In 1994, Nick Heather, Alex Wodak, Ethan Nadelmann and Pat O'Hare published the book Psychoactive Drugs and Harm Reduction: From Faith to Science (Heather, Wodak, Nadelmann & O'Hare, 1994). This edited collection, based largely upon the proceedings of the Third International Conference on the Reduction of Drug-related Harm, effectively set a new agenda for the addiction field and heralded the gradual demise of abstinence-based treatments in Europe.

In part, this slow fall from grace was due to the failure of therapeutic communities (TCs) and other abstinence-based services to respond positively to the challenges posed by the advent of HIV/AIDS. The new urgency with which governments and public health officials harnessed the work of drug treatment agencies to the broader public health imperatives of infection control and the promotion of safer sex effectively caught TCs napping. Whilst the rest of the addiction field moved to accommodate these new objectives in some form, TCs remained obdurately opposed to developments which they saw as undermining the purity of their purpose and were, as a result, diminished in relative importance.

But in part, too, this reduction in influence - and in the early days of drug treatment, TCs enjoyed influence far greater than might have been expected given their relatively small footprint (Yates, 2003) - also stemmed from the perception that the ‘new harm reductionists' had captured the higher ground of scientific evidence and that the rest was simply smoke and mirrors.

It has taken a long time for TCs and other abstinence-based services to recover from this period. But the recovery is under way and this issue is yet another step along that journey.

What is offered here is proof that the need to establish a scientific basis for our work is being taken seriously. In addition, the recent rebirth of interest in ‘recovery' sparked, at least in part, by the exponential growth of self-help recovery movements (Alcoholics Anonymous of course, but both Narcotics Anonymous and Smart Recovery and even some non-aligned groups have grown dramatically over the past few years) has allowed the TC movement to re-establish its scientific credentials.

Thus, in this issue of the journal, several articles detail work under way to improve performance and outcomes. Others examine the relationship between TCs and mutual help fellowships and their impact on the wider community.

The paper by Kressel et al. describes the trialling of a protocol to improve TC staff performance. The subsequent paper by Burkett describes a small study to examine the correlation between improved social and coping skills in TC residents and their drug/alcohol-related outcomes. Addazi, Marini and Rago examine the relationship between length of time in treatment, the nature of the termination of the treatment etc., and treatment outcomes with a sample of residents in an Italian TC. In a similar study, Freestone and Goodman explore the treatment outcomes and treatment experiences of a group of drug users entering the Ley Community in Oxford.

The drugs-crime axis has been a preoccupation of many European governments in the past few years and has, to some extent, overtaken the public health imperatives which drove the expansion of the treatment field in the last decades of the 20th century. In the UK in particular, these issues have been compounded by political and public concerns over public drunkenness and the resultant public disorder.

In her article, McDonald considers the impact of a prison-based TC in America upon the social and coping skills of the women prisoners involved, whilst Deaner and colleagues explore the impact upon local crime levels of the establishment of Oxford Houses in certain residential areas (and found that there was none!).

Oxford Houses emerged in the latter half of the last century as a fascinating example of a fusion between mutual-aid fellowships like AA and NA and self-generating sober-living movements. Certainly, mutual-aid fellowships, sober-living initiatives and TCs have much to learn from each other, and the final article by Weegmann examines the similarities and differences between AA and the British democratic TCs.

All told, we believe that this is an impressive and interesting contribution to the development of a robust and defensible evidence-base for the abstinence-based sector of the addiction treatment field both in the UK and further afield.

References

Heather, N., Wodak, A., Nadelmann, E. & O'Hare, P. (Eds.). (1994). Psychoactive Drugs and Harm Reduction: From Faith to Science. London: Whurr Publications.

Yates, R. (2003). A brief moment of glory: the impact of the therapeutic community movement on drug treatment systems in the UK. International Journal of Social Welfare, 12(3): 239-243.

______________________________________________________________________________________

Rowdy Yates is Senior Research Fellow at the University of Stirling, UK. Email: p.r.yates@stir.ac.uk
Barbara Rawlings is an Honorary Fellow at the University of Manchester, UK. Email: barbara.rawlings@talk21.com
Simon McArdle is Senior Lecturer in Mental Health at the School of Health and Social Care, University of Greenwich, London, UK. Email: s.p.mcardle@gre.ac.uk

David Kressel, Keith Morgen, George De Leon, Gregory Bunt and Britta Muehlbach

ABSTRACT: This study addresses the critical need to utilize research findings in real treatment settings. It examines the clinical utility of a protocol designed to improve the way therapeutic community (TC) treatment is provided. The protocol uses psychometrically-tested staff and client-rated instruments that measure client progress on 14 theoretically-based practitioner-validated goals of TC treatment. Clients in Daytop Village residential treatment facilities are randomly assigned to an experimental (protocol is operating; N=111) or control (protocol is not operating; N=214) condition, and compared on seven variables correlated with effective treatment. It is hypothesized that clients in facilities where the protocol is operating will exhibit greater improvements on these variables than clients in equivalent facilities where the protocol is not operating.

Repeated measures ANOVA with the between-subjects covariate of condition (protocol or control) was conducted on all the measures. Significant time by condition interactions were found for counselor competency and staff helpfulness. Within the protocol condition, counselor competency significantly increased between time point 1 and time point 2. Similarly, within the protocol condition impressions of staff, quality significantly increased between the two time points. All seven dependent measures show increases from time 1 to time 2 in the protocol group, while the same is true for only three in the control group.

The results are encouraging not only because the protocol shows promise as an effective tool for improving treatment but also because it utilizes evidence-based research instruments that measure client progress in real treatment settings.

Author disclosure

1) Acknowledgement and role of funding source

Funding for this study was provided by NIDA Grant R01 DA3075; the NIDA had no further role in the study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.

2) Contributors

David Kressel and George De Leon designed the study and developed the instruments in the protocol. Keith Morgen undertook the statistical analysis. Britta Muehlbach managed the literature searches and summaries of previous related work. David Kressel, Greg Bunt and Britta Muehlbach developed the protocol. David Kressel wrote the first draft of the manuscript.

3) Conflicts of interest

There are no conflicts of interest.

4) Acknowledgements

Special thanks to the residents of Daytop Village. Their willingness to share in their struggle will hopefully benefit those that follow.

1. Introduction

There is a growing gap between research and practice in the drug treatment field evidenced by insufficient application of current research findings in actual treatment settings. The gap is due in part to the youth of the drug treatment field and the lack of consensus on how to view the problem of drug dependence (Etheridge & Hubbard, 2000). For instance, medical, psychological, social work, addiction counseling, milieu-therapy and cognitive approaches each have a unique and varied perspective on the causes and treatments for drug dependence.

This paper presents the results of a randomized study of a new protocol for improving clinical practice (PICP) in the therapeutic community (TC) setting. The protocol, a comprehensive application of research to practice, utilizes evidence-based instruments that measure client clinical progress in real treatment settings. These instruments were constructed on the basis of a theoretical view of TC treatment (De Leon, 2000). The protocol is a counseling intervention where both client and counselor can re-focus efforts, examine client strengths and needs, mutually design treatment plans and track achievement of goals.

The seven dependent measures are chosen because the literature shows these variables are correlated with effective treatment. For example counseling requires a strong therapeutic alliance (Horvath, 1991). Within the PICP this alliance is facilitated via the counselor listening to and respecting the client's perspective of treatment progress and by working with the client to establish mutually agreed upon treatment goals.

Studies of treatment seek to identify the ‘active ingredients' that contribute to positive outcomes (Moos, Finney & Cronkite, 1990).

These ‘active ingredient' variables include therapeutic alliance, client engagement and participation, and program structure. Studies have demonstrated there are beneficial effects of increased therapeutic alliance between staff and client for treatment outcomes. Research by Simpson and colleagues with a sample of methadone maintenance clients indicates that the strength of the therapeutic relationship between client and counselor is positively correlated with client attendance, lower drug use during treatment, and retention in the program (Simpson, Joe, Rowan-Szal & Greener, 1997). A study of probationers in short-term residential treatment found that recidivism rates are negatively correlated with client ratings of counselor competence and peer support (Simpson, Joe, Broome, Hiller, Knight & Rowan-Szal, 1997). Similarly, staff perceptions of client therapeutic involvement is positively correlated with client retention in a TC. Other research has identified affiliation, connection and trust between staff and clients as essential elements of effective TC treatment (Bell, 1994).

Extent of client engagement and participation in program services is another ‘active ingredient' correlated with positive treatment outcomes. One study of TC programs shows that greater resident participation in treatment increases length of stay in the program (Condelli & Dunteman, 1993). In other treatment modalities involvement in social activities predicts length of treatment and program completion (Burling, Seidner, Robbins-Sisco, Krinsky & Hanser, 1992; Moos, King, Burnett & Andrassy, 1997). In addition, residents who develop social relationships with their peers in treatment have higher retention rates than those who do not (Bauman et al., 1988). These findings are especially relevant in TCs where interaction with peers is a major component of treatment and residents are typically held partially responsible for the recovery of their peers. Research studying treatment progress using client, staff and peer ratings reveals high correlations between the three scales (Czuchry, Dansereau, Sia & Simpson, 1998). The same study reports that progress measures have the potential to provide clients with valuable therapeutic feedback.

Besides program evaluation (which determines program effectiveness), quality assurance has become an important research area, with questions focusing on client use and satisfaction with treatment services (Wilkerson, Migas & Slaven, 2000). A positive correlation exists between client satisfaction and retention (Chan, Sorenson, Guydish, Tajima & Acampora, 1997).

Traditional psychotherapy has long been aware that client progress in treatment and treatment outcomes is correlated with the strength and quality of the therapeutic alliance between client and primary therapist (e.g. Bordin, 1979; Saunders, Howard & Orlinsky, 1989). In fact, Luborsky and Auerbach (1985) find that 70% of some 85 studies show evidence of a positive relationship between therapeutic alliance and treatment outcome. Therapeutic alliance is a predictor of outcome in different types of individual psychotherapy, in group settings, in residential treatment and across different client populations (Bergin & Garfield, 1994). In the clinical counseling literature, the counseling alliance is considered crucial to therapeutic outcomes (Gelso & Carter, 1985). This link between alliance and outcome is shown in the substance abuse treatment field (Broome, Knight, Hiller & Simpson, 1996). In work from Project Match (Connors, Carroll, DeClemente, Longabaugh & Donovan, 1997), client self-report ratings of alliance with counselors predict treatment participation and positive drinking-related outcomes. Although the TC relies heavily on peer interaction and community as method (De Leon, 2000), individual counseling still plays an important role. In the NIDA Cocaine Treatment Study, several types of counseling interventions are examined. Findings highlight that individual counseling coupled with group work is more effective in treating cocaine addiction than supportive-expressive psychodynamic and group therapies, cognitive therapy and group therapy, or group therapy alone (Crits-Christoph et al., 1999).

Similarly, brief counseling focuses on client strengths (O'Hanlon & Weiner-Davis, 1989). Goals met, progress made, and responsibilities earned within the TC are all discussed during the dialogue central to the protocol used in this study. The protocol underscores client strengths and successes, not only weaknesses. Although these matters may be discussed in regular counseling sessions, the protocol formalizes a structure and schedule for these discussions.

Clients in facilities where the PICP is operating are expected to exhibit greater improvements on these types of ‘active ingredients' than clients in equivalent facilities where the protocol is not operating.

2. Methods

2.1 The agency

Daytop Village, Inc. is one of oldest TCs in the United States. Founded in 1963 as a pilot project funded by the National Institute of Mental Health (NIMH), Daytop has grown to be one of the largest drug rehabilitation programs in the country, operating from twenty-one sites in two states. Over the course of one year, Daytop Village services up to 8,500 adult and adolescent clients as well as their families, with an average census of about 3,000 clients a day.

In New York State, Daytop's clients are placed in long-term, intermediate or short-term residential treatment, or outpatient treatment, according to their specific needs. The short-term residential program has a planned duration of one to three months, and the ‘core' phases of intermediate and long-term resid-ential treatment last approximately six and twelve months respectively. Residential treatment has three components: an assessment and referral stage, a primary treatment stage, and a re-entry stage.

Daytop employs a comprehensive approach to the treatment of substance abuse utilizing a network of clinical and auxiliary services within both residential and ambulatory settings. Within the framework of the TC approach, group therapies, individual counseling, self-help principles, behavior management tools, as well as emotional/psychological, intellectual/spiritual, and vocational interventions are applied to address and change the client's substance abuse and the accompanying behavior patterns. Daytop also provides medical services, mental health care, vocational and educational counseling and training, and legal assistance to all clients as well as specialized services for women and HIV positive clients. Daytop International has assisted 80 countries worldwide in establishing substance abuse treatment programs, adapted for the particular culture of the country in which it is located (Daytop Village, Inc., Annual Report, 2002-2003).

2.2 The protocol

The PICP has three components: a) theory-based staff training, b) monthly repeated assessment of client change by both clients and staff using instruments that measure client progress towards treatment goals, and c) procedures for tracking and utilizing the recorded progress information to improve clinical practice.

a) On-site staff training consisting of three 90-minute sessions preceded implementation of the protocol. The first session covers basic TC theory (De Leon, 2000). The second session is an overview of the instruments that measure client progress in treatment including the definitions of the competencies they measure. The third session describes the operation and logistics of the protocol; specifically how to use the progress instruments in routine treatment. The curriculum for the training sessions covers:

Session 1
1) The distinctions between the terms ‘progress in treatment', ‘treatment process' and ‘treatment outcome'.
2) The lifestyle and identity change that treatment strives to achieve in the individual.
3) The unique TC approach to treatment: ‘community as method'.
4) A description of what is expected of the client during the various phases of treatment.

Session 2
1) The definitions of the 14 behavioral, attitudinal, and cognitive competencies measured by the progress instruments.
2) A discussion of what these competencies look like in practice, how they serve as in-treatment goals, and how specific competencies are clinically relevant to individuals.

Session 3
1) An overview of the purpose of the protocol, procedures and responsibilities of clients and staff for implementing the protocol, and instructions for orienting new clients.
2) A description of the structure, time frame, and format for completing the progress instruments and the system for recording, tracking, and sharing this information.
3) Instructions to staff explaining that information from client progress instruments is not to be used as the basis for punitive measures against clients.
4) A discussion of the techniques and procedures for staff and clients to utilize the recorded client progress information to improve treatment. Topics include using the information to: clinically assess clients, enhance client problem recognition, guide clinical decisions, provide focused treatment, and direct clients towards strategies for using the program to work on needed treatment areas or deficits.
5) The schedule and format for meeting with clients and techniques for using the progress information to develop treatment plans.
6) Ethical issues and confidentiality of the data. Data are coded by number with no names or identifying information and kept in locked cabinets. In addition, client self-evaluations are not used by staff to administer privileges or sanctions.

Clients were initially skeptical about confidentiality and the potential for losing privileges but after reassurances from the research team and the facility director followed by actual experience they saw this promise was kept as evidenced by their continued willingness to sign an informed consent to participate.

