Volume 3, Issue 1 • Fall 2013

Table of Contents


The Impact of Juvenile Mental Health Court on Recidivism Among Youth

Gender-Specific Mental Health Outcomes of a Community-Based Delinquency Intervention

Predicting Recidivism Among Juvenile Delinquents: Comparison of Risk Factors for Male and Female Offenders

Building Connections Between Officers and Baltimore City Youth: Key Components of a Police–Youth Teambuilding Program

Internet-Based Mindfulness Meditation and Self-regulation: A Randomized Trial with Juvenile Justice Involved Youth

Assessing Youth Early in the Juvenile Justice System

A Jury of Your Peers: Recidivism Among Teen Court Participants

Commentary: Place-Based Delinquency Prevention: Issues and Recommendations

The Impact of Juvenile Mental Health Court on Recidivism Among Youth

Donna M. L. Heretick and Joseph A. Russell
Strategies for Solutions Consulting, Arvada, Colorado

Donna M. L. Heretick, president, Strategies for Solutions Consulting; Joseph A. Russell, research associate, Strategies for Solutions Consulting.

Acknowledgments: Appreciation is extended to Mr. Jeff McDonald, Coordinator of the First Judicial District Mental Health Court (SB94), for the opportunity to work on this important project. Mr. McDonald is a person with dedication and vision, qualities shared by others who are involved in all aspects of the Juvenile Mental Health Court program

Correspondence concerning this article should be addressed to: Donna M. L. Heretick, Strategies for Solutions Consulting, 10455 West 85th Place, Arvada, Colorado 80005. donna_heretick1@juno.com

Keywords: juvenile justice, mental health court, recidivism, program evaluation


As many as 70% of youth who enter the juvenile justice system are diagnosed with mental disorders. In 2009 alone, 1.54 million juveniles were arrested. Recidivism rates for these youth can be as high as 52%. Juvenile Mental Health Courts (JMHCs) in conjunction with Intensive Supervised Probation (ISP) is one initiative that addresses the special needs of these juveniles; however, there is limited outcome research with meaningful data comparing juveniles with and without access to JMHCs. This study employed a retrospective observational design to compare the recidivism outcomes of 81 youths (ages 10 to 17) who entered a JMHC in Colorado between 2005 and 2011 with recidivism outcomes for juveniles who entered a JMHC in California during the same time period. This study also compared the outcomes of juveniles in Colorado who were adjudicated and assigned to other forms of probation and diversion, and juveniles in the same state who were diagnosed with a mental disorder and assigned to intensive supervised probation, but who do not have access to a JMHC. Youth in the experimental group (i.e., those with access to the JMHC) showed significantly decreased recidivism rates during and following their probationary period than those in the comparison groups. Average time to reoffending for youth who completed JMHC successfully exceeded 1 year, with a significant reduction in violent/aggressive and property offenses. This article examines outcomes and includes recommendations for the future evaluation of JMHCs.


It is estimated that as many as 70% of youth who enter the juvenile justice system are diagnosed with one or more mental disorders (Hammond, 2007). To put this figure into perspective, approximately 1.54 million individuals under age 18 were arrested in the United States in 2009 alone (U.S. Census Bureau, 2012). While precise recidivism rates for this population are often difficult to determine (Roberts & Bender, 2006; Snyder & Sickmund, 2006), McReynolds, Schwalbe, and Wasserman (2010) have estimated a recidivism rate of 52.8% for these youth.

Mental Disorders Among Youth

Definitions of mental disorders among youth and adolescents in the juvenile justice system generally have followed criteria established by the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; APA, 2000). The most common diagnoses for these youth are disruptive disorders (46.5%), especially conduct disorder, followed by substance use disorders (46.2%), anxiety disorders (34.4%), and mood disorders (18.3%). Even when substance use disorders are not included, fully 61.8% of youth in the juvenile justice system meet criteria for at least one, and often more, mental disorders (Shufelt & Cocozza, 2006).

Juvenile Mental Health Courts

JMHCs are voluntary diversion programs that adhere to a paradigm of therapeutic jurisprudence, following a “nonadversarial, treatment-oriented approach when adjudicating juvenile offenders, while still upholding their due process rights” (Office of Juvenile Justice and Delinquency Prevention, OJJDP, n.d., para. 6). JMHCs rely on cooperation and collaboration among members of a multidisciplinary team who work with and for the benefit of the youth and the family. Similar to mental health courts for adults, JMHCs have two main goals: first, to decrease recidivism; and second, to increase participants’ adherence to treatment (McNiel & Binder, 2007).

The youth’s and family’s compliance with treatment is essential to the effectiveness of JMHCs. The definition of treatment compliance for youth is broader than it is for adults, and may include not only follow-up and adherence to treatment for health and mental health concerns, but expectations in relation to the youth’s behavior in school and within the family. In addition, the family is expected to comply with the JMHC program, since their compliance also affects outcomes for youth (Callahan, Cocozza, Steadman, & Tillman, 2012).

