.

THE NEED for practice guidelines on handling requests made by patients for euthanasia or assisted suicide (EAS) is gradually emerging in various countries.^1-5 From several sides, it has been emphasized that Dutch health care institutions should make clear to patients and their relatives and also to physicians, nurses, and other employees how the institution deals with requests for EAS.^1,6-9 This communication is important because of the increasing openness concerning EAS and the growing number of requests for EAS.¹⁰ Also, it is desirable for a nursing home to have written guidelines on how to handle requests for EAS, whether it accepts EAS under specific conditions or bans it completely.

At the beginning of 1995, 74% of Dutch nursing homes had already formulated a written policy on EAS, that is, a statement about the acceptability of EAS.¹¹ In 68% of these nursing homes, EAS was acceptable under specific conditions (a "tolerant" or "permissive" policy), but in 32%, EAS was never acceptable.¹¹ In practice, among all Dutch nursing-home physicians, approximately 300 requests for EAS are received each year, of which about 70 are granted.¹² Approximately 3% of the nursing-home physicians had administered EAS in 1994 and 1995.¹⁰

Although the practice of EAS is officially illegal in the Netherlands, in general, a physician will not be prosecuted if he or she acts in accordance with specific requirements for prudent practice concerning EAS. These requirements (see the "Methods" section) have been formulated by the medical profession and established in jurisprudence.⁹

To our knowledge, no published reports of any other international study on either the prevalence or content of EAS guidelines have been found in the literature. At the beginning of 1995, 36% of Dutch nursing homes had a written guideline on EAS based on the policy that EAS was acceptable under specific conditions.¹¹ Little was previously known about the variety and quality of the content of the Dutch nursing-home guidelines on EAS, apart from limited information on self-reports of nursing-home managers.¹¹ The purpose of this study was to determine again, after 2 years, the prevalence of EAS guidelines and to analyze the content of the guidelines.

This is a study of the institutional EAS guidelines. All Dutch nursing homes (328), involving all psychogeriatric (for patients with dementia), somatic (for patients with physical disorders and disabilities), and combined nursing homes were identified from the 1997 address file of the Dutch Ministry of Health, Welfare and Sports, the Hague. Four nursing homes were excluded, because of either a merger or incorrect address. Data were collected from February 12, 1997, through April 30, 1997. The directors of patient care of the nursing homes, usually a nursing-home physician, were sent a 1-page questionnaire asking them whether they had a guideline for EAS and, if so, to provide a copy of this guideline.

In this study, a guideline for EAS was defined as a written protocol that contained formalized agreements and was authorized by the management to guide physicians and other caregivers on how to handle requests for EAS, including a decision-making process or a phased plan. The definitions used for EAS were those formulated by the Dutch State Commission on Euthanasia.¹Euthanasia was defined as "the act of deliberately terminating the life of another person at his or her own request." Essential elements in the definition of euthanasia are the aspects "voluntary request," "the active termination of life," and "another person (ie, a physician) performs the act." Assisted suicide was defined as "deliberately assisting a person in a life-ending act at his or her request." Life termination without an explicit request was defined as "deliberately terminating the life of a person without his or her explicit request."

The framework is derived from different sources and contains the following elements:

• Dutch law and regulations: (a) EAS is officially illegal in the Netherlands (described in the "Introduction"). (b) Patients rights, since 1995, were embedded in The Medical Treatment Contracts Acts (WGBO).¹³ An important aspect is that informed consent is required from the patient. (c) The Care Institutions Quality Act,¹⁴ which was enacted in 1996, among other things, required institutions to describe their quality system in an annual report. Guidelines can be included as one of the elements in a quality system.

• Jurisprudential requirements for prudent practice: A distinction was made between official requirements for prudent practice (unbearable and hopeless suffering; voluntary, well-considered, and persistent requests; consultation; written report of decision making; and notification) and unofficial requirements (no alternatives for treatment available; information about diagnosis and prognosis; competency; medical-technical performance, eg, the way in which drugs are administered to hasten death; and terminal phase).

• The commonly accepted definitions of euthanasia and assisted suicide (Dutch State Commission on Euthanasia¹; see "Definitions").

• The norms and standards of the medical profession: the viewpoint of the Board of the Royal Dutch Medical Association¹⁵ with regard to euthanasia (including the requirements for prudent practice)⁹ and the model euthanasia guidelines for "physician-nurse."

