The social construction of decision-making capacity

I’ve been completing a paper on the location of rationality in medical decision-making and the relationship it has to decision-making capacity.

Decision-making capacity is now a critical concept in English law because it underpins when treatment in an individual’s best interests may be authorised under the Mental Capacity Act.

But the question I want to explore here is what kind of thing is decision-making capacity? It is clearly an assumption of the law that the capacity to make a decision inheres to the individual even if the reason the person lacks capacity relates to undue influence placed upon her by another (see A Local Authority v A). But what kind of property of the person is decision-making capacity?

It could be like height. Height is a straightforward property of bodies. Measurement may still be problematic, because we still have to choose a unit to express the measurement in: inches, cm or hands if the thing you are measuring is a horse. But the idea is uncontroversial.

But decision-making capacity involves at least two parameters. To make a decision I have to understand and retain the information and I have to be able to use and weigh the information in the process of deciding. In the MacCAT-T, a validated clinical assessment tool for measuring decision-making capacity three parameters are assessed: understanding, reasoning and appreciation.

So perhaps decision-making capacity is more like the size and shape of my body. My body has a weight which like my height can be measured in pounds or kilograms. But it also has dimensions. If I only want to know my weight then kilograms are an adequate measurement. But if I want to know if clothes are likely to fit I need to know my dimensions.

And even armed with information about the dimensions of my body I may still struggle to find clothes that fit. Clothes have properties of their own which determine their fit such as their fabric and cut.

To rationalise this process for women the clothing industry applies labels to make it easier for purchasers to work out which ones might be worth trying. Clothes are labelled 8, 10, 12 etc. But the labels can become a source of confusion if they are mistaken for information about the woman’s size. It is widely accepted that it is desirable to fit into smaller clothing sizes because our society values thinness, the number 12 must be better than 14. But sizes evolve, they change over time, according to some people because womens body shapes change and manufacturers make clothes to fit a normal distribution across the population. Others argue this phenomenon is best described as ‘vanity-sizing’ because it is motivated by the manufacturer’s desire to flatter their customers that they remain a size 10 even when they have changed shape or gained weight.

Either way labels like 8, 10 and 12 are a clothing industry construction which women might reasonably perceive as containing more information about their bodies than they actually do [note I have no reason to think the phenomenon is exclusive to women but it is less likely to affect men whose clothing is still predominently sold by reference to physical dimensions].

Finally, how do we understand if a body’s mass is healthy. Here we reach the most troubled water of all. For this the standard tool is the Body-Mass Index, famously based on a equation developed in the mid-19th century to enable comparisons across whole populations. It only began to be used in the clinical assessment of individuals in the late 20th century. The BMI is a troublesome tool because, as with the MacCAT-T, it involves measurements, calculation and then an application to the individual. This application to the individual is based on standardised tables based on actuarial assessments of risk associated with different BMIs carried out in the 1950s. There is ongoing controversy over whether the standard cut-off point for assessing a BMI as overweight actually reflects a significant level of risk to the individual’s health.

Assessments of decision-making capacity run into the same problems, with the added complexity that there is no tool available to measure decision-making capacity which has been as widely validated as the BMI (and I’m willing to assert will not be since the idea is vastly more complex). The MacCAT-T does not give an output in terms of yes or no. It merely provides some numerical outputs concerning understanding, appreciation and reasoning which the clinician can take into account alongside the wider knowledge of the patient’s history, the magnitude of the decision to be taken and any other relevant factors.

Comparing the measurement of decision-making capacity with BMI exposes the socially contestable dimensions of the idea. There will always be borderline cases because the assessment relies ultimately on a judgement that this individual has crossed a threshold which another has not. But unlike BMI, there is no comfortable possibility of recourse to international committee of experts who will defend the idea that 25 is an appropriate threshold for the definition of ‘overweight’. In borderline cases resolution can only ever be achieved through the application of more clinical judgment.

And the problem of dimensions of the body resembles the problems created by poor information sharing leading to lousy therapeutic relationships leading to mistrust. The person assessed as lacking capacity is like the woman who no longer knows if she is an 8, 10 or 12. Capacity assessment can and should be a robust process founded upon transparent norms which are communicated early on. But it frequently is not.