The JMHC attempts to meet the needs of juvenile offenders while protecting public safety. Various models of JMHCs have been implemented; evaluation of their effectiveness is ongoing. JMHCs follow the tradition of other diversion programs for youth, which are based on the philosophy that for certain youth, involvement in the courts and institutions is counterproductive to rehabilitation (Center on Juvenile and Criminal Justice, 1999; Wales, Hiday, & Ray, 2010).

Between 1977 and 1979, researchers in Colorado pilot tested an early diversion program that provided individual, parental, and/or family counseling to a general sampling of 848 youth who were diverted from processing in the formal juvenile justice system (Pogrebin, Poole, & Regoli, 1984). Although not a juvenile mental health court, youth in this initial demonstration diversion program had significantly lower recidivism rates than a control group of youth who proceeded through the juvenile justice system as usual (Pogrebin et al., 1984). Many successful diversion programs, not specific to youth diagnosed with mental disorders, have followed the methods detailed in that study (Center on Juvenile and Criminal Justice, 1999; District Attorney Jefferson & Gilpin Counties, CO, 2012).

Early JMHCs were introduced in 1998 in York County, Pennsylvania; in 2000 in Mahoning County, Ohio; and in 2001 in Santa Clara County, California. By early 2012, there were approximately 50 JMHCs in 15 states, either in operation or in the planning stages. Ohio, with nine, and California, with eight, are the states with the greatest number of JMHCs. Other states, such as Colorado, Illinois, and Idaho, have one to two districts with JMHCs (Callahan, et al. 2012; SAMHSA’s Gain Center, n.d.).

Pre-Release and Post-Release Recidivism

Recidivism, generally defined as reoffending, is classified by the Colorado Division of Probation Services (2010) according to two types: pre-release recidivism/failure and post-release recidivism. Pre-release recidivism, which occurs while the individual is on probation, is defined as “an adjudication or conviction for a felony or misdemeanor, or a technical violation relating to a criminal offense, while under supervision in a criminal justice program.” Post-release recidivism is defined as, “a filing for a felony or misdemeanor within 1 year of termination from program placement for a criminal offense” (Colorado Division of Probation Services, 2010, p. vii). Both status and nonstatus offenses are included in these definitions. A “status” offense is one which, under the current law, would not be a crime if committed by an adult (e.g., truancy or curfew violations; Colorado Division of Criminal Justice, 2011a).

Evaluations of Mental Health Courts

As Cocozza and Shufelt (2006) note, most of the evaluations of mental health courts have focused on those for adults. Recent studies of adult mental health courts have included comparison groups, as well as pre-post designs. Results indicate that they are effective both in reducing recidivism and in increasing the use of mental health services (Callahan & Wales, 2013).

By comparison, fewer outcome studies have evaluated the effectiveness of JMHCs, especially with viable comparison groups using post-participation recidivism rates (Cocozza & Shufelt, 2006; OJJDP, n.d.). One of the few that has been evaluated is California’s Santa Clara County Court for the Individualized Treatment of Adolescents (CITA; Behnken, Arredondo, & Packman, 2009). The CITA evaluation compared offense patterns of youth before they entered CITA with their offense patterns and frequency of offenses during their time in the program. The researchers found a significant reduction in recidivism during the youths’ participation in CITA. However, the researchers did not have information on a viable comparison group or on post-release recidivism patterns (Behnken et al., 2009).

A more recent report on California’s Alameda County Juvenile Collaborative Court (ACJC; National Center for Youth Law, NCYL, 2011) offers information on retrospective data (case files) for 34 participants (29 successful completers) who attended between 2007 and 2009. Their report includes patterns of offenses, and needs and uses of mental health and other services, for a period prior to the youths entering ACJC, during their participation, and for up to a year following participation. In addition, interviews and surveys of stakeholders expand on the researchers’ descriptions of experiences and needs. However, the sample size is small, and the repeated measures design does not provide for between-group comparisons on relevant outcome measures.

Target Program for Present Evaluation

The current project is an evaluation of the JMHC of Colorado’s First Judicial District. Initiated in 2005, this Colorado JMHC was modeled after the CITA program in Santa Clara County, California (Behnken et al., 2009) and follows principles of therapeutic jurisprudence and multidisciplinary cooperation:

The First Judicial District Juvenile Mental Health Court is a collaborative effort between the Jefferson Center for Mental Health, the District Attorney, the courts, the probation department, the Public Defenders’ Office, the Juvenile Assessment Center, Human Services, the school district, and the Rocky Mountain Children’s Law Center. A member from each of these agencies makes up the mental health court’s screening committee (District Attorney Jefferson & Gilpin Counties, CO, 2009, para. 4).

Referrals to the Colorado First Judicial District JMHC can come from any public agency, community provider, school, or individual when a charge has been filed against a youth for either an eligible felony or misdemeanor. In order to be eligible for JMHC, the youth must be at least 10 years of age and a legal minor. The youth is screened for mental health status, and must have either a major mental health disorder or substance abuse use with a major mental health disorder. Ineligibility also includes current or previous charge for a class 1 felony, and primary mental health diagnosis of behavior disruptive disorder (e.g., conduct disorder), mental retardation or brain injury, sexually abusive youth, substance dependence or abuse, or personality disorder, although these may be co-occurring or secondary diagnoses. With agreement and recommendation from the screening review committee, the juvenile’s case is moved from juvenile court to JMHC.

