Confusion or Conflation?

The All Party Parliamentary Group on Women in the Penal System has published a briefing on “Women’s Health and Wellbeing in Prisons”.  This post is about one specific claim only, within that briefing – that prisons are regularly but inappropriately being used a “place of safety” (PoS) and this is an unintended consequence of a change of policy to prevent women languishing in police cells (presumably, there as a place of safety). The briefing itself is short (10 pages) and this one point is only mentioned in brief, relying in turn on a report from Her Majesty’s Inspector of Prisons and Probation about an inspection which obviously must have wider implications across the prison system.

Before I get in to this: of course, prisons are over-relied upon to accommodate highly vulnerable people in circumstances where they should be able to access mental health beds in a much more timely way.  Additionally, but separately, it can be said there are difficulties securing access to MHA beds for those who require admission to hospital from police custody, but whilst being in full support of what the report is attempting to highlight (over-reliance upon the criminal justice system, about which I’ve banged on for what feels like decades), I want to argue that the cause of this over-reliance on prisons, isn’t quite right.  It’s a consequence of there being insufficient acute mental health beds, accessible in a timely way.

They key issue or potential confusion here, is what we mean when we use the term “Place of Safety”  – and neither the APPG briefing nor the HMIP report on Low Newton, define what they mean.    As they are writing about women with acute mental health problems and given the emphasis today on prisons as a PoS being a consequence of policy on police cells, it’s fair to wonder (or assume?) this may be the context: the Mental Health Act 1983.  It is just possible, however, that the phrase is being used as a non-legal term, to represent a place someone can potentially just be safeguarded and kept safe for a relevant period of time.

Here’s my main point: either way, I would argue this conclusion is not safe and I will set out why.


Few people realise, there are two defintions of a PoS in the Mental Health Act 1983 – the most common one is in Part X, outlined in 135(6) MHA and it relates to the use of police stations as one of several locations which could be considered for anyone who is detained by the police under s135 or s136 MHA – detention outside the criminal law, purely to safeguard and then assess someone at risk of harm.  It is absolutely correct to observe, things moved on in 2017 when the Policing and Crime Act changed the law, to restrict use of police stations as a Place of Safety.  This has been one of several things which means the use of police cells for woman, men and children has reduced from way over 10,000 instances a year, to under 100 a year (and at a time when use of the power has risen from just over 18,000 to well over 30,000).  Prisons are not mentioned in this definition: you cannot just use a prison for s135/136 either previously, or as a consequence of this change of law.

The PoS definition one no-one’s heard about is section 55 MHA – this definition does include prisons – it is only of relevance to the criminal courts when they are considering or making various kinds of criminal court order under Part III MHA. Use of what I will therefore call a “Part III” PoS cannot be instigated by anyone other than a criminal judge – not even mental health professionals can make this happen.  Only where a person has been through the criminal justice system to some degree and where a judge is contemplating use of a hospital order or an interim hospital order, they may choose to send the defendant to prison “as a Place of Safety” where a hopsital bed for that person is not immediately available to allow those orders to be immediately made.

So the police cannot instigate use of a prison as a Part III PoS; and the courts would not be involved in a decision about use of a Part X PoS (where often, the person detained hasn’t offended at all, never mind been prosecuted).  Any increased use of a Part III PoS cannot be directly connected to reduction in use of police stations as a Part X PoS.

The two “Place of Safety” concepts in the MHA are for entirely distinct, non-overlapping legal circumstances and cannot flow directly from each other.


What about the use of prisons as simply somewhere to keep someone safe (where it would be help if we didn’t use the term “place of safety” with all the risk that we confuse what is being referred to?!) — we know that where the police / CPS are intending to charge someone with an offence and grant bail, one basis for denial of bail is that it is “necessary for his [sic] own protection” (s38 PACE).  Once a charge and remanded defendant reaches the Magistrates or Crown Courts, the question of bail in criminal proceedings again arises and the courts must consider afresh, any denial of bail.  You won’t have to wander for very long around a police station before you’ll bump in to an officer who will have a story involving frustration of where police denied bail (for whatever reason) only for the courts to grant it the following morning (usually uncovered when the person committed a further offence or breached bail conditions to protect victims, but I digress).

So a defendant denied bail by the police may then find they are denied pre-trial bail by the Magistrates or Crown Courts and one justification for this denial can be the persons own protection (which we can immediately realise may be due to acute mental health problems, not otherwise managed via a more therapeutic and acceptable route). This is the key point, on this version of what they may mean by “place of safety”: if a defendant charged with an offence is remanded to prison for their own protection, this is not use of prison “as a place of safety” under the Mental Health Act 1983 and it does not happen as an alternative to using police custody, ever.  Courts either remand defendants to prison, or they bail them.

Remanding of un-tried, un-convicted defendants is not covered by Part III MHA – it’s a part of the Bail Act 1976  It is therefore not a consequence – intended or otherwise – of any change to the “policy” (which was, in fact, a change of law) on use of police cells as a “place of safety” under the MHA.  Police stations are never used as somewhere to just keep someone safe … or if they ever are, it’s unlawful!  You’re either under arrest for an offence and being investigated; or you’re one of the one hundred people a year under the MHA as a Part X PoS; or you’re there for other legal reasons like being arrested in warrant, under immigration law, etc..


