Professor Paul Rogers
Professor of Forensic Nursing
Faculty of Health, Sports and Science
“
Mental Health Research:
What are the risks of that happening?
Aim
Give a rough overview of my career.
Present two examples of research which examines the
evidence base examining mental health and the issue
of “risk” of violence.
Of note, these are areas where the “evidence” or the
“clinical practice” had already been very firmly
established. So there was “no need” to do the
research!!!
Available to download from - http://office.research.glam.ac.uk/
Forensic (psychiatric) Nursing
Forensic – pertaining to the law
Nursing ........those who by the nature of their health
condition are likely to come into contact with the legal
/ criminal justice system.
Historically related to mental health
Offenders of Crime (UK)
Victims of Crime (USA)
Why Psychiatric Nursing?
Nature versus Nurture
Mother and Father are
Psychiatric Nurses.
My Father was the eldest
of 9 siblings, of whom 7
became psych nurses
P
S Y
My mother has one sister
who is a psych nurse
My Maternal Grandmother was one of the first
ever Registered Psychiatric Nurses in Ireland
(Kilkenny).
C
H
I
My background
- Is there a gene for
“psychiatric nursing”?
Oliver
Monica
Joe
Sean
Rogers
Mary
Rogers
(Grennan)
Bernadette
Gabriel
= Psychiatric nurse
gene??!
Margaret
Tom
Fitzgerald Fitzgerald
(Brennan)
Mary
Jimmy
Frank
Eddie
Rogers
Margaret
Rogers
Paul
Rogers
Allison
Rogers
(Farrelly)
I was told that I had to do
………………………………………………
………………………………………………
…“the obligatory baby photos”!
The obligatory baby snapshot!!!
TO INSERT
Early career
Aged 16. Tomato picking, Southport
Aged 16. Psychiatric Nursing Home, Southport
Aged 17. General Nursing Home, Southport
Aged 17. GNC Nursing entrance test – Park Lane Hospital
Aged 18. Student Psychiatric Nurse – Fairfield Hospital, Beds
Fairfield Hospital, Bedfordshire
Fairfield Hospital
Opened in 1860 - Three Counties
Asylum (Beds., Bucks., Herts.)
350 Acres
In 1986 had 63 wards; All were
full with a patient population
in excess of 2000
Approximately 25% of Wards were “locked”
Most Secure Ward = M8 Ward
M8 Ward
36 bedded Male Ward (mostly
from Bedford Prison)
Four staff
18 Seclusion rooms
Fairfield Hospital – In short
Met Allison
Practices were “staff focussed”
Control & Restraint training
Why wasn’t violence “predicted”
How are things “prevented”
What is Psychiatric nursing?
National Brain Injury Unit
Aged 21 (1989) – Staff Nurse - National Brain Injury Rehab
Service, St Andrew’s Hospital, Northampton
Applied Behaviourism
15 minute token economy programme
Time Out for Aggression & Ind. Programmes
Last ward in the UK to use “Food” as a “reinforcer”
Moving to a culture of “positive programmes”
Became a Home Office “approved” Control and Restraint
Instructor
Caswell Clinic, Bridgend &
District NHS Trust
Aged 23 – Charge Nurse
Intensive Care Unit,
Caswell Clinic,
Interim Medium Secure Unit
“Humanistic approaches”
No seclusion rooms
Care was focussed through the Nursing Care Plans
At that time - No real “Risk assessment”
Cert ENB 650 Course
Cert ENB 650 99 Denmark Hill, Maudsley Hospital
Cert ENB 650 99 Denmark Hill, Maudsley Hospital
●
●
●
●
●
18 month, Full time course; we were Course 15
Started in 1971, National Referral Centre
Clinical Director - Prof Isaac Marks
Trained nurses to deliver Behaviour Therapy
Previous Students – Prof Kevin Gournay, Prof Charlie Brooker
● 1n 1994 - Approx 12 National Training places
● “De-constructed” what we did and then ++ training
● Treated out-patients (OCD, PTSD, Agoraphobia, Social
Phobias, Habit Disorders, Body Dysmorphic Disorder, Specific
phobias)
● The Single Case Study Experimental Design
The Single Case Study
Experimental Design
Caswell Clinic - The Single Case Study
Experimental Design
Aged 28 (1995) returned to the Caswell Clinic - CNS in CBT
Rogers, P. (1997). Posttraumatic stress disorder following male rape.
