FORUM on Corrections Research

Treatment programs for offenders with violent histories: A national survey

Efforts to develop assessment strategies to identify offenders' risks and needs have evolved considerably over the last decade.² These initiatives have led to the provision of specific treatment programs intended to target the criminogenic needs of offenders. To meet these needs, the Correctional Programs Branch has created core (Cognitive Living Skills, Anger and Emotions Management, Living Without Violence, Parenting) and supplemental (Offender Substance Abuse Pre-release Program) programs. Preliminary results on the utility of some of these interventions are now available.³ Other programs, such as those for sex offenders, are also being evaluated, as the Correctional Service of Canada moves to address concerns expressed by the Auditor⁴ regarding the provision of correctional treatment programs.

While the evaluation of program efficacy is critical to good correctional and fiscal management, other considerations also have merit. This article highlights findings from a recently completed national survey of treatment programs for offenders with violent histories.⁵ Prior reviews of published programs for violent offenders have been disappointing,⁶ yet discussions with field staff suggested there were many programs in existence. A major purpose of this review was to delineate the nature and quantity of these programs provided in the Service. This review is deliberately qualitative and descriptive, as the goal was to provide information to consumers, rather than formal program evaluation. It was our belief that the review of all programs presently being delivered to violent offenders would yield important information for those interested in the integration of programs for violent offenders. For this survey, Anger and Emotions Management (AEM) and Living Without Violence (LWV) represented 31.1% and 17.6% of the programs, respectively. Interestingly, approximately half (51.4%) of the programs reviewed were other programs. Apparently, specific sites developed them to meet their particular needs. These data, then, should inform the Service about a range of issues related to meeting the treatment needs of violent offenders. Also, as a compilation of existing treatment programs, the data represent a compendium for staff to review when making recommendations for transfer and release of violent offenders for whom treatment needs remain.

Types of programs

Those surveys completed in January 1997, were coded and entered into the present review. Overall, the national completion rate (number of returns divided by the number of sites) was 37.7% (52 responses from 138 sites). Regionally, the completion rates were Atlantic, 28.1%; Quebec, 51.6%; Ontario, 34.2%; Prairies, 38.5% and Pacific, 33.3%. Similarly, the completion rates by security level were minimum, 53.9%; medium, 72.2%; maximum, 66.7% and community 22.5%. Given the limited time frame, the completion rate was fairly good. An additional 16 surveys received after the deadline for data entry were not included. Further, several sites, particularly community sites, not offering programs may simply have failed to provide a nil response. See Table 1 for a breakdown of program type by region.

Table 1

Program Type by Region
	Atlantic		Quebec		Ontario		Prairies		Pacific		National
	%	(n)	%	(n)	%	(n)	%	(n)	%	(n)	%	(n)
AEM	37.5	(3)	18.2	(4)	42.9	(9)	20.0	(3)	50.0	(4)	31.1	(23)
LWV	25.0	(2)	18.2	(4)	9.5	(2)	13.3	(2)	37.5	(3)	17.6	(13)
Other	37.5	(3)	63.6	(14)	47.6	(10)	66.7	(10)	12.5	(1)	51.4	(38)
Total	10.8	(8)	29.7	(22)	28.4	(21)	20.3	(15)	10.8	(8)	100.0	(74)
Notes: AEM = Anger and Emotions Management; LWV Living Without Violence

Table 2

Program Type by Security Level
	Minimum		Medium		Maximum		Community
	%	(n)	%	(n)	%	(n)	%	(n)
AEM	25.0	(3)	37.0	(10)	46.7	(7)	15.0	(3)
LWV	16.7	(2)	18.5	(5)	9.5	(2)	13.3	(4)
Other	58.3	(7)	44.4	(12)	40.0	(6)	65.0	(13)
Total	16.2	(12)	36.5	(27)	20.3	(15)	27.0	(20)
Notes: AEM = Anger and Emotions Management; LWV Living Without Violence

Another consideration is the security level at which the majority of programs for violent offenders is provided (see Table 2). Minimum security includes healing lodges, and maximum security includes the treatment centres and multilevel institutions. As can be seen, programs for violent offenders are provided equally at all security levels. Targetted groups

The survey listed a number of potential target groups, or types of offenders, for whom the program was intended. These target groups were: persistently violent, criminal violence, anger, institutional violence, domestic violence and domestic exclusive. This was principally to determine whether programs were restrictive in their selection of offenders. Many programs reported using several of the targets, potentially leading to quite heterogeneous groups regarding treatment needs. Other targets such as suicidal, mentally disordered, victims of domestic abuse and lifers were also reported, but accounted for less than 7% of the total targets. These data allow us to consider whether existing programs are targetting those offenders for whom the program was originally intended and where additional programming efforts might be required in light of the prevalence of these "types" of offenders.

