Forging new paths in the development of community mental health interventions for people with mental illness at risk of criminal legal system contact

Abstract

Background

Individuals with serious mental illness (SMI) have disproportionately high rates of criminal legal system involvement. For many, this becomes a repeated cycle of arrest and incarceration. Treatments that address symptoms of mental illness are a critical component of the continuum of services for people with SMI in the legal system; yet on their own, psychiatric treatments have not been successful at reducing criminal legal system involvement for this population. Research suggests that criminogenic risk factors, the major drivers of criminal legal system involvement, are disproportionately prevalent among individuals with SMI. However, promising criminogenic-focused interventions have only just begun to be adapted for individuals with SMI. The proposed study will examine the capacity of Forging New Paths (FNP), a novel criminogenic-focused group intervention developed for individuals with SMI, to engage its primary and secondary outcomes when delivered in community mental health settings.

Methods

The proposed pilot study will engage a small-scale clinical trial comprising three cycles of FNP delivered in a community mental health center in a Southeastern state in the U.S. The anticipated total sample size is N = 72 and will consist of community-dwelling adults with SMI who have a moderate or higher criminogenic risk level and a history of criminal legal system contact. This study will examine the extent to which FNP is able to engage its primary (aggression and community tenure) and secondary (criminal attitudes and impulsivity) treatment outcomes.

Discussion

FNP provides an important new service for community-based mental health settings to reduce criminal legal system involvement (and recidivism) among the individuals they serve with SMI.

Clinical Trial Registration

NCT06290648.

Introduction

People with serious mental illness (SMI), such as schizophrenia, bipolar disorder, or major depressive disorder, are overrepresented in the criminal legal system despite decades of interventions designed to connect this population to mental health services (Crilly et al., 2009; Ditton, 1999; Prins, 2014; Steadman et al., 2009; Teplin, 1990; Teplin et al., 1996). Once involved over 60% of people with SMI return to incarceration within 3 years of release and 50% returning within the first year (Baillargeon et al., 2010; Cloyes et al., 2010; Hartwell, 2004; Lovell et al., 2002; Messina et al., 2004; (A) Wilson et al., 2014; A. (B) Wilson et al., 2011).

Traditional mental health services are a critical component of care for people with SMI, yet approaches that solely target mental health symptomatology have not been shown to significantly impact the risk of criminal legal system-involvement among this population (Morrissey et al., 2007; Osher & Steadman, 2007; Skeem et al., 2011, 2014). Researchers have posited that the lack of success of these “first generation” interventions to prevent recidivism for people with SMI may stem from the fact that they were developed in isolation from the larger body of correctional research, which has demonstrated that people with mental illness face the same risk factors for criminal legal system involvement as those without mental illnesses (Bonta et al., 1998, 2013).

A robust body of research shows that the strongest risk factors for criminal legal system involvement (i.e., “criminogenic” risk factors) among people with and without SMI include antisocial behavior, antisocial personality, antisocial cognitions, antisocial associates, substance use, family conflict, school or work problems, and lack of pro-social leisure activities (Andrews et al., 1990, 2006; Andrews & Bonta, 2010; Bonta & Andrews, 2016). Furthermore, mounting evidence suggests that compared their counterparts without SMI, legal system-involved people with SMI have higher overall levels of criminogenic risk factors, as well as higher levels of specific criminogenic risk factors such as criminal thinking and attitudes (Girard & Wormith, 2004; Skeem et al., 2014; Wilson et al., 2020). Moreover, recent research suggests that these criminogenic risk factors may mediate the risk of recidivism for people with SMI, making them potentially potent treatment targets for interventions that seek to reduce criminal legal system involvement among this population (Matejkowski & Ostermann, 2015; Morgan et al., 2010; Wilson et al., 2014; Wolff et al., 2011, 2013).

