Jansen-van der Weide MC, Onwuteaka-Philipsen BD, van der Wal G. Granted, Undecided, Withdrawn, and Refused Requests for Euthanasia and Physician-Assisted Suicide. Arch Intern Med. 2005;165(15):1698-1704. doi:10.1001/archinte.165.15.1698
The aims of this study were to obtain information about the characteristics of requests for euthanasia and physician-assisted suicide (EAS) and to distinguish among different types of situations that can arise between the request and the physician’s decision.
All general practitioners in 18 of the 23 Dutch general practitioner districts received a written questionnaire in which they were asked to describe the most recent request for EAS they received.
A total of 3614 general practitioners responded to the questionnaire (response rate, 60%). Of all explicit requests for EAS, 44% resulted in EAS. In the other cases the patient died before the performance (13%) or finalization of the decision making (13%), the patient withdrew the request (13%), or the physician refused the request (12%). Patients’ most prominent symptoms were “feeling bad,” “tiredness,” and “lack of appetite.” The most frequently mentioned reasons for requesting EAS were “pointless suffering,” “loss of dignity,” and “weakness.” The patients’ situation met the official requirements for accepted practice best in requests that resulted in EAS and least in refused requests. A lesser degree of competence and less unbearable and hopeless suffering had the strongest associations with the refusal of a request.
The complexity of EAS decision making is reflected in the fact that besides granting and refusing a request, 3 other situations could be distinguished. The decisions physicians make, the reasons they have for their decisions, and the way they arrived at their decisions seem to be based on patient evaluations. Physicians report compliance with the official requirements for accepted practice.
In many areas of the world physicians have to deal with requests for euthanasia and physician-assisted suicide (EAS).1- 8 In Oregon, Belgium, and the Netherlands, physicians are allowed to perform euthanasia (Belgium and the Netherlands) or physician-assisted suicide (Oregon and the Netherlands).9- 11 The Dutch notification procedure stipulates the official requirements for accepted practice that the patient’s situation should meet for the physician not to be prosecuted (no alternatives for treatment, well-considered explicit request, unbearable and hopeless suffering, and consultation of a colleague).11 Therefore, in deciding whether to grant a request, physicians should determine whether these requirements are met and be reluctant to grant a request if the patient’s situation does not meet these requirements. Not meeting the requirements should be a reason for refusing a request. In the time between an EAS request and a decision, several types of situations can arise.3,12 Patients can die a natural death before the performance of EAS or before the final decision is made or the patient can withdraw the request. Although several studies,3,13- 21 especially in the United States and the Netherlands, have focused on the characteristics of patients who request EAS and the reasons for their request, only one study3 distinguished between these different possible situations, albeit without describing the characteristics of the different groups.
This study aims to obtain information about the characteristics of all requests for EAS and distinguish among patients who died of EAS, patients whose request was granted but who died before the performance of EAS, patients who died before the physician had made the final decision, patients who withdrew their request, and patients whose request was refused. Beyond estimating the proportion of these groups within the total group of requests, the focus is on differences in patient characteristics among the different groups, reasons for requesting EAS, and the extent to which the patient’s situation meets the official requirements of accepted practice. Finally, factors associated with refusing a request are distinguished.
The data used in this study were collected for the project Support and Consultation on Euthanasia in the Netherlands (SCEN). In this project, general practitioners (GPs) receive specific training in formal consultation and giving expert advice to colleagues who have questions about EAS.22 This retrospective study is based on data from the posttest, which consisted of a written questionnaire sent to all GPs in GP districts in which the project was implemented before August 2001 (18 of the 23 GP districts in the Netherlands).
Data collection occurred per GP district from April 2000 to December 2002 for 1 ½ years after the start of the project. Of the 6596 GPs who received a questionnaire, 556 were not eligible for participation, in most cases because the GP was no longer in practice and in some cases owing to prolonged disability for work. In total, 60% of the GPs returned the questionnaire (n = 3614). All questionnaires returned by the GPs were processed anonymously. The responding physicians were compared with all Dutch GPs on the following variables: sex (men: 69% vs 72%), age (<40 years: both 21%; 40-50 years: 40% vs 43%; >50 years: 39% vs 36%), and type of practice (solo practice: 39% vs 42%; duo practice: both 33%; and group practice: 25% vs 27%).
