Parity Arguments for ‘Physician Aid-in-Dying’ (PAD) for Psychiatric Disorders: Their Structure and Limits

This article refers to:

Physician Aid-in-Dying and Suicide Prevention in Psychiatry: A Moral Crisis?

Kious and Battin (K&B) argue that psychiatric PAD (PPAD) should be legal in the US, based on a ‘parity’ argument (Kious and Battin 2019). This is the most popular approach to argue for PPAD (Cholbi 2013; Dembo 2010; Dembo et al. 2018; Hirsch 2016; Parker 2013; Provencher-Renaud et al. 2018; Rooney et al. 2018; Schuklenk and van de Vathorst 2015; Tanner 2018; Varelius 2016). What K&B add is that since, in their view, the parity argument is valid, there is a dilemma because PPAD conflicts with the practice of involuntary commitment in psychiatry. In this editorial, we sketch out the structure of the argument from parity, pointing out its challenges and limits. This will show that the dilemma K&B pose is actually a general problem about PPAD, not a dilemma specific to PPAD and involuntary commitment.

K&B’s parity argument has the following form: If PAD for terminal physical illness is justified on the basis of suffering, then fairness/equality/parity/non-discrimination dictates that PPAD be permitted. The argument is forceful but formal. It yields the conclusion that PPAD should be permitted only for those who also believe in four fairly controversial premises, each of which would take considerable work to defend: (P1) PAD of some form should be permitted; (P2) it must be based on suffering; (P3) situations of suffering in PAD and in PPAD are so similar such that not permitting PPAD would be arbitrary; (P4) permitting PPAD would not have negative policy and practice implications serious enough to outweigh the intended merits of PPAD. Giving a comprehensive analysis of each claim is beyond the scope of this editorial. Instead, we point out the many points of dispute that still need to be resolved for the parity argument to yield K&B’s conclusions.

THE CONDITIONAL NATURE OF PARITY ARGUMENTS

As is the case for most articles arguing in favor of PPAD based on the parity argument, the underlying assumption is that PAD for terminal physical illness is legally permitted. Thus it is a conditional argument. As a formal argument, the argument has 3 potential conclusions. One, PPAD is not permissible because PAD and PPAD are different; the parity argument does not carry. Two, PPAD is permissible because it is similar enough to PAD; parity argument carries. Three, PPAD turns out not to be permissible, but the parity argument still applies, and PAD for terminal illness is impermissible—a possibility that Foster rightly mentions but is rarely explored in parity arguments for PPAD (Foster 2019)¹.

THE MEANINGS OF ‘SUFFERING’ AND THEIR USES IN THE PARITY ARGUMENT

Most parity arguments for PPAD assert some version of ‘mental suffering is as bad as or worse than physical suffering’ (Cholbi 2013; Dembo 2010; Dembo et al. 2018; Hirsch 2016; Parker 2013; Provencher-Renaud et al. 2018; Sagan 2015; Steinbock 2017; Tanner 2018; Varelius 2016). As psychiatrists (MN and SK) we do not disagree with the statement.² What we question is the accompanying assertion that to argue against PPAD amounts to not taking suffering seriously (Cholbi 2013; Kious and Battin 2019; Provencher-Renaud et al. 2018; Schuklenk and van de Vathorst 2015; Tanner 2018). Such an assertion has rhetorical force but cannot support the parity argument since it makes sense only if the parity argument is already seen as valid. After all, it is possible to take mental suffering extremely seriously and non-callously, with skill, empathy, and resources—without permitting PPAD. ‘Suffering X is as bad as Y’ is a philosopher’s shorthand whose meaning and implications depend on how we understand the nature and source of the suffering (De Vries 2019).

The premise that all PAD regimes must be justified by (alleviating) suffering and the premise that the situations of suffering in terminal PAD and in PPAD are ‘similar enough’ are inextricably linked: how one defines and understands suffering as the basis for PAD will determine whether and how it can be used in a parity argument. Thus we examine P2 and P3 together.

Is Suffering a Necessary Basis for PAD?

The authors argue that (alleviating) suffering is the moral basis for PAD in all jurisdictions allowing the practice. While suffering is explicitly mentioned among the PAD eligibility requirements of European countries and Canada, it is not generally true of PAD laws in US states. To account for the anomaly of the US laws, K&B speculate (without explaining why such a speculation amounts to an argument) that most cases of PAD even in the US are motivated by suffering and argue that the terminal illness requirement is a mere ‘safeguard’ rather than a partial justification for PAD.