Primary studies and meta-analyses of correctional programing have found that the most effective interventions that target criminogenic risk factors adhere to the treatment principles of the Risk-Need-Responsivity model (RNR; (Bonta & Andrews, 2016), which identifies the Risk Principle (i.e., that treatment resources should focus on individuals with high criminogenic risk levels); The Need Principle, which asserts that interventions will be most effective at reducing criminal recidivism when they engage treatment targets most strongly associated with criminal behavior; and the Responsivity Principle, which specifies that interventions will be most effective when their delivery is designed to match the learning styles and abilities of offenders (Bonta & Andrews, 2016). Interventions that adhere to the RNR model have been shown to reduce recidivism by 20–55% among the general population of system-involved individuals (Bonfine et al., 2020; Bonta & Andrews, 2016; Landenberger & Lipsey, 2005; Lipsey et al., 2001; Pearson et al., 2002; Smith et al., 2009). However, existing programs based on the RNR model are typically designed for the general population, presenting obstacles for engagement for individuals with SMI, who often face neurocognitive and social impairments related to their mental illness, that hinder their ability to fully participate in such interventions (Bellack et al., 1999; Hogarty & Flesher, 1999; Wexler & Bell, 2005; Wilson et al., 2018). To address this critical gap in care, our research team previously developed the Targeted Service Delivery Approach (TSDA; Wilson et al., 2018) to adapt existing evidence-based correctional interventions for use with people with SMI in prison.

The TSDA includes a set of therapeutic strategies used during the delivery of existing evidence-based intervention to ensure the intervention’s responsivity to the specific learning needs of people with SMI (Golden et al., 2006; Morgan et al., 2010; Skeem et al., 2009; Wilson et al., 2018). A recent clinical trial has demonstrated that the TSDA can be used to adapt the delivery of existing evidence-based criminogenic-focused interventions, like Thinking for a Change (Bush et al., 1998), for people with SMI incarcerated in prison settings (Wilson et al., 2023). However, in order for these interventions to have the strongest and broadest impact possible on reducing legal involvement among people with SMI, they must also be adapted for use in community mental health settings– where the majority of people with SMI receive treatment services.

To enhance the availability of interventions for people with SMI that target risk factors for criminal legal system to community mental health settings, our team engaged user-centered design research methods (Lyon et al., 2019; Lyon & Koerner, 2016) to develop Forging New Paths (FNP), a new criminogenic-focused intervention designed specifically for use with people with SMI in community-based mental health settings. Historically, researchers struggled to adapt correctional interventions for use in community settings for people with SMI, at least in part, because retrofitting or restructuring existing correctional interventions is resource intensive (Andrews et al., 1990; Wilson et al., 2017), and copyright restrictions have made widely used correctional interventions either cost prohibitive (e.g., Reasoning & Rehabilitation; Young et al., 2010) or unavailable for use outside of correctional settings (e.g., Thinking for a Change; Bush et al., 2011). However, as 50% of people with SMI receiving mental health care in the community have had some criminal legal system involvement (Hall, 2004), making these interventions available in community-based mental health settings is critical in terms of maximizing their reach and potential benefit.

This paper presents the study protocol for a National Institute of Mental Health funded clinical trial of FNP. The primary objective of this clinical trial is to engage a preliminary evaluation of the impact of FNP on two primary treatment outcomes: levels of aggression and community tenure. The secondary objective is to assess FNP’s capacity to reduce criminogenic risk (i.e., criminal attitudes and levels of impulsivity) and explore the extent to which changes in the study’s primary outcomes are mediated through its secondary outcomes. See Fig. 1 for an illustration of the study’s primary and secondary objectives.

