- Email: [email protected]
Therapeutic communities
Journal of Substance Abuse Treatment 19 (2000) 375 ± 382
Article
Therapeutic communities Enhancing retention in treatment using ``Senior Professor'' staff George De Leon*, Josephine Hawke, Nancy Jainchill, Gerald Melnick Center for Therapeutic Community Research (CTCR) at National Development and Research Institutes, Inc. (NDRI), 16th Floor, 2 World Trade Center, New York, NY 10048, USA Received 23 December 1998; received in revised form 27 March 2000; accepted 4 April 2000
Abstract Evaluation research documents a firm relationship between retention and treatment outcomes among substance abusers in therapeutic communities (TCs). However, most admissions leave treatment prematurely, particularly in the first months after admission. This paper reports findings from a controlled study that assessed the efficacy of an intervention to reduce early dropout in a residential TC. In the ``Senior Professor'' (SP) intervention, the most experienced clinical and managerial staff in a TC program were utilized to conduct program induction seminars during the first weeks of admissions, traditionally the period of the highest rate of dropout. Rates of short - term retention (30 days of treatment) were compared for the experimental cohort (N = 362), who received the intervention, and a cohort of admissions (N = 243), who received standard or non - enhanced treatment. Findings showed that the SP intervention significantly reduced the likelihood of early dropout compared with controls. The enhanced effects are most evident for the new inductees with the lowest levels of motivation. Some theoretical and clinical implications are discussed as to the utilization of experienced staff to increase retention among new inductees with relatively lower motivational levels, who are at greatest risk for early dropout. D 2001 Elsevier Science Inc. All rights reserved. Keywords: TC staff; Early dropout; Motivation; Treatment process; TC theory
Considerable research shows that among clients who enter therapeutic communities (TCs), successful outcome is directly related to the length of stay in treatment (e.g., Barr & Antes, 1981; De Leon, 1984; De Leon et al., 1982; Holland, 1983; Hubbard et al., 1984, 1997; Simpson et al., 1997; Simpson & Sells, 1982). However, most admissions to TCs leave treatment before beneficial effects are rendered. Dropout is highest within the first 30 to 60 days of admission. Thereafter, it declines rapidly, leveling off at a low constant rate (e.g., De Leon & Schwartz, 1984). The temporal pattern of retention is characteristic for long - term residential TCs (and for drug treatment programs in general), indicating that the likelihood of remaining and completing treatment increases with time in the program itself (e.g., De Leon & Schwartz, 1984; Glaser, 1974; Hubbard et al., 1984; Joe & Simpson, 1976; Sansone, 1980). Therefore, to expand the beneficial effects of TC
* Corresponding author. Tel.: +1 - 212 - 845 - 4417; fax: +1 - 212 - 845 4698. E-mail address: [email protected] (G. De Leon).
treatment, efforts are needed to reduce the likelihood of dropout, especially during the early months of residency. As part of an ongoing program of research at the Center for Therapeutic Community Research (CTCR), a series of initial studies have explored the influences of staff factors on client retention, progress in treatment and treatment outcomes. This paper reports findings from a controlled study which assessed an intervention utilizing senior staff or ``Senior Professors'' (SPs) to enhance retention in a long term residential TC. The theoretical rationale that guided the present research views retention as reflecting interactions among three domains of variables: (1) external influences, (2) certain client characteristics and (3) the various activities and elements of the TC regimen, which are potential interventions in the treatment process. The diversity of clients with respect to their motivation and readiness to change and their suitability for the TC are particularly important in the early phases of treatment. Because relatively homogeneous program procedures are imposed upon this diversity, the process of inducting clients into their new roles as TC residents is extremely difficult.
0740-5472/00/$ ± see front matter D 2001 Elsevier Science Inc. All rights reserved. PII: S 0 7 4 0 - 5 4 7 2 ( 0 0 ) 0 0 1 2 4 - 0
376
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
The problem of role induction is further compounded by staffing patterns and functions in TCs. Traditionally, the responsibility for inducting new members into residential treatment programs belongs to relatively junior staff. This staffing pattern is based on the assumption that new admissions will be most successfully recruited into the TC regimen by staff and resident peers with whom the recruit can most readily identify, namely, recovering or recently recovered substance abusers. However, junior staff seldom has clinical experience to handle the diversity among new admissions and the problems that they experience during initiation. Therefore, it is understandable that the dropout rate is highest in the first days of treatment. Conversely, the underlying assumption of the SP intervention is that short- term retention can be enhanced if program procedures and staffing are modified to facilitate the management of diversity among new inductees. In other words, new inductees are exposed to senior staff who has the most clinical experience during the initial phase of treatment. At the same time, contact with junior staff with whom they can most readily identify and who provides positive role models for recovery is maintained. The present research utilized senior staff to facilitate the role induction of new admissions into the program. The most experienced clinical and managerial personnel in a TC program delivered a series of special seminars to first-time admissions during the initial weeks following their entry into long - term residence. The goal of this intervention was to increase the likelihood that new admissions to the residential TC would complete the initial induction stage of treatment. 1. Method 1.1. The SP intervention A faculty of senior staff or SPs was assembled from various departments within the TC agency to conduct seminars for new admissions to residential treatment. The SP seminars were carried out in a residential induction facility, which served as an assessment and orientation center for all new admissions. After 30 days in the induction facility, residents were transferred to the agency's long -term residential facilities. All residents in the induction facility convened at the usual time of the day scheduled by the program for seminars. Three SP seminars were delivered weekly by different SPs to all facility residents in a classroom setting. Each seminar lasted approximately one and a half hours. Seminar themes, though varying, all reflected the TC perspective and approach to treatment and recovery (e.g., De Leon, 1995, 2000). SPs selected their topic from a list of relevant TC themes (e.g., self -esteem, self -help and recovery concepts, the TC philosophy and its expectations, TC
retention and success rates). All SP seminars reflected problems of staying in treatment, and dealt with a wide range of concerns and questions that are often expressed by new TC members (e.g., fluctuating motivation, outside pressures and attractions, the structure and demands of the program, ambivalence about long - term residence, next steps and success rates). The seminar format followed the regular format of seminars as they are usually delivered in TCs. The format excluded formal protocols. It was didactic and experiential with considerable interaction between the seminar leader (the SP) and the resident participants. Thus, SPs used their experience, knowledge and charisma to address the concerns, questions and issues that mark the early days of the induction process. ( Note: Details of the SP procedure and seminar topics can be obtained from the first author.) 1.2. Research design Two trials of the SP intervention were conducted on separate cohorts admitted a year apart, Trial 1 ( November 20, 1984 through January 30, 1985) and Trial 2 ( February 1, 1986 through March 31, 1986). The SP staff and curriculum were the same in both trials. However, the shorter time frame in Trial 2 resulted in a different maximum number of sessions that could be attended during a 30 -day residential stay (N = 16 in Trial 1; N = 11 in Trial 2). The experimental design was a variant of time series or an ``On ± Off'' sequence. Each trial was defined as the 8 - to 10 -week period of the intervention, which was preceded by 8 to 10 weeks of the standard induction period. Retention rates between the experimental condition (the ``On'' period, which consisted of the standard induction procedures plus the SP intervention) were compared to the preceding control condition (the ``Off'' period, which consisted of the standard induction procedures only). The utility of the ``On ± Off'' design relative to matched controls and random assignment alternatives in field settings is discussed elsewhere (i.e., Staines et al., 1999). 1.3. Study sample The study site was a large northeastern drug treatment agency that was representative of traditional long - term residential TCs. The study sample (N = 605) consisted of all consecutive admissions to the induction facility during the two experimental trials of the project. The sample size of the combined trials was N = 362 admissions (Trial 1, N = 239; Trial 2, N = 123) for the experimental condition and N = 243 admissions (Trial 1, N = 173; Trial 2, N = 70) for the control condition. 1.4. Data collection Comprehensive social and psychological data were collected on all consecutive admissions to the residential
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
facility during all months of the study. Drug use, criminality, employment and other background information were obtained, as well as an extensive psychological battery that included the Beta IQ, the Tennessee Self -Concept Scale, the MMPI, the Beck Depression Inventory, the Manifest Anxiety Scale, the Socialization Scale of the California Personality Inventory, the Diagnostic Interview Schedule (DIS) (Robins et al., 1989) and an 18 - item version of the Circumstances, Motivation, Readiness, and Suitability (CMR) scales (De Leon & Jainchill, 1986; De Leon et al., 1994) to examine readiness for TC treatment. The data collection required approximately 5 h across 3 days of admission. The research assistants obtained the social background information and DIS in individual interviews and supervised the self - administered paper- and pencil psychological assessment conducted in groups of 3 ±10 inductees. The feasibility of group assessments and their reliability and validity have been demonstrated in past research in TCs. 1.5. Sample characteristics Table 1 summarizes the basic demographic and drug use characteristics of the study sample. The full social and psychological profiles and DIS results are reported elsewhere (De Leon, 1988, 1993; Jainchill, 1994). Admissions were predominantly male (74.9%), African ± American (58.6%) and over 20 years of age (60.0%).