Journal of Mental Health, 6(1), 5-9.
Rogers, P. and Darnley, S. (1997). Self-monitoring, competing
response
and response cost in the treatment of trichotillomania.
Behavioural and Cognitive Psychotherapy, 25, 281-290.
Rogers, P. & Gronow, T. (1997). Anger Management: Turn down the
heat. Nursing Times, 93(3), 26-29.
Rogers, P., Gray, N.S., Williams, T. & Kitchener, N.J. (2000).
The behavioural treatment of PTSD in a perpetrator of
manslaughter. Journal of Traumatic Stress, 13, 511-519.
.
Clinically
CBT – seeing two main sets of clients
1. Males with horrendous personal histories who
developed PTSD
2. Those with “command hallucinations” who had
acted on their commands with very serious
consequences
The psychiatric research at that time reported
that Command hallucinations were NOT
dangerous (either suicide / violence)!
Preliminary study – examining
command hallucinations and “risks”
We examined whether patients in a Medium Secure
Unit were more likely than other clients to:
1. Engage in self harm / suicidal behaviour
2. Engage in violence to others
Main issues that we found were 1. What about
the “content” of the command & 2. What about
possible confounding
Rogers, P., Watt, A., Gray, N.S., MacCulloch, MM & Gournay, K. (2002)
Content of command hallucinations predicts self harm but not violence
in a medium secure unit. Journal of Forensic Psychiatry, 13(2), 251262.
PhD - The Association
between Command
Hallucinations and Violence
1999 - PhD
The Association between Command
Hallucinations and Violence
Full time funded PhD – Wales Office of Research
and Development
Undertook PhD at the Institute of Psychiatry with
Professors Kevin Gournay and Professor
Graham Thornicroft as supervisors
Advisors – Professor Nicola Gray, Cardiff
University and Professor Glyn Lewis, Bristol
University
Command Hallucinations
“Command hallucinations are auditory
hallucinations that order particular acts,
often violent or destructive ones and
instruct a patient to act in
a certain manner”
(Hellerstein et al, 1987)
Command hallucinations and risk
By the mid 90’s, UK policy / politicians had tasked
forensic and mental health practitioners to improve
risk assessment and management
In the past 60 years, international research has
examined the associations between 1. Diagnosis and
2. Symptoms of mental disorder and violence.
No strong association between diagnosis and violence
Some positive associations found for delusions and
violence
No association found for command hallucinations
Clinical Wisdom
Clinical wisdom from the past 70 years has
assumed and directed that command
hallucinations are associated with and lead to
violence.
Bleuler, E. (1930). Textbook of Psychiatry (trans
A.A. Brill), New York, Macmillan.
Schneider, K. (1959) Clinical Psychopathology. New
York. Stratton.
Real World experiences
The research just didn’t make sense!!
Clinically, we had encountered many people who
report having been violent as a direct result of
hearing command hallucinations.
Personal experience = high proportion of clients
in forensic services report command
hallucinations.
Many of the patients in the homicide inquiries had
command hallucinations
Numerous Case Reports
Violence to others (Good, 1997)
Self amputation of a limb (Hall et al, 1981)
Swallowing objects (Karp et al, 1991)
Plucking out own eyes (Field & Waldfogel, 1995)
Self inflicted lacerations (Rowan & Malone, 1997)
Suicide (Zisook et al, 1995)
Command Hallucinations &
Violence (Pre-2000)
By year 1999, 7 controlled studies had
found NO relationship between command
hallucinations and an increased risk of
violence
Therefore, 3 systematic reviews had also
found NO relationship between command
hallucinations and an increased risk of
violence
Clinical wisdom or “psychiatric myth”
Have psychiatric practitioners and services been
unnecessarily detaining people due to a 70 year
old “myth” about the association between
command hallucinations and violence?
Bleuler, E. (1930). Textbook of Psychiatry (trans
A.A. Brill), New York, Macmillan.
