Research Reports

Evaluation of Correctional Service of Canada Substance Abuse Programs

Prepared by:

Paul Gendreau & Claire Goggin
University of New Brunswick

Research and Statistics Branch
Correctional Service of Canada

April 1991

ACKNOWLEDGMENTS

1. The project was supported by Research Contract # 21100-9-1575-INH-00, Correctional Service of Canada (CSC), Solicitor General of Canada. We appreciate the assistance of the following CSC officials on various aspects of the research: Dr. Frank Porporino and Dr. David Robinson, Research & Statistics Branch and Mr. Jon Klaus, Health Care Branch. We thank Dr. Helen Annis, Addiction Research Foundation, for her advice on the alcohol assessment and intervention literature.
s
2. Requests for further information should be directed to the first author, Division of Social Science, University of New Brunswick, P.O. Box 5050, Saint John, N.B., E2L 4L5.

ABSTRACT

 One hundred and twelve CSC substance abuse programs responded to a questionnaire, the Correctional Program Evaluation Inventory (CPEI), about their program practises. Deficits were discovered across all areas of programming activity: program implementation, classification, treatment, and evaluation.

 Programs were also assessed as to "quality". While the majority of programs were found to be less than adequate, several satisfactory programs were identified. Also, contracted and/or residential programs were rated higher on the CPEI.

 Some cautions are advances regarding the limitations of the research - nonresponding to individual items was high and instructions provided by respondents about their programs could not be verified in this study.

 Finally, some suggestions are made to improve the quality of CSC substance abuse programs.

INTRODUCTION

 One of the critical issues regarding the delivery of effective services to offenders centres on the fact that we are not the experimenting society (cf. Gendreau & Ross, 1987) that we claim to be. That is, while there exists a large data base that provides persuasive testimony to the fact that various types of service delivery reduce recidivism of offenders (Andrews, Zinger, Hoge, Bonta, Gendreau, & Cullen, 1990a; Gendreau & Andrews, 1990; Gendreau & Ross, 1987; Hill, Andrews, & Hoge, in press; Lipsey, 1990), the majority of evidence has come from studies that were "experimental" in nature.

 Critics of rehabilitation (e.g., Lab, 1990; Lab & Whitehead, 1990) have targeted this point and asserted that the successful programs are merely "utopian" and "chimerical". They reason that these programs are not reflective of the reality of government and private agency offender programming routinely found "in the field". The intent of their argument, as noted elsewhere (Andrews, Zinger, Hoge, Bonta, Gendreau, & Cullen, 1990b), is to try to deny that rehabilitation can ever be effective, which is obviously not the case¹. There are, nevertheless, legitimate concerns amongst clinicians involved in offender rehabilitation about the overall quality of services in the field. For example, there are likely 1000² offender treatment programs in existence in Canada. Only a very small percentage have ever been formally evaluated, let alone published. Even more disquieting is the fact that little information exists about the nature of the services themselves. At the present time, we are unaware of any large-scale, empirically-based surveys of the offender treatment programs in this country. This sort of information is necessary if proactive steps are to be undertaken to improve upon offender programming facilitated in the field.

 Recently the Correctional Services of Canada (CSC) has vigorously embarked upon a rehabilitation agenda. Part of this agenda involves a national strategy to alter the delivery of substance-abuse programs. The need for such a strategy was dramatically illustrated by comprehensive surveys (Research & Statistics Branch, CSC, 1990a,b) of federal inmates that found 54% of offenders had a serious substance abuse disorder and that 64% had consumed a drug the day they committed the crime for which they were incarcerated. One of the mandates of the newly created CSC Substance Abuse Task Force was to survey existing CSC substance abuse programs about the nature of their programs. In addition, a preliminary assessment of the quality of the services was requested.

PROCEDURE

 The evaluation of CSC substance abuse programs was carried out using the Correctional Program Evaluation Inventory (CPEI) as designed by Gendreau & Andrews (1990) and adapted for the purposes of this survey. The CPEI consists of a variety of items that assess several factors that have been found to be associated with the literature on "what works" with offenders (e.g., Andrews et al., 1990; Gendreau & Andrews, 1979; Gendreau &Ross, 1979). These factors are program implementation, client assessment, treatment modalities, staff characteristics/practises, and program evaluation. In summary, the CPEI allows for a comprehensive summary of the current functioning of a program as well as a rating of the program as to its potential effectiveness.

 The Research & Statistics Branch of the Correctional Service of Canada forwarded the CPEI to 170 substance abuse programs directly operated by the Correctional Service of Canada or contracted out to external agencies. Instructions were provided for answering the inventory. The completed inventories were subsequently forwarded to the authors for tabulation of the results. The last protocol was received in July of 1990.

RESULTS

1.   Response Rate

 Of the 170 programs that were requested to participate in this survey, 112 responded for a return percentage of 66%. A list of the programs that responded is included in Appendix A. Eleven respondents, six of which had identifying postmarks, returned the entire inventory unanswered. These were included in the final tally of 112. In addition, it should be noted that even among the 101 programs which did respond to the questionnaire items, there were frequent instances, ranging from 20% to 50%, where individual items went unanswered.