377
The most frequently reported primary drug was cocaine (57.2%), followed by heroin (16.6%) and marijuana (17.6%). Most (82.9%) entered treatment voluntarily and their psychological profiles (not shown here) revealed considerable psychological disturbance in addition to substance abuse. Intellectual and social functioning was low and signs of personality disorder, low self - esteem and marked levels of depression and anxiety were evident. Levels for motivation and readiness for treatment were moderately high. These profiles were similar for experimental and controls and across Trials 1 and 2, except for somewhat larger numbers of females and cocaine abusers in the second cohort. 1.6. Data analysis The SP intervention, the main independent variable, was measured in terms of presence or absence, while retention, as the main dependent variable, was measured as the proportion of residents retained. To assess the intervention's impact, a continuous measure of retention, such as total days in treatment, was considered less relevant than the proportion of dropout, especially as the major aim of the project was to reduce the actual proportion of people who leave treatment, particularly in the initial 30 days. Although not a focus of this study, longerterm retention, the proportion retained 180 and 365 days, was also examined.
Table 1 Sample characteristics, single admissions, SP intervention, and control groups (N = 605) All clients (N = 605)
SP (N = 362)
Controls (N = 243)
Characteristica
Number
Percent
Number
Percent
Number
Percent
Gender Male Female
453 152
74.9 25.1
266 96
73.5 26.5
187 56
77.0 23.0
Ethnicity African ± American Latino Caucasian Other
354 145 104 1
58.6 24.0 17.2 0.2
220 86 55 1
60.8 23.8 15.2 0.0
134 59 49 0
55.4 24.4 20.2 0.0
Primary drug Cocaine Opiates Marijuana Alcohol Other
306 89 94 22 24
57.2 16.6 17.6 4.1 4.5
189 50 49 10 4.1
60.4 16.0 15.7 3.2 4.8
117 39 45 12 9
52.7 17.6 20.3 5.4 4.1
Age at admission < 18 years 18 ± 20 years 21 ± 26 years 27 + years
93 123 165 159
17.2 22.8 30.6 29.4
35 71 95 90
15.5 22.6 30.3 28.7
58 52 70 69
18.5 23.0 30.5 30.5
a 70% or 11.5% of the cases were missing information about the client's primary drug, 65% or 10.7% were missing information on the client's age at entry to treatment and one case was missing the client's ethnic status.
378
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
For the bivariate analyses, Chi - square statistics evaluated comparisons between the experimental and control conditions. To further clarify the specific effects of the SP intervention, logistic regression analyses examined the relative contribution of client motivation and the SP intervention on dropout. Motivation was measured by the 18 -Item Circumstances, Motivation, and Readiness scale (CMR) (Melnick, 1999). Motivational levels in the sample were quadrasected into very low, low, high and very high groups, based upon scores at half an SD from the grand mean (see De Leon et al., 1994). The focus upon client motivation was based upon several reasons. Previous regression analyses of these data indicated that client factors (demographic, primary drug use, criminal background, legal status, and IQ) explained only a small amount of the variance of 30 -day retention (see De Leon, 1988). Motivation remained a large and consistent client predictor of retention in the present sample even when the contribution of the other client variables were removed (De Leon et al., 1994; Melnick et al., 1997). Second, the motivational factor has also been shown to be a consistent predictor of retention in methadone maintenance treatment programs and in other TC samples (Joe et al., 1998; Simpson & Joe, 1993). As noted above, theoretical considerations underscore the interaction between client factors such as motivation and treatment interventions in understanding retention. Thus, in the present study, logistic analyses focused on the relationship between motivation and the SP intervention. 2. Results There were no significant differences between the two trials with respect to client profiles or motivational levels and overall retention rates. However, the results for the separate and combined trials are reported in the bivariate and multivariate analyses.
2.1. Intervention effects Table 2 shows the retention rates at 30, 180 and 365 days for the experimental and control groups in the separate and combined trials. For the combined trials, the 30- day retention rates in the SP condition are significantly higher than in the control condition (p < 0.00). There were no significant differences between the two conditions at 180 or 365 days indicating that the SP intervention appears to decrease early dropout. For Trial 1, the retention difference between the experimental and control groups was significant at 30 ( p < 0.00) and at 365 days ( p < 0.02). For Trial 2, only the 30-day difference was significant ( p < 0.08). No differences were obtained at 180 or 365 days. Thus, the effects of the enhancement on short - term retention obtained in Trial 1 were replicated in Trial 2 on the new cohort a year later. 2.2. Client motivation and the SP intervention Blockwise regression equations were run for Trials 1 and 2 separately and for Trials 1 and 2 combined to test the impact of client motivation and SP intervention on short - term retention. In each analysis, the SP intervention was entered at step one, followed by the four CMR levels, and, finally, the interaction term(s) between motivation and the intervention. The control group was the reference group for the intervention analysis while clients in the highest quadrant of the four levels of motivation (greater than one standard deviation above the mean) were the reference group for the motivation analysis. Tables 3 and 4 summarize the results of the analyses for each trial and the combined trials. Trial 1 indicated that the three blocks, (1) SP intervention, (2) motivation levels, and (3) the interaction between motivation level and receiving the SP intervention, all significantly predicted 30 - day retention. Each block in the analysis significantly improved the model's overall prediction.
Table 2 Retention rates, single admissions, SP intervention, and control groups, Trials 1 and 2 SP group
Control group
SP vs. Control
N
Percentage
N
Percentage
Chi - square
Significance
Trial 1 ( N = 412) 30 - Day 180 - Day 365 - Day
182 115 83
76.2 48.1 34.7
105 70 41
60.7 40.5 23.7
11.35 2.38 5.80
0.00 0.12 0.02
Trial 2 ( N = 193) 30 - Day 180 - Day 365 - Day
95 52 37
77.2 42.3 30.1
46 30 25
65.7 42.9 35.7
3.01 0.01 0.65
0.08 0.94 0.42
277 167 120
76.5 46.1 33.1
151 100 66
62.1 41.2 35.5
14.53 1.46 2.45
0.00 0.23 0.12
Combined sample ( N = 605) 30 - Day 180 - Day 365 - Day
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
379
Table 3 The blockwise logistic analyses of the SP intervention, Trial 1, Trial 2, and combined samples compareda Chi - square
Significance of the block
Significance of the model
Percentage of predicted correct overall
Trial 1 sample ( N = 277) Block 1 Block 2 Block 3
9.31 14.34 12.12
0.002 0.002 0.007
9.31 23.66 35.78
0.002 0.000 0.000
Trial 2 sample ( N = 120) Block 1 Block 2 Block 3
3.59 6.89 1.35
0.058 0.076 0.717
3.59 10.48 11.83
0.058 0.033 0.106
13.80 17.55 9.62
0.000 0.001 0.022
13.80 31.35 40.98
0.000 0.000 0.000
Combined sample ( N = 397) Block 1 Block 2 Block 3 a
The differences in the sample sizes when compared to previous tables are due to missing data on the motivation measure.
