Recovery Management Checkups for Opioid Use Disorder Experiment (JCOIN-HUB)

Improving Retention Across the OUD Service Cascade Upon Reentry From Jail Using Recovery Management Checkups Experiment

The experiment will be conducted in collaboration with 6 jails representing discrete geographic counties in Illinois and the opioid treatment providers (OTP) that serve them. It will compare a re-entry as usual control group with two experimental groups in terms of their impact on the OUD service cascade, as well as public health and public safety outcomes. Study recruitment sites are six jails that provide treatment with medication for opioid use disorder (MOUD) to inmates with OUD prior to their release. At the time of their release to the community, 750 men and women will be randomly assigned to 1 of 3 groups: a) a re-entry as usual control, b) RMC with monthly checkups for 3 months post-release followed by quarterly checkups up to 2 years, or c) an adaptive version of RMC (RMC-A) that includes a modified checkup schedule based on each individual's pattern of treatment need. All participants will complete research interviews at release and quarterly thereafter up to 2 years post-enrollment.

This project aims to complement HEAL's Justice Community Opioid Innovation Network (JCOIN) initiative by focusing on post-release recovery in six county jails in Illinois with the goal of increasing linkage to and retention in community-based treatment with medication for opioid use disorder (MOUD) and reducing both opioid use disorder (OUD) relapse and criminal recidivism over two years. The study builds upon the platform of prior research demonstrating the effectiveness of the Recovery Management Checkups (RMC) intervention. To date, three randomized controlled trials conducted by the study investigators have demonstrated RMC's effectiveness by increasing treatment engagement and retention, re-linking recovering patients to treatment when indicated, and improving their health, psychosocial functioning, and other outcomes. Across these studies, relative to controls, participants in the RMC condition reported more treatment initiation, more days of treatment received, fewer days of substance use, and fewer health symptoms. In the proposed study, an adapted version of the RMC (RMC-A) will build upon the core components of the RMC by tailoring the checkup frequency to the individual's assessed need for treatment over time. The goal of the proposed experiment is to compare linkage and retention rates as well as public health and public safety outcomes of 750 male and female offenders who will be randomly assigned to 1 of 3 groups upon their release from jail: a) re-entry as usual (control), b) the original RMC, and c) a RMC-Adaptive version in which the frequency and content of checkups will be based upon the participant's ongoing progress as assessed by a set of indicators of treatment need. The study will be conducted in collaboration with 6 county jails in Illinois, and their local OTPs that currently provide pre-release MOUD to offenders with OUD. Pre-release participants will be screened for history of OUD and eligibility for MOUD following usual procedures. All participants will receive quarterly research follow-up assessments. Data sources include interviews with standardized measures, urine tests, treatment records to verify treatment received, recidivism, death records, and cost accounting. Research staff will administer the 25-minute Global Appraisal of Individual Needs - Quick (GAIN-Q3) and several supplemental scales at study enrollment, and again quarterly over 24 months post-study enrollment. The GAIN-Q3 includes 8 primary domains: 1) background, 2) school problems, 3) work problems, 4) physical health, 5) sources of stress, 6) HIV risk behaviors, 7) substance use, and 8) crime and violence. The response set captures the recency of these symptoms (life time, past year, past 90 days, past month) and treatment involvement, then the frequency (days) of behaviors/treatment utilization in the past 90 days. The GAIN-Q3 includes a behavioral health screener, health care utilization measures, measures of days of methadone treatment, heroin and other opioid use, days of other types of treatment, and other substance use. The GAIN-Q3 will be supplemented with the Quality of Life measure from the Patient Reported Outcomes Measurement Information System (PROMIS) - R. Data will be collected via a cloud-based computer program that controls ranges and skip-outs and identifies major inconsistencies for interviewers. All interviewers will receive extensive training and monitoring to ensure data quality. At the time of each in-person research assessment, on-site urine screens will be conducted with DrugCheck cups and fentanyl test strips using an immunoassay for rapid qualitative results based on SAMHSA-standard cutoffs for methadone and opiates/morphine in general, as well as specific tests for fentanyl, Oxycodone, and other substances. Research staff will be trained on a protocol that minimizes false negatives consistent with NIH's PhenX common data platform. The intervention focuses on improving elements in the OUD service cascade that have historically presented challenges, i.e., successful treatment linkage, engagement, and retention in community-based treatment post-release. With its emphasis on adapting an existing evidence-based intervention (RMC) to reduce high drop-out and increase retention in treatment with MOUD, it places individuals with OUD at lower risk for relapse, overdose, and death. The RMC model also proactively addresses the well-documented chronic cyclical nature of addiction, whereby individuals transition between being in treatment, in the community using, incarcerated, and in recovery. Findings from the study may help to target resources to those with demonstrated need, and reduce the intervention burden on those with lower need, thus resulting in an improved overall effectiveness and cost-effectiveness of RMC checkups.

