The Impact of Drug Treatment Courts on Recovery: A Systematic Review

Introduction. Earlier reviews regarding the effectiveness of Drug Treatment Courts (DTCs) reported a reduction in reoffending and substance use. Although substance users suffer from other difficulties than drug use and judicial issues, none of these reviews focused on outcomes or effects of DTCs on drug-related life domains, such as social relationships, employment, or health. Therefor, the present paper aims to review the impact of adult DTCs on substance use and drug-related life domains. Method. Primary studies were systematically searched in Web of Knowledge. Observational and controlled evaluation studies of adult DTCs were considered eligible if substance use and/or drug-related life domains were measured. Results. Moderately positive results were found with respect to within-program substance use. Few studies used drug-related life domains as an outcome measure and most of them yielded no effects. Employment and family relations ameliorated when specific interventions were used. Discussion. DTCs yield beneficial outcomes and effects regarding within-program substance use. However, evidence regarding the impact of DTCs on post-program drug and alcohol use and on other drug-related life domains is scarce. These life domains and thus QoL possibly can be improved by DTCs if specifically targeted. Future research is warranted.


Introduction
Substance use disorders (SUDs) are important public health problems in Western countries [1,2] and are oen regarded as chronic relapsing disorders [3][4][5][6]. e chronicity of SUDs is illustrated by the observation that most substance users cycle repeatedly between abstinence on the one hand and relapses into active substance use on the other hand before reaching the phase of stable recovery. Sometimes stable recovery is never reached [7]. In the past, lifelong abstinence was seen as the only indicator of recovery. In recent years, however, abstinence is seen as just one indicator of recovery. A shi towards a focus on other life domains and quality of life (QoL) as indicators for recovery from SUDs has been initiated [4,8]. SUDs are typically associated with severe impairments in drug-related life domains such as health, wellbeing, social network, employment, and �nancial situation [3][4][5][6]. Difficulties in these life domains remain for several years aer abstinence has been achieved, particularly in the area of employment [4]. Consequently, recovery from SUDs goes beyond abstinence, instead all associated life domains should be considered. Commitment to recovery is related to one's QoL which in its turn can be enhanced by (re)gaining and maintaining certain desired needs in life (e.g., stable housing, education and work, family, wellbeing, and stable �nancial situation) [9][10][11].
Substance users are highly prevalent in the criminal justice system [12,13] due to substance-related offences (e.g., possession and/or consumption of drugs, driving while intoxicated and the). Imposing judicial alternatives to imprisonment on substance using offenders contains an opportunity to divert them to community (drug) treatment services. ese alternative measures are associated with improvements in drug-related life domains, next to decreases in criminal offending and substance use [14][15][16]. Generally, 2. Method 2.1. Inclusion Criteria. Studies on adult DTCs, namely, standard "DTC", "Family Treatment Drug Court" (FTDC), "Dependency Drug Court" (DDC), "Driving Under In�uence Court" (DUI), and "Driving While Intoxicated Court" (DWI) were included. FTDCs (or DDCs) and DUIs (or DWIs) are modeled on standard DTCs. e former are aimed at substance abusing parents, the latter at substance abusing drivers (especially drunk drivers). Observational or controlled evaluation studies of DTCs were considered eligible if at least one indicator of substance use and/or drug-related life domains, analogous to the subscales of the Addiction Severity Index (ASI; [29]) (namely, employment, �nancial situation, housing situation, physical and mental health, family and social relationships, and leisuretime activities), was reported as an outcome measure. No restrictions regarding measurement method or instruments were imposed.

Search Strategy and Study Selection.
Web of Knowledge was consulted twice using the following search terms and Boolean operators up to December 31st, 2011: (1) a general search using drug court AND (evaluation OR effect * OR outcome), and (2) a speci�c search using drug court AND (employment OR work OR income OR �nancial OR housing OR health OR family OR social OR leisure). e "title only" option was marked and the search was restricted from 1989, since the �rst DTC was implemented in this year [20]. e general and speci�c Web of Knowledge searches together generated 576 hits. Aer removing double hits and screening title and abstract, 61 potentially relevant papers were retrieved for more detailed evaluation. ese studies were screened on meeting the inclusion criteria and subsequently 45 studies were excluded aer reading the full texts due to irrelevant subject, descriptive nature, a mere focus on recidivism, or secondary analysis of previously published data. Finally, 16 studies met the inclusion criteria, which are [30][31][32][33][34][35][36][37][38][39][40][41][42][43][44][45] in the reference list of this paper.