Selection and exclusion criteria

The survey also asked respondents to indicate specific selection criteria used to accept offenders into their program. Staff can review these results to confirm that the specific treatment needs of offenders selected to their programs are addressed in the program content. There appears to be considerable regional variation in selection criteria. On average, most programs consider current offences, but relatively few consider prior assaults or incorporate pretreatment test results into selection. Similar breakdowns are available by security level, on request.

With respect to exclusion, over 75% of the programs exclude offenders with active psychotic symptoms. Similarly, 30% exclude offenders with low motivation, 12% exclude those who deny they require treatment, and 18% exclude offenders because of low intelligence. This latter decision makes intuitive sense given the emphasis on cognitive treatment strategies in these programs. It is also understandable that poorly motivated and resistant offenders will be disruptive in groups. However, with nearly 40% of programs excluding such offenders, this identifies an emerging new treatment target group. Breakdowns by region, security level and setting are available and will be detailed in the final report.

Waiting lists

One measure of treatment needs is the number of offenders who have been referred for a particular program, but who have been placed on a waiting list until space permits their participation. Nationally, 78.9% of those programs surveyed maintain a waiting list. The average number of offenders waiting for admission to a violent offender program is 40, although the range is from 3 to 169. Regional breakdowns of offenders on waiting lists are Atlantic (8), Quebec (17), Ontario (61), Prairies (33) and Pacific (54). Further, the number of offenders waiting differs according to security level: minimum (14), medium (64), maximum (36). Setting is also important regarding waiting lists: community (11), institutional (45).

Group versus individual programs

Consistent with direction over the last decade to provide treatment in group format, 72.6%
of programs are exclusively group. Since the survey only sampled programs, individual therapy was not included, but 27.4% of the programs surveyed augment group programming with individual sessions. There are regional and setting differences, with the Pacific region only reporting group sessions, whereas the Prairies offer individual sessions in 57.1% of their programs. The other regions offer individual sessions in 13% to 32% of their programs. There are no major differences among security levels, nor community versus institution-based programs.

Program orientation and components

Not surprisingly, 85% of the programs surveyed nationally provide a cognitive behavioural model of intervention. Psychotherapeutic approaches are reported in 20.5% of the programs, suggesting that these approaches are integrated into a complementary model. Similarly, 20.5% of the programs reported their theoretical orientation as psychoeducational. Program components indicate the content presented in the various treatment programs. As with the issue of target groups, it is important for program deliverers to confirm that the content of their program addresses the needs of those offenders selected.

Treatment targets

Respondants were asked to indicate whether specific treatment targets reflected in the literature were included in their program. For these data, the "other" category includes symptom management, dealing with remorse, confronting denial, dealing with stress, self-esteem, relationships or social reintegration. The percentage of programs addressing these targets in treatment are presented in Table 3.

Table 3

Treatment Target by Location
Treatment Target	Institutional		Community		National
	%	n/52	%	n/20	%	n/72
Arousal reduction	63.5	(33)	35.0	(7)	55.6	(40)
Problem Solving	80.8	(42)	90.0	(18)	83.3	(60)
Communication Skills	88.5	(46)	85.0	(17)	87.5	(63)
Assertiveness	78.8	(41)	75.0	(15)	77.8	(56)
Insight	92.3	(48)	90.0	(18)	91.7	(66)
Cognitive distortions	88.5	(46)	80.0	(16)	86.1	(62)
Relapse prevention	63.5	(44)	60.0	(12)	62.5	(45)
Other	13.5	(7)	20.0	(4)	15.3	(11)
Note: percentages do not sum to 100 given that the response categories were not mutually exclusive

Program length

The average program length, combined with waiting list information, should help staff review resources and plan strategies for best meeting the treatment needs of violent offenders. On average, programs included 22 sessions, averaging 2.5 hours each over the course of
13 weeks. This means that a maximum of three programs could be delivered at a site if a program deliverer was dedicated to only one program. Nationally, it appears that, on average, these programs have been in place for 36 months, with only minor variation across the regions and security levels.

Changes to programs

Unfortunately, most respondents (70%) were unable to comment on the degree of change their program has undergone. However, the remaining 30% reported changes regarding program content (72.1%) and report format (34.3%). Few programs (7.2%) reported changes to admission criteria. Other changes such as procedures and scheduling were also reported (37.5%).