The final model, which included all three terms, showed a striking interaction effect. The clients with the lowest motivation levels who received the SP intervention were 47.71 times more likely than the lowest motivated clients in the control group to stay 30 days in treatment. However, within each of the experimental and control groups, the clients with low motivation were less likely to stay in treatment for at least 30 days than highly motivated clients (Table 4). For Trial 2, the best overall prediction was obtained from the model without the interaction term. Achieving statistical significance for the variables in the Trial 2 analysis was tenuous due to the sample size (N = 120). Within each group, the lowest motivated clients were significantly more likely than the highest motivated clients to separate from treatment within 30 days. The effect of the SP intervention approached statistical significance when controlling for client motivation. In the combined sample, the best model, which included (1) the SP intervention, (2) motivation levels, and (3) the interaction between motivation level and receiving the SP intervention, significantly predicted 30 -day retention. Each block in the analysis significantly improved the model's overall prediction. The final model including all three terms showed that clients with the lowest motivation levels who received the SP intervention were 18.21 times more likely than the lowest motivated members in the control group to remain 30 days (Table 4). Again, controlling for the interaction effect between motivation and the SP intervention, low and moderately low motivated clients were less likely to stay 30 days in treatment compared to the highest motivated clients. Moreover, the comparative odds of staying in treatment showed a linear trend; the lower the client's motivation level, the less likely he /she was to stay 30 days in treatment.
Thus, for the experimental and control groups in both trials, the higher motivated clients were more likely to Table 4 The best models from the blockwise logistic analyses of the SP intervention, Trial 1, Trial 2, and combined samples compareda Trial 1 sample ( N = 277) SP intervention Motivation for treatment Low motivated clients Moderately low motivated clients High motivated clients Interaction term SP low motivated clients SP moderately low motivated clients SP high motivated clients Trial 2 sample ( N = 120) SP intervention Motivation for treatment Low motivated clients Moderately low motivated clients High motivated clients Interaction term SP low motivated clients SP moderately low motivated clients SP high motivated clients Combined sample ( N = 397) SP intervention Motivation for treatment Low motivated clients Moderately low motivated clients High motivated clients Interaction term SP low motivated clients SP moderately low motivated clients SP high motivated clients
Odds
Significance
0.26 ± 0.03 0.12 0.20 ± 47.71 5.20 7.02
0.237 0.000 0.001 0.054 0.143 0.021 0.005 0.181 0.113
2.59 ± 0.09 0.22 0.13 n.s.
0.057 0.172 0.043 0.177 0.063 n.s.
0.55 ± 0.04 0.18 0.24 ± 18.21 3.18 3.28
0.481 0.000 0.000 0.032 0.076 0.039 0.008 0.225 0.212
a The differences in the sample sizes when compared to previous tables are due to missing data on the motivation measure.
380
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
remain 30 days in treatment compared to the lower motivated groups. However, among the lowest motivated clients in the study sample, retention rates in the experimental group were dramatically higher than in the controls, reflecting the impact of the SP intervention. 2.3. Long -term retention The possible long -term effects of the interaction between motivation and the intervention were examined. For this analysis, long -term retention rates were based upon the clients who remained through the initial 30 days in both groups. This base of 30 -day ``survivors'' is appropriate since long -term retention differences could reflect the initial 30 - day differences between control and experimental groups rather than the time persistent effects of the early intervention. The exclusion of the 30- day dropouts in both conditions obviously reduced the sample size for the long term analysis. Thus, motivation was dichotomized into low (very low and low motivated clients combined) and high (very high and high motivated clients combined) for each of the experimental and controls. Among the low motivated clients who had participated in the SP intervention, 63.4% (N = 45) stayed in treatment for 180 days or more and 50.7% (N = 36) stayed a full year, compared to 57.4% (N = 27) and 34.0% (N = 16) for the low motivated control group clients, respectively. The differences between the low motivated clients in experimental and control conditions approached significance ( p < 0.074) at 1 year and suggests that the intervention impacting low motivated clients during the first 30 days of their residence persisted up to a year later. 3. Discussion The bivariate and multivariate results indicated that the SP intervention significantly enhanced 30-day retention among admissions entering a residential TC. The impact of the intervention was largely confined to clients initially having the lowest motivation for treatment. This interaction suggests that the effect of the SP on retention was mediated by increasing the motivational levels among the low motivated clients to complete the first 30 days of their residency. However, a direct test of this hypothesis will be explored in later studies, which will assess changes in motivation after the intervention in relation to 30 -day retention. The present findings are consistent with a theoretical formulation outlined in other writings (e.g., De Leon, 1995, 1996, 2000) that underscores the change process in the TC as a relationship between initial motivation, particularly when early in treatment, and retention. Motivated clients are more likely to participate in program activities, which results in the initial progress that sustains motivation for further participation and remaining in treatment. The SP
intervention appears to have enhanced initial motivation to remain in treatment 30 days. In addition, although the stability of the long -term effect remains to be validated with larger samples, the 1- year retention results based upon the 30 - day survivor analyses were consistent with this process formulation. Several issues and caveats render a cautious interpretation of the present findings. One issue concerns the possible impact of the number of staff in the experimental condition. The observed effects may reflect a general impact of additional staff present in the induction facility, rather than a specific influence of the SP intervention. However, this effect is unlikely since a ``SP'' represented the addition of one staff person for approximately two h on each of the seminar nights during the period of the intervention. Another issue is the need to specify the ``active ingredient'' of the SP intervention in this research. The SPs and the curricula delivered were obviously confounded components of the intervention. Additionally, possible differential effects across SPs could not be readily identified in part because of their small numbers (N = 10). Nevertheless, the faculty and 30 -day enhancement results remained the same across the two trials, indicating that any confounded effects were replicated across different cohorts of admissions. The number of seminar sessions that clients attended would have provided some evidence of a dose ± response effect. Though not reported, the number of SP sessions attended did significantly correlate with 30 -day completion rates. However, the interpretation of this variable was considered problematical. Since every resident was required to attend all community activities including the SP seminars, the variation in client attendance resulted mostly from 30 -day dropouts. In a few cases, sessions were missed due to illness and scheduling (e.g., a new admission entering the facility in the late evening of a seminar day). Therefore, the number of seminar sessions is necessarily time- correlated such that those who received fewer sessions were the early dropouts. The generality of these findings and conclusions remains to be demonstrated since the treatment agency and the staff faculty may be unique among residential TC programs. However, analogous conclusions concerning the role of staff in the treatment process have been drawn in recent studies in methadone treatment by Simpson et al. These investigations have identified the significance of the counselor ± client ``therapeutic alliance'' in facilitating retention early in treatment (e.g., Simpson et al., 1997). In addition, the present findings demonstrate effects in group settings that are also consistent with motivational enhancement research involving the counselor ± clinician in individual settings (e.g., Miller & Rollnick 1991). A specific issue of generality in the present study is that results are based on data collected in the mid - 1980s. Several considerations bear upon this concern. First, current admissions to TCs are not remarkably different than those in the mid -1980s, although the more recent
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
entries tend to be older, less involved in opioids and appear more socially dysfunctional. Second, while the planned duration of residential treatment in current TCs has been shortened and overall retention rates are somewhat elevated compared to earlier years, the problem of high early dropout persists. Third, as noted above, a clinical and research literature has developed on motivation (e.g., De Leon et al., 2000), enhancing treatment engagement (Simpson et al., 1995, 1997) and on the role of staff in the treatment process (Center for Therapeutic Community Research, 1998). Thus, though conducted on an earlier generation of TC clients, the present findings provide a sound empirical basis for replication in current treatment populations and settings. Notwithstanding the above caveats, the findings of the present research contain broader implications for staffing composition in TC treatment settings. Generally, the most experienced personnel are not usually in the front lines serving the novice client. Their roles are usually reserved for clinical teaching activities, supervision, monitoring, evaluating and consulting. Deploying the most experienced staff to address clients at perhaps the most vulnerable point in their treatment status could yield significant benefits.