To minimize training effects both the experimental and control facilities receive session one (TC theory). Only the experimental condition receives the last two training sessions (protocol implementation) since the experimental condition is where the protocol operates.

b) The PICP uses two psychometrically tested staff and client rated instruments to measure client progress on fourteen theoretically based, practitioner validated competencies in TC treatment (Table 1). The Staff Assessment Summary (SAS) is completed by the counselor. It contains 14 items, one for each competency. It is designed to help staff document client progress and identify specific issues clients need to address. The Client Assessment Summary (CAS) is completed by the client. This self-report instrument is designed to accelerate client problem recognition. It contains the same 14 items in the SAS but the items are worded in the first person. This enables direct comparison of staff and client ratings to assess the accuracy of client self-report and to provide information for focusing treatment. All 14 items are measured along a 5-point Likert Scale ranging from Strongly Disagree to Strongly Agree.

Table 1: Competencies or goals of client change in TC treatment
______________________________________________________________________________________

Developmental Dimension: Evolution of the individual in terms of personal growth

1. Maturity (self-regulation, social management)
2. Responsibility (accountability, meeting obligations)
3. Values (integrity)

Socialization Dimension: Evolution of the individual as a pro-social member of society

4. Drug/Criminal Lifestyle (social deviancy)
5. Maintains Images (social vs. antisocial lifestyle)
6. Work Attitude (attitude appropriate for work world)
7. Social Skills (ability to relate to people)

Psychological Dimension: Basic cognitive and emotional skills
8. Cognitive Skills (awareness, judgment, insight, reality testing)
9. Emotional Skills (communication and management of feeling states)
10. Self-Esteem/Self-Efficacy (sense of wellbeing)

Community Membership Dimension: Evolution of the individual's relationship to the TC
11. Understands Program Rules, Philosophy, and Structure
12. Community Engagement and Participation
13. Attachment and Investment in the Community
14. Role Model (lives by example, teaches others)

______________________________________________________________________________________

In an earlier study the CAS and SAS were used with a sample of 346 TC clients and their primary counselors. Results show the instruments reliably measure client clinical progress in treatment (Kressel, De Leon, Palij & Rubin, 2000). The PICP utilizes these instruments in two main ways: 1) staff and clients use the SAS and CAS to measure and record client progress along the 14 competencies and, 2) staff and clients use this recorded information in groups, one-to-one sessions and other program activities to accelerate problem recognition and to develop treatment plans.

c) After thirty days and then every two months progress scores are recorded on a staff and client tracking form (Figure 1). Clients meet with their counselor to review progress scores and examine any discrepancies between counselor SAS and client CAS ratings. The counselor helps the client identify clinically relevant areas (where deficits exist) and suggests individual, group and other program activities to work on these areas. If counselor and client views are very discordant, peer support strategies are utilized to examine differences in closer detail.

Fidelity of implementation is measured by examining clinical folders of clients in the protocol group to see if they contain the progress surveys. More than 91% of the surveys were completed on a timely basis.

2.3 The design and sample

The project is conducted at adult residential Daytop TC facilities. This study compares clients (Table 2) randomly assigned to the experimental condition (protocol is operating; N=111) to clients randomly assigned to the control condition (protocol is not operating; N=214). Repeated Measure Analysis of Variance of seven dependent measures is used to compare clients in the two conditions. The sample consists of clients with CAS, SAS and dependent measure data collected at 30 days (N=660) and at 90 days (N=415). However, only 325 participants had data assessments at both 30 and 90 days.

Table 2: Client demographics (N=325)

______________________________________________________________________________________

______________________________________________________________________________________

The Swan Lake facility located in Sullivan County, New York was randomly assigned to the experimental condition. Two comparable adult residential facilities, Parksville, New York (Sullivan County) and Springwood, New York (Dutchess County), were assigned to the control condition. Random assignment is by facility rather than by client to minimize contamination of protocol implementation. The Swan Lake and Parksville facilities have an average daily census of 200, whereas Springwood's was 160 at the time of the study.

The protocol and non-protocol client samples exhibit no significant differences in ethnicity, gender, education and treatment history. Clients in the experimental and control condition also do not significantly differ in age, drug use, employment, and legal status. The average client age is 36 years. Alcohol (70%) is the most commonly reported substance used (however, mostly as a secondary or tertiary drug), followed by marijuana (54%), cocaine (48%), crack (45%), and heroin (34%). Nearly 73% of clients are unemployed or not in the labor force, and approximately 50% have not earned a high school diploma or GED when entering treatment. Fifty-nine percent report at least one active legal case when entering treatment and clients at each facility enter, transfer, and leave treatment at nearly identical rates.

Treatment is uniform across the facilities, consisting of a structured TC environment providing similar daily schedules, routines, treatment requirements, staff involvement, and client-counselor interactions. Staffing patterns are also consistent across the three facilities including: counselors, social workers, nurses, physicians, dieticians, HIV coordinator, vocational counselor, educational counselor, teachers, administrative personnel, facility director and assistant directors.

2.4 The dependent measures

The seven dependent measures include satisfaction with treatment, two self-perception measures and four measures of client perceptions of staff.

Client satisfaction with treatment. Satisfaction with treatment is measured with the short-form of the Service Evaluation Questionnaire, SEQ-8 (Nguyen, Attkinson & Stegner, 1983). Measured dimensions of satisfaction include type of treatment, amount of treatment, quality of treatment, and general satisfaction. The SEQ-8 contains eight self-administered items on a 4-point Likert scale (low to high satisfaction). Within our sample (N=325) the SEQ-8 has high reliability (α=0.91) and inter-item correlations ranging from r=0.50 to 0.72.

Client therapeutic engagement, personal progress and the four measures of client perception of staff (counselor trust, helpfulness, rapport, and competence). These are measured using subscales of the Evaluation of Self and Treatment Survey developed at the Institute of Behavioral Research at Texas Christian University (Hiller, 1996). These scales contain items measured on a 5-point Likert scale ranging from ‘strongly disagree' to ‘strongly agree'. Reliability for all these scales was strong within our sample (α ranging between .86 and .95).

The dependent measure surveys are administered to clients at two time points: 30 days and 90 days after entering treatment. This coincides with the CAS and SAS data collection time points.

2.5 The analytic approach

It is hypothesized that clients in facilities where the protocol is operating will exhibit greater improvements on the dependent measures than clients in facilities where the protocol is not operating. To test this hypothesis, repeated measures ANOVA is used. Typically at least three time points are examined with this procedure, so a between-subjects' covariate of condition (protocol or control) is included. Specifically we are looking at a between-subjects' effect interacting with time. In this way the comparison between the protocol and control group is analyzed along with the comparison between time 1 and time 2 in each group.

3. Results

Repeated measures ANOVA with the between-subjects' covariate of condition (protocol or control) was conducted on all the measures. Time 1 and time 2 means all dependent measures are presented in Table 3. Within the protocol condition, counselor competency significantly increased between time point 1 (M=29.23, SD=6.66) and time point 2 (M=30.68, SD=6.87). Similarly, within the protocol condition, impressions of staff helpfulness significantly increased between time point 1 (M=7.64, SD=1.81) and time point 2 (M=7.94, SD=1.63).

Repeated measure ANOVA results are presented in Table 4. Significant time by condition interactions were found for counselor competency (Wilk's λ=.99, F=4.62, p=.03) and staff helpfulness (Wilk's λ=.99, F=4.26, p=.04). Client perceptions of staff competency and helpfulness are the two variables that most significantly improved in the protocol group compared to the control group. Because the protocol is designed for use by both staff and clients together, this finding is relevant to the TC treatment approach.

There was no significant interaction but only a significant main effect for time for the personal progress variable (Wilk's λ=.91, F=33.53, p<.001), indicating that overall perception of personal progress significantly increased between time point 1 (M=14.26, SD=3.35) and time point 2 (M=15.52, SD=3.06).

Correlations across all seven dependent measures demonstrated significant and positive relationships between the variables, with correlation strengths ranging between .20 and .66, with one outlier at a high of .81 (see Table 5). With the exception of .81, the correlations are in a range indicating the measures relate to each other but do not overlap.

Table 3: Dependent measure means and SD

Table 4: Repeated measures ANOVA
______________________________________________________________________________________

Table 5: Correlations of dependent measures

*p<.001
______________________________________________________________________________________

All seven dependent measures show increases from time 1 to time 2 in the protocol group while the same is true for only three measures in the control group. Admittedly not all increases are significant but it is extremely unlikely that this desired directional change will occur in all seven out of seven variables purely by random chance.

4. Discussion

These findings are encouraging in that they are consistent with the hypothesis that the protocol improves treatment. The study addresses a significant issue: whether the attainment of specific milestones during treatment can be assessed in a clinically useful way. An important potential benefit is that the effectiveness of new interventions can be evaluated against the reference point of during-treatment progress rather than post-treatment outcome. This saves time and resources and yields the added benefit of providing the foundation for future studies that will examine the relationship between proximal progress and long-term outcome. Such studies could help inform the appropriateness of current treatment goals and ways to modify and improve them.

1. Assisting in the determination of when a person is ready to advance to the next phase of treatment.
2. Accelerating client ‘problem recognition' and focusing treatment on clinically relevant areas.
3. Helping inform program-wide treatment issues, e.g. appropriate planned duration of treatment and the relationship between progress in treatment and treatment outcome.
4. Assisting in staff orientation, training and paperwork preparation, e.g. treatment plans, progress reports, etc.
5. Evaluating the effectiveness of new treatment techniques in the context of in-treatment progress rather than long-term follow-up.

The clinical utility of the protocol in actual practice is supported by these findings; however the mechanism by which the protocol influences treatment is yet to be determined. Measurement of client and counselor appraisal of treatment progress may lead to a more active stance by the counselor during the period leading up to the next progress assessment. The counselor, in anticipation of this assessment, will likely pay closer active attention to clients so they can more accurately evaluate client progress. The protocol may actually enhance counselor engagement with clients in and out of counseling sessions. Goals are discussed, progress measured, and new treatment goals are reviewed as components of the protocol. This is precisely the purpose of treatment planning; therefore the protocol holds promise for writing more concise yet clinically relevant treatment plans. Improving the quality and ease of staff preparation of paperwork is a valuable benefit in itself. Although competent counselors routinely review goals with their clients, the implementation of this protocol formalizes this process. Specifically, goals are not just discussed but also tracked and measured.

The progress instruments used in the protocol may have applicability beyond the TC. The 14 competencies reflect areas of change relevant to substance abuse treatment in general. Baseline differences, change rates and the relationship between progress during treatment and treatment outcome likely vary by treatment modality but are relevant to a range of treatment approaches. These issues are currently being addressed in future studies.

References

Bauman, L., Nierporent, H.J., Ferguson, J., Klein, J., Dunne, G. & Rudoltz, C. (1988). Demystifying the patient dropout: A study of 122 brief-stay day-treatment center admissions. International Journal of Partial Hospitalization, 5, 215-224.

Bell, D.C. (1994). Connection in therapeutic communities. The International Journal of the Addictions, 29(4), 525-543.

Bergin, A.E. & Garfield, S.L. (1994). Handbook of Psychotherapy and Behavior Change (4th ed.). Oxford, England: John Wiley and Sons.

Bordin, E.S. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research, and Practice, 16, 252-260.

Broome, K.M., Knight, K., Hiller, M.L. & Simpson, D.D. (1996). Drug treatment process indicators for probationers and prediction of recidivism. Journal of Substance Abuse Treatment, 13, 487-491.

Burling, T.A., Seidner, A.L., Robbins-Sisco, D., Krinsky, A. & Hanser, S.B. (1992). Batter Up! Relapse prevention for homeless veteran substance abusers via softball league participation. Journal of Substance Abuse, 4, 407-413.

Chan, M., Sorenson, J.L., Guydish, J., Tajima, B. & Acampora, A. (1997). Client satisfaction with drug abuse day treatment versus residential care. Journal of Drug Issues, 27, 367-377.

Condelli, W.S. & Dunteman, G.H. (1993). Issues to consider when predicting retention in therapeutic communities. Journal of Psychoactive Drugs, 25, 239-244.

Connors, G.J., Carroll, K.M., DeClemente, C.C., Longabaugh, R. & Donovan, D.M. (1997). The therapeutic alliance and its relationship to alcoholism treatment participation and outcome. Journal of Clinical Psychiatry, 65, 588-598.

Crits-Christoph, P., Siqueland, L., Blaine, J., Frank, A., Luborsky, L., Onken, L.S., Muenz, L., Thase, M.E., Weiss, R.D., Gastfriend, D.R., Woody, G., Barber, J.P., Butler, S.F., Daley, D., Bishop, S., Najavits, L.M., Lis, J., Mercer, D., Griffin, M.L., Moras, K. & Beck, A. (1999). Psychosocial treatments for cocaine dependence: Results of the National Institute on Drug Abuse Collaborative Cocaine Treatment Study. Archives of General Psychiatry, 56, 493-502.

Czuchry, M., Dansereau, D.F., Sia, T.L. & Simpson, D.D. (1998). Using peer, self, and counselor ratings to evaluate treatment progress. Journal of Psychoactive Drugs, 30(1), 81-87.

Daytop Village, Inc. (2002-2003). Annual Report. Daytop Village, Inc., New York City.

De Leon, G. (2000). The Therapeutic Community: Theory, Model and Method. New York: Springer Publishing Company.

Etheridge, R.M. & Hubbard, R.L. (2000). Conceptualizing and assessing treatment structure and process in community-based drug dependency treatment programs. Substance Use Misuse, 35, 1757-1796.

Gelso, C.J. & Carter, J.A. (1985). The relationship in counseling and psychotherapy: Components, consequences, and theoretical antecedents. The Counseling Psychologist, 13, 155-244.

Hiller, M.L. (1996). Correlates of Recidivism and Relapse for Parolees who Received In-Prison Substance Abuse Treatment in Texas. Unpublished doctoral dissertation. Texas Christian University, Fort Worth, TX.

Horvath, A.O. & Symonds, B.D. (1991). Relation between working alliance and outcome in psychotherapy: A meta-analysis. Journal of Counseling Psychology, 38, 139-149.

Kressel, D., De Leon, G., Palij, M. & Rubin, G. (2000). Measuring client clinical progress in therapeutic community treatment: The therapeutic community Client Assessment Inventory, Client Assessment Summary, and Staff Assessment Summary. Journal of Substance Abuse Treatment, 19, 267-272.

Luborsky, L. & Auerbach, A. (1985). The therapeutic relationship in psycho-dynamic psychotherapy: The research evidence and its meaning for practice. In R. Hales & A. Frances (Eds.), Psychiatry Update Annual Review (pp. 550-561). Washington, DC: American Psychiatric Association.

Moos, R.H., Finney, J.W. & Cronkite, R.C. (1990). Alcoholism Treatment: Context, Process, and Outcome. New York: Oxford.

Moos, R.H., King, M.J., Burnett, E.B. & Andrassy, J.M. (1997). Community residential program.

Nguyen, T.D., Attkisson, C.C. & Stegner, B.L. (1983). Assessment of patient satisfaction: Development and refinement of a service evaluation questionnaire. Evaluation and Program Planning, 6, 299-314.

O'Hanlon, W.H. & Weiner-Davis, M. (1989). In search of solutions: A new direction in psychotherapy. New York: Norton.

Saunders, S.M., Howard, K.I. & Orlinsky, D.E. (1989). The therapeutic bond scales: Psychometric characteristics and relationship to treatment effectiveness. Psychological Assessment, 1, 323-330.

