Imagine if, when the H1N1 epidemic was at it peak, the government had passed the Swine Influenza Protection Act. Imagine that the new Act aimed to “provide for the care, treatment and control” of persons who are infected with H1N1 influenza. It allowed for the coercive treatment and isolation of infected people if “there are reasonable grounds for believing that care, treatment or control of the person is necessary” either “for the person’s own protection from serious harm,” or “for the protection of others from serious harm.” Under its provisions, any person suffering flu could be coercively assessed and, if it was thought likely that that person may become very unwell or may spread a virulent strain to others, he or she could be involuntarily detained and treated. Imagine that these provisions would apply despite there being no valid way of determining in advance who will become very unwell or who will spread serious disease. Finally, imagine that, having been promulgated, the Swine Influenza Protection Act was swiftly put to effect in hospitals and doctor’s surgeries around the nation.

Of course this sort of legislation would never have been passed. Existing legislationFootnote 1 gives the state power to compulsorily detain and treat people with specified infectious diseases, but this power is rarely used, has never been widely used in modern Australia, and would be considered a “last resort” used only as a part of a reasonable and proportionate response (Gostin 2003; Reynolds 2004, 221). The imagined heavy-handed response of the Swine Flu Act to what, at this point, seems little different to ordinary seasonal flu would surely have failed tests of proportionality and reasonableness (Cheng et al. 2009). To make matters worse, the lack of a valid marker for dangerousness would have meant that the coercive isolation would have been an essentially arbitrary intervention. Some non-dangerous flu sufferers would have been unnecessarily detained while other dangerous flu sufferers would have remained free to spread the disease.Footnote 2 This sort of legislation, for this level of public threat, could not have been introduced and implemented. Any government doing so would have feared a tremendous backlash, especially if the feared consequences of the epidemic had never been realised. While H1N1 was rapidly declared a quarantinable disease under the Quarantine Act, the wide-ranging powers of the various legislative instruments were never invoked.

There was, of course, no Swine Influenza Protection Act, but the phrasing of the provisions described in the imagined statute was taken directly from the Mental Health Act 2007 (NSW). Like all Australian mental health legislation, the New South Wales Act allows for the coercive detention and treatment of sufferers of mental illness (as defined in s 4), provided the harm criteria outlined above are applied (s 14) and no less restrictive alternative care is appropriate and available (s 12). All Australia’s mental heath acts (hereafter the MHAs) and those of many other jurisdictions have similar provisions.

Numerous authors, myself included, have argued that coercive treatment provisions of mental health acts should pivot on the psychiatrically ill person’s capacity, not upon the perceived likelihood of their coming to, or causing some harm. (See for example: Bartlett 2003; Buchanan 2002; Dawson and Szmukler 2006; Szmukler and Holloway 1998; Large et al. 2008; Richardson 2002; Dawson 2008; Wand and Chiarella 2006; Kämpf 2008; Ryan 2010). In recent months, at least two Australian jurisdictions have begun seriously investigating this sort of legislative change (Australian Capital Territory Department of Health 2009; Australian Capital Territory 2009; Victorian Government 2009), but to date no state or territory mental health act considers capacity as an important element in the justification for coercive detention (Ryan 2010).

In this paper, I provide a novel element to this debate by arguing that, since the use of the coercive treatment and detention provisions found in the Swine Flu Protection Act could not have been justified for H1N1 or for any similar medical illness, and since no special case can be made for their justification in mental illness, those provisions of the MHAs discriminate against those with mental illnesses, purely on the basis of their particular disability. I argue that this discrimination and inequality strikes at the heart of the human rights of the mentally ill. It unreasonably erodes the autonomy of some people with mental illness and is in direct contravention of numerous human rights instruments, including the Universal Declaration of Human Rights (United Nations General Assembly 1948, Art 1), the UN Principles for the Protection of Persons with Mental Illness and the Improvement of Mental Health Care (United Nations General Assembly 1991, Principle 1.4), the recently ratified UN Convention on the Rights of Persons with Disabilities (United Nations 2006, Arts 5, 12(2), 14) and Australia’s own National Mental Health Statement of Rights and Responsibilities (Mental Health Consumer Outcomes Task Force 1991, 15). I will argue therefore, that to this extent at least, the MHAs fail to meet the fundamental human rights objectives of the World Health Organisation, the United Nations, the Australian government, and even the objectives of the MHAs themselves (see for example Mental Health Act 2007 (NSW) s 3(d)). Finally, I will look briefly at how this may have come to pass and at what may be done to correct it.

