Assisted Suicide and ‘Bracket Creep’
[This is the second article on assisted suicide in the series by Peter McCullagh. Click here to read the previous article, Good Suicide Vs Bad Suicide:]
Q. What do the terms ‘climate change’ and ‘slippery slope’ (in relation to assisted suicide) have in common?
Α. Both are regularly dismissed as fictitious in ‘one liners’. In both instances, the terms predict adverse consequences. In both instances, those predictions are based on preceding events, and their value will be dependent on the accuracy of description and analysis of those events. Should the evidentiary value of relevant preceding event be poor, then the credibility of the predictions, be they concerned with climate or assisted suicide will be proportionally diminished. ‘One line’ dismissals, particularly in relation to assisted suicide, have invariably denied the existence of the preceding events on which predictions of a ‘slippery slope’ are based. Nevertheless, detailed examination of those events has invariably been absent.
The purpose of this short paper is to collate, in summarised form, some of the events involving decriminalisation and normalisation of assisted suicide/euthanasia in non-Australian jurisdictions which, I believe, raise major concerns about similar consequences following decriminalisation in Victoria. Predictions of a ‘slippery slope’ in Victoria are based on the a priori tendency of human behaviour to be replicated in equivalent situations. In the interests of brevity, detailed referencing of the sources of those events has been omitted. These details are available from the author.
Three jurisdictions, Belgium, the Netherlands and Oregon, which are commonly cited as examples of successful, robust regulation of assisted suicide/euthanasia will be considered followed by some recent events from Canada, the new kid on the block. Having visited 3 of these 4, together with Switzerland, the chair of the Victorian Select Committee asserted:
As the exercise to be undertaken will be to recall historical changes in practice in these jurisdictions, the expression ‘bracket creep’ will be employed as, perhaps having less predictive, future implications than ‘slippery slope’. Two categories of bracket creep require consideration. Both quantitative (i.e. the frequency of assisted suicide/euthanasia) and qualitative (the nature of the situations in which decriminalised procedures could be practised) data are relevant. Obviously, this separation is not absolute as numerical increases in incidents may result from recruitment of additional situations into practice.
As a further issue for inquiry, it is necessary to take account both of practices undertaken within the limits of existing regulation and those outside regulatory boundaries. This last group raises the question of ascertainment – is information retrievable about practices which were outside regulatory boundaries? Finally, it is highly relevant to identify those practices which were initially contrary to regulation but, with legislative alteration of boundaries, had legitimacy conferred upon them.
There has been considerable variation in the availability of relevant information relating to the three jurisdictions. Variations in availability reflect both characteristics of the regulatory agencies, such as reporting requirements and the extent of release and retention of reported information, and also the extent to which research, independent of regulatory agencies, has been undertaken and published. In respect of the latter, the existence, in Brussels, of a university based End of Life Care Research Group (EoLCRG) has ensured that the extent of data available concerning Belgian practice considerably exceeds that accessible from the other jurisdictions.
Regulatory Processes in the Netherlands, Belgium and Oregon
Oversight of Dutch regulation is the responsibility of 5 regional committees. The Victorian Committee explained that:
So far so good, but reports from the Netherlands paint a rather different picture.
The 2009 report on behalf of the 5 regional committees emphasised the steady increase in annual frequency of assisted suicide/euthanasia: The number of notifications received under the Act has risen sharply again, from 2,331 in 2008 to 2,636 in 2009, an increase of just over 13%. Since 2006 the number of notifications has risen by a steady 10% or more a year. By 2015, 4.5% of deaths were attributable to assisted suicide/euthanasia.
The report continued: the five regional review committees and their secretariats have now reached the limits of their capacity. The secretaries are overburdened and, despite working at maximum efficiency, are now forced to focus on their core task – supporting the committees in reviewing notified cases of termination of life – with the result that other tasks are not performed.
A similar assessment was provided by a long term member of one of the regional Committees. In 2007, 2 years after joining, Theo Boer wrote that: there doesn’t need to be a slippery slope when it comes to euthanasia. A good euthanasia law, in combination with the euthanasia review procedure, provides the warrants for a stable and relatively low number of euthanasia. Most of my colleagues drew the same conclusion (could have been written by the Victorian committee).