But does this mean that capacity isn’t just a property of the person after all. It seems to me the answer must be yes and no. We could argue that capacity is interactive and that with support at any given time most people could make a decision. I respectfully disagree. I suspect that more people can make decisions than we adequately support currently. But there are clearly many circumstances where decision-making is impossible. I have epilepsy and sometimes experience psychotic symptoms and I’d say seizures, the immediate post-ictal state and acute psychosis fundamentally undermine my decision-making capacity.

But even if we assume, as I do, that decision-making capacity is at least sometimes a property of minds as well as the social world the tools we use to measure it are inherently social constructions. Just like the tools we use to measure the body. So which tools we select and when we choose to employ them suddenly become very politically-laden decisions.

My own view is that the idea of capacity is as socially constructed as the idea of bodyweight but this does not undermine the need to assess it or make decisions on the basis of these assessments (or as my favourite social theorist Niklas Luhmann put it enigmatically ‘Reality may be an illusion but the illusion itself is real’). But we need to be constantly monitoring the way in which social and political factors interact with our understanding of what it means to have capacity to ensure that current assessment practices remain adequate.


Turning realisable rights into real rights

My understanding at this point is that AA is still trying to participate in hearings concerning the placement of her daughter P. This ongoing litigation will take place in the family courts and falls outside both my expertise, and the terms of reference of this blog. There is thorough commentary raising a range of different concerns here, here and here.

There were some really useful points raised on this post and I apologise for not replying to all of them. One which I did not reply to because I did not want to say anything glib was ‘I don’t agree that the innocent have nothing to fear. They may not have anything to fear from the law when it works as it is intended to, – but how was this woman supposed to find someone to defend her interests?’

This is the nub of the problem. As outsiders with no actual knowledge of the situation but considerable knowledge of domestic law all the mental health and family law commentators swiftly realised that the initial reports in the Daily Telegraph and Daily Mail must have been overstated or contain significant omissions. We now know this to be the case as Cardiff Law School summarises.

At the same time we know that even if the law worked as intended (and you’ll gather from my own posts I think that is a moot point) AA did not have opportunities to defend her own views of her situation. Her ‘interests’ were represented by no doubt highly skilled counsel appointed on her behalf by the Official Solicitor. So AA enjoyed a lot of rights to process, but little substantive right to challenge the decisions being made about her.

And there is a cruel irony to this because AA’s rights to a fair hearing and to respect for her private and family life – the two rights which may have been infringed here, were actually protected in turn by a complex package of international legal obligations.

AA’s protection under international law

AA was uniquely vulnerable. She was in the third trimester of pregnancy, in a foreign country away from family and friends, experiencing a mental disorder which may or may not have affected her capacity to reach important decisions about her life and she was subject to detention. Her vulnerability is recognised under domestic law – which is why we have all these procedures before she can be detained under the Mental Health Act, subject to forced treatment under the Mental Capacity Act or have her child removed from her care under the Children Act. Her vulnerability is also recognised under international law because she would have fallen within the scope of the Hague Convention on the International Protection of Adults to which Italy is a contracting party. The Convention does not, in itself, create any rights to respect for AA’s fundamental rights. But Schedule 3 MCA did create an obligation on the UK authorities to ensure that that the Italian authorities were notified of her status and that AA’s interests in her property were secured. I have no idea how these obligations are interpreted within Italian constitutional law. Within domestic law such obligations would have to be interpreted in light of s.6 of the Human Rights Act and would require consideration of the principle of respect for AA’s private and family life under Article 8 of the ECHR, i.e. there may have been a duty on the Italian authorities to consider what steps they could take to keep mother and child together.

