APPENDIX 23
Memorandum by South London and Maudsley
NHS Trust (MH 57)
EXECUTIVE SUMMARY
We address separately the four questions put to us:
1. CURRENT GOVERNMENT
DEFINITIONS AND
CATEGORISATION OF
MENTAL ILLNESS
The proposed category of "mental disorder"
is largely satisfactory, although the exclusions are rather unusual
[paras 1.1-1.2].
The proposed category "dangerous people
with severe personality disorders" (DSPD) is unsatisfactory
in a number of important respects [paras 1.3-1.6].
2. THE ABILITY
OF CARE
IN THE
COMMUNITY TO
CATER FOR
PEOPLE WITH
ACUTE MENTAL
ILLNESS
We agree with the view that community care is
at present both under-funded and under-staffed, but do not agree
that "community care has failed" [paras 2.1-2.3].
Studies are quoted showing that community care
produces better outcomes than hospital based care, and these results
have been shown to apply to standard community services. Research
that has suggested that dramatic reductions in bed numbers could
be made has always excluded from the studies many people requiring
acute hospital care [paras 2.4-2.7].
Some patients will continue to need admission
to hospital-based facilities or equivalent, however well community
facilities are developed. If the numbers of hospital beds are
to be reduced still further, then a range of alternative accommodation
is necessary in the community, community mental health teams must
be adequately staffed, and crisis houses and intensive home treatment
should be more widely available [paras 2.8-2.10].
With additional funds, we suggest improvements
that are needed to increase satisfaction of users and carers with
the service [paras 2.11-2.15].
We explain how existing funds could be distributed
more equitably, and so raise standards of care in deprived inner
city areas. The costs of secure care need to be taken into account
separately in the resource allocation formula [paras 2.16-2.17].
3. THE TRANSITION
BETWEEN ACUTE
AND SECURE
MENTAL HEALTH
SECTORS
The funding and organisation of forensic care
must not be considered in isolation from general mental health
services; the boundary between them is a flexible one and changes
in one will affect the other directly [paras 3.1-3.2].
Although it is desirable to provide beds or
equivalent care as close to people's homes as possible, it should
be borne in mind that improvements to general adult services are
likely to lead to a reduction in numbers of expensive forensic
beds [paras 3.3-3.4].
We agree that those with a severe mental illness
and a history of violent behaviour should have an enhanced level
of the CPA. The majority of patients are not violent, and the
evidence base to predict future dangerousness is not yet strong:
services should therefore continue to be focused on clinical needs
[para 3.5].
If our recommendations for taking forensic needs
into separate account in calculating a new resource allocation
formula are heeded, resources will be available for providing
a liaison forensic services to prisons—and these are greatly
needed [para 3.6].
4. THE TRANSITION
BETWEEN ADOLESCENT
AND ADULT
MENTAL HEALTH
SERVICES
We commend the proposed interventions aimed
at preventing the development of severe personality disorders,
but observe that most of them are outwith the responsibilities
of an NHS Trust, and there is a need for an improved knowledge
base [para 4.1].
In many places, adolescents with psychotic illness
continue to be admitted to adult wards, where the milieu is quite
unsuitable and they are unlikely to receive optimal treatment
[paras 4.2-4.3].
It is especially important that cases of early
psychotic illness receive the best treatment possible, as future
health costs are likely to be high, and recent research evidence
suggests that morbidity can be prevented with earlier, more effective
intervention. We describe steps taken in the South London &
Maudsley Trust to improve co-ordination between the two services
[paras 4.4-4.5].
We draw attention to poor liaison between adolescent
services and schools, and the need for better care for adolescents
with drug problems [para 4.6].
5. WE PRESENT
OUR RECOMMENDATIONS
ON THESE
POINTS
1. The proposed term "DSPD" is
unsatisfactory in several important respects, and we prefer the
Scottish approach—"serious and violent sexual offenders
who present danger to the public".
2. We are of the opinion that community
care can succeed, and that hospital bed numbers could be reduced
further—but only if considerable additional investment is
made in community based residential facilities, in the provision
of extra staff salaries, and additional staff training.
Where intensive home treatment teams are being
established, they should be properly evaluated so that the NHS
can benefit from a developing evidence base in this area.
We further recommend an important modification
to the "Resource Allocation Formula" by making separate
provision for forensic needs, and also making the funding that
has been allocated to Health Authorities for mental health explicit.
3. Although there should be a separate assessment
of forensic needs, HAs should retain their discretion over expenditure.
