Abrupt termination of care is a devastating consequence of a correctional system's inability to coordinate and link released inmates with behavioral health and social services. In 2007 an estimated 79,000 adults with serious mental illness were unable to access community-based public or private mental health services in Harris County, Texas (1). As a result, the Harris County Jail has become the de facto primary mental health care provider for the county. It serves as the largest provider of mental health beds in Texas and the second largest such provider in the nation (second only to the Los Angeles County Jail) (2).
There are roughly 2,400 people using mental health services in the Harris County Jail on any given day (2). In the absence of proper linkages to necessary behavioral health and social services, many members of the mentally ill homeless population cycle between the streets and shelters, emergency centers, and jail cells in a virtual revolving door, and the costs to the county attributable to their increased rearrest rates exceed $14 million per year. This column describes an emerging best practice—the Jail Inreach Project—to address the mounting implications of disruption or termination of care for persons who have been released from jail and are mentally ill and homeless.
The Jail Inreach Project in Harris County, Texas, provides services that strive to improve health status, support social reintegration, and reduce the costs and overall burden on the public health care system of persons who are mentally ill and homeless and being released from the Harris County Jail. In 2006 the Jail Inreach Project was established and locally funded as a demonstration project under the auspices of Healthcare for the Homeless-Houston (HHH). The Jail Inreach Project provides continuity of and access to integrated primary care and behavioral health care (mental health and substance abuse treatment) services upon release for Houston's homeless population with mental illness, who rapidly cycle through the Harris County Jail. It further provides a replicable model of care that addresses the underlying causes of increased rearrest rates and the immediate need for services after incarceration.
The program aims to reduce rearrest rates and combat the rapid cycling of the mentally ill homeless population through Harris County Jail; to collaborate with the Harris County Sheriff's Department and the local public mental health agency to develop a more coordinated system of care that supports homeless releasees in accessing resources; and to facilitate the transfer of medical records between the Harris County Jail, community-based clinics, and the public health care system to improve communication and coordination between health care providers.
HHH and the Mental Health Mental Retardation Authority (MHMRA) of Harris County met for the first time in early 2006 to solidify collaboration on the Jail Inreach Project. The meeting led to a memorandum of understanding between HHH and MHMRA that established a referral process, which allows HHH case managers to initiate contact with Harris County Jail inmates with mental illness who were homeless at incarceration or are anticipated to be homeless on release and to create a system for sharing patient records to enable continuity of care and streamline research and evaluation efforts. The memorandum further designated the placement of a licensed MHMRA clinician within HHH clinics, marking the first time that MHMRA had achieved such a placement within a Harris County federally qualified health center.
Evaluations and assessments are performed by staff while the client is still incarcerated, and linkages for needed services are established. In meeting with an inmate before his or her release, the goal is to identify the needs of the individual and to develop a discharge plan that includes initial medical and behavioral health care (including psychiatry, as needed) provided by HHH, eligibility assessments for MHMRA services, substance abuse assessment and counseling provided by HHH, and an assessment of housing and transportation needs and benefits eligibility.
In order to maximize the effectiveness of the discharge plan, case managers provide clients with the option for a "direct release" into the care of a case manager. This means that rather than being discharged from the jail to one's own care in the middle of the night (as is standard practice), participants may choose an escorted release if they agree to stay in jail until the following morning. Individuals who volunteer for daytime release are met outside the jail by an HHH case manager, who walks them to HHH's Cathedral Clinic, located just blocks from the jail. At the clinic, they receive immediate health care. Individuals with serious mental illnesses and who therefore meet diagnostic criteria (have bipolar disorder, schizophrenia or schizoaffective disorder, or major depression) may be eligible for services via the public mental health system. Access to the public system of care often requires weeks to months on a waiting list for an appointment, during which time patients might relapse because of a lack of access to medication. HHH often provides interim care until those who are eligible can be linked into larger public systems of care.
Participation in this program is based on the following four criteria. Participants must be detained in Harris County Jail; have a diagnosis of a behavioral health condition (mental illness, substance use disorder, or both); anticipated to be homeless upon release; and have a history of recidivism, defined as having had two or more bookings into Harris County Jail in the previous six months.
As of June 26, 2009, when the initial program evaluation began, 492 individuals had been referred to the Jail Inreach Project. Twenty-two percent experienced multiple encounters (those who are rearrested are contacted again by the program to try to link individuals back into community health services). For those who had multiple encounters with the program (those who are rearrested are subsequently revisited by their case manager, and a new case is opened), only the first encounter was included in this analysis. This was done to control for the possible implications that multiple encounters with the program may have on outcome and because each encounter can result in a different disposition (linked versus not linked, for example). Of the total number of first encounters (N=492), 275 (56%) resulted in successful linkage to service after their release. Twenty-four (5%) declined services, 53 (11%) were transferred to another correctional facility, and 140 (29%) engaged in the program while incarcerated but did not follow through with the program on release.
A study was conducted by HHH in collaboration with MHMRA and the Harris County Budget Office that evaluated arrest rates one year before engaging in the program and one year after engagement. Results indicate that those who were linked to services after their release had arrest rates that were 36% lower compared with the number of arrests one year before contact with the program and one year after contact with the program. Also, the average number of days spent in jail decreased from about 65 days before contact with a case manager to just less than 42 days during the year after contact with the program. Total annual criminal charges (misdemeanors and felonies) for each participant had also been reduced by 56% during the year after contact with the program.