Simpson, D.D., Joe, G.W., Broome, K.M., Hiller, M.L., Knight, K. & Rowan-Szal, A. (1997). Program diversity and treatment retention rates in the Drug Abuse Treatment Outcome Study (DATOS). Psychology of Addictive Behavior, 11, 279-293.

Wilkerson, D., Migas, N. & Slaven, T. (2000). Outcome-oriented standards and performance indicators for substance dependency and rehabilitation programs. Substance Use Misuse, 35, 1679-1704.

______________________________________________________________________________________

David Kressel, PhD and George De Leon, PhD are at the National Development and Research Institutes, New York, USA. Contact email: kressel@ndri.org
Keith Morgen, PhD, is Assistant Professor of Psychology, Department of Behavioral and Historical Studies, Centenary College, Hackettstown, NJ, USA.
Gregory Bunt, MD and Britta Muehlbach, MSc are at Daytop Village, Inc., New York, NY, USA.

Clive Burkett

ABSTRACT: This study explores the relationship between self-efficacy, performance-based coping skills and positive outcome expectancies for drugs and alcohol. It employed four questionnaires in a prospective, cross-sectional survey design utilising a between-groups sample. Participants comprised ninety-six individuals assigned to three groups ranging from active drug/alcohol users, to drug/alcohol-free clients in primary treatment, and aftercare. It predicted a significant difference between all groups on the investigated variables suggesting that, as self-efficacy and coping skills increase, positive outcome expectancies for drugs and alcohol would decrease. Results of ANOVA and correlation coefficient analysis validated all five hypotheses yielding significant difference at p<0.01 and p<0.05 levels (2-tailed). Analysis and discussion focus on a cognitive-behavioural perspective to glean greater understanding of the dynamic relationship between these variables, and attempt to improve applied practice in this area.

Full recovery from drug and alcohol addiction requires cessation of substance use and the adoption of a global lifestyle change (De Leon, 1997) if abstinence is to be maintained. This process of change requires incremental stages of learning and motivational shifts (Prochaska & DiClemente, 1984; De Leon, 1997) that promote a sense of self-control with the availability of workable coping techniques that reinforce feelings of control even further (Marlatt & George, 1988). Marlatt and Gordon (1985) put forward a model of relapse prevention, grounded in social learning theory (SLT) (Bandura, 1969, 1977) that explained how deficits in coping and self-efficacy, plus specific outcome expectancies interact and can lead to relapse.

Many theories abound as to the aetiology and precipitating factors that lead individuals on a path to substance misuse and eventual dependency. Accordingto psychoanalytic theory, Rado (1933) considered addiction as a narcissistic disorder where the use of psychoactive substances allowed the individual to return to this state of ego disruption where the full impact of reality is muted. Most psychoanalytic theories suggest factors in early childhood development as having aetiological foundations for later dependence to drugs. Savitt (1963) believed the lack of maternal love and affection plus a passive, ineffectual father result in the development of a personality that is imbued with a state of intolerable tension. The use of psychoactive substances allows for the release of this tension, which in turn perpetuates addiction as a coping strategy. Others suggest significant disappointment and unfulfilled object relations for the child with the same-sex parents throughout the Oedipus complex as influential factors in the onset of addiction (Federn, 1972).

Drive theory suggests a purely biological influence relative to the onset and development of physical and psychological addiction. In principle, any human being who is continually exposed to the euphoric effects of a psychoactive drug will develop dependence sooner or later. The stronger and more pleasurable the effect, the more rapid actual dependency will develop (Bejerot, 1972). Genes that predispose young adults towards drug misuse include those that attract them continually to excitement and thrill seeking and that manifest intolerance to frustration and vulnerability to depression. Genetic vulnerability relative to these personality traits, coupled with strong environmental influences push individuals in the direction of abuse and dependency (Berger, 2005).

SLT (Bandura, 1969, 1977) encompasses the main psychological models of learning and human social development. The basic premise is that addiction is a social phenomenon that is learned, reinforced, and ultimately conditioned within the individual through contact with his/her environment (Butcher, Mineka, Hooley & Carson, 2004). SLT also accepts the role of biological and genetic influences on addiction in conjunction with psychosocial factors (Abrams & Wilson, 1986) rendering a bio-psychosocial model (Monti, Kadden, Rohsenow, Cooney & Abrams, 2002). The concept of modelling within the SLT paradigm accounts for the basic principles of vicarious learning. Attempts to emulate and adapt certain attributes of others' behaviour become an implicit learning process (Crain, 2005). Children who are exposed to dysfunctional families where parents are emotionally absent, abusive or who have substance dependency themselves are presented with poor role models where the acquisition of the ability to monitor and judge their own conduct through internalised standards is severely impaired (Heather & Robertson, 1997).

Most theorists and practitioners are in agreement that the use of Cognitive Behavioural Therapy (CBT) as an interventional technique is applicable. CBT interventions for alcohol dependence have repeatedly demonstrated that they are effective strategies in reducing drinking in randomised trials (Morgenstern & Longabaugh, 2000). A significant number of reviews on treatment regimes have also indicated a high evidence of effectiveness in treating alcohol dependence through the medium of CBT (Holder, Longabaugh, Miller & Rubonis, 1991; Miller et al., 1995; Finney & Monahan, 1996; Roth & Fonagy, 1996). Its use is considered to be effective in that it is a collaborative approach with short-term intervention, that is flexible, structured, goal-oriented, specifically focused, and provides strong empirical support (Carroll, 1998).

The emergence of Cognitive Behavioural Coping Skills (CBCS), which has adapted the basic elements of CBT and Motivational Interviewing, also employs the essential components of functional analysis, training skills around coping with cravings, problem solving, cognitive processing of drug use, refusal skills, and client homework assignments (Dunn, Morrison & Bentall, 2002; Carroll, Nich & Ball, 2005; Gonzalez, Schmitz & DeLaune, 2006) that allow for practice and rehearsal of these skills.

Self-efficacy

In the context of recovery from addiction, an individual's belief that he can successfully cope and overcome temptation to drink or use drugs is referred to as self-efficacy (Monti et al., 2002). Introducing performance-based techniques into the recovering person's repertoire of behaviours through modelling, intrapersonal and interpersonal heightened awareness (Monti et al., 2002), cognitive processing, and mental rehearsal of coping responses can raise the belief and feelings of self-efficacy for the recovering person, particularly in high risk situations (Taylor, 2006). As self-efficacy develops through specific treatment intervention, the likelihood that performance-based coping skills will improve is further enhanced (Velicer, DiClemente, Rossi & Prochaska, 1990). The development of coping skills where individuals have strong belief in their ability to resist temptation to lapse/relapse has been negatively correlated with relapse (Abrams et al., 1991; Wunschel, Rohsenow, Norcross & Monti, 1993; Miller, Westerberg, Harris & Tonigan, 1996; Moser & Annis, 1996). Integral to this is that relapse prevention strategies are fully integrated into recovery programmes from the very outset of treatment (Taylor, 2006).

Self-efficacy is a dynamic construct that is constantly reappraised in the light of new information and experiences (DiClemente, Cabonari, Montgomery & Hughes, 1992). These self-efficacy evaluations are considered to be influential to motivation, information processing, effort, and effective action (Bandura, 1977, 1986) and form the basis of valuable information regarding behavioural change (Marlatt & Gordon, 1980; Rollnick & Heather, 1982; DiClemente, 1986). The belief of a recovering person in their ability to overcome temptation in high-risk situations is a crucial component in protecting against relapse, thus placing the development of these skills high on the list of priorities for recovering individuals (DiClemente et al., 1992). Self-efficacy theory assumes a central belief that the strength of individual self-efficacy will determine the nature of coping skills and their effectiveness as a response, across time and situations (Annis, 1986).

Coping skills

People who develop the skills to be able to perceive and overcome events within their environment as controllable and regard their coping strategies as being effective are more likely to experience less stress and its associated consequences (Benight et al., 1997). Coping is the process of managing demands both internal and external, which are evaluated as being taxing or exceeding the personal resources of the individual (Lazarus & Folkman, 1984). Treatment interventions to raise the level of awareness of exteroceptive and interoceptive cues to drink or use drugs allow the recovering person greater autonomy to make informed decisions and indicate better outcomes relative to sustained abstinence (Monti et al., 2002).

The effectiveness of these approaches in conjunction with social skills training (Freedberg & Johnson, 1978) and cognitive restructuring (Meichenbaum & Cameron, 1974; Oei & Jackson, 1982) has also led to the development of interpersonal and intrapersonal treatment adjuncts, resulting in better out-comes (Monti et al., 2002). Cognitive restructuring (Meichenbaum & Cameron, 1974), or reframing (Miller & Rollnick, 2002), involves the modification of internal monologues that are unhelpful to the individual. For example, it assists the client to overcome the abstinence violation effect (Marlatt & Gordon, 1985) by learning to perceive a lapse as a mistake rather than a complete failure, thus altering cognitive processing that sets a perception of catastrophe or ‘all or nothing' thinking (cf. Ellis, 1970; Bandura, 1978).

A study with twins by Kendler, Kessler, Heath, Neale and Eaves (1991) identified three general coping styles: direct problem solving, turning to others for support and advice, and using denial or avoidance of the problem; all constitute the main coping styles that individuals resort to in a crisis (Taylor, 2006). Other research in this area has indicated that avoidance coping techniques not only lead to higher levels of distress in the long run, but that this method of coping has serious psychological ramifications for adverse responses to stressful life circumstances (Holahan & Moos, 1986).

Problem-focused and emotion-focused coping styles generally work in conjunction with each other. Problem-focused coping involves tackling a problem or situation head on by attempting to find logical and pragmatic methods and solutions (Folkman & Lazarus, 1980). Emotion-focused coping involves efforts to self-regulate one's emotional response to a crisis. Within emotion-focused coping, two subscales of coping are identified. In the former, rumination, which involves recurrent negative thoughts that assume a more pessimistic perspective, forms the self-appraisal of low self-efficacy and coping skills to overcome the problem. Conversely, the latter, referred to as emotional-approach coping, allows for the disclosure and clarification of a problem with a view to working through the emotional experience with one's social support structures (Stanton, Danoff-Burg, Cameron & Ellis, 1994).

Outcome expectancies

Considerations pertaining to an individual's outcome expectancies (Bandura, 1977) will also determine the likelihood of lapse/relapse. Positive outcome expectancies consider the consequences and rewards associated with a given behaviour, in this case substance misuse. Individuals who have experienced and perceive the use of alcohol and drugs as pleasurable and reinforcing as a relief from social stressors, upsetting life events, or as a form of relaxation have in a sense developed their own ineffective coping strategies. Research indicates that individuals who drink more tend to have higher expectations relative to the positive effects of alcohol (Carey, 1995). Negative outcome expectancies on the other hand allow for value judgements that attempt to support and maintain the recovery process by continual review of the negative consequences associated with substance dependency. Positive outcome expectancies in relation to alcohol or drugs, which are proximal, tend to overshadow the negative outcome expectancies, which are distal, with the former playing a salient role in relapse where high-risk situations are experienced (Carey, 1995). Individuals who are engaged in a process of change can be further assisted in developing the perspective that changing their addictive habits and behaviours involves skill acquisition rather than a test of willpower (Larimer, Palmer & Marlatt, 1999). Treatment interventions that emphasise a continued series of setting small, achievable goals build upon the person's belief and ability to achieve certain objectives while nurturing feelings of empowerment and self-mastery (Bandura, 1977; Larimer et al., 1999).

The present study

Using the central components that empirical research in this area has studied and recognised as essential ingredients for sustained recovery from substance dependency, this study attempted to demonstrate the progressive and incre-mental nature of this important change process. It utilised a client base ranging from active drug/alcohol users, to drug-free clients in primary treatment and in an integration and aftercare stage of a therapeutic community (TC). It focused on, measured, and compared the coping skills' acquisition of clients at various stages of change, assuming that abstinence and the development of enhanced self-efficacy and performance-based coping skills are essential to the recovery process.

In order to fully consider the antecedent, behavioural response and consequential aspects of sabotage to recovery, this study examined the relationship between self-efficacy and performance-based coping skills with addicted and recovering individuals who were classed at different motivational stages of the Trans-Theoretical Model (TTM) of change (Prochaska & DiClemente, 1984). The relationship and interaction between the non-existence and development of self-efficacy and coping skills for these individuals was examined in conjunction with the residual existence of positive outcome expectancies for drugs and alcohol. The variables under investigation were self-efficacy, both generalised and specific to refusal confidence relative to temptation to use drugs or alcohol; performance-based coping skills that focus on avoidance-focused coping, problem-focused coping and emotion-focused coping; and positive outcome expectancies concerning drugs and alcohol that perceive these substances as pleasurable, rewarding and stress-reducing.

Programme involvement should influence self-efficacy and coping skills, while self-efficacy will ultimately influence and reinforce performance-based coping skills. Positive outcome expectancies will be determined by positive, negative and punishment contingencies relative to past conditioning and reinforcement factors. Likewise, the development of self-efficacy and coping skills will influence whether recovering individuals perceive drinking and drug use as having positive outcomes. For the recovering addict, the notion that positive outcome expectancies can be immediate, pleasurable and stress-reducing, in contrast to the long-term pay-offs of negative outcome expectancies on physical health, relationships, and mental wellbeing, plays clearly to the addicted mindset of instant gratification.

(i) It was predicted that substance abusers who are still actively using will show lower levels of self-efficacy and performance-based coping skills than those in treatment.
(ii) It is proposed that there will be a positive correlation between the development of performance-based coping skills and self-efficacy in individuals actively engaged in treatment.
(iii) There will be a negative correlation between the development of self-efficacy, performance-based coping skills in relation to positive outcome expectancies for drugs, and alcohol for individuals in treatment. Where high levels of self-efficacy and coping skills are identified, low levels of positive outcome expectancy will be predicted.
(iv) There will be a significant difference reported between positive outcome expectancies for drugs and alcohol where problem-focused and emotion-focused coping skills are employed as against avoidance-coping techniques. It is expected that active drug users will show a greater tendency than those in treatment to use coping techniques that help them to avoid their problems.
(v) There will be a significant difference between people at the early stage of treatment regarding their self-efficacy and coping skills than those at the later stages.

Methodology

Materials

The materials used in this study were four questionnaires designed to elicit generalised self-efficacy, alcohol/drug refusal self-efficacy, performance-based coping skills, and positive outcome expectancy for drugs and alcohol.

Internal consistency

An alpha reliability test was undertaken to assess alpha coefficient values (Cronbach, 1951) for all items within the four questionnaires. Apart from the Brief Cope Scale (BCS), all constructs of the alpha coefficient were found to be satisfactory with high internal consistency found above the 0.7 level. Table 1 outlines these findings.

Table 1: Alpha coefficient

The Generalised Self-Efficacy Scale (Schwarzer & Jerusalem, 1993) is a 10-item questionnaire designed to measure perceived self-efficacy with a 4-point Likert scale ranging from Not at all true=1; Barely true=2; Moderately true=3; Exactly true=4. Scores range from 10 to 40 with low scores representing low self-efficacy and high scores representing high levels.