No realistic risk assessments can be made for
people who do not have a previous history of violent behaviour;
we argue that risk assessments should be based on a full range
of needs of the patients, with careful monitoring of quality of
care.
Our financial recommendations, if accepted,
would support Trusts to set up and resource NHS forensic liaison
services with prisons across the country.
4. Health Authorities should be instructed
to provide facilities for severely ill adolescent patients to
be treated in specialised adolescent beds and day places, and
the practice of admitting adolescents to adult wards should cease.
1. CURRENT GOVERNMENT
DEFINITIONS AND
CATEGORISATION OF
MENTAL ILLNESS
1.1 There has never been a precise correspondence
between legal and medical definitions of mental disorder. The
1983 Mental Health Act recognised four categories: "mental
disorder" was an inclusive category which embraced chapters
290-318 of the International Classifications of Disease, 9th Edition
(ICD-9); "severe mental impairment" roughly corresponded
to disorders described in Chapter 318.2; "mental impairment"
was the same in milder degree—so chapter 317; while "psychopathic
disorder" referred to chapter 301. These categories were
similar to those in the previous Mental Health Act (1959).
1.2 The "Reform of the Mental Health
Act 1983" consultation document put out in November 1999
proposes "mental disorder", to correspond to chapter
F, sections 00-09 of ICD-10, and further defined as "any
disability or disorder of brain or mind, whether permanent or
temporary, which results in an impairment or disturbance of mental
functioning". This seems an acceptable definition. However,
the paper proposes to exclude conditions such as disorders of
sexual preference (F65) and drug and alcohol misuse (F10-F19)
from the category. It is not clear why these particular disorders
are excluded (such as paedophilia and sadomasochism), while others
(for example tic disorders F95; homosexuality F66 x1) are left
in. These decisions appear to us to be rather usual.
1.3 The Home Office Consultation paper proposes
the creation of a new category "dangerous people with severe
personality disorders" (DSPD). In spite of its apparent medical
flavour, this is not a recognised diagnosis: the law should classify
offences; leaving medicine to classify disorders.
1.4 Those who commit these offences should
be referred to as "serious violent and sexual offenders who
may present a continuing danger to the public", as they are
in Scotland. Those who commit such offences are not a homogeneous
group, but may have various combinations of psychosis, drug dependence,
learning disability, personality disorder, or even no mental disorder
at all.
1.5 Nor do scales of "standardised
procedures" exist that can reliably recognise dangerous and
severe personality disorder. The most reliable guide to future
offending is past offending (Heilbrun et al 1999)—which
is why it is of concern for there to be a civil way of entering
preventative detention without a limit of time. The proposal in
the Home Office document to remove the requirement that someone
assigned to the new category should be "likely to benefit
from medical treatment" (p 14, para 14d) is also of concern
to the South London & Maudsley NHS Trust. The proper place
for a serious repeat offender who is unlikely to benefit from
medical treatment may be prison, and that decision must rest with
the Courts. In a free society, there is no proper place for the
confinement of those who have committed no offence and who are
unlikely to benefit from medical treatment, whatever personality
disorder they are deemed to have.
1.6 The proposed term "DSPD" has
all the overtones of moral judgement and social rejection that
the term "psychopath" has carried for so many years.
The developments envisaged by the Home Office as "Option
B" [pp16-18] would only make sense if it had been shown that
there were better ways of assessing future dangerousness on the
basis of personality tests, than could be obtained from a consideration
of past offences. However, the fact that there are some very dangerous
people without personality disorders, as well as many with severe
personality disorders who are not dangerous to others at all,
leads one to suppose that the task may not be soluble in the current
state of knowledge.
1.7 We recommend that a programme of high
quality scientific research is commissioned to investigate the
outcomes of treatment and care for violent and sexual offenders
who constitute a danger to the public.
2. THE ABILITY
OF CARE
IN THE
COMMUNITY TO
CATER FOR
PEOPLE WITH
ACUTE MENTAL
ILLNESS
2.1 On 22 October 1998 the then Under-Secretary
of State for Health announced that "care in the community
care has failed for three reasons: it was under-funded, under-staffed
and lacked a clear and appropriate legislative framework to support
it . . ." We would agree with his assessment of the causes
of problems associated with community care, but we consider that
in overall terms community care has not failed most patients.
