New Liaison and Diversion

In the overall scheme of UK Government spending, newly allocated money for mental health Liaison and Diversion (L&D) schemes in police custody is not that much, but we can agree that £25 million is still a substantial amount of money that we don’t want to misuse.  Bearing in mind that this is money allocated for just ten pilot sites, we can predict the overall spend for L&D schemes will be way more than double this figure if they were extended to all forty-three police forces; and bear in mind that the pilot sites do not necessarily cover whole police force areas.  The scheme within West Midlands Police is for Coventry only, the scheme in the Metropolitan Police is for Bethnal Green.

I really do welcome this investment and the fact that we’re actually going to do something! – I have said for years we need to do much better in this area of offender health, to improve the outcomes we get.  But I also hope it’s not being negative to say that I have questions and concerns.  I voice them, because I have actually worked in areas with L&D services, been involved in setting one up and watched them fall flat on their faces for a variety of reasons at a cost of tens of thoustands of pounds and these questions weren’t addressed in the overall design.

So I asked the question on Twitter yesterday: “Liaison and Diversion: what, precisely, are we trying to achieve?”  I ask the question, because no-one seems to be precisely articulating it – plenty of broad-brush statements that no reasonable person could objective to about “better working” and “closer partnerships” and “better outcomes for individuals”, etc., but I remain full of particular questions about what this actually means in the real world amidst reservations I have from other L&D schemes I’ve seen, that I’ve been part of  and that I’ve read about.

I offer three things that we would all hope to see improved —

1. Appropriate identification of all those who may need MH assessment / referral.
2. Far better sharing of information between police / NHS.
3. Improved CJ decision-making – and not just about diverting people from justice.

We know that the police under-identify people in custody who may have a mental health problem.  Most research I’ve seen suggests we identify approximately 15% of people in whom we suspect a mental health problem.  Notably, I’ve seen information sharing between police and NHS which suggests that as many as 50% of people being arrested could be (currently or previously) known to a local mental health trust.  So L&D services staffed by nurses with access to records, sorts out point (1).

We know that information is under-shared between the police and the NHS, amidst all those traditional fears of violating medically confidential information and NHS Caldicott Guardians warnings about the sharing of it.  What is often missed in these debates when it relates to these kinds of MH/CJ schemes, is that you can share anything at all with the permission of the person to whom it relates – so why not ask them, especially if it may contribute to a less serious CJ response than would otherwise be the case.  Furthermore, you can share a lot of things under the Data Protection Act 1998 – yes, even where it is medical information – if it directly relates to the legal decisions the police must make under criminal law with regard to an investigation.  Section 29 and schedules 2 and 3 of the DPA ’98 refer and everyone connected to these debates should read them – links to them all are in my post on the Data Protection Act.

And let’s be frank: it is often the health information that should directly determine the criminal justice decision.  If you want to see an example of that in play, read one of my really early blogs on diversion.  That could take care of point (2).


This then brings us to the point of it all, point (3) – once you’ve screened, assessed and researched someone in terms of their mental health background, the police have to make a legal decision of one kind or another.  The person who came into custody was arrested further to an allegation of offending in the majority of cases (although L&D services will also see other categories of people in custody).  It is ultimately for the police to decide what happens with someone after an offence is alleged and having taken account of L&D professionals from MH, it will be one of three things —

  1. Some under arrest is fully and immediately diverted from justice – mental health pathways kick in for that person and the investigation is brought to an end.  This will probably mean the person is admitted to hospital and that the offence and overall background is non-serious and the argument would be that there is little public interest in pursuing a criminal prosecution in those particular circumstances.
  2. Someone under arrest is diverted from justice whilst being bailed pending further information – health pathways commence, probably admission, but the investigation is paused whilst more becomes known during assessment.  Something in the incident or background means it is not immediately seen as trivial but which can be held in abeyance whilst health issues are prioritised.
  3. Someone under arrest is still prosecuted into the justice system – notwithstanding that they are believed to have a mental health problem and any referral transfer into the mental health system occurs alongside the operation of the criminal justice system.  This could include the courts using Part III of the Mental Health Act where there is a more serious risk to health and public safety that needs managing.

