March 18, 2019
By Peter Bartlett (Nottinghamshire Healthcare NHS Trust Professor of Mental Health Law, Institute of Mental Health and School of Law, University of Nottingham)
On its face, this case considers the duty of the State to protect the lives of voluntary (or informal) psychiatric patients under Article 2 of the ECHR (right to life). Below the surface, the case raises a number of broader questions about the scope of the positive obligations under Article 2; how they relate to the autonomy-related rights in Articles 3 (prohibition of torture, inhuman or degrading treatment), 5 (right to liberty) and 8 (right to privacy and family life); the fact-finding exercises and evidential approach of the ECtHR; and the relationship between the ECHR jurisprudence and other international law, most notably the United Nations Convention on the Rights of Persons with Disabilities (CRPD).
The deceased, AJ, was first admitted to the Sobral Cide Psychiatric Hospital (HSC) at the age of twenty, in 1984. He was admitted a total of eight times between then and his death in 2000, generally for schizophrenia or major depression and often following abuse of alcohol or prescription drugs, to which he was addicted. All of these admissions were ‘voluntary’. On 27 April 2000, he left the hospital grounds without leave. Before his absence had noted and concerns raised by staff, he had died by jumping in front of a train.
The case had two elements. One was the adequacy of domestic processes: AJ’s family had commenced litigation regarding his death in March 2003, and did not get a judgment until April 2011 – a delay of more than eight years. The State allowed that this was an excessive delay; the only question was whether the violation sounded in Article 2 or, as the State claimed, only in Article 6 (right to fair trial). The Grand Chamber held unanimously for the former. While noteworthy, it is not controversial and will not be discussed further.
Of greater interest are the substantive issues surrounding Article 2. In the first instance hearing at the ECtHR, the Fourth Section held unanimously that this Article had been breached; in the Grand Chamber, it was held (Pinto de Albuquerque and Harutyunyan dissenting) that there had been no such breach.
The Majority Decision
The majority decision presents itself as simply applying the established doctrinal law related to Article 2. To that end, readers would be well to refer to this case for a clear and up-to-date statement of the positive obligations of States to protect life, both through provision of overall regulatory measures (the ‘regulatory duty’) and, for people identified as being in real and immediate risk of harm, through more direct measures to prevent that harm (the ‘operational duty’), particularly in the context of psychiatric care.
The first of these is a general duty to provide an appropriate regulatory structure. States are obliged for example to regulate hospitals to protect patients’ lives (ensuring an adequate quality of staff, for example), and to provide an effective judicial investigation into deaths. In the present context, this meant ensuring an appropriate array of legal frameworks to ensure that the lives of voluntary psychiatric patients could be protected. The Portuguese Mental Health Act allowed for the civil detention of people in appropriate cases, and a general directive required care in the least restrictive environment possible. The HSC had adopted a system where voluntary patients could be placed under a regular regime, where they were allowed to circulate freely including into the hospital grounds and where attendance was checked seven times per day, and a restricted regime, where patients were required to remain in pyjamas and not permitted to leave the ward. Formal criteria were not established for the allocation between these regimes. There was similarly at the relevant time no formal policy on the use of physical or chemical restraint in the hospital, a matter on which the CPT had been critical.
The majority held that this was sufficient to meet the regulatory duty. Following Hiller v. Austria (22 November 2016), it noted that the trend in international law and practice had been to encourage open treatment regimes. While further regulation (such as a formal restraint policy) might have been desirable, it was not obvious that it would have affected the outcome in AJ’s case. In this the majority drew a distinction between the positive obligation under Article 2, and the regulatory requirements under articles 3, 5, and 8: the latter were to prevent an arbitrary imposition of power; the former were to provide the tools necessary to protect individuals from harm at the hands of third parties or of themselves. The majority held that the existing legal framework provided sufficient mechanisms to this end.
The ‘operational duty’ requires States, when specific individuals are known to be at risk of harm from themselves or others, to take preventive action to minimise or avert the risk. In the context of detained psychiatric patients, factors giving rise to this duty included a history of mental health problems, the gravity of the mental condition, previous attempts at suicide or self-harm, suicidal thoughts or threats, and signs of physical or mental distress. The application and scope of the duty was fact-specific, and the Court acknowledged both proportionality and practicality requirements in its interpretation, in addition acknowledging the risk that overzealous application of the duty could result in issues arising under Articles 3, 5 or 8.
Prior to this case, the operational duty had been applied only to detained patients. A key legal point in the present case was therefore whether it also applied to voluntary patients. The majority held that it did, but with some qualifying language: the nature of the measures required might differ according to the legal status of the individual, and that a stricter standard of scrutiny might be applied by the Court in cases of detained patients [para 124].
In this case, the majority held that the facts did not give rise to a violation of Article 2 on this branch either. While it was clear that AJ had a long history of mental health problems, the majority did not interfere with the findings of the Portuguese court that there was insufficient evidence of suicidality to require additional protective measures on the part of HSC.
