![]()
Predictors of Engagement in Court-Mandated Treatment: Findings at the Brooklyn Treatment Court, 1996ā2000
MICHAEL REMPEL
Center for Court Innovation
CHRISTINE DEPIES DESTEFANO
Urban Institute, Justice Policy Center
ABSTRACT Past research indicates that more time in treatment yields better post-treatment outcomes, with 90 days of treatment often identified as a minimum threshold for achieving positive results, thus making it important to identify factors that predict meaningful engagement in treatment and to devise policies to assist subgroups facing a high risk of dropping out. Although a literature currently exists on voluntary treatment programs, fewer studies examine dynamics specific to court-mandated programs such as drug courts. Those programs use legal coercion applied via the threat of incarceration, coupled with ongoing court supervision, to motivate participants to succeed. Results were analyzed at the Brooklyn Treatment Court. Analyses looked at retention for at least 90 days of treatment, and engagement, defined as completing four consecutive months of drug-free and sanction-less participation. Multivariate analyses revealed that the level of
legal coercion, measured by expected incarceration time in the event of program failure, strongly predicted both retention and engagement. Also predictive was the legal/emotional coercion faced by participants who indicated
at intake that they had a pending Family Court case whose outcome might hinge on the drug court outcome. Participation during the 30-day period immediately following program entry was as important as coercion. Warranting or failing to begin treatment within 30 days of formal entry strongly predicted dropping out. The following additional characteristics predicted dropping out: younger, primary drug of heroin, prior misdemeanor conviction(s), and residence in a neighborhood characterized by greater
social isolation. Policy implications of all findings are discussed.
[Article copies available for a fee from The Haworth Document Delivery Service: 1-800-HAWORTH. E-mail address: <[email protected]> Website: <http://www.HaworthPress.com> Ā© 2001 by The Haworth Press, Inc. All rights reserved.] KEYWORDS Drug court, ethnicity, legal coercion, race, recidivism, social isolation, socioeconomic status, substance abuse history, treatment duration, treatment engagement
The substance abuse treatment literature consistently links more time in treatment to more favorable outcomes on measures such as drug use, criminal activity, and employment (Anglin et al., 1989; Collins and Allison, 1983; DeLeon, 1988(a); DeLeon, 1988(b); Hubbard et al., 1989; Lawental et al., 1996; Peters and Murrin, 1998; Siddall and Conway, 1988; Taxman, 1998; Taxman et al., 1999; Trone and Young, 1996). Some studies identify 90 days of treatment as a minimum threshold for achieving positive outcomes (e.g., De Leon 1984; Hubbard et al., 1989; Simpson, 1979, 1981; Simpson, Joe, and Brown, 1997). Yet, fewer than half of the participants in most community-based programs are retained for 90 days (De Leon and Schwartz, 1984; Joe and Simpson, 1976). Understanding why many participants abandon treatment prematurely, while others become seriously engaged, presents a critical challenge for the treatment community.
It was originally believed that participants voluntarily seeking treatment were more motivated and more likely to succeed in treatment than coerced participants. However, evidence is now mounting that coerced treatment is as effective as voluntary treatment at producing favorable outcomes (Anglin et al., 1989; Belenko, 1999; Collins and Allison, 1983; DeLeon, 1988(a); DeLeon, 1988(b); Hubbard et al., 1989; Lawental et al., 1996; Siddall and Conway, 1988; Trone and Young, 1996). Some programs, such as the Brooklyn Drug Treatment Alternative-to-Prison (DTAP), report a higher retention rate among their defendant population than non-defendants referred to comparable community-based treatment programs (Trone and Young, 1996). Responding to this evidence, court-mandated treatment programs proliferated during the 1990s, and national anti-drug policies broadened from interdiction and enforcement to put more emphasis on treatment. Court-mandated programs are characterized by the application of legal coercion, usually via the threat of incarceration, to motivate participants to succeed. Programs using the drug court model generally differ from earlier court-mandated programs by incorporating more intensive court supervision. Most drug courts require participants to report regularly for evaluation by a court-employed case manager and the drug court judge. Responding to progress, the judge uses a system of intermediate rewards or sanctions, including temporary jail stays, in an effort to motivate participants to remain in treatment.
With the recent explosion of drug courts, it is important for research to identify key characteristics associated with treatment retention among a court-mandated population and to develop effective policies to assist those facing a high risk of dropout. Yet, current studies analyzing predictors of success primarily focus on a voluntary population (Hiller, Knight, and Simpson, 1999), and few studies examine programs using the drug court model (for exceptions, see Peters, Haas, and Murrin, 1999; and Schiff and Terry, 1997). Although several locally funded drug court evaluations show predictors of retention, their methodologies are unknown, and their results were not disseminated except as project reports provided to the funder. For a summary of results from these evaluations, see Belenko (1998,1999). The existing gap in the literature is addressed here by analyzing predictors of treatment engagement at the Brooklyn Treatment Court (BTC), a program for substance-abusing persons arrested in Brooklyn, New York.
THE SIGNIFICANCE OF LEGAL COERCION
Legal coercion has repeatedly been found critical in motivating participants to begin treatment and, once there, to remain in treatment (Anglin et al., 1989; Collins and Allison, 1983; DeLeon, 1988(a); DeLeon, 1988(b); Hubbard et al., 1989; Lawental et al., 1996; Peters and Murrin, 1998; Siddall and Conway, 1988; Trone and Young, 1996). For example, Trone and Young (1996) examined repeat felony defendants (facing three to six years in prison) participating in the Drug Treatment Alternative-to-Prison (DTAP) program and found that 63% were still in treatment at one year. In drug courts nationwide, Belenko (1998) estimates that 60% of participants were in treatment at one year. By comparison, only half of the participants in outpatient drug-free programs nationwide were still in treatment after just three months (Simpson et al., 1997). Trone and Young attributed the high DTAP retention rate to the ācredible threat of incarceration to keep people in treatment until their improvement begins to motivate them to finish the programā (1996: 9). Nonetheless, other studies found no relationship between legal status and treatment retention (Joe and Simpson, 1976; Sansone, 1980; and Simpson and Friend, 1988). Hiller et al. suggest that these āinconsistencies might be due in part to the variety of indicators used as estimates of legal pressureā (1998: 466).
