A Meta-Analytic Examination of Drug Treatment Courts: Do They Reduce Recidivism?

3. Results

3. Results

3.1 Study Characteristics

Although 185 individual studies were identified through the search process, only 54 (29.2%) studies were deemed acceptable according to the study inclusion criteria. As Table 1 indicates, the research located on drug treatment courts originates almost exclusively from the United States. Only two studies were located in Canada (which is consistent with the number of Canadian DTCs that were operating long enough to generate evaluations) and two in Australia.

Table 1: Study characteristics

Country (N=54)
United States 50 (92.6%)
Canada 2 ( 3.7%)
Australia 2 ( 3.7%)

Publication type (N=54)
Academic journal article 31 (57.4%)
Other publication type (e.g., government report) 23 (42.6%)

Study design (N=54)
Random control group 7 (13.0%)
Simple comparison group 27 (50.0%)
Matched comparison group 20 (37.0%)

Control/comparison group (N=72)
Justice system (e.g., probation or custody sample) 39 (54.2%)
Drop-outs/non graduates 10 (13.9%)
Eligible but did not participate 23 (31.9%)

The studies were published between 1993 and 2005 with a median year of 2001. Just over half (57.4%) were published in peer-reviewed academic journals, which indicates that this meta-analysis is not relying solely on "traditionally published" studies. A large majority of the studies (87.0%) used a comparison group (simple or matched). Although the control/comparison groups were combined in the calculation of the ESE, there was a total of 72 unique control/comparison groups within the 54 studies. Most compared the DTC participants to offenders within the traditional justice system (54.2%) or offenders who were eligible but did not participate (31.9%) – the latter being a more appropriate comparison group as these offenders would have been screened for substance abuse problems.

One of the more important issues within program evaluation research, and particularly within DTC research, is the attrition rate (i.e., the proportion of participants who voluntarily or involuntarily leave the program before completion). The recidivism rates used in analysis do not include (or rarely include) offenders who did not complete the program. Therefore, it is important to consider attrition when examining recidivism rates. The attrition rates within the studies in this meta-analysis ranged from 9.0% up to 84.4% with a mean of 45.2% (SD=19.0), which indicates that almost half of DTC participants do not complete the program. Clearly, strategies need to be developed to decrease attrition rates within DTC programs.

A second important factor to consider in a study of recidivism is the follow-up length used to measure recidivism. Generally, longer follow-up periods produce higher recidivism rates as offenders have more time at-risk to re-offend. The follow-up length used to measure recidivism within the included studies ranged from 3 months up to 48 months with a mean of 18.7 months (SD=11.5).

Sample size is a third factor considered important in program evaluation research as larger sample sizes are considered more rigorous than smaller sample sizes. The sample sizes ranged from 39 participants up to 856 participants, with a mean of 260 participants (SD=189).

3.2 Program Characteristics

Although there were 54 individual studies, some studies reported on multiple DTCs. As such, a total of 66 individual DTC programs were included in this meta-analysis. Table 2 summarizes the DTCs described in the individual research studies. The data, however, should be viewed with some caution. There were difficulties in coding detailed information, as the programs were not adequately described in many of the studies.

Very few DTCs restricted participation based upon drug type. Some programs (4.6%) restricted access to only hard drug users (e.g., cocaine, heroin, crystal methamphetamine), and other programs (7.6%) to only soft-drug users (e.g., marijuana, hashish, alcohol). Approximately one-third of programs (31.8%) dealt primarily with repeat offenders (19.7% mostly repeat offenders and 12.1% all repeat offenders) and 19.7% dealt primarily with first-time offenders (18.2% mostly first-time offenders and 1.5% all first-time offenders). Finally, almost all DTC programs (93.9%) accepted only offenders who had been charged with non-violent offences. In summary, the programs in this meta-analysis are dealing with both first-time and repeat non-violent offenders who have substance abuse problems with numerous types of drugs.

The program lengths of DTCs (i.e., the time an offender was monitored) within this meta-analysis varied from 6 months up to 26 months with a mean of 13.4 months (SD=4.0).

Table 2: Program characteristics

Substance abuse accepted (N=-66)
Hard drugs only (e.g., heroin, cocaine) 3 ( 4.6%)
Mixed drugs/unknown 58 (87.8%)
Soft drugs only (e.g., marijuana, hashish) 5 ( 7.6%)

Offence types accepted (N=66)
Non-violent only 62 (93.9%)
Mixed/unknown 4 ( 6.1%)
Violent only 0 ( 0.0%)

Criminal history (N=66)
All first-time offenders 1 ( 1.5%)
Mostly first-time offenders (seventy percent or more) 12 (18.2%)
Mixed/unknown 32 (48.5%)
Mostly repeat offenders (seventy percent or more) 13 (19.7%)
All repeat offenders 8 (12.1%)

Program setting (N=66)
Out-patientonly 20 (30.3%)
Combined (both inpatient and outpatient) / unknown 46 (69.7%)

The identified studies sometimes indicated specific components of the DTC programs under scrutiny. For example, in recognition that substance abuse and criminality are often linked to other factors, some programs targeted a number of additional areas such as academic skills, vocational skills, and family functioning. Table 3 provides information on the additional treatment targets reported in the studies. It should be noted that the specific treatment components were only coded if there was an explicit indication in the study that the DTC targeted that particular area. Therefore, the figures in Table 3 may not necessarily reflect the treatment programs accurately. In general, the DTC programs appear to target a number of additional issues beyond substance abuse.

Table 3: Additional treatment components

Treatment component (N=66)
Academic skills (e.g., school attendance and performance) 40 (60.6%)
Vocational skills (e.g., specific trades, interview skills) 36 (54.6%)
Family functioning (e.g., communication, parenting skills) 33 (50.0%)
Aftercare (e.g., follow-up supervision after program completion) 25 (37.9%)
Social skills (e.g., social competence, ability to work in groups) 24 (36.4%)
Cognitive skills (e.g., goal setting, problem solving) 21 (31.8%)
Anger management (e.g., perspective taking, reducing aggression) 16 (24.2%)
Antisocial peers (e.g., association with criminal peers) 15 (22.7%)
Antisocial attitudes (e.g., lack of respect for authority, criminal values) 15 (22.7%)
Relapse prevention (e.g., techniques for remaining substance free) 15 (22.7%)
Psychological well-being (e.g., self-esteem, depression, anxiety) 14 (21.2%)