Developing the evidence base for reducing chronic inmate Self-injury: outcome measures behavior managern.
However, the challenges do not end there. Quite unintentionally, crisis interventions that are essential to restore safety or provide treatment can raise the risk of future inmate self-injury among this group of inmates. Removal of property and placement on observation status in a suicide smock may be experienced by the inmate as punitive. Administration of pain medication following surgical repair of self-injury and "negative" attention involved in interrupting or treating self-injury may be experienced as rewarding. Whether the inmate becomes locked in a power struggle or is rewarded with reinforcing medications and outside hospital trips, or just "feels better" after self-harm, the pattern of self-injury can intensify over time. Repetitive self-injury both raises and masks the risk for completing suicide, like "the boy who cried wolf" parable.
Behavior Management Solutions
Done properly, behavior management interventions hold much promise in addressing these challenges and improving the quality of the inmate's life. The goal of behavior management is to reinforce (or reward) inmates so desired behaviors grow stronger than the inmate's problem behaviors. In the model described here, the correctional system accomplishes this using a structured, phase-based approach that reinforces small behavior changes with small incentives and larger behavior changes with larger incentives. Rewarding behavior that the inmates should already be doing can be a tough sell, particularly if the inmates in question have had assaultive, abusive or abrasive interactions with correctional staff. In correctional settings, the first instinct may be to punish unwanted behavior, not design a behavioral program to reward positive behavior. It is natural for staff to question the wisdom of providing incentives to the very inmates who repeatedly cause the most difficulty and danger.
Implementing behavior management interventions is a boot-strap operation when there is no outcome data to support the program. Once interventions can be demonstrated to reduce self-injury and create predictable routine responses for staff, "buy in" for incentive-supported programming is easier to achieve. Support from the top--including support from central office and facility administrators--is necessary to effect culture change. Seasoned correctional staff need to see it for themselves before they are convinced. Strong policies and procedures; multidisciplinary teamwork; staff training; program monitoring to ensure fidelity of behavioral interventions; and outcome measurement are critical.
Judicious application of behavior management is also needed. Inmate self-injury is heterogeneous, and behavior management interventions are not appropriate in all cases. Figure 1 shows four possible functions of inmate self-injury. (1) Only those functions shown in the upper right and lower left quadrants are appropriate for behavioral management interventions. Self-injury that is suicidal in intent and/or that is driven by hallucination and delusions requires other types of mental health and psychiatric intervention.
Behavioral Management at MDOC
Application of behavioral interventions must not only be clinically appropriate but also ethical. The behavior management strategies of the Massachusetts Department of Correction (MDOC) involve the use of structured, preplanned incentives in response to positive behavior change. Incentives are provided in a fully transparent, predictable and consistent manner. MDOC does not use punishment, aversive stimuli, coercion, fear induction, surreptitious intervention or deprivation of basic needs in behavioral interventions. Behavior management is a multidisciplinary strategy that follows the principles of informed consent and use of the least restrictive alternative. By rewarding inmates for positive change, the time between episodes of self-injury can be increased, the severity of self-injury can be decreased, and the inmate can learn new skills to manage frustrations, communicate needs and solve problems.
Behavior management interventions can be applied either with individual inmates or with a group of inmates on a specialized mental health unit that is designated exclusively for that purpose. In MDOC, a 10-bed Behavior Management Unit (BMU) has been developed for male inmates with severe behavioral problems and very lengthy restrictive housing sanctions in the Departmental Disciplinary Unit, which is the restricted housing unit with the highest level of security. BMU was developed in response to litigation addressing treatment of inmates with a serious mental illness (SMI) who were housed in long-term, restrictive housing units. In March 2007, a lawsuit was filed by the Disability Law Center against MDOC specific to this issue. In July 2007, MDOC partnered with MHM Services Inc., and took a proactive approach by designing specialized services for SMI inmates prior to a court mandate. BMU opened in July 2010, and as a result, the case was settled in 2012.
BMU is designed as an alternative to long-term, restrictive housing for male anti-social/psychopathic inmates who meet criteria for SMI. Another secure treatment unit was designed and opened to treat long-term segregation inmates with more "classic" SMI, such as schizophrenia or bipolar disorders. BMU inmates are generally highly assaultive, often self-injurious and extremely difficult to manage from both a correctional and mental health perspective. These individuals are often strategic, calculated and instrumental in their problematic behavior. Behavior management and understanding the negative consequences of anti-social behavior are the focus of BMU, which has clearly-defined incentives and consequences and also operates using a phase system. Inmates earn points for positive behaviors, which are tallied each week and can be exchanged for incentives from a menu of options--similar to a traditional token economy. Strict adherence to rules and consistency of staff are necessary for positive behavioral outcomes. (2)
The authors have also developed individualized behavior management plans for high-profile inmates in eight correctional systems, including one large municipal jail and seven state correctional systems. Typically, by the time these inmates are referred for behavioral management consultation, correctional systems are desperate for a solution. These inmates are not only challenging to manage and worrisome to safety security staff, but these inmates are also expensive to care for. Just one visit to the community emergency room for extraction of a foreign object or medical stabilization following an episode of severe self-mutilation can cost anywhere from $10,000 to $20,000. Those costs do not include correctional officer overtime required for community escorts. Most of these inmates have had dozens of such visits. A summary of the 13 inmates for which the authors have conducted individual behavioral consultations since 2008 is shown in Table 1.
