Palliative http://pmj.sagepub.com/

End-of-life practices in palliative care: a cross sectional survey of physician members of the German Society for Palliative Medicine Jan Schildmann, Julia Hoetzel, Christof Mueller-Busch and Jochen Vollmann Palliat Med 2010 24: 820 originally published online 6 September 2010 DOI: 10.1177/0269216310381663

The online version of this article can be found at: http://pmj.sagepub.com/content/24/8/820

Published by:

http://www.sagepublications.com

Additional services and information for Palliative Medicine can be found at:

Email Alerts: http://pmj.sagepub.com/cgi/alerts

Subscriptions: http://pmj.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://pmj.sagepub.com/content/24/8/820.refs.html

Downloaded from pmj.sagepub.com at Kings College London - ISS on January 22, 2011 Original Article

Palliative Medicine 24(8) 820–827 ! The Author(s) 2010 End-of-life practices in palliative care: a Reprints and permissions: sagepub.co.uk/journalsPermissions.nav cross sectional survey of physician DOI: 10.1177/0269216310381663 members of the German Society for pmj.sagepub.com Palliative Medicine

Jan Schildmann Institute for Medical Ethics and of Medicine, -University , Malakowturm-Markstr. 258a, 44799 Bochum, Julia Hoetzel Institute for Medical Ethics and History of Medicine, Ruhr-University Bochum, Malakowturm-Markstr. 258a, 44799 Bochum, Germany Christof Mueller-Busch German Society for Palliative Medicine; University Witten/Herdecke, Witten, Ru¨sternallee 45, 14050 , Germany Jochen Vollmann Institute for Medical Ethics and History of Medicine, Ruhr-University Bochum, Malakowturm-Markstr. 258a, 44799 Bochum, Germany

Abstract Objectives: To elicit types and frequencies of end-of-life practices by physician members of the German Society for Palliative Medicine. To analyse associations between characteristics of physicians and patients and end-of-life practices with intended hastening of death. Design: Cross-sectional postal survey. Main outcome measures: Types and frequencies of end-of-life practices with foreseeable or intended hastening of patients’ death. Association between end-of-life practices with hastening of death and predefined characteristics of physicians and patients. Results: Nine hundred and one physicians participated in the study (response rate: 55.8%). There was alleviation of symptoms in 78.1% and limitation of medical treatment with possible life shortening in 69.1% of cases. In 10 cases medication had been administered by the physician (N ¼ 9) or the patient (N ¼ 1) with the intention to hasten death. Patients’ best interest and avoidance of possible harm to the patient were reported as reasons for non-involvement of competent patients in decision making. Physicians with added qualification in palliative medicine significantly less fre- quently reported end-of-life practices with intended hastening of death (p ¼ 0.003). Conclusion: Physician members of the German Society for Palliative Medicine perform a broad spectrum of end-of-life practices including intended hastening of death. The findings on patients’ non-involvement in decision making warrant further empirical–ethical analysis.

Keywords End-of-life decision, Germany, medical ethics, palliative care, survey

Introduction medical treatment as well as alleviation of symptoms In Germany as in many other countries there is with possible shortening of life is lawful provided that an ongoing debate about the appropriate ethical and such practice reflects the patient’s will. In accordance practical framework for physicians’ end-of-life prac- with legislation in most European countries it is illegal tices. With respect to legal aspects of decisions at for physicians practicing in Germany to administer sub- the end of life, the Federal Court of Justice stances which cause death even if this is the expressed (Bundesgerichtshof) has emphasized that limitation of will of the patient.1 While there is no penal statute on

Corresponding author: Dr. med. Jan Schildmann, MA, Institute for Medical Ethics and History of Medicine, Ruhr-Universitaet Bochum, Malakowturm-Markstraße 258a, D-44799 Bochum, Germany Email: [email protected]