How do we hope to see the quality of the decisions improved? — we don’t know a huge amount about the outcomes we get in custody at the moment and in the future we know that a research organisation will sit down to evaluate these £25m schemes to decide whether they should be rolled out nationally by 2017.  So the first thing to say is we’re not entirely sure what the bench-mark is for current practice.  Here is my understanding of where we start in an area where there is no current L&D provision, based upon research I’ve done of 10,000 police custody records in Birmingham –

  • 10,000 custody records were examined from two different busy custody suites.
  • 1,076 raised the ‘mental health’ question and were assessed by the FME
  • 512 were assessed by another medical professional after the FME – either CPN or AMHP-led MHA assessment.
  • (The remaining 564 of the 1,076 were deemed fit for investigation.)
  • 415 of those 512 were not in need of hospital admission on that day.
  • (The 415 were ‘managed’ in custody almost as if no mental health problems existed at all.  Appropriate adults were obtained where needed, but the overall CJ outcome was consistent with ‘normal’ CJ outcomes for other offenders.)
  • 97 were in need of hospital admission following their assessment.
  • 9 were in need of admission under s3 MHA.
  • 12 were in need of admission on a voluntary basis.
  • 76 were in need of admission under s2 MHA.
  • 97 were not prosecuted.
  • 97 were not bailed pending the outcome of their MH assessment or treatment.

It raises a load of questions that I’m interested in knowing the answers to, but I don’t know whether anyone is looking at this in appropriate detail —

  1. How many of who were not “sectionable” when assessed in police custody were referred to MH services for an outpatient appointment or follow-up by a Community Mental Health Team and how many attended and engaged?
  2. How many of those who were “sectioned” were found to have a serious mental health condition and detained for an appropriate course of treatment, for example as an informal patient or under section 3 MHA?
  3. How many of those who were “sectioned” were found to be have a mental health condition, but not serious enough to warrant ongoing treatment in hospital and were discharged to a community mental health team with which they engaged?
  4. How many of those who were “sectioned” were found to be suffering from no mental health condition whatsoever?
  5. What does all of *THAT* mean for criminal justice outcomes and re-offending rates?


If we knew the answers to questions 1-4, above, we could then reflect upon criminal justice outcomes and re-offending.  If we put that all together we would know far more about the 97 people who were admitted to hospital after assessment on the day of their arrest.  A minority of them would have been seriously unwell and further detained after initial admission; a minority would not have had any kind of mental health condition.  A bulk of people would be found to have a mental health condition, but not serious enough to require ongoing, long-term hospitalization.

  • So if we learned that one of the 97 was a very ill man, never been in trouble with the police before that he was sectioned after arrest for shoplifting a moderate amount of goods, and so ill that he ended up detained under s3 MHA for several months – we’d probably agree no particular need for a prosecution or even a caution.
  • Equally, if we learned after one man had assaulted someone before being admitted and fully assessed in hospital and it was determined after his illegal drugs had worn off that he actually wasn’t unwell at all and had been referred to his GP for follow up on his drugs usage – we may wonder the justice of things if we found out that he was never sanctioned for the assault.

What does all of that mean for criminal liability?  Well, it depends what your theory (and your politics) is!

Some people argue that if a person with a mental health problem has become unwell – let’s suggest that they have relapsed after ceasing to take medication prescribed for their condition – it is their illness that gives rise to the offending so the priority should be to treat the illness and thus reduce re-offending.  This assumes a relationship between mental distress and crime that some researchers argue is simply not there yet we can see in the language used to describe these new schemes that this presumption, very far from being evidenced by research, is believed to be true.  We heard the schemes were marketed as “mental health nurses to help police reduce re-offending.”  I vividly recall a few years ago having to legally review the detention of a man in custody being investigated for assault and thought to be mentally unwell.  As I explained the delay in him being dealt with partly arose from delays in getting the mental health team to the station he said something quite succinct in the vernacular form, “For fuck’s sake you don’t need a mental health team anyway … I punched him because he’s an arsehole not because I’ve got schizophrenia.”  Fair enough.  He was duly charged with assault with a further referral going to the mental health team responsible for his care.

The majority of people who come into police custody whilst mentally distressed are not acutely affected sufficiently to require inpatient hospital admission and they have not offended enough to justify imprisonment.  Therefore whatever health referral / assessment is perceived to be required can occur anyway.  We know that the police identify between 10-15% of those coming through custody and that with mental health professionals in custody having access to NHS records, this number could rise to between 30-50%.  In other words: the police are currently spotting about 1/3rd of the mental health related demand in detainees.  It immediately begs the question of whether the NHS is in a position to handle referrals for less acutely ill detainees that are triple the current rate.  Here’s hoping that the £25m is partly to shore up other services like drug / alcohol or personality disorder services, not just provide the staff and structures for police stations.  A potential tripling of demand! …


So if we triple, roughly, the number of people in police custody who are identified as requiring follow-up NHS care, where does that demand go?  We know that 83% of people with mental health problems are cared for by their GP, but many of those in police custody require follow-up by secondary care mental health services, ie, the mental health trust for that area.