Fig. 1

Primary and secondary objectives for forging new paths

Materials and methods

Study design

This study engages a pilot effectiveness trial of FNP to examine its impact on the primary and secondary outcomes outlined in Fig. 1. Eligible participants will be randomized into one of two study conditions: FNP and usual care, or usual care alone. Participants will be involved with the study for 9 months, during which time data will be collected through up to four face-to-face interviews as well as via administrative data. Up to 72 participants will be recruited over the course of three separate intervention cycles. The study will be active for 4 years. This clinical trial (NCT06290648) was registered with the Clinical Trials Registry at clinicaltrials.gov on February 26, 2024.

Setting and participants

The study will recruit participants through a community mental health center located in a Southeastern state in the U.S. Inclusion criteria for participation includes (a) aged 18 years or older; (b) have a diagnosis of schizophrenia spectrum and other psychotic disorders or major depression or bipolar disorder; (c) have a history of any type of criminal legal system involvement (i.e., arrest, conviction, incarceration, or probation/parole); (d) reside in the community and receive services from the community mental health center or one of their partner agencies; (e) and have moderate or higher levels of criminogenic risk factors as determined by the Level of Service and Case Management Inventory (LS/CMI; Girard & Wormith, 2004). Exclusion criteria includes (a) having an intellectual or developmental disability; (b) being incarceration at the point of study enrollment, (c) non-English speakers.

Treatment allocation

After eligibility has been determined via a screening interview, the study team will use REDCap to randomize participants into one of two study conditions: FNP plus usual care, or usual care alone. Block randomization procedures will be used to ensure an equal distribution of subjects to each condition. Given the nature of the study, neither study staff nor participants will be blinded to participant condition. Results of the recruitment process will be documented in the form of a CONSORT flow chart.

Control condition

Participants assigned to the control condition will receive usual care, which includes the full range of available community-based mental health services that they are otherwise eligible to receive during the study period. All decisions about the usual care received will be made by community mental health center staff. These services could include, but are not limited to, medication management, outpatient treatment, Assertive Community Treatment, supported employment services, integrated health services, housing navigation, and other supportive services and crisis and emergency services.

Experimental intervention

FNP is designed to enhance existing mental health services. Therefore, participants assigned to the experimental condition will receive FNP in addition to their usual care, which includes the full range of community mental health services described above.

Overview of Forging new paths(FNP)

User-centered design principles were utilized throughout the development of this new intervention to ensure its efficacy and usability in community mental health treatment settings (Lyon et al., 2019; Lyon & Koerner, 2016). Research from the RNR model (Bonta & Andrews, 2016) informed the selection of treatment targets for this newly developed intervention (impulsivity, interpersonal problem solving, and criminal attitudes and associates) and the therapeutic principles associated with the TSDA informed the development of services delivery strategies for the new intervention to ensure it is responsive to the learning needs of people with serious mental illnesses at risk of criminal legal system involvement.

Consistent with the RNR model, FNP leverages cognitive behavioral therapy techniques and social learning methods (Bandura, 1977) that provide opportunities to learn through observation and hands on practice in a group format to treat dynamic criminogenic risk factors that have been found to have a strong relationship to criminal legal system involvement (Bonta & Andrews, 2016). The content of FNP is organized into five treatment modules, illustrated in Fig. 2.

Fig. 2

The five treatment modules of forging new paths

A modular treatment format was used to create a flexible intervention structure to optimize the usability of FNP in community mental health settings. FNP’s five treatment modules include a total of 14 treatment sessions, outlined in Fig. 3.

Fig. 3

Treatment session overview for forging new paths

FNP is designed to be delivered to 8 to 12 participants by two intervention facilitators once a week in community mental health settings. Each treatment session lasts 60 min and can be delivered in-person, virtually, or in a hybrid format. To align with the RNR model’s Responsivity Principle (Bonta & Andrews, 2016), FNP’s intervention facilitators use the therapeutic strategies associated with the TSDA (Wilson et al., 2018) during the delivery of treatment content as compensatory strategies that focus on supporting participants’ ability to fully engage in and benefit from the intervention’s content. The TSDA provides facilitators with the opportunity to use their clinical judgement related to how and when to use the TSDA therapeutic strategies to optimize participants’ ability to engage in and benefit from intervention content; how to pace the delivery of the session content to meet the needs of the group; and when to utilize additional therapeutic strategies outlined in the TSDA to support participant engagement. More details on the TSDA can be found in (Wilson et al., 2018).

