Approximately one to two weeks before entering the TC, participants were asked to take part in a face-to-face interview. Information was gathered about socio-demographic background, physical and psychological health, education, employment, substance abuse history, illegal activities and family/social relationships. In-treatment assessment took place 30 days after the initial interview, gathering information on treatment process variables, psychological well-being, personality disorders and motivation. This particular time frame was chosen to minimize the effect of maturation. More specifically, as two of the five participating TCs have a welcome phase, where clients are prepared for a TC life over a period of one to two months, it is assumed that these clients will have a different perspective on TC treatment than clients who enter a TC program relatively unprepared. To be able to take into account this maturation effect, we decided that baseline data would be gathered at the moment the client decided to enter a TC program. For clients that were in a welcome phase, this moment was set around 10 to 14 days before intake. The unprepared clients – clients who entered a TC after being in a detoxification center or a crisis center, or who came straight from prison – were assessed 1 to 5 days before intake.
While most of the data was gathered by the main researcher, some EuropASI data was also collected with the help of professionals or master students in Educational Sciences, trained in EuropASI interviews. The participants were informed that the data would be processed anonymously and that the overall purpose of the study was to assess those aspects of TC treatment which might be improved. Written informed consent was obtained from each participant prior to the first interview. Ethical approval for the study was granted by the Ethical Review Board of the Faculty of Psychology and Educational Sciences at Ghent University.
Client background data, demographic data and the severity of substance use and related problems were obtained with the EuropASI, an adapted and validated version of the Addiction Severity Index (ASI) for the European context [39, 40]. The ASI explores clients’ current and lifetime functioning in seven different areas (medical status; employment and support; drug use; alcohol use; legal status; family and social relationships; and psychiatric status), displaying a multidimensional problem severity profile. An ASI composite score is calculated for each of the seven life domains (range 0–1), with higher scores indicating higher problem severity . In our study, composite scores are based on events that occurred 30 days before entering a detoxification centre and on the client’s perceived need for help at that time. However, for clients who entered the TC following a period of imprisonment or hospitalisation, the composite scores are based on the events that occurred 30 days prior to TC intake.
Treatment motivation was measured with the Circumstances Motivation Readiness and Suitability Scales (CMRS) . This is a self-administered questionnaire with 42 Likert-type items rated on a 5-point scale, which ranges from ‘strongly disagree’ to ‘strongly agree’. The instrument’s first scale, ‘Circumstances’, refers to the external conditions or reasons that influence people to enter or leave treatment. The second scale, ‘Motivation’ (internal pressures), refers to the individual’s inner reasons for change. These reasons can be initiated by feelings of guilt or self-loathing, i.e. negative feelings that are associated with a drug-related lifestyle, or by a belief in one’s own personal growth and the desire for a better life. The third scale, ‘Readiness’, underlines the perceived need for treatment in order to change. The ‘ Suitability’ items examine the individual’s perception of the appropriateness of the treatment modality. This scale determines to what extent clients think the TC treatment matches their needs. The psychometric properties of the Dutch translation were found to be acceptable and in line with the findings of the American studies . The current study obtained Cronbach alpha coefficients ranging from .67 to .83 across the four scales.
Psychological distress was measured with the Brief Symptom Inventory (BSI) [43, 44], derived from the SCL-90-R (Symptom Check List-90-R). This is a 53-item self-report scale used to measure recent psychological complaints (past 7 days) (somatization, obsessive-compulsive behavior, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation and psychoticism). Symptoms are rated on a 5-point Likert scale ranging from ‘Not at all’ to ‘Extremely’ (range 0–4). The higher the score, the greater the level of psychological distress. In this study, we used the Global Severity Index (GSI), an average rating of all 53 items and overall score of psychological functioning (Cronbach Alpha of .95). The cut-off score of the GSI (0.66 for males; 0.71 for females) is used as a general measure of psychopathology.
To measure personality traits the Assessment of DSM-IV Personality Disorders (ADP-IV) questionnaire was used . The ADP-IV is a validated Dutch self-report measure consisting of 94 Likert-type items that allows for a categorical and dimensional assessment of the 12 DSM-IV personality disorders [46, 47]. The dimensional interpretation emphasizes the continuity between normality and pathology of the DSM-IV personality ‘traits’ and is measured on a 7-point Trait (T) scale. The ‘distress’ of the subject or his/her environment as a consequence of having the trait criterion is assessed with a 3-point distress (D) scale. The categorical diagnostic evaluation is based on the following algorithm: ‘T > 4 and D > 1’; an item is scored ‘pathological’ when the trait score is larger than four and the distress score is larger than one. In accordance with the DSM-IV criteria, four or more items need to be scored positive/pathological before a diagnosis of a personality disorder can be made .