Fast forward 7 years and Boer’s assessment had changed. Writing for a British audience with a euthanasia bill about to be introduced into the UK parliament, and having resigned from his regional committee, he cautioned:
We were wrong – terribly wrong, in fact. In hindsight, the stabilization in the numbers was just a temporary pause. Beginning in 2008, the numbers of these deaths show an increase of 15% annually, year after year. The annual report of the committees for 2012 recorded 4,188 cases in 2012 (compared with 1,882 in 2002). 2013 saw a continuation of this trend and I expect the 6,000 line to be crossed this year or the next.. What was once considered a last resort, now becomes a default mode of dying for an increasing number of people. In fact unbearable suffering is now seen less in terms of physical conditions but more in terms of ‘meaningless waiting’.
The operation of the Belgian Federal Evaluation and Control Commission provides an interesting variation. The regulatory process as applied in Belgium can be concisely characterised in two words, Wim Distelmans. Following decriminalization in 2003, the Belgian Council of Ministers appointed Distelmans to serve as the chairman of the Federal Commission, with responsibility for reviewing euthanasia deaths to ensure that doctors have complied with the law.
For those not familiar with the finer points of Belgian regulation, it may come as a surprise that Wim Distelmans is reputed to be the most prolific practitioner of euthanasia in the country (as a league table of practitioners is not publicly available, this remains speculation).
An article in Der Spiegel in November, 2014 on the occasion of a study trip to Auschwitz arranged and led by Wim, he was described as responsible for the deaths of hundreds, if not thousands of people. He’s a man who scrupulously studies his field of work. Apparently, he attends throughout and participates in all meetings of the Commission, but does not participate in discussion of cases in which he was the practitioner (one has to draw a line somewhere).
Apart from the End of Life Care Research Group mentioned above, Brussels is home to the European Institute of Bioethics An Institute dossier released after the first decade of operation of the Commission concluded that its purpose had been: to put an end to semi clandestine practice thereby ensuring legal safety, first of all, for patients,. But also: for the medical practitioner who would be able to avoid facing any legal proceedings in the event of carrying out euthanasia within the strict confines of the law. The dossier concluded that no more than that had been accomplished.
As for Oregon regulation, information concerning the operation of assisted suicide remains much less accessible than it is in the Benelux jurisdictions (Oregon does not have a publicist comparable with Wim Distelmans). Reporting of assisting at suicide is not compulsory and there are no applicable penalties for failure to report a death. Reliable figures detailing the frequency of non-compliance on the part of practitioners do not exist. Funding is not specifically allocated to a government department to undertake collection and longer term retention of data.
Information about any non-compliance with the legislation is limited. It would be prudent to recall the distinction between evidence of absence and absence of evidence. A final regulatory condition, unhelpful for data collection, is that practitioners are required to ‘doctor’ the death certificate by recording, as the cause of death, the medical condition which qualified the person for assistance to commit suicide.
A report by Drum and colleagues, from the Oregon Health and Science University, Portland, concluded as follows:
Review of the empirical literature on DWD in Oregon reveals a number of potential concerns, including inadequate demographic profiling of DWD requesting patients with regard to disability, inadequate mental health evaluations, insufficient duration of physician-patient relationships, potential inaccuracy of the 6-month prognosis, and inadequate exploration of alternative treatment. Of greater concern is the lack of oversight of potential abuses of DWD. The DWD Act creates criminal liabilities for such acts as forging a request for DWD or exerting undue influence or coercing someone to request DWD but does not designate an agency to investigate potential abuses.
In order to ensure that doctors have complied with the law in terminal cases, two doctors need to confirm that the patient’s suffering stems from an incurable illness. For non-terminal cases, three doctors must agree to a patient’s request to die.
Evolution of Practice in Belgium.
Prior to the 2002 legislation, there was a substantial frequency of illicit euthanasia in Belgium. In the course of debate on the decriminalisation bill, a parallel bill to augment funding and accessibility of palliative care was foreshadowed and supported by the Belgian Health Commission. This parallel bill did not proceed. The law, when enacted, indicated that depression was specifically excluded as conferring eligibility for euthanasia.
The Commission for Control and Assessment, in its first annual report, identified its problem as being how to exercise control after the death of a patient, the only available account of which was based exclusively on a declaration by the participating medical practitioner. The Commission acknowledged the dubious likelihood of a practitioner’s preparedness to denounce himself for non-compliance with specified regulatory requirements. Whilst self incrimination by an identifiable practitioner is fanciful, a different situation applies when data is collected under the security afforded by anonymity, as practiced by the EoLCRG.
When the Commission’s annual reports are examined, exclusively on a quantitative basis, considerable bracket creep is apparent, as already referenced for the Netherlands.. Annual numbers of euthanasia cases have increased regularly – 235 (2003), 495 (2007), 704 (2008), 1133 (2011), 1432 (2012) and 1807 (2013). Whilst these numbers include only reported cases of euthanasia, the prevalence of unreported cases has been documented by the End of Life Care Research Group. In an article in the British Medical Journal, (341: c5174), Smets reported. that approximately only 50% (549 out of 1070 cases) of the estimated cases of euthanasia had been reported.