The [juvenile] mental health court requires that each child enter into a contract with the Probation Department to be supervised at the Intensive Supervision Probation (ISP) level. ISP includes much more frequent meetings with the probation officer than ordinary probation, as well as unannounced visits to home and school. The family has to agree to support the child and to the contract’s conditions. They must also be required to participate in family counseling (District Attorney Jefferson and Gilpin Counties, CO, 2009, para. 9).

Once recommended to the JMHC (and with the voluntary agreement of the parents/caregivers), the youth and family have their first of frequent, ongoing hearings for case review with the JMHC judge. The specific expectations of the court are tracked and reviewed by the youth’s probation officer and counselor, who are in contact with the youth, the youth’s school, family, and other parties who are part of the treatment/support team, and their reports are filed with the court for the judge’s evaluations and actions at each hearing. The caseload for the juvenile probation officer who works with youths in the JMHC is a maximum of 25 youths at any given time. Charges are dismissed from the youth’s record if he or she completes the terms and conditions of JMHC. If the youth violates the terms and conditions of JMHC, especially by a new offense or other problematic behaviors, the youth may be discontinued from participation (“unsuccessful”). His or her probation status may be revoked, with the youth returning to juvenile court for the usual processing of the offense(s). However, this is the very last resort and a successful completer may have offenses (“pre-release recidivism”) during their time in the JMHC.

This program evaluation was requested by the Coordinator of the First Judicial District Mental Health Court (SB94). The initial phase of the project began in 2010 when the authors (independent consultants) met with key programmatic stakeholders and staff to gather basic information on the history and characteristics of the JMHC, as well as to identify the types of data that were potentially available in order to answer key questions. By consensus, key questions were identified as follows:

  1. Who participates in the JMHC: demographics, intake diagnoses, and offenses?
  2. What are the recidivism outcomes of participation in the JMHC?
  3. How do these outcomes compare with relevant comparison groups?
  4. Are there predictors of successful or unsuccessful outcomes for youth in the JMHC?

This kind of evaluation not only provides relevant feedback for the particular JMHC, but also offers more empirical data for the general literature on juvenile mental health courts and their outcomes.


Archival Data Collection

After consultation with the JMHC coordinator and advisory committee, key resources for program data were identified by the advisory group and evaluators. JMHC and Department of Probation personnel then made plans to provide the data to the authors. Data from case files and computer databases were provided for youth who had attended JMHC between March 5, 2005 and March 3, 2011.

JMHC Data Sources and Ethical Protection of Information

Plans for and execution of the project followed professional standards of evaluation, focusing on utility, feasibility, propriety, and accuracy (American Evaluation Association, 2004; Yarbrough, Shulha, Hopson, & Caruthers, 2011). The evaluators signed confidentiality agreements regarding the contents of the case records; their access was limited to data that were provided directly by personnel of the First Judicial District. The researchers did not have direct access to any of the district’s or Colorado state’s databases.

The identities of the juveniles were maintained as confidential and the study substituted identification codes for their names. The list of names that matched the codes has been kept in a separate, secure, password-protected computer file of the first author. Similarly, all data files and all written case materials have been maintained in secure, password-protected computer files or in locked cabinets at the office of the first author. Only group data were used for reports to the District and for this article. Written permission to report the findings to professional audiences (e.g., conferences, articles in professional journals) with identification of the District was obtained from the Coordinator of the First Judicial District JMHC on September 9, 2012.

Data Collection for Comparison Groups

A key challenge for any program evaluation is to identify reasonable benchmarks or comparison groups so as to put the findings into perspective. The following groups and sources of information were selected for this study:

  1. Participants in the juvenile mental health court in California (CITA) (Behnken et al., 2009).
  2. Colorado state data for juveniles on probation (Colorado Division of Criminal Justice, 2011b; Colorado Division of Probation Services, 2010; State of Colorado, n.d.).
  3. Juveniles in diversion programs (District Attorney Jefferson & Gilpin Counties, CO, 2012).

Levels and Types of Probation

Adults and juveniles in Colorado can be placed on various types of “regular” probation, which is differentiated from intensive supervised probation. Types of regular probation include administrative probation (a probation officer usually has no direct supervision over the probationer), unclassified probation (there are too few probationers for “meaningful analysis”), and minimum, medium, and maximum levels. As the risk level of the probationer increases, which is determined by preassessment scores, the intensity of supervision increases from minimum to maximum: “Those supervised at the maximum supervision level are considered to be at the highest risk for failure” (Colorado Division of Probation Services, 2010, p. 7).

