I think this is unlikely. Although an inverse relationship between hospital beds and prisoners has occasionally been observed - on a per capita basis, world wide, richer countries institutionalise more people in jails and hospitals than poor countries. In most placed more mentally ill people in jail just reflects the growing prison population - see The Penrose hypothesis in 2004: patient and prisoner numbers are positively correlated in low-and-middle income countries but are unrelated in high-income countries.

Large MM, Nielssen O.

Psychol Psychother. 2009 Mar;82(Pt 1):113-9. 

Thanks Matthew very interesting. Anecdotally.....there are a number of people, often with complex need who seem to want some kind of institutional care, who will re-offend immediately after release from prison. Previously this might have been a need met by the Large Psychiatric Hospital. A lot of people who meet the definition of long-term mentally ill with complex needs are hidden in private sector OATs placements in England. These placements are extremely expensive. Those incarcerated (locked when service user is informally admitted) in such establishments often have a history of offending as well. 

For me, and I know you know this Charlie, in order to understand this phenomena I think we do need to strip the whole thing back. We need to know/ understand, or advertise more widely, the role of both mental health services and prisons in our society in order to understand the quantatitive expressions of what's communicated from within  them and the meaning of the costs associated with them. Prisons quite literally hold a captive audience and as we know the more you look for something, the more you are likely to find it...or convince yourself you have found it. Many would question the whole paradigm of Psychiatry as a means of understanding and quantifying distress/ illness, those sticky labels that become harder to remove...and harder to live up to than the social defecits which the prisons and mental helath units encapsulate, nurture and hide from public view . I agree with Mathew it is more complex than it at first seems and I agree with you, as the recent paper presented to the house of commons showed clearly, the cost of the purchaser/ provider split, the move into the community and the loss of instuitutions (which I, like so many others working in them, was pleased to see!) has been hidden from scrutiny, the minimum figure they suggest is that this new infrastructure of purchasing and the move to other forms of commissioning frameworks involving "other" providers accounts for at least fifteen % if the NHS costs and the report strongly suggests they believe more because the figures are not forthcoming in the way they should be ( as you no doubt know writers such as calum paton would argue aspects of this). So yes more prisoners "spotted" with mental health problems, and yes increased costs having moved away from the institutions into fee charging / contracted "other" providors! All very ironic really. So the short answer would be that I would question any "evidence" found in the UK, mainly due to the fact that there are far more complex issues at work here and just because something appears to correllate, doesn't mean, when the context and complexity of the issues are taken into account, that it does meaningfully correlate. Thanks for generating the debate.

HI Karl

I would be most interested in a reference for the House of Commons paper that you cite? Many thanks for adding to the discussion. I understand that Mental Health Strategies estimated that 25-50% of mental health budget costs were taken up with OATs. This, for a group, that seemingly are to difficult to provide a meaningful service for locally? (not that they get one out of area). The same could be said for people in the CJ system who also present with complex disorders. It's interesting to note that some providers, for example, DASH in Yorkshire, have been rebuilding in-patient rehabilitation facilities on the site of one of the old hospitals (St Catherines).

De-insitutionalising mental health was definitely a good move. Whether community mental health services have ever been adequately funded to "take up the slack" when we stopped keeping large numbers of people in psychiatric hospitals is another matter...

With respect to you Matthew, I think your comments above may be underplaying the importance of this issue. The inverse relationship between psychiatric beds and prison inmates suffering mental illness is not "occasional", rather it has been observed in many jurisdictions and is very robust.

http://journals.psychiatryonline.org/article.aspx?volume=166&page=103

.

Here are the UK stats for '82 til '97.

http://bjp.rcpsych.org/content/176/4/332/F1.expansion.html

.

That's a pretty clear inverse relationship, n'est pas?

Penrose was right to point out that there is an inverse relationship between hospital beds and prison beds. However, this relationship may not be a directly causal one.

The intuitively appealing and hopelessly simplistic explanation for the relationship  is that more people with mental illness in the community = significantly increasing crime rates.

An alternative explanation is that less investment in mental health services =  more people who exhibit deviant behaviour being incarcerated, because there is no where else for them to go.