Recruitment and informed consent

Participant recruitment will begin approximately one month prior to the initiation of each intervention cycle. Staff of the community mental health center and their partner agencies will assist study staff with recruitment efforts by distributing study flyers to potentially eligible participants. To maximize recruitment efforts, potentially eligible participants who express interest in learning more about the study will be provided with several options for scheduling a meeting with a member of the research team for further recruitment activities. They can use the contact information on the study flyer to call or email the study team or use an electronic link to a Qualtrics survey provided in the study flyer to schedule a meeting with a research team member. Additionally, when a member of the research team is available onsite, potentially eligible participants can schedule further recruitment activities. Once a meeting has been scheduled, a member of the research team will meet with potentially eligible participants for the study intervention in a private community-based setting or through a secure virtual platform, such as Zoom, to complete recruitment and obtain informed consent.

Data collection

Study participants will complete up to four face-to-face research interviews, which will take place in private community-based settings or virtually. The first study interview is the screening interview where information about participant’s demographic, clinical and criminal justice characteristics are collected to determine eligibility for participation. Participants who meet the eligibility criteria for the study intervention will be randomized to the experimental or control condition. Participants randomized to either study arm will complete up to three additional interviews: the baseline interview and two follow-up interviews. The baseline interview will occur at the beginning of each intervention period. The first follow up interview will occur at the completion of the study intervention, and the second follow-up interview occurs 3 months post intervention completion. Additionally, administrative data related to community tenure will be collected 6 months after intervention completion. Table 1 outlines the data collection schedule for the study’s primary and secondary outcomes and Table 2 provides descriptions of the study measures that will be used to assess each outcome. The measurement plans outlined in Tables 1 and 2 build on those used in prior published research (Wilson et al., 2023). All study measures are administered through face-to-face interviews, unless otherwise noted in Table 2. Community tenure, conceptualized as stability of community living, is a primary outcome in this study (Dixon, 2000; Hadley et al., 1993; Joannette et al., 2005). Given that this is a pilot feasibility study of a new intervention being delivered in a new treatment setting, the rates of hospitalization and incarceration for study participants are yet to be established. Therefore, community tenure will be measured through a dichotomous variable, described in detail in Table 2, which indicates the presence of any disruption in community tenure during the measurement period.

Table 1 Timeline for data collection

Table 2 Study outcome measure descriptions

Data analysis

Analysis of primary and secondary outcomes will be conducted on an intent-to-treat basis. Community tenure during the post-intervention period will be measured dichotomously as described in Table 2; therefore, chi-square tests and logistic regression models will be used to examine between-group differences in this outcome. Levels of impulsivity, criminal attitudes, and levels of aggression will be measured continuously and collected at both baseline and follow-up periods. Mixed effects models will be used to examine changes in these primary and secondary outcomes. Mixed effects models can account for data collection timepoint (e.g., baseline and follow-up timepoints) nested within participants. Each model will include a fixed effects term for Treatment Arm, Time, and Treatment Arm*Time interaction. This approach allows for the examination of (1) estimated mean change on continuously measured outcome variables from baseline to follow-up periods within both the intervention and control group and (2) the difference of this estimated mean change over time between the intervention and control group (i.e., the “treatment effect”). Mixed effects models in conjunction with sensitivity analyses will be used to address missing data in analyses (Chakraborty & Gu, 2009; Little et al., 2012). Additionally, the extent and nature of missing data will be assessed to inform the design of future studies.

Exploratory analyses using structural equation modelling (SEM) and generalized structural equation modelling (GSEM) will examine the mediating effects of the secondary study outcomes on the primary study outcomes. Refer to Fig. 1 for the study’s conceptual model.