Judge Pinto de Albuquerque, the Portuguese judge on the Court, wrote a blistering dissent in which Judge Harutyunyan joined. In his view, the majority opinion was ‘written for a country other than Portugal at the time of the events’ and ‘the result of a creative exercise of judicial adjudication for an imagined country.’ [para 1] The decision was ‘an ideologically charged minimalist approach to the State’s obligations in the sphere of health care law’, the effect of which was ‘downgrading the level of Convention protection to an inadmissible level of State inertia.’ [para 2] This view is supported by a number of arguments worthy of general note.
The legal regulation of voluntary patients was woefully inadequate at the relevant time. The use of ‘restrictive’ and ‘regular’ regimes had no legal basis; the CPT had for years been critical of the lack of regulation of restraint, and remained critical in its most recent (2012) visit. Restraint was entirely unregulated until 2007, well after AJ’s death. The first national suicide prevention strategy had only been introduced in 2012. ‘[I]n other words, in 2000 Portugal was in the pre-historic stages of suicide prevention of psychiatric inpatients. There was no legislation or regulation on what types of regimes could be applied, under what circumstances, by whom and until when. … To put it simply, in 2000 the HSC was in a legal black hole.’ 
Regarding the operational duty, it is not defensible that the Court would adopt a stricter level of scrutiny for detained patients: ‘this of course means that the Court will take a hands-off approach regarding the security of voluntary psychiatric inpatients at risk’. [para 20] No justification is offered by the majority for this position, in the face of a unanimous Chamber decision to the contrary. The argument that policy has moved to a less restrictive framework of psychiatric care was only half the story. States are also subject to increasing obligations regarding suicide prevention. The opinion of the majority was better understood as reflecting an increasing commodification of healthcare law, reduced state responsibility for healthcare, and the protection of healthcare professionals in ‘an untouchable legal bubble’. [para 21]
The finding that the suicide was unforeseeable was inconsistent with previous factual precedents: see Renolde v France and De Donder and De Clippel v Belgium, as well as flying in the face of the facts. In any event, even absent known risk, there are precautions involving basic surveillance that should be taken in all cases; and basic risk indicators were not identified by staff. [paras 28-38] No one seems to have noticed, for example, that AJ had not taken his medication for more than a day.
The majority decision had further entirely failed to take account of the debates surrounding the CPRD.
The Human Rights of a Fictional Country?
The dissent criticises the majority opinion for its characterisation of the facts – for findings of fact by the domestic court being taken at face value without serious interrogation, and for sympathetically constructing the behaviour of the health providers as rational and systematic in the absence of any evidence in support. This is not the place to explore the details of that. Suffice it here to say that as the dissent is written by the Portuguese judge about his own country, his views warrant serious regard.
The broader question is how the ECtHR engages in fact-finding exercises. It is no secret that the standards in many European institutions for people with mental health problems and learning disabilities would make human rights advocates cringe. The minority is right to question the reliability of domestic processes as far as fact-finding is concerned: too frequently, domestic courts and inspectorates do not highlight problematic situations. There is no easy answer to this. The ECtHR does not have the resources to engage in independent investigations into the facts, and in any event is only likely to be made aware of the case years after the relevant events occurred. At the same time, absent a robust factual basis, it is fair to wonder precisely what is being adjudicated. The dissenting judgment draws this issue to the fore in this case, but we might equally ask a similar question to some degree or other on many or most ECtHR judgments.
The Problem of ‘Voluntary’ Status
On its face, this case was about the duty of the State to ‘voluntary’ patients. ‘Voluntary’ (or, perhaps better, ‘informal’) can cover a wide array of circumstances. Sometimes it means that the patient has realised his or her problem, sought out help, and requested psychiatric admission, but this is often not the case. It may also mean admission by a legal guardian (with or without the patient’s assent or acquiescence). The patient may alternatively feel that they have no practical alternative to ‘agreeing’ to admission, as when the family with whom the patient lives tells a patient he or she will be out on the street if he or she does not agree to the admission, or where professionals tell the patient that he or she will be legally detained unless he or she agrees to ‘voluntary’ admission. Still further alternatively, patients may not realise that they have voluntary status, as when the authority to detain them has lapsed but they have not been told of this.
The degree of state involvement will therefore vary according to the facts: ‘voluntary’ admission to a state facility resulting from threatened detention by a person with power to action that detention seems to involve the state much more than a person’s own considered decision to seek help from a private hospital.
Should that matter? On the one hand, Article 2 does impose an obligation on States to make reasonable steps to protect all vulnerable citizens who are at real risk. On the other, there is a long tradition of the ECHR having particular concerns to ensure the appropriateness of State actions when the State does directly and coercively intervene in peoples’ lives. The majority decision reflects this: everyone has Article 2 rights, but particular scrutiny may be applied in cases of formal detention [para 124]. While the logic of this is clear, given the Court’s traditions, the dissenting judgment is right to ask whether this will result in a failure to protect non-detained patients – ‘rights-lite’, if you will. That is an appropriate challenge to the majority; we know better as we see case law develop on how protections are actually applied. If credence is given to the dissent’s criticisms of the factual aspects of the present case, things have perhaps not started well.
