Supportive Release Center Study (SRC)

TASC INC, Heartland Health Outreach Inc, Sheriff's Office of Cook County, Illinois, Robert R McCormick Foundation, Pritzker Family Foundation, Laura and John Arnold Foundation, Center for the Advancement of Critical Time Intervention

The Supportive Release Center (SRC) is a collaboration between the University of Chicago Urban Health Lab, Treatment Alternatives for Safe Communities, Heartland Alliance Health, and the Cook County Sheriff's office. The aim of the SRC is to identify individuals with mental illnesses, substance use disorders, and other vulnerabilities as they are released from the Cook County Jail (CCJ), provide an improved environment to assess needs of these individuals, and facilitate effective linkages with social services following release, including medical care and substance use or mental health treatment. The SRC improves the current standard of care offered at the CCJ by introducing mechanisms to facilitate engagement with post-release services and address individuals' immediate acute needs. The primary objective of this randomized controlled trial is to evaluate the impact of assignment to the SRC on the number of arrests within one year of study enrollment among eligible men being released from the Cook County Jail. Researchers hypothesize that the SRC is more effective than usual care at facilitating and ensuring receipt of transition services and care, and that receipt of this treatment will decrease the number of arrests within one year of study enrollment.

The University of Chicago Health Lab is working with service providers - Treatment Alternatives for Safe Communities (TASC) and Heartland Alliance Health (HAH) - and the Cook County Sheriff's Office (CCSO), to implement a randomized controlled trial of a Supportive Release Center (SRC) compared to usual care for persons leaving the Cook County Jail (CCJ) with mental health, substance use disorder, or acute health needs. The primary aim of this randomized controlled trial is to evaluate the impact of assignment to the SRC, compared with not being offered a placement at the SRC (and consequent usual care), on the number of arrests within one year of study enrollment among eligible men being discharged from CCJ. The recruitment strategy for the SRC intervention is integrated into the jail discharge process. Individuals are discharged from the CCJ in groups, such that they may wait several hours before being discharged. After being escorted to the discharge area, detained individuals are consented to participate in the SRC study and asked to take a survey designed to identify needs, vulnerabilities, and interest in receiving services. During consent, individuals are informed that their CCJ Booking ID will be used to acquire data from the CCSO, Chicago Police Department, Illinois Criminal Justice Information Authority, and Illinois State Police to determine their history of arrest, time spent in jail, and instances of re-arrest following study enrollment. In conjunction with demographic information linked from CCSO, this ID can be used to track healthcare and social service utilization via databases administered by Cook County Health and Hospital System (CCHHS), the Centers for Medicare and Medicaid Services, the Illinois Department of Healthcare and Family Services, the Illinois Department of Public Health, the Chicago Fire Department, and collaborating service-providers (TASC and HAH). Eligibility for the SRC intervention requires survey responses indicating both (1) vulnerability, suggesting potential need for the SRC, and (2) interest in attending the SRC. Vulnerability that suggests potential need for the SRC is ascertained by four independent criteria: a response of "more than half the days" or "nearly every day" to questions (1) and (2) from either the PHQ-9 depression or Generalized Anxiety Disorder (GAD)-7 modules, which refer to problems experienced "over the past 2 weeks" (PHQ: 1. "little interest or pleasure doing things", 2. "feeling down, depressed, and hopeless"; GAD: 1. "feeling nervous, anxious, or on edge", 2. "not being able to stop or control worrying"); two or more previous instances of incarceration over the past 12 months; history of substance use over the previous 12 months, indicated by weekly or more than weekly consumption of i) more than five alcoholic drinks per day, or ii) prescription drugs for non-medical reasons, or iii) illegal drugs; or affirmative response to questions asking about previous diagnosis by a health professional for a mental health condition, existence of urgent needs for healthcare over the three weeks following discharge, or concern about personal safety during discharge from the jail. Two questions gauge respondent interest in attending the SRC: interest in having (i) a safe place to stay upon discharge or (ii) a place to go to receive connections with social services upon discharge. An individual who responds affirmatively to both "interest" questions is automatically eligible for the SRC; to be eligible with an affirmative response to only one "interest" question, the respondent must have at least one indicator of vulnerability. Individuals who do not respond affirmatively to either question about interest are not eligible for the SRC. Detainees who are eligible and interested in services are then randomized to treatment and control groups. Individuals who are randomized to treatment have individual meetings with TASC social workers in a private office close to the discharge area, where the SRC is explained and offered. TASC provides opportunities to interview and receive service-linkages in the CCJ office for all randomized individuals. In addition to individuals randomized to control, those randomized to treatment who choose not to visit the off-site SRC facility are offered linkages during their private interview with TASC. The SRC is located on the CCJ campus, one half mile from the point of CCJ discharge. Participants accepting the SRC treatment are transported from the CCJ exit to the facility in TASC vehicles. The facility has the capacity to provide an overnight stay, meal, and shower to twelve individuals per night and provides a private environment where TASC case-workers can meet with discharged individuals. This environment provides TASC staff an opportunity to conduct a more intensive needs-screening to identify appropriate service-linkages, including connection with organizations providing substance use treatment and mental health services. An advanced practice nurse is available at the SRC to address immediate medical needs and ensure continued access to necessary prescription medications. On pre-specified days HAH recruits from the SRC to populate a non-SRC study; individuals recruited from the SRC and individuals randomized to SRC control on days HAH is recruiting are excluded from the SRC study. Intention-to-treat (ITT) and treatment-on-treated (TOT) analyses will be conducted for specified outcomes. In addition to an ITT group indicator, all models include covariates for demographic information (age, race, Hispanic/Latino indicator) acquired from the CCSO, an indicator variable for whether the individual could qualify for SRC treatment based on that measure ("qualified" (1) vs. "not qualified" or "no response (0)), and an indicator for whether that measure was missing ("no response" (1) vs. "not qualified"or "qualified" (0)) (i.e. PHQ/depression score, GAD anxiety score, previous mental health diagnosis, alcohol use, prescription drug use, illegal drug use, acute health needs, or concern for personal safety). Age (18-24, 25-35, 36-45, 46-55, over 55), race ("black", "white", "other", "unknown"), and ethnicity ("Hispanic/Latino", "not Hispanic/Latino", "unknown") are coded as categorical variables. All models will include an interaction term between age and ITT status. Arrest and hospitalization outcomes will be measured using data from the Illinois State Police and Illinois Department of Public Health acquired via probabilistic matching on demographic identifiers. As a robustness check for count outcomes, researchers will evaluate sensitivity of results to the inclusion of an offset representing the number of days individuals an individual is vulnerable to the outcome. Unexposed time (invulnerable to re-arrest or emergency department visit) is defined as time spent incarcerated or hospitalized.