How Australia’s Mental Health Acts Discriminate Against the Mentally Ill

Legislative provisions that permit the coercive detention and treatment of adult citizens are usually justified on one of two grounds. First, the provisions may be justified if wilful failure to comply with detention or treatment will place others at risk of harm—the harm justification (Mill 1859; United Nations General Assembly 1991, Principle 11.6; Gostin 2003; United Nations General Assembly 1948, Art 29(2)). Second, they may be justified if the person to be coerced lacks the capacity to accept admission or treatment, and there are grounds for believing that the person would have accepted admission or treatment had capacity been retained—the capacity justification (United Nations General Assembly 1991, Principle 11.6).

In the sections that follow I will argue that, to the extent that the MHAs rely on either of these justifications, they do so only by discriminating against people with mental illnesses.

Discrimination in the Harm Justification

The harm justification, as it applies to possible harm to others, is the rationale used to permit coercive isolation and treatment in contagious disease. Whether or not an infectious disease warrants provisions for coercion will depend upon weighing a series of factors associated with each disease. These factors include:

  1. 1.

    The reliability and validity of any method of determining who is infected and who among those infected might cause harm to others and therefore be the focus of intervention.

  2. 2.

    The severity of harm to others envisaged.

  3. 3.

    The likelihood of that harm occurring if coercive intervention is not delivered.

Governments are generally reluctant to authorise coercive detention and it is generally agreed that coercive treatment is not justified unless the method of determination is reasonably reliable, the harm envisaged severe and the likelihood of harm high.

This was arguably the case during the 2003 Severe Acute Respiratory Syndrome (SARS) outbreak. Initially SARS could only be diagnosed on clinical criteria that made the method of determination relatively unreliable; however this concern was offset by SARS’s infectivity (the chance of spread) and virulence (the consequences of spread). On average, a person infected with SARS would spread the disease to two to four others and of those who became infected around 10% of those under 60 and 50% of those over 60 died (Donnelly et al. 2003). These factors led most people to judge that it was reasonable and proportionate to order that people suffering suspected SARS could be coercively detained. SARS was added to the list of quarantinable diseases in Australia even though those powers were never used (Quarantine Regulations 2000 (Cth) cl 6(2)(i)).

These factors may be contrasted with those of swine flu. At the height of the epidemic, the diagnosis of H1N1 was fairly reliable as it made up the majority of flu cases then seen in Australia. It was also considerably more infective than SARS and on average each case spread the infection to around ten others (Cheng et al. 2009). However, H1N1 was much less virulent than SARS, with a case-mortality ratio estimated at its height at less than 0.2% (Cheng et al. 2009). It is undoubtedly this relatively low chance of any individual causing another’s death that led us to consider coercive detention and treatment unjustified.

