Intended for healthcare professionals


Physician assisted death in vulnerable populations

BMJ 2007; 335 doi: (Published 27 September 2007) Cite this as: BMJ 2007;335:625
  1. Timothy E Quill, professor of medicine, psychiatry and medical humanities
  1. University of Rochester School of Medicine, Box 601, Rochester, NY 14642, USA
  1. timothy_quill{at}

    Claims of increased risk in these groups are not supported by evidence

    Physician assisted death (both voluntary active euthanasia and physician assisted suicide) has been openly practised in the Netherlands for more than 25 years and formally legalised since 2002. The practice has been analysed in four major national studies between 1990 and 2007.1 2 A more restricted form of physician assisted death (physician assisted suicide only) was legalised in Oregon in 1997 and is the subject of an annual report ( Although these studies do little to resolve the moral and religious questions surrounding these practices, they do answer the following questions about the risks and benefits of legalisation.

    Will these practices become more common over time in a permissive environment? In Oregon, physician assisted death accounts for around one in 1000 deaths each year, with no significant change in frequency over nine years. All patients have met the necessary criteria, and more than 85% were also enrolled in hospice programmes. In Oregon, one in 50 dying patients talk to their doctors about assisted death and one in six talk to family members.3 There seems to be much conversation about end of life options, therefore, but relatively few cases of assisted death. Oregon is among the nation's leaders in other markers of good end of life care, including deaths at home, opioid prescribing, hospice enrolment, and public awareness about end of life options.4 The Dutch practices of physician assisted death have also remained stable over the duration of four studies,2 and hospice and palliative care have become more prevalent in recent years.

    Will the burdens and risks of these practices fall disproportionately on vulnerable populations? A study by Battin and colleagues published in this week's Journal of Medical Ethics that analyses existing databases from Oregon and the Netherlands dispels many of these concerns.5 They found no increased incidence of physician assisted death in elderly people, women, people with low socioeconomic status, minors, people in racial and ethnic minorities, and people with physical disabilities or mental illness. The one exception was people with AIDS, and studies from San Francisco completed before protease inhibitors were used also showed a high prevalence of physician assisted death in this population.6 These findings call into question the claim that the risks associated with legalisation will fall most heavily on potentially vulnerable populations.

    Are data available about these practices in places where physician assisted death is prohibited? Our study in 1998 assessed the secret practice of assisted death (both physician assisted suicide and voluntary active euthanasia) in the United States, and found significantly higher rates (about one in 50 deaths) than in Oregon after legalisation.7 The data are not directly comparable, as the study strategy we used safeguarded the surveyed doctors to ensure anonymity (similar techniques are used to study other illegal practices). This may have meant that the participating doctors were less representative and that they reported their practice differently than if the practice were legal. None the less, it raises the possibility that legalisation and regulation with safeguards may protect rather than facilitate the practice.

    Are there some cases in legal environments that do not meet the criteria and are not reported? The most controversial cases in the Netherlands are the life ending acts that have no explicit requests (about 1000 cases each year).1 2 8 Most, but not all, of these patients were suffering greatly and had lost the ability to make decisions for themselves, and many had previously given consent for physician assisted death under such circumstances. The number of such cases, known as LAWER cases, has decreased over time,2 but they still account for about 0.4% of deaths that fall outside the Dutch guidelines on voluntariness. It is tempting to attribute such cases to legalisation becoming a slippery slope, but a recent study of six Western European countries—using the same format and questions as the Dutch studies—showed that four of the six countries where assisted death is illegal had a much higher incidence of LAWER cases than is seen in the Netherlands. In fact, such cases were more common than cases of assisted death where voluntary consent was given (either voluntary active euthanasia or physician assisted suicide).9

    What happens in the US to patients without mental capacity who are dying and whose suffering cannot be relieved by usual palliative measures? Evidence based answers to this question are unknown, but there is likely to be extreme variability in the face of the legal and moral uncertainty about responsibilities, risks, and acceptable approaches.10 Clinical experience suggests that we deal with many of these patients using terminal sedation,11 a last resort that has been legal in the US since the 1997 US Supreme Court ruling. No formal tracking is available for this practice in Oregon or elsewhere in the US. Limited data suggest that the practice of terminal sedation is highly variable and accounts for 0-44% of deaths, depending on definitions and programmes.12 In the Netherlands, terminal sedation accounted for 5.6% of deaths in 2001, compared with 7.1% in 2005 (it was not measured in the first two studies).2 Many patients who receive terminal sedation are actively dying, experiencing severe physical suffering, and have lost capacity, so some were probably categorised as LAWER cases in previous Dutch studies. Terminal sedation is a legal practice in the US that could be improved if directed by carefully crafted guidelines and reporting.

    These days, patients who are dying are faced with a wide array of uncertainties and choices, and the physical and psychological challenges they experience are more complex. Available data suggest the risks and benefits of controversial practices like physician assisted death or terminal sedation are more favourable when practitioners work together with patients and families in an open and accountable environment. Secret practices and arbitrary restrictions should be avoided whenever possible.

    Studies such as those by Battin and colleagues from Oregon and the Netherlands help clarify the actual risks and benefits of legalisation of physician assisted death to vulnerable populations. We should ensure that pseudoscientific arguments are not used to promote particular moral values and associated restrictions. Patients who are dying and their families need us to be as objective and honest as possible in these deliberations.


    • Competing interests: None declared.

    • Provenance and peer review: Commissioned; not externally peer reviewed.


    View Abstract