Results for youth in minimum, medium, and maximum levels of regular probation, JISP, and in diversion programs were considered to be meaningful alternative groups, and we assumed that mental health diagnoses occur in these groups with frequencies similar to those generally reported for youth in the juvenile justice system (Hammond, 2007).

Data Analysis

We computed descriptive statistics for participants’ demographics and outcome variables. Where appropriate, we used t-tests to compare group means across programs. We compared frequencies/proportions using the chi-square test with Yates correction and two-sample z-tests. We used binary logistic regression to evaluate predictors of successful and unsuccessful outcomes in juvenile mental health court. Two-tailed alpha levels were set at .05.


Characteristics of Participants in the Colorado JMHC

We compared demographics for 81 youths who participated in the Colorado JMHC from its inception in 2005 through March, 2011 and who had either successfully or unsuccessfully terminated the program, with the available published data for the general population of Colorado juveniles who were arrested and had their cases adjudicated (Colorado Division of Criminal Justice, 2011b), and with 133 total number of youths who entered the Santa Clara CITA program between 1996 and 2008. As a reminder, the CITA program was selected as a comparison group for two reasons: first, the Colorado JMHC program was modeled on the CITA program; and second, the CITA program was the only report on a JMHC that yielded sufficient and relevant quantitative data.

In general, youths from the Colorado JMHC were younger (M = 14.47 years, SD = 1.62) than the CITA youths (M = 15.0, SD = 1.45; t(212) = 2.48, p = .014, Cohen’s d = .34) and the adjudicated juveniles, according to state estimates (M = 15.5 years; SD not reported). Males comprised the majority of youth in all groups: Colorado JMHC (72%), CITA (66%), and those in statewide probation (76.2%). While White youth comprised the majority of JMHC participants (75%), Whites comprised only 34% of the CITA group (z = 5.75, p < .001). Hispanics comprised the second largest CITA group (33%). The majority of all juvenile arrests in Colorado are White (approximately 82%; Colorado Division of Criminal Justice, 2011b). Thus, it would appear that the JMHC youth were fairly representative of Colorado juveniles, but younger than and not as racially/ethnically diverse as those in the CITA program.

Intake mental health diagnoses. Most youth in both the Colorado JMHC and CITA programs received multiple mental health diagnoses, even within the same general diagnostic category. In fact, the total for various diagnoses in subcategories of mood disorders (e.g., various forms of depression and bipolar disorder) was 112.04% for the CITA group and 95.62% for JMHC youth. Among JMHC youth, anxiety disorder/posttraumatic stress disorder (PTSD) (25.97%) were the second most common diagnoses, followed by attention deficit hyperactivity disorder (ADHD) (20.78%) and developmental disorders (JMHC, 25.97%).

In contrast, a diagnosis of ADHD (60.15%) was considerably more frequent among CITA youth, followed by developmental disorders (38.34%) and anxiety disorder/PTSD (13.53%), respectively. The differences between CITA and JMHC youth in terms of the frequency of ADHD and developmental disorders are best explained by differences in eligibility criteria for the two programs: unlike CITA, JMHC did not consider a single diagnosis of ADHD or developmental disorders as meeting eligibility criteria. To be eligible for the JMHC, Colorado youth were required to have ADHD or a developmental disorder as a co-occurring, rather than a primary, diagnosis. Comparable data are not available for other Colorado youth.

Participant offenses. As a reminder, major (class 1) felony offenses and extremely high-risk youth, such as those who are sexually violent, are not considered eligible for JMHCs or similar probation diversion programs. In order to compare rates of offenses by youth in the JMHC and CITA programs, we grouped offenses into the four major categories used in the CITA report (see Behnken et al., 2009). In general, rates of types of pre-intake offenses did not differ between the two groups: violent/aggressive (CITA: 158 or 408 total offenses, 38.72%; JMHC: 68 of 187 total charges, 36.36%) and miscellaneous (CITA: 127/408, 31.13%; JMHC: 65/187, 34.76%) offenses were the two most common, with property offenses (CITA: 83/408, 20.34%; JMHC: 32/187, 17.11%) and those that were substance-related (CITA: 40/408, 9.8%; JMHC: 22/187, 11.76%) following in frequency (χ(3) = 1.9, n.s). Thus, the background offenses for youth in the two programs were generally similar.

Recidivism Rates for JMHC and Comparison Groups

The following discussion of recidivism is consistent with the terminology and reporting method used by the Colorado Division of Probation Services since 1998: “pre-release recidivism” includes new charges filed for offenses (technical violations or status or nonstatus offenses) that occur while the youth is still participating in the JMHC or another form of probation; “post-release recidivism” includes new charges that are filed after the completion of the JMHC or another form of probation (Colorado Division of Probation Services, 2010). These sources do not include any statistics regarding offending or reoffending that occurred prior to entry into the probation/diversion program.

We compared participants in the Colorado JMHC program with Colorado youth who had been placed on minimum, medium, or maximum levels of “regular” supervision; Colorado probationers who were on JISP; Colorado youth diagnosed with mental disorders who were on probation (levels not specified); and Jefferson County (First Judicial Court District) youth who were in other diversion programs. Post-recidivism rates for Colorado youth were calculated for any new criminal charge within 1 year following successful termination of probation. Unfortunately, comparable post-release recidivism rates were not reported for the CITA youth (Behnken et al., 2009).

