Rhetoric: Suicide 'contagion'

0
"HOPE's" Branka van der Linden and the ACA misrepresent figures, again

Here we go again. Branka van der Linden of Catholic anti-VAD website “HOPE”, and the Australian Care Alliance — endorsed by a number of well-known, committed Catholic doctors — have just published more egregious misinformation against VAD. This time they've collectively piled it on Victoria's general suicide statistics, recently updated by the Victorian Coroner. So what did they say, and how did it misrepresent the actual situation? Let's take a look.

The reason the statistics are being discussed is because in 2017, Victoria's parliament legalised voluntary assisted dying (VAD) for the terminally ill. The law came into effect halfway through 2019, and 2020 was the first full year of its operation.

Australian Care Alliance gets the basics wrong

Here's the Australian Care Alliance's (ACA) splashy page trumpeting that Victoria's suicide rate has jumped 21.2% from 694 in 2017 to 842 in 2020.

acagetsstatswrongmarch2021.jpgFigure 1: ACA's splashy page trumpeting a 21.2% increase in Victorian suicides

That's... interesting. According to the Victorian coroner's official figures, there were indeed 694 suicides in Victoria in 2017. However, in 2020 the coroner's figure is actually 698, not 842 as claimed by the ACA. According to the ACA, Victoria's suicide count data looks like this (Figure 2).

acasuiciefigurewrong2021.gifFigure 2: The ACA polemically claims that Victoria's suicide count has increased 21.2%

So, how did the ACA reach a count of 842? Well, their argument is to shamefully and humiliatingly disrespect Victoria's terminally ill who died peacefully under its VAD law in 2020 — 144 of them according to the official 2020 reports of Victoria's Voluntary Assisted Dying Review Board — and add them to the coronial count of 698 suicides.

The ACA points out that VAD supporters have said that legalising VAD should decrease Victoria's general suicides by about 50 cases a year, but say the count's gone up substantially instead. See how they craftily deploy logical fallacy to fabricate a crisis?

Arguing that VAD law must reduce the suicide count by 50 cases a year (but seemingly didn't) and at the same time adding VAD cases to the suicide count to complain that it's gone up, requires at least three assumptions:

  1. that all terminally ill violent suiciders now automatically qualify for and easily gain access to VAD; and
  2. that nobody else with a terminal illness who would not have chosen violent suicide, should or would use the law; and
  3. that no other factors make a significant difference to trends in general suicides.

 
All these assumptions are patently false.

Obviously, some people will not legally qualify for VAD; for example, amongst its restrictions it requires death to be expected with 6 months; 12 months for a small set of specific illnesses.

Obviously, some who would not have suicided but instead would suffer intolerably and against their wishes until death, will now choose to pursue VAD.

And obviously, well-known factors such as rates of mental illness, substance abuse, intimate relationship troubles, bullying, financial or legal difficulties, and other factors are major influencers of general suicide rates. But to the ACA, the only factor that supposedly has any effect is the one they are ideologically opposed to: VAD.

It's worrisome that this nonsense is sold to the public by ACA's supposed experts: “health professionals and lawyers”.

Cherry-picked overseas data, too

The ACA's ideological bias is further revealed by their website page about the “social contagion of suicide”. In it, they cite as authoritative, the 2015 Jones and Paton (both firm Catholics) article purporting to show 6.3% suicide contagion from VAD to the general population. I've comprehensively exposed that article as an ideologically-driven mathematical farce fuelled by no fewer than ten major scientific offenses. It's interesting that the ACAs methodology is just like Jones' and Patons': reporting VAD supporter statements that legalisation should decrease the general suicide rate, and then adding VAD deaths to conclude the opposite.

They also commit one of Jones' and Patons' other offences: selectively quoting data from other studies that might be seen to support their theory, but excluding critical alternative information from the same study that runs counter to the theory.

The ACA cites a Swiss study to breathlessly report that 6.5% of those who witnessed an assisted death in that country experienced sub-threshold PTSD, and 13% full PTSD. The ACA expressly states:

“Like any other suicide, assisted suicide can profoundly affect surviving family members and friends.” — The Australian Care Alliance

There you have it: the ACA draws a direct equivalence between peaceful VAD deaths in the face of terminal illness and with loved ones present, and lonely, violent deaths by general suicide.

The ACA cites no other relevant material from the Swiss article. That's revealing, because the article clearly reported that the PTSD rates were higher than in the general population. There's what the ACA left out: the PTSD rates were higher than for almost everyone else who hadn't just suffered the loss of a loved one.

To draw valid and meaningful interpretations, it is necessary to compare the bereavement challenges of VAD family versus families of general deaths, deaths in the face of extreme suffering without hastened death, and cases of violent suicide. As I've published before from peer-reviewed studies, bereavement symptoms of VAD family are at least as good as, and can be better than those where the deceased has suffered in extremis at the end of life, and certainly relative to violent suicides.

The ACA also doesn't mention that the Swiss study found a "prevalence of complicated grief ... comparable to that reported for the general Swiss population". It's not like the information was hard to find. It's right there in the Abstract on the front page of the article.

That the ACA cherry-picked a couple of Swiss data points while omitting key “unhelpful” information, and argued, by linking the selected cherries with the above quote, that said Swiss data established something it clearly did not (that VAD deaths supposedly cause similar family trauma as violent suicides), suggests an astonishing degree of ignorance.

The ACA's cherry-picking of data, while omitting key unhelpful information, suggests an astonishing degree of ignorance.

Enough of that.

Branka van der Linden cherry-picks, too

I've crossed pens (or is that keyboards?) with Ms van der Linden several times before in regard to misinformation. She misinforms on this matter, too.