In fact, rather than ‘suffering as the basis and terminality as a safeguard,’ the Oregon style laws can just as well be interpreted as based on autonomy with terminality as a co-justification. A doctor in Oregon is permitted to provide PAD to a patient (who otherwise meets criteria for PAD) whose reason for requesting PAD is a desire to control his exit from life. Would an advocate of PAD in Oregon think providing PAD to this man violates some principle behind the law? Without having to refer to suffering,³ one can coherently say: the justification for PAD is that at the end of one’s life, how one dies should be determined by the person’s ‘values and beliefs.’ Or, as some have put it, “we want that last act to reflect our own convictions” (Dworkin et al. 1997). This certainly sounds like addressing terminality itself—the fact that one’s “very existence as a singular entity is ending”(Bartlett and Finder 2019)—is the very point of PAD (Campbell 2019; De Vries 2019; Foster 2019; Ho and Norman 2019; Lemmens 2019).

Indeed, some suffering-based regimes might more accurately be re-framed as ‘autonomy-based with suffering as a safeguard’ regimes. If a jurisdiction leaves the determination of suffering as the justification for PAD entirely up to the patient, as in Canada, the ultimate justification for PAD seems to be autonomy and such a law, as den Hartogh observes, may only pay “lip-service to its commitment to compassion as the basic justifying ground for PAD” (den Hartogh 2019). In such cases, suffering might better be seen as a safeguard, not a justification.

Is Suffering a Sufficient Basis for PAD?

Even if suffering is taken as the basis for PAD, no suffering-based regime treats alleviation of suffering as a sufficient basis: they all add additional eligibility criteria. These additional restrictions seem just as arbitrary or just as necessary, depending on one’s point of view, as the terminal illness requirement; or at least it would require an argument to support either view.

First, PAD based on purely existential suffering without a medical basis (e.g., tired of living or completed life) does not qualify. What is the suffering-based moral principle that excludes all non-medically based suffering as a basis for PAD?(Gaignard and Hurst 2019; Steinbock 2017) It is arbitrary and inconsistent to expand the meaning of suffering to include one’s inability to exercise control over one’s death (as a means of arguing for suffering as basis for PAD) but then to restrict its meaning by insisting that it must be medically based.

Second, in all suffering-based PAD regimes, intractability/irremediability is a requirement. It is a separate, additional restriction (van Veen and van Delden 2019). But why is irremediability necessary? If the experience of suffering is the same in person A and person B but their prognosis is different (an irremediable condition versus a slow to recover but not irremediable condition), why should only person A qualify?

Third, all PAD regimes permit only voluntary PAD. But a welfare based justification for PAD, such as suffering (in contrast to an autonomy based regime), does not rule out non-voluntary PAD (Jones 2011; Keown 2002; Varelius 2016). One may question why only voluntary PAD should be permitted when two people have identical suffering.

Can the Parity Argument Work When Suffering is Neither Necessary Nor Sufficient Basis for PAD?

First, if suffering is not a necessary basis for PAD, then a suffering-based parity argument is irrelevant. Second, the various extra restrictions on PAD that we observe even in suffering-based regimes clearly violate parity. One might therefore argue that terminality violates parity but is just as necessary; at any rate, if one accepts the other parity-violating restrictions, parity per se would not be an argument against terminality.

But if one maintains the primacy of parity, the differential treatments required by the above restrictions are illegitimate. One would then be committed to expanding PAD to include not only suffering from non-terminal disorders, but also non-medical suffering, non-irremediable suffering, and non-voluntary PAD. Or one might instead, seeing that parity requires such an expansion, reject the original premise of permissibility of PAD.

WOULD PERMITTING PPAD NOT HAVE CONSEQUENCES WORRISOME ENOUGH TO COUNTER ITS INTENDED MERITS?

Up to this point, we have largely focused on normative, conceptual problems for the parity argument. But (just for the sake of argument) even if we imagine that the parity argument for PPAD were conceptually sounder than it is, the question still remains: how does one justify going from an idealized conceptual argument to a public policy? K&B, like other proponents of the parity argument, write as though showing the philosophical plausibility of PPAD in a single ideal case provides sufficient basis for a public policy (Cholbi 2013; Dembo et al. 2018; Kious and Battin 2019; Provencher-Renaud et al. 2018; Rooney et al. 2018; Schuklenk and van de Vathorst 2015; Steinbock 2017; Tanner 2018). But as De Vries reminds us, such arguments are idealized, “with no appreciation of the way such judgments are shaped by the context in which they are made” (De Vries 2019). Or, as Foster describes it, “considering only that patient is a philosophical indulgence not available to legislators” (Foster 2019). To be fair, even if some proponents have neglected the real world context and consequences of normative concepts, other commentators have pointed this out. While some argue that there is no “principled basis” for excluding psychiatric patients (Schuklenk and van de Vathorst 2015), others state that “there is a gap between acknowledging that there are cases in which [PPAD] is justified and creating a law or policy that reliably identifies such cases” (den Hartogh 2015; Steinbock 2017).