Comparing JMHC and CITA participants

Youth in both programs were considered “successful” if they met the program and court requirements, “graduated” from the program, and completed their probation. The JMHC data gave us some information on those who had been “unsuccessful”; that is, those who committed new offenses and were deemed inappropriate to continue in the program. Their cases were then referred back to juvenile court for further action. Of the 18 unsuccessful JMHC youth, six were placed directly into custody of the Colorado Division of Youth Corrections.

CITA reported only the number of offenses during participation in the program for their successful completers (M = 1.14, SD = 1.41, range: 0-6 offenses). Although the mean number of offenses during the program was higher for those who completed JMHC successfully (M = 1.31, SD = .82, range: 0-3 offenses), the difference was not statistically significant (t(123) = .82, n.s.).

Comparing JMHC and other Colorado programs

Recidivism reports for youth in other Colorado programs considered the number of youth in each probation program who incurred new charges, rather than the number of new charges, and included those who were not successful completers. Data are summarized in Table 1, with programs listed from lowest to highest for recidivism rates.

Table 1. Comparison of Recidivism Rates for JMHC with Other Colorado Juvenile Probation Programs

Probation Program N of All Participants N Who Reoffended % Who Reoffended Comparison of % with JMHC



Reg. Minimum3




z = - 11.04, p < .0001

Reg. Medium3




z =   - 3.22, p = .0013

JC Diversion5




z = - 1.31, n.s.






M H Dx on Probation4




z = .81, n.s.   





z = 2.84, p = .005

Reg. Maximum3




z = 3.3, p < .0001











Reg. Minimum3




z = 1.44, n.s.

Reg. Medium3




z = 2.26, p = .024





z = 2.64, p = .008

Reg. Maximum3




z = 2.99, p = .003

M H Dx on Probation4




z = 2.87, p = .0041





z = 3.60, p = .0003

1 These offenses occurred while the youth was in the respective probation program. The reports of the various Colorado probation programs indicate the proportions of youths who were terminated from the probation status (“pre-release failure”) due to pre-release offenses, making them unsuccessful completers of the particular program (Colorado Division of Probation Services, 2012). However, pre-release offenses did not automatically lead to termination for the JMHC or CITA youth.

2 Post-release recidivism rates are for successful completers of all Colorado probation programs, including the JMHC. All Colorado groups are for a period of 1 year following successful completion of the probation program.

3 Mean annual rates for 2005-2011 reports computed from Colorado Division of Probation Services (Tables 11 and 12; 2012).

4 Estimate based on random sample of 20% of cases from 2001 to 2007, provided by Probation Department from internal records.

5 Pre-release recidivism and unsuccessful completion of program only available for 2010; post-release recidivism for successful completers is an average of the annual rates provided for 2006 to 2009 cohorts (District Attorney Jefferson and Gilpin Counties, CO, 2012).

6 These figures are for youths who participated and terminated (successfully, 61, or unsuccessfully, 18) from JMHC from 2005-2011. During the post-release period, one additional youth had a charge filed after completion of JMHC, but the offense was committed prior to entry to JMHC.

Pre-release recidivism rates. Recidivism rates for JMHC youth while on probation fell between recidivism rates for the other Colorado groups, being statistically significantly higher than those for youth on minimum (p < .0001) or medium (p = .0013) regular probation, but statistically significantly lower than youth on JISP (p = .005) or maximum regular probation. While higher than the rate for youth in other diversion programs, and lower than the rate for juveniles diagnosed with mental disorders not in JMHC, these differences did not reach statistical significance.

Post-release recidivism rates. By contrast, successful completers of JMHC fared better than all other Colorado groups on post-release recidivism. Only the difference in recidivism rates with youth who had been on minimum regular probation failed to reach statistical significance.

Changes in Offense Patterns

One of the notable findings of this study is that there were significant declines in the incidence of violent/aggressive (χ2(2) = 24.3, p < .0001) and property/theft offenses (χ2(2) = 12.3, p < .01) from intake to JMHC to while active in JMHC, and following successful completion of JMHC. Changes in miscellaneous and substance-related offenses were not significant (see Figure 1). However, ongoing substance use/abuse problems by a few youths accounted for a notable portion of the post-release offenses: of the 24 post-release charges reported for the 16 youths who were successful completers of the JMHC, 10 (45%) of the charges were drug/alcohol related and committed by only six (37.5%) of these 16 youths.

Figure 1. Frequency and types of offenses at intake, during, and following participation in the Colorado JMHC for successful completers.