Data Extraction.
Eligible studies were independently screened by two researchers (C. Wittouck and A. Dekkers) using a checklist to extract data on the following variables: (1) author, publication data, and country where the study was conducted, (2) DTC characteristics and procedures, (3) study design and follow-up period, (4) description of intervention and control group (if present), (5) sample size and dropout rate, (6) demographic and substance use characteristics of participants, and (7) outcome variables, measurement instruments, and study �ndings. Due to the heterogeneity in study methodologies and the variety in data reporting, a narrative review was regarded as most appropriate. e individual study �ndings were grouped according to outcome measure e Scienti�c World �ournal 3 and were tabulated in a separate table to facilitate comparison and discussion. Table 1 provides an overview of included studies, according to location of DTC, study design, participant characteristics, and outcome measures.

Study
Quality. Study quality was assessed using the Maryland Scienti�c Methods Scale (MSMS) [46] and was also based on the quality assessment criteria described in the guidance for undertaking systematic reviews [47].
e follow-up period in the 16 included studies differed to a large extent. Some studies carried out a follow-up measurement aer a particular time post-admission [36,38,41,45]. Within these studies the follow-up period ranged from four months aer baseline [45] up to three years aer randomization [34]. In other studies, the outcomes were measured on several follow-up moments, both within-and post-program [30-32, 35, 39, 42, 43]. Some studies only compare post-program outcomes [33,34,37,40,44].

Effects on Drug Use, Alcohol Use, and Other Drug-Related
Life Domains. In Table 2 the individual study �ndings are displayed according to study design and outcome measure. Outcome measures were drug use, alcohol use, family and social relationships, employment and income, and mental and physical health. None of the included studies reported on housing situation or leisure-time activities as outcome measures. Five studies using a RCT [35,36,38,42,43] and one study using a QED [41] were treated as observational studies since they randomized between a standard DTC and an "enhanced" DTC. Some studies only reported on postprogram outcomes regarding drug use and family reuni�cation but were still included in the review, since it could be assumed that all participants were drug users at baseline or risked losing parental rights at baseline [33,34,37,40,44]. Substance use signi�cantly improved over time in almost all studies, regardless of study design. With respect to other drug-related life domains, the difference in results between observational studies and studies using a comparison condition is very obvious. All observational studies, except one [35], conclude that DTC participants improved on all these drug-related life domains from pre-to post-test. ese bene�cial results were not found by controlled studies. Most of the latter types of studies only found favourable results for DTCs on drug and alcohol use. Only FTDCs found bene�cial results regarding family and social relationships.

