Meta-Analysis Snapshot
|
|
Literature Coverage Dates
|
Number of Studies
|
Number of Study Participants
|
|
Meta Analysis 1
|
2001-2010
|
20
|
12394
|
|
Meta Analysis 2
|
2003-2008
|
18
|
10235
|
|
Meta Analysis 3
|
2004-2011
|
6
|
0
|
|
Meta Analysis 4
|
2003-2012
|
9
|
11071
|
Meta Analysis 1
Cross (2011) conducted a meta-analysis of mental health courts (MHCs) to examine their effect on recidivism and clinical outcomes. A comprehensive literature search, which included databases, journal article references, and author queries, was conducted from 1997 through April 2011. Studies were included if they were 1) confined to the United States, 2) written in English, 3) focused on individuals ages 17 years and older with a mental illness, and 4) reported at least one quantifiable recidivism or clinical outcome that allowed for computation of an effect size statistic. Only experimental and quasi-experimental studies were included.
Roughly 45 studies were reviewed, and a total of 20 studies were found to meet the inclusion criteria. Almost all of the studies (19 out of 20) were published journal articles. The remaining study was a thesis/dissertation. All of the studies were published after 2000. Most of the studies (18 out of 20) were quasi-experimental designs, whereas the other two studies were experimental designs. The 20 studies included more than 12,000 participants. All of the studies included males and females, and the average age of participants ranged from 25 to 40 years. The majority of studies had a sample with over 50 percent of participants who were white (the race/ethnicity of nonwhite participants was not specified).
Recidivism outcome measures were coded according to the following priority rankings: 1) bookings rates, 2) rearrests, 3) new convictions, and 4) jail days. Clinical outcomes were coded using the following priority rankings: 1) established measures (such as the BASIS-32); 2) hospitalization days; and 3) psychiatric visits. The Maryland Scale of Scientific Methods (MSSS) was used to assess the methodological quality of each evaluation using a 5-point scale. The standardized mean difference was calculated to examine the effect of MHCs on recidivism and clinical outcomes. Effect sizes reported as positive values indicated positive outcomes (i.e., a reduction in recidivism).
Meta Analysis 2
Sarteschi, Vaughn, and Kim (2011) conducted a meta-analysis to assess the effectiveness of MHCs. A comprehensive literature search was conducted through July 2009 of databases such as Social Science Citation Index, Sociological Abstracts, Social Science Abstracts, and others. Emails were also sent out to more than100 MHCs across the United States, to increase the possibility of locating unpublished studies. Studies were included in the review if they were conducted in the United States, written in English, focused on individuals ages 17 years and older with a mental illness, reported at least one quantifiable MHC clinical or recidivism outcome that permitted computation of an effect size statistic, and used experimental or quasi-experimental designs to compare an MHC treatment condition with a control, comparison, or wait-list group. Pretest-posttest, one-group, or multigroup study designs were also included but were analyzed separately. About 74 relevant citations were reviewed, of which 18 were found to meet the inclusion criteria.
Of the 18 studies, 11 were journal articles or dissertations and seven had not been published in a journal. All of the studies had been published after 2002. Twelve studies were quasi-experimental designs, two studies were experimental designs, and three studies were single-group designs. The authors conducted analyses on recidivism outcomes looking at the effect sizes by the type of study design. The review of this meta-analysis focused on the overall effect size for the quasi-experimental design studies, because single-group design studies do not include comparisons groups and there were too few experimental-design studies to present an overall effect size.
Across the 18 studies, there were more than 10,000 study participants. The average age of treatment-and control-group members ranged from 30 to 40 years old. Most of the studies included male and female participants (usually male). In almost all of the studies, the majority of the participants were white, except for two studies in which the majority of participants were African American, and one study in which about 50 percent of the participants were white and 50 percent were African American.
Meta Analysis 3
Lee and colleagues (2012) conducted a meta-analysis to examine the economic value of programs that reduce crime, including MHCs. A comprehensive literature search was conducted, which involved looking at bibliographies of systematic and narrative reviews, examining the citations in the individual studies, searching research databases, and contacting authors of primary research. The analysis only included studies that had a control or comparison group. Random assignment and quasi-experimental studies were included. Studies that had a treatment group made up solely of program completers were excluded.
A total of six studies were included in the analysis to examine the effect of MHCs on crime. Five of the six studies were published in peer-reviewed journals, and one study was a technical report. All studies were published after 2003. The average age of program participants was 28 years. However, no information was provided on the number of study participants or the gender and race/ethnicity of participants.
A 6-point scale (with values ranging from 0 to 5) was used to assess the quality of the research design and adjust the reported results. A rating of “5” reflects a well-implemented random assignment study, whereas a rating of “0” reflects an evaluation that does not have a comparison group or that has a comparison group that is not equivalent to the treatment group. Studies with ratings of “0” were not included in the analysis. The mean-difference effect size was calculated based on the data coded from the included studies.
Meta Analysis 4
The Utah Criminal Justice Center (2012) conducted a meta-analysis to examine the economic and behavioral outcomes of interventions designed to prevent criminal behavior, including MHCs. A comprehensive search of the literature was conducted between 1987 and 2011 to identify studies of adult MHCs. To be eligible, studies had to meet the following criteria: 1) both the treatment group and comparison/control group must consist of adults (ages 18 years and older) with an identified mental illness; 2) the study must evaluate an MHC program with the following elements: comprehensive supervision, treatment services, and immediate sanctions and incentives; 3) the study must include an outcome measure of recidivism, which could be arrest, conviction, or incarceration; and 4) both experimental and quasi-experimental studies were eligible for inclusion.
Forty-two studies were found, of which nine met the inclusion criteria. Of the nine studies, seven were published in peer-reviewed journals and two were unpublished technical reports. There were about 11,000 participants across all nine studies. All of the studies were published after 2002 and included male and female participants. No information was provided on the age or race/ethnicity of study participants.
All data was coded and transformed into an odds-ratio, with values above 1 indicating a negative-treatment effect and values below 1 indicating a positive-treatment effect (i.e., reduced recidivism rates for those who participated in an MHC). A random-effects model was used to analyze the data. The MSSS was used to assess the methodological quality of each evaluation using a 5-point scale. Two studies received a 5 for methodological quality and the remaining studies received a score of 3 or 4.