Acknowledgments This study was supported in part through a grant from the National Institute on Drug Abuse (NIDA), Grant No. R01 DA03617.
References Barr, H., & Antes, P. (1981). Factors related to recovery and relapse in follow - up. Final report of project activities (Grant No. 1H81 DA01864). Rockville, MD: National Institute on Drug Abuse. Center for Therapeutic Community Research. (1998). Communications, (June, 2), 2. De Leon, G. (1984). Social and psychological adjustment of 400 dropouts and 100 graduates from the Phoenix House therapeutic community. In The therapeutic community: study of effectiveness ( National Institute on Drug Abuse Treatment Research Monograph Series, DHHS Publication No. [ADM] 85 - 1286). De Leon, G. (1988). The therapeutic community: enhancing retention in treatment. Final report of the National Institute on Drug Abuse (Grant No. R01 - DA3617). De Leon, G. (1993). Cocaine abusers in therapeutic community treatment. In F.M. Tims, & C.G. Leukefeld (Eds.), Cocaine treatment: research and clinical perspectives. NIDA Research Monograph 135, NIH Publication No. 93 - 3639 (pp. 163 ± 189). Superintendent of Documents, US Government Printing Office, Washington, DC. (1990). [paper presented] Advances in cocaine treatment. NIDA Technical Review Meeting, Bethesda, MD. De Leon, G. (1995). Therapeutic communities for addictions: a theoretical framework. The International Journal of the Addictions, 30 (12), 1603 ± 1645. De Leon, G. (1996). Integrative recovery: a stage paradigm. Substance Abuse, 17 (1), 51 ± 63.
381
De Leon, G. (2000). The therapeutic community: theory, model, and method. New York, NY: Springer. De Leon, G., & Jainchill, N. (1986). Circumstances, motivation, readiness and suitability as correlates of treatment tenure. Journal of Psychoactive Drugs, 18 (3), 203 ± 208. De Leon, G., Jainchill, N., & Wexler, H. (1982). Success and improvement rates 5 years after treatment in a therapeutic community. International Journal of Addictions, 17 (4), 703 ± 747. De Leon, G., Melnick, G., & Hawke, J. (2000). The motivation - readiness factor in drug treatment: implications for research and policy. Advances in Medical Sociology, 7, 103 ± 129. De Leon, G., Melnick, G., Kressel, D., & Jainchill, N. (1994). Circumstances, motivation, readiness, and suitability (CMRS) as correlates of treatment tenure. American Journal of Drug and Alcohol Abuse, 20 (4), 495 ± 515. De Leon, G., & Schwartz, S. (1984). The therapeutic community: what are the retention rates? American Journal of Drug and Alcohol Abuse, 10 (2), 267 ± 284. Glaser, F. B. (1974). Splitting: attrition from a drug - free therapeutic community. American Journal of Drug Alcohol Abuse, 1, 329 ± 348. Holland, S. (1983). Evaluating community based treatment programs: a model for strengthening inferences about effectiveness. International Journal of Therapeutic Communities, 4 (4), 285 ± 306. Hubbard, R. L., Craddock, S. G., Flynn, P. M., Anderson, J., & Etheridge, R. M. (1997). Overview of 1 - year follow - up outcomes in the drug abuse treatment outcome study (DATOS). Psychology of Addictive Behaviors, Special Issue: Drug Abuse Treatment Outcome Study (DATOS), 11 (4), 261 ± 278. Hubbard, R. L., Valley Rachal, J., Craddock, S. G., & Cavanaugh, E. R. (1984). Treatment outcome prospective study (TOPS): client characteristics and behaviors before, during, and after treatment. In F. M. Tims, & J. P. Ludford (Eds.), Drug abuse treatment evaluation: strategies, progress, and prospects. NIDA Research Monograph No. 51, Research Analysis and Utilization System Review Report, (pp. 42 ± 68). DHHS Publication No. (ADM) 84 - 1329. Rockville, MD: National Institute on Drug Abuse. Jainchill, N. (1994). Co-morbidity and therapeutic community treatment. In F. M. Tims, G. De Leon, & N. Jainchill (Eds.), Therapeutic community: advances in research and application. NIDA Research Monograph 144, NIH Publication No. 94-3633 (pp. 209 ± 231). Washington, DC: Superintendent of Documents, US Government Printing Office. Joe, G. W., & Simpson, D. D. (1976). Relationship of patient characteristics to tenure. In Sells, & Simpson (Eds.), Studies in effectiveness of treatment of drug abuse. Cambridge, MA: Ballinger. Joe, G. W., Simpson, D. D., & Broome, K. M. (1998). Effects of readiness for drug abuse treatment on client retention and assessment of process. Addiction, 93 (8), 1177 ± 1190. Melnick G. (1999). Assessing treatment readiness in special populations. Final report of project activities. National Institute on Drug Abuse (Grant No. RO1 DA07377). Rockville, MD: National Institute on Drug Abuse. Melnick, G., De Leon, G., Hawke, J., Jainchill, J., & Kressel, D. (1997). Motivation and readiness for therapeutic community treatment among adolescent and adult substance abusers. American Journal of Drug and Alcohol Abuse, 23 (4), 485 ± 506. Miller, W. R., & Rollnick, S. (1991). Motivational interviewing: preparing people to change addictive behavior. New York: The Guilford Press. Robins, L., Helzer, J., Cottler, L., & Goldring, M. A. T. (1989). NIMH diagnostic interview schedule, version 111 revised. Rockville, MD: National Institute on Drug Abuse (Monograph). Sansone, J. (1980). Retention patterns in a therapeutic community for the treatment of drug abuse. International Journal of the Addictions, 15, 711 ± 736. Simpson, D. D., & Joe, G. W. (1993). Motivation as a predictor of early dropout from drug abuse treatment. Psychotherapy, 30, 357 ± 368. Simpson, D. D., Joe, G. W., & Rowan - Szal, G. A. (1997). Drug abuse treatment retention and process effects on follow - up outcomes. Drug and Alcohol Dependence, 47, 227 ± 235.
382
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
Simpson, D. D., Joe, G. W., Rowan - Szal, G., & Greener, J. (1995). Client engagement and change during drug abuse treatment. Journal of Substance Abuse, 7, 117 ± 134. Simpson, D. D., Joe, G. W., Rowan - Szal, G. A., & Greener, J. M. (1997). Drug abuse treatment process components that improve retention. Journal of Substance Abuse Treatment, 14 (6), 565 ± 572. Simpson, D. D., & Sells, S. (1982). Effectiveness of treatment for drug
abuse: an overview of the DARP research program. Advances in Alcohol Substance Abuse, 2 (1), 7 ± 29. Staines, G. L., McKendrick, K., Perlis, T., Sacks, S., & De Leon, G. (1999). Sequential assignment and treatment - as - usual: alternatives to standard experimental designs in field studies of treatment efficacy. Evaluation Review, 23, (1).