Time for some critical appraisal
• None of the studies
were prospective.
• Research evidence was based upon a
total population of 237!
• Only 13% of cases had reported
commands directing violence
Study examples
Study
Sample
size
Command content
Hellerstein et 58
al, (1987)
30 suicide
7 self-harm
3 homicide
8 non violent acts
10 unspecified
Zisook et al,
(1995)
20 violent (self/others)
19 benign
7 unspecified
46
Trying to make sense of the “evidence”
Are the pre-2000 studies that found no
association accurate ???
or …...................
Could they possibly be misleading
Clinicians into discharging people who
“may” be a risk?
SECONDARY ANALYSES OF
THE MACARTHUR VIOLENCE
RISK ASSESSMENT DATA
Research grant provided by the Wales Office of
Research and Development for Health and Social
Care (S98/004)
Research Questions
• Are violent-content command
hallucinations associated with 1yr
FU violence compared with all other
patients?
• What happens to the association if
we examine those with non-violent
content command hallucinations?
Sample
Secondary data epidemiological analysis of the
MacArthur Violence Risk Study Data
Largest worldwide study of its kind which used
multiple methods for determining whether
violence occurred after discharge
1,136 patients were randomly selected from
12,873 patients who were admitted to any of
three large US hospitals
Patients followed up every 10 weeks for 1 year
Outcome Measure: Violence
All participants were followed up every 10 weeks
for 1-year after discharge
Subject self-reports
Collateral informant reports
Official arrest records
Hospital admitting incident chart information
Rehospitalisation records
All violent incidents were systematically reviewed,
independently coded and a decision was made as
to whether it occurred
Exposure: Command Hallucinations
THE AUDITORY HALLUCINATIONS SCHEDULE
Have you more than once had the experience of
hearing things or voices other people couldn't
hear?
Do the voices tell you to do anything?
What is the highest level of violence they have
commanded?
This allowed us to categories commands in to two
groups: - violent or non violent
Adjusted confounders
Age
Gender
Marital status
Ethnicity
Beaten as a child
Beaten as a teenager
History of drug abuse
History of alcohol abuse
Severity of symptoms (BPRS)
Impulsivity (BIS)
Any delusions
Persecutory delusions
Psychopathy (PCL:SV)
Living with relative post discharge
Prior arrests for “crimes against the person”
Statistical Analysis
Random-effects, repeated measures, logistic
regression
Unadjusted and then adjusted Odds ratios adjusted for time, and a range of
confounders (with 95% C.I.’s)
Odds ratios:
Anything above “1” = an increased risk of
violence.
Anything below “1” = a reduced risk of violence
Results
The words - “all your eggs”
and...............
“one basket” - spring to mind!
Proportion who
had a violent
incident over 1
yr
Unadjusted OR
(95% C.I.)
Fully adjusted (time
and confounders)
OR (95% C.I.)
All others
26.7%
(Reference group) (199/745)
(n=887)
1.00
1.00
Non-violent
content command
hallucinations
(n=125)
23.9%
(26/109)
0.87 (0.531.41)
0.66
Violent content
command
hallucinations
(n=105)
44.4%
(36/81)
2.03 (1.251.30)
1.86
 .005
 .05
 .0001
(0.37-1.16)
(1.09-3.18)
Conclusions
There was good evidence that violentcontent command hallucinations were
associated with future violent incidents
both before and after adjustment for
confounding.
There was no evidence that any-content
command hallucinations were
associated with future violent incidents
either before or after adjustment for
confounding.
Real world implications
We need to consider the content of command
hallucinations when making decisions about
future violence risk
We need to trust our clinical “uncertainty”
We need to critically appraise the quality of
research when making conclusions about their
findings
What training are staff getting regarding risk
assessments and the evidence base???
MRC Post Doctoral Fellowship
2002. Left the NHS for University Life! And properly joined
the Section of Nursing, Health Service Research
Department, Institute of Psychiatry, Kings College.
Professor Kevin Gournay, Dr Sue Plummer, Dr Richard Gray,
Dr Mark Haddad, Jimmy Noak, Edwin Gwenzie, et al.
