Jail-to-community treatment continuum for adults with co-occurring substance use and mental disorders: study protocol for a pilot randomized controlled trial

The objectives of the current study are to (1) develop research protocols to obtain access to a jail population, implement a rigorous research design, including randomization to a pilot trial, and ensure that participants do not perceive any undue coercion to participate; (2) develop communication protocols so that jail and community treatment are coordinated and linked for effective community re-entry; and (3) adapt the clinical content of extant evidence-based programs to better address both behavioral health and criminogenic needs of jail inmates with CODs, including anger management, harm reduction, illness management, criminal thinking, antisocial peer networks, and treatment motivation. The following research aims were designed to address these objectives:

Dual-diagnosis motivational interviewing (DDMI) [25‐27] and integrated group therapy (IGT) [28‐31] are two integrated dual disorder treatment (IDDT) programs with strong evidence bases. IDDT programs are the “standard of evidence-based treatment” for adults with CODs ([32], p. 317). After a review of multiple IDDT and non-IDDT treatment programs, our partner agencies believed DDMI and IGT to be appropriate candidates for adaptation to justice-involved populations.

Our treatment adaptation and implementation and communication protocol frameworks are described next.

The study consists of two phases: (1) an uncontrolled open trial and (2) a pilot RCT. During the open trial phase, we will collect data to inform the refinement of the research and communication protocols and the adapted interventions—broadly, the jail-to-community treatment continuum. For example, results from the uncontrolled open trial may result in changes to the inmate participant recruitment and consent process, procedures for facilitating inter-agency communication, and additional adaptations to the DDMI-IGT continuum training materials, manuals, or session handouts. All modifications to the research and communication protocols and to the adapted interventions will be made in collaboration with community partners throughout the uncontrolled open trial and implemented in the pilot RCT.

The study will be conducted at a large urban county jail in the southeastern USA. The jail’s average daily census ranges between approximately 1300 and 1450 with one-quarter of those inmates identified as being in need of mental health evaluation beyond standard screening completed during booking. Within 24–48 hours, on average, all inmates have their first court appearance. From there, inmates are either (1) bonded out or have their charges dropped or (2) detained and sent to Misdemeanor or Felony Court. From Misdemeanor or Felony Court, inmates are released or sentenced.

Inmate participants will be broadly representative of adults with CODs in a large urban county jail. Twelve inmates will be enrolled in the open trial and 60 in the pilot RCT. Eligibility criteria are: 18 to 65 years of age; incarcerated in the county jail; a resident of the county; able to speak and read English; and meet Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) diagnostic criteria for (1) drug or alcohol abuse or dependence and (2) a serious mental illness, including major depressive disorder, depressive disorder not otherwise specified (NOS), bipolar disorder I, II, or NOS, schizophrenia-spectrum disorder, schizoaffective disorder, schizophreniform disorder, brief psychotic disorder, delusional disorder, or psychotic disorder NOS, based on review of clinical records and input from available informants, including jail-based clinicians.

Prior to the start of the open trial, we will provide 1 day of training on IDDT principles, frameworks for treatment adaptation and implementation, and the DDMI and IGT programs to all study clinicians. We also will provide a second day of training separately for the DDMI and IGT clinicians, focusing on jail- and community setting-specific issues relevant to each of the interventions. Clinicians will receive a booster training during the open trial that will address modifications and refinements to DDMI and IGT and the communication protocols, as informed by open trial data. Clinicians will be retrained prior to the start of the pilot RCT and will receive another booster session midway through the RCT.

Participants will be identified for potential participation in the study through routine screening measures at booking, including the Brief Jail Mental Health Screen [45] and other jail-specific mental health and substance use screening protocols. The Jail Mental Health Liaison will identify participants who meet the study inclusion criteria, which were noted above, and the jail-based clinician will approach potential participants (after booking but before the first court appearance) to inquire about interest in study participation.