2.2 The Government were concerned about
the number of incidents involving homicides by schizophrenic patients,
as well as those with severe personality disorders who were in
the community but clearly were not receiving adequate care in
the community. The numerous deficiencies of the mental illness
service in London were documented in the King's Fund Report (Johnson
et al 1997), but there are similar problems in other large cities
in the United Kingdom. These defects are not confined to community
services: hospital facilities are too often overcrowded, understaffed
and violent, and services for those with chronic disability have
too often been curtailed or relatively neglected.
2.3 Yet the King's Fund Report did not conclude
that community care had failed, as it described numerous positive
features of the present system, and made positive recommendations
for the future. However, it agreed with him that community care
was both under-funded and understaffed, and that further reform
of the Mental Health Act would be helpful.
2.4 There have been numerous papers showing
that community-based services give a better service to users and
carers than the hospital-based services that they replace, showing
that there are both health and social gains that are replicable
in ordinary clinical settings (Thornicroft et al 1998). These
gains are less marked than those reported in earlier cost-efficacy
studies, but they are there—and seen in a range of outcome
measures, including more users receiving services, an improved
quality of life, and better social networks.
2.5 Earlier studies, reviewed in Thornicroft
and Goldberg (1998) had shown that services based on modern acute
units are more cost effective than those based upon a large mental
hospital, in that they produce better social and clinical outcomes
at no greater cost. This paper showed that community care can
prevent the accumulation of long-term damaging effects of institutional
care, can produce better social adjustment of patients, and is
greatly preferred by both users and their carers.
2.6 The early studies of the cost efficacy
of community care gave some commentators the impression that hospital
beds could be dispensed with entirely. It was perhaps not appreciated
that all these studies excluded numerous patients from entry to
the study [for example, those with homicidal or suicidal tendencies;
those with drug dependency in addition to their psychotic illnesses,
those who were homeless, or those with organic brain syndromes
complicating their management]. Also, nearly all these studies
did admit most patients to hospital, but discharged them after
a few days to intensively staffed care in a community setting.
Thus, a more cautious conclusion from them would be that length
of stay can be drastically shortened for many users, providing
that community services are well staffed, and that there is a
readily available range of alternative residential placements
in the community.
2.7 For these large and important groups
of patients who were often excluded from the randomised controlled
trials, there continues to be a need for some form of in-patient
care or intensively staffed equivalent. There are indeed other
groups as well: those needing complex brain imaging techniques
need to be in hospital, as do those mentally ill people with violently
disturbed behaviour. In each area, there should be access to at
least one ward for more behaviourally disturbed patients, and
another for less disturbed patients.
2.8 A reduction in the number of hospital
beds could be made if two important conditions were to be met:
first, there should be a full range of residential facilities
to the required capacity available in the community—some
staffed mainly by the NHS, others by social services or the voluntary
sector—and secondly, there should be adequate numbers of
community staff available to look after them.
2.9 Many patients with established illnesses
need re-admission from time to time to deal with a relapse in
their illness, or to re-establish treatment. Many such admissions
need not be to hospital premises, and can take place in facilities
called "crisis houses" (Sledge et al 1995). In the South
London & Maudsley Trust we have opened in Croydon a "women's
crisis house", that provides such care outside the hospital,
for women who are at least as unwell as those who are voluntarily
admitted, and a further crisis house for women is planned in Lambeth
as well. However, very important additional measures should be
made by social services, housing departments and the voluntary
sector, working in partnership with Health Authorities, to provide
intermediate care, a range of residential accommodation, some
of it staffed 24 hours a day, some staffed only during the daytime,
and yet others with low staffing levels.
2.10 There is evidence from randomised controlled
trials that offering intensive treatment at home to people with
severe mental illness during crises can result in less use of
hospital beds. Though up to half the patients in these trials
require at least a short admission, the overall reduction in bed
use may be up to 60 per cent (Joy et al 1999; Smyth and Hoult
2000). However, attempts are now being made to introduce intensive
home treatment into areas where community care is already functioning,
and to combine it with other community treatment approaches such
as assertive outreach (AO). While early reports of services such
as those in Birmingham (Minghella et al, 1998) and in Islington
suggest that savings in bed use can be made, there is a need for
randomised controlled trials. The South London & Maudsley
Trust is planning to introduce a home treatment service in North
Southwark, and will be evaluating its impact on bed use.
The necessary funding arrangements to improve
the quality of care
2.11 The National Service Framework for
Mental Health refers to a number of current challenges that need
to be met if we are to improve the quality of mental health services.