 The following tabulations are based on data from only those programs that responded to the items, with the exception of number 2a), which also includes the six unanswered questionnaires noted above.

2.  Program Demographics

		N
a) Response by region:
	Pacific	5
	Prairie	30
	Ontario	20
	Quebec	23
	Atlantic	28
b) Program setting:
	Institution	80
	Residential	12
	Therapeutic community/other	9
c) Security level:
	Minimum	18
	Medium	32
	Maximum	22
	Combinations of above	8
	Community	17
d) i) number of programs and client composition	Alcoholics	14
	Substance abusers	11
	Both	75
ii) eighty-eight % of programs were male only
iii) of the institutional programs, 5 housed their clients separately from the main population
e) Contracted/operated by CSC:
	Contracted	56
	Operated	36
	Both	1
f) i) mean length of program operation:		7.51 years
	SD	10.01
ii) one-third (33%) of programs have been in operation for one year or less
iii) mean program duration:		45.32 days
	SD	65.69

3.   Program Implementation

 The rate of non-responding to items in this category ranged from 24 - 57%.

 Thirty-six percent of the persons primarily responsible for designing and establishing their program were professionally trained, while 33% had been involved in conducting similar programs in the past. In 30% of cases, that individual was directly involved in the selection and training of staff, and 36% indicated that person was also directly involved in running some of the therapeutic components of the program. Thirty-five percent said that the person integral to program design and implementation continues to play an active role in the program.

 Forty-five percent reported that they had conducted a literature search prior to program implementation and 58% of respondents indicated that they had conducted a program needs assessment before putting the program into effect.

 A majority of respondents said that their program was generally perceived by both the institution (76%) and the community (64%) as being cost-efficient and sustainable. In only a distinct minority of cases, just 16%, was a pilot program conducted before implementation of the formal program.

4.   Client Assessment

 Despite the diversity of client characteristics available for assessment, approximately 35%-55% of those surveyed did not respond to the items in this section. A further 11% indicated that formal client assessment was not a regular part of their program. The analysis of the remaining responses provides an indication of those client assessment variables which are most commonly assessed among substance abuse treatment programs that did report. It should be noted that, of the characteristics which were assessed "regularly", a majority (75%) indicated the information upon which the assessment was based. The most common response was file information, personal interview, and case management documents. Only a handful among those measures cited, however, were recognized, standardised psychometric measures, e.g., MAST-DAST, MMPI, Buss-Durkee.

 The following percentages are generated only from those programs that checked off one of the three available response categories. For example, 45% of the programs stated they assessed the variable Aggression "regularly".

a)    Client Assessment Variables

Variable Assessed	Not assessed	Rarely/ Occasionally	Regularly
Aggression	17%	38%	5%
Alienation	19%	53%	28%
Anti-social attitudes	16%	29%	55%
Anxiety	17%	39%	45%
Cognitive - reasoning skills	19%	47%	34%
Copying styles	19%	34%	47%
Depression	17%	38%	45%
Diet	20%	64%	16%
Educaiton	18%	45%	37%
Empathy	18%	42%	40%
Employment	17%	38%	45%
Family factors	16%	34%	50%
Family history: substance abuse	15%	26%	59%
Family history: criminality	20%	31%	49%
Harm caused to victim	17%	46%	37%
Intelligence	20%	58%	22%
Learning disability	20%	60%	20%
Leisure/recreation	18%	43%	39%
Medical status	19%	45%	36%
Mental disorder	20%	58%	22%
Motivation	16%	24%	61%
Moral development	19%	48%	33%
Peer group association	19%	40%	41%
Psychopathy	22%	56%	22%
Religious values	19%	60%	21%
Self-esteem	16%	25%	59%
Sexual beliefs	20%	50%	30%
Situational factors re: substance abuse	20%	30%	50%
Social support	16%	23%	61%
Socialization	18%	32%	50%
Other	47%	18%	35%

 

b)    Assessment of risk level of client:

no response no yes

35% 37% 29%

5.   Treatment Modalities

 Excluding the "other" category the percentage of "no response" for the following modalities ranged from 24%-39%. Therefore, the following percentages are based on those programs that responded to the three categories noted below.