AA’s protection under EU law

AA was an EU citizen and a worker when she arrived in the UK and thus exercising rights under the Treaty on the Functioning of the European Union. Her access to emergency healthcare was a right exercised under Regulation (EC) 883/2004 (warning pdf) and presumably Essex Health Trust sought reimbursement for her care from the Italian authorities. If they did then decisions made about her care undoubtedly fell within the scope of EU law. This matters because after the decision in Akerberg-Fransson we know that the scope of the Charter of Fundamental Rights of the European Union (the Charter) extends to the interpretation of domestic legislation which is being used to implement an EU right, so in this case it could be applied to the interpretation of the Mental Health Act and Mental Capacity Act insofar as they are being used to define the scope of AA’s rights of access to treatment under EU law. The key right here is Article 47 of the Charter which protects the right to a fair hearing and incorporates all the procedural rights guaranteed under Article 6 ECHR. You might ask (as Mostyn J. did in another of his recent judgments) how the Charter is of relevance when the UK explicitly derogated from it by negotiating the famous opt-out: Protocol 30 TFEU. The answer is that in a close analysis of the text the Advocate-General in N.S. v Secretary of State for the Home Department found that the Protocol primarily reaffirmed the content of Art 51 of the Charter and clarified the application of the remaining provisions, but was clearly not intended to disapply the Charter to those states which had negotiated it. There is a crucial difference between fundamental rights as recognised under EU law and under the ECHR however. If it is found that a state’s action in implementing EU law has been inconsistent with the Charter then the requirements of EU primacy require the national court to disapply the domestic law, and not merely find the provision incompatible.

This protection is far more problematic, I do not know if Essex Health Trust conceptualised AA as an EU citizen protected by the Charter, but I would argue they certainly should have done. AA’s rights were distinctively affected because she was an EU citizen working in another member state. Had she remained at home in Italy her vulnerability would have been substantially diminished.

These international legal mechanisms for recognising rights apply before you even consider the application of the ECHR to AA’s case, or the rights contained in the UN Convention on the Rights of Persons with Disabilities which the UK has also ratified.

But to transform these realisable rights into real rights AA first had to be an actor with legal capacity and able to define her own interests.

If there is a larger moral to this then it is that the political significance of legal capacity cannot be overstated. Defining rights alone is of little value. The large and as yet unanswerable questions all remain. How do we define who cannot exercise legal capacity or provide appropriate assistance to those who need it? I know of no supported decision-making regime, for example, that has solved the problem of supporting decision-making when someone has current psychosis, and as someone with personal experience of psychosis my imagination struggles to accomodate that.

And how can we begin to ensure that the failsafes our system applies to ensure that when individuals are vulnerable for multiple reasons (pregnant, foreign, seriously ill, detained) we are able to interpret their interests and represent them robustly?

A final point on orders

In this case Mostyn J. had at least four options available:

1. To consider the local authority’s request for a police protection order under s.46 Children Act. This would allow the police to remove the baby from AA’s care immediately after the birth.

2. To suggest the local authority apply under s.38 Children Act after the birth seeking an interim care order for the child. This would allow a hearing where AA’s interests could be represented by the Official Solicitor and the child’s welfare could be discussed.

3. To make an order under s.16 Mental Capacity Act finding that it is in the mother’s best interests to be cared for in a Mother and Baby Unit. It is not clear if this was properly advocated for by the Official Solicitor. The practical effect of this decision would be that the child would remain with the mother, but would not at this point be subject to any order under the Children Act. The Mother and Baby Unit option was recommended by the clinician working for the Health Trust and would allow AA and her baby to remain together consistent with her wishes. If it was necessary to detain AA to make this option possible then this could be done under the Mental Health Act (there is precedent for this, in Re E(Medical Treatment) the CoP ordered that the patient receive treatment for her mental disorder in her best interests, but in order to practically achieve this she was subsequently detained under s.3 MHA).

4. To make no order or recommendations at all.

Since the baby was to be born the following day we will assume 4 was not a realistic option. Mostyn J. ruled out 1 on the basis that it was heavy-handed. He took the initiative, indicating that he is capable of taking initiative and advocated 2. Why didn’t he take the initiative and apply 3 instead which was consistent with the medical evidence, evidence of AA’s wishes and presumably consistent with the child’s welfare?

There might have been strong reasons to suppose AA posed an immediate risk to her child after the birth. But this evidence did not appear to have been advanced by the local authority, instead the evidence was she might not be capable of caring for the child due to an earlier history of neglecting her other children.

In any event if AA posed immediate risks to her child after birth there would also be risks involved in option 2 which he did choose.

Choosing Option 3 would not prevent the local authority applying for an order under s.38 at any later point if they became concerned about the child’s welfare, or in an emergency asking the police to apply for an order under s.46. It only facilitated choices, it did not limit them.

