From Cheshire West to Northern Ireland – A new Dawn for Deprivation of Liberty

Executive summary

• The Supreme Court has now overruled its landmark judgment in Cheshire West (2014), which had set the definition of deprivation of liberty (DoL) so widely that rights under Article 5 of the European Convention of Human Rights (the right to liberty) were engaged for hundreds of thousands of people across the health and care landscape.  
• The Supreme Court clearly intends to set the bar for a DoL much higher now, with fewer people defined as being deprived of their liberty, especially in their own homes, or in hospital for medical treatment. But in doing so it may have created fresh uncertainty - and is likely to generate more litigation - over where exactly the new line is to be drawn.  
• It will be important to keep on top of that emerging further case law, and to share best practice to minimise inconsistency in application across and within organisations.  
• In particular, the idea that someone who lacks capacity to make decisions about their care and accommodation may nonetheless “consent” to the arrangements by showing they are “happy” there, may be problematic.  
• But there is no further appeal, and we now have to adapt quickly to the new, more complex legal test – reviewing workforce, resources and approach across health and social care providers and commissioners, updating policies, updating training, and reassessing cases to consider afresh whether there is a DoL that needs to be authorised.  
• This might pave the way for the government to introduce the Liberty Protection Safeguards, having announced its intention to further consult on this later this year. (Or perhaps there may now be less perceived need for reform if the number of people regarded as DoL is brought back into numbers considered more manageable with the system that we have?)  

Background 

We all know that the legal framework around deprivation of liberty is in desperate need of reform; tens of thousands of people are considered to be unlawfully deprived of their liberty, whether waiting too long for an authorisation under the Deprivation of Liberty Safeguards (“DoLS” - for people over 18 years old, in care homes and hospitals), or when an application ought to be made to court in other situations. 

The current DoLS system, launched in 2009 and funded for an estimate of a few tens of thousands of cases per year, has been swamped by the hundreds of thousands of cases a year there have been since the Supreme Court decision in Cheshire West in 2014. In that case the Supreme Court determined that the test for identifying a confinement for the purposes of a DoL should be that someone is “under continuous supervision and control and not free to leave” (the acid test), and — crucially — that good intentions, care that is appropriate for the person’s needs, or that they are content there – do not stop it being considered a DoL. 

As well as there being a confinement, (and any DoL being imputable to the state — everything in health and social care meets that test!), there is a third part to the definition of a DoL– there must also be no valid consent. This is one reason why things are more complicated for children, raising whether or not the consent of a parent ought to mean that there is no DoL, even if the restrictions meet the acid test for a confinement - broadly, parents’ consent can negate a DoL where the child is under the age of 16, but not after that. 

Cheshire West didn’t consider the question of valid consent at all – it was common ground that each of the individuals in that case lacked capacity to make decisions about their care and residence, and therefore there was no valid consent to the restrictions, allowing the court to focus on the question of whether the restrictions amounted to a confinement, according to the acid test. 

The Attorney General of Northern Ireland (AGNI) made an application to the Supreme Court all about consent. In particular, the AGNI proposes to amend the Code of Practice that goes with the DoLS system in Northern Ireland to say that there can be valid consent (and therefore there is no DoL), despite the person lacking capacity to make decisions about their care and residence, if they “actively express positive wishes and feelings about the arrangements”. 

Under devolution arrangements, the AGNI asked the Supreme Court if it will be lawful to make this change, and whether such a Code would still be compliant with Article 5 of the European Convention of Human Rights, given the inconsistency with Cheshire West. 

The Lord Advocate of Scotland supported the AGNI’s case, which the AGNI estimated would reduce the number of people currently regarded as DoL by around 25%, allowing the system to focus its resources more appropriately. 

The Secretary of State for Health and Social Care (SSHSC) also intervened in the Supreme Court to support the AGNI’s application but invited the court to go even further. SSHSC argued that the court should revisit the Cheshire West judgment itself, going back to the drawing board on how we define confinement, giving a chance to narrow the goalposts again, after the floodgates of the Cheshire West decision. 

Ranged against the AGNI, the Lord Advocate and the SSHSC, were the Official Solicitor and a group of charities. 

The Official Solicitor’s role is to act as a litigation friend for individuals in Court of Protection cases. She argued that there are all sorts of problems with the AGNI’s proposed approach – including the logical challenge of reconciling that someone lacks capacity to make decisions about the care and residence, but can nonetheless “consent” to it; the practicality of establishing whether someone is expressing “positive feelings” about a placement; and the wisdom of removing someone from these safeguards based on something so subjective, and likely to depend heavily on the opinion of those who are providing the restrictive care.