Modality	Not Important	Moderately Important	Very Important
AA	7%	32%	61%
Advocacy	20%	68%	12%
Chemical	71%	25%	4%
Confrontation	37%	50%	13%
Cognitive Behaviour Modification	28%	47%	25%
Controlled drinking	71%	20%	9%
Covert sensitization	70%	21%	9%
Criminal thinking	59%	34%	7%
Detoxification	58%	25%	17%
Education	12%	32%	56%
Family therapy	42%	41%	17%
IPPS (Platt)	23%	42%	35%
Literacy	33%	45%	22%
Marital therapy	43%	52%	5%
Moral - development	28%	49%	23%
Client-centred councelling	25%	49%	26%
Operant strategies	56%	38%	6%
Positive peer culture	13%	51%	36%
Psychoactive drugs	85%	5%	10%
Psychodynamic therapy	60%	27%	13%
Recreation	37%	45%	18%
Restitution	62%	37%	1%
Social - cognitive skills	25%	55%	20%
Spiritual	20%	51%	29%
Stress management	23%	46%	31%
Surveillance	72%	19%	9%
Vocational	44%	43%	13%
Other	13%	7%	80%

b)    Matching

 The following information applies only to respondents that answered the items. Sixty percent of the programs did not vary strength of treatment with client risk level. In 52% of the cases client characteristics were not matched with the treatment and 46% of the programs did not match client with the personal and professional skills of the therapist.

 Staff were allowed discretion in the management of exceptional cases for 21% of the programs. Finally, of the institutional programs, only 13% scheduled the program within the last three months of the inmates' sentence.

c)    Relapse prevention

 The following relapse techniques involving training the client 1) to monitor and anticipate problem situations, 2) to rehearse alternatives to problem situations involving substance abuse, and 3) to practise new behaviours in increasingly difficult situations, were responded to affirmatively by 52%-56% of the programs that answered these items.

 Two other techniques, "booster sessions" and "using friends as co-therapists", were used in 17% to 22% of instances, which is not surprising given 71% of programs were institution based.

6.    Staff Characteristics

a) Seventy-three percent of programs answered the staff characteristics questions. The "average" program, of which 63% were staffed by men, had the following demographics.

Education	%
High school	19
Community college	34
BA	27
Bsc	6
MA / Msc	4
PhD	1
Other	7

Profession	%
Clergy	13
Criminology	16
Education	8
Medicine	1
Nursing	11
Psychology	17
Other	7
Social Work	27
Sociology	6
Other (i.e. addictions counsellor)	30

 

b) The mean number of staff years working with substance abusers and offenders was 1) n = 82, X = 3.72, SD = .98 and 2) n = 77, X = 3.71, SD = 1.01, respectively.

c) Several other staff characteristics of importance were:

	No response	No	Yes
i) staff hired for characteristics other than experience and training	44%	11%	45%
ii) staff skills assessed periodically	40%	22%	38%
iii) staff input into program design	37%	13%	50%
iv) staff input into program functioning	37%	3%	50%
v) staff training workshops	38%	15%	47%
vi) staff hiring: of 58 programs, the average number of staff hired for the last three years was one per year.
vii) Director hiring: of 44 programs, the similar statistic was 1 per 3 years.

7.    Evaluation/Accountability

	No response	No	Yes
i) Board of Directors	13%	62%	25%
ii) Advisory Committee re: programming	35%	54%	11%
iii) Quality Assurance assessment	41%	20%	39%
iv) Client satisfaction	25%	34%	41%
v) Client follow-up	25%	46%	29%
vi) Formal program evaluation	25%	53%	22%

8.    Region, Operated/Contracted and Program Setting

 The analysis included a determination as to whether any of the above factors were associated with the results reported on the CPEI. The following associations noted below were significant at the .05 level using Pearson's

Region

1. Ontario contracts out more programs.
2. Prairie programs tend to be shorter in duration.
3. Ontario rejects more clients from their programs.
4. The Prairie region, compared to the others, places less emphasis on assessing anti-social attitudes, anxiety, cognitive-reasoning skills, education, and learning disabilities.
5. The Atlantic region places the most emphasis on the assessment of education, intelligence, and psychopathy.
6. Differences in the emphasis of treatment modalities were reported for the following regions:
   • more emphasis - chemical (Atlantic), detoxification (Atlantic, Quebec), positive peer culture (Atlantic and Prairie)
   • less emphasis - social-cognitive skills training (Pacific and Prairie), vocational (Prairie).
7. The Pacific and Prairie regions have staff with a greater degree of previous experience working with offenders.
8. The Quebec region, in contrast to others, expressed less interest in hiring on the basis of staff characteristics related to treatment, periodic assessment of staff skills (also the case in the Atlantic region), and providing on-going staff training.
9. Quebec region programs reported less concern with follow-up evaluation.

Program Setting

 Program setting was sub-divided into institution vs. community-based.

 A number of differences existed between the two settings. First, as to program demographics, community programs are contracted out more frequently, are more likely to have a Board of Directors and program advisory committee, and are of less duration.

 In regard to assessment, community programs place more emphasis on alienation, anxiety, education, family histories of substance abuse and/or criminal behaviour, medical state, peer group association, and social supports. Institutional programs assessed aggression more often.

 Community programs are more involved in matching therapist and client characteristics as well as allowing for more deviations from the treatment norm for exceptional cases. Relapse prevention strategies, as expected, are utilized more often in community settings.

 There are important staffing differences. Community programs are:

• more concerned with hiring staff on the basis of skills related to treatment other than only experience and training
• provide more periodic assessment of staff skills
• allow staff more input into program functioning
• provide more on-going training
• are involved in more hiring and have tried to hire more females and recruit more among the social work and addictions counsellor professions.