Effects on Drug
Use. In the majority of the included studies ( , participants ) illicit drug use was an outcome measure. In the Freeman study [32], illicit drug use was measured indirectly by looking into self-reported weekly spending. Weekly spending signi�cantly decreased while weekly legal income did not change signi�cantly, indicating a reduction in drug use over time. In the other studies drug use was measured directly either by urine test results [30,31,42] or self-reported data on drug use [34,38,45] or both [35,39,43].
Self-reported data on drug use showed positive results for DTC participants within-program [35,43,45], at discharge [39] and post-program [34,35,39,43]. In addition, DTC participants reported signi�cantly less illicit drug use than drug using offenders diverted to standard adjudication [34] and methamphetamine-dependent outpatients with no drug court supervision [39]. Leukefeld and colleagues [38] found that drug use decreased for all study groups. However, the study�s high upgrading group performed signi�cantly better than the drug court as usual and the low upgrading group.
Urine test results also showed bene�cial outcomes for DTC participants during the DTC trajectory [39,42,43]. Only one study reported on urine test results aer a DTC program. Speci�cally, Dakof and colleagues [43] found that mothers were less likely to provide a positive urine test aer the DDC program ended. Deschenes et al. [30] found that about half of DTC participants as well as probationers delivered a positive drug test within-program. Signi�cant differences between these two study groups emerged when type of drugs was considered. Probationers tested signi�cantly more positive for cocaine and heroin, whereas drug court participants were signi�cantly more likely to test positive for marijuana.
Not all studies concluded unanimously positive. Although both DTC study groups in the Marlowe et al. study [35] signi�cantly improved with respect to self-reported drug problems, these results were not mirrored by the urine drug screen results. Positive urine drug screens increased, although not signi�cantly, over time for both conditions. Most of these positive drug tests were related to cannabis use. But then, the vast majority of study participants were cannabis users at baseline. Brewster [31] found that, although the rate of positive drug tests was lower for DTC participants in comparison with probation participants, a comparable amount of participants from both study groups remained drug-free throughout the study period. Possibly, the higher  1 In this study, there were two comparison groups ("treatment refusal group" and "traditional child welfare") next to the intervention group ("FTDC"). However, comparison group "treatment refusal" and it's results were unclear, therefore this group was not included in this review. 2 Participants were randomly assigned at intake either to attend judicial status hearings on a bi-weekly basis throughout their enrollment in drug court ("bi-weekly condition"), or to be monitored by their treatment case managers who petitioned the court for status hearings as needed in response to serious or repeated infractions ("as needed condition'). 3 Enhanced employment intervention aimed at obtaining, maintaining, and upgrading employment and attending upgrading sessions. e number of upgrading sessions attended by each participant was divided by the total number of possible upgrading sessions that a participant could have attended. e resulting percentages were then split in half, with those below the median in the low upgrading group and those above the median in the high upgrading group. 4 Engaging Moms Program(EMP) was adapted for use in a family drug court context. EMP was designed to help mothers succeed in drug court by helping them comply with all court orders, including attending substance abuse and other intervention programs (e.g., domestic violence counselling, parenting classes, etc.), attending court sessions, remaining drug free, and demonstrating the capacity to parent their children.
e only difference between the FTDC and EMP groups was the working relationship between the drug court caseworker and the mothers; all other aspects of the programs, including overall requirements, phases, and sanctions and rewards, were exactly the same.
A "+" indicates a signi�cant difference in favor of the DTC, a "−" indicates a signi�cant difference in favor of the control group, a "�" indicates no signi�cant difference between the DTC and the control group, and "#" indicates that the outcome variable was not reported. 1 non-controlled pre-post design. 2 randomized controlled trial. 3 quasi-experimental design. amount of positive drug tests among probationers can be attributed to drug testing at irregular intervals (e.g. in case of suspicion) of drug use), while testing of DTC participants at regular intervals is standard practice.

Effects on Alcohol Use.
Only six studies (participants ) included alcohol use as an outcome measure. Eibner and colleagues [36] focused exclusively on alcohol use since they evaluated a DUI court. All six studies found that DTC participants reported signi�cantly less alcohol use over time. DTC participants as well as outpatients with no DTC supervision, both suffering from methamphetamine dependence, reported signi�cantly less alcohol use over time [39].

Effects on Family and Social
Relationships. In ten of the included studies (participants 2 ) family and social relationships were used as an outcome measure. While in some studies social functioning [32], family functioning, and parenting practices [41,43] of DTC participants signi�cantly improved over time, other studies found no difference regarding family and social relationships [35].
In comparison with non-DTC methamphetaminedependent outpatients [39] and drug using offenders proc-essed through standard adjudication [34], family and social relations of DTC participants improved signi�cantly.
Slightly more than half ( , participants 2) of these ten studies evaluated FTDC [33,37,40,41,43,44]. Dakof and colleagues [41,43] found that FTDC interventions can positively in�uence child welfare outcomes. Besides, at follow-up, family reuni�cation was signi�cantly more likely in FTDC cases than in cases processed through traditional child welfare services [37,40,44], except in the Ashford study [33]. Children of parents managed by traditional child welfare services were signi�cantly more likely to be allocated to out-of-home-placement [37] and to reach permanency faster [40,44]. Ashford [33], on the other hand, found that FTDC children reached a permanency decision sooner than children in traditional child welfare services. Furthermore, time to family reuni�cation was signi�cantly longer for children in traditional child welfare services than children in the FTDC group [40]. In addition, Burrus and colleagues [44] found that FTDC children spent less time in nonkinship foster care. Surprisingly, Boles and colleagues [37] also found that DDC participants had a higher re-entry rate to out-ofhome care a�er family reuni�cation than comparison participants. When outcomes of these re-entries were examined, however, it was found that all these comparison children e Scienti�c �orld �ournal 9 moved to permanency, while almost one third of DDC children were subsequently reuni�ed with their parents.