A Data Safety and Monitoring Board (DSMB), whose members will be independent of the funding agency, the National Institute on Drug Abuse (NIDA), will be brought together to oversee the study’s activities and to ensure the safety of human subjects, validity of findings, and need for further data collection. The study principal investigator (PI), Dr. Van Dorn, will interact with the DSMB at the DSMB’s discretion (at least two meetings per year), providing them with material to review, monitor, evaluate, audit, and make recommendations regarding: (1) protocols, informed consent procedures, and safety plans; (2) study progress (i.e., recruitment and retention, risk/benefit ratio for subjects, adherence to timetable, quality of data); (3) the impact of new treatment developments on the risk/benefit ratio of the study; (4) continuation, modification, or termination of ongoing studies based on adverse events or beneficial outcomes; (5) interim analyses; (6) confidentiality of trial data and results of monitoring; and (7) procedures likely to increase subjects’ burden, to raise ethical concerns, or to give the appearance of a conflict of interest.

The PI will make the following available to the DSMB: (1) all adverse events (RTI International protocol requires the reporting, in writing, of all adverse events within 5 days of the study team becoming aware of the adverse event. All adverse event forms will be made available to the DSMB as well as tables summarizing the occurrence of specific events.); (2) all interim data analyses; (3) analyses requested by the DSMB; and (4) all reports to NIDA and all publications. Finally, data management procedures, including those for data entry and quality checks, coding, security, and storage are available from the study PI.

The proposed study will have a small, but appropriate [65] sample size of 60. This study is developmental in nature, which informs how we have chosen to address the choice of error rates and power to detect effects in our clinical and public health outcomes. Prior research suggests small to medium substance use and psychiatric symptom effects of DDMI [25, 27] and IGT [28‐30]. Therefore, type I error rate α = 0.15 was chosen so as not to limit power to conduct inferential analyses, giving an 80% chance (1 – β = .80) to detect a difference between the treatment conditions that would explain 9% of the total variance [θ² = .09] in an outcome variable if the null hypothesis were false (calculated using G*Power 3.1 [66]). This approach allows for increased opportunity to detect a smaller effect, while running the risk that the effect will not replicate. This course of action is appropriate for a developmental study and is consistent with prior research involving adults with SMI and preliminary-stage multivariable models [61].

Our study will advance current research and practice regarding evidence-based treatment for inmates with CODs and associated research methodologies. We will adapt two interventions, DDMI and IGT, for justice-involved adults with CODs. Our project is grounded in evidence-based conceptual and treatment processes and will have the potential to improve substance use, psychiatric, arrest, and routine outpatient treatment uptake outcomes among an underserved and high-risk population. This type of research is needed to address the increasing rates of persons with CODs in jail populations and associated treatment failure. Second, the development and implementation of RCT methodologies is needed within and across jail and community settings to demonstrate the feasibility, acceptability, and effectiveness of adapting treatments and conducting RCTs with this population. Accordingly, this study may influence the progression of subsequent CODs and jail research and encourage others to adopt similar rigorous and system-bridging approaches.

Our study has several limitations in the design and implementation. Although court orders can be successful in increasing treatment participation among adults with CODs, substance use disorders (SUDs), and SMIs [67‐69], we do not propose any judicial oversight. This decision is in line with the preferences of our community partners and other ongoing studies [70] and has proven to be successful in retaining participants. Given our focus on one jail, the generalizability of our findings may be limited; however, generalizability of our findings can be examined in a future full-scale RCT. Allowing access to TAU in both experimental conditions also presents several potential limitations. As much as possible, our community agencies will have different clinicians deliver the DDMI-IGT and usual care services to experimental and control participants; however, DDMI-IGT and usual care will be provided in the same agencies, which may result in treatment contamination. Fidelity assessments will allow us to monitor DDMI-IGT treatment integrity and to address this potential contamination. It also is likely that, across both experimental conditions, inmates will receive different types and intensities of usual care services, resulting in treatment heterogeneity. This will not affect study interpretability, however, because service intensity will be titrated to individual client needs, consistent with the RNR framework [39]. If our DDMI-IGT continuum increases the amount or duration of usual care service uptake, the extent to which improvements in outcomes can be “explained” by amount or duration of usual care services could be explored by statistically controlling for these services and comparing the treatment conditions, e.g., through analysis of covariance (ANCOVA) models. Lastly, given that our study is focused on the preliminary treatment adaptations and feasibility, we have not focused on differential effects of DDMI and IGT within our treatment continuum.

Participant recruitment for the RCT is ongoing.