Curiously, like the ACA and also without explanation, she cherry-picks just the 2017 and 2020 suicide counts from the Victorian coroner's report (Figure 3). You'd think this was the only data in the report, but no, it isn't.

vanderlindenvicsuicidestats2021.gifFigure 3: Branca van der Linden's version of Victorian suicide counts by year

She uses these two figures to argue that said drop of 50 cases per year hasn't happened. This employs the same fallacies as the ACA: suggesting that two single data points strongly support a hypothesis, and assuming that the thing one is ideologically opposed to, VAD, is the only thing to alter the rate of general suicides over time.

Like the ACA, she also suggests adding the VAD figures to the coroner's general suicide data to say that in that case, suicides have increased significantly.

Both the Australian Care Alliance and Branca van der Linden cherry-pick just two data points from more full and robust longitudinal data to try and argue their case against VAD.

So what does the coroner's full data set look like?

The actual numbers

The Victorian coroner's 2021 report into suicides contains data for all years 2016 to 2020 inclusive. And it looks like this (Figure 4).

viccoronersuicides2021.gifFigure 4: The complete set of data from the Victorian coroner's report on suicide counts per year

Now we're beginning to see a possible reason as to why the ACA and Ms van der Linden chose just two data points. Remember that VAD was legalised by the Victorian parliament in 2017. The law was not in effect for 2017, 2018, or the first half of 2019.

Well, the data clearly suggests an increasing suicide count trend up to 2018. The upward trend stops in 2019, when VAD was in operation for the second half of the year. And in 2020, the first full operational year of VAD, the upward trend has been interrupted by a downward result. Neither the ACL nor Ms van der Linden mention this.

Neither the Australian Care Alliance nor Branca van der Linden mention the fuller, longitudinal data that doesn't support, and indeed appears hostile to, their hypothesis.

Update 19-Mar-2021

I thought it so obvious that I didn't write it up, but a colleague points out it's important to highlight, that in picking just two data points to stake their claim, the ACA and Ms van der Linden chose 2017, and not 2018, as their reference year. To compare “after” with “before” in the most basicly valid manner (full longitudinal data is better), it is appropriate to compare the last data point that completely excludes the new condition (VAD law in operation), with the first data point that fully includes it.

Those years are 2018 (none of the year) and 2020 (all of the year). But the ACA and Ms van der Linden didn't pick 2018, they picked 2017.

What possible reason might explain that? Well, by comparing 2017 with 2020, they got to say that the general suicide count increased by 2 from 694 to 698. However, had they more validly compared 2018 with 2020, they would have had to report a drop of 19 from 717 to 698.

And that would have contradicted their flimsy confection that suicides hadn't gone down after VAD was introduced.

But even the raw suicide count statistics are a bit misleading.

Interpreting suicide data correctly

Using raw counts to compare suicide statistics (e.g. year to year or place to place) is lazy and wrong. All other things being equal, if you had twice the population, you'd expect twice the suicide count. To make valid comparisons, you have to compare rates, not raw counts. This is relevant because populations obviously change over time, and Victoria between 2016 and 2020 was no exception.

I've retrieved the official Victorian population figures by year and computed the standard official suicide rate statistic: suicides per 100,000 population. The Victorian suicide rates look like this (Figure 5):

vicsuiciderate2016-2020.gifFigure 5: Victorian suicides per 100k population by year

The data shows a rising suicide rate from 2016 to 2018, a levelling off in 2019 in which VAD was operational for half the year, and a fall back to the 2016 rate in 2020, the first full operational year of VAD.

Computing from the rate drop between 2018 (11.4 with no VAD law) and 2020 (10.8, first full year of VAD law), the equivalent count of suicide decrease in 2020 was 38 persons. And that's without assuming the general suicide rate would have continued its rising trend.

The equivalent suicide decrease from 2018 to 2020 was 38 persons.

Getting all the numbers right

The ACA correctly cites then Minister for Health, Ms Jill Hennessy, as stating in 2017 that "Evidence from the coroner indicated that one terminally ill Victorian was taking their life each week." That would be 52 cases a year, which the ACA rounds out to 50 a year. The headline figure from the coroner's report actually calculates to 48. No biggie, just round numbers.

But the figure is quite wrong. You have to read the coroner's special 2017 report to the Victorian parliament regarding suicides in cases of illness, to calculate the correct numbers.

The coroner's report didn't just include suicide data for terminally ill people. It also included cases of advanced incurable but not terminal illness, and cases of severe suffering resulting from injuries. So the terminal illness data (to which the VAD law is relevant) is a fraction of the total. We can calculate from the Tables in the report that 23% of the cases were in respect of injuries, so that leaves 77% for terminal and other advanced illnesses.

Of the illnesses listed, the relevant one as a proxy measure for terminal illness is “cancer”, and that comprises 50% of the illness cases. So, 50% of 77% of 48 cases a year = 19 cases a year in respect of terminal illness.

So that's an actual likely decrease of 19 suicide cases a year, compared with an equivalent drop in the actual data of 38 persons in the first full year.

The actual annual count of general suicides in respect of terminal illness, as reported to the Victorian parliament by the state coroner in 2017, was 19 persons a year, and not 50 as widely stated.

Don't get carried away

It's imporant to note that citing this interesting numeric analysis as “proof” of the law's effectiveness in respect of reducing Victoria's suicide rate, would, at present, be an overconfident claim. While far more firmly based in proper forms of evidence than the vapid nonsense promoted by the ACA and Ms van der Linden, this is a correlation. Correlation does not equal causation: the ACA and Ms van der Linden should remember that. For example, 2020 was a very unusual, Covid-19-dominated annus horribilis, which may have affected suicide rates in unexpected ways.

While the coroner's fuller data set so far is consistent with reasoned expectations of suicide substitution, it is premature to conclude the data proves the principle. More years' data, and more detailed, causative analysis involving the control of confounding factors, is necessary before reaching greater certainty in the association.