Choosing Option 3 would rely on a bed being available. This would be an obligation on the Health Trust and not the local authority. Since AA was about to undergo major surgery Mostyn J. could easily have said that he was minded to make an order under the Mental Capacity Act but invited the Health Trust to ensure that the relevant service was available for AA and her child before doing so and then reconvened the hearing the following day.

In any event Option 3 was not chosen, despite the Official Solicitor representing the view that it should be. A further question then is why the Official Solicitor did not appeal this decision? It was of grave significance to AA, and as Celticknot pointed out in an earlier comment on this blog the senior courts have been critical of judgments made in the early stages after a child is born which lead inexorably to adoption. An advocate concerned with promoting AA’s best interests should have been vigilant regarding this point.

It has been said this is an extempore judgment. I agree. But I don’t think that justifies a failure to survey the options which has such serious implications for AA.

A post which is about caesarean without consent

‘I might, if it is published, deplore Mostyn J’s reasoning about the mother’s best interests or her capacity…’

Showing uncanny prescience Mostyn J’s judgment and the transcript of the case of Re AA have now been published and I deplore them (warning pdf).

I wanted them to be better than this. But the judgment does not really explain how the test of capacity in s. 2 Mental Capacity Act applies to AA and reading the transcript my suspicion is that he has not assessed AA’s capacity properly, and may even have confused it with her best interests.

What did the judge know at the point he made his judgment?

AA was detained in hospital under s.3 of the Mental Health Act. She had a ‘schizophrenic disorder which was psychotic in nature’, and she was experiencing delusions.

She had had two previous births by caesarean section increasing the risk of a uterine rupture in this pregnancy to something close to 1%.

At the time of the hearing she was 39 weeks pregnant and it was proposed that the caesarean be performed the following day.

Her doctors favoured a planned caesarean performed under a general anaesthetic because they feared that if she was allowed to go into labour spontaneously she might lie about the onset of labour and interfere with efforts made to help her. Planned caesarean would allow them total control of the process and thus make the process maximally physically safe.

The finding of lack of capacity

In the judgment Mostyn J. tells us that he finds AA lacks capacity ‘within the meaning of s.2(1) MCA. But s.2(1) MCA only tells us that for a person to lack mental capacity in English law they must have a ‘an impairment of, or a disturbance in the functioning of, the mind or brain.’

AA certainly met this requirement. But the judgment does not tell us why she met the further requirements of being unable to make a decision under s.3 MCA. To be unable to make a decision a person must be unable to understand, retain, use or weigh or communicate their decision even after all necessary practical assistance has been provided to them. The judgment contains no discussion of what efforts have been made to discuss birth planning with AA, nor of whether she understands the consequences of refusing a caesarean section.

This does not mean AA had capacity at the time. Only that the judgment does not communicate that she did not.

The finding that a caesarean was in AA’s best interests

He further fails to apply the best interests test in s.4 correctly. He correctly identifies that the interests he is meant to take into account are primarily those of the mother and not those of the child who is, as yet, unborn. This is clearly a difficult task since in late pregnancy the interests of most women are intimately intertwined with those of their baby. There is a suggestion this was true for AA, because it is recorded in the transcript that she wants to see and hold her baby after it is born. Nevertheless the position in law is that we cannot start from the position of healthy baby = happy mother and work backwards to establish what the mother’s interests are.

We can however, identify in some cases that the mother’s best interests will prospectively be achieved through a healthy delivery and take that into account when making a best interests assessment. This is true when giving birth to a healthy child will be critical to the mother’s mental health for example. Unfortunately, Mostyn J. places a high value on this point but fails to consider any other evidence concerning the mothers past wishes, feelings, beliefs or values. Nor is any evidence from her family concerning her views on childbirth cited. Of course, this may not have been available, but its absence should have been noted and strongly regretted since it makes the decision reached under s.4 MCA largely artificial.

The actual best interests assessment to be reached here was a subtle one. Vaginal birth after caesarean is a normal practice and the risks of uterine rupture cited in the judgment are those given in standard literature given to women considering this. So if AA was refusing a caesarean this did not by itself indicate anything amiss. Many women with capacity with a history of previous caesareans choose to give birth vaginally every day. The case also reinforces the view that adults lacking capacity are not allowed to make the same risky decisions that are permitted to the rest of society.