So, perhaps it's just that more police (and more police powers/prohibitory laws/mandatory sentencing) = more incarcerations, and less mental health services = more disadvantaged people running foul of the criminal justice systemn?

I agree, Paul, that a casual relationship is 'hopelessly simplistic'. The funding was never targeted to community mental health services when hospitals closed, it was often used to prop up the acute sector. The promises that ex-residents would have homes for life was reneged on as well. Policy is an issue too. We had the National Service Framework for Mental Health from 1999-2009 and rehabilitation is barely mentioned in it. As far as policy-makers were concerned the hospitals were closed and that was it (see attached graph of dates of closure in relation to policy). I also agree with you on the 'nowhere else to go to' explanation which, despite the huge costs, has led to massive expansion of OATs in England. All grist to the mill for the private sector which is barely regulated.  

Having worked in psychiatric hospitals (in the US) during the deinstitutionalization my observation was that the community support never materialized as it was promised.  Many, many people where placed in the community and are doing well.  But a number have ended up in prison.  One thing I have observed is that a good number of psychologists and psychiatrists who worked in those hospitals are now working in prison (with similar patients and problems)

I tend to disagree with Matthew.  Living in NYC since 1979 I have seen the increase in the homeless mentally rise with the decrease of facilities that provide long-term hospitalization.  Many psychiatrically ill patients self-medicate with street drugs, get arrested and end up in jail.  The number of mentally ill prisoners has increased over the past few decades and I believe that is because the community mental health service were never put in place after the emptying of state hospitals in the late 1970's and early 1980's, at least on the east Coast.

De-institutionalization in Canada has put more pressure on the criminal and penal systems. We have had several episodes of police killing people with mental illness in Toronto as they did not know what else to do. Where police have training there are less such incidents, such as is the case in Chatham Ontario where they have a 'help' team of trained officers. Special courts have been developed to divert these cases in the hope they can get appropriate supports in place. We remain short of adequate housing and community supports for the population in need. Many are now numbered among the homeless that seem to grow in number every year. Both prisons and the street have become options but they are not helpful options. What is needed is more safe supported housing and opportunities to be as engaged in work and life as symptoms allow. This of course also requires a reduction of stigma and the fear that the mentally ill are 'dangerous' . In fact people with mental illness are much more likely to be victims that dangerous offenders, with the reverse being the case for the rest of society.

This is a rich and fascinating conversation with lots of ends to explore. Some of which I'd like to address:

1. Charlie Brooker mentioned that the needs of institutionalised people may drive them to prisons when the MH institutions close. It's an astute observation. One of the characteristics of institutionalised lifestyle is that it's really simple. You don't have to make many decisions on your own steam. You just do as you're told. This may make life more livable for people who can no longer (if ever) handle the complexity of existence. 

2. Whilst I join the overwhelming chorus of those people who believe in deinstitutionalisation, I do concede that for some people it's better than the alternative: living outside the institution on what's often paltry resources and in miserable conditions. There's a very good research article on this by Baltazar et al. (2013).

3. The ideal direction, I believe is to have MH Facilities that are genuinely recovery-focused, and are geared to help people slowly learn to cope with real life. And that means giving people increased responsibility. 

4. One of the problems with the argument above is that both imprisonment and severe mental illness (of the kind that will bith have a person sectioned in a facility for a length of time and in a sufficient state of confusion where they have difficulty coping outside) is that both phenomena are rare. What this means is that both statistical sets represent tails of the normal curves of society. There are statisticians who are very good at analysing these data sets using Bayesian methods, but standard correlations are going to be very difficult to establish, and might be meaningless even if they are. I would expect that more direct causal influences such as the temperament and ideology of the local authorities (the District Attorney, local Police Chief etc.) will influence who ends up in prison far more than the availability of beds in MHf's. 

Baltazar, Ana, Kapp, Silke, Tugny, Augustin, & Furtado, Juarez. (2013). Spaces for differences: dwelling after deinstitutionalization. Facilities, 31(9/10), 407-417.