The Low Newton inspection and the APPG report are not clear what precisely they mean when they say “Place of Safety” and it’s impossible to guess, but I submit it’s one of the two interpretations covered above.  Neither of them can be correct, so in one sense it doesn’t matter which of them they mean, but given “place of safety” is a legal term and these are inspectorate and formal reports, I’d expect clarification of any use of that term in a non-legal way.  Perhaps I missed it …

Use of prison as a place of safety (under either circumstance outlined above) is matter for a criminal judges, pre-conviction or post-conviction – use of a police station as a place of safety is a matter for police officers in circumstances where no offence has been committed (or where it’s set aside to prioritise the mental health of someone who is acutely unwell, usually because the offence was very minor and wouldn’t see the inside of a prison anyway).  These two things do not overlap

  • One is about people who are not in the criminal justice system at all, merely in contact with the police as part of their expanding role as an emergency healthcare service, along side the NHS system;
  • The other is about criminal prosecution and post-trial processes where the police have long since ceased to be decision-makers of any kind.

But Part III X Place of Safety concepts are almost invisible – we rarely hear of them being used and during training presentations I’ve often used the trick question “What is the maximum time someone can be kept in a police station as a Place of Safety?”  I normally get the answer “36hrs” because this is the maximum for Part X.  However, police stations can be used as a Part X place of safety as well as prisons and the maximum time is 28 days — that said, I’ve only ever heard of it happening once in 24yrs (and it confused the life out of the custody officer who rang up to ask “what the hell is this? – how can this be right?!  28 DAY!!!”

The point about over-use of prisons is correct and should be flagged and questioned: my point here is merely to argue, it’s not the unintended consequence it’s claimed to be.  It’s far more likely just the consequence of not having enough acute, inpatient mental health beds for women who need them – same applies to children and to men, sadly.


After writing the post, I came across another briefing note from the Howard League for Penal Reform, which might unlock some of the confusion so I thought I’d add to the post, at the risk of unmanageable length!

It states (p4) that “prisons are not places of safety for the mentally unwell”.  Could still mean either of the two things, above – but it goes on, emphasising the place of safety concept from Part X, referenced above and it mentions s136 before adding —

“Unlike police cells, prison have never been designated in law as ‘places of safety’ and for very good reason.  THey are not equipped to care for people with significant mental or physical health issues.  Prison officers and prison medical staff do not have powers under the MHA 1983 to treat a person suffering from a mental disorder against their will.  Not do they routinely receive specialist training for dealing with the mentally ill.  So managing extremely vulnerable prisons in the harsh surrounds of a prison can be very difficult and distressing, both for the individual and those caring for them (Pattison, 2016).  Indeed, being detained in the noisy, coercive and unpredictable surroundings of a prison is highly likely to exacerbate a mental illness.”

First things first, you could leave out the first sentence of that quote, replace the ‘prisons’ for ‘police’ and it would still make sense.  But the first sentence is important: because it’s flatly wrong, in law!”  As in the original post, prisons have been designated in law as a ‘place of safety’, in the specific contexts of s37, 38 and 45A of the Mental Health Act.  Now, I agree this is nothing at all to do with the s136 Place of Safety idea under Part X, it is the Part III idea no-one seems to know about!

But it perhaps shows the source of the confusion for the incorrect legal claim and indicates we are talking more about the second scenario, which the Howard League calls “remand for protection”.  This means that one other observation of mine by reading this further report is key to unlocking what they mean by this.


The other briefing note is also published with badging from the APPG on Women in the Penal System

“Research suggests that, in order to avoid police cells being used as places of safety, instead of using their s136 MHA powers, the police are simply charging individuals with low level offences (such as a public order offence, arising from the person’s distressed behaviour) and passing the challenge of addressing and individual’s mental health crisis to the courts — Pattison, 2016.”

I can’t say much more about this research, because the link provided in the bibliography from the Howard League, no longer works.  All I can say generally, is that it’s my instinct is wonder about the research methods behind that claim and to point out it’s entirely at odds with the oft-heard mental health systems claim that s136 MHA is wildly over-used – something which is also attributed to police policy and the legal change from 2017.  But here’s a key question: how can any research completed before 2016 tells us anything about the impact of a policy change that didn’t really kick in until 2018 (the MHA legal changes took effect in December of 2017).

My own view: this whole discussion is confused, legally uninformed and I doubt the research behind any of it would stand up.  I’m happy to be convinced otherwise, if someone has access to the deleted document which appears to be the source of the thinking here.

Winner of the President’s Medal,
the Royal College of Psychiatrists.

Winner of the Mind Digital Media Award


All views expressed are my own – they do not represent the views of any organisation.
(c) Michael Brown, 2022

I try to keep this blog up to date, but inevitably over time, amendments to the law as well as court rulings and other findings from inquests and complaints processes mean it is difficult to ensure all the articles and pages remain current.  Please ensure you check all legal issues in particular and take appropriate professional advice where necessary.

Government legislation website –