Retsudvalget 2015-16, Social- og Indenrigsudvalget 2015-16, Ligestillingsudvalget 2015-16
REU Alm.del Bilag 254, SOU Alm.del Bilag 245, LIU Alm.del Bilag 52
Offentligt
1618579_0001.png
September 2015
1
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0002.png
This report was prepared jointly by the Centre for Forensic Behavioural Science and the Victorian Institute
of Forensic Mental Health (Forensicare). The views of the authors do not necessarily represent the views of
the Government of Victoria or the Department of Health and Human Services, or Department of Justice and
Regulation, whose joint funding of the Program is acknowledged.
Suggested Citation:
McCarthy, J., McGrail, J., McEwan, T., Ducat, L., Norton, J., & Ogloff, J. R. P. (2015).
Evaluation of the
Problem Behaviour Program: A Community Based Program for the Assessment and
Treatment of Problem Behaviours.
Melbourne, Victoria: Forensicare and Centre for Forensic
Behavioural Science, Swinburne University of Technology.
Contact:
Centre for Forensic Behavioural Science
505 Hoddle Street
Clifton Hill 3068
Victoria Australia
[email protected]
+61 3 9947 2606
+61 3 9947 2650
2
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
Contents
PART 1: Describing Problem Behaviour Program (PBP) Clients ....................................................................................... 4
Key Findings ................................................................................................................................................................. 4
Background .................................................................................................................................................................. 5
Results ......................................................................................................................................................................... 5
Referral information ................................................................................................................................................ 5
Presenting problem behaviours .............................................................................................................................. 7
Mental health information ...................................................................................................................................... 7
Treatment clients .................................................................................................................................................... 9
Criminal History and Recidivism ............................................................................................................................ 10
Mental health system contacts ............................................................................................................................. 10
Conclusions ................................................................................................................................................................ 12
PART 2: Offending and mental health outcomes for individuals assessed and treated at the Problem Behaviour
Program (PBP) ................................................................................................................................................................ 14
Key Findings ............................................................................................................................................................... 14
Background ................................................................................................................................................................ 15
Purpose of the evaluation ......................................................................................................................................... 16
Key Questions ........................................................................................................................................................ 16
Definition of terms ................................................................................................................................................ 16
Methodology ............................................................................................................................................................. 16
Data linkage procedure ......................................................................................................................................... 16
Sample Descriptives .............................................................................................................................................. 16
Question 1: Is the PBP effective in reducing the frequency, nature and time to reoffence? ............................... 17
Question 2: What impact does contact with the PBP (assessment and/or treatment) have on mental health
outcomes for clients? ............................................................................................................................................ 17
Question 3: Who are PBP treatment clients, what is the impact of treatment, and what is their experience? ... 18
Results ....................................................................................................................................................................... 18
Question 1: Is the PBP effective in reducing the frequency, nature and time to reoffence? ............................... 19
Question 2: What impact does contact with the PBP (assessment and/or treatment) have on mental health
outcomes for clients? ............................................................................................................................................ 26
Question 3: Who are PBP treatment clients and what is their experience? ......................................................... 28
Discussion .................................................................................................................................................................. 30
Reduction in the nature, frequency and time to reoffending ............................................................................... 30
Mental health outcomes for PBP clients ............................................................................................................... 31
Efficacy of PBP treatment ...................................................................................................................................... 32
References ................................................................................................................................................................. 34
3
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0004.png
PART 1: Describing Problem Behaviour Program
(PBP) Clients
Key Findings
In an analysis of all clients attending the PBP over a five year period to 2011, it was found that
violence was the most common reason for referral to the PBP, accounting for 46% of overall
referrals. Harmful sexual behaviour accounted for a further third of referrals and stalking for
a quarter (with a number referred for multiple problem behaviours).
More than 40% of referrals to the PBP came from Community Correctional Services, followed
by Area Mental Health Services (30%) and self-referral (6%).
The vast majority of PBP clients have had previous contact with the Victorian public mental
health system (90%). PBP clients were diagnosed with a range of psychiatric disorders
including psychotic disorders (28%), depressive disorders (15%) and paraphilias (13%). This
suggests that, although not based on a typical forensic mental health service model, the PBP
routinely provides assistance to mentally disordered offenders and others experiencing
serious mental health difficulties.
Individuals seen at the PBP have diverse offending histories. Nearly 20% were versatile
offenders (index problem plus five or more offence types). Approximately a third (37.2%) of
clients who had contact with the PBP were charged for a subsequent offence; this offending
included violent offences (16%); breaching a legal order (3.6%); sexual assault (3.3%) and
weapons offences (2.2%)
Approximately 25% of clients seen for assessment at the PBP went on to receive individual
treatment. However, 60% of clients recommended for treatment dropped out prior to
commencement or satisfactory completion. This likely relates to the persistent and
challenging nature of many clients’
behaviours.
37.2% of all PBP clients (attending for either assessment only, or assessment plus treatment)
reoffended within the follow-up period. Analyses separating rates of reoffending for
assessment compared to assessment plus treatment are provided in Part 2.
Dr Jennifer McCarthy, Dr Jennifer McGrail, Dr Troy McEwan, Dr Lauren Ducat, Mr Jonathan
Norton & Professor James Ogloff.
4
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0005.png
Background
The PBP is an internationally recognised community-based service that was established in 2004 and
provides assessment and treatment to individuals with high-risk problem behaviours, including harmful
sexual behaviours, violence, threatening, stalking and fire-setting. The PBP is a unique forensic mental
health service in Australia as it expands the scope of the traditional community forensic mental health
service model beyond a focus on psychopathology to other psychosocial needs. The PBP does not require
clients to have a diagnosable mental disorder or current legal order to attend. Instead, clients are accepted
to the PBP if they are engaging in a problematic behaviour that has the potential to cause harm to the
community and to the client themselves. Clients of the PBP engage in a wide range of problem behaviours
including harmful sexual behaviours such as rape, child molestation, internet child pornography use,
exhibitionism, and actual or threatened violence. The PBP is the only service in Victoria to provide specialist
assessment and treatment to stalkers, fire-setters and individuals engaging in unreasonable complaints.
In 2012 Forensicare committed to a comprehensive research evaluation of the program. This report is the
first of a series investigating the operation of the PBP and its efficacy.
Part 1: Describing Problem Behaviour
Program Clients
provides a summary of the clients who were referred to the PBP for assessment or
assessment and treatment between January 2006 and January 2011. This is part of the first evaluation of
the PBP since its inception and this report provides necessary descriptive information regarding referral
sources, presenting problem behaviours, mental health diagnoses and history, criminal history, length and
type of treatment received and general recidivism of PBP clients.
While the term ‘problem behaviour’ is used, it
should be noted that individuals do not have to have
committed
a violent or other criminal act to be referred for treatment. For example, a number of individuals
are taken on for treatment for violent or homicidal thoughts, making threats, problematic sexual thoughts
or urges, or internet child pornography. The PBP plays a pivotal role in the assessment and treatment of
client groups that are often unable to access treatment elsewhere, and whose behaviours impact on
community safety. The results of this evaluation will inform service development to ensure the program is
employing best practice in the assessment, treatment and management of problem behaviours, as well as
meeting the needs of key stakeholders in reducing harm to victims, the community and clients.
Results
Referral information
During the five year period of the study, 824 individuals were assessed and/or treated at the Problem
Behaviour Program. Figure 1 depicts the number of individuals seen by the PBP per year (note that 2011 is
not represented as the study ended in January of that year by which time only 14 referrals had been made).
200
150
100
50
0
2006
2007
2008
2009
2010
Number seen per year
Figure 1. Number of individuals assessed per year at the PBP
Clients were on average 36.9 years old (range: 15-83; SD 12.7) at the time of the assessment or treatment
episode with PBP. In total, 25 (3%) clients were 18 or under
i
at the time of PBP contact (15 years: 1; 16
5
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0006.png
years: 4; 17 years: 11; 18 years: 9). The vast majority were male (731, 88.7%; female 91, 11%) and did not
self-identify as Aboriginal or Torres Strait Islander (774, 93.9%).
During the study period 610 (74%) clients were seen for an assessment only, while 214 (26%) were taken
on for treatment. Figure 2 depicts the number of individuals either assessed or referred for treatment per
year.
160
140
120
100
80
60
40
20
0
2006
2007
2008
2009
2010
Assessment Only
Treatment
Referred but never commenced
treatment
Figure 2. Number of individuals seen for assessment only compared with those taken on for treatment, or referred for
treatment but never attended.
Referrals came from a range of services (depicted in Table 1.), although the majority were from community
corrections services or Area Mental Health Services. A small number of individuals were referred from a
number of other services, including Thomas Embling Hospital, the Adult Parole Board, Youth Criminal
Justice Services, private psychological services, prisons, brain injury services, mental health court liaison
service and legal services. The purpose of the referral was mainly primary consultation (627, 75.7%),
secondary consultation (186, 22.5%) and tertiary consultation (12, 1.4%).