The dynamics of regulation
The majority distinguishes between the human rights issues of regulation under Article 2, where it is characterised as providing a tool box for protecting people who are vulnerable, and under Articles 3, 5 and 8, where it is about restricting human rights and where the Court’s role is to ensure that such restrictions are justified, proportionate and non-arbitrary. Intellectually, one can see the logic of that, but the practicality is not so simple. Protections against arbitrariness require the provision of substantive criteria, and particularly when the failure to take action may give rise to complaint under Article 2, those criteria are likely to be read instructively rather than proscriptively by professionals: enhancing Article 2 protections will mean that if Article 5 criteria are met, compulsion is likely to be imposed. For this reason, the effective requirement of the majority that coercive mental health detention and care must be available to meet the regulatory requirement of Article 2 is somewhat startling, and will certainly constitute fighting words for those CRPD advocates who call for the abolition of such coercive structures.
Predicting suicide (or violence – the same issue arises) of people with mental health problems is extremely difficult: suicide is an extraordinarily rare event, and risk assessment systems are of highly doubtful validity and accuracy. Already, we detain large numbers of people who were not going to go on to be dangerous or to commit suicide, on the basis of ‘risk’: this is not, and cannot be, an exact science. The question as regards the scope of Article 2 is therefore how far, if at all, the scope of compulsion should be expanded to include yet more people who will not commit suicide, to ensure that we control an additional few that do. Political noise needs to be acknowledged here: people who do end up committing suicide are a great deal more visible in public discourse and in the media than people whose rights may be removed and who were not going to go on to commit suicide. That is not, of course, to say that suicide prevention programmes should not be made available on a supportive basis; it is to say that what is under discussion in the present case is compulsion, and that must be approached with considerable care.
The majority is aware of that difficulty, and acknowledges the moves in international human rights law and medical policy towards care in the least restrictive environment possible [para 124]. The dissenting judgment dismisses this as a ‘recent trend’ [para 21]. It is actually a challenge to the overuse of institutionalisation and coercion, flowing from ongoing efforts of human rights activists, patients and their advocates over more than half a century. As the dissenting judges must be aware, such over-institutionalisation and coercion for convenience continue to exist in many Council of Europe countries, and it is surprising and at best unhelpful for the dissenting judgment to use language which could be taken to support such practices. The majority is right to be worried on this point.
The dissenting judgment is more interesting when it challenges the conceptualisation of the problem. Whether one agrees that the majority decision ‘reflects a hidden social-welfare disengagement policy, which aims at the maximum commodification of health-care services and above all at the protection of health professionals in an untouchable legal bubble, shirking State responsibilities for health-system and hospital-related death or serious injury under the Convention and consequently limiting the Court’s jurisdiction in this area’, [para 21] the question of how the ECHR should deal with health care decisions is potentially interesting.
The dissenting judgment does not speak directly in terms of compulsion, although on many of the specific interventions it seems to propose for specific people believed to be at risk, it would seem to be implied. Rather, it finds that AJ needed ‘a specific personal care regime accompanied by measures of restraint that would provide a sufficient guarantee of his medical as well as his safety needs.’ [para 14] States are expected to design and implement suicide reduction policies. [para 21] The dissent explicitly questions whether the ECHR jurisprudence is consistent with the CRPD. That approach raises tantalizing possibilities, as however one understands the CRPD (and interpretations are diverse), the provision of as full as possible an array of supportive services is certainly a requirement. The question could cease to be ‘did the individual need to be detained and/or restrained’, and instead become ‘was the individual getting what he or she needed?’
Meaningful engagement with that could open a much larger set of questions: if it becomes appropriate to explore supportive services within hospital under Article 2, why not more broadly? In cases such as Kolanis v UK (21 September 2005) and Johnson v UK (24 October 1997), release of detained patients was not possible because the appropriate services were not available in the community. Given the right to live in the community contained in Article 19 of the CRPD, does the ECtHR need to re-think those judgments? Indeed, should the ECtHR be imposing requirements on States to provide supportive services in the community to prevent suicide under Article 2, or to prevent the need for institutionalisation under Article 5? In much of Europe, particularly for people with learning difficulties, large-scale institutional living (often on a ‘voluntary’ basis) is a fact of life. To characterise the human rights problem as whether it is a ‘nice’ institution, rather than whether they should be there at all, seems to be colossally missing the point. Why should the ‘untouchable legal bubble’ be limited to Article 2?
That would involve a significant re-thinking of ECHR jurisprudence, and it is at best unclear whether there is the will on the Court to do this. And to be fair to the Court, overthrowing forty years of jurisprudence since Winterwerp is not a project to be undertaken lightly. At the same time, however one reads the CRPD, it does pose profound challenges to what has come before. At some point, somehow, the ECtHR needs to take meaningful account of those challenges.
 See, eg., Matthew Large, Christopher Ryan and Olav Nielssen, ‘The Validity and Utility of Risk Assessment for Inpatient Suicide’, 19(6) Australasian Psychiatry (2011) 507-512; L Quinlivan, J Cooper, L Davies, et al., ‘Which are the most useful scales for predicting repeat self-harm? A systematic review evaluating risk scales using measures of diagnostic accuracy’ BMJ Open 2016.