Downloaded from pmj.sagepub.com at Kings College London - ISS on January 22, 2011 Schildmann et al. 821 assisted suicide in Germany the highest court points out countries so far.9,12 In this paper we present the first that an attempt of suicide in general is viewed as an results of this study. accident which creates a duty for physicians to try to The aims of this study are: save the life of a patient.2 The German debate on normative and practical (1) to describe different types and respective frequen- aspects of end-of-life care has been influenced by the cies of end-of-life practices by physician members of historical experiences during the Nazi regime. the German Society for Palliative Medicine; Discussions on ethical and clinical aspects of end- (2) to describe characteristics of the decision-making of-life practices have been a taboo for a long time in process (i.e. patients’ involvement); postwar Germany.3,4 The first hospices and palliative (3) to identify factors, such as physicians’ and patients’ care wards in this country have been established in the characteristics, which have been described as possi- 1980 s and in 1994 the German Society for Palliative bly associated with end-of-life practices resulting in Medicine was founded. In 2003 an added certificate of hastened death qualification in palliative medicine had been introduced and since 2009 palliative medicine has been an obliga- The results will be discussed in light of available data tory part of the undergraduate medical curriculum. The of other countries as well as the current international political and public support for palliative care is rele- debate on an appropriate ethico-legal framework for vant to the current discourse on practical as well as end-of-life decision making. ethical aspects of end-of-life decision making. With respect to practice, the development of palliative care Methods structures improves clinical competency at the end of life. On a normative level the possibility to avoid suffer- The authors conducted a postal cross-sectional sur- ing at the end of life by means of palliative care mea- vey on end-of-life practices among all physician mem- sures is cited as argument against physician-assisted bers of the German Society for Palliative Medicine dying (i.e. ending a patient’s life on request or assisted (N ¼ 1645). In line with the research proposal accepted suicide). In accordance with this line of argument by the research ethics committee of the Medical Faculty, the German Society for Palliative Medicine takes a Ruhr-University Bochum (Registration Number 3373- clear stance against hastening of death in its statutes.5 09), the contact details of the sample of all physician Similarly the German Medical Association in its guid- member of the German Society for Palliative Medicine ance points out that ending a patient’s life on request were provided by the secretary of the society as address is illegal and rejects physician-assisted suicide as labels. For the protection of anonymity there was no unethical.6 identifying code on the questionnaire. The whole In contrast to many other countries,7–12 there is scar- sample received the questionnaire for the first time in city of robust empirical data on end-of-life practices of February 2009 and a second time in March 2009 as part physicians in Germany.13–15 However, such research of a reminder for the study. In the cover letter of this contributes to the scientific and public debate for a follow up, potential participants were informed that number of reasons. First of all such studies can they should only respond once. inform about the spectrum of different types of end- As a survey instrument we used the German version of-life practices and the respective frequencies. Second, of the EURELD questionnaire, which had already been empirical research may direct the focus of attention used in the German-speaking part of Switzerland.9 to end-of-life decision making in clinical practice Following a pilot study, few smaller formal adaptations which deserves further ethico-legal analysis. Finally, regarding the 47-item questionnaire had been made. All interdisciplinary empirical research on clinical and eth- changes had been discussed with members of the Swiss ical aspects at the end of life contributes to an informed group of the EURELD study (S. Fischer and G. (continuing) medical on end-of-life issues. Bosshard) to avoid changes regarding the content. In light of the relevance of palliative care in the cur- Following the procedure described by Seale12 potential rent debate on end-of-life decision making and scarce participants had been informed on the first page of the data which are currently available on end-of-life prac- questionnaire that all questions of the survey instru- tices performed by specialists in palliative care, we have ment refer to the patient who had most recently died selected physician members of the German Society of under their care. Participants of the study who indi- Palliative Medicine as the target population for a cated on the cover page of the questionnaire that they survey on end-of-life practices. As a survey instrument had not cared for a dying patients within the last we used the questionnaire which has been developed by 12 months or that they had worked abroad during the EURELD (European end-of-life decision) consor- the last year were asked to return the questionnaire tium and which has been used in seven other European without any further information.

Downloaded from pmj.sagepub.com at Kings College London - ISS on January 22, 2011 822 Palliative Medicine 24(8)

Box 1. Key questions on physicians’ end-of-life practices

4 Did you or a colleague perform one or more of the following actions - or ensure that this action/these actions would be performed - taking into account the probability or certainty that this action would hasten the end of the patient’s life?:

(Please answer all three questions 4a, 4b, and 4c) 4a Withholding of treatment*? yes no 4b Withdrawal of treatment*? yes no 4c Intensification of pharmacological pain relief yes if you answered “yes”, go to and/or symptom relief? question 5 no if you answered “no”, go to question 6 *In this study ‘treatment’ includes artificial feeding and/or hydration.

