Connor Sparrowhawk died in an assessment and treatment centre in Oxford in July 2013 after suffering an epileptic seizure whilst he took a bath. Connor had been diagnosed as having epilepsy for two years, was receiving medication for his condition and there were many signs in the weeks preceding his death that his seizure activity might have increased.
An independent report commissioned by Southern Healthcare NHS Trust which runs the centre where Connor was a patient has just been published. It has found that Connor’s death could have been prevented if staff had recognised that his epilepsy created risks around taking baths and conducted a proper risk assessment.
This post is not about the report. Instead it is a reflection about the current state of the law on the liabilities of state institutions responsible for detaining people who subsequently die in their care.
Has the Trust violated Connor’s right to life under Art 2 ECHR?
There are two aspects to this question:
Did Southern Healthcare owe an operational duty under Art 2 ECHR to Connor?
At the time of his death did Southern Healthcare know of a real and immediate risk to Connor’s life which they could have taken reasonable steps to avert?
The answer to the first question must be yes. Connor’s legal status at the time of his death was that of an informal patient. His admission history had been chequered, he had been admitted as an informal patient, later detained under the Mental Health Act and then discharged from MHA detention although an assessment completed at the point of his discharge from detention concluded that he lacked capacity to make decisions regarding admission. A standard authorisation for his detention had been sought under Schedule A1 MCA but this had been refused so there were no formal controls in place regarding his detention status. However, after the unanimous Supreme Court decision in Rabone v Pennine Care NHS Trust it is clear that the approach to be taken to deciding the question of operational duty under Art 2 ECHR is one of substance not form. Connor lacked the capacity to make decisions about his admission and he was on a locked unit. He was manifestly deprived of his liberty and under the control of the state. As such he could not have protected his own right to life and it was the state’s duty to do so instead.
The content of the report only tell us that Southern Healthcare staff could have prevented Connor’s death which is not the same as saying the death was predictable. Liability under Art 2 ECHR does not arise simply because the state has taken on the care of someone who has an illness and subsequently dies. Nor does it arise because the state has taken on the care of an individual within a well-organised healthcare system but one rogue professional acts negligently.
The starting point remains the infamous ‘Osman’ test – that the state can only be liable if it knew or ought to have known of a real and immediate risk to the individual’s life and had failed to do all that could reasonably be expected of them to avert that risk.
The fact of Connor’s de facto detention is not decisive although it is persuasive. His family would also need to demonstrate that the risk of death was ‘real’ in this sense meaning serious or significant and ‘immediate’ which has been glossed as meaning ‘present and continuing’. Threats of suicide 18 days prior to a suicide were held in Renolde v France to have been adequate evidence that a risk was immediate. Repeated warnings by family members that an individual is showing signs of seizure activity would seem to fall into a similar category.
The fact that the risks posed by Connor’s epilepsy were not known to staff are a matter of evidence, but ignorance is irrelevantif it is found that the staff should have known about the risks.
Finally the fact that small measures could avert the risk weighs strongly in favour of finding a substantive breach. The reason the test in Osman v UK is typically described as a high threshold to reach is political rather than legal. Neither the UK not the Strasbourg court wants human rights law to become the vehicle for extending the liability of public authorities to cover liability for the actions of criminals or individual incompetent professionals. Such an approach could subvert the ends of justice. But systemic failures are different. Here there is a policy argument in favour of using human rights law as the tool to force healthcare providers to prioritise patient safety or the police to maximise the safety of detainees. This is reasonable provided the changes in practice this requires are those which are included in the definition of good healthcare – like assessing the risk of drowning if someone with epilepsy has a bath unsupervised.
The decision in Rabone has made it easier for these substantive claims to be taken, notwithstanding Lord Dyson’s emphasis on the language in Osman and Van Colle that the test of ‘real and immediate’ risk was significantly higher than that employed in the common law of negligence.
