Intended for healthcare professionals

Primary Care

National survey of medical decisions at end of life made by New Zealand general practitioners

BMJ 2003; 327 doi: (Published 24 July 2003) Cite this as: BMJ 2003;327:202
  1. Kay Mitchell, research fellow (k.mitchell{at},
  2. R Glynn Owens, professor of forensic clinical psychology2
  1. 1 Department of Psychology, University of Auckland, Private Bag 92019, Auckland, New Zealand
  2. 2University of Wales, Bangor, Gwynedd LL57 2AS
  1. Correspondence to: K Mitchell
  • Accepted 4 July 2003


Several attempts have been made in New Zealand to liberalise laws prohibiting euthanasia. Surveys in other countries where legalisation of euthanasia was being considered have found illegal provision of physician assisted death.13 However, none of these studies investigated the availability of palliative care services, which arguably make euthanasia unnecessary.4 We investigated the prevalence of physician assisted death in New Zealand within the context of availability of palliative care services.

Participants, methods, and results

We obtained an English version of the anonymous questionnaire previously used in Dutch and Australian studies1 3 and sent it to 2602 general practitioners. The general practitioners were in a commercially supplied address list, comprising 87% of the general practitioners in New Zealand in 2000. To retain consistency with previous studies, we sought details of medical decisions made for the last death attended in the preceding 12 months.

Of 1302 returned questionnaires, 1255 were usable. An interdisciplinary palliative care team was available in 1116 (89%) cases. Eleven hundred (88%) doctors reported attending a death in the past 12 months, and 693 (63%) had made a medical decision that could hasten death (table).

Medical actions that could hasten death taken by 693 New Zealand general practitioners on behalf of last dying patient cared for in past 12 months

View this table:

In 39 (5.6%) cases, death was attributed to actions consistent with physician assisted suicide or euthanasia. In 17 of these cases there was no discussion with the patient, and in 34 (87%) palliative care services were available (table). The doctor alone administered the drug in 13 cases, the nurse alone in 15 cases, the patient alone in two cases, and more than one person in eight cases; no information was available for one case.

Ninety four respondents (13.6%) reported final actions that were taken partly with the intent of hastening death. Fifty (53%) had not discussed this with the patient beforehand. A further 132 (19%) had withdrawn or withheld treatment explicitly to hasten death, 63 (48%) without discussion with the patient. Palliative care was available in over 85% of cases. Other actions were defensible under the principle of double effect—withholding or withdrawing treatment (55) or increasing medication to alleviate symptoms (373) knowing that death could be hastened (table).

Providers of physician assisted death were significantly older and less religious than those who had made other medical decisions. No other significant demographic differences were noted.


Thirty nine respondents had provided some form of physician assisted death, and 226 had taken actions partly or explicitly with the intention of hastening death. These actions would be indefensible under the principle of double effect.5 Our figures could be an underestimate of the number of physician assisted deaths as the phrasing of the question asked only about the last death attended. The fact that euthanasia is illegal in New Zealand could also have contributed to the low response rate (48%).2

Our results are consistent with research elsewhere.13 The proportion of physician assisted deaths in which there was no prior discussion with the patient was higher than reported in the Netherlands,1 but similar to that in Belgium and Australia.2 3

The intent to hasten death does not seem to be a consequence of the non-availability of palliative care. This suggests that palliative care did not meet patients' needs. We do not know whether this is because the best palliative care was not available or because even best care was insufficient to meet patients' needs. Either way, New Zealand requires greater commitment to the provision of quality palliative care or a re-examination of the legal framework that surrounds decision making at the end of life, or both.


  • Contributors KM was the principal investigator of the study, which formed part of her PhD. RGO supervised the study. KM is guarantor.

  • Funding None.

  • Competing interests None declared

  • Ethical approval University of Auckland Human Subjects Ethics Committee, 10.02.00, Ref. 1999/Q032