5 Was hastening the end of life partially the yes intention of the action indicated in 4c no

6 Was death of the patient caused by one of the following actions or omissions which you or a colleague decided to take with the explicit intention of not prolonging life or hastening the end of life**? (please answer both questions 6a and 6b) 6a Withholding of treatment***? yes no 6b Withdrawal of treatment***? yes no ** Either ‘hastening death’ or ‘not prolonging life’. *** In this study ‘treatment’ includes artificial feeding and/or hydration.

7a Was death caused by the use of a drug yes if you answered “yes”, go to prescribed, supplied or administered by you question 7b or a colleague with the explicit intention of no if you answered “no”, go to hastening the end of life (or of enabling the question 8 patient to end his or her own life?) 7b Was a dying organization (for example yes Dignitas or Exit) involved at this death? no 7c Who administered this drug (that is to say, the patient imported it into the body)? You or another physician (please tick as many answers as apply) nurses someone else

End-of-life practices in this questionnaire are defined are provided as total numbers or percentage of all as physicians’ actions which are associated with the responding physicians working in Germany who had possible or intended shortening of patient’s life. Box 1 cared for a patient who died within the 12 months prior summarizes the translated version of key questions of to the survey (N ¼ 780). Descriptive and statistical anal- the survey instrument related to end-of-life practices. ysis was performed with SPSS version 18.0 for Windows Respondents could indicate more than one end-of-life (SPSS Inc, Chicago, IL, USA). Statistical analysis to practice if this was the case with regards to the respec- test for possible associations between certain end-of-life tive patient. practices and characteristics on the side of the patient or Data were collected in a SPSS data file by the second the physician had been conducted based on findings author. A random sample of 15% of the data was of earlier studies.9,16,17 Chi-squared tests were used for checked by a second person (SD) for correctness. The univariate comparisons. In addition, we performed logis- results of the descriptive analysis of end-of-life practices tic regression analysis for multivariate testing. In this

Downloaded from pmj.sagepub.com at Kings College London - ISS on January 22, 2011 Schildmann et al. 823 analysis all end-of-life practices with intended shorten- Table 1. Demographic details of patients ing of life (e.g. limitation of treatment with intended N % hastening of death or ending patients’ life on request) were summarized and defined as dependent variables. Studied Based on the findings of univariate analysis conducted cases 780 100 prior to the regression analysis, physicians’ qualifica- tion, cancer as a cause of death and preceding end- Gender Male 407 52.2 of-life discussions between physicians and patients Female 363 46.5 were explored as potential factors relevant to the inci- Missing 10 1.3 dence of end-of-life practices with intended hastening Age in years <1 3 0.4 of death. p-values <0.05 were considered significant. 1–17 16 2.1 18–64 298 38.2 Results 65–79 291 37.3 80 and older 157 20.1 Nine hundred and one questionnaires had been returned Missing 15 1.9 following the initial mailing of questionnaires and one Cause of Cardiovascular disease 52 6.7 follow up. In addition 31 questionnaires were returned death* Cancer 584 74.9 to sender because the physician could not be located Respiratory disease 55 7.1 under the given address. The response rate was 55.8%. One hundred and fourteen respondents (12.7%) indi- Disease of the 56 7.2 nervous system cated that they either had not treated a patient who had died within the last 12 months (N ¼ 84) or that Other/unknown 53 6.8 they had not worked in Germany during this time span Place of death Hospital 433 55.5 (N ¼ 30). These questionnaires were excluded as well as Home 218 27.9 seven other questionnaires in which further analysis was Care home 59 7.6 not possible due to a significant lack of data. In total Hospice 57 7.3 780 questionnaires were eligible for further analysis. Other 9 1.2 There were 447 male (57.3%) and 332 female Missing 4 0.5 (42.6%) respondents eligible for further analysis. In one case the gender was not indicated. Internal med- *Multiple answers possible. icine was named by 28.5% of these physicians as their speciality, followed by anaesthesiology (19.1%) and participants report that in 213 patients (27.3%) end- general practice (14.4%). A certificate of added qualifi- of-life practices had been performed with the intention cation in palliative medicine had been acquired by to shorten life. This includes limitation of treatment 77.9%. With respect to their religious affiliation, with the intention to shorten of life as well as adminis- 39.5% indicated that they were protestant, whereas tration of substances with partial or explicit intent to 36.9% were catholic, 16.9% indicated that they are shorten life reported by physicians. Medical treatment not affiliated to a religion and 4.0% indicated that was withheld with the explicit intention to hasten death they belong to other religious affiliations. by 21.3% of the respondents, and 13.7% of the physi- The majority of patients who had been cared for by cians stated that medical treatment was withdrawn with the responding physicians had been reported as dying such intention to shorten the patients’ life. of cancer (74.9%). Table 1 outlines further demo- In 353 cases physicians described their respective graphic details of the patients. patients as capable of judging their situation adequately In 740 cases (94.9%) respondents indicated that the and to make a decision. In 47 of these cases patients patients’ death was neither sudden nor unexpected and were not involved in decisions about end-of-life prac- in 692 patients (88.7%) at least one of the predefined tices. Table 3 summarizes the reasons given for the end-of-life practices (see Box 1) had been performed. In non-involvement of these patients by the respective 7 cases respondents report ending a patient’s life on physicians. request, 2 physicians report ending of patient’s life Patients dying at home received alleviation of symp- without explicit request and one physician has been toms significantly more often with possible shortening involved in a case of assisted suicide. Table 2 summa- of life (85.8%) compared with patients dying in hospital rizes the different types of end-of-life practices and the or other places (79.2%) (p ¼ 0.040). Alleviation of respective frequencies. symptoms with possible shortening of life had been When asked about their intention regarding hasten- conducted significantly more often by physicians who ing death associated with certain end-of-life practices had acquired the certificate of added qualification in