But as I've published in detailed and extensive analyses based on robust official data, so far all the longitudinal data on suicide rates in jurisdictions where VAD is lawful is consistent with suicide substitution, not suicide contagion. Some VAD opponents just cherry-pick their way through tidbits to try and argue the opposite.

To date, all the robust, longitudinal data on suicides in jurisdictions where VAD is lawful is consistent with suicide substitution, not suicide contagion.

Conclusion

The Australian Care Alliance and Ms van der Linden disgracefully cherry-pick and misrepresent Victoria's recent suicide data in a manner consistent with their own theories, while proper and appropriate analysis of the full data available shows results inconsistent with their hypothesis, and currently consistent with the opposite.

To paraphrase Ms van der Linden's own statement: “It is unfortunate that the deaths of terminally ill Victorians were politicised so shamelessly by [anti-]euthanasia activists for their own ends.”

These continued cherry-picked data gaffes are an embarrassment to their promoters.


Share This Post:
0
A forensic analysis exposes Theo Boer's smoke and mirrors on 'suicide contagion'

In my most recent article in the Journal of Assisted Dying, I forensically analyse Dutch ethicist Professor Theo Boer’s 2017 paper purporting to find suicide contagion from assisted dying in the Netherlands. It doesn’t go well for Professor Boer, to put it mildly. You can find the full article here.

I also find an astonishing coincidence that occurred in 2014, the year Boer went feral against the Dutch euthanasia law.

Multiple fatal flaws

In the ‘analysis’ outlined in his article, Boer commits a number of fatal scientific no-noes, including failing to analyse the variable he actually surmised might cause suicide contagion, cherry-picking data that supported his conclusion while ignoring or offhandedly dismissing data at odds with his conclusion, and wrongly forming a causative conclusion from a simple correlation while failing to control for any confounding variables of which there are many.

A litany of scientific offences

In addition to the fatal flaws, Boer’s article contains numerous other scientific and academic offences. My forensic analysis concludes:

“In summary, Boer’s article contains a litany of scientific and scholarly failures. Its speculations are ill-informed, poorly-assembled, incoherent in places and mostly uncited, the data cherry-picked and invalidly interpreted, and the laissez faire methodology incapable of validly supporting its conclusion.
 

Boer conjures up mere smoke and mirrors to argue suicide contagion from VAD in the Netherlands. The article should be retracted.”

The article also reflects badly on the journal that published this smoke and mirrors: the Journal of Ethics in Mental Health. Neither peer review nor editorial effort identified or attempted to correct any of the nonsense in the article.

What was he thinking?

Professor Boer is an expert in Reformist Protestant theology. As a religious ethicist, it’s astonishing that he considered himself suited to conducting and publishing a ‘causative’ scientific study.

In his article, Boer proposed VAD as the only factor to contribute to changes in the Netherlands’ general suicide rate (and dismissed the Belgian data which contradicted his theory).

In reality, numerous risk and protective factors affect the suicide rate, and in the Netherlands as I’ve established using their official government data, just one factor — unemployment — explains 80% of the variance in the Dutch suicide rate since 1960. Boer casually dismisses this without providing the faintest fume of an empirical analysis himself.

Boer’s article did little but amply demonstrate his underlying anchoring and confirmation bias on the subject, his unfamiliarity with the complexity of suicide, and ignorance of proper scientific principles.

For good measure, he casually threw in a comment about “suicide contagion” or copycat suicides, without understanding that in suicide, copying is the method of causing death. But by definition, general suiciders don’t follow the provisions of the euthanasia Act.

His endeavour made as little sense as me writing a conclusive article about Reformist Protestant theology, about which I know very little.

A copycat analysis?

Coincidentally, the structure of the storyline, the litany of scientific offences committed, and the conclusions reached in Boer’s article were surprisingly similar to those in an ‘analysis’ of Oregon’s suicide rate in another paper by Jones and Paton. Like Boer, Jones and Paton start out by surmising that assisted dying ought to lower the general suicide rate, and conclude the opposite.

Boer approvingly cites the Jones and Paton article, even though a forensic analysis found no fewer than ten major scientific flaws in it and provided multiple sources of empirical evidence at odds with the article’s conclusions.

But Boer manages to cock even the citation up, referring to the article’s authors as Holmes and Paton.

Will the real Theo Boer please stand up?

Boer notes that he’s always been a euthanasia sceptic. Nevertheless, as a Reformist Protestant, he had long accepted assisted dying in “emergency” situations, of which intolerable and otherwise unrelievable suffering is a ‘qualifying’ criterion, and which is the substance of the Dutch euthanasia law (it’s regarded in legal circles as a law of “necessity”). He also opined that the Dutch model was a decent one that other jurisdictions could emulate.

Boer served as the ethicist member of one of the five Dutch euthanasia review commissions, examining every case reported to it between 2005 and 2014.

In 2014 he publicly quit his post on the review committee, slamming the Dutch assisted dying system. He’s been badmouthing it to anyone who will listen, since.

In preparation for this analysis, I asked Boer if his vocal opposition to the Dutch assisted dying model was now based on an in-principle opposition to assisted dying, or only in regard to more recent practice under the Dutch euthanasia Act. Despite a couple of iterations, I didn’t get a specific answer.

The law hasn’t changed

Here’s the point. While Boer repeatedly opines that things changed radically in the Netherlands around 2007, the country’s euthanasia Act hasn’t changed since it was passed in 2001 (and came into effect in 2002). Not. One. Word.

In addition, the Dutch Supreme Court determined in 1994 that individuals with mental (in the absence of concomitant physical) illness could qualify under the then regulatory euthanasia framework, and it was found that cases occurred every year.