Table 1. Number of individuals referred by various sources
Type of service
Community Corrections Services
Area mental health services- Inpatient
Area mental health services- Outpatient
Self-referral
Youth mental health services
Private psychiatric services
DHS- child protection
Other
GP
Other community health services
Number (%)
346 (41.8%)
112 (13.5%)
134 (16.2%)
53 (6.4%)
35 (4.2%)
25 (3.0%)
20 (2.4%)
26 (3.1%)
13 (1.6%)
13 (1.6%)
6
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
Presenting problem behaviours
Violence was the most common referral reason (384, 46.6%), followed by harmful sexual behaviour (266,
32.3%), stalking (213, 25.8%), firesetting (52, 6.3%), and other, for instance gambling and persistent
complainers (95, 11.4%). In understanding these figures, it is noted that a minority of the sample were
referred for multiple problem behaviours (131, 15.9%). For referrals related to violence, specific referral
reasons included: violent behaviour (236, 28.6%), violent or homicidal ideation (74, 9%) and threats (74,
9%). Specific harmful sexual behaviours included: problematic sexual behaviour (209, 25.4%), internet child
pornography (24, 2.9%), and problematic sexual thoughts or fantasy (33, 4%).
Mental health information
Primary and secondary diagnoses made by PBP clinicians at the time of assessment are shown in Table 2.
As these diagnoses were made between 2006 and 2011, they are reported here according to DSM-IV-TR
Axis I and Axis II categories; however, DSM 5 no longer makes these distinctions.
7
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0008.png
Table 2. Primary and secondary diagnoses made by clinicians at time of assessment
Disorder
Primary
Diagnosis in
Assessment
only group N
(%)
188 (30.5)
202 (33.1)
28 (4.6)
76 (12.5)
28 (4.6)
16 (2.6)
24 (3.9)
11 (1.8)
22 (3.6)
11 (1.8)
4 (0.7)
10 (1.6)
2 (0.3)
3 (0.5)
36 (5.9)
28 (4.6)
1 (0.2)
6 (1)
23 (3.8)
Primary
Diagnosis in
Treatment
group N (%)
Total Primary N
(%)
Total
Secondary
N (%)
Axis I
None
Psychotic disorders
Bipolar affective
Depressive
Anxiety
Paraphilia
Substance misuse
Developmental disorder
Cognitive disorder
Other (DID, eating, impulse,
adjustment)
Mood NOS
Axis II
None
Personality Disorder
Paranoid
Schizoid/schizotypal
Antisocial/Psychopathic
Borderline
Histrionic
Narcissistic
Other (avoidant,
dependent, obsessive-
compulsive)
Total Personality Disorder
Intellectual disability
Axis II traits
Paranoid
Schizoid/schizotypal
Antisocial
Borderline
Histrionic
Narcissistic
Psychopathic
Other (avoidant, dependent,
obsessive-compulsive)
Total Personality Disorder
55 (25.7)
30 (14)
6 (2.8)
28 (13.1)
7 (3.3)
63 (29.4)
9 (4.2)
6 (2.8)
2 (0.9)
7 (3.3)
1 (0.5)
2 (0.9)
2 (0.9)
3 (1.4)
15 (7)
10 (4.7)
0
1 (0.5)
11 (5.1)
243 (29.3)
232 (28)
34 (4.1)
104 (12.6)
35 (4.2)
79 (9.6)
33 (4)
17 (2.1)
24 (2.9)
18 (2.2)
5 (0.6)
670 (80.9)
4 (0.5)
6 (0.7)
51 (6,1)
39 (4.7)
1 (0.1)
7 (0.8)
34 (4.1)
658 (79.5)
6 (0.7)
5 (0.6)
18 (2.2)
31 (3.7)
27 (3.3)
40 (4.8)
12 (1.4)
17 (2.1)
11 (1.3)
3 (0.4)
807 (97.4)
0
1 (.01)
3 (0.4)
7 (0.8)
1 (0.1)
2 (0.2)
4 (0.6)
99 (16.3)
13 (2.1)
42 (19.6)
3 (1.4)
142 (17)
16 (1.9)
18 (2.2)
2 (0.2)
0
0
5 (0.8)
5 (0.8)
0
1 (0.2)
0
6 (1)
17 (1.9)
2 (0.9)
4 (1.9)
8 (3.7)
3 (1.4)
0
1 (0.5)
1 (0.5)
10 (4.7)
29 (13.8)
2 (0.2)
4 (0.5)
13 (1.6)
8 (1)
0
2 (0.2)
1 (0.1)
17 (2.1)
47 (5.7)
1 (0.1)
1 (0.1)
3 (0.4)
4 (0.5)
0
3 (0.4)
0
2 (0.2)
14 (1.7)
A small number of individuals were diagnosed with 3-4 Axis I disorders at time of assessment; the majority
of the third and fourth diagnoses related to developmental or substance use disorders. When individuals
did not meet full criteria for a personality disorder but were exhibiting personality traits consistent with the
personality disorders represented in Axis II of the DSM-IV-TR, clinicians recorded the presence of these
traits only. These are also shown in Table 2.
8
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0009.png
Treatment clients
The length of the episode of treatment at the PBP ranged from 0 to 38 months (M = 7.29, SD = 7.76, median
= 5 months). Of those who were taken on for treatment 63 (29.4%) completed the course of treatment, 19
(8.9%) were still in treatment at the time of the study collection, and the remainder did not complete
treatment for a range of reasons. Refer to Figure 3 for the break-down. For those who dropped out of
treatment the reasons were diverse: Failure to attend (68, 31.8%), being taken into custody (10, 4.7%),
correctional order expiring (3, 1.4%), hospitalisation (1, 0.5%), and other reasons (9, 4.2%).
25
20
Never attended
15
Failure to engage
Drop out
10
Completed
Ongoing
5
0
2006
2007
2008
2009
2010
Figure 3. Treatment outcomes by year
Based on the referral question and outcome of the assessment a range of targets were identified as
priorities for treatment. These were both offence-specific and offence-related targets. Many clinicians
identified a number of targets; up to four were recorded for the purposes of this evaluation. Treatment
targets by treatment clients are shown below in Table 3.
Table 3. Percentage of treatment clients by specific targets
Target 1
N= 197 (%)
Offence-specific
Cognitive distortions
Victim empathy
Fantasy work
Violence reduction
Antisocial attitudes
Risk management
Other
Offence-related
Social skills
Emotion regulation
Prosocial relationships
Substance misuse
Coping skills
ADLs
Suicide
Other
60 (28)
12 (5.6)
28 (13.1)
4 (1.9)
1 (0.5)
16 (7.5)
1 (0.5)
14 (7.1)
49 (22.9)
3 (1.4)
1 (0.5)
8 (3.7)
0
0
0
Target 2
N= 182 (%)
22 (10.3)
18 (8.4)
14 (6.5)
7 (3.3)
1 (0.5)
19 (8.9)
1 (0.5)
26 (14.3)
27 (12.6)
3 (1.4)
11 (5.1)
24 (11.2)
5 (2.3)
1 (0.5)
3 (1.4)
Target 3
N = 138 (%)
14 (6.5)
3 (1.4)
6 (2.8)
2 (0.9)
0
22 (10.3)
1 (0.5)
28 (20.3)
28 (13.1)
7 (3.3)
5 (2.3)
17 (7.9)
2 (0.9)
2 (0.9)
1 (0.5)
Target 4
N= 71 (%)
4 (1.9)
2 (0.9)
5 (2.3)
1 (0.5)
3 (1.4)
9 (4.2)
1 (0.5)
13 (18.3)
5 (2.3)
6 (2.8)
3 (1.4)
8 (3.7)
3 (1.4)
1 (0.5)
7 (3.3)
9
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0010.png
Criminal History and Recidivism
Criminal history was obtained from Victoria Police’s Law Enforcement Assistance Program (LEAP) for all
individuals in the study, including offences committed two years prior to the study period (01.04.2004) and
follow-up data for offences committed up to 30.06.2012. These data show that in addition to being referred
for a diverse range of problem behaviours, individuals seen at the PBP had diverse offending histories. While
573 (69.5%) were not versatile offenders (index problem plus two or fewer offence types), a sizeable
proportion were. 101 (12.3%) clients were moderately versatile (index problem plus 3-4 other offence types
in history) and 150 (18.2%) were highly versatile (index problem plus five or more offence types). On
average, clients had a mean of 12.5 (SD = 16.8; range 1 - 112) total charges and a mean of 2.5 (SD = 6; range
0
–
53) subsequent charges after the PBP contact. There were no significant differences in the number of
offences committed by assessment or treatment clients.
More than one third (n= 299, 36.3%) of the individuals were charged for an offence after any contact with
the PBP
ii
. Each
individual’s most serious offence committed after the index episode of contact was coded
according to the Cormier Lang system (Quinsey, Rice, Harris, & Cormier, 2006). The most serious offences
individuals reoffended with included violent offences (132, 16%), breaching a legal order (30, 3.6%), sexual
assault (27, 3.3%), weapons offences (18, 2.2%), and a range of stalking, drugs, property damage, theft,
deception, threats, bad public behaviour, possess child pornography and arson offences. Further detail on
client recidivism can be found in Part 2 including analysis of subsequent offending comparing assessment
only clients versus those who received assessment plus treatment.