The questionnaire consisted of questions commonly used in end-of-life research in the Netherlands.12,16,23 To calculate the precise number of requests received in 1 year, the questionnaire contained questions about the amount of requests received and the way the GP dealt with them in the previous 12 months. Furthermore, the questionnaire contained questions about the characteristics of the most recent explicit request within the last 1 ½ years (if any). They answered questions regarding a patient’s reasons for requesting EAS, symptoms, the extent to which the patient’s situation met the official requirements for accepted practice, and the decision-making process. The GPs were asked to base their answers on the patient’s situation at the end of their decision-making process.
Logistic regression analysis was used to identify variables associated with refusing a request. A selection was made of the cases in which there was no doubt about the GP’s decision: cases of EAS vs refused requests. All possible factors were analyzed using univariate logistic regression. Then, all significant variables (P<.05) were included in a multivariate analysis. Stepwise backward logistic regression was applied to make a predictive model. Variables were removed if P>.05 (2-tailed, unpaired). Since having cancer is probably not a reason in itself to grant or refuse a request but is associated with the clinical characteristics of the patient, a diagnosis of cancer was not considered as a possible factor. Internal validation of the final model that was constructed took place by splitting the sample in half. The first half was used as a calibration sample, constructing a model in the same way as in the total sample. The second half was used as a validation sample by entering only the variables that were retained in the multiple stepwise backward analysis of the first sample. The analyses for the total sample resulted in a model with 12 variables, whereas the calibration sample consisted of 9 variables that all also were part of the final model in the total sample. All but 1 of these variables were also significant in the validation sample.
Of the respondents, 78% had ever received a request for EAS from a patient. In the 12-month period before they filled out the questionnaire, 55% had not received a request, 26% had received 1 request, 17% had received requests from 2 or 3 patients, and 2% had received requests from more than 3 patients. This resulted in 2658 explicit requests for EAS received by the 3614 respondents in the 12-month period. Of these requests, 44% resulted in EAS and 13% were granted but the patient had died before the performance. In another 13% of the requests, the patient died before the GP had made the final decision to grant or refuse the request. In 12% of the cases, the GP refused the request. In 13% of all requests, the patient no longer wanted EAS. Finally, in 3% of the requests, the patient was still alive and the decision-making process was still ongoing, and in 2% it was not clearly stated by the respondent why EAS was not performed (data not shown). Of all respondents, 1681 described their most recent request for EAS that they received at most 1½ year before filling out the questionnaire.
Most patients who requested EAS were between the ages of 40 and 79 years (Table 1). Patients whose requests were refused were more often older than 80 years. Compared with other groups, patients whose requests had been refused less frequently had cancer (57% vs 84%-90%). Compared with other groups, feeling bad (92%-77%), tiredness (89%-70%), pain (50%-21%), nausea (42%-15%), difficulty with breathing (34%-11%), and vomiting (25%-16%) mostly occurred in patients whose request was granted and administered and least in patients whose request was refused. Anxiousness and depression occurred the least in EAS cases (13% and 9% vs 24%-30% and 16%-32%). The 3 most frequently mentioned important reasons for requesting EAS were pointless suffering, loss of dignity, and weakness, but the frequency differed per group, being highest in the EAS group and lowest in the refused requests group (pointless suffering, 75% vs 38%; loss of dignity, 69% vs 51%; and tiredness, 60% vs 39%). Not wanting to be a burden on the family, tired of living, and depression were mentioned more often by patients whose request had been refused (33%, 37%, and 18%, respectively), followed by patients who changed their mind (27%, 22%, and 10%, respectively), than by patients in the other groups (13%-18%, 14%-15%, and 4%-8%, respectively).
No alternatives for palliative treatment existed in 74% of all explicit requests, ranging from 88% in EAS cases to 32% in refused requests (Table 2). Requests were mostly completely explicit or to a high degree explicit, ranging from 100% in EAS cases to 74% in patients who withdrew their request. Most of the patients were fully competent; however, patients who died before the final decision, patients who no longer wanted EAS, and patients whose request was refused were sometimes not (fully) competent (15%, 20%, and 29% vs 0% and 3%, respectively). The patient’s unbearable suffering to a (very) high degree differed among the groups, ranging from 93% for EAS cases to 30% for refused requests.