And the 2001 Act formalised in statute the regulatory framework that had existed since at least 1984, when the Dutch medical association first published guidelines for euthanasia.

Thus, the Act reflects very long-standing practice, and it hasn’t changed since it was enacted, in contrast to Boer’s claim that things have radically changed.

Flimsy and incoherent ‘ethics’ part 1

This brings us to the first fatal incoherence of Boer’s “ethics”: that he now opposes the law because people with psychiatric illness and other conditions are, in slightly increasing numbers, availing themselves of the euthanasia law. It is these cases against which Boer rails, despite having previously said the Dutch model is a good example for the world, and having actively participated in the system.

Boer’s flip flop is to argue that a law that permits assisted dying under a range of medical conditions (and has done so for decades) is a good law, provided some of those who might qualify (like psychiatric cases) never use it.

Try and explain the ethics behind that position.

Flimsy and incoherent ‘ethics’ part 2

The second fatal incoherence of Boer’s ‘ethics’ is his repeated complaint that until around 2007, the numbers of euthanasia cases was “somewhat steady”, but increased after that. Never mind that the majority of the increase was still in relation to terminal cancer: Boer simply railed at the increased numbers as a major problem.

But, try and explain using ethical principles, why it is appropriate for 2,000 people a year to avail themselves of the euthanasia law, but inappropriate for 4,000 (who all qualify)?

Indeed, the Dutch euthanasia Act makes no mention of numbers: there is no legislated limit on the count of people who might choose to use the law. Rather, it is based on due care criteria, outlining the circumstances of who may qualify, and the process by which they may.

The legislature’s intent remains unchanged and is still being adhered to, though more people, the majority of whom have terminal cancer, are using the law.

It’s astonishing that a Professor of Ethics fails to reflect on the fatal incoherence of his own ‘ethical’ arguments.

What happened?

Boer, who had supported and promoted the Dutch euthanasia model suddenly and incoherently changed his position to vocally opposed in 2014. What happened?

One factor might shed some light. In 2014, Boer was appointed to the endowed professorship of Lindeboom Chair in Ethics in Healthcare at Kampen Theological University.

While Kampen Theological University is a Dutch Reformist Protestant institution and therefore may support assisted dying in “emergency” cases, the Lindeboom Institute, which endows Boer’s eponymous professorship, is less understanding.

The Lindeboom Institute was co-founded by several orthodox Christian institutions and cooperates with the Netherlands Evangelical University which studies science from an creationist Biblical perspective.

The Institute demands “biblically sound medical ethics” along with “Christian norms and values”. You’d be left wondering what that actually means, until you find on its website that the Board’s role is “the protection of people at all stages of life”.

In addition, participating organisations that fund the Lindeboom endowment, like the Dutch Patients Association, Pro Life Health Insurance and the Foundation for Christian Philosophy, are strongly opposed to assisted dying in any form.

It turns out that the authors of that other ‘analysis’ that commits numerous similar scientific offences which generate smoke and mirrors, Jones and Paton, are devout conservative Catholics.

Gosh. What a coincidence.


Share This Post:

Netherlands 'suicide contagion' from assisted dying: Theo Boer's smoke and mirrors


Author(s)

Neil Francis

Journal

Journal of Assisted Dying, vol. 4, no. 1, pp. 1–11.

Abstract

Background: Concerns had been raised about the scientific quality of a 2017 article by ethicist Theo Boer in which he theorised that lawful voluntary assisted dying (VAD) would potentially ‘dampen’ suicide rates, but drew the opposite conclusion: the suggestion that VAD cases have caused higher suicide rates.
Methods: A structured, forensic examination of the article was conducted.
Results: Numerous serious shortcomings were found, including (a) profound unfamiliarity with the complexity of suicide; (b) lack of a clear and specific pre-hoc methodology; (c) numerous unsupported speculations; (d) cherry-picked data and casual dismissal of data at odds with the conclusion; (e) a simple correlation interpreted as causation while failing to control for any confounding factors; (f) incoherent, contradictory and misleading statements; and (g) multiple editorial errors.
Conclusions: Boer’s article is poorly conceived and carelessly assembled, revealing unfamiliarity with both the subject matter and with scientific principles. The conclusions drawn are not supported by the article’s methodology or data. The article offers mere smoke and mirrors to conclude that VAD may increase suicide rates, at odds with wider evidence.

Article keywords

voluntary assisted dying, euthanasia, suicide contagion, Werther effect, Netherlands, methodology

Full PDF

Download the full PDF: Download the full article (5.4Mb)

Citation

Francis, N 2019, 'Netherlands "suicide contagion" from assisted dying: Theo Boer's smoke and mirrors', Journal of Assisted Dying, vol. 4, no. 1, pp. 1-11.

Download the citation in RIS format: RIS.gif


Share This Post:
0
Jones, Paton and Kheriaty's articles demonstrate poor science and multiple, egregious instances of bias.

In 2015, Dr David Jones and Prof. David Paton published an article titled “How does legalization of physician-assisted suicide affect rates of suicide?” in the Southern Medical Journal. The article purported to establish suicide contagion from Oregon and Washington Death With Dignity Act (DWDA) deaths to “total suicides.”  It also purported to establish no decrease in general suicide rates, which Jones & Paton argued should occur by substitution of assisted death for some general suicides. (Notice how these two ‘expected’ results — an anticipated rise and an anticipated fall in suicide rates — are at odds in principle.)

In my thorough and empirically-backed response, I expose the disgraceful playbook of these authors as they shambolically commit no fewer than ten deadly sins against science in the pursuit of their opposition to lawful assisted dying.