Mental health system contacts
Information about lifetime contact with mental health services was collected from the Department of
Health and Human Services. 744 (90.3%) individuals were registered on the statewide mental health
database (known as RAPID) and had accessed a range of services across the lifespan, with outpatient
contact being the most common. Types of services accessed are displayed in Table 4. When comparing
lifetime registration between clients referred to the PBP by mental health services and those referred via
legal channels, the former group were significantly more likely to have been registered with psychiatric
services; although the proportion amongst those referred through legal services was also high (316, 82.7%).
Table 4. Type of lifetime mental health service contact
Type of service
Outpatient
Inpatient
CCU/ongoing care
Community Treatment Order
Crisis Assessment and Treatment Teams
Child/Adolescent
Number (%)
710 (86.2)
390 (47.3)
418 (50.1)
205 (24.9)
361 (43.8)
89 (10.8)
A comparative analysis was undertaken to compare the service usage of individuals who were only seen for
an assessment to those referred for ongoing treatment. As shown in Table 5, assessment only clients were
significantly more likely than treatment clients to have had all types of mental health contact (with the
exception of mental health contact as children).
10
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0011.png
Table 5. Comparison of the type of contacts with MHS by treatment and assessment clients
Type of service
RAPID registered
Outpatient
Inpatient
Community Treatment Order
Child/Adolescent
*** p < .001
Assessment only
Number (%)
535 (64.9)
502 (60.9)
316 (38.3)
180 (21.8)
66 (8)
Treatment
Number (%)
209 (25.4)
208 (25.2)
74 (19)
25 (3)
23 (2.8)
χ²
17.92***
25.51***
18.82***
26.94***
0.001
Of those registered on RAPID, 634 (76.9%) received an Axis I diagnosis, 305 (37%) received an Axis II
Personality diagnosis and 46 (5.6%) received an Intellectual Disability diagnosis. Diagnostic information
from RAPID showed that individuals have received a range of diagnoses across the lifespan (see Table 6). A
large proportion had lifetime diagnoses of substance misuse (305, 37%). Despite the range of diagnoses
received in adulthood, only 69 (8.6%) received some form of diagnosis in childhood. Personality disorder
diagnoses (as recorded on RAPID) were predominantly of the antisocial (110, 13.3%), borderline (66, 8%)
and unspecified (58, 7%) types, with a number receiving more than one personality disorder diagnosis over
time (28, 3.4%).
Table 6. Lifetime Axis I diagnoses received by individuals
Most severe diagnosis ever received
None
Schizophrenia spectrum
Other Psychosis
Bipolar affective disorder
Depressive disorder
Anxiety Disorder
Eating Disorder
Paraphilias
Impulse control
Substance use disorder
Other
N
200
218
34
30
120
97
1
72
5
33
14
(%)
(24.3)
(26.5)
(4.1)
(3.6)
(14.6)
(11.8)
(0.1)
(8.7)
(0.6)
(4.0)
(1.7)
11
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
Conclusions
The Problem Behaviour Program was established in 2004 to fill a perceived gap in existing mental health
and justice services. Clients of the PBP are those who often cannot access other services either because the
nature of their behaviour means there are no available services (e.g., stalking, firesetting, internet child
pornography); they have mental disorder that impacts on their ability to engage with other forensic
services; their behaviour places them at risk of being excluded from mental health services and incurring
legal penalties (e.g., threats/violence); or because they do not qualify for other services due to the absence
of mental disorder. This first survey of PBP clients over a five year period between 2006 and 2011 showed
that the largest single problem behaviour exhibited by PBP clients is violence, accounting for almost half of
all referrals. Harmful sexual behaviour was present in one third of referrals, stalking in one quarter, and
firesetting in six percent. Notably, one in six clients were referred with multiple different problem
behaviours.
Nearly half of PBP referrals came from Community Corrections Services during the data collection period.
One in twenty clients self-referred and one in four had no mental health diagnosis. These figures suggest
that the PBP is meeting its stated purpose: to fill a gap by providing assistance to clients who engage in
problem behaviours but due to the absence of a diagnosable mental disorder or current legal order, cannot
access other services. Referrals to the PBP remained largely steady over the five year period of data
collection, with a trend towards increasing referrals since 2007. It should be noted that since the end of the
data collection period, the relationship between the PBP and Corrections Victoria has been formalised and
Corrections clients now constitute an even larger proportion of the PBP workload than is represented in
this report.
The majority of clients seen by the PBP have had contact with the public mental health system.
Furthermore, approximately three quarters of clients were diagnosed with an Axis I psychiatric disorder
and a quarter with an Axis II disorder. This highlights that whilst the single largest group of referrals come
from the legal system, the PBP is responding to clients with major mental health difficulties. Given high
rates of psychiatric disorder in PBP clients, there is a clear need for staff to have clinical training and skills
in addition to specific forensic skills and training. It should be noted that the figures regarding the
prevalence of personality disorder are likely to underestimate the true prevalence of those disorders
amongst PBP clients, as personality disorders often go unrecorded on RAPID. Studies of clients referred to
the PBP for stalking show that approximately 45% of this sub-group have identifiable personality disorder
or problematic personality traits that contribute to their behaviour.
One quarter of clients assessed at the PBP were taken on for treatment. The target population for the PBP
are those individuals whose problem behaviours place them at risk of engaging in offending behaviour.
More specifically, they pose a high risk of harm to the community, their needs cannot be met elsewhere
and they show some sign of willingness to engage in treatment and are likely to benefit from treatment. Of
this group, approximately one third completed treatment as recommended, with the majority dropping out
of treatment or treatment being terminated due to failure to engage. Those who dropped out did so for a
range of reasons including simply failing to attend, ceasing attendance after the expiry of their legal order,
being re-apprehended or being hospitalised. In the time since 2011 a range of strategies have been put in
place to improve client attendance, including a text message service to remind clients of appointments and
increased liaison with Community Corrections staff. Despite these efforts, treatment drop-out remains an
area for further consideration and review. There are a range of issues that require additional attention
including the effectiveness of current administrative procedures, closer consideration of suitability of
clients for treatment, and routine exploration of issues related to the engagement of clients in treatment
(e.g., motivation to engage).
12
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0013.png
A range of treatment targets were identified by clinicians, including both offence-specific (e.g., antisocial
attitudes) and offence-related (e.g., emotion regulation) targets. Anecdotal recognition by clinicians of the
need to address offence-related targets of interpersonal skills and emotion regulation has resulted in the
establishment of group programs (e.g., the Handling Anger Wisely and Positive Relationships groups).
Perhaps one of the more interesting findings from this review of PBP clients is that approximately one
quarter are versatile offenders who engage in a wide range of offences aside from those arising from the
problem behaviours leading to referral. Moreover, almost 40% of PBP clients reoffended in the 2
–
6 year
follow-up period. This reflects broader reoffending rates reported in the literature and indicates that while
PBP clinicians clearly have a responsibility to understand and address the specific problem behaviours that
led to referral, they also need to be mindful of the broader criminogenic needs of their clients.
i
The PBP provides service to adults; however, over the time period of the study, decisions have on occasion been
made to accept referrals pertaining to clients aged under 18. These exceptional cases have generally involved
circumstances where the person in question was near the age of 18, and/or displayed problem behaviours directly
and clearly within the scope of service of the program and with no alternative expert assessment options available.
ii
25 cases had no reoffending data available and were excluded from the denominator in these analyses
13
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0014.png
PART 2: Offending and mental health outcomes for
individuals assessed and treated at the Problem
Behaviour Program (PBP)
Key Findings
The PBP is effective in reducing reoffending. Two-thirds of all clients who received
assessment and/or treatment with the PBP did not receive any further charges after PBP
contact. PBP clients had on average 4.9 offences prior to contact with the PBP; post PBP
contact this dropped to 2.5 offences.
For individuals who re-offended (33% of total), the average time to reoffence was just
over one year (14.41 months). Of these, two-thirds had no change or decrease in offence
severity from their pre-referral offence, and only one third (i.e., one ninth of the total
sample) had an increase in severity of offence.
Clients completing treatment reoffended at significantly lower rates than other clients.
Average time to reoffence for the treatment group (785 days) was significantly longer
than for the other client groups. A higher proportion of clients in the
treatment
group
compared to the
failed to attend
group had a reduction in the severity of their
reoffending.
Contact with the PBP also resulted in more positive mental health outcomes for clients.
Overall, there was a significant reduction in the number of outpatient contacts following
service provision from the PBP.
Even for those clients who were seen for
assessment
only, there was a significant
reduction in inpatient admission, CATT contact and CTO status.
There was no significant difference in inpatient contact for the other clients (treatment
and
failed to attend
groups); this may be due to the more complex nature of these higher
risk clients. Overall the
failed to attend
group continued to utilise mental health
services at relatively the same level as prior to contact with the PBP, suggesting that
assessment alone is not sufficient to result in positive mental health outcomes for many
higher risk clients and they require specialist treatment.
