eMJA     The Medical Journal of Australia

Home | Issues | eMJA shop | My account | Classifieds | Contact | More... | Topics | Search   

Abstract

The use of drugs to intentionally hasten the death of a terminally ill patient is prohibited in most countries, including Australia. The only country that has openly allowed medically assisted deaths is the Netherlands, where 3.4% of all deaths are reported as (intentional) medically assisted deaths.¹ Most of these are voluntary euthanasia or assisted suicide, but about a quarter are "life-terminating acts without explicit and persistent request".² The most recent survey indicates that 53% of Dutch doctors have practised euthanasia or assisted suicide and 23% report that they have performed "life-terminating acts without explicit and persistent request".¹

Medically assisted deaths also occur in countries where they are prohibited and the figures have been remarkably consistent — in the United States,^3-6 Denmark,⁷ England⁸ and Australia,⁹ between 2.2% and 12.3% of doctors report that they have assisted death in response to an explicit request. Outside of the Netherlands, however, few studies have broadened the question of assisted death to include instances where there has been no explicit request. In a study comparing North American and Dutch physicians, 2% and 15%, respectively, reported "ending of life without an explicit request from the patient", but the numbers were small and the difference not statistically significant.¹⁰ In Australia, it has been claimed that 3.5% of all deaths are cases of "ending life without explicit request".¹¹

A potentially confounding issue faced by all researchers of assisted deaths is that of intention. Doctors sometimes give large doses of potentially lethal drugs to terminally ill patients to treat symptoms, foreseeing but not necessarily intending a medically hastened death. This kind of action has been shown consistently to have the approval of more than 80% of doctors.^1,5,7,12 However, there may be considerable ambiguity about a doctor's intention,¹³ and some studies have indeed noted partial or dual intentions (to relieve pain and to hasten death) when analgesic drugs are given.¹ An intention to hasten death has been suggested as being best distinguished by the use of drugs in doses greater than those required for symptom control.¹⁴

Our study incorporates such a distinction. Our objective was to conduct a survey of attitudes to and practices regarding assisted death using questions that were absolutely explicit about the agent's intention.

Methods

Sample

A list was obtained of all doctors with Australian mailing addresses registered as general surgeons with the Royal Australasian College of Surgeons (n = 1218). No attempt was made to exclude those who had recently retired or who had subspecialised. After excluding 200 surgeons who had been randomly selected for pretesting and those who had moved, were ill or deceased (26), a final eligible sample of 992 remained.

Survey instrument

The survey instrument was an anonymous, self-administered, mail-out questionnaire (available on the MJA website at <http://www.mja.com.au>. The questionnaire was developed from a review of the literature, discussion within a multidisciplinary research group and extensive pretesting, including 13 interviews and consistency checks on the responses to 200 mailed questionnaires. Advice was sought on specific questions from three independent ethicists with substantially different ethical backgrounds in ethics. All questions were closed (mostly "Yes/No"), but respondents were invited to make additional comments on the final page of the survey. The survey instrument included a clinical vignette (see Appendix), and some of the questions alluded to this vignette.

Our main question on experience with assisted death (Question 1, Box 2) was presented alone under a separate heading and was prefaced by the comment "All further questions address general issues and are not specific to the scenario [clinical vignette] . . .". Key words in Question 1 ("greater" and "intention") were printed in bold and underlined. Further testing of the understanding of this question was undertaken by interview with 10 general physicians after they had completed the entire questionnaire.

Administration of questionnaire

The questionnaire and three subsequent reminder letters were sent according to a set protocol¹⁵ commencing in August 1999. Intention to participate was indicated by return of a labelled consent or refusal card separate from the unmarked questionnaire, and reminders were sent to those who had not returned a consent or refusal card.

Statistical analysis

Affirmative responses are reported as a proportion of all respondents (not just those answering the question), except where explicitly stated. The rate of missing data was less than 4.4% for all questions and less than 2.3% for questions reported here.