Sample size and power calculation

Given the preliminary nature of this pilot study, our goals are not to fully test our hypotheses nor estimate effect sizes associated with FNP. Rather, our main purpose is to collect data on available sample size, sample attrition and sample retention, as well as to generate the necessary data to conduct a comprehensive power analysis for a future R01 grant.

Data Management, sharing, and safety monitoring

All electronic data will be maintained in secure university platforms and access will be restricted to IRB approved project staff. All paper copies of consent documents and interview forms will be filed securely in locked file cabinets in locked offices restricted to research staff. Study data will be periodically uploaded to the National Institute of Health’s National Data Archive as per this study’s Data Sharing Agreement with the study funder.

The study PI will oversee the safety of the study. Research data will be reviewed in a timely manner. The PI will comply with all existing policies for both documenting and reporting unanticipated or adverse events associated with this study. In addition, an Independent Safety. Monitor (ISM) will be utilized for the clinical trial. The ISM does not have any professional or financial conflicts of interest with this research project or study investigators. The ISM will review the study protocol prior to the enrollment of the first participant in the RCT and will also develop a charter that outlines which data points will be monitored, how they will be assessed, and a schedule for monitoring. The scope of monitoring for the RCT will include enrollment data, data on participant safety, and data integrity.

Ethical approval and informed consent

All research procedures have been approved by an Institutional Review Board (IRB) at the University of North Carolina at Chapel Hill. Research staff will engage informed consent with all potential participants who express an interest in participation in the study during recruitment. Participants will also be asked to sign a HIPAA authorization form allowing the study team to collect administrative data pertaining to the study outcomes. All researchers involved in recruitment will be trained to cover consent information clearly and comprehensively. Furthermore, study flyers, consent forms, and HIPAA authorization forms will be made available to all potential participants. While this study includes individuals with current and/or prior criminal legal system involvement, individuals who are incarcerated will not be enrolled or assigned to study arms.

Multiple strategies will be employed to reduce the potential risk of coercion. The informed consent document will clearly state that participation in this study is voluntary, can be discontinued at any time, and will not impact the participant’s status in the criminal legal system or any other community mental health services they may be eligible to receive. The recruitment and consent process will emphasize that information about participation will not be shared with either criminal legal system authorities or community mental health center staff. Finally, the study also has a certificate of confidentiality.

Discussion

The study tests the ability of an innovative intervention to engage novel treatment targets for people with SMI who have moderate or high criminogenic risk levels and a history of criminal legal system involvement in a potentially high impact setting. Findings from this research will lay the groundwork for larger-scale testing of FNP in an R01 grant that will engage a Type 2 hybrid implementation-effectiveness trial (Curran et al., 2012), thus broadening the public health impact of our research. If successful, FNP has the potential to address criminogenic risks in the community and prevent future criminal legal system involvement before people with SMI re-engage with the criminal legal system. Without such programming, people with SMI might only be able to receive criminogenic programming within carceral settings and from programs that may not be able to address their neurocognitive and social impairments.

Aside from expanding the reach of needed interventions for people with SMI with criminal legal system involvement, there are a number of other potential benefits to delivering criminogenic-focused interventions for system-involved people with SMI in community-based mental health settings. For example, community-based interventions can harness existing relationships between service providers and clients, make use of therapeutic practices already present in community mental health services that compliment effective criminogenic-focused interventions (e.g., cognitive restructuring techniques), and provide an opportunity to intervene earlier in people’s trajectory of criminal legal system involvement to prevent initial engagement or re-engagement with this system (Bonfine et al., 2020). Tailored, evidence-based interventions such as FNP can help the community mental health system to become the “Ultimate Intercept” by equipping it with a new service needed to address the complex needs of people with SMI in an effort to prevent criminal legal system involvement (Bonfine et al., 2020).