The type and level of services provided to individuals at re-entry will vary across jails and will be carefully documented. For the most part, individuals released to the community will receive a referral to an opioid treatment provider (OTP) for treatment with MOUD, and a subset may potentially be mandated to participate in community based treatment and/or recovery programs such as recovery coaching, and/or sentenced to varying levels of probation.

In the RMC condition, participants will have access to services provided as a part of re-entry as usual. In addition, checkups will be provided on a fixed schedule that includes face-to-face monthly checkups for the initial 3 months, and quarterly for the rest of the two years. Participants will have access to referrals and services provided by the jail and linkage to an OTP as part of their usual re-entry procedures. Individuals will meet with a Linkage Manager (LM) upon study enrollment and during each quarterly checkup, during which they will complete a Brief Treatment Needs Assessment, receive motivational interviewing, linkage assistance, or a check-in on continuing care and recovery support. The priority is to engage the individual into treatment with MOUD as soon as possible at the time of release, however, if individuals express a preference for another form of SUD treatment, the LM will work with that individual to link, engage, and retain them in that form of treatment.

In the RMC-Adaptive condition, checkups will be provided based on the participant's current need for treatment and will be adapted in three ways. First, the interval between RMC-A check-ups will vary (in 1-month increments) depending upon the individual's assessed need for treatment at the prior check-up. Second, in cases where participants have 3 consecutive checkups in which they need treatment, the LM and treatment provider will discuss how to better meet the participant's needs, e.g., a different treatment provider, different type of MOUD or other types of treatment, and/or additional services. Third, if RMC-A participants are re-incarcerated at the time of their checkup, the LM will meet with the individual while incarcerated to discuss a recovery plan, which may include initiation of treatment with MOUD while incarcerated and re-linkage to an OTP upon release.

The RMC model was designed to improve treatment linkage, engagement, and long-term treatment retention. The conceptual framework was based on the public health theory of chronic disease management, which utilizes ongoing assessment and monitoring through regular face-to-face checkups and early (re)intervention to facilitate detection of relapse, reduce the time to treatment re-entry, and consequently, improve long-term outcomes. The original RMC model includes: 1) a fixed schedule of face-to-face quarterly checkups to assess need for SUD treatment, 2) personalized feedback based on an assessment and motivational interviewing to increase treatment motivation, 3) problem solving around barriers to treatment access and retention, and 4) assistance with scheduling and linkage to treatment. Individuals were deemed to be in "need of treatment" if they reported weekly, or more frequent, substance use since the last checkup, any past-month SUD symptoms, or self-perceived need for treatment.

Participants in the RMC-Adaptive group will receive face-to-face monthly checkups during the first 2 months post-release with additional face-to-face checkups dependent upon the participant's progress. Based on the data presented in the prior section, as long as participants need treatment, they will receive monthly checkups. In addition, for every checkup that a participant does NOT need treatment, the number of months before the next checkup will be increased by 1 month (e.g., after 2 checkups without need, they will receive the next checkup 2 months later, after 3 it will be 3 months, after 4 it will be 4 months, and so on). Based on these decision rules, individuals who have NO treatment need over 24 months will still receive 5 checkups.