Our review revealed primarily two analytic strategies for measuring legal coercion, and each has methodological limitations. One popular strategy is to study participants enrolled in community-based programs, distinguishing between those who are and are not subject to outside legal supervision (Anglin et al., 1989; Hiller et al., 1998). For example, Hiller and colleagues classified participants as facing low, medium, and high legal pressure and find that those under moderate and high pressure were more likely to be retained at 90 days than were those under low pressure. However, the low pressure group did not have any formal legal status, thus making it difficult to establish whether the moderate and high pressure groups were more likely to be retained due to legal coercion per se or due to other aspects of their legal involvement (e.g., court supervision, encouragement by parole or probation officers, regular drug-tests, etc.).
A second analytic strategy is to focus on participant perceptions as to the level of coercion facing them. For example, Young, Dynia, and Belenko (1996) found, using a sample consisting only of court-involved participants, that those perceiving greater legal pressure were likely to remain in treatment longer, even though objective program features did not vary systematically across participants. While clearly informative, this strategy leaves untested the impact of more objective differences in the level of coercion facing different participants: e.g., whether participants factually facing more incarceration time do better. Testing this is important, since it is easier for programs to manipulate their objective features than to affect the psychological processes underlying each participantās perceptions. Accordingly, we developed a measure classifying the objective level of coercion facing different Brooklyn Treatment Court participants.
OTHER PREDICTORS OF TREATMENT ENGAGEMENT
Much research has studied other factors affecting treatment success outside a court-mandated population. This section reviews those factors found generally relevant. Some of the contradictory findings reported below underscore the difficulty in distinguishing which factors predict treatment success and failure. In some instances, better theorization is needed to explain why certain relationships hold, as opposed to merely reporting results and then relating them to past results (which sometimes differ).
Previous findings suggest that persons who are young, female, less educated, unemployed, and dependent on a highly addictive drug are at higher risk of treatment dropout.
Age. Research consistently shows that throughout the United States and other industrialized societies, criminal behavior peaks in late adolescence and gradually declines thereafter (e.g., Hirschi and Gottfredson, 1983; Farrington, 1986). Most explanations note that adolescents and young adults are particularly likely to seek autonomy and status through involvement in deviant peer groups, whereas deviant peer involvement diminishes as persons age (Hirschi and Gottfredson, 1983; Moffitt, 1993). Correspondingly, it is plausible that older persons tend to be more receptive to rehabilitative policy interventions such as drug courts. Another explanation for the āaging outā phenomenon among a substance-abusing population may be that over time, persons grow tired of their addicted lifestyle. Saxon reasons, āit makes sense that as opioid addicts grow older, increasing dissatisfaction with their addict life-styles, health concerns, and other factors associated with aging may make them more amenable to treatmentā (1996: 1206). Indeed, several studies report that older participants are retained in treatment longer than younger participants (Grella et al., 1997; Mammo and Weinbaum, 1991; Sansone, 1980; Saxon, 1996).
Sex. In general, treatment outcomes for women are poorer than for men (Beckman, 1979; Mammo and Weinbaum, 1993; Wilsnack, 1982). In their analysis of 12,697 alcoholics admitted to outpatient treatment centers, Mammo and Weinbaum (1993) found that 48.8% of the women dropped out of treatment whereas 38% of the men dropped out. However, Mammo and Weinbaum cautioned that educational background may be a confounding variable, since the women in their study were more likely to lack college training (1993: 96). Poorer outcomes may also result because women are more often the primary caretakers of young children and sick family members, possibly creating external barriers to their ability to attend treatment. The literature suggests a need to control more carefully for confounding variables in pinpointing if and why sex-based differences exist. (See Beckerman & Fontana, this volume.)
Race/Ethnicity. To date, studies assessing race and treatment outcome have produced different results. Some conclude that race is significant (Peters and Murrin, 1998; Steer, 1980; Sansone, 1980; Saxon et al., 1996), while others report no relationship (Condelli and Hubbard, 1994; McFarlain et al., 1977). Mammo and Weinbaum (1991) found that it is more likely for white and āotherā race/ethnic groups than blacks and Hispanics to complete treatment. However, when social, demographic, and economic variables were controlled, they found that race is no longer significant. As with the analysis of sex, this highlights the need to control for socioeconomic variables. Additionally, it is important for research to study treatment programs designed specifically for different ethnic groups, since nonwhites may be particularly prone to drop out when treatment programs are not culturally sensitive to their needs (Beauvais, 1998; Fisher et al., 1996; Westermeyer, 1995).
Socioeconomic Status. Research indicates that higher socioeconomic status predicts better outcomes. Several studies report a positive relationship between educational background and treatment success (Hiller et al., 1998; Mammo and Weinbaum, 1991; Sampson et al., 1978). Other studies report a positive relationship between employment and treatment success (Hiller et al., 1999; Hser et al., 1990-91; Mammo and Weinbaum, 1991; McLellan et al., 1983(a); Siddall and Conway, 1988; Steer, 1980). One explanation may be that those higher in socioeconomic status have more to lose economically by continuin...