Get the full report here

Executive Summary

In 2015, Dr David Jones & Prof. David Paton published an article in the Southern Medical Journal titled “How does legalization of physician-assisted suicide affect rates of suicide?” This study examines the article, as well as an enthusiastic editorial of it by Dr Aaron Kheriaty in the same journal issue, both of which portray “suicide contagion” from Oregon and Washington’s death with dignity acts (DWDA).

However, while contagion from general suicides is a well-established phenomenon, there are multiple sound reasons to reject contagion theory in relation to assisted deaths, including:

  • Most healthcare professionals readily acknowledge key differences in the characteristics of assisted deaths: for example, a fully informed, tested and rational decision with shared decision-making.
  • Those using Oregon and Washington’s DWDAs are, by qualifying for it, already actively dying. Thus, they are choosing between two ways of dying rather than between living and dying.
  • Most of those using the DWDA discuss it with their families (expected, peaceful death), whereas most general suicides occur in isolation and without discussion (unexpected, often violent death).
  • Multiple studies show that while families of general suicide experience complicated bereavement, families of assisted dying cope at least as well as, and in some cases better than, the general population or those who considered but did not pursue assisted death.

 
Even if “suicide contagion from assisted dying” theory were sound, direct evidence from official government sources shows that the number of potential suicides in Oregon in 2014 would have been fewer than 2 in 855 cases: undetectable by general modelling methods.

Jones & Paton’s article title conveys an air of skilled and scientific neutrality. However, close examination of the article, and Kheriaty’s editorialisation of it, reveals least ten serious flaws or ‘scientific sins.’

The authors demonstrated little understanding of the complex issues surrounding suicide, willingness to unjustifiably equate assisted dying with general suicide, contentment with failing to search for, consider or include contrary evidence including from sources they cite to argue their case, unreasonable trust in a model that couldn’t hope to legitimately resolve their premises, satisfaction with executing their model amateurishly, a disposition to overstate confidence of causation in the absence of meaningful statistical correlations, and an inclination for emphasising results in accordance with their theories while de-emphasising or ignoring others.

Any of these flaws was serious enough to invalidate Jones & Paton’s article and Kheriaty’s conclusions of it, yet there is not one deadly flaw: there are at least ten.

Their claim of a supposed 6.3% suicide contagion rate from assisted dying in Oregon and Washington is a conceptual and mathematical farce.

The Southern Medical Journal is a peer-reviewed journal. However, it is difficult to reconcile the rigorous standards and sound reputation that peer review is intended to maintain, with the numerous, egregious flaws in this study and its dissemination.

Rather than inform the ongoing conversation about lawful assisted dying, the Jones & Paton and Kheriaty articles misinform and inflame it.

Given the numerous egregious flaws, both articles ought to be retracted.

 

Get the full report here

Share This Post:
0
The deeply-flawed Jones & Paton, and Kheriaty articles purporting to show suicide contagion.

In the ongoing political campaign against assisted dying law reform, opponents have spread one piece of egregious misinformation after another. One of the most common is supposed “suicide contagion” from assisted dying laws to general suicide, a theory popularised by Catholic Prof. Margaret Somerville. Despite the nonsense of her claim being comprehensively exposed, she still believes that her opinion “will prove to be correct.” Two journal papers published in 2015 purported to, but didn't, establish suicide contagion in Oregon and Washington states.

Note: the report is now published here.

Assisted dying law reform opponents are still relying on a 2015 paper by Catholics David Jones and David Paton, bolstered by a glowing editorial of it written by Catholic psychiatrist Aaron Kheriaty, published in the Southern Medical Journal, as continued ‘proof’ of suicide contagion theory, at least in respect of USA states Oregon and Washington (since data from other lawful jurisdictions contradicts the theory).

Jones & Paton’s article reported the use of econometric modelling to test for ‘suicide contagion’ from Oregon and Washington’s Death With Dignity Act (DWDA) laws. But, in an exposé to be published this week, no fewer than ten ‘deadly sins’ of the study are peeled back to reveal the rot within.

The very deep flaws and biases of the original articles include:

  • Cherry-picking information from cited sources to argue their case, while omitting information from the same sources that contradicted their case;
  • Including test and control subjects whose consequence was likely to maximise the likelihood of finding a positive association;
  • Demonstrating a poor understanding of suicide and its risk and protective factors and failing to control for most confounding effects in their econometric model ‘pudding’;
  • Overegging the “causative suicide contagion” interpretation when no correlation between assisted dying and general suicide rates was found; and
  • Failing to use direct, robust and readily-available evidence that showed their study couldn’t possibly have hoped to return scientifically valid “contagion” proof.

 
The USA’s National Violent Death Reporting System (NVDRS), of which Oregon is a founding member, shows that even if “assisted dying suicide contagion theory” were true, fewer than 2 of 855 Oregon “total suicides” in 2014 could have been attributed to “contagion” from DWDAs.

Further, both Oregon and Washington state rankings for suicide rates have improved, not deteriorated, since their DWDAs came into effect, while the suicide ranking for a relevant control state — Oklahoma — has deteriorated substantially over the same time.

Ultimately, through numerous and deep methodological flaws, the Jones, Paton and Kheriaty articles reveal a bias to promote “assisted dying suicide contagion theory” while ignoring the robust evidence from multiple lawful jurisdictions, including in their own ‘study,’ that contradict it.

The exposé, titled “The ten deadly sins of Jones, Paton and Kheriaty on ‘suicide contagion’,” will be published by DyingForChoice.com later in the week.

Note: the report is now published here.


Share This Post:
0
Professor Margaret Somerville makes an indefensible 'suicide contagion' claim.

Catholic ethicist Professor Margaret Somerville claims that every assisted suicide jurisdiction shows 'contagion' to the general suicide rate. The empirical evidence contradicts her claim.

Get the full report here.

Professor Margaret Somerville, currently Professor of Ethics in the School of Medicine at the Catholic University of Notre Dame Australia,[1] has enjoyed ongoing publication of her opinions, with few challenges published to date.