Effects on Employment and Income. About half of the included studies (
, participants ) used employment as an outcome measure, and two of these studies also reported on income. Only two studies found bene�cial outcomes for DTC participants regarding employment situation. Dakof and co-authors [43] found that employment problems of mothers in DDC signi�cantly decreased. In their evaluation of an enhanced employment intervention, Leukefeld and colleagues [38] found that high upgrading participants reported signi�cantly more full-time employment, signi�cantly less unemployment, and more days of working when compared to both the no-intervention and the low upgrading group. Participants in the no-intervention group worked signi�cantly more in the past 30 days than those in the low upgrading group. In concordance with the abovementioned results, participants in the high upgrading group reported more income from a legitimate job in the past year than those in the low upgrading group. e no-intervention group did not differ from either of the upgrading groups.
In the Marlowe et al. [35] study, reported employment problems were not signi�cantly reduced over time for participants from both DTC conditions. Furthermore, employment situation of drug court participants did not differ signi�cantly when compared to non-drug court participants with methamphetamine dependence [39], standard adjudication [34], and probationers [31]. Deschenes and colleagues [30] even concluded that probationers were signi�cantly more likely to be employed than drug court participants. However, during the follow-up period, DTC participants were signi�cantly more likely than probationers to be involved in counselling and outpatient treatment sessions [30]. Although Gottfredson and co-authors [34] did not �nd a signi�cant difference regarding employment situation between drug court participants and drug offenders processed through traditional adjudication, they did �nd the former relying on welfare signi�cantly less than the latter, which could be due to a reduction in money spent on drug use since drug use decreased signi�cantly in DTC participants.

Effects on Mental and Physical Health.
In only �ve studies (participants ) the effects of drug treatment court on mental and physical health was examined. Mental health [32,43], general health and vitality [32] of DTC participants signi�cantly increased over time, while medical problems [35,43], bodily pain, and emotional and physical role limits [32] of DTC participants signi�cantly decreased over time. Marlowe and colleagues [35], however, did not �nd improvement for mental health and Freeman [32] found no improvement regarding physical functioning.
Furthermore, no differences were found regarding mental and physical health between DTC participants and non-DTC methamphetamine-dependent outpatients [39] nor between DCT participants and drug using offenders processed through standard adjudication [34].