For the bivariate and multivariate analysis we used the dimensional assessment by summing the ADP-IV trait scores for the 3 clusters. Cluster A represents disorders that are marked as ‘odd or eccentric behavior’: paranoid, schizoid and schizotypical personality disorders. Cluster B refers to those disorders that manifest ‘dramatic, emotional or erratic behavior’, i.e. antisocial, borderline, narcissistic and histrionic personality disorders. Finally, Cluster C corresponds to disorders that are marked as anxious or fearful behavior, i.e. avoidant, dependent and obsessive-compulsive personality disorders. The Cronbach’s alpha coefficients ranged from .85 to .88 across the three clusters.
The treatment process was assessed with the Dimensions of Change Instrument (DCI) (cf. background). It is a 54-item questionnaire that assesses clients perceptions on various components of the TC treatment process. All items are positively worded and ask respondents to indicate their extent of agreement on a 5-point scale (1 = Not at all to 5 = Completely) with higher scores indicating a greater extent of agreement. The instrument consists of eight different subscales . These are: (1) Community Responsibility (CR) – the client personally accepts the rules of conduct; (2) Clarity and Safety (CS) – the client has a good understanding of the goals, structure, patterns of interpersonal interaction and feels safe in the community environment; (3) Group Process (GP) – the client observes the group meetings as helpful and perceives that residents actively participate in group therapy activities; (4) Resident Sharing, Support, and Enthusiasm (RS) – the client perceives residents as being enthusiastically engaged in sharing of personal feelings and being supportive in social interactions; (5) Introspection and Self-Management (IS) - the client engages in personal self-awareness and reflection, and adopts self-management enhancement activities; (6) Positive Self-Attitude and Commitment to Abstinence (PS) – the client admits to feelings of self-efficacy and commitment to achieving abstinence; (7) Problem Recognition (PR) – the client recognizes that his/her personal behavior and attitudes can lead to personal and interpersonal problems; (8) Social Network (SN) – the client believes he or she has a supportive social network outside of the TC community) . The first four subscales are clustered in the Community Environment (CE) summary dimension whereas the latter four are grouped in the ‘Personal Development and Change (PDC) summary dimension.
For the present study, the instrument has been translated into Dutch using back and forward translation. Subscale scores were calculated as the mean of the respective items, while summary scores for the two DCI dimensions represent the mean scores of the respective subscales. The internal consistency of the Dutch version of the DCI shows alpha reliability coefficients of .87 for the Community Environment summary dimension and .82 for the Personal Development and Change summary dimension. For the separate scales the Cronbach’s alpha ranges from .61 to .81.
To verify clients’ early perceptions on the TC treatment process we computed the means and standard deviations for the two DCI summary dimensions and the eight DCI subscales. Multiple linear regressions were used to determine the fixed and dynamic predictors for the two DCI summary dimensions. We first used bivariate analyses, including Pearson product–moment correlations, independent t-tests and one-way ANOVA’s, as appropriate to the level of measurement, to determine any relations between potential predictor variables and the two DCI summary dimensions, or at least one of the dimension’s subscales. These tests were performed on treatment site, client demographics (e.g. age, gender and ethnicity), all important EuropASI items, the EuropASI composite scores, the three ADP-IV clusters, the BSI total average score, and the four scales of the CMRS. To help control for the inflated alpha levels due to multiple testing and to focus results on the larger effect sizes for clinical significance, we only withheld the variables that were associated at the 0.01 level of significance.
Each time, the variables were entered in the regression equation in one single step using the default method. Data analysis was conducted using the SPSS 19.0 statistical program. Visual examination of the standardized residuals (the errors) by the regression standardized predicted values indicate that both, the assumption of linearity and homoscedasiticty, was met; the residual plot was rectangular with concentration of points around zero, respectively. Also the collinearity diagnostics revealed no difficulties. Variance inflation factors (VIF) and tolerance values were within the acceptable ranges; all VIF values were below 10.0 and all tolerance values were above 0.10.