The Wilson procedure with correction for continuity was used to calculate 95% confidence intervals (CI) for single proportions.¹⁶ To determine the influence of the five demographic variables (Box 1) on attitudes and practice, logistic regression analysis was performed using SAS for Windows.¹⁷ Variables which were significant at = 0.2 (Pearson's ² or Fisher's exact test) were entered into the logistic regression model and then eliminated in a backward stepwise procedure until only those variables remained that were statistically significantly associated with an affirmative response.

Ethical approval

Ethical approval for our study was obtained from the Hunter Area Research Ethics Committee, from the Human Research Ethics Committee of the University of Newcastle, and from the Ethics Committee of the Royal Australasian College of Surgeons.

Results

Of the eligible sample of 992 surgeons, 683 returned questionnaires (response rate, 68.9%). This sample size was associated with a precision of 4% (95% CI). Six hundred and fifty-four surgeons (65.9%) returned a separate consent card or other communication indicating intention to participate; 166 (16.7%) indicated that they did not wish to participate and 172 (17.3%) did not respond. Of those who declined to participate, 25 volunteered reasons. Of the respondents, 210 (30.7%) volunteered additional comments.

Demographic features of respondents are summarised in Box 1. Only age, sex and years in practice were available for non-respondents. There were no sex differences between respondents and non-respondents, but older surgeons and those who had been in practice for longer were slightly less likely to respond. Results for selected questions are given in Box 2, with wording and textual emphasis unchanged from that in the questionnaire.

Use of drugs with the intention of ending life or hastening death

Twenty-nine respondents (4.2%) reported having given a bolus lethal injection "in response to a sincere and unambiguous request", 13 (1.9%) reported assisting with suicide (Questions 3, 4, 5 and 6 in Box 2), and 36 respondents (5.3%; 95% CI, 3.8%-7.3%) had done one or both of these. Two hundred and forty-seven respondents (36.2%) reported that they had, for the purpose of relieving a patient's suffering, given drugs in doses greater than those required to relieve symptoms with the intention of hastening death (Question 1, Box 2). Of these, 139 indicated (in response to questions 3, 5 and 6, Box 2) that they had never received a sincere and unambiguous request for a lethal injection, and had never granted a request for assisted suicide. Thus, at least 20.4% of the entire sample (139/683; 95% CI, 17.4%-23.6%) have apparently given drugs with the intention of hastening death, but without the explicit request of the patient. Of the remaining 108 respondents who reported having given drugs with the intention of hastening death, it is unknown whether they have ever done so in the absence of a request.

Effect of religion

Religious affiliation was a significant predictor of response to questions on attitudes to and practice of intentionally assisted death. Roman Catholics were about 4-10 times more likely, and Protestants about 2-3 times more likely, to give a negative answer than colleagues who had no religious affiliation (Box 3).

Discussion

Our finding that very few doctors report having given a bolus lethal injection in response to a patient's request agrees with the findings of previous reports.^3-9 Our study also reveals that many doctors report giving drugs in doses greater than those required to relieve symptoms, with the intention of hastening death, often in the absence of an explicit request. Outside the Netherlands, this has not been widely reported.

Our main question on experience with assisted death was deliberately written to include the use of infusions of drugs, with or without a request. That some doctors are prepared to hasten death by infusion (but not by bolus) was confirmed by volunteered comments:

"It is difficult to actually administer a lethal injection, but setting up a potentially lethal system allows a degree of psychological and physical separation from the actual event."

"The giving of a single lethal injection would be unusual. Increasing infusion is a far preferable and controllable method."

"I also appreciate the inconsistency between being prepared to 'up the dose', but not being prepared to give it as a bolus — but that's the way I feel . . ."

"I have frequently used large doses of morphine (previously heroin!) to hasten death . . . I can't see the ethical difference between this and a bolus injection in a fully informed patient . . . but simply would not be capable of the deed myself."