Phases of Behavior Management
Both in BMU and when developing individual behavior management plans, the authors have collaborated with correctional systems to structure behavior management interventions using "phases" or "levels." These
phases dictate, in advance, the specific positive behaviors the inmate must demonstrate, the specific length of time in which these behaviors must be demonstrated, and the specific incentive(s) the inmate earns as a result of positive behavior. Phases are structured to start with the most restrictive level of incentives and the greatest level of safety precautions, with incremental increases in incentives during the following days, weeks and months as the inmate demonstrates progressive behavior change. For example, an inmate may be allowed to have time-limited access to a book or magazine while on watch if he has been free from self-injury for three days and has been willing to track his moods during that same period. After three more days, he may be allowed time-limited access to television or an additional 15 minutes of recreation time. For some inmates, additional individual contacts with a mental health professional or an additional treatment group can be used as incentives to provide more out-of-cell time and positive attention.
Figure 2. Results of Individualized Behavior Management Plans Pre-Behavior Post-Behavior Management Management Self-Injurious Behavior 228 103 On-Site Medical 99 63 Outside Hospital 55 12
The phases also dictate, in advance, what happens if the inmate reinjures himself or herself or engages in other problematic behavior. In almost every case, the authors' experience is that inmates exhibit problem behaviors after behavioral interventions are implemented. In some cases, the problem behaviors initially escalate as the inmate "tests the limits" and increases self-injury because it has worked in the past. The phase system anticipates problem behaviors and makes clear to staff and to the inmate the external responses to these behaviors. A return to a lower, more restrictive phase occurs in response to self-injury or threats of self-injury. In the case of actual self-injury, the clock resets to the beginning phase, and the behavioral program starts over. These set-backs are expected in inmates who have engaged in recurrent self-injury, but they are not considered failures. Most inmates move up and down the phase system repeatedly before achieving sustained behavior change. The elegance of the phase system is that it provides a simple roadmap that guides progress and is clear to everyone--inmates and staff alike. Treatment and management decisions that were once "crisis-driven" become routine, matter-of-fact and highly predictable. For many inmates, this takes the excitement out of self-injury. It becomes boring because the consequences are so consistent.
During the last several years, MDOC has implemented and measured the impact of the behavioral interventions in correctional settings. (3) If a correctional system is pouring resources into mental health services and there are no clinical improvements, it is likely time to go back to the drawing board. However, clinically meaningful outcome measures can be challenging to identify for specialized mental health services. There are no laboratory tests to measure mental health or risk for self-injury. Correctional facilities rarely develop systems for routinely measuring mental health outcomes, and data collection can be time-consuming. When it comes to self-injurious inmates, there may be many "behavioral relapses"--recurrence of self-injury--before sustained progress is made. A long-distance approach to measuring behavioral change is needed.
In light of these challenges, the authors have tried to keep its outcome measures simple and straightforward. Once behavioral programming is in place, it is important to ask the following questions:
* Are inmates engaging in self-injury less frequently?;
* Are inmates going out to the community hospital as much?; and
* How often do the inmates need on-site emergency medical care?
For the 13 inmates who have received individual behavior management interventions, the authors measured these three parameters retrospectively for the six months prior to implementation of interventions, and then followed the inmates' progress for six months following implementation. Figure 2 shows that, aggregated across all 13 inmates, the individualized incentive plans reduced outside hospital trips by more than 75 percent, reduced self-injury by more than 50 percent and reduced on-site emergency medical care by more than 33 percent.
In BMU, the authors had access to a wider range of outcome data as a result of the close collaboration with MDOC. The authors also accumulated a larger sample of inmates in this group setting. To date, the authors have 29 inmates representing 33 admissions in the BMU sample. Four inmates required readmission following release. For each inmate, the authors measured a range of behavioral indicators for the six months prior to admission to BMU, for the first six months of treatment in BMU (most inmates require one to two years of BMU treatment) and for the first six months following completion of the BMU program and release into general population. Results are shown in Table 2. Using diverse measures of behavior and institutional responses, the data confirm the effectiveness of behavior programming both while the inmate is housed in BMU and after release to general population housing. Similar to the authors' outcome data for individual cases, the unit-based data show a reduction of self-injury by half.