The work of my colleague Jan Cassidy and others revealed a very high prevalence of mental disorders in prisoners, with especially high rates among women in custody. My own work in prisons was back in the 1990s, but I attended a 2014 meeting on health in the justice system and depressingly little has changed yet, except a bit more access to drug treatment for opiate users. However, more recently than my work with adults I did a study with young offenders where 37/37 young people assessed had untreated mental illness, often in combination with mild learning disabilites and hazardous substance use. There was co-occurrence in England of decreasing inpatient beds for mental illness and  rising numbers of adults in prison (circa 85,000 now - it was 37,000 in my time !). However the age profile confirms my old observations that most prisoners in their early 20s had no prior NHS care (although it was striking that many had sought some sort of support from the Voluntary sector) - in general the inpatient population is a little older than the prisoners. In some NHS services like drug dependence the majority of patients had prior experience of custody (90% in one service) but the number of inpatient beds for detox is MINISCULE now. So, overall, we have a load of young adults with mental illness (especially females) in custody, of whom a minority will later have an NHS admission but most of whom never see a psychiatric ward.

Thanks, Woody, interesting observations. Our prevalence study of mental health disorders in a probation service showed that 40% had a diagnosable mental health problem and, of these, only 47% were known to formal mental health services. However, alcohol problems were seven times more prevalent than drug problems (measured using DAST and AUDIT). In probation, privatisation looms, the CRCs should be announced soon, some will say this is an opportunity others that disaster is on the horizon. One change that will occur with 'Transforming Rehabilitation' is that prisoners serving sentences of 12 months or less, unlike the past, will receive a community order. There is an urgent need to increase the use of mental health treatment requirements and St Andrews are running an innovative scheme in this regard.  

Some comments on the various threads in the responses so far. De-emphasizing institutionalization would have been preferable to deinstitutionalization. There are still many people with chronic and severe mental illness who are not benefiting from the "new world order” of mental health. They need longer term care options which are scare and often viewed as inconsistent with "recovery." Most are homeless, some incarcerated, and others must run the gamut of serial acute hospitalizations that just shelter them for brief involuntary stays or in some cases mini-institutionalizations in settings not designed for stays of several months or more. The chorus of woe about the incidence of mental illness in prisons tends to be myopic in view of the problem. In the US at least, there is a strong jail diversion effort ranging from mental health crisis training for police to mental health courts for offenders found to be mentally ill, These measures are largely ignored in the "prisons are the new psychiatric hospitals" mantra. More to the point is the naiveté displayed about how inmates come to be "diagnosed" with mental illness in corrections. The role of malingering and manipulation is overlooked. Mental illness is easily faked and can lead to special housing and other privileges. The fact that anybody who seeks let alone receives psychotropic meds in jails is ipso facto mentally ill is apparently irrelevant. Then there are the prison systems (e.g., PA Dept. of Corrections in the US) that establish their own criteria for determining which of their prisoners has a psychiatric diagnosis. I don't impugn the validity of the studies cited but there are significant limitations on generalizing their findings to correctional systems world-wide. Concern should first be given to those incarcerated inappropriately as a consequence of behavior related to mental illness as opposed to losing sight of them in the rush to save criminals who may coincidentally have a psychiatric disorder or only are reputed to be so.

Hi Tony,

The surge of deinstitutionalization in the 1960’s eliminated state psychiatric hospitals and contributed to the movement of federally funded mental health facilities. The federal funding was proposed to last for a period of time and then transfer continued funding responsibility to the communities for support (Browning, Van Hasselt, Tucker, & Vecchi, 2011) to no avail. According to a 2006 Bureau of Justice Statistics (BJS) report, mentally ill adults incarcerated in local, state and federal facilities were estimated at 1,264,300. These statistics have risen compared to literature documented in 1999 by Ditton surmising 283,600 incarcerated individuals were diagnosed with mental illness (Sarteschi, 2013). The most common mental illnesses found in the criminal justice system include depression, schizophrenia, anxiety, bipolar and substance abuse disorders (Brandt, 2012).

The demand in forensic mental health services as seen in the United States, Canada, and internationally describes statistics doubling annually, upwardly of 110% increase, in over a decade. The statistics reflect individuals requiring mental health housing, institutionalization and resource support (Jansman-Hart et al., 2011). The interface between community safety and mentally ill offenders is poorly addressed by intensified sentences and reinventing pre-emptive imprisonment (Sullivan & Mullen, 2006).