Back in 2007, Somerville, then a Professor of Ethics at McGill University in Montreal, Canada, appeared as an expert witness in an Iowa District Court case. The court comprehensively rejected her testimony, determining that she:

“…specifically eschews empirical research and methods of logical reasoning in favour of ‘moral intuition.’ She has no training in empirical research…”

Professor Somerville, I argue, has again fallen short on empirical research and logical reasoning. To illustrate, I will analyse her claim, published in an opinion piece in ABC Religion and Ethics that:

“…the general suicide rate has increased in every jurisdiction that has legalized assisted suicide.

While her claim may be her own personal opinion, she has presented it expressly stating that she is a Professor of Ethics at her current university of employment, lending the claim perceived authority.

This report demonstrates how her claim and her defence of it are contradicted by multiple sources of empirical government and other primary research evidence. It also demonstrates that she failed to engage appropriate scholarly standards that require the active search for, acquisition and analysis of all reasonably available relevant data in an attempt to answer a particular question.

In making her claim, Prof. Somerville:

  • Cites ‘supportive’ data from lawful jurisdictions while overlooking other data, sometimes even in the same data set, that are inconsistent with her claim;
  • Cites as supporting evidence an econometric modelling study that did not find a statistically-significant relationship between assisted dying law and the general (non-assisted) suicide rate;
  • Fails to consider data from all jurisdictions with assisted suicide laws while making a claim about them all — overlooking Switzerland, whose empirical data is clearly at odds with her claim;
  • Repeatedly cites non-academic anti-euthanasia lobbyist Mr Alex Schadenberg (who also cites her) as a source of evidence for her claim and who in turn quotes a television source and another lobbyist’s opinion to underpin his own beliefs about ‘suicide contagion’; and
  • Conflates voluntary euthanasia (physician-administration) with assisted suicide (patient self-administration) such that her argument, at least in the context of Belgium and the Netherlands, is substantially about the novel concept of ‘euthanasia contagion’ rather than the more familiar ‘suicide contagion’ expression she uses.
     

These findings are consistent with the Iowa court’s ruling that Prof. Somerville sometimes relies on ‘moral intuition’ rather than sound empirical research and logical reasoning.

My report also draws a number of connections between those advancing misinformation on assisted dying ‘suicide contagion,’ and Catholic identity. Catholic identity is not a reason to reject arguments, but it does help identify the source of a majority of ‘suicide contagion’ misinformation.

Finally, I argue that the appropriate course of action for Prof. Somerville is to retract her ‘suicide contagion in every jurisdiction’ claim.

 

Get the full report here.


[1]   Not to be confused with another Professor Margaret Somerville, who is Director of the Centre for Educational Research at Western Sydney University.

Share This Post:
0
Brad Mattes' blog containing misleading and evidentially false statements.

The latest misinformation employed by assisted dying opponents is to imply that Belgium’s general suicide rate is high as a consequence of its assisted dying law: i.e. to argue the discredited 'suicide contagion' line which has in the past been peddled about the USA state of Oregon. I have demonstrated that it was false in Oregon, and I equally demonstrate here that it's false in Belgium.

Mr Brad Mattes recently published emotional anti-assisted-dying nonsense in LifeSiteNews. (LifeSiteNews is a Canadian blog site that was established by the conservative Christian Campaign for Life Coalition and which has a primary principle of promoting “traditional Judeo-Christian principles”. Mr Mattes is radio host for Ohio’s Life Issues Institute, an anti-abortion lobby group established by conservative American John C. Willke who claimed that women’s bodies were resistant to pregnancy as a result of rape.)

Putting on the misinformation running shoes

In his blog Mr Mattes quickly establishes a practice of communicating misinformation by first claiming that assisted dying laws have “devastating effects” around the world including Japan and Albania—which don’t have such laws.

He then sprints onwards to the ‘650 babies euthanized in the Netherlands’ claim—which I have already comprehensively exposed as fake in the Journal of Assisted Dying—and then onto the bogus claim that the Netherlands has descended into a mire of ‘killing’ without the patient’s ‘consent,’ contrary to empirical evidence I've published that such actions occur around the world regardless of assisted dying laws, and which have decreased significantly in the Netherlands and Belgium since their assisted dying statutes came into effect in 2002.

Let’s add fries to that

Having served up a lot of fat and salt that might appeal to those on a fast-food anti-assisted-dying diet, he then offers the unqualified statement:

By the way, Belgium has the second-highest suicide rate (nonrelated to euthanasia) in Western Europe.” — Brad Mattes

The ‘informational’ consequence is unequivocal: by Mr Mattes failing to contextualise this ‘incidental factoid’ in any way, the reader is destined to deduce that it is Belgium’s euthanasia law that causes Belgium’s suicide rate to be the second-highest in Western Europe. In other words, Mr Mattes is another poorly-informed commentator using the 'suicide contagion' line.

But what are the facts?

Depending on the source and year of data, one can certainly argue that Belgium’s general suicide rate is the second-highest in Western Europe. Setting aside for now the serious question of why it is valid to exclude all of the world’s other countries from the comparison, WorldLifeExpectancy.com reports figures that were published in 2014 (Table 1).

Table 1

Country

Suicides*

Finland

15.11

Belgium

14.64

Iceland

14.06

France

12.84

Austria

11.87

Sweden

11.43

Ireland

11.06

Germany

9.59

Switzerland

9.56

Norway

9.28

Denmark

9.19

Luxembourg

9.14

Netherlands

8.54

Portugal

8.49

Turkey

7.92

UK

6.28

Malta

5.75

Spain

5.23

Italy

4.76

Greece

3.86

* Suicides per 100k population, age-adjusted

The table includes all the countries in the wider definition of “Western Europe”, bar four: no suicide statistics are published for Liechtenstein, Monaco, Andorra or San Marino.