In the case of people detained by the state the threshold set for establishing a real and immediate risk is only notionally different from the very low threshold set for foreseeability in tort. It is not clear from the facts of the decided cases that the two areas of law actually operate differently. I cannot identify any decided cases which did not meet the requirements for ‘present and continuing’ risk under Art 2 but which would still give rise to a successful action in negligence (because the test for foreseeability in tort is lower than the test of immediate harm) and sadly the issue is unlikely to be tested. Deaths in state detention tend to involve people who do not leave dependants (in the narrow sense defined by the Fatal Accidents Act 1976) and the damages that can be recovered for their death under an action in tort will be small so in practice actions are always taken using the Human Rights Act with small additional claims made for the costs arising from the individual’s death.
Taking an action for a substantive violation under the Human Rights Act may not be what the family want to do.
The state also has an obligation to investigate all deaths which arise when a person is under state responsibility. In England and Wales the primary mechanism for achieving this is via the coroner’s inquest. Recent caselaw has held that investigations such as the one published today are not essential for meeting the requirements of Art 2 although publishing them is still good practice.
I get the very strong impression that when an individual dies in the care of the state the responsible authorities are deeply conflicted about how to meet their investigative duty under Art 2 in a way which also meets the needs of the family and how to avoid admitting liability for failing to meet positive duties under Art 2. If I was advising a healthcare provider I could not tell them how to do both.
The issue is not damages. The quantum awarded for actions taken under the Human Rights Act are typically small, and pragmatic legal advice on any borderline case would be to suggest the authority apologise and settle an action as early as possible to avoid legal costs and the costs of staff time involved in giving evidence. A bigger issue is reputational harm. The current uncertainty about what actions may give rise to liability after Rabone means that honestly acknowledging poor care for an in-patient who lacked capacity may be tantamount to acknowledging liability for their death and a violation of their right to life. This raises the stakes considerably.
At the same time families have consistently demonstrated their need to understand what happened to their loved one. The facts of Powell v UK describe a nearly 10 year long battle by a couple desperate to establish the facts concerning the medical decision-making prior to the death of their 10 year old son who died as a result of medical negligence. In the subsequent caselaw concerning deaths in prison, police custody and most recently in mental health services families have made it clear that they need to understand the circumstances which lead up to their loved ones death as soon as possible after the death. Waiting to know fuels fear, anger and the plausible belief that the provider must have something to hide and that others may be in danger. All of this at a time when the family are grieving.
Can anything be done to improve this?
Two things could possibly improve it. One is emphasising that establishing how a death occurred is distinct from the question of fault. This distinction is actually made in coroner’s practice, but is confused at the level of some of the leading judgments on how coroner’s inquests serve the purpose of Art 2.
Current caselaw on the function of the inquest has emphasised that its role in meeting the state’s investigative duties under Art 2 ECHR are ‘parasitic’ on the state’s positive duties under Art 2. In other words the investigative duty exists to serve the positive duty, and it is primarily necessary to perform the investigative duty in circumstances when we think the positive duty may have been compromised (Middleton and R(Smith) v Deputy Coroner for Oxfordshire). The situation is subtly different where the death occurs outside detention. In Takoushis it was argued that the investigative duty still arose when the person died outside custody, but here the policy reasons for it (based on close analysis of ECtHR caselaw) were different. The object was to ensure that whether by empowering family members to take a civil action or by holding an inquest the legal system must ensure that all deaths which might be attributable to negligence can be investigated. The judgment in Takoushis found that the English legal system met this requirement. So the current position in English law does not support the view that when the operative duty under Art 2 is engaged investigations into the death should be an end in themselves.
But the behaviour of families does not suggest they need to know how the death occurred only or mainly because they need to know if the death was unlawful. They need to know for a wide range of reasons which our legal system does not currently address.
Another question that arises is whether we need a procedure which falls between inquests held under the Coroner and Justice Act 2009 and the root cause analysis internal investigations conducted under the current guidance on serious incident investigations. The objects of the inquest are to serve the needs of the community and in most cases to meet the formal requirements of an independent investigation required to satisfy the investigative duty under Art 2. And the objects of the internal investigation are to ensure healthcare providers identify dangerous practices and potential improvements for the future. But neither procedure is adequately oriented to meet the needs of bereaved families.
I don’t have a good conclusion for this blogpost. But Connor Sparrowhawk’s mother does. It is a blogpost called Imagine.