The Alcohol/Drug Refusal Self-Efficacy Scale is a modified version of the Situational Confidence Questionnaire (Annis, 1982), which is a 100-item scale to test levels of confidence in high-risk situations. The ADRSE is a 12-item questionnaire that identifies a series of recognised antecedents of an extero-ceptive and interoceptive nature (Monti et al., 2002) that generally precipitate relapse. This questionnaire employs a 6-point scale ranging from Not confident at all=0; Little confidence=2; Slightly confident=4; Reasonably confident=6; Confident=8; Extremely confident=10. Scores on this scale range from 0 to 120, with low scores representing little or no confidence and high scores representing good confidence to resist temptation.

The Brief Cope Scale (Carver, 1997) is a 21-item questionnaire that identifies specific coping skills and strategies that can be employed in a wide range of difficult situations. It combines a subscale of avoidance coping, and problem-focused and emotion-focused strategies (Taylor, 2006). Avoidance-coping items within the questionnaire required a re-coding as this form of coping is acknowledged as an ineffective coping strategy. Items are scored by a 5-point Likert scale ranging from Never=0; Seldom=1; Sometimes=2; Often=3; Always=4. Scores on this scale range from 0 to 84, with low scores representing poor coping skills and high scores representing good coping skills.

The Alcohol/Drug Expectancy Questionnaire III (Brown, Christiansen & Goldman, 1987) is a 120-item questionnaire designed to assess positive effects attributed to alcohol consumption. For this particular study the need for some modification of the ADE was deemed necessary. Firstly, the use of the full instrument would have implications for client responses based upon the fact that three other questionnaires were also employed. It was considered that the accumulated battery of questionnaires would create some frustration for respondents if the full 120-item questionnaire was used. With this in mind, only sections 1, 3, 4, and 5 from the subscale were used, thus rendering a 48-item questionnaire. Additionally, the inclusion of the words Drugs and or Alcohol into each of the 48 items broadened the scope of the questionnaire in order to make it applicable to respondents. This questionnaire employed an Agree or Disagree response for each item, with an Agree response weighted as one point and a Disagree response weighted as two points. Where all 48 items were responded to, a range of 48 to 96 points was calculated, with high scores representing high positive outcome expectancy and low scores representing low outcome expectancies.

Participants

Participants comprised 96 individuals (24 female and 72 male) involved at various stages of a continuum of care at Coolmine Therapeutic Community in Dublin. This continuum commences with outreach and drop-in services engaging with active drug users to assess individual readiness and suitability to engage. Clients involved at this stage fall into the pre-contemplation, contemplation, and preparation stages of the TTM (Prochaska & DiClemente, 1984). This group was referred to as Group A or Active Drug/Alcohol Users. The majority of this group were categorised as having poly-drug use difficulties as assessed by the Maudsley Addiction Profile (MAP) (Marsden et al., 1998). Participants numbered (N=28) twenty-three male and five female.

Once clients have achieved drug/alcohol-free status they are ready to move to primary treatment. Primary treatment in this case includes full-time residential care or community-based day programme options, both having a minimum stay of six months. Clients involved at this stage are classed at the action stage of the TTM and fall into Group B or drug/alcohol-free clients in primary treatment. Participants numbered (N=42) 30 male and 12 female.

Clients who successfully complete primary treatment move to the next phase of the Coolmine continuum by entering an Integration and Aftercare (IAC) service, which is divided into two stages, both involving a six-month stay. These clients have re-entered mainstream society and are in full-time employment or further education. Drug/alcohol-free status is a requirement of these phases and these individuals fall into the maintenance stage of TTM. This group is referred to as Group C or Aftercare clients and participants numbered (N=26) nineteen male and seven female.

All clients in Groups B and C have experienced a standardised approach of therapeutic intervention and involvement that included ongoing assessment, supervised case management with regular review, group therapy, psycho-educational workshops, community living based upon SLT (Bandura, 1969, 1977), relapse prevention, and the concerted development of self-efficacy and performance-based coping skills.

A cross-sectional survey design was utilised with the use of four separate questionnaires. This design was quantitative and prospective in nature and focused on three separate groups using an independent or between-groups sample. A correlation coefficient method using Pearson's r and a one-way analysis of variance (ANOVA) was also undertaken in order to establish any significant relationships between the investigated variables of self-efficacy, performance-based coping skills, and positive outcome expectancies. The mean age of clients participating in this study was 28 years.

Procedure

Participants in Groups A, B, and C were all interviewed in small groups. In all interview scenarios, participants were instructed to carefully read all questionnaire details and complete all items within each questionnaire. Participants were also briefed in advance on the confidentiality aspects and were instructed not to place their names on the questionnaires. Female participants were asked to place the letter ‘F' on the top right-hand page of each questionnaire to distinguish gender. On completion of all questionnaires respondents were debriefed, at which point they were given an opportunity to withdraw their specific questionnaires.

Results

Descriptive statistics outlined in Table 2 present the mean scores, standard deviation, and score range on each questionnaire for the three individual groups.

In the Alcohol/Drug Expectancy (ADE) questionnaire, a summed score range from 48 to 96 indicates that high scores on this questionnaire demonstrate higher positive outcome expectations for drugs and alcohol. Conversely, lower scores indicate lower expectations as a desired outcome for psychoactive substances. Group A shows a mean score of 85.14 (SD=5.66) in comparison with Group B at 77.40 (SD=8.83) and Group C at 64.16 (SD=9.71). A total mean score of 76.20 (SD=11.41) was reported for all three groups. Results indicate that Group A shows higher scores on positive outcome expectancies in comparison to Groups B and C, and Group B also shows higher scores in comparison to Group C.

In the Generalised Self-Efficacy (GSE) questionnaire notable levels of self-efficacy were reported between the three groups with high scores indicating high levels of self-efficacy and low scores indicating lower levels. With a score range of 10 to 40, the mean score for Group A was reported at 18.71 (SD=5.30); for Group B the mean score was 28.33 (SD=3.52); and for Group C the mean score was 30.23 (SD=3.70). A total mean score of 26.04 (SD=6.31) was reported for all three groups on this questionnaire. Results here indicate an ascending level of self-efficacy for Groups B and C in comparison to Group A.

Table 2: Descriptive statistics for group scores on the four questionnaires

______________________________________________________________________________________

The ADRSE questionnaire presents a summed score range of 0 to 120 with high scores indicating greater self-confidence to overcome temptation to use drugs or alcohol. Conversely, lower scores indicate poor levels of self-efficacy and coping skills. The reported mean score for Group A was 36.71 (SD=16.22); for Group B the mean was 79.23 (SD=19.88); and for Group C it was 88.69 (SD=19.46). A total mean score of 69.39 (SD=28.36) was reported for all three groups on this questionnaire. These results indicate a marked difference in the three groups' ability and confidence to deal with and overcome temptation to use when presented with high-risk situations.

A one-way analysis of variance (ANOVA) was carried out on the total scores for each questionnaire. This was performed in order to establish any variance of a significant difference between responses on the four questionnaires pertaining to the three groups.

On the ADE questionnaire a statistical significance of positive outcome expectancy of F(2,92)=43.07, p<.001 was reported, indicating a significant difference between groups on this questionnaire. Of the 96 respondents involved in this study, only 95 fully completed the ADE questionnaire.

The GSE questionnaire also produced a significant difference between groups F(2,93)=62.90, p<.001 indicating ascending levels of self-efficacy as clients progress.

In relation to the BCS a statistical difference of F(2,93)=7.81, p<.001 was reported between Group A and Groups B and C, with lesser variation between Groups B and C. Variation on scores between Groups B and C will be discussed in greater detail in the post hoc report to follow.

Finally, statistical significance was also reported between all three groups on the ADRSE questionnaire with F(2,93)=61.90, p<.001. These results also indicate a significant difference between groups regarding the levels of confidence and the performance-based coping skills required to overcome temptation. Table 3 sets out the source of variation of the above results.

Table 3: One-way ANOVA results

_____________________________________________________________________________________

Post hoc analysis using Bonferroni correction was also performed to examine the multiple comparisons between groups on all four questionnaires. On the ADE a mean difference of 7.74* between active drug users and primary treatment clients was reported. When clients in primary treatment are compared with aftercare clients, a mean difference of 13.24* was reported. In relation to Group A and Group C, a mean difference of 20.98* was reported. These figures indicate a gradual and incremental improvement in relation to all three groups in the way that drugs and alcohol are perceived as a result of programme intervention.

[* Indicates mean difference is significant at the .05 level.]

Examining the total BCS mean scores for all three groups, a mean difference of -5.10* was found between Group A and Group B. The mean difference between Group B and Group C was reported at -2.08 showing no significant difference in relation to these two groups. Group A and Group C showed a mean difference of -7.19* indicating a significant difference between these two groups in coping skills.

Post hoc analysis on the ADRSE yielded a significant mean difference of -42.52* between Group A and Group B. The mean difference reported between Group B and Group C was -9.45* indicating a significant statistical difference between these groups. Finally, results of the mean difference between Group A and Group C indicated a significant difference of -51.97*. In all cases the mean difference is significant at the .05 level. Table 4 indicates the mean difference between groups on all four variables.

Table 4: Mean difference between groups on all four variables

_____________________________________________________________________________________

Correlations

A Pearson's coefficient analysis was conducted on all four total scores of each questionnaire with the addition of the three subscales associated with the BCS questionnaire, which included avoidance-focused coping, problem-focused coping, and emotion-focused coping factors. The strongest significant association was found between problem-focused coping relative to the overall coping strategies with r=.738 (p=.000) indicating the use of problem-focused coping as an overall efficacious coping strategy in comparison to avoidance and emotionfocused coping. Refusal self-efficacy was also reported to be highly significant with a correlation of r=.698 (p=.000) indicating that those scoring higher on this measure would have higher levels of refusal self-efficacy. The GSE also showed highly significant correlation with r=.694 (p=.000) again indicating strong association between coping skills and self-efficacious beliefs.

In relation to the ADE, a negative correlation of r=-.685 (p=.000) was reported indicating an overall negative association between generalised, refusal self-efficacy and performance-based coping skills and its associated subscales. The ADE showed other strongly significant negative correlations in relation to GSE with r=-.503 (p=.000); the ADE when contrasted against the BCS showed a weak to moderate correlation of -.244 (p=.000); against the ADRSE, the ADE reported a moderate negative correlation of r=-.480 (p=.017). All this indicated that, where self-efficacy and performance-based coping skills are high, positive outcome expectancies towards drugs and alcohol are low.

When the subscales within the BCS were analysed, the variable of avoidance-focused coping showed some significant negative correlations in relation to GSE with r=-.385 (p=.000); in relation to ADRSE a moderate to strong correlation with r=-.451 (p=.000) was reported; avoidance-focused coping also showed significant negative correlations in relation to problem-focused coping and emotion-focused coping with r=-.413 (p=.000) and r=-.442 (p=.000), respectively. Additionally, a strong positive correlation was reported between problem-focused coping and emotion-focused coping of r=.664 (p=.000) supporting previous research by Folkman and Lazarus (1980). Table 5 shows Pearson's correlation coefficient between the measured variables.

Table 5: Correlation between measured variables

** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
_____________________________________________________________________________________

This study set out to investigate the relationships between self-efficacy, performance-based coping skills, and positive outcome expectancies towards drugs and alcohol. It proposed that the incremental and ongoing development of self-efficacy and performance-based coping skills were essential ingredients in relation to sustained recovery from drug/alcohol dependency. The study postulated the assumption that, because these substances are perceived as pleasurable, rewarding and stress reducing, individuals who are in early recovery from substance dependency still crave for and perceive these substances as rewarding.

Five main hypotheses were set out. Firstly, that there would be a significant difference in levels of self-efficacy and coping skills between active drug users and individuals engaged in treatment. It claimed that as refusal self-efficacy developed and grew so too would performance-based coping skills. It also predicted that where high levels of self-efficacy and coping skills were identified, low levels of positive outcome expectancies for drugs and alcohol would also be found. It identified three sub-factors within coping strategies that would have positive and negative implications for desired outcomes. It predicted that avoidance-focused coping would be more synonymous with proximal and ineffective outcomes and employed more extensively by active drug users. Finally, it predicted a significant difference between levels of self-efficacy and coping skills relative to individuals in early treatment and those in aftercare.

Results from this study validated all hypotheses. Analysis yielded a significant difference between all three groups on the acquisition and development of self-efficacy and performance-based coping skills. The most obvious and significant difference relates directly to the variation between Group A (active drug users) and Group B (primary treatment) on levels of confidence and ability to cope and deal effectively with high-risk scenarios. In comparing results of self-efficacy and coping skills for active drug users and aftercare clients, a significant difference was reported also. The results exceeded the mean difference between active drug users and clients in primary treatment, not only suggesting a greater overall gap in learning and time spent in a recovery process, but also clearly indicating the transitory and evolving dynamics of Group B.

A positive correlation between the development of self-efficacy and performance-based coping skills was reported. A strong positive correlation was found between the GSE and the ADRSE. Further comparisons of the GSE against the sub-factors of the BCS yielded much stronger correlations in relation to problem-focused coping and emotion-focused coping. Analysis found a significant negative correlation between GSE and avoidance-focused coping, suggesting that as self-efficacy increases ineffective coping strategies diminish. When compared with the BCS, the ADRSE reported a weak correlation, significant at the 0.05 level (2-tailed). Although this association may be deemed as weak, much stronger correlations on problem-focused and emotion-focused coping were identified. A moderate to strong negative correlation between the ADRSE and avoidance-focused coping was reported indicating that as refusal self-efficacy improves more positive and viable coping strategies manifest.

From perusal of the results section of Table 4, one can see the validation of the predicted negative correlation between positive outcome expectancies and the development of self-efficacy and coping skills. When the ADE is contrasted against GSE and the ADRSE, negative correlations are also reported. These findings clearly suggest that as self-efficacy grows and develops a significant decrease in positive outcome expectancies for drugs and alcohol is perceived. These findings are supported by the assertion that self-efficacy evaluations are hypothesised to mediate all behavioural change by having a direct influence on motivation, information processing, personal effort, and effective action (Bandura, 1977, 1986).

A study by Hasking and Oei (2004) concluded that coping, self-efficacy and outcome expectancies are crucial variables in governing drinking behaviour and that the development of self-efficacy is integral to self-regulatory mechanism by which humans exhibit control over their behaviour. An earlier study by Annis and Davis (1988) found that self-efficacy increased significantly from intake to a six-month follow-up treatment trial among problem drinkers. Research on alcohol and drug expectancies (Goldman, Brown & Christiansen, 1991; McMahon, Jones & O'Donnell, 1994) suggests that higher positive outcome expectancies are associated with higher levels of consumption and, conversely, higher negative outcome expectancies are associated with lower levels or no consumption at all.

As predicted, a positive relationship between positive outcome expectancies and avoidance-focused coping in contrast with problem-focused and emotion-focused coping was reported. This indicates that individuals who tend to avoid their problems rather than deal with them directly are more likely to turn to drugs or continue using as a solution to their problems. Individuals who have learnt other ways of coping, such as addressing their problems pragmatically and regulating their emotional responses to crises are less likely to lapse. The development of pragmatic, solution-focused coping to overcome encountered stressors is a vital deterrent in preventing the escalation of positive outcome expectancies for substance use. Furthermore, it suggests that the combination of problem-focused coping and emotion-approach coping further reduces stress and upset, thus enhancing sustained recovery. The relationship between posi-tive outcome expectancies and avoidance-focused coping in this study suggests that this style of coping feeds into and increases positive outcome expectancies towards substance misuse.