Awarded £220,000 by the Medical Research Council to
conduct secondary analysis of two cross sectional surveys
in order to investigate the aetiology of high rates of
psychiatric morbidity and suicidal thoughts among
prisoners. MRC Special Training Fellowship (Health
Services Research and Public Health). (2002-2006).
Post Doc / University of Glamorgan
Involved in a range of studies:
Connolly A, Rogers P, Taylor D. (2007). Antipsychotic prescribing quality and
ethnicity: a study of hospitalized patients in south east London. Maudsley
Hospital (£60,000)
£34,000 study from the National Programme on Forensic Mental Health Research
& Development. Assessing the utility of the Offenders Group Reconviction
Scale-2 in predicting the risk of reconviction within 2 & 4 years of discharge
from English & Welsh Medium Secure Units. (2004-06).
Amos, T., et al. A review of forensic and prison reviews. (£35,000). The National
Programme for Forensic Mental Health Research, Department of Health.
(2005-2006).
Harrison, G. et al. The DEBIT Trial (A Clustered Randomised Controlled Trial to
reduce anti-psychotic polypharmacy (£430,000). Funded by an NHS Regional
R&D grant (2000-2005).
Developed a bid to WAG for £90,000 to set up WARRN through the NHS
Current main research interest
Main theme / My main interest
● Examining issues related to imminent or real
violence.
● What does one actually do in these circumstances?
● Linked to my earlier experiences on “M8” ward at
Fairfield Hospital and my “Control and Restraint”
Instructor training
Latest Research Findings
on “Breakaway” training
We know that violence to
healthcare staff is a major
problem
Health
Business
Sic/Tech
World
Front
UK
Education
Politics
Page
SHOCKING NHS VIOLENCE FIGURES
RELEASED (2002)
The NHS executive has
reported upon a national
cross sectional survey and
found that in the last year
there were 65 000 violent
incidents reported against
staff in the NHS.
Healthcare & Violence
Scottish Health Service Management Executive (1996)
Royal College of Psychiatrists (1998)
NHS Executive (2000)
Nursing & Midwifery Council (2001)
NHS Security Management Service (2001)
National Audit Office (2003)
World Health Organization (2003)
Welsh Assembly Government (2004)
National Institute for Clinical Excellence (2005)
National Institute for Mental Health England (2005)
Wales Audit Office (2005)
Protecting NHS staff from
V&A – Welsh violence data
8,000 incidents of violence and aggression,
in Welsh NHS Trusts = 22 incidents per
day (2003-04 )
Mental Health staff most likely to be
assaulted, followed by Learning Disability
then A+E
Cost due to consequences of violence or
investment in training- £6.3 million in (03-04)
Lets recommend “breakaways”
Scottish Health Service Management Executive (1996)
Royal College of Psychiatrists (1998)
NHS Executive (2000)
United Kingdom Central Council (1999)
Nursing & Midwifery Council (2001)
NHS Security Management Service (2001)
Welsh Assembly Government (2004)
National Institute for Clinical Excellence (2005)
National Institute for Mental Health England (2005)
Welsh Assembly Government
- Passport scheme
Breakaway training must be available to
all employees who require it
Aims of breakaway
training - To provide
practical techniques
enabling breakaway
from violent/aggressive situations
History of breakaway training
“Breakaway training” is a part of the wider
“Control & Restraint training” (from Ju
Jitsu)
Home Office adoption for Prison Service in
1981
4 UK High Secure Hospitals 1985 onwards
Cascaded downwards
Breakaway refresher training
Scotland - 1 year
England - 1 year
Northern Ireland - 1 year
Wales - 2 years!!!!
Examples of breakaways
Breaking away from …..
Wrist grab
Bear hugs
Hair pull
“Standing up” Strangle / neck locks
Clothes grab
What is the evidence base
supporting “breakaway
training”?
NICE Guidelines/ systematic review
5 UK studies which attempted to
evaluate
the
effectiveness
of
breakaway training in mental health
Only one found any difference; that staff
felt
satisfied
and
slightly
more
confident as a result of the training
(Southcott, et. al. (2002).
Study 1 – Do staff recall their
breakaway training?