Operated/Contracted  

 There is overlap with the previous category as 90% of community programs were contracted out.

 Contracted out programs differed on some important dimensions from those operated directly by CSC. Administratively, they were more likely to have a Board of Directors and have combined alcohol/substance abuser clientele. As well, their programs were of shorter duration.

 In regard to treatment, contract programs relied more on the cognitive behaviour modification and client centred counselling modalities. They were more concerned with matching i) strength of treatment with risk level, ii) therapist and client characteristics, and iii) type of treatment with the client. They have more staff training, more periodic assessment of staff skills, and allow staff more input into programs. Contracted out programs are more likely to conduct literature reviews of treatment and hire persons with criminology degrees.

 Finally, in contrast to the differences in emphasis on client assessment characteristics in the institution-community program comparison, the differences between the operated and contracted programs on this dimension were minimal.

9. Program Quality   In reviewing the results in this section the following qualification is stressed. While the CPEI affords a score and a resultant classification of a program's quality, the instrument is experimental in nature and work is in progress to further refine the instrument. Thus, the classification data is best considered preliminary. The Discussion section denotes further cautions in this regard.

 Forty-four items on the CPEI were designated as critical indices of program quality. Each program was scored on these items and the scores expressed as a percentage of 100. The mean percentage for all programs on the CPEI was 25%, with 10 programs scoring 50% or better.

Region	n	CPEI%
Pacific	5	40%
Prairie	30	25%
Ontario	20	28%
Quebec	23	26%
Atlantic	28	20%
TOTAL	112	25%

There were no significant differences across regions (F = 1.76, df =4/101, p > .05).

Operated / Contracted	n	CPEI%
Operated	36	20%
Contracted	56	32%

Contracted programs scored significantly higher on the CPEI (F = 12.65, df = 1/90, p < .05).

Operated / Contracted	n	CPEI%
Institution	80	23%
Community	20	37%

Community programs scored significantly higher on the CPEI (F = 11.78, df = 1/98, p < .05).

 It should be noted that the significant effects reported for contracted and community programs were independent of each other. Partial correlations indicated that the correlation between higher scores on the CPEI and "community" was r 13.2 = .28, df = 109, p < .01 with the operated/contracted variable held constant. Similarly, the correlation between higher scores on the CPEI and "operated" was r 23.1 = .45, df = 109, p < .01 with the institution/community factor held constant.

DISCUSSION

 Some important caveats must be noted before proceeding further. First, while the questionnaire return rate was an impressive 66%, non-responding to individual items ranged from 20%-50%. It is possible that, had a higher percentage of questions been completed, a somewhat different picture of the type and quality of service delivery would have emerged. The same rational applies as to how the information was gathered. It is mandatory that further investigations along this line utilize a key-informant format. On-site reviews would ensure that all pertinent data is gathered. We have only sampled the "paper quality" of programs. No doubt a few programs were done a disservice as a result. Finally, the actual measure of program "quality", the CPEI, has face and content validity but, as yet, no criterion validity³. Indeed, while it is quite probable that programs with elevated scores on the CPEI may be more effective in reducing the recidivism rates of their clients, we must emphasize that some of the lower scoring programs may also be effective in this regard. From the perspective of CSC and their desire to modify substance abuse programming, it is crucial that programs collect follow-up recidivism rates in the future. Only 25 programs reported follow-up evaluations and it is unclear as to what the follow-up information consisted of.

 Granted the above limitations, however, it would be fair to conclude that the data gathered suggest that CSC substance abuse programs require revision and upgrading in almost all areas. This conclusion comes as no surprise to those observers who have expressed reservations about the general quality of offender services in the field. We address the concerns raised in this study from a general perspective, touch upon some issues internal to CSC, then provide suggestions for improvements in programming.

General Issues

 First, the implementation of programs requires more attention. Several factors have to be in place to increase the chances that a program will be established and, most importantly, maintained (cf. Gendreau & Andrews, 1979).

 Secondly, the approach taken to classification has been a bit haphazard. Some variables among those most favoured for assessment e.g., anxiety, depression, self-esteem, motivation, are unreliable, if not poor, predictors of recidivism (see Andrews, Bonta, & Hoge, 1990). How the assessments are tabulated and scored is extremely problematic. As a case in point, no mention was made of employing two superior measures of criminal behaviour (LSI, PCL) or Annis' (1990) measure of substance abuse.

 The situation is also serious when it comes to treatment modalities. Alcoholics Anonymous was the most popular treatment modality; evidence for it's success is very sparse indeed (Miller & Hester, 1985). Similarly, positive peer culture, spiritual, and client-centred counselling are usually ineffective strategies for offenders (Andrews et al., 1990a). Meanwhile, the following modalities - controlled drinking, covert sensitization, operant techniques, and surveillance - were little employed by respondents. All have promise in the treatment of alcoholism according to experts in the area (Miller & Hester, 1985).