Methodology

Many of the programs reported maintaining a control group (37%) and database (73.2%). Similarly, 67.1% reported having an advisory committee. Interestingly, only 17% reported a formal evaluation, with 30% reporting published findings (10.8 formal, 18.5 informal), yet most (53.8%) indicated that evaluations were in progress. The majority of programs (87.7%) have treatment manuals which reflect content (100%), rules (92%), admission criteria (94%), expulsion criteria (88%), report format (86%), homework assignments (84%) and a test battery (78%). Additional information contained in
the treatment manuals, such as participant contracts, session goals, session summaries and guidelines for staff, were also reported, albeit infrequently.

Assessment of treatment needs

Various strategies for the assessment of treatment needs were recorded by program deliverers. The endorsement frequency of these methods is presented in Table 4.

Table 4

Assessment Method by Location
Assessment Method	Instutional		Community		National
	%	n/53	%	n/19	%	n/72
Interview	96.2	(51)	100.0	(19)	97.2	(70)
File review	77.4	(41)	68.4	(13)	75.0	(54)
CTP	58.5	(31)	68.4	(13)	61.1	(44)
Collateral information	30.2	(16)	47.4	(9)	34.7	(35)
Testing	62.3	(33)	47.4	(9)	58.3	(42)
Note: percentages do not sum to 100 given that the response categories were not mutually exclusive

These data provide evidence that program staff use multimethod assessment in identifying treatment needs. It is disappointing, however, that these data suggest that correctional treatment plans appear to be only modestly linked to the identification of treatment needs for violent offenders.

Assessment of treatment gain

Respondents were provided with eight choices for assessing treatment gain. In order of most frequently to least frequently endorsed, these were: offender satisfaction, role plays, knowledge questionnaire, test battery interactions with staff, behavioural rating, institutional performance and institutional charges. By combining these categories into a composite score, we can conclude that, on average, programs use four different methods for the assessment of treatment gain.

Risk assessment

Incorporation of risk considerations into programs for violent offenders was reviewed in several ways. First, respondents were asked to indicate whether a risk assessment is conducted; 83.6% reported it was. Fewer respondents (63.3%), however, indicated a risk assessment was part of the initial assessment. Risk was rarely considered as part of the admission criteria for the program (16.7%). Further, assessments of risk were often (53.3%) not considered in post-treatment reports.

Summary

The findings of this survey are informative and encouraging. First, even allowing for modest completion rates because of time constraints, there are a significant number and variety of programs offered to violent offenders. Second, the majority of programs report relevant treatment targets, multimethod assessments of treatment need and treatment gain. Surprisingly, almost half of the programs surveyed were not core programs, indicating considerable initiative in the field to meet treatment needs. Offenders admitted to these surveyed programs are likely heterogeneous regarding type of offences. However, program content and treatment targets are diverse, optimally providing responsive intervention. For most of the questions considered, there were minor regional variations. Since offender profiles also vary among the regions, this is not necessarily problematic. Finally, these results suggest the need for improved integration with correctional treatment plans and greater emphasis on risk assessments by program staff.

---

1. Research Branch, Correctional Service of Canada, 340 Laurier Avenue West, Ottawa, Ontario, K1A 0P9.

2. J. Monahan and H.J. Steadman, eds., Violence and Mental Disorder: Developments in Risk Assessment (Chicago, Illinois: University of Chicago Press, 1994). See also L.L. Motiuk, Assessment Methods in Corrections, Paper presented at the 4th Annual International Community Corrections Association Research Conference (Austin, Texas, September, 1996). And see C.D. Webster, G.T. Harris, M.E. Rice, C. Cormier and V.L. Quinsey, The Violence Prediction Scheme: Assessing Dangerousness in High Risk Men (Toronto, Canada: Centre of Criminology, University of Toronto, 1994).

3. W.A. Millson, J.R. Weekes and L.O. Lightfoot, The Offender Substance Abuse Prerelease Program: Analysis of Intermediate and Post-release Outcomes, Research Report R-40 (Ottawa: Correctional Service of Canada, 1995). See also D. Robinson, The Impact of Cognitive Skills Training on Postrelease Recidivism among Canadian Federal Offenders, Research Report R-41 (Ottawa: Correctional Service of Canada 1995).

4. Auditor General of Canada, Correctional Service of Canada - Rehabilitation Programs for Offenders (Ottawa: Minister of Public Works and Government Services Canada, 1996).

5. We would like to acknowledge those program staff who completed the survey. Their support and comments were greatly appreciated.

6. R.C. Serin, Treating Violent Offenders: A Review of Current Practices, Research Report R-38 (Ottawa: Correctional Services of Canada, 1994). And see R.C. Serin and S.L. Brown, "Strategies for Enhancing the Treatment of Violent Offenders," Forum on Corrections Research, 8, 3 (1996): 45-48.

7. One program reported 520 sessions. This was five times the next highest number, so was deleted for these analyses.