Article
Therapeutic communities Enhancing retention in treatment using ``Senior Professor'' staff George De Leon*, Josephine Hawke, Nancy Jainchill, Gerald Melnick Center for Therapeutic Community Research (CTCR) at National Development and Research Institutes, Inc. (NDRI), 16th Floor, 2 World Trade Center, New York, NY 10048, USA Received 23 December 1998; received in revised form 27 March 2000; accepted 4 April 2000
Abstract Evaluation research documents a firm relationship between retention and treatment outcomes among substance abusers in therapeutic communities (TCs). However, most admissions leave treatment prematurely, particularly in the first months after admission. This paper reports findings from a controlled study that assessed the efficacy of an intervention to reduce early dropout in a residential TC. In the ``Senior Professor'' (SP) intervention, the most experienced clinical and managerial staff in a TC program were utilized to conduct program induction seminars during the first weeks of admissions, traditionally the period of the highest rate of dropout. Rates of short - term retention (30 days of treatment) were compared for the experimental cohort (N = 362), who received the intervention, and a cohort of admissions (N = 243), who received standard or non - enhanced treatment. Findings showed that the SP intervention significantly reduced the likelihood of early dropout compared with controls. The enhanced effects are most evident for the new inductees with the lowest levels of motivation. Some theoretical and clinical implications are discussed as to the utilization of experienced staff to increase retention among new inductees with relatively lower motivational levels, who are at greatest risk for early dropout. D 2001 Elsevier Science Inc. All rights reserved. Keywords: TC staff; Early dropout; Motivation; Treatment process; TC theory
Considerable research shows that among clients who enter therapeutic communities (TCs), successful outcome is directly related to the length of stay in treatment (e.g., Barr & Antes, 1981; De Leon, 1984; De Leon et al., 1982; Holland, 1983; Hubbard et al., 1984, 1997; Simpson et al., 1997; Simpson & Sells, 1982). However, most admissions to TCs leave treatment before beneficial effects are rendered. Dropout is highest within the first 30 to 60 days of admission. Thereafter, it declines rapidly, leveling off at a low constant rate (e.g., De Leon & Schwartz, 1984). The temporal pattern of retention is characteristic for long - term residential TCs (and for drug treatment programs in general), indicating that the likelihood of remaining and completing treatment increases with time in the program itself (e.g., De Leon & Schwartz, 1984; Glaser, 1974; Hubbard et al., 1984; Joe & Simpson, 1976; Sansone, 1980). Therefore, to expand the beneficial effects of TC
* Corresponding author. Tel.: +1 - 212 - 845 - 4417; fax: +1 - 212 - 845 4698. E-mail address: [email protected] (G. De Leon).
treatment, efforts are needed to reduce the likelihood of dropout, especially during the early months of residency. As part of an ongoing program of research at the Center for Therapeutic Community Research (CTCR), a series of initial studies have explored the influences of staff factors on client retention, progress in treatment and treatment outcomes. This paper reports findings from a controlled study which assessed an intervention utilizing senior staff or ``Senior Professors'' (SPs) to enhance retention in a long term residential TC. The theoretical rationale that guided the present research views retention as reflecting interactions among three domains of variables: (1) external influences, (2) certain client characteristics and (3) the various activities and elements of the TC regimen, which are potential interventions in the treatment process. The diversity of clients with respect to their motivation and readiness to change and their suitability for the TC are particularly important in the early phases of treatment. Because relatively homogeneous program procedures are imposed upon this diversity, the process of inducting clients into their new roles as TC residents is extremely difficult.
0740-5472/00/$ ± see front matter D 2001 Elsevier Science Inc. All rights reserved. PII: S 0 7 4 0 - 5 4 7 2 ( 0 0 ) 0 0 1 2 4 - 0
376
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
The problem of role induction is further compounded by staffing patterns and functions in TCs. Traditionally, the responsibility for inducting new members into residential treatment programs belongs to relatively junior staff. This staffing pattern is based on the assumption that new admissions will be most successfully recruited into the TC regimen by staff and resident peers with whom the recruit can most readily identify, namely, recovering or recently recovered substance abusers. However, junior staff seldom has clinical experience to handle the diversity among new admissions and the problems that they experience during initiation. Therefore, it is understandable that the dropout rate is highest in the first days of treatment. Conversely, the underlying assumption of the SP intervention is that short- term retention can be enhanced if program procedures and staffing are modified to facilitate the management of diversity among new inductees. In other words, new inductees are exposed to senior staff who has the most clinical experience during the initial phase of treatment. At the same time, contact with junior staff with whom they can most readily identify and who provides positive role models for recovery is maintained. The present research utilized senior staff to facilitate the role induction of new admissions into the program. The most experienced clinical and managerial personnel in a TC program delivered a series of special seminars to first-time admissions during the initial weeks following their entry into long - term residence. The goal of this intervention was to increase the likelihood that new admissions to the residential TC would complete the initial induction stage of treatment. 1. Method 1.1. The SP intervention A faculty of senior staff or SPs was assembled from various departments within the TC agency to conduct seminars for new admissions to residential treatment. The SP seminars were carried out in a residential induction facility, which served as an assessment and orientation center for all new admissions. After 30 days in the induction facility, residents were transferred to the agency's long -term residential facilities. All residents in the induction facility convened at the usual time of the day scheduled by the program for seminars. Three SP seminars were delivered weekly by different SPs to all facility residents in a classroom setting. Each seminar lasted approximately one and a half hours. Seminar themes, though varying, all reflected the TC perspective and approach to treatment and recovery (e.g., De Leon, 1995, 2000). SPs selected their topic from a list of relevant TC themes (e.g., self -esteem, self -help and recovery concepts, the TC philosophy and its expectations, TC
retention and success rates). All SP seminars reflected problems of staying in treatment, and dealt with a wide range of concerns and questions that are often expressed by new TC members (e.g., fluctuating motivation, outside pressures and attractions, the structure and demands of the program, ambivalence about long - term residence, next steps and success rates). The seminar format followed the regular format of seminars as they are usually delivered in TCs. The format excluded formal protocols. It was didactic and experiential with considerable interaction between the seminar leader (the SP) and the resident participants. Thus, SPs used their experience, knowledge and charisma to address the concerns, questions and issues that mark the early days of the induction process. ( Note: Details of the SP procedure and seminar topics can be obtained from the first author.) 1.2. Research design Two trials of the SP intervention were conducted on separate cohorts admitted a year apart, Trial 1 ( November 20, 1984 through January 30, 1985) and Trial 2 ( February 1, 1986 through March 31, 1986). The SP staff and curriculum were the same in both trials. However, the shorter time frame in Trial 2 resulted in a different maximum number of sessions that could be attended during a 30 -day residential stay (N = 16 in Trial 1; N = 11 in Trial 2). The experimental design was a variant of time series or an ``On ± Off'' sequence. Each trial was defined as the 8 - to 10 -week period of the intervention, which was preceded by 8 to 10 weeks of the standard induction period. Retention rates between the experimental condition (the ``On'' period, which consisted of the standard induction procedures plus the SP intervention) were compared to the preceding control condition (the ``Off'' period, which consisted of the standard induction procedures only). The utility of the ``On ± Off'' design relative to matched controls and random assignment alternatives in field settings is discussed elsewhere (i.e., Staines et al., 1999). 1.3. Study sample The study site was a large northeastern drug treatment agency that was representative of traditional long - term residential TCs. The study sample (N = 605) consisted of all consecutive admissions to the induction facility during the two experimental trials of the project. The sample size of the combined trials was N = 362 admissions (Trial 1, N = 239; Trial 2, N = 123) for the experimental condition and N = 243 admissions (Trial 1, N = 173; Trial 2, N = 70) for the control condition. 1.4. Data collection Comprehensive social and psychological data were collected on all consecutive admissions to the residential
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
facility during all months of the study. Drug use, criminality, employment and other background information were obtained, as well as an extensive psychological battery that included the Beta IQ, the Tennessee Self -Concept Scale, the MMPI, the Beck Depression Inventory, the Manifest Anxiety Scale, the Socialization Scale of the California Personality Inventory, the Diagnostic Interview Schedule (DIS) (Robins et al., 1989) and an 18 - item version of the Circumstances, Motivation, Readiness, and Suitability (CMR) scales (De Leon & Jainchill, 1986; De Leon et al., 1994) to examine readiness for TC treatment. The data collection required approximately 5 h across 3 days of admission. The research assistants obtained the social background information and DIS in individual interviews and supervised the self - administered paper- and pencil psychological assessment conducted in groups of 3 ±10 inductees. The feasibility of group assessments and their reliability and validity have been demonstrated in past research in TCs. 1.5. Sample characteristics Table 1 summarizes the basic demographic and drug use characteristics of the study sample. The full social and psychological profiles and DIS results are reported elsewhere (De Leon, 1988, 1993; Jainchill, 1994). Admissions were predominantly male (74.9%), African ± American (58.6%) and over 20 years of age (60.0%).