We know that beds are currently in very short supply and that the CQC are actually starting to push in areas that have decommissioned too many beds and can no longer meet levels of demand.  We are also starting to hear how equivalent pressures are affecting community mental health teams because of the increased push to provide community care, in light of the bed crisis.

So Liaison and Diversion: very welcome, but it needs to be widely known that the “Criminal suspects suffering mental distress” group of in custody is SEVEN times the size of the “section 136 Mental Health Act” group and we know how difficult it is to ensure effective 136 provision.  It also needs to be known, that this will increase demand for NHS services that we know are already saying they struggle and it needs to be known precisely what we are asking them to do better or differently and with what objective in mind.

How will we know if this “works”?



IMG_0053IMG_0052Winner of the President’s Medal from
the Royal College of Psychiatrists.

Winner of the Mind Digital Media Award.

5 thoughts on “New Liaison and Diversion

  1. Michael,
    Excellent blog, glad to see you support the L&D principle, I certainly do. Your blog raises numerous issues, but two key ones for me:
    1. You touch briefly on the role of CPS. One thing I find as a PCC is a real disconnect between CJ initiatives and CPS. Quite frankly, it seems people (or Govt) forget their relevance. Neighbourhood Justice Panels is another good example. Your two examples are a good place to start. The shoplifter example- should that person be criminalised for a one off offence caused by an illness? Arguably not. If that example involved a person of tender age, attempts would be made to avoid criminalisation. If we had the right processes in place (to ensure people who are mis-diagnosed or feigning do not avoid the CJ system) surely people who commit offences through mental illness could be totally diverted, not convicted then diverted?
    2. Data Collection – Your questions 1-5 highlight something I have also spoken and written about. We simply do not know enough of what happens to people who are dealt with (or refuse to engage) in the community, after detention. It totally skews outcomes, especially in relation to reducing re-offending and Section 136. We need to get better at it.
    Happy New Year, and thanks for what you do.
    Martyn Underhill
    Dorset Police and Crime Commissioner

  2. I am interested in the identified pilot Trusts…historically they have many problems and I just hope this is not a case of shutting door after horse has bolted. Plainly people detained by State are the responsibility of State but the State must also be given the right tools for the job! We keep hearing what a success Olympics were and “there was even an underspend” Nothing mentioned on OVERSPEND enabling effective services. How will cutting MH Beds and not fully implementing services like dual diagnosis enable full support. …..but its a tiny step along a very long road!
    Happy New Year and thank you for opportunity to comment!

  3. Michael,
    great blog as usual.

    I’m interested in your thoughts on the stage prior to the courts. A huge number of the MH cases will be taking adult cautions to avoid the stress of going to court where they might be offered help via the L&D process. Can you see a way in which custody sergeants/inspectors could pass on people to MH teams where it might be more appropriate instead of just issuing an adult caution ? Any room in the law for this type of process ?

    Chris Crook
    Carers Ambassador – Carers UK
    Trustee – Action for Carers Surrey

  4. thank you in particular for your extensive piece of research into 10,000 custody records in Birmingham, which certainly gives me a better understanding of the situation

  5. There are a few things wrong with this blogpost.

    The pilots and rollout (courtesy of NHS England, at some point, maybe) are the result of the Bradley report, and are based on work that has been going on for over twenty years. It is not a new thing, nor is it a simple phone call to a MH Trust, it is partly aimed at those who have not been identified or, have not sought help before. It is not about a custody sergeant saying “I think this one’s a nutter!”
    There are a number of other models and the pilots are intended to determine the most effective one prior to rollout (allowing for local needs). Again, it isn’t necessarily about you guys identifying someone who may be unwell.

    I agree that there are issues with lack of provision when people are referred to MH services (which is standard across the board in MH now), which you have not focussed on, aiming more for non-concordance.
    For example, someone is referred by L&D to a psychiatrist who agrees that a MH unit would be the best place, but it is decided that the threshold for Section isn’t met and there are no informal beds available (let’s not go into whether a bed should be secured for a Section2/3 to be legal). So, the person is remanded by Magistrates. The psychiatrist attempts to ensure the person is carefully monitored in prison. It doesn’t happen. That person dies by their own hand in prison.
    It has happened and will continue to do so.
    L&D has not failed in such cases, the lack of provision has.

    Yet you seem to have an issue with L&D itself. The rollout was supposed to be in 2017, but has been delayed and included in Dave’s “new money for MH” prior to the EU referendum and might happen in 2020. Bear in mind that the Bradley report was published in 2009.

    So, yes, lack of provision is an issue, but that does not make the concept of L&D an automatic fail unless we lie down and allow the lack of beds, specialist and community support to continue.

    People who care about MH and CJ need to come together and make a case for decent resources on both fronts.

Comments are closed.