PBP treatment clients report high levels of satisfaction with the service. The majority of
clients surveyed report that they have had a positive experience at the PBP and felt
supported. More specifically, they reported that the PBP has helped them to understand
their problem behaviour, reduce this behaviour, understand unhelpful thinking habits,
and feel better.
The very positive results for those who attended and completed treatment suggests that
an increased focus of the work of PBP clinicians should be on engaging clients to the
extent possible to retain them in treatment, including further exploration of the reasons
that clients cease treatment prior to an agreed end and the kinds of barriers that exist
that prevent successful completion.
14
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
Background
Part 2 of this Report looks in detail at the treatment and outcomes for PBP clients. The nature of the
PBP makes it challenging to evaluate. By its very nature “successes” are cases in which risk is successfully
managed and nothing happens. That is, there is no further instance of the problem behaviour and no
further involvement with the criminal justice system. The current evaluation therefore examines the
reoffence rates and mental health system contacts of PBP clients. Importantly the evaluation also
sought feedback from clients about their experience of treatment at the PBP.
In the absence of the PBP, individuals who are at risk of engaging in complex criminal behaviour, but
have not yet done so, would not be able to access specialist assessment and intervention unless or until
they committed and were prosecuted for an offence. Even then, their access to correctional
rehabilitation services may be hampered by the presence of mental disorder or by a lack of service
specific to their behaviour. Acceptance into the PBP for many individuals is also often a first step
towards establishing links with other key services such as mental health, community health, or other
social services. “By taking a lead clinical role in the management and treatment of high-risk
behaviours,
the PBP strengthens referral pathways to those services that may have previously been apprehensive
about taking on such clients” (McEwan et al.,
2014, p. 363).
Treatment at the PBP is evidence-based, adhering to
Andrews and Bonta’s (2010) risk, needs and
responsivity principles and using structured risk assessment to identify clients who present as moderate
or high risk and so are appropriate for behaviour-specific treatment. The modality of treatment,
although somewhat eclectic, is based on cognitive behaviour therapy and relapse-prevention
approaches. Relevant risk factors
(i.e., ‘criminogenic needs’)
that may be pertinent treatment targets
are identified from the risk assessment tools, supplemented with functional analyses of the problem
behaviour. Treatment is oriented towards the cessation of the problem behaviour and the formulation
is used to prioritise treatment targets and responsivity factors. This allows treatment plans to focus on
the client’s criminogenic needs, but also to be individualised and tailored to the
context of the specific
problem behaviours (McEwan et al ., 2014). Clients who attend for treatment at the PBP are deemed
to have completed treatment once they have progressed in their treatment gaols to a satisfactory level,
as agreed upon by both the clinician and client.
A typical assessment (primary consultation) lasts between two and six hours duration. This takes the
form of a semi-structured interview, covering areas such as childhood, adolescence and adulthood,
educational and employment history, relationship and sexual history, psychiatric and medical history,
drug and alcohol use, and offense history. It is essential that corroborative history is obtained, including
criminal history reports, police charge sheets and previous mental health assessment/reports. Further
collateral information is frequently obtained from the family or friends of the client, with the client’s
consent. In the case of a psychological assessment, psychometric testing is likely to be conducted. This
testing is tailored to the individual and their presenting problem behaviour. Typically this comprises of
some measure of socially desirable responding, personality testing, and other supplementary tests as
required. In most cases a structured risk assessment using a set of professional judgement guidelines
(e.g. the HCR-20
V3
[Douglas, Hart, Webster & Belfrage, 2013] or RSVP [Hart et al., 2003]) is also
completed and informs the results of the wider assessment (McEwan et al 2014).
This comprehensive assessment process allows the assessor to develop an explanatory formulation of the
problem behaviour, including a functional analysis, which considers the psychological, psychiatric, and
social determinants of the behaviour. A written report is provided to the referrer outlining the
conclusions of the assessment and providing recommendations for management. In some cases, where
risk level and lack of other support services warrants, this may include a recommendation to attend the
PBP for ongoing treatment (McEwan et al., 2014, p. 365).
15
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0016.png
Purpose of the evaluation
A primary aim of the evaluation is to assess the impact of the PBP in reducing offence-specific recidivism.
Part 2 of the Report provides an analysis of offending patterns before and after contact with the PBP for
all individuals who were assessed at the PBP between January 2006 and January 2011
1
. It also discusses
the mental health outcomes of individuals who were assessed and/or treated at the PBP to determine
if treatment has an impact on patterns of mental health service usage. Finally, the report examines
characteristics of clients who receive treatment at the PBP and includes direct feedback from PBP clients
to further understand their experience.
The results of this evaluation will inform service development to ensure the program is employing best
practice in the assessment, treatment and management of problem behaviours, as well as meeting the
needs of clients in reducing harm to victims, the community and clients.
Key Questions
1. Is contact with the PBP effective in reducing the frequency, nature and time to reoffending?
2. What impact does receiving services from the PBP (assessment and/or treatment) have on
mental health outcomes for clients?
3. Who are PBP treatment clients, what is the effect of treatment, and what is their experience?
Definition of terms
Reoffence
is measured by charges incurred after the index assessment date, including charges that
were laid during the treatment period.
Time at risk
is the period of time between the date of assessment at the PBP and the date the individual
first reoffended or the end of follow up period, whichever comes first. Time incarcerated for other
offences was subtracted from this period. We also removed people who died from the follow-up
analyses, ensuring that those in the community were still ‘at risk’ for re-offending.
Methodology
Data linkage procedure
To determine rates of reoffending and mental health service usage, Forensicare data were linked with
data from the Department of Health (CMI), Victoria Police (LEAP), Corrections Victoria (PIMS) and the
National Coronial Information Service (NCIS). Cases were matched using identifying information (first
name, surname, aliases, date of birth, age range and gender). Once linked, the identifying information
was removed for subsequent analyses.
Sample Descriptives
During the study period from January 2006 until January 2011, 901 individuals were assessed and/or
treated at the PBP. Data could not be collected for all clients, reducing the final sample to 824.
Individuals were on average 36.9 years old (range: 15-83; SD 12.7) at the time of the assessment or
treatment episode with PBP. In total, 25 individuals were 18 and under at the time of PBP contact (15
years: 1; 16 years: 4; 17 years: 11; 18 years: 9). The majority were male (731, 88.7%; female 91, 11%)
and did not self-identify as Aboriginal or Torres Strait Islander (774, 93.9%). A full account of the sample
is contained in Report 1:
Describing Problem Behaviour Program Clients.
1
Individuals referred to the PBP who failed to attend for assessment are not discussed in this report.
16
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
Question 1: Is the PBP effective in reducing the frequency, nature and time to reoffence?
To answer this question, data were collected of all individuals who were assessed at the PBP between
2006 and 2011. Clients were grouped into three categories;
1.
Not recommended
(i.e. assessment only): Those who were assessed and not recommended for
PBP treatment (n=612)
2.
Failed to attend:
Those assessed and recommended for treatment but who failed to attend
(n=130). This includes those who failed to attend any treatment sessions as well as those who
dropped out of treatment prior to completion. It is noted that clients who attend for treatment
at the PBP are deemed to have completed treatment once they have progressed in their
treatment gaols to a satisfactory level, as agreed upon by both the clinician and client.
3.
Treatment:
Those assessed and recommended for treatment who attended treatment (n=84).
This includes those who were determined to have completed treatment (n=65) as well as those
who (n = 19) remained in treatment at the time of the study, with the latter group comprising
2.3% of the total treatment group.
Offending data were collected from Victoria Police for all clients from 2004 and 2012. This ensured that
all clients had at least a two year period prior to assessment at the PBP, and gave a follow-up time of 6
months to 7.8 years (mean 3.89, SD 1.49).
A number of between-group and within-individual comparisons were made to determine the efficacy of
PBP assessment and treatment in reducing reoffending. Individuals’ reoffending rates and the severity
of offending were compared before and after the PBP assessment. Reoffending rate was calculated
using the average number of offences per month in the 2 years prior to PBP assessment and the 6
months to 2 years after PBP assessment. Severity of offending was coded using the Cormier-Lang System
(Quinsey, Rice, Harris & Cormier, 2006) and severity of offending before and after PBP contact was
compared and analysed.
In addition to evaluating the number of people who reoffended and the types of offences, time to
reoffence was examined using Kaplan-Meier survival analyses (controlling for time in custody). Survival
time comparisons were made between the
not recommended
and
failed to attend
groups, and between
the
failed to attend
and
treatment
groups (neither analysis controlled for time in treatment). Log rank
statistics were used to determine statistically significant between-group differences in time to reoffence.
Question 2: What impact does contact with the PBP (assessment and/or treatment) have on
mental health outcomes for clients?
Mental health information was taken from the Victorian Case Psychiatric Register (VPCR, otherwise
known as Client Management Index CMI). All information regarding contacts with the public mental
health service (across the lifespan) was collected, including: the service contacted (e.g., CATT, Inpatient
Acute), the specific program (e.g. PARC or EPPIC), types of contact (e.g. direct), duration of contact, and
whether the person has been or was placed on a CTO. Diagnostic information was also coded (all
diagnoses are recorded on CMI using ICD-10).