Usual care. These individuals will receive linkages to community-based services only. These linkages will be made while in the Discharge Area of the Cook County Jail. Individuals randomized on days Heartland Alliance Health recruits at SRC for non-SRC study are excluded from population. Receives: TASC Service Linkages - Discharge Area Only

These individuals will receive SRC services: an overnight stay at the SRC, linkages to community-based services made at the SRC or in the discharge area of the CCJ if they choose not to visit the SRC facility, and access to an Advanced Practice Nurse. Individuals randomized on days Heartland Alliance Health recruits at SRC for non-SRC study are excluded from population. Receives: TASC Service Linkages - Discharge Area Only -OR- SRC Overnight Stay, TASC Services Linkages - SRC Onsite, APN Appointment

Individuals will meet with TASC Care Coordinators in the discharge area of the Cook County Jail. Individuals will be screened for housing, health and behavioral health needs and then linked to appropriate community-based services.

Individuals receiving SRC services will have the option to stay at the SRC overnight. At the SRC, individuals will be able to rest, have a meal, receive a change of clothes, and take a shower in a therapeutic, non-coercive environment.

While at the SRC, individuals will meet with TASC Care Coordinators. Individuals will be screened for housing, health and behavioral health needs and then linked to appropriate community-based services.

Individuals at the SRC will be able to meet with an Advanced Practice Nurse. As part of that visit, the APN will be able to prescribe medications or reroute prescriptions previously administered by the Cook County Jail's hospital, Cermak, to a participant's desired location as needed.

Number of re-arrests within 365 days of study enrollment, measured at the individual level using Illinois State Police arrest records. Follow-up will begin the day an individual is first randomized (index randomization) and ITT group membership will be based upon index treatment assignment. Given expected dispersion and excess zeroes in the final dataset, the ITT treatment effect will be modeled using zero-inflated negative binomial regression. ITT analyses will be supplemented by a TOT analysis treating randomization as an instrument. Covariates for this outcome will include ordinal measures of arrest history (0, 1-2, 3-5, 6-10, and >10 arrests) and days incarcerated (0, 1-7, 8-14, 15-30, and >30 days) during the 365 days prior to index randomization.

Number of emergency department visits within 365 days of study enrollment, measured at the individual level using hospitalization records from Illinois Department of Public Health. Follow-up will begin the day an individual is first randomized (index randomization) and ITT group membership will be based upon index treatment assignment. Given expected dispersion and excess zeroes in the final dataset, the ITT treatment effect will be modeled using zero-inflated negative binomial regression. ITT analyses will be supplemented by a TOT analysis treating randomization as an instrument. Covariates for this outcome will include ordinal measures of ED utilization (0, 1-2, 3-5, 6-10, and >10 visits) and days hospitalized (0, 1-7, 8-14, 15-30, and >30 days) during the 365 days prior to index randomization. Emergency department visits that occur during periods of incarceration will be included in the count outcome.

Days until first re-arrest, measured at the individual level using Illinois State Police arrest records. Follow-up will begin the day an individual is first randomized (index randomization) and last for 365 days, after which all individuals not re-arrested will be treated as censored. The hazard ratio associated with treatment assignment (ITT) will be estimated via Cox proportional hazard model. ITT analyses will be supplemented by a TOT analysis treating randomization as an instrument. Model covariates are identical to those included in the primary outcome analysis (count, re-arrests).

Number of re-arrests within 365 days of study enrollment, measured at the randomization level using Illinois State Police arrest records. Follow-up will begin the day a randomization occurs and ITT group membership will be based upon randomization status. Given expected dispersion and excess zeroes in the final dataset, the ITT treatment effect will be modeled using zero-inflated negative binomial regression. To account for correlation between randomizations of the same individual, researchers will employ a 2-level model with random effects at level 2 (individuals). ITT analyses will be supplemented by a TOT analysis treating randomization as an instrument. Model covariates are identical to those included in the primary outcome analysis (count, re-arrests).

Existence of formal contact within 30 days of study enrollment, measured at the individual level using utilization records from the All Chicago Homeless Management Information System (HMIS). Follow-up will begin the day an individual is first randomized upon discharge from jail (index randomization) and ITT group membership will be based upon index treatment assignment. The ITT treatment effect will be modeled using a simple logistic model, regressing engagement status on treatment status. We will conduct secondary analysis including covariates for age, race, and offense history.

Inclusion Criteria: Male adult being discharged from Cook County Jail Individuals indicating a need for services Individuals interested in receiving services Exclusion Criteria: Individuals who are not leaving Cook County Jail Female individuals Individuals without acute social needs Individuals below 18 years of age