We need to employ more staff providing evidence based psychological
treatments than we have at present; we should routinely provide
prompt care after discharge to reduce the suicide rate; we should
provide training to the majority of CPNs who have not yet had
training in family interventions for schizophrenic patients; we
should offer therapy aimed at improving acceptance of treatment
generally available—since it has been shown to improve symptoms
and reduce admissions (Kemp & David 1997); we should provide
adequate numbers of permanent, trained nursing staff on our in-patient
units; and we should provide intensive home treatment and assertive
outreach—since these cause increased user satisfaction,
and can also reduce admission rates.
2.12 Resources are allocated to Health Authorities
(HAs) using the "York formula", which contains a psychiatric
component based upon census data. This results in a three-fold
variation between HAs in the psychiatric component of their allocations.
However, HAs have discretion over the way they spend their allocated
resources, and actual expenditure varies widely, not always proportionately
to the needs implied by the formula.
2.13 When one compares the amount allocated
to HAs per capita for mental illness with the amount spent, there
is enormous variation—with some spending very much less
than their notional allocation, and others rather more. In practice,
per capita expenditure on all mental illness has a five-fold variation
between HAs, and expenditure on mental disordered offenders (MDOs)
has a 20-fold variation (Glover et all 1999).
2.14 Since deprived inner city areas inevitably
have the highest forensic expenditures, this often has the effect
of starving other parts of local health services on funds.
2.15 Outside London, secure costs
impinge on the resources available for other forms of mental health
care. Within London, secure care costs are exceptionally
high, but expenditure on other forms of mental health care is
also higher than predicted by the resource allocation formula,
and the demand for secure care is therefore drawing resources
away from other parts of the health service. There are strong
reasons why inner London HAs are obliged to spend more than their
allocation on mental health, due to the extent of the drug problem
in London, the needs of refugees, the large immigrant populations
with higher rates of illness and specific needs, and the particularly
high rate of single person households compared with other deprived
inner cities.
2.16 If the implications of the resource
allocation formula for mental health spending were to be made
explicit to Health Authorities, while leaving them with discretion
over their expenditure, under-funding of mental health services
in the most deprived areas might be corrected over time.
2.17 These problems could be addressed by
revising the resource allocation formula to take account of the
highly variable needs for forensic services, and other relevant
factors such as the size of the local prison population, and the
number of high support hostels in each local area. This would
result in a shift in resources towards those areas where needs
for forensic services are greatest. HAs should have the option
of spending mental health resources directly on forensic services,
or on other forms of service aimed at reducing the total need
for beds. The current review of the resource allocation formula
needs to take these issues into account.
3. THE TRANSITION
BETWEEN ACUTE
AND SECURE
MENTAL HEALTH
SECTORS
3.1 The interface between general and secure
mental health services may be of two types:
integrated, in which individuals requiring secure
in-patient care are cared for by forensic services but return
to general services as soon as secure care in no longer required
(transition between the two therefore depending on the nature
of the care required), and
parallel, in which mentally disordered offenders
(MDOs) are cared for by forensic services in a range of settings
including the community (transition being dependent on the offending
history or perceived risk characteristics of the patient). In
practice, local services tend to contain elements of both integrated
and parallel services in varying proportions.
3.2 No systematic evidence favours the adoption
of one or other model. Whichever model is adopted, it is important
to consider the fact that most patients of forensic services will
at times be in need of services provided by acute mental health
services, as well as by social services and the voluntary sector.
3.3 Although forensic needs must be included
in a future allocation formula, forensic services should not be
provided with funds in isolation from acute services. General
services provided effectively at an earlier stage in a patient's
illness may help prevent later offending and the need for acute
care.
3.4 At present, many patients in need of
secure care are placed in private facilities at great distances
from their homes. Standard 5 of the National Service Framework
for Mental Health states that hospital care should be provided
as close to home as possible, is not met. During long admissions,
vital links with family and the local community are lost, and
this complicates attempts at rehabilitation through community
services. Current programmes such as those funded from the NHS
Modernisation Fund to build more secure units may go some way
to addressing this problem. However, a more efficient way of containing
the demand for secure beds is to improve the quality and comprehensiveness
of acute general services.
3.5 It is reasonable that those with a severe
mental disorder and a history of violent behaviour should receive
an "enhanced level of the Care Programme Approach",
as has been suggested. However the majority of patients are not
violent, and the evidence base to predict future dangerousness
is not strong, and is likely to produce far too many "false
positives". Standard adult mental illness services should
continue to be focused on the patient's needs, and the quality
of care that is being received.
3.6 There is a high level of psychiatric
morbidity and suicide in prisons (ONS 1998). We are pleased that
the government has recently responded to this and given the Department
of Health a lead role in providing medical services to prisons.