Resources:

Brandt, A. (2012). Treatment of persons with mental illness in the criminal justice system: A literature review. Journal of Offender Rehabilitation, 51, 154-558. http://dx.doi.org/10.1080/10509674.2012.693902

Browning, S. L., Van Hasselt, V. B., Tucker, A. S., & Vecchi, G. M. (2011). Dealing with individuals who have mental illness: The crisis intervention team (CIT) in law enforcement. The British Journal of Forensic Practice, 13(4), 235-243. http://dx.doi.org/10.1108/14636641111189990

Jansman-Hart, E. M., Seto, M. C., Crocker, A. G., Nicholls, T. L., & Cote, G. (2011). International trends in demand for forensic mental health services. The International Journal of Forensic Mental Health, 10(4), 326-336. http://dx.doi.org/10.1080/14999013.2011.625591

Sarteschi, C. M. (2013). Mentally ill offenders involved with the U.S. criminal justice system: A Synthesis. Sage Open. http://dx.doi.org/10.1177/2158244013497029

Sullivan, D. H., & Mullen, P. E. (2006). Forensic mental health. Australian and New Zealand Journal of Psychiatry, 40, 505-507. http://dx.doi.org/10.1111/j.1440-1614.2006.01851.x

Good question, non-straightforward answer. In general, the growth of prisons has had little to do with the closure of hospitals if one thinks of this as a process of 'transinstitutionalisation'. The people who were in the old hospitals and the people who are in our larger prisons are not the same people, in the main. There are problems of an over-representation of people with serious mental illness (SMI) in prisons and jails, but there is little evidence that the proportion of prison inmates who suffer SMI has changed over the period of de-institutionalization. It remains commonly approximately 15%, regardless of when the study has been performed. 

There are certain patterns, however. As we emptied our hospitals, we did not necessarily create service models or capacities for community care to meet the need of people with serious mental illness in the community. This has resulted in increase police involvement with people with SMI, but very little evidence of rising crime rates by people with SMI (see especially the homicide studies, Schanda 2005 in the Lancet summarizes them nicely).

There is evidence though that the characteristics of mental health services have changed, especially institutional mental health services that have remained or have been built in the last 15 years. Rather than criminalisation, what we have seen is a progressive 'forensification' of mental health services: rising resources for forensic mental health teams, increasing forensic inpatient beds and community services. This is widely documented in Europe, Australia and New Zealand [see Priebe and Schanda and colleagues for the European studies on these trends].

I believe though that the US must be seen as a different epidemiological domain than Europe and the Commonwealth countries. The reason for this is that the US has had very particular distortions to de-institutionalization because of the perverse incentives for States to close hospitals because they could access federal money for the patients in the community, and the fact that the US imprison approximate 7 times more citizens than does most of Europe and the Commonwealth. The policies that drove de-insitutionalisation and the policies that drove mass incarceration are very different ones. But if you choose to incarcerate large numbers of people on the fringes of society you can bet there will be an over-representation of people with SMI amongst them. And so the US has found.

Was Penrose right? Penrose said, first, that small hospitals are associated with higher homicide rates. That is not true. He also said that smaller hospitals are associated with bigger prisons. In many countries that is not true. Much of Europe closed hospitals and made no change in their prison numbers, that is true in Canada also. So Penrose's second point is generally not true. That hospitals have shrunk and prisons have exploded is true of the US, but for very different reasons. 

Hi Alexander

Much of what you say above rings true for England. I also understand now why it is the US literature that makes the most claims of the link between hospital closure and increase in prison numbers. Thanks

The Guardian today (16/9/14) says

"Chris Grayling has ordered justice ministry officials to start work on developing a network of specialist mental health centres within prisons in England and Wales....'.