Figure showing Frequency and types of offenses at intake, during, and following participation in the Colorado JMHC for successful completers

Similar declines in the frequency of violent/aggressive and property/theft pre-release offenses were reported for the CITA program. The 64 successful CITA graduates also showed significant decreases in both the frequency of violent/aggressive offenses at intake (65) and during participation in the program (16; χ2(1) = 28.44, p < .0001), and in the number of property/theft offenses from intake (39) to participation in the program (8; χ2(1) = 19.14, p < .0001). The frequency of miscellaneous offenses did not change (intake = 45; during = 43 offenses). The CITA program did not report these data for substance-related offenses (Behnken et al., 2009, Table 5).

Predictors of Outcomes for JMHC Participants

Because the criterion variable is dichotomous (successful, unsuccessful completion of JMHC), we used a simultaneous logistic regression to model participants’ outcomes. The predictor variables were: age (in years) at intake screening; race/ethnicity (White, nonWhite); intake primary diagnostic code (bipolar, mood disorder, anxiety/PTSD, other); previous Individualized Education Plan (IEP; yes/no); primary intake charge code (violent/aggressive, property/theft, miscellaneous); as well as length of time in JMHC, number of charges during JMHC, and number of court reviews during JMHC (all continuous scale of measurement).

Results of the logistic analysis indicate that the predictor model provided a statistically significant improvement over the constant-only model (χ2(9) = 28.92, p = .001). The model accounted for 57.2% of the total variance, indicating that the model discriminates between successful and unsuccessful participants in the JMHC. Prediction success was relatively high, with an overall prediction success rate of 82%, and 91.5% correct for successful completers and 50% correct for unsuccessful completers. Table 2 presents the regression coefficients (B), the Wald statistics, significance levels, and odds ratios [Exp(B)]. Only number of reviews and number of new charges during JMHC were statistically significant predictors (p = .005 and p = .003, respectively); the number of days in JMHC was a meaningful predictor, although it did not reach statistical significance (p = .063).

Table 2. Logistic Regression Results for Predictors of Successful or Unsuccessful Completion of JMHC (Dependent Variable = Successful)

Predictor B S.E. Wald df Sig. Exp.(B)















Intake Diagnosis














Intake offenses







Intake offenses (1)







Intake offenses (2)







Days in JMHC







Number of reviews







Number of new offenses during JMHC







Additional t-tests found statistically significant differences between successful and unsuccessful participants on all three of these variables: number of days in JMHC, successful participants (M = 301.2, SD = 133) versus unsuccessful (M = 407.7, 174.8; t(77) = - 2.77, p = .007; number of new charges while in JMHC, successful (M = .4, .7) versus unsuccessful (M = 1.9, 1.0; t(75) = - 7.20, p = .000); and, number of reviews while in JMHC, successful (M = 11.8, 6.9) versus unsuccessful (M = 20.1, 11.7; t(66) = - 2.61, p = .018). Thus, those who were ultimately unsuccessful had more charges, more court reviews of their cases, and were in JMHC longer than those with successful outcomes. This highlights the program’s commitment to working with the youth as far as possible to achieve success.

Summary and Discussion

The results of this program review clearly support the efficacy of the Colorado First Judicial District’s juvenile mental health court for reducing both recidivism while in the program and during at least 1 year following successful completion. There also is preliminary evidence that post-release recidivism rates of youth in the JMHC were significantly lower than those for a sample of other youth in the juvenile justice system diagnosed with mental disorders.

A contribution of this evaluation is that we considered the outcomes of this program in the context of other relevant comparison markers, including another juvenile mental health court program (CITA; Behnken et al., 2009) and Colorado youths in other probation programs (Colorado Division of Probation Services, 2010; District Attorney Jefferson & Gilpin Counties, Co., 2012; state of Colorado, n.d.). This allowed us to consider the breadth of risk levels of the youth and the related levels of supervision during probation.

Results of this kind of research help us to compare the effectiveness of JMHCs with the effectiveness of adult mental health courts. For example, we can now note that graduates of JMHC show significant post-release reduction in violent offenses, which also are reported for successful completers of adult mental health courts (McNiel & Binder, 2007).


Study limitations are common for program evaluations that rely on archival, rather than prospectively planned, data; that is, some data may not be documented sufficiently or may be unavailable (Posavac & Carey, 2007). One notable example of such a limitation was not having information on the pre-entry offense histories of youth who participated in the JMHC or for those in other probation programs in Colorado. Another limitation that may affect the interpretation of results is the possible impact of nonrandom assignment by JMHCs; participation is voluntary. It is possible that youth and/or their families who participate in JMHC differ in their motivation for treatment or some other unknown variable from those who are eligible but choose not to participate. Regrettably, there were no data for a comparison sample of youth who were eligible for the JMHC but chose not to participate. In fact, most families who were accepted to JMHC did participate, unless they left the district or otherwise became ineligible. Furthermore, our case study focused on one judicial district in one state. It is difficult to know how these findings may generalize to other locations and other programs.

While we interpreted our evaluations in comparative contexts, there is always the risk of variations in diagnostic procedures, eligibility criteria, and definitions of recidivism that may limit direct comparison (Cocozza & Shufelt, 2006). Although it was clear that successful completion of JMHC was positively related to decreased recidivism, we were not able to identify predictors of successful completion. Was it a matter of youth or family readiness, or was it due to other support systems and resources (e.g., Cottle, Lee, & Heilbrun, 2001)?