Finally, a significant difference between primary treatment clients and aftercare clients in relation to self-efficacy, coping skills, and positive outcome expectancies was also found. However, a significant difference was established in relation to positive outcome expectancies for drugs and alcohol in relation to refusal self-efficacy, suggesting that as one's refusal confidence grows when presented with high-risk situations positive outcome expectancies for alcohol and drugs diminishes.

Conclusion

The term ‘ambivalence', often referred to as the dilemma of change (Miller & Rollnick, 2002), highlights the inherent contradictions often associated with human behaviour and change. Individuals in the throes of addiction frequently vacillate between over-indulgent behaviour and repeated attempts to give up (Miller & Rollnick, 2002). Temporal considerations must also be factored into this equation, where proximal and distal influences have direct effect on behavioural intention and response. Reconciliation of this ambivalence requires individuals to maintain an unequivocal position that results in overall improvement and quality of life, or alternatively return to the addicted lifestyle.

The removal of substance misuse from an individual's life creates a vacuum that must be replaced by other activities; likewise the removal of cognitive processes and associated conditioned responses that reinforce positive outcome expectancies towards substance misuse must also be balanced and replaced by the ongoing development of negative outcome expectancy, self-efficacy and performance-based coping skills that sustain recovery. According to The Addicted-Self Model (Fiorentine & Hillhouse, 2003), the onset of stressful stimuli, situational risks and maladaptive coping responses may not necessarily lead to relapse. Fiorentine and Hillhouse claim that in some cases relapse may be precipitated more so by positive outcome expectancies and that a negative expectancy balance in relation to mood altering substances may prevent relapse even when high risk or situational stress is experienced.

The overarching hypothesis clearly indicates that ongoing development of self-efficacy and performance-based coping skills are crucial components to sustained recovery from active substance dependency. Further research in this area may well be directed towards the ambivalent dynamics of negative and positive outcome expectancies for drugs and alcohol in conjunction with self-efficacy and coping skills. This may allow for some form of psychometric assessment that could evaluate a plotted balance between positive and negative outcome expectancies, using the TTM to plot temporal considerations regarding the stages of change.

References

Abrams, D.B. & Wilson, G.T. (1986). Habit disorders: Alcohol and tobacco dependence. In A.J. Frances & R.E. Hales (Eds.), American Psychiatric Association annual review (Vol. 5, pp. 606-626). Washington, DC: American Psychiatric Press.

Abrams, D.B., Binkoff, J.A., Zwick, W.R., Liepman, M.R., Nirenberg, T.D., Munroe, S.M. & Monti, P.M. (1991). Alcohol abusers' and social drinkers' responses to alcohol-relevant and general situations. Journal of Studies on Alcohol, 52, 409-414.

Annis, H.M. (1982). The Situational Confidence Questionnaire. Addiction Research Foundation, Toronto, Ontario M5S 2S1.

Annis, H.M. (1986). A relapse prevention model for treatment of alcoholics. In W.R. Miller & N. Heather (Eds.), Treating Addictive Behaviours: Processes of Change (pp. 407-433). New York: Plenum Press.

Annis, H.M. & Davis, C.S. (1988). Self-efficacy and the prevention of alcoholic relapse: Initial finding from a treatment trial. In T.B. Baker & D. Cannon (Eds.), Assessment and Treatment of Addictive Disorders (pp. 88-112). New York: Praeger Pubs.

Bandura, A. (1969). Principles of Behaviour Modification. New York: Holt, Rinehart and Winston Inc.

Bandura, A. (1977). Self-Efficacy: Toward a Unifying Theory of Behavioural Change. Psychological Review, 84, 191-215.

Bandura, A. (1978). Reflections on self-efficacy. Advances in Behavioural Research and Therapy, 1, 237-269.

Bandura, A. (1986). Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs: NJ Prentice-Hall Inc.

Bejerot, N. (1972). Addiction: An artificially induced drive. Springfield, Ill: Charles C. Thomas.

Benight, C.C., Antoni, M.H., Kilbourn, K., Ironson, G., Kumar, M.A., Fletcher, M.A. et al. (1997). Coping self-efficacy buffers psychological and physio-logical disturbances in HIV-infected men following a natural disaster. Health Psychology, 16, 248-255.

Berger, K.S. (2005). The Developing Person. Through the Life Span. (6th ed.). NY 10010: Worth Publishers.

Brown, S.A., Christiansen, B.A. & Goldman, M.S. (1987). The Alcohol Expectancy Questionnaire: An instrument for the assessment of adolescent and adult alcohol expectancies. Journal of Studies on Alcohol, 48(5): 483-491.

Butcher, J.N., Mineka, S., Hooley, J.M. & Carson, R.C. (2004). Abnormal Psychology (12th ed.). Boston/New York/London: Pearson Publication.

Carey, K.B. (1995). Alcohol-related expectancies predict quantity and frequency of heavy drinking among college students. Psychology of Addictive Behaviours, 9(4): 236-241.

Carroll, K.M., Nich, C. & Ball, S.A. (2005). Practice Makes Progress? Homework Assignments and Outcome in Treatment of Cocaine Dependence. Journal of Consulting and Clinical Psychology, 73(4), 749-755.

Carver, C.S. (1997). You want to measure coping but your protocol is too long: Consider the Brief COPE. International Journal of Behavioural Medicine, 4, 92-100.

Cronbach, L.J. (1951). Cronbach Alpha Coefficient Reliability Test. Psychometrika, 16(3), 297-334.

Crain, W. (2005). Theories of Development. Concepts and Applications (5th ed.). New Jersey 07458: Pearson Prentice Hall.

De Leon, G. (1997). Community As Method. Therapeutic Communities for Special Populations and Special Settings. Westport, CT/London: Praeger Publishers.

DiClemente, C.C. (1986). Self-efficacy and the addictive behaviours. Journal of Social & Clinical Psychology, 4, 302-315.

DiClemente, C.C., Cabonari, J.P., Montgomery, R.P.G. & Hughes, S.O. (1992). The Alcohol Abstinence Self-efficacy Scale. Journal of Studies on Alcohol, 55, 141-148.

Dunn, H., Morrison, A.P. & Bentall, R.P. (2002). Patient's Experiences of Homework Tasks in Cognitive-Behavioural Therapy for Psychosis: A Qualitative Analysis. Clinical Psychology and Psychotherapy, 9, 361-369.

Ellis, A. (1970). The essence of rational psychotherapy: A comprehensive approach to treatment. New York: Institute of Rational Living.

Federn, E. (1972). A psycho-social view of drug abuse in adolescence. Child Psychiatry and Human Development, 3, 10-20.

Finney, J.W. & Monahan, S.C. (1996). The cost-effectiveness of treatment for alcoholism: a second approximation. Journal of Studies on Alcohol, 5, 229-243.

Fiorentine, R. & Hillhouse, M.P. (2003). When Low Self-efficacy is Efficacious: Toward An Addicted-self Model of Cessation of Alcohol and Drug-dependent Behaviour. The American Journal of Addictions, 12, 346-364.

Folkman, S. & Lazarus, R.S. (1980). An analysis of coping in a middle-aged community sample. Journal of Health and Social Behaviour, 21, 219-239.

Freedberg, E.J. & Johnson, W.E. (1978). The effects of assertion within the context of a multi-modal alcoholism treatment programme for employed alcoholics. Toronto: Addiction Research Foundation.

Goldman, M.S., Brown, S.A., Christiansen, B.A. et al. (1991). Alcoholism aetiology and memory: Broadening the scope of alcohol expectancy research. Psychological Bulletin, 110, 137-146.

Gonzalez, V.M., Schmitz, J.M. & DeLaune, K.A. (2006). The Role of Homework in Cognitive-Behavioural Therapy for Cocaine Dependence. Journal of Consulting and Clinical Psychology, 74(3), 633-637.

Heather, N. & Robertson, I. (1997). Problem Drinking (3rd ed.). New York: Oxford University Press and Oxford University Press Inc.

Holahan, C.J. & Moos, R.H. (1986). Personality, coping, and family resources in stress resistance: A longitudinal analysis. Journal of Personality and Social Psychology, 51, 389-395.

Holder, H., Longabaugh, R., Miller, W.R. & Rubonis, A.V. (1991). The cost-effectiveness of psychotherapy for alcoholism: a first approximation. Journal of Studies on Alcohol, 52, 517-540.

Kendler, K.S., Kessler, R.C., Heath, A.C., Neale, M.C. & Eaves, L.J. (1991). Coping: A genetic epidemiological investigation. Psychological Medicine, 21, 337-346.

Larimer, M.E., Palmer, R.S. & Marlatt, G.A. (1999). Relapse Prevention: An Overview of Marlatt's Cognitive-Behavioural Model. Alcohol Research & Health, 23(2), 151-160.

Lazarus, R.S. & Folkman, S. (1984). Stress, appraisal and coping. New York: Springer.

Marlatt, G.A. & George, W.H. (1988). Relapse prevention and the maintenance of optimal health. In S. Shumaker, E. Schon, & J.K. Ockene (Eds.), The adoption and maintenance of behaviour for optimal health. New York: Springer.

Marlatt, G.A. & Gordon, J.R. (1980). Determinants of relapse: Implications for the maintenance of behaviour change. In P.O. Davidson & S.M. Davidson (Eds.), Behavioural Medicine; Changing Health Lifestyles (pp. 410-452). New York: Brunner-Mazel, Inc.

Marlatt, G.A. & Gordon, J.R. (1985). Relapse Prevention: Maintenance Strategies in The Treatment of Addictive Behaviours. New York: Guildford.

Marsden, J., Gossop, G., Stewart, D., Best, D., Farrell, M., Lehmann, P., Edwards, C. & Strang, J. (1998). The Maudsley Addiction Profile (MAP): A brief instrument for assessing treatment outcome. Addiction, 93(12), 1857-1867.

McMahon, J., Jones, B.T. & O'Donnell, P. (1994). Comparing positive and negative alcohol expectancies in male and female social drinkers. Addiction Research, 1, 349-365.

Meichenbaum, D.H. & Cameron, R. (1974). The clinical potential and pitfalls of modifying what clients say to themselves. In M.J. Mahoney & C.E. Thoresen (Eds.), Self control: Power to the person (pp. 263-290). Monterey, CA: Brooks-Cole.

Miller, W.R. & Rollnick, S. (2002). Motivational Interviewing. Preparing People for Change (2nd ed.). New York/London: Guilford Press.

Miller, W.R., Westerberg, V.S., Harris, R.J. & Tonigan, J.S. (1996). What predicts relapse? Prospective testing of antecedent models. Addiction, 91 (Suppl.) S155-S171.

Miller, W.R., Brown, J.M., Simpson, T.L., Handmaker, N.S., Bien, T.H., Luckie, L.F., Montgomery, H.A., Hester, R.K. & Tonigan, J.S. (1995). What works? A methodological analysis of the alcohol treatment outcome literature. In R.K.Hester & W.R. Miller (Eds.), Handbook of Alcoholism Treatment Approaches: effectiveness alternatives (pp. 12-44). Boston: Allyn & Bacon.

Monti, P.M., Kadden, R.M., Rohsenow, D.J., Cooney, N.L. & Abrams, D.B. (2002). Treating Alcohol Dependence. A Coping Skills Training Guide (2nd ed.). New York/London: Guilford Press.

Morgenstern, J. & Longabaugh, R. (2000). Cognitive-behavioural treatment for alcohol dependence: a review of evidence for its hypothesized mechanisms of action. Addiction, 95(10), 1475-1490.

Moser, A.E. & Annis, H.M. (1996). The role of coping in relapse crisis outcome: A prospective study of treated alcoholics. Addiction, 91, 1101-1113.

Oei, T.P. & Jackson, P.R. (1982). Social skills and cognitive behavioural approaches to the treatment of problem drinking. Journal of Studies on Alcohol, 1, 111-120.

Prochaska, J.O. & DiClemente, C.C. (1984). The Trans-theoretic approach: Crossing traditional boundaries of therapy. Chicago: Dow Jones/Irwin.

Rado, S. (1933). The psychoanalysis of pharmacothymia (drug addiction). Psychoanalytic Quarterly, 2, 1-23.

Rollnick, S. & Heather, N. (1982). The application of Bandura's self-efficacy theory to abstinence-orientated alcoholism treatment. Addictive Behaviour, 7, 243-250.

Roth, A. & Fonagy, P. (1996). What Works for Whom, a critical review of psychotherapy research. New York, NY: Guilford Press.

Savitt, R.A. (1963). Psychoanalytic studies on addiction. Psychoanalytic Quarterly, 32, 43-57.

Schwarzer, R. & Jerusalem, M. (1993). The Generalized Self-Efficacy Scale (GSES). From Measurement of Perceived Self-Efficacy: Psychometric Scales for Cross-Cultural Research. Berlin: Freie Universitat.

Stanton, A.L., Danoff-Burg, S., Cameron, C.L. & Ellis, A.P. (1994). Coping through emotional approach: Problems of conceptualisation and confounding. Journal of Personality and Social Psychology, 66, 350-362.

Taylor, S.E. (2006). Health Psychology (6th ed.). Boston/New York/London: McGraw-Hill International Edition.

Velicer, W.F., DiClemente, C.C., Rossi, J. & Prochaska, J.O. (1990). Relapse situations and self-efficacy: an integrative model. Addiction Behaviour, 15, 271-283.

Wunschel, S.M., Rohsenow, D.J., Norcross, J.C. & Monti, P.M. (1993). Coping strategies and the maintenance of change after inpatient alcoholism treatment. Social Work Research and Abstracts, 29, 85-103.

__________________________________________________________________________________________________

Clive Burkett, BA psy. is Programme Development Manager for Coolmine Therapeutic Community, Dublin, Republic of Ireland. He has 20 years' experience of work within the TC and has memberships of the IAAAC, EFTC and WFTC. Email: clive@coolminetc.ie

Effective Treatment, Ineffective Treatment: What Makes the Difference?
Some Indices of Predictability Taken from a Follow-up Study

Anna Addazi, Roberto Marini and Nicolino Rago

ABSTRACT: The article summarises the most salient data of a follow-up study that constitutes the third part of a more in-depth investigation into a sample of drug-addicted people from Città della Pieve residential community. We focused our attention first on the outcome of the period of therapy in terms of whether the therapeutic aims of treatment were met, roughly defined by the way in which individuals concluded the programme, and subsequently we measured the efficacy of the treatment through a follow-up interview of the remaining 63 of the original 121 members of the initial sample. Finally, we looked for those variables which correlated significantly with the results of the treatment, with reference both to the variables of the process and to variables relating to the history and psychological diagnoses of the individuals concerned.

Introduction

The follow-up study that this paper focuses on is the third part of a more in-depth investigation into a sample of drug-addicted people from Città della Pieve residential community. Due to obvious space limitations we are unable to set forward the research in its entirety. We limit ourselves to point out that, after having gathered information regarding the general condition of 121 people who took part in the service from 1999-2002 (first part), and, after having carried out a diagnostic valuation comprising every area of their lives - revealing in particular the gravity of their psychological problems (second part), we extended our analysis to the results of the residential treatment carried out by the subjects themselves, concentrating our analysis as much on the outcome of their treatment path as on its efficacy (third part). Besides including a descriptive analysis of the results of the follow-up research, in this article we will include the most significant evidence obtained by cross-referencing this data with some variables that are relative to the history and psychopathological condition of the sample. These were indicated in the previous two sections of the research.