"Talk of bolus injections in fully competent patients is not the real-life situation. We help very ill patients to die by a combination of sustenance withdrawal, increasing analgesia and 'masterly inactivity'."

Clearly, surveys that have limited their inquiry to the administration of a bolus lethal injection are likely to have underestimated doctors' involvement with assisted death. It may be that researchers have avoided addressing the use of infusions because of uncertainty about a doctor's intentions in such circumstances. However, it is possible to be unambiguous. Our question specified an intention to hasten death, and a dose of drug greater than that required to treat symptoms. Physician interviews confirmed that the question was indeed understood by most respondents, but there is also quantitative evidence of this from the survey itself.

Firstly, there was internal consistency: 95% of those who answered affirmatively to Question 1 also answered affirmatively to a question on the morality of giving drugs by slow intravenous infusion with the intention of hastening death (Question 2, Box 3), although the questions were separated in the questionnaire.

Secondly, there was a profound effect of religious affiliation on responses to both Questions 1 and 2, with odds ratios that were similar to those measured for questions relating to euthanasia by bolus lethal injection or assisted suicide (Box 3). The only plausible explanation for this strong association is that the respondents understood Questions 1 and 2 to be about the intentional hastening of a patient's death. In contrast, responses to a question about the use of an infusion of drugs that might incidentally hasten death (Question 7, Box 3) showed no effect of religion, with more than 90% of respondents supporting such action regardless of religious affiliation.

Euthanasia and palliative care — same drugs, same doses?

Our question specified a dose of drugs greater than that required to relieve symptoms, but it may be difficult to assess symptoms once consciousness has begun to deteriorate in a dying patient. Possibly the only way to be sure that a patient is not suffering at this point is to render him or her deeply unconscious by giving generous doses of opiates and/or sedatives. It would then be probable, but not certain, that the doses used were greater than those required to relieve symptoms. One respondent volunteered a comment to this effect:

"Intravenous infusion may be used to induce an unconscious state at a rate equal or greater than that to relieve symptoms, whereby the practitioner and family are then guaranteed that all the patient's symptoms are relieved . . ."

Whether the use of generous doses of analgesic or sedative drugs constitutes "good palliative care" or "non-voluntary euthanasia" depends, according to a widely held view, on the doctor's self-professed intention.¹⁴ Question 1 clearly specified an intention to hasten death. Doctors who responded affirmatively to this question have therefore crossed a legal threshold and, according to some, a moral threshold. However, it is not clear that they have acted differently from their colleagues other than by reporting their own mental state differently. Furthermore, it may be hard to distinguish many of their actions from those of Dutch doctors who have performed "life-terminating acts without explicit request". At least 20% of our entire sample appears to have given drugs with the intention of hastening death in the absence of an explicit request, similar to the 23% of Dutch doctors who report performing "life-terminating acts without explicit request".¹

There is a discrepancy between the relatively large proportion (36.2%) of surgeons who report giving drugs with the intention of hastening death, and the small proportion (5.3%) who report giving a bolus lethal injection or assisted suicide in response to an explicit request. We believe that many of those who make up this difference have given generous doses of analgesics or sedatives by infusion to dying patients. The circumstances of these deaths, other than in the agent's reported intention, may not differ substantially from what is widely accepted as good palliative care.

Acknowledgements

We would like to thank Professor Miles Little for critical reviews of our questionnaire and methodology, and Professor Grant Gillett and Dr Bernadette Tobin for providing opinions on the wording of key questions.

This research project was conducted with the assistance of a Royal Australasian College of Surgeons research scholarship.