The charities involved (Mind, Mencap and the National Autistic Society) broadly backed the Official Solicitor and added their concerns about the extent to which people with disabilities might be (mis)perceived to be happy about their situation. They opposed the “highly subjective and hopelessly vague” idea that someone might be denied independent scrutiny as a result of an apparent “positive attitude”. 

View the written submissions.

The result was a big win for AGNI and SSHSC.

Decision 

The Supreme Court heard the case in October 2025 with an expanded bench of seven justices (there are usually five) as it was always clear that it was being asked to overturn an earlier Supreme Court decision.  It is therefore an unexpected blessing, at least, that there is a single judgment (on 2 June 2026) setting out the Court’s unanimous decision, albeit at some length (the judgment is 82 pages, but there is also a helpful 5 page press summary, both available here).  

The nutshell is that the AGNI is right that where someone is “happy” with arrangements that include restrictions, then their “consent” can mean that there is no DoL, even though they lack capacity for the relevant decisions.  And the DHSC was right that Cheshire West went too far, setting the bar too low for a DoL, and in oversimplifying an “acid test”. 

“Incapacitous consent”

The idea that someone can “consent” to arrangements, so stopping there being a DoL, despite having no capacity to make the relevant decisions about their care and accommodation, is difficult to swallow for someone used to working with the Mental Capacity Act, or in health and care decision making more generally.  

A fundamental principle is that “consent” to care or medical treatment, for instance, needs the person to have capacity to make that decision.  Without capacity to make that decision, we might at most have “compliance”, and we should be making a best interests decision rather than using the language of consent.

The Supreme Court explain this by saying that this is about the requirement for a DoL under Article 5 ECHR that there is “no valid consent” to the restrictions that amount to a confinement.  And that means something entirely different to our idea of consent in our domestic law to the care or treatment being given (though, confusingly, the word is the same!). And in ECtHR case law, it has often been held that someone can “consent” to restrictions even though they lack capacity to make the decisions for themselves.

The court sets the bar for being able to give that Article 5 “consent” to negate a DoL quite low – “A person may not have capacity to make decisions about their care and residence arrangements, but if they have a basic level of awareness and consciousness of their living arrangements that is sufficient to enable them to know and communicate whether they are happy or unhappy with them, they may be treated as able to give or withhold valid consent to confinement by an expression of their wishes and feelings” (para 201 – but this is expressed in slightly different versions in at least five places in the judgment). 

The fact that this might be difficult to assess in practice for some people, as the charities argued, should not mean that it is ignored for those who are able to give that “consent”, which should be respected, the court says.  

But what is meant by being “happy”? Elsewhere this is put in terms of “manifesting their acceptance”, or even that they do “not express or manifest any objection” (para 172): Though the judgment says that “it may be that … mere compliance or acquiescence… is not enough”, the examples actually given include many cases where there was either overt objection to the arrangements, or what the court called “tacit consent”. (Are we now closer to saying that in practice active objection is needed before Article 5 safeguards are triggered?).  

Bearing in mind that this decision is based on an abstract referral rather than a real case with a set of facts, it is likely to need many more judgments to delineate how this should be applied in practice. 

It is telling, though, that the judgment ends with reference to the individuals in the Cheshire West case. The court says that the two sisters, MiG and MeG, were not deprived of their liberty as in both case there were sufficient indications of them being happy with the arrangements (this is despite MEG’s “occasional outbursts”, and the fact that she received tranquilising medication – which was OK because it “was not administered with a view to disabling her from forming a view about her living circumstances and does not seem to have had that effect”). They also did not meet the objective test for confinement, since the arrangements were “relatively normal”.  

However, we approach the uncertainties of drawing this line for the subjective element, the concept of “incapacitous consent” is going to take some getting used to, and the risks are obvious of compounding already far too common misunderstandings about the MCA and consent to treatment.  

We are going to have to work with a system where someone lacks capacity for a decision about care or treatment, so a best interests decision must be made (which is not “consent” to that care); but if they are “happy” with the restrictions involved in that care or treatment this is “consent” to the restrictions and is therefore not a DoL.  This is going to be challenging at the frontline in practice, to say the least.

Cheshire West was wrong in every way

It was “both unfortunate and confused” that the question of consent was ignored by the Supreme Court in Cheshire West (para 84), but that judgment was also wrong in pretty much everything it said about the objective definition of DoL as well: 