Primary psychiatric diagnoses were coded into categories, replicating previous research undertaken in
the CFBS (Cutajar et al., 2010; Ducat, Ogloff & McEwan, 2013; Short et al., 2010; Wallace et al., 2004).
The categories included: schizophrenia spectrum (e.g., schizophrenia, schizoaffective disorder, shared
psychotic disorder), other psychotic (e.g., delusional disorders, and unspecified non-organic psychosis),
bipolar affective disorder, depressive disorder, anxiety, eating disorders, paraphilias, impulse control,
and substance use (as primary and any
lifetime). A ‘substance-use disorder’ was defined as any type of
17
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0018.png
substance abuse, substance dependence, or substance-induced disorder (such as substance induced
psychosis), excluding nicotine-related disorders.
Given the large number of potential diagnoses an individual may receive over a lifetime, diagnoses were
only coded when they were upheld in 75% of the diagnoses given, or there was a clear diagnostic
progression over time resulting in a clear diagnosis. This method has been used by several studies and
demonstrates good reliability (Bennett et al., 2009; Krupinski et al., 1982; Short et al., 2010).
Frequency and nature of mental health service usage was compared for the period prior to and after
PBP assessment.
Question 3: Who are PBP treatment clients, what is the impact of treatment, and what is their
experience?
A repeated-measures design using the sign test was used to examine the changes in offending and
mental health service usage pre- and post-treatment at the PBP. The sign test is used to determine
whether there is systematic change over time in offending and mental health service usage, depending
on treatment
2
. The sign test compares differences in paired values before and after treatment, and
examines the overall direction of change in the entire group. Where treatment has not affected
offending or service usage, the number of people with improvements should be approximately equal to
the number of people without improvements. A significant sign test indicates that treatment has had
some effect across the entire group.
Treatment clients attending the PBP during a two week period from 17 November 2014 to 28 November
2014 were asked to complete a written survey in which they rated their experience of the PBP and their
perceptions of its effectiveness in targeting their problem behaviours. Participation in the survey was
voluntary. Twenty-six clients attended for PBP treatment during the data collection period and 15
completed the survey (58% response rate). Those who participated were offered the opportunity to
participate in a structured telephone interview regarding their experience of the program. Three clients
consented to the telephone interview (20%).
Results
The reasons for referral to the PBP are included in Part 2 in greater detail. In summary, violence
accounted for 46% of overall referrals; harmful sexual behaviour for a further third and stalking for a
quarter (with a number of clients referred for multiple problem behaviours). Referrals to the PBP came
from a range of sources, and for a range of reasons, as displayed in Table 7.
2
The Wilcoxon signed-rank test was not used because the assumption of symmetrical distribution of differences
between two related groups was violated.
18
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0019.png
Table 7. Breakdown of referrals to the PBP by referral source and presenting problem behaviour(s)
Referral
Source
Community
Corrections
N (%)
Mental
Health
Services
N (%)
Self-Referral
N (%)
Other
N (%)
Total
a
N (%)
a
Number of
referrals
346 (42.1)
Violence
Harmful
sexual
behaviour
82 (23.7)
Stalking
134
(38.7)
Firesetting
Multiple
Problem
Behaviours
c
67 (19.4)
155 (44.8)
18 (5.2)
280 (33.9)
151 (53.9)
94 (33.6)
53 (18.9)
23 (8.2)
45 (16.1)
51 (6.2)
144 (17.4)
821 (99.6 )
b
9 (17.6)
61 (46.6)
384 (46.6)
37 (72.5)
50 (38.2)
266 (32.3)
5 (9.8)
20 (15.3)
213
(25.8)
2 (3.9)
9 (6.9)
52 (6.3)
4 (7.8)
14 (10.7)
131 (15.9)
Total is greater than 100% as clients can be referred for multiple problem behaviours
b
Includes three
cases with unknown referral source so total percentage ≠ 100%
c
Cases in this column are also represented in the relevant specific problem behaviour columns
The vast majority of PBP clients have had previous contact with the Victorian public mental health
system (90%). PBP clients were diagnosed with a range of psychiatric disorders including psychotic
disorders (28%), depressive disorders (15%) and paraphilias (13%). This suggests that, although not
based on a typical forensic mental health service model, and not a requirement for service, the PBP
nonetheless routinely provides assistance to mentally disordered offenders and others experiencing
serious mental health difficulties.
Question 1: Is the PBP effective in reducing the frequency, nature and time to reoffence?
Frequency of offending (pre- and post-contact with the PBP) across the Not Recommended,
Treatment and Failed to Attend groups
Five hundred and twenty-five (63.7%) individuals did not reoffend after the PBP contact and almost half
the sample had no formal offence history in the two years prior to the index PBP contact (400, 48.5%).
A large proportion of reoffenders had a prior offence history (241, 80.6% of reoffenders). Conversely,
only 58 (19.4% of reoffenders) individuals without a prior offending history went on to reoffend (χ² =
159.59,
p <
0.001, Φ = 0.44).
For those with an offending record (either pre- or post-PBP contact; n = 495, 60.1%), their offending
careers were diverse. Nearly 20% were versatile offenders (five or more offence types ever), 12.3%
were moderately versatile (3-4 offence types), and 69.5% were not versatile (1-2 offence types). A small
proportion of individuals had only one offence type (108, 13.1%), while 329 (39.9%) individuals had no
offences ever recorded during the data collection period. On average individuals had 4.9 (SD 10.5, range
0-110) prior offences recorded in their criminal history for the two years prior to the index PBP contact.
Figure 4 shows criminal versatility by referral problem behaviour.
19
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0020.png
100
90
80
70
60
50
40
30
20
10
0
4.9
11.7
20.8
0.8
6
19.9
2.8
8
22.5
1.9
9.6
11.5
High versatility in reoffending
Moderate versatility in
reoffending
%
70.7
57.8
64.3
69.2
Low versatility in reoffending
No reoffence
Violence
Harmful sexual
behaviour
Stalking
Firesetting
Figure 4. Criminal versatility in subsequent offending by referral problem behaviour
Two hundred and ninety-nine (36.3%) clients were charged with a subsequent offence; including violent
offences (16%); breaching a legal order (3.6%); sexual assault (3.3%) and weapons offences (2.2%). The
average time to reoffence was 14.41 months (range 0-69, SD 14.53). Reoffending patterns by referral
problem behaviour were examined. For those with a presenting problem (PP) relating to violence (n =
384), 114 reoffended with violence (29.7% of individuals with PP relating to violence). Of these,
approximately one fifth committed one to two other offence types in addition to violence, 11.7%
committed 3-4 other offence types in addition to violence and the remaining 5% were considered highly
versatile offenders, committing 5-6 other offence types in addition to violence. Of the total 266
individuals who were referred for harmful sexual behaviour or ideation 22 (8.3%) went on to reoffend
with a sexual or child pornography offence. This group was comparatively less versatile than the violent
offenders. While 20% committed 1-2 other offence types, only 6.7% committed 3-4 other types, and
only 1% were highly versatile. Of the 213 who were referred for stalking 27 (12.7%) reoffended by
stalking, and of the 52 referred for firesetting 3 (5.8%) reoffended by arson. Versatility statistics were
not calculated for these groups.
Across the entire sample, the average number of offences post-PBP contact was 2.6 (SD 6.4, range 0-
66). To take account of time available to offend (time
at risk),
both pre-and post-assessment, the
average number of offences committed per month were calculated. Clients committed on average 0.09
(SD 0.19, range 0-1.83) offences per month prior to contact with the PBP, which reduced significantly to
0.06 (SD 0.13, range 0-1.02) offences per month after contact with the PBP (Sign
test =
-7.11,
p
< 0.001).
Time to reoffence across the Not Recommended, Treatment and Failed to attend groups
Amongst those who received subsequent charges (n = 299; 36.3%), time to reoffending differed
significantly depending on whether or not clients were referred to and attended for treatment (-2log
χ²
= 6.55,
p <
0.05). As shown in Figure 2, the
treatment
group took significantly longer to reoffend (average
23.52 months, range
–0-68,
SD 18.82) than the other two groups (not
recommended:
average 13.87
months, range 0
–
69, SD 14.48;
failed to attend:
average 17.05 months, range 0 -62, SD 15.19). In Figure
5, vertical lines represent censored cases. Cases are censored or excluded from the analysis where the
line crosses the survival curve, as that is the maximum follow-up time available for that case.
20
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0021.png
Figure 5: Survival curves representing days to reoffence for each of the study groups
Severity of offending (pre- and post-contact with the PBP) across the Not Recommended,
Treatment and Failed to Attend groups
Just over half of the sample (435, 52.8%) had no change in the severity of offending pre- and post-
assessment. This can likely be accounted for the large proportion of the sample who had no offending
history. Just over one third of the sample (382, 34.2%) showed a decrease in severity (for example
committing theft from assault), while the remaining 107 (13%) showed an increase (for example
progressing from stalking to sexual assault). When individuals had no prior offending there was mostly
no change in severity (n = 342, 85.5%), with the remaining 14.5% (58) showing an increase in the severity
of offending. For those who reoffended after the index PBP contact, the most common types of
reoffence were violence (138, 46.2%), breach of a legal order (134, 44.8%) and theft (92, 30.8%).
