SUMMARY OF THE WORKSHOP:
Why They Run From Treatment, and What We Can Do

Introduction

Starting in the 1970s, status offenders (i.e., children who are truant, stubborn, or ran away) were no longer treated as criminals in Massachusetts. Rather than being contained in secure institutions, status offenders started to receive services in community-based programs because they run away repeatedly. Hence, containment is again being considered as an option for these youth.
Although youth who run away from treatment centers pose serious problems for both the community-based program and the referring government agency, very little information exists regarding these youth. No studies exist concerning how large the problem is, what program characteristics are associated with reduced running away, or if these youth can be served effectively outside secure institutions. Such information is needed by both program directors and policy planers.

The purpose of our study was to develop data that would increase our understanding of these issues.

Why They Run
Running away can be best understood as any other form of acting out - as a coping mechanism, a way to reduce anxiety. Adolescents use running away as a coping mechanism because it can be so effective.

First and foremost, running away is the most complete way to escape from the stress of milieu treatment. Moreover, running away can be positively rewarding. Adolescents often have fun while they are absent and may experience a sense of adventure that further removes them psychologically from the emptiness of their lives. Also, their need for affiliation can be met from those whom they meet while absent and from the approval they receive from their peers at the center when they return.
Running away also is a most effective way to act out their angry defiance toward staff. When they are absent, staff cannot control them and criticize them. Running can also be a way for them to demonstrate that they are not dependent on staff, that they do not really need them.
When milieu therapy promotes personal growth through confronting maladaptive behaviors, it also creates the stress that motivates adolescents to run away. This is especially true for youth who have a history of coping with stress by running away. Hence running away can be understood as a learned response to the stress that should exist in residential treatment programs.

Running Away: A Special Problem For Group Care
When children in residential treatment express their emotional distress by running away, a variety of problems are created.
The first, most obvious, problem is that the child's safety joins caring adults in the common cause of funding and "rescuing" the child from danger. Eventually, however, this worry and concern leads to conflict among the many members in the child's system of relatives and advocates.
For example, parents may feel the agency is not providing enough protection; funding agencies who are paying for services question the effectiveness and appropriateness of placement; police departments get annoyed. They expect us to be able to handle our problems within our own boundaries; staff second-guess themselves and each other and also look to place blame.
The second major problem is that staff cannot provide treatment to a youth who is absent. Unlike other forms of acting out, when a child runs away, communication and thus learning, are not possible.
Finally, running away is a special problem because of the way it impacts staff. Concerned for the child's safety, staff often feel scared and guilty that they could not protect the child. They may feel incompetent for not stopping the child, or blame others for not intervening effectively. Often staff are left feeling vulnerable, rejected, and experience a sense of loss.
However, staff also often become very angry and rejecting toward the child who has made them feel this way. Due to the distance created by running, staff find it more difficult to be empathetic with the runaway. Hence, running away often does more damage to therapeutic relationships than other maladaptive behaviors.
In summary, running is a behavior that evokes little or no empathy. Moreover, it places the child in danger and creates conflict among helping adults. It always creates distance and often engenders anger and hostility in staff who feel rejected and defeated.

Program Characteristics Related To The Incidence of Running Away - Quantitative Analysis
The Clinical Directors or their designees of the eighteen residential schools in Massachusetts serving adolescents with emotional disturbances and/or behavior disorders were surveyed by telephone and a written questionnaire.
The primary "caveat" or limitations to this study are related to its small sample size, focus on Massachusetts programs, and the descriptive nature of its design.
The results of the study should not be generalized as representing causal relationships, or as representing what would be found in other states. However, given these constraints, there is a great deal of usefulness in the results when interpreted by program decision-makers.

The variables examined were:

(A) percentage of youth "on the run" on a given day (March 9, 1987);

(B) percentage of youth currently enrolled in the program who have run away from the program since their admission;

(C) a combination index of (A) + (B) represents a program's incidence of running as "hi" or "lo";

(D) percentage of youth who terminated after unsuccessful placements for whom running contributed to their unsuccessful placement;

(E) whether or not the program had a policy that instructed staff to restrain youth (under any circumstances) from leaving the program without permission;

(F) the extent to which a program's geographic location is a deterrent to a youth running away;

(G) the average staff-to-client ratio on evening and/or weekend shift (these were then recorded as "Hi" if better than one staff to four youth, or "Lo" if equal o or worse than one to four youth; and

(H) the population served in terms of sex; male only, female only, or co-ed.

Findings

(A) Twenty-six of the 562 adolescents enrolled in the eighteen residential treatment facilities surveyed were on the run on March 9, 1987. This represents between 4 percent and 5 percent of the total enrollment of these facilities.

(B) Twenty-eight percent of the adolescents enrolled in these programs on March 9 have run from the programs at some time during their stay.