As you can see, Belgium is indeed the second-highest. But this begs the question:

If the reason Belgium is the second-highest is due to its assisted dying law, how come Finland, which Mr Mattes doesn't mention by name and which has no such law, is higher?

It also begs the question:

If assisted dying law were the fundamental cause of a high general suicide rate, how come Switzerland (statute since 1942), Luxembourg (statute since 2009) and the Netherlands (practice since the early 1980s and statute since 2002) have rates that are much lower, all key facts that Mr Mattes also fails to mention?

These are of course indications that the factoid has been included as a cherry-picked morsel of 'proof' because it sounds so compelling as a throw-away sentence amongst the other (false) statements.

But wait, there’s more

We can go further than merely wondering about the country comparisons, by comparing Belgium’s general suicide rate before and after assisted dying law reform. A critical step in establishing causation is to first establish correlation. If there is no correlation, there can be no causation.

Published OECD data shows that in 2013 (the most recent available data), Belgium’s general suicide rate was 16.7 per 100,000 population. What was it before their 2002 law reform? Well in 2000 it was 20.5, in 1990 it was 19.2, … you get the idea.

Has Belgium’s general suicide rate soared (or even increased modestly) since their 2002 assisted dying law came into effect? No. It’s dropped. Indeed, the slight downward trend apparent before the statute came into effect in 2002 has accelerated downward since (Figure 1).

Belgium's suicide rate since 1987Figure 1: The Belgium general suicide rate before and after assisted dying law reform

Even the headline is misleading

Mr Mattes fails to point out in his blog that assisted dying statutes in Benelux countries (Belgium, the Netherlands and Luxembourg) never restricted access to only the 'terminally ill,' that is, those imminently dying. He also fails to point out that USA states whose laws do restrict assisted dying to the terminally ill—most notably Oregon and Washington—have not changed their statutes in this regard since they came into effect.

Therefore, his headling implying that lawful jurisdictions have broadened their laws from 'only the terminally ill' is also wrong.

Conclusion

Mr Mattes makes multiple false and misleading claims and it’s easy to see his opinion for what it is: an emotional dump that fails to engage with and indeed flies in the face of actual evidence. The latest 'suicide contagion' implication, that assisted dying law causes Belgium’s general suicide rate to be the “second-highest in Western Europe” is evidentially false.

I call on LifeSiteNews to withdraw Mr Mattes' article because it breaches their primary principle:

“1. Accuracy in content is given high priority. News and information tips from readers are encouraged and validated. Valid corrections are always welcome. Writing and research is of a professional calibre.” — LifeSiteNews

 

Summary of facts

Belgium's general suicide rate is one of the higher ones in Western Europe. However:

  1. At least one country without an assisted dying law has a higher suicide rate, inconsistent with 'suicide contagion' theory.
  2. Other Western Europe countries with assisted dying laws have suicide rates much lower than Belgium's, also at odds with 'suicide contagion' theory.
  3. But the clincher is that the suicide rate in Belgium has dropped, not risen, since their 2002 assisted dying law came into effect.

Share This Post:
0

Wesley Smith never seems to tire of spreading opinion. In another piece of published nonsense, he's proposed that USA's rise in national suicide rate is in significant part a consequence of assisted dying law in those few states that permit it (up until the most recent general suicide data that's Oregon, Washington state, Vermont and Montana). His claim flies in the face of actual evidence.

Wesley Smith is a Senior Fellow at the Discovery Institute. Remember that? It's the organisation that a USA Federal court ruled pursues "demonstrably religious, cultural, and legal missions," and which comprehensively lost a test case in which it tried to have 'intelligent design' (that's creationism with lipstick) taught as a 'scientific' alternative to evolution.

In a piece recently published by conservative blog NationalReview and reprinted in pro-life LifeSiteNews, Mr Smith has asserted that assisted suicide has a significant part to play in the rising USA national suicide rate. "Color me decidedly not surprised. We are becoming a pro-suicide culture," he asserts.

"I am convinced that the correlation [between assisted suicide advocacy and the general suicide rate] could also be at least a partial causation."

On the matter of rhetoric, notice how Mr Smith cleverly mixes certainty ('convinced') with uncertainty ('could') in order to hedge his literal argumentative bets while giving the impression of valid authority. Ultimately, however, being certain about uncertainty can only be... uncertain.

Mr Smith argues from a USA Centers for Disease Control (CDC) report that the national suicide rate increased more after 2006, "the very time when the assisted suicide movement has become the most vigorous and made its most dramatic advances [emphasis is Smith's]." He offers not one shred of quantitative empirical evidence to support his contention that the assisted dying movement's 'vigor' changed suddenly and substantially from 2006.

So, what does analysis of relevant and readily-available data show? I've reproduced the USA national general suicide rates obtained from the USA government's CDC online database, plus unemployment rates obtained from the USA government Bureau of Labor Statistics online database in Figure 1. (The suicide data does not include deaths under state Death With Dignity Acts because under these Acts such deaths are not suicides.)
 

USA national suicide and unemployment rates

Figure 1: USA national suicide and unemployment rates

 

The national annual suicide rates are shown in red, and the unemployment rates are shown in blue with linear regression lines for before and after 2006. It's easy to see that prior to 2006 the unemployment rate peaked at around 6%, while after 2006—and clearly in response to the global financial crisis (GFC)—the rate peaks much higher at nearly 10%.

Suicide is indeed a complex phenomenon, with a wide range of both risk factors (e.g. unemployment, mental illness, substance abuse) and protective factors (e.g. mental illness mitigation programs, unemployment benefits), and it would be glib to assert only one or a few factors. Nevertheless, Figure 1 demonstrates a clear correlation between trends in unemployment and the overall suicide rate.