Differences in reoffending for clients from different referral sources
As shown in Table 8, of those clients who reoffended, there were some differences in patterns of
reoffending depending on referral source.
21
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0022.png
Table 8. Comparison of reoffending rates by referral source
Number of
Reoffenders
N (% of referral
source)
156 (45.1)
86 (30.7)
57 (31.3)
Average months to
reoffence
Referral Source
Number of subsequent offences
M
SD
9.80
7.39
9.06
Range
1 - 66
1 - 31
1 - 43
Community Corrections
Area Mental Health Services
(including youth)
Other
8.01
5.53
6.98
15.32
15.36
14.84
As can be seen in Table 8, clients referred from Community Corrections who reoffended committed
significantly more offences after contact with the PBP than clients from Area Mental Health Services (U
=
5181.5,
p
< 0.01,
θ
= 0.61). However, there were no significant differences in the time to reoffence
when comparisons are made by referral source (CCS, AMHS, Other), nor when only comparing CCS with
AMHS clients. As can be seen in Figure 6 the severity of offending post-assessment between the CCS and
AMHS groups is largely consistent with the overall group findings. While not significant, proportionally
more of the AMHS clients experienced no change in severity than the CCS clients, a factor that is likely
explained by the higher proportion of AMHS without an offending record. Overall, these between group
differences were not significant
(χ² =
4.33 (df 2),
p <
0.12,).
100
90
80
70
60
CCS
AMHS
All
%
50
40
30
20
10
0
Reduction in severity
No change in severity
Increase in severity
Figure 6. Comparison of severity of offending, pre- and post-PBP contact, by referral source.
Comparison of reoffending data for Not Recommended and Failed to attend groups
To determine whether the PBP is identifying appropriate clients for treatment, reoffending patterns
between the
not recommended
and
failed to attend
clients were compared. In most cases, individuals
who are assessed as moderate to high risk of offending behaviour where no other appropriate services
are available are recommended for treatment following the primary consultation. As such
failed to
attend
clients are moderate to high risk clients who are non-compliant with treatment and would be
expected to offend more often than clients assessed and not recommended for treatment (in most cases
lower risk clients or those engaged with other appropriate services).
22
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0023.png
Number and frequency of offending (pre- and post-contact with the PBP) for Not
Recommended and Failed to attend groups
Overall, 46.2% of the
failed to attend
group reoffended, compared to 35.3% of the
not recommended
group (χ² = 5.41,
p <
0.05,
ϕ
= 0.09). The
not recommended
group also had significantly fewer offences
post-PBP assessment (m = 2.40, range 0
–
49, SD = 5.87) compared with the
failed to attend
group (m =
3.93, range 0
–
66, SD = 9.05) (U
=
34874,
p
< 0.01,
θ
= 0.44). The prior and reoffending patterns of the
two groups were compared to determine if assessment at the PBP impacts upon rates of reoffending.
In the
not recommended
group, offences per months fell from an average of 0.10 (SD 0.19) to an average
of 0.06 (SD 0.14) after the PBP assessment (Sign
test
= -6.5,
p < 0.001).
In contrast, there was no
significant change in the number of offences pre- and post-PBP episode for those who were referred for
treatment but did not attend or dropped out of treatment prior to completion (Before: m = 0.09, SD
0.19; after: m = 0.09, SD 0.20;
sign test =
-1.30,
p =
0.20).
Time to reoffence for Not Recommended and Failed to attend groups
When examining time from assessment to reoffence for the
not recommended
and
failed to attend
groups there was no significant difference, as seen in the survival curve in Figure 7 (-2log
χ² = 2.58,
p =
0.11). Individuals in the
not recommended
group had an average time to reoffence of 13.87 months
(range 0
–
69, SD 14.48), compared with the
failed to attend
group’s average time of 17.05
months
(range 0 -62, SD 15.19).
Figure 7: Comparison of the survival curves representing days to reoffence for the
Not Recommended
and
Failed
to Attend
groups.
23
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0024.png
Severity of offending (pre- and post-contact with the PBP) for Not Recommended and Failed
to Attend groups
Comparison of the severity of offending pre- and post-PBP contact in the
not recommended
and
failed
to attend
groups, as shown in Figure 8, indicated that the majority of individuals in both groups had no
change in severity of offending, one third had a decrease, and a small proportion showed an increase.
These differences were not significant (χ² (df 2) = 4.20,
p =
0.12.
100
90
80
70
60
53.8
34.3
51.5
Not recommended
30
18.5
11.9
Failure to attend
%
50
40
30
20
10
0
Reduction in severity
No change in severity
Increase in severity
Figure 8: Comparison of severity of offending, pre- and post-PBP contact, between Not Recommended and
Failure to Attend groups
Comparison of reoffending data for Failed to attend and Treatment groups
To determine whether treatment is effective when it has been recommended by the PBP following a
primary consultation, comparison was made of the reoffending rates of clients who completed
treatment or who remained in treatment at the time of the study (treatment
group)
and those failed to
attend or who dropped out of treatment prior to satisfactory completion (failed
to attend group).
This
compares groups of similar profile in terms of risk, complexity and lack of availability of other
appropriate services. It was expected that the
treatment
group would have lower rates of reoffending
compared to the
failed to attend
group.
Approximately 25% of clients seen for assessment at the PBP were recommended to receive individual
treatment. However, 60% of these clients recommended for treatment dropped out prior to
commencement or satisfactory completion. This relates to the persistent and challenging nature of the
behaviours of many of these clients.
Number and frequency of offending (pre- and post-contact with the PBP) for Failed to attend
and Treatment groups
Overall, 59 (72%) of the
treatment
group had not reoffended by the follow-up end date. In relation to
those in the
treatment
group that did reoffend during the follow-up period (28%), this result was in
comparison to 46.2% of the
failed to attend
group who reoffended (χ² = 6.92,
p <
0.01). The
treatment
group also had significantly fewer offences post-PBP assessment (m = 1.8 range 0
–
34 SD = 5.18)
compared with the
failed to attend
group (m = 3.93, range 0
–
66, SD = 9.05) (U
=4297.5, p
< 0.01). The
prior and reoffending patterns of the two groups were compared to determine if treatment at the PBP
impacts upon rates of reoffending. In the
treatment
group, offences per month fell from an average of
24
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0025.png
.08 (SD = .00) prior to PBP assessment, to .04 (SD = .11) after PBP treatment (sign
test =
-2.01,
p =
0.05).
In contrast, there was no significant change in the number of offences pre- and post-PBP episode for
those who were referred for treatment but did not attend or dropped out of treatment prior to
completion (Before: m = 0.09, SD 0.19; after: m = 0.09, SD 0.20;
sign test =
-1.30,
p =
0.20).
Time to reoffence for Failed to attend and Treatment groups
When examining time from assessment to reoffence for the
treatment
and
failed to attend
groups there
was a significant difference, as seen in the survival curve in Figure 9 (-2log
χ² = 6.55,
p =
0.01). Individuals
in the
treatment
group had an average time to reoffence of 23.52 months (range 0
–
68, SD 18.82),
compared with the
failed to attend
group’s average time of 17.05 months (range
0 -62, SD 15.19).
Figure 9: Comparison of the survival curves representing days to reoffence for the
Treatment
and
Failed to Attend
groups.
Severity of offending (pre- and post-contact with the PBP) for Failed to attend and Treatment
groups
Comparison of the severity of offending pre- and post-PBP contact was examined by comparing the
percentage of individuals who had an increase, decrease or no change in severity of reoffence depending
on whether they completed treatment or not.
25
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0026.png
100
90
80
70
60
%
50
40
30
20
10
0
47.6
40.2
30
51.5
Treatment
Failure to attend
18.5
12.2
Reduction in severity No change in severity
Increase in severity
Figure 10: Comparison of severity of offending, pre- and post-PBP contact, between Treatment and Failure to
Attend groups
As can be seen in Figure 10 both the
failed to attend
and
treatment
groups showed a reduction or no
change in the severity of offending, while a small group showed an increase in severity. While
proportionally more of the
treatment
group reduced the severity of their offending, overall changes in
severity were not statistically significant (χ² (df 2) = 2.94,
p =
0.23).
Question 2: What impact does contact with the PBP (assessment and/or treatment) have on
mental health outcomes for clients?
This part of the report will examine the mental health service usage of individuals who were seen at the
PBP. To determine the impact of mental health service usage of individuals before and after contact
with the PBP (for all clients), contacts with the mental health system were examined.
As shown in Table 9 there was a significant reduction in outpatient contact pre-and post-PBP contact in
each of the three groups. There was a significant reduction in inpatient admissions pre-and post-PBP
contact for
not recommended
clients. Whilst there was a decrease in inpatient admissions for treatment
clients, this did not reach significance. There was a significant difference in CATT usage for
not
recommended
group and
treatment
group such that CATT usage decreased. There was a significant
reduction in CTO usage pre- and post-PBP contact for the
not recommended
group.