There is a wide range of policy challenges related to allowing and implementing PPAD. The most frequently cited challenge is that there is a greater potential for error in evaluating patients with nonterminal, psychiatric disorders: even if ideal cases exist, there is the question of reliably identifying those cases. As den Hartogh points out, policymakers could reasonably think that no “institutional arrangement will guarantee us to sufficient extent that the exceptional cases are properly identified,” hence we should err on the side of safety (den Hartogh 2015; Foster 2019; Lemmens 2019; Miller and Appelbaum 2018; Steinbock 2017; Vandenberghe 2018; Zuradzki and Nowak 2019). The specific difficulties relate to reliably and objectively assessing irremediability and decision-making capacity in persons with a PPAD request (Broome and de Cates 2015; den Hartogh 2015; Lemmens 2019; Miller and Appelbaum 2018; Steinbock 2017; van Veen and van Delden 2019; Zuradzki and Nowak 2019). Indeed, what we mean by irremediability and ‘treatment-refractory’ in psychiatry is ill-defined (Blikshavn et al. 2017; Jansen et al. 2019; Kim and Lemmens 2016; Kissane and Kelly 2000; Miller 2015; Schoevers et al. 1998; Simpson 2018; Steinbock 2017) and predictions about prognosis can be unreliable when causation is poorly understood and diagnosis mostly descriptive (Blikshavn et al. 2017; den Hartogh 2015; Kelly 2017; Kelly and McLoughlin 2002; Kendler 2019; Naudts et al. 2006; Pearce 2017; Schoevers et al. 1998; Simpson 2018; van Os et al. 2019). Furthermore, there is a challenge of defining what counts as an informed request for PPAD and how each of the criteria for capacity should be interpreted (Kim 2016; Owen 2016).

Finally, there are broader policy concerns about allowing PPAD and its potential societal consequences: the impact on the patient-physician relationship (Blikshavn et al. 2017; Calkins and Swetz 2019; Olié and Courtet 2016; Schoevers et al. 1998) and on the profession of psychiatry (Calkins and Swetz 2019; Jansen et al. 2019; Kissane and Kelly 2000; Miller 2015; Simpson 2018), the role of social determinants contributing to mental health (Ho and Norman 2019; Pearce 2017; Simpson 2018), and the expressivist consequences such as the implicit message that may be conveyed to vulnerable populations (Appelbaum 2018; De Vries 2019; Foster 2019; Kim 2019; Kim and Lemmens 2016; Le Glaz et al. 2019; Simpson 2018).

CONCLUSION

K&B’s parity argument is similar to many parity arguments in that it advocates legal policy based on an idealized conceptual argument. What is different about their paper is that, after concluding that PPAD should be legal, they then go on to consider some serious policy and practice considerations, namely, the problem of involuntary commitment. Indeed, K&B’s discussion of the difficulties of using a capacity standard or a suffering metric articulate important policy and practice problems. But that discussion makes perfect sense without juxtaposing it to the issue of involuntary commitment at all. The authors do not see this because, by this point in their paper, they have already accepted the parity argument as valid. They locate the dilemma (‘a moral crisis’) in the wrong place. Their problem is about PPAD itself.

The parity argument is ultimately a test of the validity of the content that we put in it. If it yields a conclusion that conflicts with our moral and policy considerations, then we should at least revisit the argument’s starting point. What we should not do is to be so committed to an outcome of the argument that we lose sight of its double-edged nature.

ACKNOWLEDGMENTS

The authors thank Dominic Mangino for his helpful comments on the paper.

DISCLOSURE STATEMENT

The views expressed by the authors are their own and do not represent the views or policies of the NIH, DHHS, or the US government.

FUNDING

Funded in part by the Intramural Research Program of the National Institutes of Health, USA (S.K. and M.N.). ▪

Notes

1. For example, K&B state, under the section on ‘Severity of Suffering’: “We would also invite interlocutors to imagine someone with a relatively painless, terminal physical illness who chooses PAD to end or prevent the emotional or existential suffering that her illness brings. If that is justifiable (and it seems to be permitted by PAD statues in the U.S.), PAD in mental illness should sometimes be justifiable, too.” If however it turns out there are good reasons to not permit PAD for mental illness, then it would seem the parity argument should persuade K&B that permitting PAD for terminal illness may be a mistake. (We here ignore arguments for PPAD not using the parity argument which are hard to find. Does the relative rarity of such arguments indicate the difficulty of constructing them?)

2. We have not found a commentator who argues PPAD should not be allowed on the basis that mental suffering is not serious enough.

3. K&B may object that in our example the person who does not get to control how he dies will experience suffering because of this and that is the basis for allowing PAD. But this is like saying that coercing another person is wrong only because of the suffering it would cause. The determining value is autonomy itself. ‘Suffering’ by itself is too malleable a label. It can too easily apply to cases we would intuitively resist (Cowley 2013; Lemmens 2019).