The delivery of health care in the correctional environment has many challenges. Those who are incarcerated and who have a behavioral health diagnosis tend to have more comorbid chronic medical problems much earlier than those who are not incarcerated. Because many detainees have poor health and inadequate diets, have risky lifestyles, and abuse substances, they are likely to have a higher incidence of medical conditions in addition to their mental health needs. Delivery of care within the jail and provision of discharge planning services are complicated by complex social and health needs, the usual short length of incarceration, and a sometimes uncertain date of release.
Many logistical considerations and adjustments had to be taken into account as the program developed early on; much time was spent experimenting with different procedures for visiting detainees in order to increase efficiency. Initially, detainees were called from their cells to an interview room on the first floor of the jail. The process of moving a detainee from his or her cell to the interview room often took up to 30 minutes. In the first six months, with over 85 referrals and nearly 200 visits, it became clear that a more efficient system was needed. We discovered that more detainees could be visited in a shorter period if the case manager went to the detainee's cell and met with him or her either in the cell or in another suitable location on the same floor. Similarly, the issue of dealing with uncertain release dates resulted in the development of a policy mandating that after the initial meeting, clients were seen only after they had an established release date. This provision aided in successfully establishing a postrelease plan and in case managers' serving more clients because case managers could maximize the efficiency of how their time was allocated.
Outreach strategies that provide continuity of care have generally been supported as a best practice. The most effective strategies seem to be those that introduce personal connections and reduce the distinction between inpatient and outpatient services.Linkage is better served when the step to what has traditionally been called aftercare is treated more as a transition than as a change (3—5). Practices that include contact with the ongoing care agency before discharge have proven to be beneficial. For the Jail Inreach Project, the contribution of the negotiated direct daytime release to the care of a case manager cannot be overstated.
The most recent quality assurance analysis of the program indicated that 129 of 150 (86%) participants who opted for direct release to the care of a case manager were successfully linked to services, compared with only 51 of 181 (28%) of those who opted for a self-release. Further, those who opted for a "self" nighttime release were 6.2 times less likely than those who opted for a direct release to a case manager to be linked to services (6). Agreeing to spend a few more hours in jail in return for direct release may be an indicator of greater motivation for change but may also reflect the perceived importance of the health care and social service linkages provided by the program. Early indications from our data suggest that linkage, in its turn, appears to reduce the likelihood of rearrest, especially for participants with fewer and milder previous charges (misdemeanors versus felonies).
The importance of efficiently and effectively tracking client information and data in a way that is accessible to multiple HHH staff members, including case managers and administrators, as well as for research and evaluation purposes, led to the development of an online database, which is housed on a secure server at Baylor College of Medicine. It is used to track release dates, diagnoses, initial referral plans, postrelease service linkages, and case status of the individual. The database can also generate reports for quality assessment and evaluation.
We believe that the design and implementation of the Jail Inreach Program constitutes a best practice that should be subjected to further services research. One of the most successful components of the Jail Inreach Project is in bridging gaps between services provided in the jail with services provided in the community. However, the program is restricted by the limited capacity of community resources to provide services after a detainee is released. An alarming shortage of affordable housing, psychiatric services, substance abuse treatment services, and other medical and social support services has put a ceiling on the capabilities of the safety net programs to benefit this population and the community at large.
Stemming from this assessment of the program, there are two subsequent evaluation projects examining whether there are differences in outcomes, procedures, or client characteristics between individuals who are linked to services and those who do not follow through with the program after release. Results will be informative in terms of tailoring better interventions and in replicating these findings.
On a systemic level, our findings suggest that effective and replicable mechanisms for reducing utilization of correctional institutions as behavioral health treatment facilities include offering early prerelease planning, engaging clients in their treatment plan, offering daytime release, and providing a full health evaluation, including medical case management and behavioral health assessments. These elements provide continuity of care for those who require medication and help to prevent lapses in treatment. In addition, the activation of social services (shelter and housing, job training, and life skills counseling) and health services (primary health home with behavioral health care) in the framework of enhanced integrated care appears to decrease arrest rates (thus decreasing utilization of mental health services within the correctional system) and increase the possibility of transitioning out of homelessness.
Another opportunity for change in the collaboration between HHH and MHMRA would be to allow for integration of a behavioral health specialist in the primary health home. This would provide immediate behavioral health resources to stabilize clients adjusting to the challenges common to releasees. Clients could then be provided a permanent health home where their full health needs are addressed. Behavioral health exacerbations requiring internal referral or consultation could be obtained as needed. Future research will be necessary to validate our observations and recommendations.
In 2008 the Medallion Foundation, Inc., funded the initiation of a pilot project using intensive health care-based case management. Subsequent funding for the evaluation of and research associated with the program was funded by general operating funds of HHH, where the program is housed. The authors thank the staff of HHH, MHMRA of Harris County, and the Harris County Budget Office, who have helped make this project possible. The authors appreciate the editorial input of William E. Fann, M.D.
The authors report no competing interests.