377
The most frequently reported primary drug was cocaine (57.2%), followed by heroin (16.6%) and marijuana (17.6%). Most (82.9%) entered treatment voluntarily and their psychological profiles (not shown here) revealed considerable psychological disturbance in addition to substance abuse. Intellectual and social functioning was low and signs of personality disorder, low self - esteem and marked levels of depression and anxiety were evident. Levels for motivation and readiness for treatment were moderately high. These profiles were similar for experimental and controls and across Trials 1 and 2, except for somewhat larger numbers of females and cocaine abusers in the second cohort. 1.6. Data analysis The SP intervention, the main independent variable, was measured in terms of presence or absence, while retention, as the main dependent variable, was measured as the proportion of residents retained. To assess the intervention's impact, a continuous measure of retention, such as total days in treatment, was considered less relevant than the proportion of dropout, especially as the major aim of the project was to reduce the actual proportion of people who leave treatment, particularly in the initial 30 days. Although not a focus of this study, longerterm retention, the proportion retained 180 and 365 days, was also examined.
Table 1 Sample characteristics, single admissions, SP intervention, and control groups (N = 605) All clients (N = 605)
SP (N = 362)
Controls (N = 243)
Characteristica
Number
Percent
Number
Percent
Number
Percent
Gender Male Female
453 152
74.9 25.1
266 96
73.5 26.5
187 56
77.0 23.0
Ethnicity African ± American Latino Caucasian Other
354 145 104 1
58.6 24.0 17.2 0.2
220 86 55 1
60.8 23.8 15.2 0.0
134 59 49 0
55.4 24.4 20.2 0.0
Primary drug Cocaine Opiates Marijuana Alcohol Other
306 89 94 22 24
57.2 16.6 17.6 4.1 4.5
189 50 49 10 4.1
60.4 16.0 15.7 3.2 4.8
117 39 45 12 9
52.7 17.6 20.3 5.4 4.1
Age at admission < 18 years 18 ± 20 years 21 ± 26 years 27 + years
93 123 165 159
17.2 22.8 30.6 29.4
35 71 95 90
15.5 22.6 30.3 28.7
58 52 70 69
18.5 23.0 30.5 30.5
a 70% or 11.5% of the cases were missing information about the client's primary drug, 65% or 10.7% were missing information on the client's age at entry to treatment and one case was missing the client's ethnic status.
378
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
For the bivariate analyses, Chi - square statistics evaluated comparisons between the experimental and control conditions. To further clarify the specific effects of the SP intervention, logistic regression analyses examined the relative contribution of client motivation and the SP intervention on dropout. Motivation was measured by the 18 -Item Circumstances, Motivation, and Readiness scale (CMR) (Melnick, 1999). Motivational levels in the sample were quadrasected into very low, low, high and very high groups, based upon scores at half an SD from the grand mean (see De Leon et al., 1994). The focus upon client motivation was based upon several reasons. Previous regression analyses of these data indicated that client factors (demographic, primary drug use, criminal background, legal status, and IQ) explained only a small amount of the variance of 30 -day retention (see De Leon, 1988). Motivation remained a large and consistent client predictor of retention in the present sample even when the contribution of the other client variables were removed (De Leon et al., 1994; Melnick et al., 1997). Second, the motivational factor has also been shown to be a consistent predictor of retention in methadone maintenance treatment programs and in other TC samples (Joe et al., 1998; Simpson & Joe, 1993). As noted above, theoretical considerations underscore the interaction between client factors such as motivation and treatment interventions in understanding retention. Thus, in the present study, logistic analyses focused on the relationship between motivation and the SP intervention. 2. Results There were no significant differences between the two trials with respect to client profiles or motivational levels and overall retention rates. However, the results for the separate and combined trials are reported in the bivariate and multivariate analyses.
2.1. Intervention effects Table 2 shows the retention rates at 30, 180 and 365 days for the experimental and control groups in the separate and combined trials. For the combined trials, the 30- day retention rates in the SP condition are significantly higher than in the control condition (p < 0.00). There were no significant differences between the two conditions at 180 or 365 days indicating that the SP intervention appears to decrease early dropout. For Trial 1, the retention difference between the experimental and control groups was significant at 30 ( p < 0.00) and at 365 days ( p < 0.02). For Trial 2, only the 30-day difference was significant ( p < 0.08). No differences were obtained at 180 or 365 days. Thus, the effects of the enhancement on short - term retention obtained in Trial 1 were replicated in Trial 2 on the new cohort a year later. 2.2. Client motivation and the SP intervention Blockwise regression equations were run for Trials 1 and 2 separately and for Trials 1 and 2 combined to test the impact of client motivation and SP intervention on short - term retention. In each analysis, the SP intervention was entered at step one, followed by the four CMR levels, and, finally, the interaction term(s) between motivation and the intervention. The control group was the reference group for the intervention analysis while clients in the highest quadrant of the four levels of motivation (greater than one standard deviation above the mean) were the reference group for the motivation analysis. Tables 3 and 4 summarize the results of the analyses for each trial and the combined trials. Trial 1 indicated that the three blocks, (1) SP intervention, (2) motivation levels, and (3) the interaction between motivation level and receiving the SP intervention, all significantly predicted 30 - day retention. Each block in the analysis significantly improved the model's overall prediction.
Table 2 Retention rates, single admissions, SP intervention, and control groups, Trials 1 and 2 SP group
Control group
SP vs. Control
N
Percentage
N
Percentage
Chi - square
Significance
Trial 1 ( N = 412) 30 - Day 180 - Day 365 - Day
182 115 83
76.2 48.1 34.7
105 70 41
60.7 40.5 23.7
11.35 2.38 5.80
0.00 0.12 0.02
Trial 2 ( N = 193) 30 - Day 180 - Day 365 - Day
95 52 37
77.2 42.3 30.1
46 30 25
65.7 42.9 35.7
3.01 0.01 0.65
0.08 0.94 0.42
277 167 120
76.5 46.1 33.1
151 100 66
62.1 41.2 35.5
14.53 1.46 2.45
0.00 0.23 0.12
Combined sample ( N = 605) 30 - Day 180 - Day 365 - Day
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
379
Table 3 The blockwise logistic analyses of the SP intervention, Trial 1, Trial 2, and combined samples compareda Chi - square
Significance of the block
Significance of the model
Percentage of predicted correct overall
Trial 1 sample ( N = 277) Block 1 Block 2 Block 3
9.31 14.34 12.12
0.002 0.002 0.007
9.31 23.66 35.78
0.002 0.000 0.000
Trial 2 sample ( N = 120) Block 1 Block 2 Block 3
3.59 6.89 1.35
0.058 0.076 0.717
3.59 10.48 11.83
0.058 0.033 0.106
13.80 17.55 9.62
0.000 0.001 0.022
13.80 31.35 40.98
0.000 0.000 0.000
Combined sample ( N = 397) Block 1 Block 2 Block 3 a
The differences in the sample sizes when compared to previous tables are due to missing data on the motivation measure.