I agree Us and Europe differ greatly.

as such, It might also be interesting to check the Austrian situation, documented by Schanda. its a small country with a stable crime-figure, and where a rise in criminalization of mental health clients was also reported.  the same was documented by Peter Kramp in Denmark. Both agree that the reasons must be multi-factorial, and just a decrease in beds will not be sufficient to explain. As u point out, they stress the importance of service delivery models.

in Belgium, the decrease of beds only started since 2011 and no data is yet available, but it will be interesting to see if here also, we see a rise in mentally ill people entering the justice department, i guess.  so far, figures on NGRI-clients remain stable.

many thanks

You are right, Thomas, the Austrian experience is fascinating. Stable crime rates, stable imprisonment rates, but rapidly rising forensic populations. And the rise commenced abruptly with a change in the Mental Heath legislation [see Schanda et al, 2009, International Journal of Forensic Mental Health Services, http://www.tandfonline.com/doi/full/10.1080/14999010903199290#.VBrtM5RdVWI ] 

I think that the research on this is not the best way to spend money. What we know ias that there are too many people in prison who have mental illness. Many, the majority, have a short sentence. There needs to be diversion at court stage. the more serios crime commmited by the mentally ill needs specialist units. the need for clinics in prison is easily made with the multiple paers on the level of MI in prisoners. Lets look at models of care and their outcomes and not more epedemiological papers telling us more versions of what we know.

Seriously John the const of research is straw man, this research costs nothing. My paper that found that patient and prisoner numbers are positively correlated using international data sets from 2004 cost precisely 0 dollars 0 cents. The real reason not to re-research Penrose  is that it was a fluke, a contingent accident of history in pre war Europe  - there is simply no evidence for such a relationship as Penrose formulated - intact the reverse is true , countries that can afford to build Jails also can and do afford to build psychiatric hospitals. 

Unfortunately there does not need to be closures of hospitals to increase the number of prisoners with mental health.  There is a high number of prisoners who suffer from a mental illness.  Unfortunately the number of prison guards that have training in mental health is very low.

Hospitals and their building vs jails are not the problem as Matthew says . The problem is strategies to manage mentally ill people in the community with models of care that are effective. There is  a dearth of staff who are trained and feel confident in the community work required. Where i work there are some "shining lights" re. skills but they are often in teams 33% staffed so the work load and hence time is limited. The issue is not re money but the time of researchers I think it is better spent on researching models that reduce the number of mentally ill in prison who have remediable conditions. In NSW we know this is around 10-20% of prisoners (1-2000) people per year. They are treated in clinics in prison and in the prison hosptal (correctional patients). This work should be done in the community as most serve less than 6 months.

In line with some of the previous responses,

we are looking at two systems that work independently, but impact directly on each other.

Health is a system focussed on that, and Law is a system focussed on that. As a practising Doctor in Psychiatry, I have seen many cross-overs between people who have committed crimes and have no formal mental illness, but complex mind-frames that have assisted in their crimes, that don't require a mental health ward, but require some legal intervention AND I have seen the persons who are in prison systems who would be much better rehabilitated through mental health services.

there is the cross over ward of Forensic Psychiatry, but this sadly ends up receiving the title of the patients that are in the "too hard" basket of both mental health and legal. There's not enough bed space funded for patients with petty crimes and less serious mental health issues, but could actually be intervened on earlier, and possibly limit the outcome of them ending up in forensic psych 10years later.

a lot of this research would be hypothetical as well, considering it would be pretty unethical to close down a mental health ward and just wait at the prison to see what happens.

sadly, retrospective data is pretty limited too, as it focuses on assumed cause and effect, which cannot truly "code" cultural change - when a city changes one of its major systems, it will not only impact another system, but EVERY other system available.

Consider the history of Morisset hospital - in its glory days it had approx 1500 patients. the patients kept many of their husbandry skills by working the cattle, pigs and other yard-work. There was an industrial kitchen, and ability to learn how to drive a car and a boat. The patients were given all their clothes, all their cigarettes and all their food. Since changes in funding occurred, and the idea that patients were being exploited, the change resulted in patients having disability support pensions to pay for their cigarettes and clothes, with food being funded by the hospital, which would either be paid for by medicare of private health care, and no more skilling up of patients preparing them for a successful return home, instead, it allows a form of counter-dependance in the mental health world.

not a simple change, not a simple result.

Its interesting that the original question was about research in England that might have shown links between de-institutionalisation and the growth in numbers of MI people in prison, yet the bulk of the replies have come from US, Canada, and Australia.