The ELCRG data was collected from practitioners on the basis of anonymity and it also included information about compliance with regulations. The BMJ article concluded that: Unreported cases were generally dealt with less carefully than reported cases: a written request for euthanasia was more often absent, other physicians and caregivers specialised in palliative care were consulted less often. The Victorian Select Committee’s presumption that decriminalisation would eliminate illegal assisted suicide and euthanasia appears to be quite dubious.
Another of the Victorian assertions, namely that decrimnalisation would reduce the frequency of non-consensual euthanasia was evidently made in ignorance of the EoLCRG’s research. In a 2010 group survey published in the Canadian Medical Association Journal, Chambaere reported that euthanasia was undertaken in the absence of a request in 66 out of a series of 208 cases.
So much for quantitative evidence of Belgian bracket creep. What about qualitative creep? Legalisation of euthanasia for subjects under the age of 18 was precluded in the original legislation. Subsequent parliamentary submissions asserting that this constituted a breach of their rights proved persuasive with legislators, a conclusion reinforced by requests from paediatricians at the Queen Fabiola Hospital for legal immunity.
A second instance of legislated qualitative creep concerned people with dementia. Original requirements were that demented patients could only be eligible if a request had been made in the preceding 5 years, in anticipation of dementia. This timing constraint was abolished together with the original requirement that the patient be comatose by the time that euthanasia was undertaken.
The original regulation in Belgium required that, if there was any question about a patient’s mental competence, or if a request for euthanasia was possibly influenced by a mental condition, especially depression, then assessment by a psychiatrist was indicated. Fast forward 13 years and mental illness, far from constituting an impediment to euthanasia, had become a justification for it (more of a leap than a creep). The irrepressible chair of the Belgian Commission, Wim Distelmans, in a 2015 interview published in the magazine Humo, disclosed that: At least 50 patients are euthanased each year for purely psychiatric reasons.
This is a small group – 50 to 60 patients – but it is not a negligible number: 2 to 3 percent of the 1,924 people who were euthanased last year. Usually they are not old, but they have suffered for a long time. They feel that they are no longer comfortable in this world: “All my relationships have failed.” They are in their 20s or 30s.
Qualitative bracket creep has been reflected not only in the scope of Belgian patient eligibility but also in the procedural requirements for undertaking euthanasia. For example, in a 2010 report in the Canadian Medical Association Journal, Inghelbrecht examined compliance with the regulatory requirement that the lethal medication could only be administered by a medical practitioner.
It was reported that:
The life-ending drugs were administered by the nurse in 12% of the cases of euthanasia, as compared with 45% of the cases of assisted death without an explicit request. In both types of assisted death, the nurses acted on the physician’s orders but mostly in the physician’s absence The authors concluded that: By administering the life-ending drugs in some of the cases of euthanasia, and in almost half of the cases without an explicit request from the patient, the nurses in our study operated beyond the legal margins of their profession.
In another qualitative deviation from legislated requirements, Smets queried in the British Journal of General Practice, the extent to which the regulatory requirement for consultation with a second practitioner was ignored when euthanasia was undertaken in a patient’s home. Commenting on these questionable deviations from regulations, it was pointed out that: Physicians may therefore be inclined to report only those cases where due care has been taken, or present their cases as compliant with the law.- After 7 years of legalised euthanasia, it is thus still unknown how it is actually practised by physicians in Belgium
It has been suggested, on the basis of reports of the frequency with which continuous deep sedation is often provided in nursing homes, although not reported by practitioners as euthanasia, that many instances of interventions which could qualify as euthanasia are not notified as such. A 2014 survey was reported in the Journal of the American Medical Directors Association (titled “Continuous sedation until death, with or without, the intention to hasten death – A nationwide study in nursing homes in Flanders, Belgium”)
Whilst the reported frequency of euthanasia in nursing homes in Flanders did not increase to the same extent as was reported in the community at large, the institutional frequency of death following continuous deep sedation increased (5.6% – 2001, 8.2% – 2005 and 12.3% – 2010). This item is recorded separately in the official statistics and the extent to which it has progressively replaced euthanasia in nursing homes is unclear.