 There also appears to be a conceptual chasm, in the minds of most respondents, regarding classification and treatment. The correctional and substance abuse treatment literatures (Andrews, et al., 1990a; Annis, 1990) are replete with examples of the importance of matching client characteristics i.e., risk, with type and strength of treatment. Less than 30% of surveyed programs said they did either.

 Fourthly, the evaluation component has been virtually ignored, although admittedly, quite a few programs (33%) were within their first year of operation.

Internal Issues

 Regional differences amongst substance abuse programs were not profound. The Pacific region was difficult to assess given its low response rate. The fact that some assessment characteristics were weighted differently across regions likely reflects differences in personnel and training amongst other things. Some diversity is probably quite healthy.

 It was noted that Ontario seemed to contract out a good deal, of the programming. Somewhat disconcerting was the evidence from the Quebec region indicating that it's programs tend to de-emphasize certain staff development areas.

 The most striking result was the superiority of contracted and community based programs on the basis of CPEI scores. There are presumably some systemic reasons for this that are well known to CSC cognoscenti. The major area of improvement for most institution/operated programs seems to be the matching of client and treatment factors, and staff development. The reader must bear in mind that having community and/or contracted programs are no panaceas The average CPEI score for contracted and/or community programs was less than 40%. In addition, well rated programs can exist in institutions and be operated directly by CSC.

Suggestions for Effective Programming

A.    Program implementation and maintenance In establishing a substance abuse programming strategy the following factors should be considered:

1. staff (IC) responsible for the programs should have professional training, preferably a post-B.A. degree, with experience/training in the substance abuse or offender treatment area.
2. a detailed cost breakdown of the program be provided for a three year period and some guarantee be obtained of "intent to support" from the administration to fund the program.
3. the IC have access to a consultant (e.g., based in a university or an ARF-like provincial organization) knowledgeable in the area for advice. The creation of a Board, similar to some mental health models, is another model in this regard.
4. a professional role be designated within each CSC region to provide direction, support, training for field-based programs.
5. the IC and associates conduct a literature search of the relevant treatment research literature covering at least the last five years. The source materials recommended are the Psychological and Criminal Justice Abstracts. Documentation of the search should be provided. .
6. the IC is directly involved in the hiring and supervision of line staff. Line staff should be selected on the basis of characteristics predictive of job success and therapeutic effectiveness. The relevant variables are a) cognitive ability, b) relevant life experiences, c) education and training in either education, nursing, psychology, or social work, d) relationship characteristics - clarity, empathy, fairness in supervision and employs problem solving approach. Structured interviews, which can include situational assessments such as role-playing, in contrast to unstructured interviews, should be employed for staff selection.
7. where feasible a pilot program should be run before beginning the program officially.
8. the IC should contribute, in a tangible and practical way, to some therapeutic aspect of the program. The IC also carries out a quality assurance audit each year.
9. inmates in the program should be housed separately from the "main" population unless the entire institution is run as a therapeutic setting.
10.  enrolment in the program (for incarcerated inmates), where sentence length allows, is for both the first and last 3-6 months of their sentence.
11.  finally, a mechanism is put in place whereby staff are allowed direct input re: program modification.

B.    Classification

1. all substance abuse programs should employ classification techniques of some recognized validity and submit their classification protocol for peer review.
2. the types of individual factors that should be given priority in assessment are anti-social attitudes, cognitive skills, attitudes towards education, employment and leisure, situational analysis of substance abuse behaviour, psychopathy, socialization, social skills, and awareness of harm to others.
3. the classification instruments should be a)administered by a trained psychometrist or someone supervised by a psychologist, b) the data stored in a confidential file.
4. upon the initial administration of the classification test battery the protocol must be scored as to risk level. Those clients scoring "medium" to "high" risk should be admitted to the program. An override clause should be in place, however, for exceptional cases. That is, for the occasional "low" or very "high" risk cases, exceptions can be made for admittance to the program. The reasons for the override should be documented.

C.    Treatment

 The offender treatment literature clearly denotes that successful programs share the following general principles. Substance abuse programs should consider these guidelines:

1. Intensive services, cognitive-behaviour modification in nature, provided to higher risk clients. Risk is defined, as noted previously, by an objective measure.
2. Explicit reinforcement, modelling, problem solving, skill training of alternatives to pro-criminal styles of thinking, feeling, and acting.
3. Therapists relate to offenders in interpersonally sensitive and constructive ways while strongly supporting anti-criminal modelling.
4. Program contingencies enforced, firm but fair approach.
5. Program activities disrupt the delinquency network either within the prison or in the community.
6. Transfer training techniques (for institution programs) are employed to prepare inmates for community adaptation.
7. High levels of advocacy and brokerage as long as community agency offers appropriate services.

Unsuccessful programs generally display the following features:

1. Intensive services provided to low risk clients.
2. Traditional psychodynamic, client-centred, non-directive relationship oriented therapies.
3. Traditional medical model approaches - pharmacological, diet.
4. Punishment strategies and/or control techniques, i.e., electronic monitoring, without provision of any sort of service delivery.