A continuous measure of months of treatment with MOUD received from opioid treatment provider records (Interclass Correlation Coefficient [ICC]=.37 when measured quarterly). The count will be based on days of medication received (including any take-home dosages); injectable naltrexone will be counted as treatment for 30 days. Missing records data will be estimated from self-report.

Continuous time-to-event measures (based on date of discharge from jail) and dichotomous measures (yes/no) of whether each of the following events happen: a) initiation of treatment with MOUD, b) engagement in MOUD treatment for at least 6 weeks, c) retention in MOUD treatment for at least 90 days (median in national data), and d) retention for at least 6 months, as well as times from post-drop out and relapse to re-linkage to MOUD; largely used descriptively.

A self-reported count 0 to 90 days of using any kind of opioids each quarter (test-retest rho=.95; ICC=.34 across quarters). There will also be additional measures of the days of using heroin, fentanyl, and prescription opioid misuse for descriptive use. All come from a standardized assessment tool, the Global Appraisal of Individual Needs (GAIN).

A self-reported count of the 0 to 11 the Diagnostic and Statistical Manual version 5 (DSM-5) symptoms of Opioid Use Disorders (OUD) each quarter from the GAIN (alpha=.95; ICC=.28 across quarters). There will also be alternative measures for OUD symptoms in the past month, year, and lifetime, as well as a measure of DSM-5 symptoms across other SUD each quarter.

This is a quarterly self-reported frequency of problems in with reference to the past 7 days in 8 domains: physical function, anxiety, depression, fatigue, sleep disturbance, ability to participate in social roles and activities, pain interference, and cognitive function ability, and an overall measure of pain intensity from the Patient Reported Outcomes Measurement Information System (PROMIS). Scored on a scale from 0 (near death) to 100 (best health).

This cost measure is based on the self-reported frequency of healthcare services related to substance use, mental health, and physical health, measured in units such as "days," "visits," or "episodes", which are multiplied by corresponding monetary conversion factors (price weights for a unit of service) to estimate a continuous "total healthcare costs" variable and trimmed at the 99% percentile due to sharp right skews (ICC=.27 across quarters). The self-report questions come from the GAIN.

This is a self-reported count of 19 items across different types of illegal activities related to property crimes (e.g., vandalism, bad checks, theft, breaking and entering), personal crime (e.g., assault, rape, murder), and substance use (driving under the influence, distribution, prostitution, gang membership, gambling) in the past quarter (alpha=.9; ICC=.20 across quarters). These self-report questions come from the GAIN. A count of the number of types of crimes committed in the 24 months.

A dichotomous measure of whether the person has been re-arrested or re-incarcerated on any charge within 24 months.

This cost measure is based on the self-reported frequency of 12 criminal offenses times their respective societal cost per offense based on the most recent economic estimate, which will be used to estimate a continuous "total cost-of-crime" variable in whole dollars and trimmed at the 99% percentile due to sharp right skews (ICC=.03). The self-report questions come from the GAIN.

This measure is based on the Patient Reported Outcomes Measurement Information System (PROMIS) - R. It is the self-reported frequency of problems in with reference to the past 7 days in 8 domains: physical function, anxiety, depression, fatigue, sleep disturbance, ability to participate in social roles and activities, pain interference, and cognitive function ability, and an overall measure of pain intensity. QALY (0-100%) times 0.25 years for each quarter and summed across 8 quarters/24 months post-release. While missing data will be imputed, from a point of death on will be treated as 0. In addition to being compared between groups, data can be scaled to U.S. norms using the AHSR data.

Inclusion Criteria: meets DSM-5 opioid use disorder criteria in the past year reports heroin or other opioid use in the 90 days prior to entering jail is released from 1 of the participating jails. Exclusion Criteria: is under age 18 has cognitive impairment that precludes ability to give informed consent resides outside the service area.