An opportunistic sample of 47
nurses in a MSU
We would expect these 47 nurses to
be able to breakaway from holds
as the service they work within
holds the most dangerous
psychiatric patients in Wales.
Do staff recall their breakaway training?
Nurses approached on the ward with no warning
Asked to participate in a study evaluating
breakaway techniques
Picked one of 6 envelopes which contained a
named “hold” (strangle, grab, hair-pull)
Nurse had 10 seconds to prepare
One staff initiated the hold
2 staff recorded time and whether
the correct technique was used
Do staff recall their breakaway training?
50 nurses approached / 47 agreed (94%) to
take part.
One of the nurses who refused was a “C&R
Instructor”
All had had previous breakaway training.
11 staff had received the full breakaway
training more than once.
24 had at least one update since their original
breakaway training course.
Do staff recall their breakaway training?
Forty percent (19/47) were unable to breakaway
within the ten second period.
Of those that did breakaway - 60% did not
employ the “correct” breakaway technique.
One of the sample who did not employ the
correct technique was one of the Instructor’s
Although violence was a problem within the
Clinic, none of the sample (0/47) had used a
breakaway technique in the preceding 12
months!!!!!
A big surprise!
Despite exposure to violence
(mostly kicks and punches)…NONE
of the 47 nurses had needed to
use a breakaway technique in
the last year!
Maybe this was a weird sample?
St. Andrews Hospital (Northampton)
have replicated the study with a
larger sample.....
Of 147 healthcare staff only 15%
were able to breakaway from a
hold within 10 seconds using
correct technique.
Is breakaway training sufficient?
Staff just can’t remember it…. But
we shouldn’t really be determining
the training that staff need until we
know more about what the
violence that staff have to face
actually is…..
What are the realities of
NHS Violence?
Problem = Despite the headline
news items about NHS Violence,
no responsible body is able to
provide detailed data on the type
of assaults staff face!
Research question = What is the
reality of violence to NHS staff??
Study 2
We surveyed all mental health nurses in 2
Welsh NHS Trusts in all clinical areas
Total n = 471
340 from Trust 1
121 from Trust 2
Over 75% return rate
Percentage of staff who have had access
to breakaway training in last 2 years
45, 10%
390, 90%
yes
no
Study 2 - ? Assaulted in last 2 years
Some members of staff who did report an
assault also reported being subjected to
more than one type of assault and
experiencing same type of assault on
more than one occasion.
Therefore, over the two year period a
total of 5866 assaults were reported
This reflects an average of 6.74 assaults
per staff member over a 12 month
period.
Study 2 - ? Assaulted in last 2 years
However, majority of these assaults were
reported by staff working with elderly
patients
Elderly = total of 5626,
25.8 assaults per person per 12 months
Adult = total of 207
0.38 assaults per person per 12 months
Study 2 – assaults by type
1
2
3
4
5
6
7
8
9
10
11
12
Grabbed
Punched
Pushed
Kicked
Slapped
Spat at
Pinched
Hair pull
Head butt
Weapon
Strangle
Other
(Yes)
(No)
(No)
(No)
(No)
(No)
(No)
(Yes)
(No)
(No)
(Yes)
(No)
Breakaway studies - Conclusions
We now have evidence that:
Where nurses are “held” and try to breakaway,
they are unlikely to use the correct
technique
Even if they could remember all of the correct
techniques, it is likely that they won’t
actually be able to deal with the majority of
NHS violence (blows and strikes)
Therefore, evidence base for “Breakaway
training” is poor
Where next?
Large £40,000 funded study into
breakaway retention of approx 160
participants
Ensuring that our courses are linked to
our research and the evidence we
develop – (BSc in Violence Reduction)
Violence Research Group (N=90) –
Academics and Clinicians
Inaugural summary
It has been an interesting career thus far; with many
excellent influences.
The studies I have most enjoyed are the ones
presented – where “wisdom” or practice had
already been “established”.
Critical appraisal and knowledge of the real world
experiences can lead to the challenge of existing
“wisdom” / evidence.
The evidence base for most mental health nursing
activity is tenuous due to research funding
access and due to design issues
Thank You
perogers@glam.ac.uk