Downloaded from pmj.sagepub.com at Kings College London - ISS on January 22, 2011 824 Palliative Medicine 24(8)

Table 2. Frequencies of end-of-life practices to hasten death compared with those physicians who had not acquired the certificate (9.1% versus 15.7%, N % p ¼ 0.048). Following bivariate statistical analysis with Studied respect to possible factors relevant to the incidence of cases 780 100 end-of-life practices in which shortening of life was intended by the physician we conducted one logistic Sudden and 26 3.3 regression. Based on the findings which were significant unexpected death determinants for end-of-life practices with intended Non-sudden death 740 94.9 shortening of life we analysed for independent associa- Missing 14 1.8 tions between the incidence of reported practices with Total cases* with (at least) 692 88.7 intended hastening of death such as limitation of treat- one end-of-life practice ment with hastening of death or ending patients’ life on Doctor assisted dying: 10 1.3 request and possible predictive variables on the side of Ending of life 7 0.9 the physician (added certificate in palliative medicine, on request religious affiliation) and on the side of the patient Ending of life 2 0.3 (cause of death). In addition, a possible association without between the incidence of reported practices with request intended hastening of death and prior discussions Physician 1 0.1 between physician and patient about possible life short- assisted ening was tested as part of the multivariate analysis. suicide Logistic regression indicates that being a physician Symptom alleviation 609 78.1 with an added qualification in palliative medicine with possible reduced the odds (odds ratio [OR] 0.519) of end- shortening of life of-life practices with intention of hastening death Non-treatment decision 539 69.1 (95% CI 0.337–0.799, p ¼ 0.003). There was also a sig- with possible nificant reduction (OR 0.552) of these end-of-life prac- shortening of life tices associated with cancer as cause of death (95% CI *Multiple answers possible. 0.361–0.843, p ¼ 0.006) whereas there was a positive association (OR 1.302) between end-of-life practices with shortening of life and preceding end-of-life discus- Table 3. Reasons given for non-involvement of competent patients in end-of-life decision making sions between physicians and patients (95% CI 1.069– 1.585, p ¼ 0.009). N % Non-involvement of Discussion competent patients: 47 100 In this paper we have presented the first data on the Reasons*: Patient was too young 1 2.1 types and frequencies of end-of-life practices reported Obviously in best interest 11 23.4 by physician members of the German Society for of patient Palliative Medicine. In addition, the study has provided More harm than good 11 23.4 information about ethically relevant characteristics of Patient was unconscious 5 10.6 the decision-making process as well as factors associ- Patient was suffering 1 2.1 ated with those end-of-life practices in which hastening from dementia of death is intended by the physician. Patient was mentally disabled 0 0 Physician members of the German Society for Patient had psychiatric illness 0 0 Palliative Medicine are a relevant and interesting Other reasons 15 31.9 group with respect to empirical research on end-of-life practices. However, while there is a reasonable good Missing 8 17.0 response for this target group, the sample is not repre- *Multiple answers possible. sentative for all physicians working in Germany. Therefore, our results cannot be generalized in this palliative medicine (84.3%) compared with physicians respect. Owing to the aforementioned limitation it is who did not have this qualification (69.2%) (p < 0.001). also not possible to compare our results with the results Physicians with the certificate in palliative medicine of the EURELD study9 or either of the studies con- indicated significantly less often that alleviation of ducted by Seale in the United Kingdom,12,18 which symptoms had been performed with the partial intent are based on a representative sample of physicians in