While not concerned with contagious diseases, the same three-factor approach may be used to assess the reasonableness and proportionality of the mental health legislation provisions for the coercive detention and treatment of the mentally ill. While most people with a mental illness will not cause harm to others, some will. Using published epidemiological studies it is possible to make reasonable estimates of the chances of a person with a mental illness causing harm to another as a result of that illness. Figures vary from illness to illness and envisaged harm to envisaged harm, but to take one example that fits well with our considerations of SARS and influenza, the annual probability of a person with schizophrenia (or a similar psychosis) killing a person previously unknown to him or her has been estimated at around 1 in 140,000 (Nielssen et al. 2010). This figure could play a role analogous to the case-infectivity and case-fatality in a contagious disease.Footnote 3

Such a low prevalence demands a mechanism to try to determine who, among the population of people with schizophrenia, is more likely to cause harm in the future. There are numerous questionnaires and tools that aim to categorise patients by their likelihood of committing future harm, but to date none has been able to achieve a sensitivity or specificity of greater than 80% (Hare 1991; Harris et al. 1993; Webster et al. 1997; Monahan et al. 2005).

Knowing the base rate of stranger homicide in schizophrenia and assuming we were able to predict future homicide with a sensitivity and specificity of 80%,Footnote 4 it is possible to calculate the number of individuals with schizophrenia it would be necessary to detain in order to prevent one person with schizophrenia from committing stranger homicide. That number is 35,000. The state would have to detain 35,000 individuals with schizophrenia, who would never have committed stranger homicide, in order to prevent each preventable death. Moreover, the 80% sensitivity of our detection method means that one in five such deaths would occur anyway (Large et al. 2010).Footnote 5

It is possible to repeat this exercise for a range of lesser harms than homicide (Large et al. 2010) and for a range of other mental illnesses; however, the point should by now be clear.

Australia’s mental health legislation allows coercive treatment of people with mental illnesses at a threshold for harm to others that is far, far lower than the threshold used to justify the coercive handling of people with contagious disease. Unless we are prepared to lower the threshold for coercion in those with contagious diseases, or unless we can make some additional argument to justify the special treatment of the mentally ill, we must acknowledge that threshold for the harm justification of coercion, as set out in mental health legislation, discriminates against people with mental illnesses.

Note that the harm justification in mental health acts typically has another arm—harm to self. Though beyond the scope of this particular essay, it is worth noting in passing that the existence of this arm will do nothing to quell a suggestion that mental health acts are discriminatory. It is a well-established principle in common law that, as a general rule, a competent person may not be coercively treated for a medical illness even if failure to accept that treatment might result in the person’s death (Hunter and New England Area Health Service v A [2009] NSWSC 761; Re T (Adult: Refusal of Treatment) [1992] EWCA Civ 18; Airedale Hospital Trustees v Bland [1992] UKHL 5; Re PVM [2000] QGAAT 1).

Discrimination in the Capacity Justification

People with a mental illness may in some circumstances, because of the severity of their condition, be unable to weigh the pros and cons of adopting a treatment, which they might, without illness, have regarded as being in their own best interests. The UN Principles for the Protection of Persons with Mental Illness and for the Improvement of Mental Health Care suggest that the protection of individuals who lose decisional capacity in this way should be central to the purpose of mental health legislation (United Nations General Assembly 1991, Principle 6.1).

As stated above, however, Australian MHAs are almost silent on the question of a mentally ill person’s capacity, and no Australian jurisdiction demands that a person lose capacity before they be subject to the detention provisions of an MHA (Ryan 2010).

It might be argued that the determination of the presence of a “mental illness,” as defined in the MHAs, is supposed to stand in the stead of the determination of capacity. However, there is little to suggest that this was intention of any Australian parliament (Ryan 2010), and even if mental illness were intended as a proxy for incapacity, there is little to support this line of reasoning. Legal capacity is recognised to be task specific (Gibbons v Wright [1954] HCA 17, [7] (Dixon CJ, Kitto and Taylor JJ)), and there is ample evidence that the occurrence of a serious mental illness alone does not universally and completely erode an individual’s capacity to make decisions (Grisso and Appelbaum 1995; Owen et al. 2009; Bellhouse et al. 2003a, b; Milne et al. 2009).