• Opinions of experts and literature sources, including opinions of the authors themselves.^6-8,11,16

Guidelines were analyzed on the basis of a structured list with items (questionnaire) developed by the authors and based on the framework described herein. The structured list used for the content analyses of the guidelines contains the following main topics: type of guideline (eg, specifically for EAS only or integrated in terminal care policy), definitions used, policy on acceptability, way of handling a request, requirements for prudent practice, type of patients, incompetency of patients, living wills, responsibility for decision making, procedure concerning in-principle objections, people involved in the decision making (eg, nurses and family), consultation of colleagues, the role of the management in EAS, and procedures for solving disagreements.

As the interpretation of some of the items on the questionnaire is potentially subjective, the content analysis was performed by 4 independent reviewers (I.H., M.T.M., F.J.J., and M.C.). Each guideline was judged by 2 reviewers. The initial crude percentages of agreement between each pair of 2 reviewers were based on the total number of variables agreed on, minus the total number of variables disagreed on, divided by the total number of variables contained in the questionnaire. Differences were solved in consensus meetings. The topics of disagreement were extensively discussed, and the guidelines in question were reviewed again. The overall approach of the analyses was descriptive.

Of the 324 nursing homes approached, 313 directors (97%) responded, of which 206 (66%) indicated that in their nursing home a guideline for EAS was present. Only 183 guidelines (58%) were valid for content analyses because 23 documents were policies only (ie, statements about the acceptability of EAS within the institution), and therefore they did not meet the criteria of our definition of a guideline. There was considerable variation in the nature and volume of the guidelines, ranging from 1 page to approximately 35 pages (eg, including mission statements or involving a guideline for terminal care). The initial raw percentages of agreement between the 4 pairs of reviewers were between 87% and 90% before 100% consensus on assessment of the guidelines was reached.

Of the 183 guidelines involved, 18 guidelines (10%) were based on the policy that EAS is never acceptable, and the other 165 guidelines (90%) were based on the policy that EAS is acceptable under specific conditions.

Table 1 shows that 74% of all guidelines included EAS only, and 26% involved a guideline in which the subject of EAS was integrated in a guideline for terminal care policy. In 26% of all guidelines, the subject of assisted suicide was also discussed, and in 7%, the termination of life without the explicit request of the patient was discussed. Furthermore, Table 1 shows that in 90% of the guidelines a definition of euthanasia was given. In 87%, this definition was similar to the definition formulated by the Dutch State Commission on Euthanasia and complied with current jurisprudence.¹ In 21 guidelines (13%), the definitions of euthanasia that were given deviated in one or more aspects: the essential element "on request" was missing in 6 guidelines; "deliberately" (relating to the intention of terminating life) was missing in 2 guidelines, and "of another person" was missing in 11 guidelines. In 6 guidelines, "terminal phase" was added to the definition, and in 1 definition, "withholding action" was included. In 25% of the guidelines, definitions of assisted suicide were given; definitions deviated in only 3 guidelines (2 guidelines added "terminal phase" and 1 added "for physical reasons"). In 19%, definitions of termination of life without the explicit request of the patient were given, all of which were in accordance with the definition of the Dutch State Commission on Euthanasia, except 1, in which the element "deliberately" was missing. In 6 (3%) of the guidelines, a distinction was made between "active" and "passive" euthanasia (no longer customary in the Netherlands).

Of the 165 guidelines, 78% were based on the policy that EAS is acceptable under specific conditions; it was stated that informed consent from the patient was required before other caregivers could be informed (Table 2). In 64% of these guidelines, it was stipulated that the nurse should inform the physician if a patient initiates a discussion concerning EAS. The reverse situation that the physician should inform the nurse was included in fewer guidelines (19%). Twenty-one percent of the guidelines stated that a specific team had to be formed when a patient makes a request for EAS. In 45% of the guidelines, it was explicitly stated that the request must be taken seriously. In 6% of the guidelines, there was no explicit description of how to handle a patient's request for EAS.

In 12 of the 18 guidelines that were based on the policy that EAS is banned within the nursing home, it was stated that the patient should be informed about the possibility of referral to another nursing home. In 5 of the 18 guidelines, it was explicitly stated that the request must be taken seriously. In another 5, no statement was made about the handling of a request for EAS.

In 81% of the guidelines based on the policy that EAS is acceptable under specific conditions, one or more procedures were described for cases in which caregivers have in-principle objections to EAS. Table 2 shows that the most frequently described procedure was that the physician can abstain from decision making or performing EAS (61%). In 44% of the guidelines, it was stated that a physician with in-principle objections is obliged to refer the patient to a colleague; in another 25%, it was stated that the physician has the opportunity to do so.