Downloaded from pmj.sagepub.com at Kings College London - ISS on January 22, 2011 Schildmann et al. 825 the respective countries (or parts of these countries). against the statutes of the German Society for Palliative Next to the well-known limitations of postal surveys Medicine as well as statements of international profes- such as social desirable answers, recall and response sional bodies such as the European Association of bias, the framing of questions on end-of-life practices Palliative Care.5,20 In contrast to this normative state- in such a survey is associated with reductions and ment on the side of the representatives of palliative potential bias of results.19 Next to the advantage of medicine and in line with our findings the results of extensive validation of the EURELD questionnaire in an earlier survey on physicians’ attitudes conducted many studies one advantage of the survey instrument is by one of the authors (CM-B), this indicates that its terminology which avoids terms easily to be misun- there is support for physician-assisted dying among a derstood such as ‘(passive) euthanasia’. However, a smaller part of members of the German Society for problem with the EURELD survey instrument in this Palliative Medicine.21 The obvious differences between respect is that end-of-life practices are defined as certain empirical data on end-of-life practice and views among practices associated with potential shortening of life. palliative care physicians on the one hand and the nor- This compound statement which combines an action mative statements forwarded by palliative care associ- with an expectation may have led to a misreporting ations on the other hand may serve as starting point for of end-of-life practices.18 Even in light of the aforemen- further discourse regarding an appropriate normative tioned limitations we believe that this study informs the and practical framework on end-of-life practice. current debate on ethically as well as clinically relevant In addition to information about the types of end- aspects of end-of-life decision making in several ways. of-life practices, this study shed light on the decision- First of all our data indicate that although the mur- making process between patients and physicians. dering of patients during the Nazi regime has influenced In this respect it is an interesting finding that 22 patients the discourse on end-of-life decision making in the who were deemed to be capable by their physicians to sense of putting a taboo on this topic,3 German pallia- judge their situation were not involved in decision tive care physicians, comparably to physicians in other making on end-of-life practices. The patient’s ‘best countries,9 perform all end-of-life practices. The results interest’ (N ¼ 11) or because ‘more harm than benefit’ of this survey support findings in earlier studies con- would have been done by involving the patient (N ¼ 11) ducted in Germany using different methods,13–15 as well were named as reasons for such practice (Table 3). as the conclusions of international research that con- Although the survey instrument in the respective ques- troversial practices such as physician-assisted suicide or tion does not use a wording which clearly reflects the ending patients’ life on request do not only take place in concept of legal capacity, it seems likely that, in these countries which have legalized these practices.9 cases, competent patients had not been involved in deci- Second, the survey provides information about end- sion making. To use the best interest of the patient as of-life practices of physicians with an interest in the an argument for end-of-life practice without involve- field of palliative medicine. On this background it is a ment of patients judged as competent by their remarkable finding that a relevant proportion of physicians resembles the paternalistic model of respondents, and even more of those who have not physician–patient relationship. This is also valid for acquired the added certificate of qualification in pallia- the use of the ‘therapeutic privilege’ as an argument tive medicine, not only foresee but also intend shorten- not to involve patients in end-of-life decisions because ing of life as part of their practice. This is valid not only of doing harm. From an ethical perspective such prac- for the small number of physician-assisted dying (assis- tice runs clearly contrary to the respect for autonomy of ted suicide, ending patients’ life) events, but also for patients at the end of life. Further qualitative research example for those cases in which treatment was either on the perspective of physicians involved in such deci- withheld (21.3%) or withdrawn (13.7%) with the expli- sion making seems warranted to explore in more detail cit intention to hasten death. Similarly to the data the context of and possible reasons for such clinical recently published by a study group in Belgium, this practice.22 finding may be interpreted as evidence on a clinical In light of the scarcity of empirical data on end- practice of co-existence of accepted palliative care prac- of-life practice in Germany and based on international tice (i.e. limitation of treatment or symptom treatment end-of-life research, we conducted a statistical analysis with possible life-shortening effect) and physicians’ to explore the relevance of factors which have been action with intended shortening of life (e.g. assisted sui- identified as relevant to end-of-life decision cide, ending patients’ life on request or limitation of making.9,16,17 Owing to multiple testing, the p-values treatment with intended hastening of death).7 While given in this paper must be interpreted with caution our data should be interpreted cautiously in the light and need confirmation based on predefined hypotheses of the aforementioned challenges regarding the wording in future studies. An interesting finding is the significant of the survey questions it is clear that such practices are association of respondents who have acquired an added