Correlation is of course not causation: though I will in a future report show how extensively common this correlation is around the world and over time. Nevertheless, the data, had Mr Smith bothered looking for it, offers a vastly more rational and compelling explanation of the rise in suicide rate than does some hokey theory about how just 325 rational adults in two states (Oregon and Washington state Death With Dignity Act deaths in 2014) who were already dying and quietly and privately chose to go a little early in response to intolerable suffering, caused the suicide rate amongst 319 million inhabitants (2014) across a nation of fifty states, to rise by a "huge and alarming" amount.

Mr Smith backhandedly acknowledges that there are multiple causes of suicide. "There is no question that assisted suicide advocacy is not the only factor causing this alarming increase in suicides," he says, presumptively positioning his hypothetical reason as definitely one of them.

Mr Smith does refer to a recent journal article by David Jones and David Paton that purports to show a weak link between assisted dying and the total suicide rate (with the weak link appearing only if assisted deaths are counted as suicides). I have analysed that paper in detail and shall deal with it in due course. How it passed peer review (if it was indeed peer reviewed) remains a mystery. The study is of an unacceptably poor standard on a range of facets as I will demonstrate.

In conclusion, did the CDC report that Mr Smith cites suggest that 'assisted suicide contagion' was a possible cause of the increase? Nope.

It's really time that Mr Smith and colleagues gave the misinformation campaign a rest.


Share This Post:
0
Dr William Toffler (left) acknowledges no slippery slope cause-and-effect evidence, with Drs Bentz and Stevens

A new scholarly journal focused on end-of-life ethics, decision-making and practice has just been launched: the Journal of Assisted Dying. In the first article, claims by Oregon lobby group Physicians for Compassionate Care (PCC), including Doctors Bill (William) Toffler and Ken Stevens (and others) are assessed against empirical evidence and found to be completely wrong, or highly misleading as a result of selective use of data.

The new scholarly journal, the Journal of Assisted Dying, is dedicated to careful and holistic analysis of evidence in regard to the various forms of assisted dying that are lawful in a number of jurisdictions around the world... and to practices in jurisdictions where assisted dying remains illegal.

In the first article of a series on Oregon, I examine claims and speculations made by various doctors (and others who quote them), that Oregon has the second-highest suicide rate in the USA (or is always in the top 10), that Oregon's Death With Dignity Act has resulted a massive increase in the state's general suicide rate, and other astonishing statements.

Of course, the empirical evidence from the Oregon Health Authority and from the USA Centers for Disease Control and Prevention does not support these statements and interpretations, and I analyse and critique the evidence.

Some of the claims are just plain factually and hugely wrong. Others are the result of failing to read their sources more carefully, misunderstanding what the data actually represents. Still further claims are made on the basis of selectively-chosen statements from government reports, while omitting statements that are contrary to, or provide alternative and well-researched explanations for Oregon's recently rising general suicide rate.

Drs Toffler and Stevens have even published some of their claims and speculations in the British Medical Journal.1 It goes to show that even good journals sometimes publish bunkum:  their article was a letter to the editor rather than peer-reviewed research. Great care is required to sort real evidence from hype and opinion.

Ultimately, Dr Bill Toffler of PCC has acknowledged on video that there is no cause-and-effect evidence between Oregon's Death With Dignity Act and Oregon's suicide rate, an acknowledgement that went unchallenged by his two PCC colleagues present at the time, Dr Ken Stevens and Dr Chuck (Charles) Bentz. You can see Dr Toffler's statement here (at 10'50").

The Journal of Assisted Dying is an open-access journal, and you can read the full article here.

-----

1. Toffler, WL & Stevens, K 2015, 'Re: Assisted dying: law and practice around the world', BMJ, vol. 351, 19 Aug, p. h4481.


Share This Post:

Physician use of misinformation to speculate 'assisted dying suicide contagion' in Oregon


Author(s)

Neil Francis

Journal

Journal of Assisted Dying, vol. 1, no. 1, pp. 1–6.

Abstract

Background: Several physicians have speculated that Oregon’s general suicide rate is evidence of suicide contagion as a result of Oregon’s Death With Dignity Act (‘the Act’).
Methods: Search and analysis of physician and related online sources of Oregon suicide contagion speculation; retrieval and analysis of cited Centers for Disease Control and Prevention (CDC) and other publications relied upon; analysis of authoritative, public Government mortality data for Oregon and other USA states.
Results: Several physicians have speculated about Oregon suicide statistics in a manner that is not supported by the cited publications, or by public CDC mortality database data. The claims variously (a) misrepresent key data in the publications, (b) omit information in the publications that is at variance with suicide contagion speculation, and (c) overlook other significant information at variance with speculation. The physicians have previously acknowledged inability to prove perceived “slippery slope” effects of the Act. Other opponents of the Act have republished the physicians’ erroneous information.
Conclusions: Evidence advanced by several physicians to speculate that Oregon’s Death With Dignity Act causes suicide contagion in Oregon is variously false, misleading or highly selective—omitting key facts—and has arisen even though the physicians acknowledge they have no proof of ‘slippery slope’ effects.

Article keywords

suicide contagion, copycat suicide, Werther effect, slippery slope, misinformation, Oregon, Dr William Toffler, Dr Kenneth Stevens, Physicians for Compassionate Care

Full PDF

Download the full PDF: Download the full article (390Kb)

Citation

Francis, N 2016, 'Physician use of misinformation to speculate 'assisted dying suicide contagion' in Oregon', Journal of Assisted Dying, vol. 1, no. 1, pp. 1-6.

Download the citation in RIS format: RIS.gif


Share This Post:

Pages

Subscribe to RSS - Rhetoric: Suicide 'contagion'