• It was wrong to say that there is a simple “acid test”, which is “too crude” and “ignores all other potentially relevant factors”. 
• In particular, it was wrong for this to exclude the question of P’s compliance or lack of objection – which is relevant to whether or not there is an objective DoL / confinement. This matters because there is more likely to be conflict and coercion where there is objection. 
• It was wrong to hold that the “normality” of the situation is not relevant to whether there might be a DoL. Where there is doubt, cases are dismissed as being not a DoL where the situation was around provision of care and “is very far removed from the paradigm case of confinement in a prison cell” (para 203, 205).  
• Cheshire West takes no account of the limitations imposed by the person’s disabilities. The Supreme Court in AGNI approved of the recent High Court decision by Lieven J that a profoundly disabled 12 year old girl was limited not by her care, but by her disabilities, which was criticised at the time for being incompatible with Cheshire West.  We know now that Lieven J was right, and the Cheshire West decision was wrong.  A number of other judges who have pushed back against Cheshire West over the years will be feeling vindicated! The court approves Professor Eldergill’s analysis that “Mere inability is not lack of freedom”, but a DoL is instead about stopping someone from doing something that they want to do, and are physically able to do (see the point about coercion, below).  
• Cheshire West was also wrong in disregarding the “purpose” of the arrangements.  It matters, says the Supreme Court now, that restrictions are imposed with good intentions to provide care, and not to punish the person. (When this point was made in the Court of Appeal in Cheshire West it felt like the “death of DoLS” – if it is right that providing well-intentioned, appropriate care, which is the least you would expect, surely, would not be regarded as a DoL, who will say that they are not doing this, by seeking a DoLS authorisation?). 
• DoL refers to “physical liberty” and has no “sensible meaning” in the context of someone who is “catatonic, for example due to severe dementia or in the aftermath of suffering a stroke or traumatic head injury”. They are, essentially, too disabled to have liberty, or to be deprived of it (para 142). 

Having said that MiG and MeG were not DoL, the court reserved judgment on the other individual (P) in the Cheshire West case, but hints that he might not have been DoL either (he was under continuous supervision and control and was regularly physically restrained to remove dangerous material from his mouth, or put in clothing to prevent him getting to his incontinence pads to do this). 

The court sees the Cheshire West decision as having been policy driven. Lady Hale, famously, said that “a gilded cage is still a cage”, and that we should “err on the side of caution” in identifying a potential DoL to ensure the necessary safeguards are available to a vulnerable group where our best interests decision-making, no matter how well intentioned, is leading to arrangements that are particularly restrictive.  The Supreme Court now do not share those concerns; instead, it worries that scarce resources are being wasted on the DoL process, which itself generates intrusive and burdensome assessments for the person involved, when there are other more appropriate safeguards in place which are more proportionate than Article 5 (such as support plans, Care Act assessments which should be reviewed annually, and local authority safeguarding duties, as well as the obligations under the MCA itself and Code of Practice for proper best interests decision-making).

Wider comments – the family home to acute medical treatment  

Without the anchor of a particular factual scenario, the Supreme Court ranges widely across the relevant elements of deprivation of liberty. 

With the new focus on P’s expression of being happy, or not, there is also emphasis on the use of sedative medication as a red flag where this might affect the ability to express wishes and feelings.  

The court clearly sees “coercion” as key to a DoL, talking about “someone being compelled to live somewhere they do not want to live” (para 187), and contemplates this typically happening in an institutional setting.  

Though it says that it is possible for there to be a DoL in any setting, including in a private home (para 54), it goes on to put this in terms that raises the bar there in practice: 

“The parties identified no decision of the European Court to date which has held that an individual living in their own home is deprived of liberty”. And restrictions in a family home would “need to be more severe or extensive to amount to [a DoL], such as, for example, a combination of restraint, medication, and seclusion”. (para 193).  

There are also interesting comments on the scope of DoL in the context of medical treatment. The court endorsed the view of King LJ in Briggs that someone “being cared for in a minimally conscious state” is not deprived of their liberty, but also went much further:

“In ordinary circumstances, an individual in hospital to receive treatment (and who, for their own protection, would be prevented from leaving while, say, confused and unable to think because of medication) would not be regarded as suffering a deprivation of liberty… The presumption is that a person who is unconscious and kept in care to have medical treatment administered does not suffer a deprivation of liberty, unless there is evidence that they are in fact in detention or there is an established pattern of behaviour to show that they clearly do (or would) object to being cared for in this way”. (para 145). 

This seems to go well beyond the Court of Appeal line in Ferreira, which was expressed in terms of “life-saving medical treatment” generally being no DoL, and finding the edges of this line now will be an important part of the inevitable next wave of case law.  

Implications 

It seems obvious that the court intended to raise the bar for defining a DoL, and it might well have that effect.  In doing so, however, the court has made the defining line much less clear. You might think that the Cheshire West acid test caught too many people, but it had the virtue of relative simplicity.  

The test for an objective DoL is now much more complex. Now that we know that “continuous supervision and control” is not enough for confinement, we have to consider other factors to identify a DoL – the purpose of the restrictions, P’s compliance, the context, the relative normality, and so on. But how do we apply those, and what weight do we give to each? 