Downloaded from pmj.sagepub.com at Kings College London - ISS on January 22, 2011 826 Palliative Medicine 24(8) certificate in palliative medicine and who more fre- by the research ethics committee of the Medical Faculty, quently conducted alleviation of symptoms with possi- Ruhr-University Bochum (Registration Number 3373-09). ble life shortening than those physicians without this qualification. At the same time physicians with this Funding qualification significantly less frequently report the This research received no specific grant from any funding intention to shorten life when alleviating symptoms agency in the public, commercial, or not-for-profit sectors. than the group of physicians without formal qualifica- tion in palliative medicine. In light of the clear stance of Conflict of interest statement the German Society for Palliative Medicine against has- Professor Mueller-Busch (current president) and Dr. Jan tening death, the results need to be interpreted cau- Schildmann are members of the German Society for tiously on the background of possible socially Palliative Medicine. desirable answers. However, it is also possible that training regarding the medical aspects of alleviation References of symptoms as well as ethical teaching on the doctrine of double effect as part of the palliative care curriculum 1. Bosshard G, Broeckaert B, Clark D, et al. A role for have an influence on the reported intentions regarding doctors in assisted dying? An analysis of legal regulations and medical professional positions in six European coun- end-of-life practices. tries. J Med Ethics 2008; 34: 28–32. In summary this study indicates that members of the 2. Weber M and Kutzer K. Ethische Entscheidungen am Ende Germany Society for Palliative Medicine perform a des Lebens - Grundsa¨tze, Unsicherheiten, Perspektiven. broad spectrum of end-of-life practices including Dtsch Med Wochenschr 2002; 127: 2689–2693. those with hastening of death. The data on non- 3. Scho¨ne-Seifert B and Rippe KP. Silencing the Singer. involvement of patients deemed competent and the for- Antibioethics in Germany. Hastings Cent Rep 1991; 6: warded reasons (i.e. ‘best interest’ and ‘avoidance of 20–27. harm’) give rise to further empirical–ethical analysis 4. Vollmann J. Sterbebegleitung. Gesundheitsberichterstat- on the underlying reasons for such practice. Finally tung des Bundes Heft 02/02, 2nd ed. Berlin: Verlag the indicated associations between patients’ (i.e. dis- Robert Koch-Institut, 2003. ease) and physicians’ characteristics (i.e. qualification 5. GSPM [Deutsche Gesellschaft fu¨r Palliativmedizin]. Satzung der Deutschen Gesellschaft fu¨r Palliativmedizin in palliative medicine) and end-of-life practices with e.V. (Statutes of the GSPM, current version of hastened death may inform continuing medical educa- 27 September 2008), http://www.dgpalliativmedizin.de/ tion with respect to clinical and ethical aspects of end- images/stories/pdf/DGP-Satzung%202009%20_091003_ of-life practices and should be explored in more detail .pdf (accessed 7 August 2010). in future empirical–ethical research. 6. German Medical Association [Bundesa¨rztekammer]. Grundsa¨tze der Bundesa¨rztekammer zur a¨rztlichen Sterbebegleitung (German Medical Association’s guide- Acknowledgements lines on aid in dying). Dtsch Arztebl 2004; 95: We would like to thank all participants for their time and A2366–A2367. consideration in taking part in this study. We thank 7. Van den Block L, Deschepper R, Bilsen J, et al. Dr. Tim Holland Letz, Department for Medical Euthanasia and other end of life decisions and care pro- Informatics, Biometry and Epidemiology, Ruhr-University vided in final three months of life: nationwide retrospec- Bochum for statistical support and Dr. Georg Bosshard, tive study in Belgium. BMJ 2009 (online publication doi: Dr. Susanne Fischer, Dr. Agnes van der Heide and 10.1136/bmj.b2772). Professor Clive Seale for their support in preparing the 8. Emanuel EJ. Euthanasia and physician-assisted suicide: study. We also would like to thank Stephanie Dahm (SD) a review of the empirical data from the United States. for data checking and Dr. Thomas Schindler for his support Arch Intern Med 2002; 162: 142–152. on the side of the secretary of the GSPM. Dr. Schildmann’s 9. van der Heide A, Deliens L, Faisst K, et al. EURELD and Ms Hoetzel’s work has been funded by a start-up grant of consortium. End-of-life decision-making in six European the (‘Programm zur countries: descriptive study. Lancet 2003; 362: 345–350. Anschubfinanzierung des wissenschaftlichen Nachwuchs’). The 10. Kuhse H, Singer P, Baume P, et al. End-of-life decisions sponsor had no role in the study design, in the collection, in Australian medical practice. Med J Aust 1997; 166: analysis or interpretation of data, in the writing of the 191–196. report, or in the decision to submit the article for publication. 11. Mitchell K and Owens G. National survey of medical All authors (JS, JH, CM–B, JV) have contributed a significant decisions at end of life made by New Zealand general amount to the conception, design, analysis and interpretation practitioners. BMJ 2003; 327: 202–203. of data. They were all involved in the production of the drafts 12. Seale C. National survey of end-of-life decisions made by of the article and have given their approval of the submitted UK medical practitioners. Palliat Med 2006; 20: 3–10. version to be published. There is no one else who fulfils the 13. Dornberg M. Angefragt: Sterbehilfe. Behandlungsbe- criteria of authorship. The study protocol has been accepted grenzung und Sterbehilfe aus der Sicht internistischer