However, mental health services to prisoners remain poor (Birmingham
et al 1998; Robertson et al 1994), and liaison between prisons
and local mental health Trusts is highly variable. We accept in
principle the responsibility of Trusts to take responsibility
for improving liaison services to local prisons. However, given
the failure of resource allocation to HAs to make adequate allowance
for the need for forensic services in deprived areas (2.7 above)
Trusts with large local prison populations are often constrained
in their ability to offer an adequate liaison service. Improving
mental health care for prisoners should be a priority. The changes
to the resource allocation formula which we have proposed would
to some extent help to achieve this.
4. THE TRANSITION
BETWEEN ADOLESCENT
AND ADULT
MENTAL HEALTH
SERVICES
4.1 The measures set out in "Managing
Dangerous People with Severe Personality Disorders" aimed
at preventing children and adolescents from developing severe
abnormalities of personality are admirable in their intentions.
Those aimed at supporting families, improving conditions in schools,
preventing bullying and providing better education on parenting,
are especially welcome. However, the research base for these various
interventions requires improving, and many of the suggested improvements
do not lie within the remit of an NHS Trust.
4.2 There is considerable variation across
the country concerning how well the interface between adolescent
and adult services works, but we are unaware of any empirical
data concerning the interface. However, adolescents are still
admitted to adult wards in many parts of the country—and
this is usually highly undesirable, depending upon how mature
they are for their years. Figures exist in Wales in 1996-98 published
by the Welsh Office show that a total of 69 adolescents were admitted
to adult wards in South Wales, and 63 in North Wales. Some of
these were admitted to paediatric wards, but these are also unsuitable,
since they are unlikely to offer the patients the range of specialised
services that are necessary.
4.3 It has been estimated that about 4 per
cent of schizophrenic episodes have their onset before the age
of 17 (Cannon et al 1999), so this is a problem which requires
a solution. There is also a lack of age-appropriate day care facilities
for adolescents with severe mental disorders; since not all of
these people need to be admitted to hospital for in-patient treatment.
The newly established Lambeth Early Onset (LEO) project has been
designed to offer optimum treatment and care to people who develop
a first episode of psychosis, many of whom are teenagers. It will
create a range of services acceptable to people who have traditionally
found mental health services inappropriate to their needs, or
culturally inappropriate. This service will also have a newly
developed designated in-patient ward providing safe areas for
women. It is especially important that adolescents with severe
mental illness receive the best possible treatment, and do not
have to take their chances on adult wards where the milieu may
be quite unsuitable for them. There are at present insufficient
numbers of in-patient adolescent beds and day places to allow
this highly undesirable practice to cease.
4.4 Not all adolescents who have been treated
by specialist adolescent mental health services wish for or need
to be seen by the adult services: it is often completely appropriate
to promote resumption of development towards independent functioning,
with social/educational support as necessary, but without the
need for indefinite involvement from mental health services. However,
it is important that those with complex and long-term needs (in
particular those with severe psychotic illnesses) do receive continuity
of care—up to 70 per cent of adolescents experiencing a
psychotic episode will experience another episode within five
years (Kraus & Muller-Thomsen 1993). Another recent European
study showed that half of those developing psychosis before adult
life had a severe, continuous course to their illness, with those
developing the illness before the age of 12 often having an insidious
onset (Eggers et al 1999). Birchwood (1999) has described the
"critical period" early in the course of the illness,
where it is desirable that in addition to optimal pharmacological
management, the patient receives cognitive-behavioural treatment
as well, and family interventions are tried in an attempt to reduce
the [otherwise high] probability of relapse.
4.5 In the South London & Maudsley NHS
Trust, in the newly established Croydon Adolescent Community Service,
for example, there are between five and eight referrals per year
for psychotic adolescents, for a unit with approximately 120 referrals
per year—the remainder being mainly conduct disorders, severe
depression and deliberate self-harm. Our service in Croydon is
arranged on multidisciplinary lines similar to an adult community
mental health team, and the consultant psychiatrist and community
psychiatric nurses devote regular time to liaison between their
own team and the appropriate adult mental health team. Also within
our Trust the work of Professor Hilton Davies indicates that early
parenting assistance to families may help to reduce the occurrence
of mental health problems in adolescents.
4.6 Nationally, the interface between specialist
Mental Health, Education and Social Services is poorly developed,
and protracted hospital admissions are most likely for those with
severe and complex needs, where policies for multi-agency co-ordination
of assessment and provision are lacking and there is a lack of
joint adolescent/adult services for those with drug dependence.