  Therefore, unless there are well documented mitigating circumstances, it is recommended that the above noted classes of service delivery be abandoned.

 While cognitive behaviour modification programs have been shown to be the most effective they require certain features in their operation:

1. Highly structured with program content and contingencies under control of therapist.
2. Clients and staff share equally in program design, maintenance and enforcement of appropriate behaviours.
3. Clients re-labelled as co-therapists.
4. Negative peers are prevented from taking over program.
5. Effective internal control established to detect anti-social activities within the client group.
6. Limits placed on client turnover.
7. Positive reinforcers > punishment by 3:1.

 Types of effective behaviour modification modalities recommended, where appropriate, for offender substance abuse treatment programs are:

1. anger management training
2. controlled drinking
3. covert sensitization
4. family systems, family therapy
5. operant procedures
6. social-cognitive skills training
7. surveillance

 For community based programs there is one component that should be employed. That is, relapse prevention which involves the following steps:

1. Monitoring and anticipating problem situations.
2. Planning, rehearsing alternative responses.
3. Practising new behaviours in increasingly difficult situations and rewarding improved competencies, i.e., thinning and delay of reinforcement and b) fading.
4. Booster sessions.
5. Training significant others to provide reinforcement.
6. Finally, in regard to effective treatment there are three more important points.

 First, client factors must be matched with the strength of treatment. Medium to higher risk offenders profit most from the more intensive forms of service.

 Secondly, whenever possible, therapist characteristics must be matched with client factors. Certain types of offenders respond better to different sorts of therapists. If a program does not attempt to match clients and therapists in some manner, reasons should be provided.

 Thirdly, the recommended length of treatment should range from three to six months.

D.   Training and Evaluation

1. CSC should consider establishing a centralized institute for their substance abuse program staff. There are similar models in the U.S. (e.g., National Institute of Justice) and in Canada, for criminal justice agencies such as the police. The training institute might also be affiliated with an academic setting.
2. Some of the programs surveyed are experienced, knowledgeable, and competent. They are a valued resource and can serve as useful training models.
3. As noted previously, basic evaluation information is lacking. Many programs do not have the resources to carry out this role. The Research & Statistics Branch of CSC should facilitate and contribute to research on programming efforts. While the above model is centralized there may be instances whereby a region could assume some of the above tasks if the interest and adequate resources (e.g., university, CSC staff) were available.
4. All research and/or clinical information be categorized as to confidentiality and staff and client access.

CONCLUSION

 The rehabilitation initiatives of CSC have been timely. They have led to the generation of a meaningful amount of normative information as well as a tentative index of "quality" for offender substance abuse programs within CSC's domain. Deficits, some extensive, in program quality were reported.

 Rather than an occasion for despair, however, the knowledge produced now allows policy makers and clinicians to target deficits in a rational, constructive, and positive manner. Substance abuse programs can only benefit from this process.

FOOTNOTES

1. For examples of successful government/private agency studies, some of long standing, see Ross & Gendreau (1980). For another view on the matter raised by Lab see Gendreau & Ross, (1987) regarding maintenance of successful programs.
2. CSC has approximately 330 treatment programs for offenders.
3. The criticism of CPEI re: criterion validity, can also be applied to any of the standard audit methods employed to assess programs. We hope to confirm the predictive validity of the CPEI in the near future.

REFERENCES

Andrews, D.A. (1989). Recidivism is predictable and can be influenced: Using risk assessments to reduce recidivism. Forum on Corrections Research, 1, 11-18.

Andrews, D.A., Bonta, J., Hoge, R.D. (1990). Classification for effective rehabilitation: Rediscovering psychology. Criminal Justice & Behaviour, 17, 19-52.

Andrews, D.A., Zinger, I., Hoge, R.D., Bonta, J., Gendreau, P., & Cullen, F.T. (1990). Does correctional treatment work? A Clinically-relevant and psychologically-informed meta-analysis. Criminology, 28, 369-404.

Annis, H. (1982). Inventory of Drinking Strategies. Toronto: Addiction Research Foundation.

Annis, H.M. (1990). Effective treatment for drug and alcohol problems: What do we know? Forum on Correctional Research, 2, 18-23.

Cullen, F.& Gendreau, P. (1989). The effectiveness of correctional rehabilitation: Reconsidering the "Nothing Works" doctrine. In L. Goodstein & D.L. Mackenzie (Eds.), The American Prison: Issues in Research Policy, N.Y.: Plenum.

Gendreau, P. (1990). An outline of the principles of effective intervention. Division of Social Sciences, University of New Brunswick, Saint John, N.B.

Gendreau, P. & Andrews, D.A. (1979). Psychological consultation in correctional agencies: Case studies and general issues. In J.J. Platt & R.W. Wicks (Eds.) The Psychological Consultant, New York: Grune & Stratton.

Gendreau, P. & Andrews, D.A. (1990). Tertiary prevention: What the meta-analysis of the offender treatment literature tells us about "what works". Canadian Journal of Criminology, 32, 173-184.