Downloaded from pmj.sagepub.com at Kings College London - ISS on January 22, 2011 Schildmann et al. 827

Krankenhausa¨rzte. Ergebnisse einer Befragung und 19. Hagelin J, Nilstun T, Hau J and Carlsson HE. Surveys on medizinethische Bewertung. Frankfurt a. M: Peter Lang, attitudes towards legislation of euthanasia: importance of 1997. question phrasing. J Med Ethics 2004; 30: 521–523. 14. Maitra RT, Harfst A, Bjerre LM, et al. Do German gen- 20. Materstvedt LJ, Clark D, Ellershaw J, et al. Euthanasia eral practitioners support euthanasia? Results of a and physician-assisted suicide: a view from an EAPC nation-wide questionnaire survey. Eur J Gen Pract ethics task force. Palliat Med 2003; 17: 97–101. 2005; 11: 94–100. 21. Mu¨ller-Busch HC, Klaschik E, Oduncu F, et al. 15. Wehkamp K-H. Sterben und To¨ten. Euthanasie aus der Euthanasie bei unertra¨glichem Leid? Ergebnisse und Sicht deutscher A¨rztinnen und A¨rzte. Ergebnisse einer Erkenntnisse der DGP-Mitgliederbefragung im Mai empirischen Studie. Dortmund: Humanitas Verlag, 1998. 2002. Z Pallmed 2003; 3: 75–84. 16. Fischer S, Bosshard G, Faisst K, et al. Swiss doctors’ 22. Schildmann J and Vollmann J. Empirische Forschung in attitudes towards end-of-life decisions and their determi- der Medizinethik: Methodenreflexion und forschung- nants: a comparison of three language regions. Swiss Med spraktische Herausforderungen am Beispiel eines mixed- Wkly 2006; 136: 370–376. method Projekts zur a¨rztlichen Handlungspraxis am 17. Lo¨fmark R, Nilstun T, Cartwright C, et al. Physicians’ Lebensende [Empirical research in medical ethics: a experiences with end-of-life decision-making: survey reflection on methods and practical challenges of a in 6 European countries and Australia. BMC Med mixed-method project on physicians’ end of life prac- 2008; 6: 4. tices]. Ethik Med 2009; 21: 259–269. 18. Seale C. End-of-life decisions in the UK involving med- ical Practitioners. Palliat Med 2009; 23: 198–204.

Downloaded from pmj.sagepub.com at Kings College London - ISS on January 22, 2011