Competing Interests

References

1. van der Maas PJ, van der Wal G, Haverkate I, et al. Euthanasia, physician assisted suicide, and other medical practices involving the end of life in the Netherlands, 1990-1995. N Engl J Med 1996; 335: 1699-1705.
2. Pijnenborg L, van der Maas PJ, van Delden JJM, Looman CWN. Life-terminating acts without explicit request of patient. Lancet 1993; 341: 1196-1199.
3. Meier DE, Emmons C, Wallenstein S, et al. A national survey of physician-assisted suicide and euthanasia in the United States. N Engl J Med 1998; 338: 1193-1201.
4. Back AL, Wallace JI, Starks HE, Pearlman RA. Physician-assisted suicide and euthanasia in Washington State. JAMA 1996; 275: 919-925.
5. Fried TR, Stein MD, O'Sullivan PS, et al. Limits of patient autonomy. Arch Intern Med 1993; 153: 722-728.
6. Lee MA, Nelson HD, Tilden VP, et al. Legalizing assisted suicide - views of physicians in Oregon. N Engl J Med 1996; 334: 310-315.
7. Folker AP, Holtug N, Jensen AB, et al. Experiences and attitudes towards end-of-life decisions amongst Danish physicians. Bioethics 1996; 10: 233-249.
8. Ward BJ, Tate PA. Attitudes among NHS doctors to requests for euthanasia. BMJ 1994; 308: 1332-1334.
9. Kuhse H, Singer P. Doctors' practices and attitudes regarding voluntary euthanasia. Med J Aust 1988; 148: 623-627.
10. Willems DL, Daniels ER, van der Wal G, et al. Attitudes and practices concerning the end of life: a comparison between physicians from the United States and from The Netherlands. Arch Intern Med 2000; 160: 63-68.
11. Kuhse H, Singer P, Baume P, et al. End-of-life decisions in Australian medical practice. Med J Aust 1997; 166: 191-196.
12. Emanuel EJ, Fairclough DL, Daniels ER, Clarridge BR. Euthanasia and physician assisted suicide: attitudes and experiences of oncology patients, oncologists, and the public. Lancet 1996; 347: 1805-1810.
13. Quill TE. The ambiguity of clinical intentions. N Engl J Med 1993; 329: 1039-1040.
14. Gillon R. Foreseeing is not necessarily the same as intending. BMJ 1999; 318: 1431-1432.
15. Dillman DA. Mail and telephone surveys: The total design method. New York: Wiley, 1978.
16. Newcombe, Robert G. Two-sided confidence intervals for the single proportion: comparison of seven methods. Stat Med 1998; 17: 857-872.
17. SAS system for Windows [computer program], version 6.12. Cary, NC: SAS Institute Inc, 1998.

(Received 28 May, accepted 3 Sep, 2001)

Authors' details

Hunter Centre for Health Advancement, Wallsend, NSW.
Katherine J Rainbird, BA(Hons), PhD, Research Associate.

Reprints will not be available from the authors.
Correspondence: Dr C D Douglas, c/- Professor A D Spigelman, Discipline of Surgical Science, Faculty of Medicine and Health Sciences, University of Newcastle, Locked Bag No 1, Hunter Region Mail Centre, Newcastle, NSW 2310.
cdouglasauATyahoo.com.au

©MJA 2001
Make a comment

Other articles have cited this article:

  David C Currow . MEDicine or MADness Med J Aust 2002; 176 (4): 190-191. [Letters] <http://www.mja.com.au/public/issues/176_04_180202/cur_hun_letter.html>

  Mathew Piercy, Gerald B Fogarty, Aubrey W Jansz and David M Gawler. The intention to hasten death of terminally ill patients Med J Aust 2002; 177 (3): 165-167. [Letters] <http://www.mja.com.au/public/issues/177_03_050802/piercy_050802.html>

  Roger W Hunt. "Death talk": debating euthanasia and physician-assisted suicide in Australia Med J Aust 2003; 179 (1): 57-60. [Matters Arising] <http://www.mja.com.au/public/issues/179_01_070703/matters_arising_070703.html>

  Malcolm H Parker, Colleen M Cartwright and Gail M Williams. Impact of specialty on attitudes of Australian medical practitioners to end-of-life decisions Med J Aust 2008; 188 (8): 450-456. [Research] <http://www.mja.com.au/public/issues/188_08_210408/par10958_fm.html>