Table 2. Behavioral Outcomes for 29 Behavior Management Unit Inmates Pre-BMU BMU Percent Post-BMU Percent Placement Placement Decrease Placement Decrease (Average (Average During BMU (Average Post-BMU Per Per Placement Per Discharge Inmate) Inmate) Inmate) Number of 1.00 0.70 [down 0.24 [down Use-of-Force arrow] arrow] Incidents 30% 76% Number of 0.31 0.27 [down 0.05 [down Assaults on arrow] arrow] Staff 13% 84% Number of 0.62 0.06 [down 0.20 [down Inmate-on- arrow] arrow] Inmate 90% 68% Assaults Number of 1.69 0.00 [down 1.17 [down Restrictive arrow] arrow] Housing 100% 31% Placements Number of 93.92 0.00 [down 5.15 [down Days in arrow] arrow] Restrictive 100% 95% Housing Number of 0.38 0.12 [down 0.10 [down Inpatient arrow] arrow] Hospital 68% 74% Transfers Number of 20.46 3.42 [down 9.00 [down Days at arrow] arrow] Inpatient 83% 56% Hospital Number of New 2.46 1.85 [down 1.05 [down Mental Health arrow] arrow] Watches 25% 57% Number of 13.84 7.48 [down 7.65 [down Days on arrow] arrow] Mental Health 46% 45% Watch Number of 7.11 3.73 [down 1.10 [down Disciplinary arrow] arrow] Reports 48% 85% Number of 1.08 0.87 [down 0.50 [down Self- arrow] arrow] Injurious 19% 52% Behaviors* One extreme outlier from the data was excluded in the final row. The excluded inmate had required two admissions to BMU and showed a unique pattern. In his case, a burst of self-in-jurious behavior occurred shortly after his first release to BMU and prompted his readmission to BMU, where he continued to for self-in-jure some time. As of this writing, he has been behaviorally stable and free from self-injury for 12 months.
Both individual- and unit-based behavioral interventions can have a large impact on inmate self-injury and other problem behaviors that are otherwise tenacious and "treatment resistant." The data confirms that large reductions in problem behaviors can be realized within the first six months of behavioral programming. The data shows significant reductions in sell-injury, on- and off-site medical care, uses of force, assaults and mental health crises. The authors believe the results reported here contribute to the evidence base supporting behavior management as an effective treatment for these high-need, high-risk inmates. The authors also hope that the results provide support for correctional systems working to develop behavioral programs where they are needed. Although development of these interventions is time- and resource-intensive, inmate self-injury is something correctional systems will either pay for "at the front end"--in advance of the problem behavior--or "at the back end," once inmates have already put their bodies and lives in jeopardy. Implementing such strategies has shown to be effective in reducing overall costs and "unrealized costs," such as staff injury. Planned interventions at the front end work better than crisis interventions at the back end to reduce chronic self-injury. To the authors' knowledge, behavior management is the only evidence-based intervention for reducing chronic self-injury among inmates.
Table 1. Individual Behavior Management Plans: 13 Inmates
* 62% male
* 62% Caucasian, 38% African American
* 77% have committed a violent crime
* Average age: 29 (range: 19 to 39)
* Average length of sentence: 16.5 years
* Average time served: four years
(1.) Jeglic, E.L., Vanderhoff and P.J. Donovick. 2005. The function of self-harm behavior in a forensic population. International Journal of Offender Therapy and Comparative Criminology, 49(2):131-142.
(2.) Andrade, J.T. 2009. Psychopathy: Assessment, treatment and risk management. In Handbook of violence risk assessment and treatment: New approaches for mental health professionals, 241-290. New York: Springer Publishing Company.
Barboza, S.E. and J.S. Wilson. 2011. Behavior management plans decrease inmate self-injury. Corrections Today, 73(5):34.
Wilson, J.S., J.T. Andrade and E. Franko. 2014. Measuring effectiveness of behavior management: Individual- and unit-based outcome data. Workshop presented at 143rd Congress of Correction.
Joel T Andrade, Ph.D., LICSW, is the program manager and director of clinical programs at the Massachusetts Partnership for Correctional Healthcare. John S. Wilson, Ph.D., CCHP-MH is the senior clinical operations specialist at MHM Services Inc. Emily Franko, LICSW, is a clinical operations associate at MHM Services Inc. Jennifer Deitsch is the mental health CQI coordinator at the Massachusetts Partnership for Correctional Healthcare. Sharen Barboza, Ph.D., CCHP-MH, is the director of clinical operations at MHM Services Inc.
|Printer friendly Cite/link Email Feedback|
|Author:||Andrade, Joel T.; Wilson, John S.; Franko, Emily; Deitseh, Jennifer; Barboza, Sharen|
|Date:||Nov 1, 2014|
|Previous Article:||Can corrections heal: Reducing Recidivism and Increasing Public Safety in Virginia.|
|Next Article:||Theater production teaches inmates about teamwork.|