Discussion
e results of the present paper show that, in general, DTCs appear to produce a decrease of illicit drug use in substance using offenders during a DTC trajectory, which was also found by earlier review studies [21,[23][24][25]. DTC participants, who did not reach abstinence, oen moved away from more harmful substances as heroin to less harmful substances as marijuana [30]. Little information is currently available on the long-term effects of DTCs on illicit drug use. Although long-term self-report data show promising results, information from urinalysis con�rming these results is lacking. Alcohol use of DTC participants also decreased. Surprisingly alcohol use was an outcome measure in only half of the included studies, given the well-established link between drug and alcohol use [48][49][50][51].
Although the importance of drug-related life domains, for example employment and housing, is recognized in the recovery and desistance literature [9-11, 14, 15, 17-19], only a small amount of DTC evaluation studies focusing on these outcomes were found. e lack of empirical data on the effects of DTCs on drug-related life domains might be explained in two ways. Or DTCs focus predominantly on substance use and drug-related crime which results in a lack of attention to other drug-related life domains. Or DTC research has not yet caught up with the state of the art on recovery, resulting in a lack of focus on these drug-related life domains. Consequently, DTC research should broaden its focus and systematically record information on DTC interventions aimed at these drug-related life domains, and study the short-term and long-term effects of DTCs on these life domains and QoL of DTC participants. Subsequently, evidence-based recommendations can be made in order to improve DTC policy and practice.
As for the studies that do focus on drug-related life domains, observational studies found bene�cial results, which were not consistently demonstrated by (randomized) controlled studies. As opposed to standard adult DTCs, FTDCs do appear to positively in�uence child welfare outcomes (e.g., family reuni�cation) [37,40,44]. Next, Leukefeld and colleagues [38] found that DTC participants who received an enhanced employment intervention aimed at obtaining, maintaining, and upgrading employment and attended su�cient sessions, had more bene�cial outcomes regarding their work situation (e.g., more full-time employment and less unemployment).
Following the (moderately) positive results of DTCs on substance use, of FTDCs on child welfare outcomes and of the DTC with an enhanced employment intervention on work situation, it can be hypothesized that providing speci�c interventions for each of the drug-related life domains will bene�cially affect these domains. Indeed, considering the chronicity of SUDs and the complexity of associated problems [3][4][5][6][7], one cannot expect that a mere focus on substance use as such will automatically entail improvements in other life domains. Moreover, by offering interventions speci�cally aimed at improving drug-related life domains, attention is given to outcomes which are reported as desired by drug users themselves. In other words, such interventions start e Scienti�c World Journal from drugs users' own experiences and expectations (e.g., QoL) [52]. Aer all, the recovery model is a widely used and accepted approach for supporting persons with mental health and/or addiction problems, ultimately aimed at improving QoL [9,53,54]. When substance users experience progress in those drug-related life domains they consider important, and subsequently in their QoL, their drug use and criminal offending could be positively in�uenced.
No information was available regarding the effects of DTCs on housing situation or leisure-time activities, although research shows that moving away from drugusing friends and acquaintances supports the maintenance of abstinence [17] and permanent housing is associated with a reduction in recidivism [14,15]. In addition, non-substancerelated leisure-time activities contribute to a higher QoL [55].
Some limitations of the present paper should be addressed. First, one should be cautious when generalizing the results of the present paper. e restriction of the search strategy to peer reviewed publications could have induced publication bias, since oen only favorable results are published. Although peer review also guarantees some form of quality control. Mitchell and colleagues [27, p. 69] also found that the results of published and unpublished studies on the effects of DTCs are "roughly similar. " Next, only a small number of individual studies reporting on the effects of DTCs on drug use, alcohol use, and/or drug-related life domains were available. e included studies are also marked with substantial heterogeneity regarding methodology, sample size, and measurement instruments. Comparability of individual study results is further compromised since comparability between DTCs from different jurisdictions is low (e.g., inclusion criteria, procedures, treatment services, and treatment providers) [21]. Evaluations are thus very site-speci�c, which gives little insight into the overall effectiveness of the system [56]. In fact, a general DTC framework and general DTC theoretical underpinnings are lacking [56,57]. Generalizability is further limited since bene�cial outcomes are more prevalent in DTC graduates than in dropouts [25,27], and in general about half of DTC participants drop out of the trajectory [21,[23][24][25]. e dropout rate varied greatly between the included studies, and intent-to-treat analyses were not carried out or were impossible to carry out. In addition, the included studies were almost exclusively USA based. e adoption of USA DTC practices and the generalizability of USA DTC results to other countries, especially European, is problematic because of fundamental differences in the law system. Second, RCTs, which receive the highest score on the Maryland Scienti�c Methods Scale [46], are rare in evaluating DTC's effectiveness. In the present review less than half of the 16 included studies used a RCT, and only two of these RCTs randomly assigned participants to DTC or another kind of judicial processing. e observational and quasi-experimental studies do not allow to conclude that the observed bene�cial outcomes are attributable to DTCs, since improvement due to confounding factors cannot be ruled out. Finally, to our knowledge, the longest follow-up period in DTC evaluation studies is three years [34]. Empirical data on the long-term effects of DTCs on drug-related life domains and substance use (and criminal offending) is thus lacking since studies with extended follow-up periods are non-existent.
To conclude, although one should consider the abovementioned limitations when generalizing the present paper's results, some important conclusions should be highlighted. First, through a dominant focus on substance use and criminal offending, DTCs and DTC research possibly suffer from a lack of attention and interest for other drug-related life domains and QoL of substance users. Second, these life domains can be improved as long as they are addressed. Consequently, DTC policy and practice should be adapted according to the recent �ndings of recovery and desistance research by focusing on improvement in drug-related life domains and by targeting these domains using speci�c interventions thus improving QoL of substance using offenders. In addition, each DTC trajectory should be tailored to the unique problems a DTC client faces, herewith assuring a more individual approach. As research has shown that great interpersonal variability exists between DTC participants, and that the effectiveness of DTCs differs according to these differences [38,58,59]. Finally, it can be expected that a decrease in substance use and criminal offending results from better life circumstances for substance users. erefor, future research on the effectiveness of DTCs should use a more comprehensive focus and study the short-term and long-term effects of DTCs on drug-related life domains and QoL next to the effects on substance use and criminal offending, which are, aer all, socially desirable outcomes [52].