Table 3 shows that in 96% of the 165 guidelines based on the policy that EAS is acceptable under specific conditions one or more official requirements for prudent practice were described. In 65% of the guidelines, all official requirements were mentioned. The most frequently described requirements were procedural: consultation (93%) and written reports of the decision making (89%). Furthermore, in 90% of the guidelines, unofficial requirements were described, such as information about diagnosis and prognosis (72%). With regard to "consultation of another physician," the following items were described in the 165 guidelines based on the policy that EAS is acceptable: "written report-making" (45%), "the consultant must be from another institution" (35%), "the consultant must have no personal relationship with the patient" (28%), "the consultant must not be a co-attending physician or assistant" (25%), "the consultant must see the patient" (12%), and "a psychiatrist must be involved if there appear to be psychiatric problems" (12%). In 2 guidelines (1%), it was stated that a psychiatrist must always be involved in the decision-making process.

Table 3 shows that, of the 183 guidelines, 18 (10%) were based on the policy that EAS is banned within the nursing home, 4 of the 18 included one or more official requirements for prudent practice, and 3 provided one or more unofficial requirements.

Table 4 shows that in 135 (82%) of the 165 guidelines the role of the nurse was explicitly described. The most frequently described role of the nurse with regard to EAS decision making was "involvement" (without further explication) (32%), followed by "discussion (but with no responsibility for the decision)" (25%). In 50% of the guidelines, the role of the nurse in the administration of EAS was explicitly described, the most frequently mentioned statement being that a nurse was never permitted to administer EAS (32%). Also, Table 4 shows that in 75% of the guidelines the role of the family was described. The most frequently mentioned statement was that discussions must be held with the family (47%).

In 145 (88%) of the 165 guidelines based on the policy that EAS is acceptable under specific conditions, the role of the management was described. The most frequently mentioned statement was "the management must be informed about the planned administration of EAS" (78%). Some other aspects were described: supervision of the procedure (22%), the management must be informed after the administration of EAS (21%), and the management must act in case of publicity (10%). In 3 guidelines, it was stated that in a conflict situation the management makes the final decision, and in 1 guideline it was stated that the management has to give permission for the actual administration of EAS (not included in the tables).

In 118 (72%) of the 165 guidelines based on the policy that EAS is acceptable under certain conditions, the responsibility for the decision concerning EAS was described: in 70%, only the physician was responsible, and in 2%, a special team was responsible. In 76% of the guidelines, it was explicitly stated that only the physician is permitted to administer the lethal drugs.

In 97 (59%) of the guidelines, remarks were made about excluding patients from EAS. Forty-five percent excluded incompetent patients (in general), 32% excluded patients with psychogeriatric conditions, 18% excluded patients in comas, and 1% excluded patients with psychiatric conditions.

In 35% of the guidelines, it was stated that a written request can be used to support the actual request; 30% stated that a written request was obligatory, and 5% stated that only a recent written request was valid. One of these guidelines stated that the written request should be "not older than 1 year" and in another one "not older than half a year," while 7 guidelines did not indicate a time.

In 20 (12%) of the guidelines it was stipulated that, in the case of an incompetent patient, a written living will was only valid "if it had recently been written," and in 2 guidelines (1%), "providing the patient was competent during the writing of the living will." Of the 20 guidelines, 11 specified that a written living will should be not older than 5 years, and 9 guidelines did not define recently. In 2 guidelines, it was stated that a living will was always valid, and 7 (4%) stated that it was never valid. In 134 (81%), no stipulations were made in the guidelines in this respect.

In 8 guidelines it was stated that there must be a discussion included in the admission procedure about the patient's wishes concerning the end of life. This discussion was mentioned in 2 of the 18 guidelines that were based on the policy that EAS is banned within the nursing home.

The major strength of this descriptive study of the prevalence and content of Dutch nursing-home guidelines regarding EAS is its high response rate of 97%, which suggests that the results of the study may be representative of all nursing homes in the Netherlands. Its major drawback is that the content analysis of the guidelines is potentially subjective. An attempt was made to overcome this subjectivity by having 2 reviewers analyze each guideline. The initial percentage of agreement on the assessment of the guidelines among the reviewers varied between 87% and 90%, and all disagreements were solved by consensus.

At the beginning of 1995, 36% of Dutch nursing homes had, according to the directors, a written guideline on EAS based on the policy that EAS was acceptable under specific conditions in the nursing home.¹¹ Two years later, the prevalence of these guidelines had increased to 53%. Despite the rapidly rising number of EAS guidelines, the extent to which they differed from various essential elements stipulated in the written model documents was surprising.

Of the guidelines that included a definition of euthanasia (90%), 13% gave a definition that deviated, in one or more crucial aspects, from the commonly accepted definition of the Dutch State Commission on Euthanasia. Ambiguous terminology should be avoided so that confusion in discussions on this sensitive topic does not occur.