As Lucy Series predicted, if AA had been present a better judgment might have been reached. There might have been practical difficulties in securing the physical presence of a woman with serious mental distress, currently detained who was also 39 weeks pregnant at this hearing. But the sheer absence of her voice in the decision being reached about her life is utterly inexcusable. We must do better than this.

Finally, the thing that utterly baffles me about this judgment is Mostyn J’s objection to the placement in a Mother and Baby Unit, which is what her treating clinician had recommended and which would have been by far the best option for her since it would have respected the one wish we know her to have had – to have contact with her baby. I cannot make head nor tail of paragraph 7 of the judgment and can only identify uncharitable (to Mostyn J.) interpretations to place upon it. If anyone could help me out I’d be grateful.

A post that is not about forced caesarean sections

Commenting on this case is uncharacteristically fashionable for this blog.

But it has to be said that even after the publication today of the family court judgment concerning the placement of the little girl, P, for adoption there is still far too little which we can seriously evaluate here.

What can’t we say?

We cannot evaluate the original decision to admit the mother to hospital. It would appear that she has a history of significant mental disorder and has continued to receive subsequent specialist mental health care in Italy but the actual symptoms that precipitated her admission in June 2012 are not in the public domain and probably never will be since decisions to detain someone in hospital in this country are clinical rather than judicial and remain confidential.

We cannot evaluate Mostyn J’s judgment in a Court of Protection hearing held in August 2012. The hearing presumably found that the mother did not have capacity to consent to a caesarean section but that it would be in her best interests for the procedure to be performed. This could be for any number of reasons. That judgment is not in the public domain although it may yet be published. It is not clear why the decision about caesarean section needed to be made in court. Many pregnant women with mental disorders detained in hospital will retain capacity to make decisions about their care during the birth so this was not an inevitable or essential step by the hospital. There must therefore have been a specific aspect of the mother’s care that precipitated the move to seek a court order specifying that a caesarean section would be in her best interests.

It is worth noting that by taking this action the hospital were in fact opening their decision-making up to scrutiny, not concealing their actions in order to facilitate a secret adoption. Ss.5 and 6 of the Mental Capacity Act codify the doctrine of necessity – they allow clinicians confident that a medical procedure which is in an incapable adult’s best interests to carry it out and provide a general defence against actions under the civil or criminal law for doing so. If the mother’s clinicians were confident that she lacked capacity to consent to a caesarean and nevertheless would (or might) require one they did not need to apply to the Court of Protection for scrutiny of that decision.

It is also significant and welcome that the hearing took place in the Court of Protection. There is highly controversial caselaw that a caesarean section may be administered to a woman detained in a psychiatric hospital under s.63 of the Mental Health Act as treatment for her mental disorder (Tameside and Glossop v CH [1996] 1 FLR 762). In theory the hospital could, if the mother had refused a caesarean section for reasons relating to her mental disorder, have performed the procedure without her consent as long as it was in her best interests. This is an alarming state of affairs, but it is nevertheless a case that whilst criticised has not been overturned (though I’d strongly urge clinicians reading this not to administer caesarean sections to their detained but capable and refusing patients on ethical grounds alone).

Instead the hospital took a far more responsible course of action and asked the Court of Protection to clarify whether the mother had capacity and whether the intervention was in her best interests – thus providing two safeguards which do not apply to treatment administered under the Mental Health Act. I might, if it is published, deplore Mostyn J’s reasoning about the mother’s best interests or her capacity but I welcome the fact that the issue was decided in the Court of Protection.

We can evaluate the fact that soon after the birth of her child the mother returned to Italy whilst P remained in England. This was regrettable. Why this took place is not clear. The mother may have had strong reasons for wanting to return to Italy. P was with foster carers and it may not have been in her best interests to be removed and placed with a mother who was still seriously unwell at that time. Nevertheless allowing the mother and child to become separated at that time was a decision (or possibly omission) on the part of clinicians which has had significant implications for mother and child subsequently.

And we can evaluate the judicial reasons given for allowing P to be placed for adoption. These are simple. P was six months old at the time of the hearing in Feb 2013. The object of a Family Court hearing is to secure the welfare of the child. There is evidence that if adoption is agreed to be the best means of securing the child’s long-term welfare then the child’s best chances of achieving a stable placement are attained if the child is settled before nine months. So if adoption was agreed to be the best outcome time was of the essence for P. If on the other hand P was to return to the care of her mother, as her mother wished, her hope was that P would remain with foster carers in the UK for a further year before making the transition to life with her mother in Italy.