And we now need to consider as well whether P is “consenting” to the restrictions, such that it is not a DoL, even though they lack capacity for the decisions about their care and residence. There will be complications about what this means, whose evidence of “happiness” ought to be relied upon (bearing in mind the potential conflicts of interest for providers involved, who are less motivated to identify those in their care as “unhappy”), and how do we deal with inconsistency – someone who is happy with some but not all aspects of their care?  Or someone who is happy most of the time, but sometimes is not? Does a “bad” afternoon, when P objected to their care, mean that they were (unlawfully) deprived of their liberty for that time, and how are providers supposed to deal with such variation in short periods of time? The court doesn’t really deal with duration / timescales of restrictions before Article 5 is engaged, and this might become a more important topic now in this context, where there are short periods of “unhappiness”.    

Given the greater complication, now, of assessing the “multi-factorial” objective part of the test for DoL without the benefit of the acid test, there will be a temptation to start instead with the subjective part – a binary question of “are they happy”? as a way of excluding from the safeguards altogether those who can be considered as “consenting”, ending the process there.  There must be a temptation as well over time for that to slip into a tick box approach, and to effectively say that there is at least “tacit consent” where there is no overt opposition.  

There is likely to be significant work involved in these assessments, and in guarding against that risk in particular.  Local authorities and ICBs will have to think about how their teams who had been used to working with the simple acid test, now deal with this very different role, and making sure that they are resourced and trained appropriately (mindful of this being a time of huge change in the workforce in ICBs in particular already).  

One area where the decision may have a positive impact on workload (you will take your own view on the impact on the level of safeguards) is perhaps in deprivation of liberty in the community, and especially the streamlined Re X (or COPDOL11) process. Partly given the court’s reluctance to identify a DoL in a private home (in other than the most extreme situations), and partly as the Re X process itself is intended for only the most vanilla cases in which there is no objection to the care, we might say that anything suitable for the streamlined process is now unlikely to be considered a DoL at all. This of course should be reviewed on a case by case basis. (And, perhaps, the compensation claims we see which include the costs of a court authorised DoL for the impact of a care plan which amounts to meeting what was the Cheshire West acid test, often in a specially purchased / adapted property, ought to be reviewed).  

Elsewhere, though, we anticipate that there will be a great deal more uncertainty.  Instead of the Court of Protection being mostly busy with cases thought to be a DoL which just needs authorisation, as has been the case since Cheshire West, it is likely to be busy dealing with cases about how we draw the lines to define it, as it was before that judgment.  

Setting aside the COPDOL11 process, the guts of the Court of Protection caseload in health & welfare is the s21A challenges to DoLS authorisations, which are triggered, typically, by P’s objection and so are unlikely to be reduced by this judgment.  The rest is mostly s16 applications about substantive issues about the health and care decisions, which won’t be reduced.  Overall, there might be more litigation, and it might be more complex than before.

It remains to be seen whether the government still has the appetite to continue with implementation of the Liberty Protection Safeguards, having promised a consultation later this year, in light of this judgment. Either way, it would be very useful to get an updated Code of Practice for England & Wales (while the AGNI completes the Code for Northern Ireland) for both the MCA and the DoLS system as soon as possible, to give some more authoritative practical guidance. 

Now what? 

For now, there is no further appeal, and no grace period for implementation.  The Cheshire West judgment is explicitly overruled with immediate effect, and this Supreme Court decision now sets out the law on deprivation of liberty.  

No doubt plenty more case law will follow trying to clarify all the uncertainties that this has (re)introduced. We can expect this to be contradictory and inconsistent for a while, taking some time for a clearer picture to emerge, with first instance judges likely to be able to cherry-pick parts of the wide-ranging AGNI decision to reach their own conclusions.  

For those who always thought that Cheshire West had gone too far, this might be a chance to draw the line nearer the right place in defining a DoL, but we have turned the clock back to the uncertainty before the acid test, and it will take some to find our way forward again from here.  

In the meantime, it will be important to:

• Review the structure and resource and expertise of the professional teams working in health and social care, as this judgment will change the scope of their roles significantly.
• Update policies.
• Update awareness, and training. And appreciate that this will need to be tiered, with different concerns for different roles in the organisation – from senior management needing to think about the shape and role of teams, to professionals thinking about how to assess and document whether someone is “happy”, and what this means.  
• Benchmarking will be important – with so much uncertainty again now, some inconsistency is inevitable, both within and between organisations, and best practice will emerge more quickly the more this is shared.   
• Review afresh work being done on cases that were considered to be a DoL that might no longer be.
• Show appropriate urgency – but don’t panic! Not everything can be done at once or immediately.  We will need to prioritise.  
• The consequences of this will unfold over months and years, and (we hope) some guidance from an updated Code of Practice will be available soon, and it will be important to keep on top of the emerging case law, with regular legal updates.