There was no statistical association with reoffending and having had a prior psychiatric inpatient
admission, CATT contact, or other outpatient contact prior to the PBP assessment. Reoffending was
significantly related to previous CAMHS contact, with a greater number of reoffenders having had prior
contact
with a child or adolescent mental health service than not (χ² = 7.47,
p <
0.01).
26
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0027.png
Table 9. Analysis of differences in the nature and number of contacts with mental health services before and after contact with the PBP, depending on treatment recommendation
and attendance
Groups
Outpatient
before
(M, SD)
Sign test
Sign test
Outpatient
after (M,
SD)
Inpatient
before
(M, SD)
Inpatient
after
(M, SD)
CATT
before
(M, SD)
CATT after
(M, SD)
Sign test
CTO
before
(M, SD)
CTO after
(M, SD)
Sign test
Not
recommende
d
1.53, 2.08
1.27, 2.13
-5.3***
2.91, 7.74
1.24, 4.27
-9.3***
0.72, 2.11
0.49, 2.96
-6.1**
1.06, 3.80
1.03, 2.75
-3.0**
Treatment
1.40, 0.95
0.49, 1.06
-5.7***
0.60, 1.54
0.28, 0.95
W
-1.7
0.37, 0.92
0.20, 0.76
W
-2.1*
0.18, 1.10
0.21, 1.26
W
-0.1
Failed to
attend
1.72, 1.36
1.18, 2.24
-5.8***
1.37, 3.84
0.82, 2.53
-1.9
0.37, 1.08
0.38, 1.28
W
-0.03
0.48, 2.60
0.31, 1.31
W
-0.2
Total
1.55, 1.89
1.17, 2.07
-
8.87***
2.43, 6.88
1.07, 3.83
-9.33***
0.63, 1.90
0.44, 2.61
-6.24***
0.88, 3.45
0.84, 2.48
-2.77*
Note:
W
indicates use of Wilcoxon test for within group comparisons
*
p < 0.05
**
p < 0.01
*** p < 0.001
27
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0028.png
Question 3: Who are PBP treatment clients and what is their experience?
Demographic Information
The majority of the 214 individuals who were referred for treatment at the PBP were male (n = 196,
91.6%; female n = 18, 8.4%). Consistent with the
not recommended
group, nearly half of the sample
were referred by Community Corrections Services (n = 94,
43.9%), one-fifth came from Area Mental Health Services (n = 41,
“Always felt safe
19.2%) and the remaining 79 (36.9%) came from a range of other
referral sources, including private and other community health,
and supported”
self-referrals and legal services. There were differences between
the
treatment
group and the
failed to attend
group in terms of
referral sources (see Figure 11).
%
100
90
80
70
60
50
40
30
20
10
0
58
31
41
130
Failed to attend
38
35
82
Treatment
9
CCS
AMHS
Other
Total
Figure 11: Referral source by
Treatment
and
Failure to Attend
groups (n (%))
Individuals were referred for treatment of a range of problem behaviours, as can be seen in Figure 12.
As can be seen in this figure, there was little difference across the types of problem behaviours for
treatment attenders (treatment group) compared to the
failed to attend
group.
100
90
80
70
60
50
40
30
20
10
0
48
64
34
3
%
3
24
36
18
Failed to attend
Treatment
Violence
Harmful sexual
behaviour
Stalking
Firesetting
Figure 12: Referral problem behaviour by
Treatment
and
Failure to Attend
groups (n (%))
28
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0029.png
The average length of the treatment episode (i.e. the date the individual actually commenced
treatment to separation from the service) ranged from 0 to 47 months (M = 8.44, SD = 9.03, median =
6 months) (Note. The end of the data collection period (30.12.2011) was taken as the treatment end
date for individuals still in treatment at the time of data collection). Figure 13 shows the dispersion of
months in treatment between those who completed treatment or were ongoing at the time of data
collection and those who failed to complete treatment. Those in the
failed to attend
group spent an
average of 6.59 (SD 7.10, median 4) months in treatment, while those who completed (treatment)
spent an average of 11.24 (SD 10.49, median 8) months in treatment.
Figure 13. Dispersion of months spent in treatment for those who failed to attend any treatment as well as those
who dropped out (Failed
to Attend group)
versus those who successfully completed or are continuing in
treatment
(Treatment group).
“They helped me
understand”
As can be seen in Figure 13, the highest rates of treatment failure
occurred within six months of commencement of treatment,
although some of the treatment failure group still dropped out
of treatment after one to two years. There was considerable
variability in the length of treatment for treatment completers
with most completing treatment within two years.
Of those who had completed treatment or were ongoing treatment clients, 9 (11%) committed
offences during their treatment episode (mean number of offences= 0.51, range of offences 0-16, SD
= 2.32), and 23 (28%) committed offences after the treatment episode (mean number of offences
1.29, range 0 -18, SD = 3.14). However, 40 (48%) of these clients had committed offences in the two
29
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
years prior to the treatment episode (mean number of offences = 4.3, range 0-110, SD = 13.04).
Overall, the number of offences after the treatment episode was significant less than the number prior
to treatment (Z
sign test =
-2.56,
p =
0.01). There was also a significant reduction in the number of
offences committed during the treatment episode compared with the number committed prior to (Z
Sign test
= -4.90,
p <
0.001) or after finishing treatment (
Z sign test =
-3.74,
p <
0.001).
Client experience
Fifteen of the 31 clients who attended for treatment between 17 and 28 November 2014 responded
to a feedback survey. Of these, 53.3% were referred from CCS, 26.7% were self-referred, 13.3% were
referred by their Area Mental Health Service and 6.7% were from another referral source.
Clients were asked to rate their overall experience at the PBP. The vast majority indicated that they
found the service to be ‘very good’ (73%), and almost all indicated that they felt ‘very supported’ by
the service (80%). No clients felt the PBP provided a poor service or was unsupportive. Clients were
also asked some specific questions about whether they perceived PBP treatment has assisted them.
Eighty percent said that treatment helped them to understand the problem they were referred for
(one individual felt it had not helped in this regard, others were neutral); 86% reported that treatment
helped them to manage the problem (others were neutral); 93% indicated that treatment helped
them to understand their offending behaviour and 80% believed it helped them to reduce their
offending behaviour (these questions were not applicable for one client).
Discussion
This report aimed to answer three questions:
1. Is contact with the PBP effective in reducing the frequency, nature and time to reoffending?
2. What impact does receiving services from the PBP (assessment and/or treatment) have on
mental health outcomes for clients?
3. Who are PBP treatment clients, what is the effect of treatment, and what is their experience?
Results relating to each question are discussed below, in addition to implications and
recommendations for practice.
Reduction in the nature, frequency and time to reoffending
The results of this evaluation of the efficacy of the PBP in reducing the frequency, nature and time to
reoffending shows that the Program is in fact successful by all three measures. These findings are
particularly important given the high risk nature of most of the referrals received by the PBP.
Specifically, the average number of offences per month reduced after PBP contact in the total sample
and in all three sub-groups. The fact that all clients of the PBP experienced a reduction in reoffending,
regardless of whether they were recommended for or attended treatment suggests that PBP
assessments themselves have some impact on reoffending. It is likely that this is due to appropriate
identification of risk reduction strategies, and implementation of these strategies by referring
agencies.
Between group comparisons of time to reoffending showed that amongst those referred for
treatment, clients who actually attended and completed treatment reoffended significantly more
slowly than other PBP clients: on average 6 months later than those who did not attend for treatment,
and 10 months later than those not recommended for treatment. Amongst those referred for
treatment, approximately half experienced no change in the severity of offending post PBP contact,
30
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0031.png
while amongst the other half of this group, the vast majority experienced a reduction in severity and
only a small minority committed further offences of a more severe nature.
The results suggest that PBP clinicians were generally making
appropriate recommendations targeting higher risk clients
“I am grateful to have
for offence-specific treatment at Forensicare. Clients who
been referred to
were recommended for treatment but who did not attend
Forensicare; it has
(the
failed to attend
group), reoffended significantly more
changed my life”
often than those who were not recommended for
treatment. This finding can be taken as a proxy for the fact
that those not recommended for treatment generally
present a lower risk of reoffending than those who are recommended. Notably, while these two
groups differed in the frequency of reoffending following their PBP assessment, there were no
significant differences in the time to their first reoffence after PBP contact. In both groups, PBP clients
who reoffended did so between 12 and 18 months post assessment or final treatment session.
Moreover, there were no differences between these two groups in the number of clients who
reoffended in a more severe way following PBP assessment.
Mental health outcomes for PBP clients
With regards to mental health outcomes and service usage of PBP clients, there were a range of
different findings that reflect the heterogeneity of the sample. Perhaps the most important is that PBP
clients frequently have contact with public mental health services, regardless of whether they are
referred to Forensicare by Community Corrections, Area Mental Health Services or other services.