A formal consultation with a colleague is an official procedural requirement for accepted practice. In the group of EAS cases and the group of patients who died before the performance, the GPs had more often had a formal consultation than in the other 3 groups (97% and 84% vs 66%-72%, respectively) (Table 3).
At the moment of requesting a formal consultation, GPs were already certain about whether to grant the request in 3 of the groups (EAS cases, 95% [probably] yes; patients died before performance, 79% [probably] yes; refused requests, 73% [probably] no). With regard to requests from patients who died before the final decision and requests from patients who no longer wanted EAS, the answers were more evenly spread. The GPs (in all 5 groups) who at the time of the consultation had not already decided to grant the request most frequently mentioned “doubts about hopeless and unbearable suffering” (34%), “availability of alternatives for treatment” (27%), and “doubts of the physician in this particular case” (23%) as reasons for their reluctance. For patients who no longer wanted EAS and patients whose request was refused, the GPs also frequently mentioned “doubts about well-considered and persistent request” (25% and 18%) or depression (22% and 29%). In cases in which the patients died before the final decision, the GPs more frequently mentioned “pressure of next-of-kin on request” (21%). In EAS cases and cases in which the patient died before the performance, most consultants concluded that the patient’s situation met the official requirements for accepted practice (96% and 84%). In contrast, for patients who died before the final decision and patients who no longer wanted EAS, they concluded this in 36% and 41% of the cases, respectively, and in refused requests in only 11% (Table 3).
Table 4 gives the predictive model for refusing a request for EAS. The strongest positive associations with refusing a request were “the patient being not (fully) competent” (odds ratio [OR], 21; 95% confidence interval [CI], 4.2-1096), “a lesser extent of unbearable suffering” (OR, 15; 95% CI, 6.9-33), “a lesser extent of hopeless suffering” (OR, 11; 95% CI, 3.3-39), and “the availability of alternatives for treatment” (OR, 4.4; 95% CI, 2.1-9.1). Three reasons for requesting EAS made it less likely for a GP to refuse a request: fear of suffocation (OR, 0.24; 95% CI, 0.08-0.73), pointless suffering (OR, 0.30; 95% CI, 0.15-0.61), and loss of dignity (OR, 0.32; 95% CI, 0.15-0.67).
In total, 44% of the explicit requests for EAS were granted and EAS was performed. This percentage is in accordance with the results of another Dutch study.23 It is much higher than in Oregon, where it was 10%. Approximately 1 of 2 requests is refused in Oregon, compared with approximately 1 of 8 in the Netherlands. For the other 3 groups the percentages found are similar in both places.3 This might be partly explained by the fact that patients in Oregon less frequently had cancer and more frequently felt that they were a burden to the family and partly by the less positive attitude toward EAS of physicians in Oregon. Although 37% of the physicians in Oregon who received a request would never be willing to prescribe a lethal medication, only 12% of all Dutch physician would never perform EAS.3,23
The most prominent symptoms at the end of the decision-making period were feeling bad, tiredness, and lack of appetite. The most frequently mentioned reasons for requesting EAS were pointless suffering, loss of dignity, and weakness. Not wanting to be a burden on the family, tired of living, and depression were more often reasons to request EAS in refused requests than in other groups. The patients’ situation met the official requirements for accepted practice most in EAS cases and least in refused requests. Before consulting a colleague, most GPs were already certain about whether to grant the request. Reasons most frequently mentioned for reluctance to grant a request were “doubts about hopeless and unbearable suffering,” “availability of alternatives for treatment,” and “doubts of the physician in this particular case.”
The complexity of EAS decision making is reflected by the fact that besides requests granted and requests refused, 3 other situations could be distinguished. Requests that resulted in EAS and refused requests most clearly differ from each other. Patients who died before the performance seem to resemble patients who received EAS concerning the patient characteristics and reasons for requesting EAS. Since in this group the patient’s situation met the official requirements for accepted practice somewhat less, it is possible that the patients generally did not die before the performance because they were more ill but rather because the decision-making process took more time (possibly because these patients seem to have more mental health problems than patients who received EAS). Similarities, especially in symptoms, were apparent in patients who died before the final decision and patients who withdrew their request, although patients of the latter group generally seem to have a less strong will to die. It is possible that reluctance of a physician in granting a request sometimes influences the patient to withdraw a request, since these groups are alike when looking at the frequency of consultation and doubts about granting the request. Unfortunately, we have no information on the background of the decision of the patients to withdraw their request and of the kind and extent of negotiation between patients and physician about the request. In any case, the fact that approximately 1 of 8 people who request EAS change their mind indicates that it is extremely important to continue ensuring that EAS is really what the patient wants until the moment when EAS is finally performed.