As is well known, there is a considerable overlap between drug
use and severe psychotic illness. Recent studies in London, for
example, show that at least 30 per cent of patients with psychotic
disorders also misuse drugs, or alcohol, or both.
REFERENCES
Birmingham, L; Mason, D; Grubin, D. (1998) A
follow-up study of mentally disordered men remanded to prison.
Criminal Behaviour and Mental Health. 8(3), 202-213.
Bindman J. (1999) Allocation and expenditure
on mental health services in England: The implications for equity.
MSc dissertation: London School of Hygiene and Tropical Medicine.
Bindman J, Beck A, Thornicroft G, Knapp M, Szmukler
G. Which psychiatric patients are at greatest risk and in greatest
need? Impact of the Supervision Register Policy. British Journal
of Psychiatry (in press).
Bindman J, Beck A, Glover G, Thornicroft G,
Knapp M, Leese M, Szmukler G. (1999) Evaluating mental health
policy in England: The Care Programme Approach and Supervision
Registers British Journal of Psychiatry 175 327-330.
Birchwood M (1999) Early intervention in psychosis:
the critical period. In Ed McGorry PD and Jackson HJ "The
recognition and management of early psychosis. Cambridge: Cambridge
University Press pp 226-264.
Cannon M, Jones P, Huttunen MO, Tanskanen A,
Huttunen T, Rabe-Hesketh S, Murray RM (1999) School performance
in Finnish children and later development of schizophrenia: a
population-based longitudinal study. Arch Gen Psychiatry 56(5):457-63.
Eggers C, Bunk D, Volberg G, Ropcke B (1999)
The ESSEN study of childhood-onset schizophrenia: selected results.
Eur Child Adolesc Psychiatry 1999;8 Suppl 1:I21-8.
Glover G (1999) Can we say how much English
HAs are allocated for mental health? British Journal of Psychiatry
175 402-406.
Gunn J, Maden A & Swinton M (1991) Treatment
needs of prisoners with psychiatric disorders. British Medical
Journal 303, 338-40.
Heilbrun K, Ogloff JRP, Picarello K (1999) Dangerous
Offender statutes in the USA and Canada: implications for risk
assessment. Int J Law Psych: 22; 393-415.
Joy C, Adams C, Rice K (1999) Crisis intervention
for people with severe mental illnesses. Cochrane Library.
Johnson S, Ramsay R, Thornicroft G, Brooks L,
Lelliott P, Peck E, Smith H, Chisholm D, Audini B, Knapp M, Goldberg
DP (1997) London's Mental Health. London: The King's Fund.
Kemp R; Kirov G; Everitt B; Hayward P; David,
A (1998). Randomised controlled trial of compliance therapy: 18-month
follow-up. British Journal of Psychiatry. 172; 413-419.
Kraus M, Muller-Thomsen T (1993) Schizophrenia
with onset in adolescence—an 11 year follow-up period. Schizophrenia
Bulletin 19; 831-834.
Maden T; Swinton M; Gunn J (1994) Therapeutic
community treatment: A survey of unmet need among sentenced prisoners.
Therapeutic Communities: the International Journal for Therapeutic
and Supportive Organizations. 15(4), 229-236.
MILMIS Project Group (1995) Monitoring Inner
London Mental Illness Service Psychiatric Bulletin 19: 276-280.
Minghella E, Ford R, Freeman T et al (1998)
Open all hours: 24 hour response for people with mental health
emergencies. London: Sainsbury Centre for Mental Health.
Office of National Statistics (1998). Psychiatric
Morbidity among prisoners in England and Wales—a survey
carried out in 1997 on behalf of the Department of Health, London:
The Stationery Office.
Sledge W, Tebes J, Rakfeld J (1995) Acute crisis
respite care. In ed Phelan P, Strathdee G, and Thornicroft G "Emergency
mental health services" Cambridge: Cambridge University Press.
pp 233-258.
Smyth M, Hoult J (2000) The home treatment enigma.
British Medical Journal 320 305-307.
Thornicroft G, Goldberg DP (1998) Has community
care failed? Maudsley Discussion Paper No.5. London: Institute
of Psychiatry.
Thornicroft G, Wykes T, Holloway F, Johnson
S, Szmukler G (1998) From efficacy to effectiveness in community
mental health services. Prism psychosis study 10. British Journal
of Psychiatry 173; 423-427.
| 