In winding up this catalogue of varieties of bracket creep a question could arise of how the Belgian government has reacted. In 2015, in the course of parliamentary debate about bringing regulation into line with the evolved practice, Christian Democrat MP Els Van Hoof was reported in Knack as having: repeatedly asked the Commission why should we decide to expand if an assessment had not been made of the existing law on euthanasia. For it was clear that the law showed serious gaps, particularly with regard to the ‘safety valves’ that are built in to prevent abuses. Regulation of this, and many other aspects of the operation of the existing law, had been patently non-existent.
Evolution of Practice in the Netherlands.
As a detail additional to those presented above about frequency of Dutch assisted suicide/euthanasia cases, in 2015 these accounted for 4.5% of all deaths. Recalling the Belgian statistics, the question of actual ascertainment of the frequency of medically undertaken deaths in the Netherlands indicates this to have been substantially in excess of that of reported ones. A 2012 article in The Lancet concluded that 23% of cases were not reported by the responsible practitioner.
. A 2007 paper in the New England Journal of Medicine dealing with involuntary euthanasia reported that, of 2410 Dutch euthanasia deaths in 2005, more than 560 people received lethal substances without having given explicit consent. Of relevance to this omission, it was reported in the British Medical Journal in 2007 that, in 35% of cases of involuntary euthanasia, the opinion of a second independent physician, required by regulation, was not sought.
Instances of amendment of regulations in response to published surveys of their non-observance in practice have not been unknown. An example of this was a report in a paper in JAMA Psychiatry emanating from the US National Institutes of Health (NIH) which documented the use of assisted suicide/euthanasia for Dutch psychiatric patients. As usual, this discrepancy between regulation and practice was remedied by legislated adjustment authorising the euthanasia of severely demented patients.
The clinic referred to in the original NIH article and in the editorial commentary on the article was comprehensively described in 2012:
The Dutch right-to-die organization (NVVE ) is now offering a mobile euthanasia service, prompting accusations that the law has been pushed too far. Teams will be travelling around the country assisting patients whose own doctors refuse to help them to die. The new units consist of a doctor, a nurse and all the medical equipment required to carry out euthanasia.
The operation of mobile clinics has generated considerable concern on the part of many Dutch physicians because of its effective elimination of any possibility of a previous doctor-patient relationship. Nevertheless, the Victorian Select Committee claimed that:
Now for some breaking news. In 2013 the question of whether ‘completed life’ (meaning that a person considered that s/he had had enough) could provide grounds for euthanasia was examined. The resulting Schnabel report considered that existing provisions for eligibility for euthanasia were adequate, incidentally provoking considerable criticism from the NVVE Jongeren, the youth wing of the Dutch right-to-die association (perhaps there’s something to the notion of accelerated inheritance).
Subsequently, in October 2016, a press release on behalf of the Dutch Government informed the nation that:
People who have come to a well-considered decision that their life is completed must be allowed, under strict and carefully identified criteria, to end their life with dignity. The government intends to develop new legislation based on this principle. The system would exist alongside and separately from the current legislation on euthanasia.
In July 2017, another press release indicated that a bill to enable this practice will be introduced by the end of 2017. The press release continued:
The government is firmly committed to protecting human life. This means that the preconditions under which people would be allowed to exercise their freedom to choose how their life ends will be an essential component of the solution the government seeks. The voluntary, well-considered nature of such decisions, as well as safety and due care, are the guiding principles in this respect.
Dutch regulation is certainly not static.
Evolution of Practice in Oregon.
The practice of assisted suicide in Oregon differs from that in Benelux in the comparative paucity of information available. The Oregon Science and Health University at Portland, cited above, highlighted lack of mandatory reporting and of enforceable penalties for failure to report. Frequency of non-compliance is unknown .and there is no funding specified for collection and analysis of such data. Nevertheless, some concerning qualitative aspects of assisted suicide, as practised in Oregon have been documented.
In the first place, the term ‘suicide’ is as verboten in Oregon regulations as it was in the Victorian Select Committee report. It is a regulatory requirement that the word ‘suicide’ is not to appear on any death certificate (this has to be ’doctored’ to record as the cause of death, the medical condition which qualified the person for assisted suicide) – not conducive to data collection.
One serious concern in any system in which a patient is provided with lethal medication to be ingested privately at a time and place of his/her choosing relates to risks of coercion. Whilst much talk about regulation to avoid coercion relates to circumstances at the time of prescription, a more realistic risk assessment would place the risk at the time of ingestion. In 2013 the prescriber attended in 15.7% of Oregon cases whilst, in 2014, the attendance rate was 13.9%.
A requirement of the Oregon Death with Dignity Act is that the decision of the primary practitioner is to be confirmed by an independent second practitioner. This requirement was achieved in 58 of one reported group of 61 consecutive patients by a member of the Compassion and Choices group (originally the Hemlock Society). Whilst the adjective ‘independent’ appears almost universally in ‘robust’ regulatory frameworks, it remains unclear from what the second practitioner is meant to be independent.