We also were short on quantifiable information on treatment compliance and the juveniles’ use of mental health services, a second criterion typically noted for evaluating outcomes of JMHC treatment (McNiel & Binder, 2007; Office of Juvenile Justice, n.d.). We discovered that although the JMHC probation officer and counselor/navigator kept case notes on treatment activities and clearly followed up with the youth and families regarding care, there was no systematic central record from which these details were readily available for external analysis. Recommendations have been made to the First Judicial District for ways to improve prospective data collection and maintenance for future evaluations of this JMHC. Improved data collection will allow evaluators to consider possible predictors of outcome, such as successful completion of JMHC and recidivism rates.

Recommendations for Future Research

Procedures for evaluation of mental health courts have been described elsewhere (e.g., Steadman, 2005). Clearly, a key recommendation is preplanning for prospective data collection aimed at offering relevant, usable, and comparable data, both for participants and nonparticipants of JMHC. Systematic, even standardized, operational definitions and measurements of key variables need to guide information gathering and data management.

Although randomized assignment of eligible youth to the JMHC as opposed to other post-adjudication options usually is not possible, systematic data for viable comparison groups—including juveniles who are screened and are eligible, but who choose not to participate in the juvenile mental health court—offer opportunities for quasi-experimental designs using matching methods, with regression-based model adjustments, which could be used for reliable estimation of causal effects (Stuart & Rubin, 2008). It is important to interpret the outcomes of JMHCs within the larger context of how other youth in the juvenile justice system are faring, especially those who are diagnosed with mental disorders.

State, local, and even national resources should keep the same kind of details on the dispositions and outcomes for youth who are diagnosed with mental disorders and eligible for JMHCs as they do for the general population of youth who enter the juvenile justice system. This would help researchers compare outcomes for youth when they enter various levels of regular probation, JISP, and other diversion programs. Similarly, we need to be able to track recidivism, compliance, and other meaningful indicators for juveniles who are eligible for and enter JMHC, but who do not successfully complete the program.

While the current focus of research seems to be on recidivism, more attention needs to be given to collecting and managing meaningful data on mental health status and use of services for youth diagnosed with mental disorders, and comparing their offense histories and rates prior to entry, during, and post-release from JMHC. The ACJC report (National Center for Youth Law, 2011) offers one attempt to do this with juveniles, but with retrospective data. Examples from research on adult mental health courts can provide guidance. Similar also to research on adult mental health courts, evaluations should begin to study multiple sites (Callahan & Wales, 2013). This helps to identify not only generalizable findings, but also those that may be program—or location—specific.

About the Authors

Donna M. L. Heretick, PhD, is a social and clinical psychologist and president of Strategies for Solutions Consulting, located in Arvada, Colorado. In addition to research and consulting, Dr. Heretick teaches and mentors graduate students in doctoral psychology programs at several universities.

Joseph A. Russell, MA, is completing a PsyD in clinical psychology at Argosy University, Denver, Colorado, after more than 20 years in law enforcement.


American Evaluation Association. (2004). Guiding principles for evaluators. Retrieved from http://www.eval.org/p/cm/ld/fid=51

American Psychiatric Association. (2000). Diagnostic and statistical manual for mental health disorders (4th ed., text revision). Washington, DC: Author.

Behnken, M. P., Arredondo, D. E., & Packman, W. L. (2009). Reduction in recidivism in a juvenile mental health court: A pre- and post-treatment outcome study. Juvenile and Family Court Journal, 60(3), 23–44.

Callahan, L., Cocozza, J., Steadman, H. J., & Tillman, S. (2012). A national survey of U.S. juvenile mental health courts. Psychiatric Services, 63(2), 130–134.

Callahan, L., & Wales, H. (2013, Mar. 26). Mental health court research roundup: Applying research to practice. [Powerpoint/Webinar presentation.]. Washington, DC: Justice Center of the Council of State Governments. Retrieved from http://consensusproject.org/features/webinar-archive-mental-health-courts-research-roundup-applying-research-to-practice

Center on Juvenile and Criminal Justice. (1999, Sept.). Diversion programs: An overview. Juvenile Justice Bulletin. Retrieved from www.ncjrs.gov/html/ojjdp/9909-3/div.html

Cocozza, J. J., & Shufelt, J. L. (2006). Juvenile mental health courts: An emerging strategy. Research and Program Brief, National Center for Mental Health and Juvenile Justice. Retrieved from www.ncmhjj.com/pdfs/publications/JuvenileMentalHealthCourts.pdf

Colorado Division of Criminal Justice (2011a). Compliance monitoring policy and procedure manual. Retrieved from http://dcj.state.co.us/oajja/ComplianceMonitoring/Compliance_Monitoring.html

Colorado Division of Criminal Justice (2011b). Crime and justice in Colorado, 2008-2010. Retrieved from http://dcj.state.co.us/ors/pdf/docs/CJ08-10.pdf

Colorado Division of Probation Services. (2010, October 20). Pre-release termination and post-release recidivism rates of Colorado’s probationers: FY2009 releases. Denver, CO: Evaluation Unit, Division of Probation Services, State Court Administrator’s Office, Colorado Judicial Branch. Retrieved from http://www.courts.state.co.us/Administration/Custom.cfm?Unit=eval&Page_ID=189

Cottle, C. C., Lee, R. J., & Heilbrun, K. (2001). The prediction of criminal recidivism in juveniles: A meta-analysis. Criminal Justice and Behavior, 28(3), 367–394.