The complexity of the EAS decision-making process is also indicated by the fact that physicians of patients who received EAS mentioned aspects that caused them to doubt granting requests, albeit to a lesser extent than in the other situations. Physicians also consult colleagues when still deciding on granting or refusing a request or, although somewhat less frequently, when they are doubtful of whether they should grant the request. This is an improvement compared with the period between 1990 and 1995, when consultation took place in only 16% of refused requests.16
From the predictive model it is clear that not being (fully) competent and a lesser degree of unbearable and hopeless suffering have the strongest association with refusing a request for EAS. The variables associated with refusing a request predominantly concerned the official requirements for accepted practice. Depression seems to be associated with refusing a request, probably because it causes doubts about the extent to which the request is well considered. Fear of suffocation and vomiting make it less likely for a physician to refuse a request, possibly because these symptoms are associated with unbearable suffering or with death being very near. These symptoms have also been found to be important in the decision-making process in a US study,13 although pain was also mentioned in this study. It is noticeable that younger physicians are more likely to refuse a request. The increased attention to palliative care in the Netherlands in the past decade might have made physicians who completed their training during this period more aware of additional treatment alternatives. However, it is also possible that this is not a cohort effect but an age effect.
The SCEN project, of which this study is a part, is nationwide, and almost all the GPs in the Netherlands received a questionnaire. Therefore, data on many cases are available. In addition to increasing the power of the present study, this also made it possible to obtain information on all groups of requests. Although the response rate was 60% and a nonresponse study was not feasible, the comparison between participating GPs and all Dutch GPs suggests that responding GPs did not differ from nonresponding physicians concerning sex, age, and type of practice. The percentages of missing data (well less than 5% for almost all questions) are comparable to studies with less controversial data. Possibly this is due to the Dutch situation, in which euthanasia is regulated.
This study has several limitations. The study focuses only on GPs. However, GPs receive the most requests for EAS.23 Furthermore, the study is retrospective, but a request for EAS is an exceptional occurrence, and the GPs were asked to recall a maximum of 1 ½ years. Another limitation is that the GPs provided the information about patient characteristics and patients’ reasons for requesting EAS, so it is possible that their decision influenced their description of the patient characteristics. It might be that the degree of depression is underestimated, because patients might censor themselves if they realize that mentioning depression could disqualify them from EAS. This possible effect could not be detected, since no objective depression scale was used. Finally, the specific Dutch situation concerning EAS may limit the generalizability to other countries. Although the various studies were based on different methods, populations, and questions, it appears that there are great similarities between patients who request EAS in the Netherlands and in the United States. Most patients have cancer, are relatively infrequently older than 80 years, and have many physical symptoms (approximately 1 in 5 has depressive symptoms). Important reasons for requesting EAS are loss of dignity, pointless suffering, and tiredness. However, there also seem to be differences, with patients who request EAS in the United States somewhat less frequently having cancer and more frequently requesting EAS because of being a burden to the family than in the Netherlands.3,21
In conclusion, this study shows that the decisions physicians make, the reasons they have for their decisions, and the way they arrive at their decisions seem to be based on patient evaluations. Physicians report compliance with the official requirements for accepted practice. This is a prerequisite for the safeguarding of the practice of EAS.
Correspondence: Bregje D. Onwuteaka-Philipsen, PhD, VU University Medical Centre, Institute for Research in Extramural Medicine (EMGO)/Department of Public and Occupational Health, Van der Boechorststraat 7, 1081 BT Amsterdam, the Netherlands (firstname.lastname@example.org).
Accepted for Publication: November 16, 2004.
Financial Disclosure: None.
Funding/Support: This study was funded by the Royal Dutch Medical Association (Utrecht) and the Dutch Ministry of Health, Welfare, and Sports (The Hague).
Acknowledgment: We are indebted to the thousands of GPs who provided the data.