Considerable evidence has emerged from a number of studies that, compared with non-depressed patients, patients who are depressed are more likely to request euthanasia and that treatment for depression will often result in the patient rescinding the request. Whilst Oregon regulation requires that, if either physician believes the patient’s judgement is impaired by a psychiatric or psychological disorder, the patient must be referred for a psychological examination, this rarely occurs. For instance, only 2 of the 71 patients who died in 2013 were actually referred.
When considering the attitudes of people seeking assisted suicide, the official Oregon report scores the most frequent end of life concern cited is not pain but ‘loss of autonomy’ (91.5%) whilst inadequate pain control or concern about it was much less common (24.7%).
One of the most subtle, albeit possibly the most ominous, evolving features of assisted suicide as practised in Oregon has been the apparent trend for its infiltration into, and effect on, the health system. An early indication of this trend was provided in the course of the Californian senate hearings preceding the decriminalisation of assisted suicide in that state. Whilst the reference to Oregon which received the most publicity on that occasion was undoubtedly the testimony of Brittany Maynard prior to her return to that state to undertake assisted suicide very little attention was directed to testimony from 2 other Oregon residents. Their concern was that they had been informed that, while the department:
… wouldn’t pay for prescribed chemotherapy it could fund their assisted suicide.
under ORS 127.800-127.897 (Oregon Death with Dignity Act).
Whilst insufficient medical detail was provided. to assess the potential value of the prescribed medication for them, the location of decision making about medical treatment as against assisted suicide within a common department was worrying and subsequent events have served to reinforce that.
An indication of the interrelated sourcing of decisions about assisted suicide and ‘conventional’ medical treatment was documented, perhaps inadvertently, in the Victorian Select Committee report. Common perceptions about the background to predicting that a patient was within 6 months of death would, I suspect, be based on considerable medical experience of the likely prognostic implications of specific features of particular medical conditions. Observations by the Victorian Select Committee suggest otherwise:
That categorization as ‘terminal’ is acknowledged to be a bureaucratic rather than a clinical judgment, given that this categorization confers eligibility for assisted suicide with substantial consequent budgetary savings to the health system is further cause for concern
The combined consideration of funding of expensive medical care versus assisted suicide, notwithstanding some denials from the industry, could be becoming a feature of US health insurance practice. In mid 2017, the Washington Times carried a recent report from a Nevada practitioner who claimed that two of his patients were denied medication which offered a better than 50% chance of cure but were instead offered insurance funding for assisted suicide.
Evolution of Practice in Canada.
Whilst euthanasia has only been decriminalized for little more than a year in Canada, several events are worth noting. The Canadian Supreme Court explicitly dismissed any risk of ‘slippery slope’ leading to increase in prevalence as had occurred in other jurisdictions: the permissive regime in Belgium is the product of a very different medico-legal culture. (para 112). Twelve months after decriminalization, Dr Ellen Wiebe who was reputed to be the practitioner in more than 40 of the 970 enabled deaths was reported as predicting that Canadian prevalence would rapidly rise to equate with that of Belgium.
The Ontario parliament rejected a motion to the effect that practitioners who were unwilling to refer patients to another doctor for euthanasia would not be subject to prosecution. Other provincial parliaments had moved to confer immunity from prosecution.
Quite recently, an Ontario court upheld the request for euthanasia of a 77 year old patient who had suffered from osteoarthritis for 4 decades. The judgment asserted that, as she was approaching the end of life because of her age, she qualified. As Canadian regulations require that the cause of death be certified as the pre-existing condition which led to a patient’s request, osteoarthritis shall appear on her death certificate. Watch this space.
This paper has been concerned with documented historical events, not with predictions. Such events are critical in relation to ‘one line’ dismissals of any possibility of ‘slippery slope’ consequent on the passage of proposed Victorian legislation. Whether one line dismissals are based on ignorance of recent history or reflect intentional misrepresentation is not the subject of this paper. Irrespective of which is the explanation, I submit that failure to take account of that history renders assertions concerning the implausibility of slippery slopes inherently invalid. Furthermore, it calls into question the credibility of the Victorian Legislative Council Select Committee in general.
The contention of that committee’s report that: Concerns raised in arguments against legalising assisted dying — such as the inability to implement and maintain effective safeguards — have not eventuated in jurisdictions where assisted dying is legal. The Committee did not find compelling evidence to support the negative consequences predicted by these claims… lacks a factual basis.