Gendreau, P. & Ross, R.R. (1979). Effective correctional treatment: Bibliotherapy for cynics. Crime & Delinquency, 25, 463-489.

Gendreau, P. & Ross, R.R. (1987). Revivification of rehabilitation: Evidence from the '80s. Justice Quarterly, 2, 18-23.

Hill, J.K., Andrews, D.A., & Hoge, R.D. (in press). Meta-analysis of treatment programs for young offenders: The effect of clinically relevant treatment on recidivism with controls for various methodological variables. Canadian Journal of Program Evaluation.

Lab, S. (1990). Discussant to symposium by M.W. Lipsey, "What works in delinquency treatment: Results from 400 studies". Academy of Criminal Justice Science, Denver, March 1990.

Lab, S.P. & Whitehead, J.T. (1990). From "nothing works" to the "appropriate works": The latest step in the search for the secular grail. Criminology, 28, 405-417.

Lipsey, M. (1990). What works in delinquency research: Results from 400 studies. Claremont Graduate School, Claremont, CA, 91711.

Miller, W.R. & Hester, R.K. (1985). The effectiveness of alcoholism treatment methods: What research reveals. In W.R. Miller & N. Heather (Eds.). Treating addiction behaviors: Process of change. N.Y.: Plenum.

Research & Statistics Branch (1990a). Prevalence, Nature, and Severity of Mental Health Disorder Amongst Federal Male Offenders. Correctional Services Canada.

Research & Statistics Branch (1990b). Assessing offender substance-abuse problems at reception: Preliminary findings from the computerized lifestyle assessment instrument. Forum on Corrections Research, 2, 11-14.

Ross. R.R. & Gendreau, P. (1980). Effective Correctional Treatment. Toronto: Butterworth.