The basic prerequisite of a good guideline on EAS is that it include all the official requirements for prudent practice. In 96% of the guidelines that were based on the policy that EAS is acceptable under specific conditions, one or more official requirements were described, while all the official requirements were mentioned in only 65%. Since 1984, the Royal Dutch Medical Association⁹ no longer considers the "terminal phase" to be feasible for the acceptability of euthanasia, a view that is supported by jurisprudence. However, this informal requirement was mentioned in 19% of the guidelines and sometimes even formed part of the definition.

The value attached to the medical-ethical aspects concerning the autonomy of the patient varies. The "request for EAS must be voluntary" and "informed consent is required from the patient before other caregivers can be informed" were described in the majority of the guidelines (86% and 78%, respectively). However, the unofficial requirement of "competency of the patient" was mentioned in only 42% of the guidelines. Furthermore, only 45% of the guidelines explicitly stated that the request "must be taken seriously." Thus, the guidelines can be improved with regard to various aspects related to autonomy.

A current topic of discussion is the living will, in which a person can request active life termination in a future situation when he or she is no longer able to communicate this wish. In this study, 13% of the guidelines stated that written living wills were only valid under certain conditions, and 4% stated that they were never valid. The viewpoint of the Royal Dutch Medical Association is that, in principle, if all the requirements for prudent practice are met, the living will can be considered a valid declaration that can form a basis for the acceptability of active life termination. However, physicians, lawyers, and nursing-home policy makers clearly have reservations with regard to this subject, and there is no relevant jurisprudence yet.¹³

A patient has a right to request EAS but no right to have this request granted.¹³ Since EAS is not part of normal medical practice, every physician has the right to reject EAS on the grounds of principle. However, according to professional medical standards, physicians should communicate their opinion on EAS, if requested, to their patients.⁹ In 29% of the guidelines, it was explicitly stated that caregivers must inform the patient about their possible objections. Furthermore, physicians with in-principle objections should at least give the patient the opportunity to contact a colleague who has no such objections, as was stated in 44% of the guidelines. Both aspects are in need of improvement in the guidelines.

The position of the nurse with regard to EAS deserves extra attention.^15-18 Important elements of the Dutch model guidelines are that the decision making falls under the responsibility of the physician and that the actual administration of EAS must be performed by a physician.¹⁴ These aspects were mentioned in 70% and 76%, respectively, of the guidelines evaluated in this study. In our opinion, the implication that nurses are not permitted to administer the lethal drugs, which was only stated in 32% of the guidelines, should be stated explicitly.

Good communication between physician, patient, and family members is an essential element of medical practice, especially concerning the end of life. In most cases, the family will be closely involved with the request for EAS. Indeed, 75% of the guidelines included the aspect of communication with the family, mainly by stipulating that discussions must be held with the family (47%).

Guidelines might be a helpful tool in the complex EAS decision-making process and could provide a standard for the quality of practice. In cases of EAS, physicians are confronted with complicated ethical, emotional, and practical issues for which guidelines might provide some support. However, the mere presence of guidelines does not guarantee that they are actually followed or that they contribute to improvement in the quality of care. On the basis of this study, it is not known how the guidelines were developed or implemented. In an earlier study,¹⁹ carried out in 1995 and 1996, compliance of physicians with EAS guidelines was estimated to be 75%. Moreover, 82% of the physicians who used the guidelines found them supportive. However, more research is needed to gain insight into the additional value of the guidelines in practice.

Although the quality of the existing guidelines for EAS in the nursing homes seems to be promising, they can be improved in many aspects. We recommend that the definition of EAS should comply with the commonly accepted definition formulated by the Dutch State Commission on Euthanasia.¹ In our opinion, the guidelines should be explicit about implications; for example, with regard to the role of the nurse in the administration of EAS. Guidelines should at least include the requirements for prudent practice and provisions for in-principle objections of caregivers. When institutions are considering the development of a guideline on EAS they should first examine the existing model documents to determine whether adaptation to the local situation is desirable. Because of ongoing changes and developments in medical practice, in jurisprudence, and possibly in legislation, the guidelines should be updated on a regular basis.

Accepted for publication May 27, 1999.

This study was funded by the Ministry of Public Health, Welfare and Sports, Amsterdam, the Netherlands.

Corresponding author: Ilinka Haverkate, PhD, Vrije Universiteit, Institute for Research and Extramural Medicine, Van der Boechorststraat 7, 1081 BT, Amsterdam, the Netherlands (e-mail: I.Haverkate.gpnh@med.vu.nl).