The three research branches had the same sample of subjects in common, even if in the case of the follow-up investigation the number of people traced in the final interviews was almost cut in half compared to that of the first two parts (63 out of 121 subjects).

There were therefore three goals of the third part of the research.
1) To verify the treatment's efficacy by monitoring the process of emancipation from drug addiction in the time following the treatment, the average time from programme closure being about 30 months. It was important to take behaviour resulting from drug abuse into account as well as a series of indicators of the psychosocial situation, family and working conditions of subjects. This information was gathered using the EUROPASI (European adaptation from the Addiction Severity Index) follow-up interview predictions (Pozzi & Tempesta, 1995).
2) To more deeply investigate the outcome of the programme, which intended to reach goals of change that were fixed before and during the therapeutic path. This complex procedural variable was to be largely measured by the different methods through which people completed their stay on the programme.
3) More specifically: having codified information regarding the users' conditions for variables that rested along three temporal dimensions (before, during and after the treatment), our objective was to verify which of these variables were most closely correlated with the efficacy of the treatment path.

A selection of the most significant data regarding the above-mentioned points will be put forth in this publication.

Presentation of the service

The research targeted the community for drug addicts of the municipality of Rome, which is managed by the organisation (Cooperativa Sociale) ‘Il Cammino' in convention with the city's Agency for Drug Addiction Policies (Agenzia Comunale per le Tossicodipendenze, Comune di Roma). The community has been active since 1982 and rests in the territory of Città della Pieve, a town in the province of Perugia, from which it takes its name. It is made up of three operatives of rural homes situated in the Umbrian countryside and which house 45 users from the city of Rome.

The first phase of the community programme (residential welcoming) takes place in the rural home ‘La Villa', which also serves as a Diagnostic Observatory: the other two rural homes have the same function and are appointed to the second and third phases (treatment and clearance). Average programme length oscillates around 24 months, with staying periods in the community that alternate with periods outside the community of a progressively longer duration (from 48 hours to 20 days, starting from the seventh/eighth month of residence).

Once the residential phase is concluded, a re-entry phase begins, in which the individual is placed once again in urban territory. This uses resources ranging from psychological support (interviews as well as individual, family and couple psychotherapy) to work training and temporary relocation in protected apartments for those lacking housing. The service involves fourteen workers, of which six are psychologists, five are psychotherapists, two are community workers and one is a professional educator.

Materials and methods

The follow-up sample was made up of 63 out of the 121 from the original sample; for the remaining 58 it was not possible to administer the research instrument for the following reasons:
- for 21 people (17.4%) the telephone number was wrong or had been changed
- 12 people (9.9%) were in communities, the same or others
- 3 people (2.5%) were in prison
- 9 people (7.4%) had passed away
- 13 people (10.7%) refused to answer.

Out of the 63 subjects interviewed, 51 were male (81%) and 12 female (19%). At the time they entered the community their average age was 34.2 years (d.s.=5.3; min 22-max 45), while at the time of the follow-up interview their average age was 36.8 years.

EUROPASI's Follow-up Interview was the instrument used, which is made up of twenty-six items that investigate the following five areas of the subject's life: working conditions, alcohol and drug consumption, legal status, family and social relations, and psychological condition.

The questionnaire was administered via telephone interviews carried out over a five-month period. The two interviewers - not strangers to the line of work - remained strangers to most of the subjects interviewed. They introduced themselves as colleagues of a worker known to the ex-user, explaining the point and objective of the interview and guaranteeing that all received information would be used anonymously.

The data analysis was carried out according to a descriptive standard based on percentage estimates; the chi-square (χ2) test was used in order to verify the differences regarding the way attendance was distributed in the various contingency tables.

Figure 1 shows how the traced and non-traced populations are distributed based on the number of months they stayed in the community (in columns, from 1 to 42 separated by class). The figures in parentheses indicate the number of subjects.

It was evidently easier to trace those who had stayed longer in the structure, the difference being statistically significant [χ2(6)=19.642, p=.003]; quite probably, those who had an important and lasting relationship with community workers tended to stay in touch with them for a longer period of time after the end of the programme. This fact takes on a certain relevance for our investigation; in fact, it is important to keep in mind that most of our data refer to those who went through long-lasting programmes who were therefore more likely to have positive results (Figure 3).

In this section the data describing the conditions of the sample will be listed according to every variable investigated by the interviewer.

Working conditions

In the post-treatment period 69.8% of those interviewed were hired full-time, 22.3% were hired part-time, while only 7.9% didn't work; 54% of the subjects were satisfied with their occupation.

Alcohol and drug consumption

As far as relapse was concerned, the questionnaire asked about possible return to abusive behaviour in the following arcs of time:
1. in the period extending from the end of treatment to the 30 days preceding the interview (for brevity's sake we'll call it the post-residence period);
2. in the last 30 days before the interview;
3. in both periods.

Based on this temporal distinction, the following observations came forth: out of 63 ex-users interviewed, 30 had never relapsed (47.6%) and 33 had instead started up their abusive behaviour once again (52.4%): of these, 17 ‘relapsed' in the post-residence period, 2 in the last 30 days before the interview, and 14 had abused substances in both periods. At the time of the interview, there were therefore 47 (30+17) who were abstinent, or 74.6% of the sample, against 16 (2+14) who still had drug addiction problems (25.4%). This means that the 17 ‘relapsed' subjects after the end of treatment had already reached a drug-free condition at the time of the interview.

Furthermore:
- the percentage of relapses passes from 52.4% to 44.4% if those who chose to abuse only cannabis are extracted from the number of relapses;
- a good part of the sample that ‘relapsed' went back to abuse more than one substance, sometimes simultaneously (‘mix of substances' in 12.5% of the cases) and sometimes in succession. (The substances that were used the most were: heroin in 37.5% of cases, methadone (18.8%), cannabis (14.6%), alcohol (8.3%), cocaine (6.3%) and psycho-pharmaceuticals (2.1%).)
- 83.9% of those interviewed did not return to drug addiction treatment after resigning from the programme; most of those who requested intervention (12.9%) turned to territorial services to ask for non-residential substitution programmes, while the remaining 3.2% turned to semi-residential services.

Legal condition

The items that indicated the legal condition of subjects gave information on the relationship between the people in the sample and the justice system at the time of the interview: 79.4% of the sample left the programme free, that is, they were not put under restrictive measures, while 69.8% had no legal proceedings underway. All 19 subjects (30.2%) with pending lawsuits said that the problems related to crimes committed before beginning treatment.

Familial and social relations

Almost two-thirds (63.5%) of the subjects were single or celibate. Most of those interviewed (49.2%) lived with a partner or with a partner and children; another 22.2% lived alone or with friends and 25.4% lived with their parents. Over half (58.7%) of the sample were satisfied with their situation while 33.4% said they would like to change it. A large number (66.6%) of those interviewed affirmed that they have close friends, 44.4% spent free time with family members who did not use substances, while 41.3% spent time with friends who did not have drug addiction problems. In 65.1% of the cases, the subjects interviewed were satisfied with how they spent their free time.

Psychological condition

Just over a third (39.9%) of the sample demonstrated symptoms of serious depression in the post-residence period, 42.9% complained about intense anxiety and nervous tension, while 4.8% of those interviewed claimed to experience concentration, comprehension and memory difficulty. In 20.6% of the sample, they had trouble controlling violent behaviour, while 1.6% had attempted suicide or had had serious suicidal thoughts.

B - Results: treatment outcome

Outcome of treatment was evaluated according to the programme's various closure methods. Outcomes were as follows: twenty-one people (33.3%) stopped treatment (dropout); thirty-five (55.6%) finished it (‘programme end') and seven people (11.1%) resigned from the programme (‘spontaneous closure').

As previously stated, the high percentage of users in the sample that concluded the programme fully is explained by the fact that these were the easiest to trace. Spontaneous closure or self-resignation differs from pro-gramme end in that it is a form of closure that isn't supported by a joint goal-reaching valuation (social worker-user), but is instead initiated by the subject, who experiences a condition of greater wellbeing and considers it is good enough or not liable to further improvement. At times this form of resignation coincides with recognising a limit that can't be overcome at the time, such as a belief in the impossibility of freeing oneself completely from substance use. This is the case for those who decide to turn towards substances that are seen as less dangerous and taken in more controlled ways (i.e. in greatly reduced quantities or with a medical prescription). This conclusion to treatment differs from abandonment because it occurs in an advanced phase of the programme and because it implies a negotiation with the team, which results in a form of official recognition by the workers of the progress that has been made by the user.

Using the closure methods of the programme, the research discovered the following:
- gender of interviewees - women were found to be more prone to abandon the programme (50% versus 28.6% of men), but this item is not statistically relevant;
- number of months' stay in the community, shown in columns and subdivided in classes of six (Figure 2).

The relationship between the ‘programme end' and the number of months' stay appears curvilinear: 66.7% of those who completed the programme stayed in the community for a period of between 19 and 30 months; the percentage of people who finished treatment at programme end decreased if the period of stay was shorter or longer. Obviously, programme abandonment is concen-trated in the first year and a half. The percentage of spontaneous closures had peaks that corresponded with the third and sixth semester.

The reason why only two resignations from the programme are registered in the first semester is due to the fact that the sample is composed only of those who have decided to enter the treatment phase after having passed the residential welcoming phase, the phase with the greatest number of dropouts.

Residential treatments already carried out

At the ‘programme end' there are many more subjects who are having their first experience of residential treatment, 41.7% against 33% of those who have already had an experience of such treatment. There is a progressive reduction in the percentage of ‘agreed upon closures' and this reduction increases with the number of treatments already experienced. Those who have already gone through residential treatment are mostly distributed within ‘spontaneous closure' methods (29.7%).

C - Analysis of the results: treatment efficacy

The phenomenon of relapse will now be more deeply examined by analysing the difference between the group that relapsed (33 subjects) and the drug-free group (30 subjects) according to the following variables:
- gender of interviewees - a higher percentage of women relapsed (58.3% compared to 51% of men), although the difference was not statistically significant;
- number of months' stay in the community (Figure 3): if those who stayed in the community for less than six months are excluded (there were only two of these), those who stayed for a period of time between thirteen and eighteen months had a higher probability of relapsing (90%) and the difference between all length groupings is statistically significant [χ2(6)=15.27, p=.018].

• Programme closure standards (Figure 4): 80% of the subjects who concluded their relations with the community in the time and way agreed upon did not go back to abusing substances in the period following programme resignation whilst 19% of those who interrupted treatment did. This difference is statistically significant [χ2(2)=18.327, p=.0001].
• Residential treatments already carried out: the highest percentage of drug-free subjects (52.4%) is traced to those who are in their first community programme, a percentage that decreases the more programmes a subject participates in.

• Presence of alcoholic family members: those who have family members with alcohol problems are less likely to ‘relapse' (30.6% versus 84.2%). This difference is statistically significant [χ2(1)= 11.050, p=.001].
• Presence of family members with drug addiction problems: those with family members who had drug problems relapsed less (41.7%) compared to those that didn't (66.7%). This difference is statistically significant [χ2(1)=3.866, p=.049].
• Serious depression in the post-treatment period: relapse in substance use happened to 60.2% of those who suffered from serious depression against 43.8% of those who claimed not to suffer from depression. This difference is statistically significant [χ2(1)=3.384, p=.05].
• Intense anxiety in the post-treatment period: this variable doesn't seem to discriminate between those who relapsed and those who remained abstinent (51.9% of subjects who complained about frequent anxiety relapsed, while 52.8% of those that did not mention it relapsed).
• Psychological diagnosis formulated during the sample's stay in the community (Figures 5 and 6).

The presence or absence of an Axis I mental disorder* does not help us foresee substance abuse behaviour in the period after leaving the programme. Even though it missed statistical significance, the diagnosis of depression was the component that appeared to be the most associated with relapse. Other relevant mental health categories could not be taken into consideration due to the small numbers of subjects who exhibited them.

* This diagnosis was arrived at as the result of an evaluation done by a team of psychotherapists based on clinical observation, on the relationship between therapist and patient and on the results of a battery of psychological tests. These were the Minnesota Multiphasic Personality Inventory – 2 (Butcher, Dahlstrom, Graham, Tellegen & Kaemmer, 1989), the Defence Mechanisms Inventory (Ihilevich & Gleser, 1986) and the Holtzman Inkblot test (Holtzman, 1974).

The presence or absence of an Axis II personality disorder does not help us foresee substance abuse behaviour in the period after leaving the programme. Although all those diagnosed with paranoid and narcissistic disorders relapsed, the numbers were too small to be helpful in making predictions. The high number of subjects in the sample with a narcissistic personality disorder made it the only diagnosis from which it was possible to trace an indication: those with this diagnosis relapsed less compared to the total sample, but this does not reach statistical significance.

Discussion

Programme length and closing methods

The best results in terms of reaching a drug-free condition after treatment were obtained by those who had participated for a period of between 25 and 30 months (only 23.5% of ‘relapses'; Figure 3), a population that was largely the same as that which completed the programme in the time and way agreed upon by social workers: 80% of those who duly completed the programme did not ‘relapse' (Figure 4). Figure 2 helps us in this regard by clearly showing how the closure agreed upon by the programme reaches the peak of its concentration (41.7%) during the time band that goes from a 25- to a 30-month stay in the community. The joint valuation of reaching goals of change - that shared by both social workers and users - seems therefore to be the most trustworthy base on which the highest chance of a drug-free future can be based.

The percentage of relapses increased remarkably in groups of those who stayed in the programme for less than 25 months or more than 30 months, and the curvilinear relation between the two variables indicated that the probability of success is not directly proportional to the length of treatment (Figure 3). The fact that programmes longer than 30 months do not manifest better results than those that last less time (19-30 months) can be attributed to a greater resistance to treatment in those who, for self-perceived difficulties, prolong their stay in the community without this improving the efficacy of the therapeutic process.

Spontaneous closures

According to Figure 2, 30% of spontaneous closures are spread out over a 31-36 month stay in the community; this could be attributed to a perceived lack of strength and lack of faith in bringing the programme to a successful close. This perception could be linked to repeated relapses in the ‘clearance' phase and was often resolved by falling back on spontaneous closure followed by a new entrance to the programme a few months later. Another 30% of spontaneous closures were concentrated in the temporal interval extending from 13 to 18 months and were probably sustained by a wish to end the need for treatment because one is convinced of having reached therapeutic goals.

Treatment interruptions

From Figure 2 it is deduced that the highest percentage of abandonment (31.1%) is situated around the second temporal interval (7-12 months in the community); abandonment between seven and eighteen months in the community (51.4%) should probably be related to factors that are linked to the asperity and effects of the therapeutic path in course. Programme interruptions are less frequent after 18 months, and the reasons are probably traced to reinforcing support and in the fact that those who do not manage to achieve the ‘programme end' goal are nevertheless able to put other forms of closure in place such as resignation, which is more considered and less based on impulse. From a statistical point of view, another significant fact is the increase in the percentage of ‘relapses' in those who leave the programme during the third semester compared to those who leave during the second semester (90% in the first case and 58.3% in the second). This can mean that a therapeutic process that is launched and endures for some time, but is not consolidated, has a greater relapse risk than a process which has just been put into place. This is perhaps because the first is associated with a bigger vulnerability of those who no longer recognise themselves in what they had been before, but have not yet formed a new and more functional self-identity.