Appendix A
Substance Abuse Program
Respondents

Addictions Education	Prince Albert,Sask.	Prairies
Journey to Wholeness	Drumheller, Alberta	Prairies
Twilight Serenity Retreats	Drumheller, Alberta	Prairies
Smoking Cessation	RPC, Prairies	Prairies
Drug/Alcohol Education	RPC, Prairies	Prairies
Adult Children of Dysfunctional Families	RPC, Prairies	Prairies
AA, Sask. Farm Institution	Prince Albert, Sask.	Prairies
Narcotics Anonymous, Sask. Farm Institution	Saskatchewan	Prairies
Returned unanswered	Unknown	Unknown
Returned unanswered	Unknown	Unknown
Returned unanswered	Unknown	Unknown
St. Norbert Foundation	St. Norbert, Manitoba	Prairies
Returned unanswered	Unknown	Unknown
Skill Development Program	Stony Mountain, Manitoba	Prairies
Chemical Dependency Awareness	Stony Mountain, Manitoba	Prairies
Getting It Straight	Stony Mountain, Manitoba	Prairies
Novalco Alcohol Program, Sask. Penitentiary	Saskatchewan	Prairies
ADD-CAN Drug Abuser Program	Saskatchewan	Prairies
Returned unanswered	Unknown	Unknown
AA/NA, Westmorland Institution	Dorchester, N.S.	Prairies
Chemical Dependency Awareness	Drumheller, Alberta	Prairies
AA, Edmonton Institution	Edmonton, Alberta	Prairies
NA, Edmonton Institution	Edmonton, Alberta	Prairies
Lifestyle Assessment, Dorchester	Dorchester, N.S.	Prairies
PACADA, Addiction Education Program	Prince Albert, Sask.	Prairies
Getting It Straight	RPC, Prairies	Prairies
Alc/Sub Abuse Counselling for Natives	Bowden Institution	Prairies
Arrows to Freedom	Drumheller, Alberta	Prairies
AA, (Campus AA Group)	Drumheller, Alberta	Prairies
NA, (NA Freedom Group)	Drumheller, Alberta	Prairies
Familyships Program	RPC, Prairies	Prairies
Alcan & Novelco 12 Step Program	RPC, Prairies	Prairies
Individual Counselling	RPC, Prairies	Prairies
Returned unanswered, Salvation Army, TO	Toronto, Ontario	Ontario
Mann House Corp.	Charlottetown, P.E.I.	Atlantic
Sand River, CCC	Parrsboro, N.B.	Atlantic
Sobriety House	Ottawa, Ontario	Ontario
Programme d'information, Archambault	Montréal, Québec	Québec
AA Partage Archambault	Montréal, Québec	Québec
Atlantic Substance Abuse Program Springhill	Springhill, N.S.	Atlantic
NA, Springhill Institution	Springhill, N.S.	Atlantic
AA, Springhill Institution	Springhill, N.S.	Atlantic
Life Styles Projects (Computers)	Springhill, N.S.	Atlantic
Queen's Co. Addiction Services	Charlottetown, P.E.I.	Atlantic
La Maisonée d'Oka	Oka, Québec	Québec
Alcohol/Substance Abuse Program
Addiction Follow-Up	Kingston Pen., Ontario	Ontario
Pre-Release Substance Abuse Program, Joyceville	Kingston, Ontario	Ontario
JI Recovery Program Joyceville Institute	Kingston, Ontario	Ontario
Bibliotherapy, Dorchester Library	Dorchester, N.B.	Atlantic
NA, Dorchester Penitentiary	Dorchester, N.B.	Atlantic
Native Drug & Alcohol Workshop, Dorchester	Dorchester, N.B.	Atlantic
AA, Dorchester Penitentiary	Dorchester, N.B.	Atlantic
Centre de Traitement Toxicomanie,
Pavillon E. Grégoire	Québec	Québec
Salvation Army Yukon Adult Resource Centre	Whitehorse, Yukon	Pacific
Programme Portage, Lac Echo Prévost	RCSCC Laurentides, Québec	Québec
Substance Abuse Pre-Release Program Warkworth	Gravenhurst, Ontario	Ontario
Établissement Résidentielle Communautaire l'Étape	Sherbrooke, Québec	Québec
Programme Virage, Unité "L"	Québec	Québec
St-Leonard's Substance Abuse Treatment Centre	Hamilton, Ontario	Ontario
Atlantic Substance Abuse Program	St-John, N.B.	Atlantic
Pro-Soft Substance Abuse Program	Surrey, British Columbia	Pacific
Pro-Soft Training Institute Substance Abuse Program	Province not specified	Pacific
Alcohol & Drug Program, Wm. Head Inst.	Victoria, B.C.	Pacific
Skill Development Program	Rockwood Institution, Prairies	Prairies
Getting It Straight	Rockwood Institution, Prairies	Prairies
Chemical Dependency Awareness	Rockwood Institution, Prairies	Prairies
Substance Abuse Program, Atlantic Institution	Renous, N.B.	Atlantic
AA, Atlantic Institution	Renous, N.B.	Atlantic
NA, Atlantic Institution	Renous, N.B.	Atlantic
Clean & Sober	Unknown	Unknown
Alcare Place	Halifax, N.S.	Atlantic
Établissement carcéral: AA	Quebec	Quebec
Sub Abuse Program, Westmorland Institution	Dorchester, N.B.	Atlantic
Christian Education Program, Chapel	Springhill Institution	Atlantic
Camillus Centre, St. Joseph's General Hospital	Elliot Lake, Ontario	Ontario
Talbot House	North Sydney, N.S.	Atlantic
Substance Abuse Relapse Prevention Program	Kentville, N.S.	Atlantic
Alternatives de la Toxicomanie, Étab. Drummond	Frontenac	Québec
Groupe l'Éclaircie, Établissement Drummond	Drummondville, Québec	Québec
AA, Établissement Drummond	Drummondville, Québec	Québec
Royal Ottawa Hospital Addiction Services	Ottawa, Ontario	Ontario
AA, français et anglais, E.M.S.F.	Québec	Québec
Toxicomanie, Établissement Montée St-François	Québec	Québec
St-Leonard Society	Brantford, Ontario	Ontario
W.L. Judson, Beaver Creek Institution	Gravenhurst, Ontario	Ontario
Alcohol & Drug Education/Counselling	Gravenhurst, Ontario	Ontario
AA Astra & Discussion Group, Warkworth	Warkworth, Ontario	Ontario
Drug Addiction Studies Program, Warkworth	Warkworth, Ontario	Ontario
Alcohol & Drug Education Program, Kingston	Kingston, Ontario	Ontario
NA, Établissement Leclerc	Québec	Québec
AA, Établissement Leclerc	Québec	Québec
HAPEC House	Belleville, Ontario	Ontario
Returned unanswered	Dorchester, N.B. (postmark)	Atlantic
Returned unanswered	Dorchester, N.B. (postmark)	Atlantic
Returned unanswered	Dorchester, N.B. (postmark)	Atlantic
Returned unanswered	Unknown	Unknown
BIIPMAD, Bowden Institution	Alberta	Prairies
Returned unanswered	Dorchester, N.B. (postmark)	Atlantic
Centre Correctionnel Communautaire Ogilvey	Montréal, Québec	Québec
Salvation Army Harbor Light	Sault Ste-Marie, Ontario	Ontario
C.B.I. Recovery - Brentwood	Ontario	Ontario
Kingston, Collins Bay Inst.
Programme de Toxico, Étab. Ste-Anne-des-Plaines	Québec	Québec
Enfants Adultes de Parents	Québec	Québec
Alcooliques, La Macaza
AA Francophones, Établissement La Macaza	Québec	Québec
Journée Intensive AA Francophone, La Macaza	Québec	Québec
Journée Intensive AA Anglophone, La Macaza	Québec	Québec
AA Anglophone, Établissement La Macaza	Québec	Québec
Cours Toxicomanie, Étab. La Macaza	Québec	Québec
Harbor Light Centre (Addictions Program)	St-John's, Newfoundland	Atlantic
Native Substance Abuse Program, Kent	British Columbia	Pacific
Women's Substance Abuse Program	Kingston, Ontario	Ontario