The final model, which included all three terms, showed a striking interaction effect. The clients with the lowest motivation levels who received the SP intervention were 47.71 times more likely than the lowest motivated clients in the control group to stay 30 days in treatment. However, within each of the experimental and control groups, the clients with low motivation were less likely to stay in treatment for at least 30 days than highly motivated clients (Table 4). For Trial 2, the best overall prediction was obtained from the model without the interaction term. Achieving statistical significance for the variables in the Trial 2 analysis was tenuous due to the sample size (N = 120). Within each group, the lowest motivated clients were significantly more likely than the highest motivated clients to separate from treatment within 30 days. The effect of the SP intervention approached statistical significance when controlling for client motivation. In the combined sample, the best model, which included (1) the SP intervention, (2) motivation levels, and (3) the interaction between motivation level and receiving the SP intervention, significantly predicted 30 -day retention. Each block in the analysis significantly improved the model's overall prediction. The final model including all three terms showed that clients with the lowest motivation levels who received the SP intervention were 18.21 times more likely than the lowest motivated members in the control group to remain 30 days (Table 4). Again, controlling for the interaction effect between motivation and the SP intervention, low and moderately low motivated clients were less likely to stay 30 days in treatment compared to the highest motivated clients. Moreover, the comparative odds of staying in treatment showed a linear trend; the lower the client's motivation level, the less likely he /she was to stay 30 days in treatment.
Thus, for the experimental and control groups in both trials, the higher motivated clients were more likely to Table 4 The best models from the blockwise logistic analyses of the SP intervention, Trial 1, Trial 2, and combined samples compareda Trial 1 sample ( N = 277) SP intervention Motivation for treatment Low motivated clients Moderately low motivated clients High motivated clients Interaction term SP low motivated clients SP moderately low motivated clients SP high motivated clients Trial 2 sample ( N = 120) SP intervention Motivation for treatment Low motivated clients Moderately low motivated clients High motivated clients Interaction term SP low motivated clients SP moderately low motivated clients SP high motivated clients Combined sample ( N = 397) SP intervention Motivation for treatment Low motivated clients Moderately low motivated clients High motivated clients Interaction term SP low motivated clients SP moderately low motivated clients SP high motivated clients
Odds
Significance
0.26 ± 0.03 0.12 0.20 ± 47.71 5.20 7.02
0.237 0.000 0.001 0.054 0.143 0.021 0.005 0.181 0.113
2.59 ± 0.09 0.22 0.13 n.s.
0.057 0.172 0.043 0.177 0.063 n.s.
0.55 ± 0.04 0.18 0.24 ± 18.21 3.18 3.28
0.481 0.000 0.000 0.032 0.076 0.039 0.008 0.225 0.212
a The differences in the sample sizes when compared to previous tables are due to missing data on the motivation measure.
380
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
remain 30 days in treatment compared to the lower motivated groups. However, among the lowest motivated clients in the study sample, retention rates in the experimental group were dramatically higher than in the controls, reflecting the impact of the SP intervention. 2.3. Long -term retention The possible long -term effects of the interaction between motivation and the intervention were examined. For this analysis, long -term retention rates were based upon the clients who remained through the initial 30 days in both groups. This base of 30 -day ``survivors'' is appropriate since long -term retention differences could reflect the initial 30 - day differences between control and experimental groups rather than the time persistent effects of the early intervention. The exclusion of the 30- day dropouts in both conditions obviously reduced the sample size for the long term analysis. Thus, motivation was dichotomized into low (very low and low motivated clients combined) and high (very high and high motivated clients combined) for each of the experimental and controls. Among the low motivated clients who had participated in the SP intervention, 63.4% (N = 45) stayed in treatment for 180 days or more and 50.7% (N = 36) stayed a full year, compared to 57.4% (N = 27) and 34.0% (N = 16) for the low motivated control group clients, respectively. The differences between the low motivated clients in experimental and control conditions approached significance ( p < 0.074) at 1 year and suggests that the intervention impacting low motivated clients during the first 30 days of their residence persisted up to a year later. 3. Discussion The bivariate and multivariate results indicated that the SP intervention significantly enhanced 30-day retention among admissions entering a residential TC. The impact of the intervention was largely confined to clients initially having the lowest motivation for treatment. This interaction suggests that the effect of the SP on retention was mediated by increasing the motivational levels among the low motivated clients to complete the first 30 days of their residency. However, a direct test of this hypothesis will be explored in later studies, which will assess changes in motivation after the intervention in relation to 30 -day retention. The present findings are consistent with a theoretical formulation outlined in other writings (e.g., De Leon, 1995, 1996, 2000) that underscores the change process in the TC as a relationship between initial motivation, particularly when early in treatment, and retention. Motivated clients are more likely to participate in program activities, which results in the initial progress that sustains motivation for further participation and remaining in treatment. The SP
intervention appears to have enhanced initial motivation to remain in treatment 30 days. In addition, although the stability of the long -term effect remains to be validated with larger samples, the 1- year retention results based upon the 30 - day survivor analyses were consistent with this process formulation. Several issues and caveats render a cautious interpretation of the present findings. One issue concerns the possible impact of the number of staff in the experimental condition. The observed effects may reflect a general impact of additional staff present in the induction facility, rather than a specific influence of the SP intervention. However, this effect is unlikely since a ``SP'' represented the addition of one staff person for approximately two h on each of the seminar nights during the period of the intervention. Another issue is the need to specify the ``active ingredient'' of the SP intervention in this research. The SPs and the curricula delivered were obviously confounded components of the intervention. Additionally, possible differential effects across SPs could not be readily identified in part because of their small numbers (N = 10). Nevertheless, the faculty and 30 -day enhancement results remained the same across the two trials, indicating that any confounded effects were replicated across different cohorts of admissions. The number of seminar sessions that clients attended would have provided some evidence of a dose ± response effect. Though not reported, the number of SP sessions attended did significantly correlate with 30 -day completion rates. However, the interpretation of this variable was considered problematical. Since every resident was required to attend all community activities including the SP seminars, the variation in client attendance resulted mostly from 30 -day dropouts. In a few cases, sessions were missed due to illness and scheduling (e.g., a new admission entering the facility in the late evening of a seminar day). Therefore, the number of seminar sessions is necessarily time- correlated such that those who received fewer sessions were the early dropouts. The generality of these findings and conclusions remains to be demonstrated since the treatment agency and the staff faculty may be unique among residential TC programs. However, analogous conclusions concerning the role of staff in the treatment process have been drawn in recent studies in methadone treatment by Simpson et al. These investigations have identified the significance of the counselor ± client ``therapeutic alliance'' in facilitating retention early in treatment (e.g., Simpson et al., 1997). In addition, the present findings demonstrate effects in group settings that are also consistent with motivational enhancement research involving the counselor ± clinician in individual settings (e.g., Miller & Rollnick 1991). A specific issue of generality in the present study is that results are based on data collected in the mid - 1980s. Several considerations bear upon this concern. First, current admissions to TCs are not remarkably different than those in the mid -1980s, although the more recent
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
entries tend to be older, less involved in opioids and appear more socially dysfunctional. Second, while the planned duration of residential treatment in current TCs has been shortened and overall retention rates are somewhat elevated compared to earlier years, the problem of high early dropout persists. Third, as noted above, a clinical and research literature has developed on motivation (e.g., De Leon et al., 2000), enhancing treatment engagement (Simpson et al., 1995, 1997) and on the role of staff in the treatment process (Center for Therapeutic Community Research, 1998). Thus, though conducted on an earlier generation of TC clients, the present findings provide a sound empirical basis for replication in current treatment populations and settings. Notwithstanding the above caveats, the findings of the present research contain broader implications for staffing composition in TC treatment settings. Generally, the most experienced personnel are not usually in the front lines serving the novice client. Their roles are usually reserved for clinical teaching activities, supervision, monitoring, evaluating and consulting. Deploying the most experienced staff to address clients at perhaps the most vulnerable point in their treatment status could yield significant benefits.