Other family members with drug addiction problems

Another variable that is worth considering in depth is the presence of family members with alcohol and drug problems in the families of our sample subjects. Statistical evidence shows that those who have one or more persons in their family with alcohol problems (84.2% of drug-free subjects) or drug problems (66.7% of drug-free subjects) were more able to abstain from substance abuse in the post-residential period. This was not a foreseen result and posits questions that deserve additional reflection. For now, we limit ourselves to suggest that in multi-problematic families the possibility of taking leave of the role of the carrier of the symptom is more easily contemplated, a possibility that is presumably more accessible to the member that actively and knowingly searches for it and is possibly supported by a relational network of orientation such as the community. On the other hand, the systemic-relational theory gives us an interpretive key that confirms the importance of the ‘rigidity/flexibility of roles' variable in configuring family dynamic and movement: we can hypothesise that a family with only one problematic member has less flexibility regarding family roles due to the concentration of negative expectations on one of its members.

Weight of the psychological diagnosis

From Figures 5 and 6 it is deduced that there are no important differences between the various diagnostic groupings as far as programme end standards are concerned. If the diagnosis of major depression is excluded (68.2% of relapses between those who suffered depression in the post-residential period; Figure 5), all the other diagnoses are irrelevant in determining the probability of relapse in the post-treatment period. In other words, they don't have value in predicting return to abusive behaviour. Relapse seems more greatly associated with variables that have to do with the environment or processes such as the socio-familial condition and the treatment path (Figures 3 and 4). In any case, it seems that it is the quality and dynamic of life relations rather than individual characteristics fixed by diagnostic features that weighs on the futures of our users, creating chances for change which are as rich as the potentialities contained in relationships and occasions of encounters. On this subject we suggest consulting the following authors: Galanter and Kleber (2006), Clerici, Bertolotti Ricotti and Malagoli (2000), Gori (2001), Putt (2001) and Serpelloni (2004).

Acknowledgements

Those who conducted the research have worked for various years in the community of Città della Pieve; they consulted Massimo Clerici while carrying out this study, whom they thank; a warm thanks is also extended to Federica Volpi for precious statistical consultation.

Butcher, J.N., Dahlstrom, W.G., Graham, J.R., Tellegen, A. & Kaemmer, B. (1989). The Minnesota Multiphasic Personality Inventory-2 (MMPI-2): Manual for administration and scoring. Minneapolis, MN: University of Minnesota Press.

Clerici, M., Bertolotti Ricotti, P. & Malagoli, M. (2000). Epidemiologia e diagnosi della comorbilità psichiatrica nei disturbi da uso di sostanze. Quaderni di Itaca numero 6. Roma: Cedis.

Galanter, M. & Kleber, H.D. (2006). Trattamento dei disturbi da uso di sostanze. Masson Italia.

Gori, E. (2001). L'esito a distanza dei trattamenti riabilitativi per chemiodipendenti. Quaderni di Itaca numero 7. Roma: Cedis.

Holtzman, W.H. (1974). New developments on the Holtzman Inkblot Technique. In P. McReynolds (Ed.), Advances in Psychological Assessment, Vol. 3. San Francisco: Jossey-Bass.

Ihilevich, D. & Gleser, G.C. (1986). Defense Mechanisms. Their Classification, Correlates, and Measurement with the Defense Mechanism Inventory. Psychoanalytic Quarterly, 57, 459-461.

Pozzi, G., Tempesta, E. et al. (1995). Europasi. Adattamento europeo dello Addiction Severity Index. Bollettino per le Farmacodipendenze e l'Alcoolismo, XVIII (2), 7-41.

Putt, C.A. (2001). La ricerca sull'esito nel trattamento dell'abuso di sostanze. In E.T. Dowd & L. Rugle (Eds.), La tossicodipendenza. Trattamenti a confronto di (pp. 9-52). Milano: McGraw-Hill Libri Italia.

Serpelloni, G. (2004). Modello concettuale per la valutazione dell'outcome nei trattamenti delle tossicodipendenze. Quaderni di Itaca numero 11. Roma: Cedis.

__________________________________________________________________________________________________

Anna Addazi is a psychologist and coordinator at Città della Pieve TC, Italy. Email for correspondence: addazi.anna@libero.it
Roberto Marini and Nicolino Rago are psychologists and operative unit co-ordinators at Città della Pieve TC, Italy.

Mark Freestone and Paul Goodman

ABSTRACT: This paper considers outcomes data from the Association of Therapeutic Communities' research project relating to the Ley Community, a therapeutic community (TC) for the treatment of drug addictions located near Oxford, UK. A total sample of 150 clients were recruited into the study from the community, 49 of whom were in treatment at the commencement of the study (the ‘baseline' sample) and 101 of whom began treatment after commencement (the ‘sequential' sample). The sample populations are described in terms of basic demographics, presenting psychopathology and presence of personality disorder. Longitudinal outcomes data is then considered relating to length of stay and change in psychopathology. Broad conclusions are then drawn about the effectiveness of the Ley Community as an intervention for clients with complex needs.

Introduction

In 1999, the Association of Therapeutic Communities, together with the University of Nottingham, funded by the then National Lottery Charities Board, began a naturalistic, comparative, cross-institutional study ‘in the field’ to evaluate the effectiveness of therapeutic community (TC) treatment for people with personality disorders (Lees, 2004). Twenty-one TCs took part in the study, drawn from the following categories: Day TC, Residential TC, Prison TC, Addiction TC and Private Sector TC. Data were returned from 507 clients.

This paper will concentrate on the data returned from the 150 clients who were treated at the Ley Community, an Addiction TC located on the outskirts of Oxford, UK. The Ley Community’s contribution to the overall n of the study was large (over a fifth of the total respondents with excellent survivability) and the data of a very high quality due to the involvement of an internal ‘on-site’ researcher in the project; thus we have elected to use this data to present a fuller picture of the impact of residence at the Ley Community that may not have been apparent from previous papers (Lees, Evans, Freestone & Manning, 2006; Freestone et al., 2006).

The Ley Community

The Ley Community was established in 1971 at Littlemoor Hospital in Oxford under the guidance of Drs Bertram Mandelbrote and Peter Agulnik. The programme was based on that developed in New York at Phoenix House. The first Director, John McCabe, was a graduate from Phoenix House, New York, who had been involved as a staff member in setting up the London Phoenix House in 1969 (see Agulnik & Wilson, 2007, for an account of the early history of the Ley Community). In 1979, the Ley Community moved to its current site now containing residential accommodation for up to fifty-eight residents in six acres of land fifteen minutes by bus from the centre of Oxford City.

The programme in an Addiction TC is based on the theoretical approach of ‘Community as Method’ (De Leon, 1997). There is a substantial body of research, mainly from the United States, that points to the efficacy of this treatment approach. The concept of a drug-free TC is defined as:

a drug-free environment in which people with addictive (and other) problems live together in an organised and structured way in order to promote change and make possible a drug-free life in the wider society. The therapeutic community forms a miniature society in which residents, and staff in the role of facilitators, fulfil distinctive roles and adhere to clear rules, all designed to promote the transitional process of the residents. Self-help and mutual help are pillars of the therapeutic process in which the resident is the protagonist principally responsible for achieving personal growth, realizing a more meaningful and responsible life, and of upholding the welfare of the community. The programme is voluntary in that the resident will not be held in the programme by force or against his/her will.
(Ottenberg, Broekaert & Kooyman, 1993)

Kaplan, Broekaert and Morival (2001) add that

every (addiction) therapeutic community has to strive towards integration into the large society; it has to offer its residents a sufficiently long stay in treatment; both staff and residents should be open to challenge and to questions; ex-addicts can be of significant importance as role models; staff must respect ethical standards, and therapeutic communities should regularly review their ‘raison d’être’.

The Ley Community is targeted at entrenched addicts who have reached a point where they desperately want to change. The treatment programme is divided into four phases that cover twelve months, though no two residents will go through an identical programme. Central to the process is building relationships in peer groups based on ‘trust, respect and honesty’ where residents can safely challenge each other’s behaviour and attitudes, and change. All activities – therapeutic groups, work on maintaining the grounds and fabric, cooking, housework, leisure and therapeutic games – provide material for learning and developing new skills. As trust grows, residents start to work on their ‘issues’ and frequently share early traumas of neglect and abuse. Through acceptance by others, a gradual process occurs whereby acceptance of ‘self’ takes place.

After nine months or so within the safety of a ‘closed community’, residents start preparing to move on. At this point, residents undertake voluntary work and begin socialising in Oxford with their peers. They take on the responsibility within the Community as senior peers, but begin looking outwards and start the preparing to look for work through putting together a CV and practising interviews. Once a full-time job has been secured (and approved by the staff team), a resident will move into the resettlement phase of the programme and for three months continue to live at the Ley Community, cater and budget for themselves, attend two weekly therapeutic groups, and act as auxiliary staff on a rota to support paid staff at weekends and in the evenings. Within three months, residents will have saved sufficiently to put a deposit down on a privately-rented house, usually with a couple of their peers. Over 50% of residents admitted into treatment at the Ley Community leave the programme in full-time employment having been drug free for over 12 months.

The Ley Community programme attempts to provide residents with the ‘opportunity to overcome serious drug and/or alcohol problems, and subsequently to lead full and fulfilling lives’ (Ley Community Mission Statement). Research undertaken at the Ley Community has demonstrated that, after six months in treatment, residents felt less anxious and depressed, less angry, more in control of their lives, and significantly more self-confident (Small, 2001; Small & Lewis, 2003, 2004). Research carried out in 1978 and repeated in 2001 demonstrated that the longer a resident remains in treatment, the less likely they are to subsequently be convicted of a further offence, with the outcome for those who complete the programme being particularly impressive (Wilson & Mandelbrote, 1978; Small, 2001).

The residential treatment sector for substance misuse in the UK has struggled in the last few years in face of the huge expansion of community treatment, mainly centred on substitute prescribing. Within the residential sector, TCs have been particularly hard hit with many commissioners not prepared to fund residential treatment for more than three to six months. The lack of objective research within the field in the UK into the treatment process leaves Addiction TCs vulnerable. Hence the involvement of the Ley Community in this research project. The current study provided an opportunity to compare the characteristics of the client group in treatment at the Ley Community with those in treatment at other TCs, and explore the impact that the programme had on them. This paper will highlight the most significant findings from the study.

Method

The ATCs’ study had a number of research aims (see Lees, 2004, for a full description of all the research aims of this project), of which five will be partially addressed in this paper, namely:

1. describe and measure the populations of the TCs involved, in terms of their background characteristics and their symptoms and behaviours at admission
2. compare the population characteristics of the different subgroups of TCs in the sample (e.g. non-secure/secure; residential/day)
3. particularly measure and discuss the profiles, or social climates, of these communities, and compare them to typical TC profiles
4. measure changes in symptoms and behaviours over time in treatment, and, for some of the population sample, after discharge
5. lastly, attempt to identify how the population characteristics at intake, and the various elements of the treatment environment and process are related to good outcome for members.

Thus the study would be a within-subjects’ (many administrations over time) and between-subjects’ (comparison of different subject groups) multivariate design.

These aims were to be met in two ways: firstly, there was to be a snapshot, or baseline, sample including everyone who was in the 19 TCs (i.e. those 19, out of the 21 TCs providing both environmental and client member data, which provided useable baseline or sequential admission client member data), when the data collection started on 1 April 2001, and who was willing to participate in the research project; and the second, sequential, sample included everyone who was admitted to these TCs after 1 April 2001 for a period of 18 months. Both of these samples were followed up, and retested at three-monthly intervals until the end of the data collection period, 30 September 2002. Members of both these samples were requested to complete a complete battery of environmental and psychometric measures at their first test, either on day one of the project (1 April 2001) or on their admission to the Ley Community, during the life of the project (i.e. between 2 April 2001 and 30 September 2002).

Ethical clearance for this study was obtained through multi-site clearance from the National Research Ethics Committee (NREC), and all respondents, including those at the Ley Community, were required to have signed consent forms before participating in the study.

An interesting aspect of the ATC project in relation to an addiction TC is that no psychometric outcomes measures dealing with substance misuse per se were employed; thus, the data presented here are at once somewhat tangential to the essential aims of the Ley Community, whilst also providing an interesting insight into the ‘generic’ effects of involvement in a TC where the focus is not directly on improving psychopathology or everyday functioning.

Initial description of the sample population

The ATC National Lotteries project surveyed a total of 150 residential clients at the Ley Community between April 2001 and October 2002. One hundred and twenty-two clients (85%) were male and twenty-two clients (15%) were female. The mean age of the clients at point of entry to the study was 29 years (M=28.97, SD=5.98).

For this sample, the mean length of stay in the community was 272 days (SD=184), a figure that reflects well on the Ley Community as a 12-month time-bound intervention, and there was no significant difference in length of stay between the two gender groups, nor was length of stay correlated with age (F(141)=.39, p=.53). The distribution of lengths of stay is represented graphically in Figure 1.

Cursory examination of Figure 1 shows that the length of stay of client members is not normally distributed (K-S D=1, p<0.001). Consistent with other findings (Chiesa, Drahorad & Longo, 2000; Lees et al., 2003; Birtle et al., 2007), there seem to be two ‘peaks’ in discharges: one at around the three-month mark (the ‘early dropouts’) and a second at formal discharge at around 500 days (the ‘completers’); this lends the histogram in Figure 2 a ‘saddle’ or inverse Gaussian (‘normal’) shape. This might suggest two separate populations.

In terms of presenting difficulties, the Ley Community sample was compared with the overall sample of the ATC/NLCB project for the four main change measures considered: the Clinical Outcomes Routine Evaluation: Outcomes Measure (CORE-OM) (Evans et al., 2000); Borderline Syndrome Index (BoSI) (Conte, Plutchik, Karasu & Jerrett, 1980); the EuroQol (EuroQol Group 1990); and the Brief Symptoms Inventory (Derogatis, 1983).

Table 1: Comparison of the Ley Community sample with other TCs in the NLCB study Outcomes measure


These initial descriptors would tend to indicate that the Ley Community population presented with significantly less severe symptoms across the measures used, than the remainder of the study sample.

Data were also available on the presence of personality disorders in the TCs considered for the study, using the Personality Diagnostic Questionnaire 4+ (Hyler, Reider, Spitzer & Williams, 1987): although prevalence of individual personality disorders varied (e.g. incidence of Antisocial PD was noticeably higher – see also Lees, Manning, Freestone & Evans (2003, unpublished)), overall prevalence of the ten DSM–IV Personality Disorders was not significantly different in the Ley Community from the remaining sample (χ2(10)=12.942, p=.227).