Without any attention to the matter of capacity, the MHAs lose the second possible justification for their coercive provisions. In all Australian jurisdictions guardianship legislation provides considerable protection for those who lose their capacity in circumstances not related to mental illness, but the MHAs offer no similar protection. This differential treatment discriminates against those with a mental illness.

Sanism as a Source of Discrimination

Assuming the MHAs are discriminatory in the way that I have suggested, other questions present themselves. Why, in the case of NSW for example, wasn’t this issue addressed at all in the consultative documents released as part of the Act’s recent revision (New South Wales Health 2004a, b, 2006)? Why didn’t this issue arise in the parliamentary debate that surrounded the relevant Bill (Kelly 2007)? Why is it only very recently that Australian legislators have even begun to consider using capacity as a fulcrum for coercive attention in the MHAs?

I suspect that the answers to these questions lie in a profound prejudice against the mentally ill that Perlin has termed “sanism.” Perlin defines sanism as:

An irrational prejudice of the same quality and character … [as] racism, sexism, homophobia, and ethnic bigotry, … that is largely invisible and largely socially acceptable, and that is based predominantly upon stereotype, myth, superstition, and deindividualization, and is sustained and perpetuated by our use of alleged “ordinary common sense” and heuristic reasoning in an unconscious response to events both in everyday life and in the legal process (Perlin 2008, 1).

As conceptualised by Perlin, sanism could lead to this sort of legislation and to its general acceptance in three ways. First, it is an intrinsic part of the sanist schema that people with mental illnesses are “invariably more dangerous than non-mentally ill persons, and such dangerousness is easily and accurately identified by experts” (Perlin 1992, 394). Second, another part of the sanist schema sees mentally ill individuals as “presumptively incompetent to participate in ‘normal’ activities, to make autonomous decisions about their lives (especially in areas involving medical care)” (Perlin 1992, 394). Third, and most pernicious of all, sanism, like other “-isms”, leads us to see the mentally ill as “the other,” makes us less able to empathise with their circumstances, and more likely to completely overlook even systematised discrimination (Perlin 1992, 389). Perlin believes sanism is endemic in legislators, academics and in legal and psychiatric professionals, and that inherent prejudice is a root cause of discrimination in not only statutory law, but the whole working of the legal system (Perlin 1999).

I have suggested that the MHAs, at their core, ignore the basic human rights of those whose rights they are supposed to defend. If this is true, and if it has been allowed to occur so recently in New South Wales, for example, with barely a word of protest from legislators, academics, or legal and psychiatric professionals, it is hard to do other than agree with Perlin’s formulation of the nature and scale of the sanist stain.

Repairing the MHAs—Replacing Harm with Capacity

In the real world it is not possible to predict, in any useful way, whether a particular person with a mental illness is likely to harm another person, at any time beyond the immediate future. The coercive provisions of the MHAs cannot be reasonable and proportionate mechanisms for the protection of society, any more that the Swine Influenza Protection Act would have been.

In contrast, it is possible to judge with a reasonable degree of reliability, validity and utility whether a person has the capacity to make decisions such as whether they might benefit from treatment, or a brief period in a secure environment (Grisso and Appelbaum 1995). People with mental illness often lose their capacity as a result of the effects of that illness, but the mere presence of mental illness does not imply a loss of capacity (Melamed et al. 1997).

People should have access to legal mechanisms that will activate when their capacity is impaired and that will enable decisions to be made on their behalf on the grounds that it is likely that they would have opted for a particular course had their capacity not been lost (Ryan 2010). When capacity is lost for reasons other than mental illness, Australia’s guardianship legislation provides exactly that mechanism, but when people lose capacity because of mental illness, there are no similar provisions.

There may be policy and logistical reasons for housing these legal provisions relevant to mental illness within a free standing mental health act, but the core nature of those provisions should not differ substantially from those that would apply to any other incapacitated person.

Australia’s mental health acts should be re-written to remove their discriminatory coercive provisions and to apply provisions based on lost capacity.