District Attorney Jefferson & Gilpin Counties, Co. (2012). Juvenile diversion statistics. Retrieved from http://jeffco.us/da/da_T99_R284.htm

District Attorney Jefferson & Gilpin Counties, Co. (2009). Juvenile mental health court. Retrieved from http://jeffco.us/da/da_T99_R203.htm

District Attorney Jefferson & Gilpin Counties, Co. (2013). Juvenile mental health court. Retrieved from http://jeffco.us/da/da_T99_R307.htm

Hammond, S. (2007). Mental health needs of juvenile offenders. Denver, CO: National Conference of State Legislators. Retrieved from www.ncsl.org/print/cj/mentaljjneeds.pdf

McNiel, D. E., & Binder, R. L. (2007). Effectiveness of a mental health court in reducing criminal recidivism and violence. American Journal of Psychiatry, 164(9), 1395–1403.

McReynolds, L. S., Schwalbe, C. S., & Wasserman, G. A. (2010). The contribution of psychiatric disorder to juvenile recidivism. Criminal Justice and Behavior, 37, 204–216.

National Center for Youth Law (NCYL). (2011). Improving outcomes for youth in the juvenile justice system. Oakland, CA: author. Retrieved from http://consensusproject.org/announcements/report-on-juvenile-mental-health-court-suggests-promising-outcomes

North Carolina Department of Juvenile Justice and Delinquency Prevention (n.d.). Risk and protective factors. Retrieved from http://www.ncdjjdp.org/community_programs/gsm/risk_factors.html

Office of Juvenile Justice and Delinquency Prevention, OJJDP. (n.d.). Mental health court. Retrieved from http://www.ojjdp.gov/mpg/progTypesMentalHealthCourt.aspx

Pogrebin, M. R., Poole, E. D., & Regoli, R. M. (1984). Constructing and implementing a model juvenile diversion program. Youth and Society, 15(3), 305–324.

Posavac, E. J., & Carey, R. G. (2007). Program evaluation: Methods and case studies (7th ed.). Upper Saddle River, NJ: Pearson/Prentice Hall.

Roberts, A. R., & Bender, K. (2006). Overcoming Sisyphus: Effective prediction of mental health disorders and recidivism among delinquents. Federal Probation, 70(2). Retrieved from http://www.uscourts.gov/uscourts/FederalCourts/PPS/Fedprob/2006-09/sisyphus.html

SAMHSA’s Gain Center. (n.d.). Juvenile mental health treatment courts database. Retrieved from http://gainscenter.samhsa.gov/grant_programs/juvenilemhc.asp

Shufelt, J. L., & Cocozza, J. J. (2006). Youth with mental health disorders in the juvenile justice system: Results from a multi-state prevalence study. NCMHJJ Research and Program Brief. Retrieved from www.unicef.org/tdad/usmentalhealthprevalence06(3).pdf

Snyder, H. N., & Sickmund, M. (2006). Juvenile offenders and victims: 2006 national report. Washington, DC: National Center for Juvenile Justice. Retrieved from http://www.ojjdp.gov/ojstatbb/nr2006/downloads/NR2006.pdf

State of Colorado. (n.d.). Annual recidivism study reports. Retrieved from http://www.courts.state.co.us/Administration/Custom.cfm?Unit=eval&Page_ID=189

Steadman, H. J. (2005). A guide to collecting mental health court outcome data. New York: Council of State Governments.

Stuart, E. A., & Rubin, D. B. (2008). Best practices in quasi-experimental designs: Matching methods for causal inference. In J. W. Osborne, Best practices in qualitative methods (Chap. 11). Thousand Oaks, CA: Sage.

U.S. Census Bureau. (2012). Law enforcement, courts, & prisons: Arrests. Retrieved from http://www.census.gov/compendia/statab/cats/law_enforcement_courts_prisons/arrests.html

Wales, H. W., Hiday, V. A., & Ray, B. (2010). Procedural justice and the mental health court judge’s role in reducing recidivism. International Journal of Law and Psychiatry, 33(4), 265–271.

Yarbrough, D. B., Shulha, L. M., Hopson, R. K., & Caruthers, F. A. (2011). The program evaluation standards: A guide for evaluators and evaluation users (3rd ed.). Thousand Oaks, CA: Sage.

OJJDP Home | About OJJDP | E-News | Topics | Funding
Programs | State Contacts | Publications | Statistics | Events