Ninety percent of PBP clients have had contact with public mental health services prior to their PBP
assessment. Overall, contact with the PBP coincided with a reduction in contact with community area
mental health services. This was true both in the whole sample, and in each of the three
recommendation/treatment subgroups. It is possible that this reduction reflects general improvement
in clients’ wellbeing (including in their mental health) associated with management strategies
implemented post PBP assessment (whether by PBP treatment or other services). It could, however
also be associated with an overall decrease in crisis-driven contacts. Pre-PBP crisis Assessment Team
(CAT) and inpatient contacts in the
not recommended
group in particular may reflect a period of
difficulty in which the problem behaviour was accompanied by these mental health contacts. Amongst
this seemingly lower risk group, it is possible that with the resolution of the problem behaviour, the
mental health contacts also reduced in frequency (or vice versa).
A particularly positive mental health outcome for PBP
treatment
clients is the reduction in CATT and
outpatient contacts in this group. While it is impossible to be certain, the reduction may indicate the
development of more effective strategies for managing mental health crises secondary to PBP
treatment. Notably, a similar reduction was not observed in the
failed to attend
group. Overall the
failed to attend
group continued to utilise mental health services at the same rate as prior to their
contact with the PBP. This difference suggests that assessment alone or partial treatment is not
sufficient to result in positive mental health outcomes for these more complex and higher risk clients,
and the specialist treatment service provided by the PBP generates positive outcomes for this category
of clients when they complete treatment.
31
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0032.png
Efficacy of PBP treatment
Amongst those who were referred for and who attended
treatment
at the PBP, reoffending rates were
significantly lower than in all other groups. Perhaps most importantly, clients who had completed or
remained in
treatment
reoffended at significantly lower rates compared to clients who fail to
attend/drop out of treatment. Furthermore, within the treatment group itself, there was a significant
change in the average number of offences pre- and post-treatment, with significantly fewer offences
per month in the time since treatment was completed. Interestingly, the period of lowest risk of
offending appeared to be while the client was in treatment. Although reoffending rates increased after
treatment was completed, rates nonetheless remained lower than pre-treatment offending. There
was no significant change in the number of offences pre- and post-PBP contact for those who
failed
to attend
as recommended.
One of the most concerning findings of this evaluation was the high rate of treatment drop out in the
sample. Sixty percent of clients who were recommended to attend treatment either did not attend or
ceased treatment prior to the recommended time. Amongst those who commenced but dropped out
of treatment, actual time in treatment varied greatly. Fifty-nine individuals ceased treatment within
the first six months
–
of these 14 never attended at all and a further 30 individuals ceased within three
months. The remaining 32 clients dropped out of treatment after attending for between six months
and almost four years, similar to the length of time that most treatment completers remained in
treatment. The very positive results for those who attended and completed treatment suggests that
an increased focus of the work of PBP clinicians should be on engaging clients to the extent possible
to retain them in treatment. Moreover, where clients are recommended by external agencies (e.g.,
community corrections) to attend treatment, the expectation should be reinforced by staff in those
agencies.
There is wide variation in the duration of treatment episodes at the PBP, with most clients successfully
completing treatment within 6-12 months, while others remain with the service for 2-4 years or even
longer in a very small number of cases. Length of a treatment episode is based on a number of factors
including assessed risk level, engagement in treatment, and the possibility of making offence-related
treatment gains. Regular clinical reviews determine whether or not treatment will continue to be
offered in the face of poor engagement. Amongst the group who ceased treatment within the first six
months, it is likely that there are a sizeable number of clients who failed to engage in treatment (e.g.,
they continued to have low motivation to attend or to
change their behaviour). In these cases a clinical decision is
“…I didn’t know I had a
made to cease treatment despite there being few gains and
problem till I got told”
potentially with risk level remaining high. Amongst those
who drop out at a later point, there are likely to be a number
of Corrections-referred clients who chose not to voluntarily
continue treatment after the end of their correctional order. It is possible that this late drop-out group
also includes clients who have not engaged well but have been maintained in treatment due to anxiety
about discharging someone who is assessed as presenting a high risk of future offending.
It should be noted that this research was based on formal treatment commencement and separation
dates recorded in the public mental health database, CMI. It was not possible to collect information
about the actual number of treatment sessions attended by clients and it is possible that some clients
32
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
1618579_0033.png
ceased attending mental health services some weeks or even months prior to being formally
separated, artificially inflating their recorded length of treatment.
Given the high rate of recorded drop out, it will important to devote closer attention to the reasons
that clients cease treatment prior to an agreed end, and the kinds of barriers that exist that prevent
successful completion. This is particularly important given the demonstrated positive outcomes
associated with treatment completion in this study. It may be that a combination of systemic measures
(e.g., closer monitoring or liaison with Corrections Victoria officers regarding completion of ordered
treatment),
and psychological interventions (e.g., greater overt attention to clients’ motivation and
readiness to change prior to commencing offence-specific interventions) may go some way towards
reducing drop out. This clearly needs to be a focus for the PBP and PBP stakeholders in the future. In
addition, given the high success of treatment in significantly reducing re-offending among this high-
risk sample, if funding was available, a greater number of clients should be referred for treatment
In the survey of current PBP treatment clients, almost all reported high levels of satisfaction with the
service. Most importantly, the majority of clients reported that they have had a good experience at
the PBP and felt supported. More specifically, they reported
that the PBP has helped them to understand their problem
“… [they’re] helping
me
behaviour, reduce this behaviour, understand unhelpful
with a problem I don’t
thinking habits, and feel better. It is recommended that
want to have any more”
client feedback frequently be sought as part of an ongoing
evaluation of the program. Of course, it is possible that only
well engaged clients who value the service chose to respond
to the survey. It is noteworthy that the one client who
provided negative evaluations identified that they did not understand why they had been referred to
the program and did not believe that their behaviour was a problem.
In summary, the results of this evaluation study reinforce the pivotal role played by the Problem
Behaviour Program in the assessment and treatment of client groups that are often unable to access
treatment elsewhere, and whose behaviours impact on community safety. The PBP has demonstrated
efficacy in reducing offending, particularly amongst clients who successfully complete treatment in
the program. The positive results highlight the importance of the program not only continuing but
expanding and developing so as to further meet the needs of the mental health and justice systems in
reducing harm to victims, the community and clients.
33
 LIU, Alm.del - 2015-16 - Bilag 52: Baggrundsmateriale til brug for udvalgets høring om stalking onsdag den 13. april 2016.
References
Andrews, D. A., & Bonta, J. (2010).
The psychology of criminal conduct.
(5th ed.). New Providence,
NJ: Matthew Bender & Company, Inc.
Bennett, D. J., Ogloff, J. R. P., Mullen, P. E., Thomas, S. D. M., Wallace, C., & Short, T. (2011).
Schizophrenia disorders, substance abuse and prior offending in a sequential series of 435
homicides.
Acta Psychiatrica Scandinavica, 124,
226-233.
Cutajar, M. C., Mullen, P. E., Ogloff, J. R. P., Thomas, S. D., Wells, D. L., & Spataro, J. (2010).
Psychopathology in a large cohort of sexually abused children followed up to 43 years.
Child
Abuse and Neglect, 34,
813-822.
Douglas, K.S., Hart, S.D., Webster, C.D., & Belfrage, H. (2013).
HCR-20v3: assessing risk for violence:
user guide.
Burnaby, BC: Simon Fraser University, Mental Health, Law, and Policy Institute
Ducat, L., Ogloff, R.P., & McEwan, T. (2013). Mental illness and psychiatric treatment amongst
firesetters, other offenders, and the general community.
Australian and New Zealand Journal
of Psychiatry, 47,
945
–
953.
Hart, S.D., Kropp, P.R., & Laws, D.R. with Klaver, J., Logan, C., & Watt, K.A. (2003).
The Risk for Sexual
Violence Protocol (RSVP).
Burnaby, BC: Simon Fraser University, Mental Health, Law, and
Policy Institute
Krupinski, J., Alexander, L., & Carson, N. (1982). Patterns of psychiatric care in Victoria: Special
publication No. 12. Melbourne: Health Commission of Victoria.
McEwan, T., MacKenzie, R. D., & McCarthy, J. (In press). The Problem Behaviour Program: Threat
assessment and management in a community forensic mental health context.
Quinsey, V., Rice, M. E., Harris, G. T., & Cormier, C. (2006).
Violent offenders: Appraising and
managing risk.
Washington, DC: American Psychological Association.
Short, T., Thomas, S., Luebbers, S., Ogloff, J. R. P., & Mullen, P. (2010). Utilization of public mental
health services in a random community sample.
Australian and New Zealand Journal of
Psychiatry, 44,
475-481.
Wallace, C., Mullen, P. E., & Burgess, P. (2004). Criminal offending in schizophrenia over a 25-year
period marked by deinstitutionalization and increasing prevalence of comorbid substance
use disorders.
American Journal of Psychiatry, 161,
716-727.
Webster, C., Douglas, K., Eaves, D. and Hart. S. (1997).
HCR-20. Assessing risk for violence, version 2.
Burnaby, BC: Simon Fraser University, Mental Health, Law, and Policy Institute
34