P’s mother had made a remarkable recovery and was doing well, but this was against a backdrop of recent serious mental illness. She did not currently care for her two older children. She did not have extended family who were in a position to care for P and it did not appear that P’s father was able to care for P. In those circumstances there was a risk that if P’s mother relapsed she would not be able to provide the stable placement P now needs if she is to have the maximum chance of having a good start in life. Nor was the transition plan proposed one which maximised P’s chances of stability.

At this point the circuitous route by which P had reached local authority care is irrelevant. What matters is reaching the decision which will reflect P’s best interests between now and when she turns 18. On the basis of the evidence the judge found that she would be more likely to find stability with an adoptive family than with her mother, even though adoptive families can be imperfect and unstable. As a result he over-rode the requirement for the parent’s consent to the adoption and allowed Essex County Council to start seeking a permanent family for the child.

What can we say?

There may well have been some cock-ups in this case. In particular, para 9 of the judgment describes a family court hearing in October 2012 when the mother appeared unwell but her doctors asserted that she had both legal capacity and the capacity to decide to return to Italy which she subsequently did. This practical separation of mother and child only seven weeks after the birth must, presumably, have made re-establishing contact difficult and raises serious and troubling questions about how we define best interests in cases concerning families. P’s interests were paramount and her mother at that time may not have been able to care for her, but it is not obvious on the basis of what has been reported so far that anyone involved was making decisions with a view to promoting the mother and child relationship.

At the same time whilst we do not know why the application for an order finding that a caesarean section was in the best interests was made we do know that making the application suggests a concern to uphold the law.

Nothing in the facts currently in the public domain suggests a conspiracy. I could try to excuse the terrible sensationalist reporting on this case on the basis that it does involve the complex operation of three different statutory jurisdictions. But I cannot grant that excuse. The underlying statutes are not difficult to understand. Decisions to admit people to hospital on the basis of a mental disorder without consent must be made under the Mental Health Act 1983, decisions taken about treatment for people lacking capacity must be made under the Mental Capacity Act 2005 and decisions made regarding the subsequent welfare of any child born as a result of those interventions must be made under the Children Act 1989.

The people who make decisions about Mental Health Act admission and treatment are psychiatrists and social workers and therefore employees of the NHS. The decisions made about the welfare of the child were made by social workers working in two different teams within Essex County Council. In the family court the judge was assisted by the evidence by P’s guardian and in the Court of Protection the mother will have been represented by the Official Solicitor. The conspiracy theorists are asking us to believe that multiple psychiatrists and social workers working for both the NHS and Essex County Council, assisted by judges sitting in two separate courts, as well as independent representatives of CAFCASS and the Official Solicitor colluded to acquire a baby. These people have clearly never tried to organise a meeting in the public sector.

There is a lot wrong with our legal system. But misrepresenting it as a closed system capable of a monumental conspiracy on this scale is not only a calumny it is dangerous. It leaves the people who have most cause to rely on it with unjustified fears – instead of focusing on the perfectly justified fears we should be addressing.

Pregnant women with mental disorders, even if currently detained, should not be given the impression that substitute decisions about their clinical care during childbirth are likely to be made on the basis of a social worker’s view about the best interests of the child, rather than what is in their clinical best interests. This is not true. Social workers were not involved in that decision.

Nor is it true that the judge made a decision to override the mother’s wishes for fear that she would stop taking her medication in the future as was reported in both the Daily Telegraph and Daily Mail. The far more profound issue was that the mother’s plans for the child involved a long period of uncertainty and the judge took the view that a stable placement through adoption was to be preferred.

There is a final question this case raises. We know that reporting judgments may be a good thing because transparency in judicial reasoning is right in principle. In the family courts and Court of Protection it may come at a cost to vulnerable participants who have no power to control the information which enters the public domain in this way.

One justification for reporting more cases is of course that we can scotch pernicious inaccurate reporting of this nature by ensuring the facts are available from the outset. But a trend worth monitoring when hysteria of this nature is so easy to whip up is the relationship the facts have to the commentary.