Acknowledgments This study was supported in part through a grant from the National Institute on Drug Abuse (NIDA), Grant No. R01 DA03617.
References Barr, H., & Antes, P. (1981). Factors related to recovery and relapse in follow - up. Final report of project activities (Grant No. 1H81 DA01864). Rockville, MD: National Institute on Drug Abuse. Center for Therapeutic Community Research. (1998). Communications, (June, 2), 2. De Leon, G. (1984). Social and psychological adjustment of 400 dropouts and 100 graduates from the Phoenix House therapeutic community. In The therapeutic community: study of effectiveness ( National Institute on Drug Abuse Treatment Research Monograph Series, DHHS Publication No. [ADM] 85 - 1286). De Leon, G. (1988). The therapeutic community: enhancing retention in treatment. Final report of the National Institute on Drug Abuse (Grant No. R01 - DA3617). De Leon, G. (1993). Cocaine abusers in therapeutic community treatment. In F.M. Tims, & C.G. Leukefeld (Eds.), Cocaine treatment: research and clinical perspectives. NIDA Research Monograph 135, NIH Publication No. 93 - 3639 (pp. 163 ± 189). Superintendent of Documents, US Government Printing Office, Washington, DC. (1990). [paper presented] Advances in cocaine treatment. NIDA Technical Review Meeting, Bethesda, MD. De Leon, G. (1995). Therapeutic communities for addictions: a theoretical framework. The International Journal of the Addictions, 30 (12), 1603 ± 1645. De Leon, G. (1996). Integrative recovery: a stage paradigm. Substance Abuse, 17 (1), 51 ± 63.
381
De Leon, G. (2000). The therapeutic community: theory, model, and method. New York, NY: Springer. De Leon, G., & Jainchill, N. (1986). Circumstances, motivation, readiness and suitability as correlates of treatment tenure. Journal of Psychoactive Drugs, 18 (3), 203 ± 208. De Leon, G., Jainchill, N., & Wexler, H. (1982). Success and improvement rates 5 years after treatment in a therapeutic community. International Journal of Addictions, 17 (4), 703 ± 747. De Leon, G., Melnick, G., & Hawke, J. (2000). The motivation - readiness factor in drug treatment: implications for research and policy. Advances in Medical Sociology, 7, 103 ± 129. De Leon, G., Melnick, G., Kressel, D., & Jainchill, N. (1994). Circumstances, motivation, readiness, and suitability (CMRS) as correlates of treatment tenure. American Journal of Drug and Alcohol Abuse, 20 (4), 495 ± 515. De Leon, G., & Schwartz, S. (1984). The therapeutic community: what are the retention rates? American Journal of Drug and Alcohol Abuse, 10 (2), 267 ± 284. Glaser, F. B. (1974). Splitting: attrition from a drug - free therapeutic community. American Journal of Drug Alcohol Abuse, 1, 329 ± 348. Holland, S. (1983). Evaluating community based treatment programs: a model for strengthening inferences about effectiveness. International Journal of Therapeutic Communities, 4 (4), 285 ± 306. Hubbard, R. L., Craddock, S. G., Flynn, P. M., Anderson, J., & Etheridge, R. M. (1997). Overview of 1 - year follow - up outcomes in the drug abuse treatment outcome study (DATOS). Psychology of Addictive Behaviors, Special Issue: Drug Abuse Treatment Outcome Study (DATOS), 11 (4), 261 ± 278. Hubbard, R. L., Valley Rachal, J., Craddock, S. G., & Cavanaugh, E. R. (1984). Treatment outcome prospective study (TOPS): client characteristics and behaviors before, during, and after treatment. In F. M. Tims, & J. P. Ludford (Eds.), Drug abuse treatment evaluation: strategies, progress, and prospects. NIDA Research Monograph No. 51, Research Analysis and Utilization System Review Report, (pp. 42 ± 68). DHHS Publication No. (ADM) 84 - 1329. Rockville, MD: National Institute on Drug Abuse. Jainchill, N. (1994). Co-morbidity and therapeutic community treatment. In F. M. Tims, G. De Leon, & N. Jainchill (Eds.), Therapeutic community: advances in research and application. NIDA Research Monograph 144, NIH Publication No. 94-3633 (pp. 209 ± 231). Washington, DC: Superintendent of Documents, US Government Printing Office. Joe, G. W., & Simpson, D. D. (1976). Relationship of patient characteristics to tenure. In Sells, & Simpson (Eds.), Studies in effectiveness of treatment of drug abuse. Cambridge, MA: Ballinger. Joe, G. W., Simpson, D. D., & Broome, K. M. (1998). Effects of readiness for drug abuse treatment on client retention and assessment of process. Addiction, 93 (8), 1177 ± 1190. Melnick G. (1999). Assessing treatment readiness in special populations. Final report of project activities. National Institute on Drug Abuse (Grant No. RO1 DA07377). Rockville, MD: National Institute on Drug Abuse. Melnick, G., De Leon, G., Hawke, J., Jainchill, J., & Kressel, D. (1997). Motivation and readiness for therapeutic community treatment among adolescent and adult substance abusers. American Journal of Drug and Alcohol Abuse, 23 (4), 485 ± 506. Miller, W. R., & Rollnick, S. (1991). Motivational interviewing: preparing people to change addictive behavior. New York: The Guilford Press. Robins, L., Helzer, J., Cottler, L., & Goldring, M. A. T. (1989). NIMH diagnostic interview schedule, version 111 revised. Rockville, MD: National Institute on Drug Abuse (Monograph). Sansone, J. (1980). Retention patterns in a therapeutic community for the treatment of drug abuse. International Journal of the Addictions, 15, 711 ± 736. Simpson, D. D., & Joe, G. W. (1993). Motivation as a predictor of early dropout from drug abuse treatment. Psychotherapy, 30, 357 ± 368. Simpson, D. D., Joe, G. W., & Rowan - Szal, G. A. (1997). Drug abuse treatment retention and process effects on follow - up outcomes. Drug and Alcohol Dependence, 47, 227 ± 235.
382
G. De Leon et al. / Journal of Substance Abuse Treatment 19 (2000) 375± 382
Simpson, D. D., Joe, G. W., Rowan - Szal, G., & Greener, J. (1995). Client engagement and change during drug abuse treatment. Journal of Substance Abuse, 7, 117 ± 134. Simpson, D. D., Joe, G. W., Rowan - Szal, G. A., & Greener, J. M. (1997). Drug abuse treatment process components that improve retention. Journal of Substance Abuse Treatment, 14 (6), 565 ± 572. Simpson, D. D., & Sells, S. (1982). Effectiveness of treatment for drug
abuse: an overview of the DARP research program. Advances in Alcohol Substance Abuse, 2 (1), 7 ± 29. Staines, G. L., McKendrick, K., Perlis, T., Sacks, S., & De Leon, G. (1999). Sequential assignment and treatment - as - usual: alternatives to standard experimental designs in field studies of treatment efficacy. Evaluation Review, 23, (1).