Results and analysis

Psychometric change

As described above, client members’ global symptomatology was assessed using the Brief Symptoms Inventory (BrSI) (Derogatis, 1983) global severity index (GSI) at three-monthly intervals. Boxplots showing longitudinal change in scores across these time intervals are shown in Figure 2. Each boxplot shows the median score (the thick central black line) of all participants at each administration; the ‘notches’ show the confidence interval for the median; and the blocky ‘shoulders’ show the inter-quartile range. Broadly speaking, if the ‘notches’ do not overlap, there will be a statistically significant difference between the two medians shown in the boxplots; so, from this figure, we would expect administration 2 to be significantly lower than administration 1 and administration 3 lower than 2.

In these circumstances, we can test for statistically significant change by comparing pre- and post-treatment measures, i.e. those taken at the first administration and then again at the nine-month stage for those in the sequential sample (i.e. those admitted after the study commenced and hence would not have any ‘prior’ treatment at the Ley). The underlying population is approximately normally distributed (K-S Z(133)=1.169, p=.13) and so a paired samples t-test was employed. This test suggests that, between the first administration and the nine-month stage, there is a significant mean reduction in GSI total score by 0.682 (t(21)=5.802, p<.001), implying an effect size of r=.78.

A second outcome measure used in the ATC study is the Borderline Syndrome Index (BoSI), a brief and fairly sensitive measure of general presenting difficulties (Conte et al., 1980) with excellent validity within this study (α=0.95). Figure 3 shows notched boxplots for client scores at each administration (‘1’ being entry to the study and ‘6’ being at 15 months).

Outcomes data generated by the BoSI are ordinal and do not follow a normal distribution (K-S D=0.97, p<0.001), meaning that non-parametric tests were used to ascertain the significance of change over time in individuals’ scores.

Value of n for the sixth administration is very small (n=3), so we looked at the significance of the reduction in BoSI total scores over a nine-month period, or to the fourth administration (n=21). This change was highly significant at the 99% level (Wilcoxon V(37)=604, p<0.001), and suggests an effect size of r=.78, or a strong reduction in Borderline symptomatology following nine months of treatment. For the three clients who stayed on in treatment for fifteen months, this effect size increases to r=–.92, although the result is not statistically significant (p=.11).

The final longitudinal outcome measure considered by this study is the CORE-OM (see Figure 4). Again, the data were normally distributed so parametric tests were employed: these suggested a significant reduction in mean score after nine months (t(22)=6.507, p<0.001) with effect size r=.70.

Quality of life

Another way of looking at Ley Community clients’ improvement over time is to explore the relationship between length of time spent in the community and self-reported improvement in clients’ quality of life, as measured by the EuroQol tool (EuroQol Group, 1990). The EuroQol tool features a graphical ‘barometer’ that requests clients mark their overall quality of life on a scale of 0–100. The EuroQol was administered to each client twice: once at entry to the community, and once at discharge; Figure 5 shows a scatterplot of discharge scores against length of stay.

The correlation in Figure 5 is fairly weak, but it is positive (τ=.173) and significant (one-sided p<0.05), which would tend to indicate that a longer stay in the Ley community was associated with improved self-reported quality of life.

Early dropouts and long-stayers

Previously cited work completed at the Ley community (Wilson & Mandelbrote, 1978) suggested that ‘long-stay’ members completing six months or more of treatment obtained better outcomes in relation to drug relapse and criminality. We can identify from the sequential sample (that is, those clients who joined the community during the study) a subset of cases who then stayed on through treatment for over 180 days after joining the study. For this sample n=36, and the presenting difficulties were not significantly different from the remainder of the Ley community sample for the EuroQol, BoSI, CORE-OM or BrSI–GSI; nor were outcomes.

However, research performed in other TCs (Chiesa, Wright & Need, 2003) has suggested a link between borderline personality disorder and dropout rates in a TC. When we analyse the prevalence of personality disorders in the ‘long-stay’ sample, we find significantly lower incidence of Narcissistic (χ2(1)= 5.118, p=.025) and Borderline (χ2(1)=5.697, p=.020) personality disorders in this population.

Reliable change

A final test we can make, based on a method developed by Jacobson, Follette and Revenstorf (1984) and further clarified by Evans, Margison and Barkham (1998), utilises the reliability coefficients for the various outcome measures for the study (previously published in Lees et al., 2006) in order to calculate whether or not the levels of change experienced by those in the treatment sample are reliable; that is to say, they are not simply due to inconsistencies or reporting quirks in the measures. Using this technique, a threshold for mean change is calculated, above which the change experienced by the clients becomes ‘reliable’ (i.e. there is a 95% chance that this is not a spurious finding due to unreliability of the psychometric itself). The results of these calculations are given in Table 2.

Table 2: Reliable change for Ley Community clients using the BrSI and CORE-OM Outcome measure

Measures of therapeutic environment

A final aspect of the ATC/NLCB study worth considering in understanding outcomes at the Ley Community would be to look at measurements of the therapeutic environment, in an attempt to understand ‘what kind of place’ clients find when they are accepted for treatment. Of particular interest might be how the Ley differs from other TCs in the various cultural aspects of its programme delivery.

The measure used by the study to assess the ‘quality’ of the therapeutic environment was the Community Oriented Programmes Environment Scale (Moos, 1997). This measure poses a series of 100 true/false statements about the environment of their community to client and staff members (e.g. ‘This place always looks clean and tidy’, ‘Client members are not afraid to speak their minds’). The responses are then factored into 10 subscales, relating to various aspects perceived as relevant to a ‘healthy’ community.

For the purposes of this paper, the largest available data ‘snapshot’ – from October 2001 – was taken of completed COPES measures for the Ley Community and for the rest of the ATC/NLCB sample. These were then plotted onto a line graph (Figure 6) and differences tested. Significant differences at the 95% level were present for all subscales except for ‘Support’ (p=.08) and ‘Staff Control’ (p=.133).

Discussion and conclusions

Several findings of interest arise from the foregoing study that certainly merit further discussion and possibly further investigation, including the following.

1. There are substantial differences between the severity of presenting psychological differences within clients attending the Ley Community and those within the national sample, with the Ley clients presenting with much lower levels of psychopathology.
2. There is significant and reliable clinical improvement in the presenting difficulties of clients who remained in treatment at the Ley Community for at least nine months (aggregate effect size r=.75).
3. There was a modest but significant correlation between length of stay in the Ley Community and improved self-reported quality of life.
4. There were significant differences between various facets of therapeutic environment at the Ley Community when compared to the overall ATC/NLCB sample; of particular note were the higher levels of Involvement, Practical Orientation and Order and Organisation in the Ley programme relative to the sample.
5. The significantly higher prevalence of narcissistic and borderline personality disorders in clients who leave treatment before the expected duration (six months or more) suggests that individuals presenting with these difficulties are at particular risk of ‘early dropout’.

These results are encouraging, as they suggest that treatment in the Ley Community is associated with a strong, reliable reduction in presenting psychopathology across a number of variables and outcome measures for clients who remain in treatment for at least nine months. The extant question for the community is not so much ‘is this intervention effective?’ (repeated studies have now shown that it is), but rather ‘why do some people leave the community early?’ This is not a question that can be explored through psychometric outcomes data; it necessitates another investigation in its own right, one that might be based on long-term follow-up and collaborative, non-judgemental exit interviews.

This study may be perceived as limited in the sense that the outcomes considered are largely only proxies for the true outcomes of the Ley Community: i.e. enabling clients to remain drug-free post-treatment. Reduction in psychological symptomatology may be a necessary and desirable part of such an outcome, but it is not sufficient in its own right. As well as repeated work on reconvictions, such as that performed by Wilson and Mandelbrote (1978) and Small (2001), it would be desirable to consider follow-up factors such as employment, relapse into drug use and hospital admissions.

References

Agulnik, P. & Wilson, S. (2007). The early history of the Ley Community – Personal accounts. Therapeutic Communities, 28(1), 11–16.

Birtle, J., Calthorpe, B., McGruer, F., Adie, L., McCullagh, G. & Kearney, S. (2007). Preparing to go to main house – this week, maybe next week! Therapeutic Communities, 28(2).

Chiesa, M., Drahorad, C. & Longo, S. (2000). Early termination of treatment in personality disorder treated in a psychotherapy hospital. British Journal of Psychiatry, 177, 107–111.

Chiesa, M., Wright, M. & Need, R. (2003). A description of an audit cycle of early dropouts from an inpatient psychotherapy unit. Psychoanalytic Psychotherapy, 17(2), 138–149.

Conte, H.R., Plutchik, R., Karasu, T.B. & Jerrett, I. (1980). A self-report borderline scale: discriminative validity and preliminary norms. Journal of Nervous Mental Disorders, 168, 428–435.

De Leon, G. (1997). Community as Method: Therapeutic Communities for Special Populations and Special Settings. New York: Praeger.

Derogatis, L.R. (1983). SCL–90R: Administration, Scoring & Procedures Manual for the R(evised) Version. Towson, MD: Clinical Psychometric Research.

EuroQol Group (1990). EuroQol--a new facility for the measurement of health-related quality of life. Health Policy, 20(3), 321–328.

Evans, C., Margison, F. & Barkham, M. (1998). The contribution of reliable and clinically significant change methods to evidence-based mental health. Evidence Based Mental Health, 1, 70–72.

Evans, C., Mellor-Clark, J., Margison, F., Barkham, M., McGrath, G., Connell, J. et al. (2000). Clinical Outcomes in Routine Evaluation: The CORE-OM. Journal of Mental Health, 9, 247–255.

Freestone, M., Lees, J., Evans, C. & Manning, N. (2006). Histories of Trauma in Client Members of Therapeutic Communities. Therapeutic Communities: The International Journal for Therapeutic and Supportive Organizations, 27(3), 411–434.

Hyler, S., Reider, R.O., Spitzer, R.L. & Williams, J. (1987). Personality Diagnostic Questionnaire – Revised. New York: New York State Psychiatric Institute.

Jacobson, N.S., Follette, W.C. and Revenstorf, D. (1984). Psychotherapy outcome research: methods for reporting variability and evaluating clinical significance. Behaviour Therapy, 15, 336–352.

Kaplan, C.D., Broekaert, E. & Morival, M. (2001). Improving social psychiatric treatment in residential programmes for emerging dependence groups in Europe: cross-border networking, methodological innovations and substantive discoveries. International Journal of Social Welfare, 10(2), 127–133.

Lees, J. (2004). Practice Evaluation of Therapeutic Communities. In J. Lees, N. Manning & N. Morant (Eds.), A Culture of Enquiry: Research Evidence and the Therapeutic Community (pp. 76–90). London: Jessica Kingsley.

Lees, J., Evans, C., Freestone, M. & Manning, N. (2006). Who comes into therapeutic communities? A description of the characteristics of a sequential sample of client members admitted to 17 therapeutic communities. Therapeutic Communities, 27(3), 411–434.

Lees, J., Manning, N., Freestone, M. & Evans, C. (2003, unpublished). Some Preliminary findings from the ATC/NLCB Therapeutic Communities Project – A comparative evaluation of Therapeutic Community effectiveness for people with personality disorders. Association of Therapeutic Communities Annual Conference, Windsor.

Moos, R.H. (1997). Evaluating Treatment Environments: the quality of psychiatric and substance abuse programs. New Brunswick, NJ: Transaction Publishers.

Ottenberg, D.J., Broekaert, E. & Kooyman, M. (1993). What can and cannot be changed in a therapeutic community. Orthapedagogische Reeks Gent, 2, 51–63.

Small, M. (2001). Two Year Reconvictions in a Rehabilitation centre. Therapeutic Communities, 22(2), 153–166.

Small, M. & Lewis, S. (2003). Six-month follow-up of anxiety and depression in polysubstance misusers undergoing treatment in a therapeutic community. Therapeutic Communities, 24(4), 142–160.

Small, M. & Lewis, S. (2004). Changes in hostility in residents undergoing treatment for drug and/or alcohol misuse in a residential centre. Therapeutic Communities, 25(3), 219–235.

Wilson, S. & Mandelbrote, B. (1978). The relationship between duration of treatment in a therapeutic community for drug abusers and subsequent criminality. The British Journal of Psychiatry, 132, 487–491.

______________________________________________________________________________________

Mark Freestone is Senior Clinical Forensic Research Officer, East London Forensic Personality Disorder Service, London, UK. Email: mark.freestone@eastlondon.nhs.uk
Paul Goodman was Chief Executive at The Ley Community, UK from 1998 to 2008.

Empowering Female Inmates: An Exploratory Study of a Prison Therapeutic Community for Substance Abuse and its Impact on the Coping Skills of Substance Abusing Women

Danielle McDonald

ABSTRACT: Robert Johnson's model of mature coping was applied to a therapeutic community (TC) for women to better understand if the women were learning positive coping skills that would replace their prior negative coping skill of substance abuse. Face-to-face interviews were conducted to examine the differences between women in the TC and women on the waiting list for treatment. Data from this study suggest that women involved in the TC were able to improve their problem-solving and seeking social support skills, while many still struggled with avoidance techniques.

The Relationship Between Neighborhood Criminal Behavior and Oxford Houses

Jeffrey Deaner, Leonard A. Jason, Darrin M. Aase and David G. Mueller

ABSTRACT:

Experts by Experience: Discovering the Heart of a Therapeutic Group IQ

Peter R. Holmes and Susan B. Williams

ABSTRACT: The present study investigated crime rates in areas surrounding 42 Oxford Houses and 42 control houses in a large city in the Northwestern United States. A city-run Global Information Systems' (GIS) website was used to gather crime data including assault, arson, burglary, larceny, robbery, sexual assault, homicide, and vehicle theft over a calendar year. Findings indicated that there were no significant differences between the crime rates around Oxford Houses and the control houses. These results suggest that well-managed and governed recovery homes pose minimal risks to neighbors in terms of criminal behavior.

Is Alcoholics Anonymous a Therapeutic Community?

Martin Weegmann

ABSTRACT: Alcoholics Anonymous and other Fellowship groups are effective and sustaining organisations for many addicted individuals. This paper notes elements in common between such mutual-help groups and formal therapeutic communities, arguing that they can be regarded as being special therapeutic communities (TCs) in their own right. Premised on recognition of mutual vulnerability and appreciation of the role of human defences and assets, AA addresses change through lay processes of group psychology and involvement in a sense of community

John Dowsett and Jackie Craissati (2007) Managing Personality Disordered Offenders in the Community - A Psychological Approach. Published by Routledge, pp. 215, price p/b £19.99, Paperback, ISBN: 1-58391-739-8. Reviewed by Simon McArdle.

A. Washton and J. Zweben (2006) Treating Alcohol and Drug Problems in Psychotherapy Practice. Published by Guildford Press (New York), pp. 312, Paperback, ISBN: 978-1-59385-980-0. Reviewed by Martin Weegmann.

John Gordon and Gabriel Kirtchuk (2008) Psychic Assaults and Frightened Clinicians: Counter-transference in Forensic Settings. Published by First edition published by Karnac, pp. 152, Paperback, £????, ISBN: 978-1-85575-562-8. Reviewed by Phil Garnham.

L.A. Jason, J.R. Ferrari, M.I. Davis and B.D. Olson (Eds.) (2006) Creating Communities for Addiction Recovery: The Oxford House Model. Published by Binghampton, NY: Haworth Press Inc., pp. 180, £24.99, ISBN-10: 0789029308, ISBN-13: 978-0789029300. Reviewed by Rowdy Yates.