(C) Nine programs were determined to be "Hi Run" programs; nine were determined to be "Lo Run" programs.

(D) For twenty-nine percent of the youth who terminated after unsuccessful placements, running was considered to be a problem that contributed to their unsuccessful placement.

(E) Ten programs out of the eighteen surveyed reported that they instruct their program staff to restrain physically a youth who is leaving the facility without permission. (Note: Most of those Clinical Directors responding affirmatively indicated that other circumstances must be in effect in order for the youth to be restrained, such as sufficient staffing for a safe restraint, and/or perception that the youth is a danger to him/herself or others, etc.)

(F) Six of the eighteen programs reported that their geographic location is a deterrent to a youth running away from their program.

(G) Nine of the programs reported an average evening and/or weekend shift staffing pattern of better than one (ten staff person to every four residents ).

(H) Programs were represented as serving the following populations:

Co-ed 6
Female 5
Male 7

Analysis
Two levels of analysis are presented in the tables located at the end of this paper. First, the associations between the various program characteristics and the incidence of running are displayed in Tables 1-4.
Then, more complicated "three-way" associations appear in Tables 5-10.

From reviewing Tables 1-4, it is evident that the incidence of running is strongly correlated with staffing patterns and restraint policy, somewhat correlated with geographic location, and not apparently correlated with the sex of the population served.
However, from reviewing Tables 5-10 it is clear that no one single factor alone is demonstrating a unilateral relationship with running incidence. Rather, a combination of factors is what is actually represented. For example, a restraint policy without a high staffing pattern is not effective in stopping run-away behavior, nor is an isolated location without a restraint policy.

The implications of these finding are that the approach to reducing run-away behavior must be comprehensive, and must focus on many variables simultaneously. Apparently there are no "quick fixes."
In fact, these variables are enmeshed with each other and probably reflect the underlying program philosophy and organizational culture regarding how to work with all youth in the program not just potential runaways.

Program Characteristics Related To The Incidence of Running Away: Qualitative Analysis
We investigated how programs respond to running away through telephone interviews and questionnaires.
We found a wide variety of interventions that programs use to try to work with clients that run away.
The most common interventions used were loss of privileges, behavioral contracts to reinforce positively not running and reinforce negatively running away, and more intensive counseling.
Based on the variety of interventions used, it appears that programs work very hard trying to work with clients who run away. In fact, very few programs terminate a client's placement because he/she ran away. They tend to try to find an effective intervention.
However, these interventions creatively still have high rates of running away. We wanted to know what, if anything, programs can do to decrease drastically running away and work effectively with chronic runaways.
Through telephone interviews and field visits we studied programs that have low rates of running away. We examined more than the interventions these programs used in response to running away.
We focused on the total design of the programs in order to understand what aspects of these programs were different from programs with higher rates of running away.
In analyzing our findings we discovered eight key program variables that influence the rate of running away - regardless of population or location.

(1) Admission criteria and the intake process. Programs with few runs tend to have involuntary admissions policies in which the child receives a strong, clear message that the adults are in control and will keep the child safe. These programs tend to believe that children needing residential treatment cannot be expected to ask for help and placement should not be dependent on their decision.

(2) Approaches to engaging the child in treatment. Because the transition to the residential treatment center is stressful, new residents are most likely to run away. Hence, enormous attention must be paid to structuring the first month of treatment in order to contain and engage the new resident. Low-run programs use a number of practices "to maximize buy-in". Specific practices used by programs successful with reducing runaways include: - limited access to the community, family and friends i.e., no visits from family during first six weeks - limited access to clothes, shoes, and money. - highly controlling and directive child care practices so that the child learns that the adults are in control, consistent and can keep him or her safe.

(3) Level of staff awareness and surveillance of residents. Programs successful with runaways tend to create an aura of security and control by constantly supervising and monitoring residents. Residents always tend to be in sight of staff or they are allowed very little time alone in their rooms; group conversations are monitored as are telephones calls; the bed-checks occur regularly throughout the night.

(4) Limit setting and time-out procedures. The successful programs set limits quickly and consistently. Compliance and safety, rather than increased insight, is the immediate objective. Staff give a clear message that the resident must comply with limits. When necessary, staff do not hesitate to use physical restraint to obtain compliance. Prolonged discussion and negotiation is not tolerated. Clients comply or staff help them comply. Low-run programs tend to confront all unacceptable behavior quickly, and ensure compliance with consequences. Firm limits, consistently implemented by staff, seem to result in more responsible behavior in the program and a significant decrease in all acting out behaviors including running away.

(5) Staff response to attempts to run and successful runs. In low-run programs staff give residents clear messages that they will be stopped from running and are prepared to physically restrain residents if necessary to stop them from leaving the program without permission. If a child is successful at running away, staff almost always are immediately aware of the event and then follow the child and notify the police. In general, staff are very addictive at every stage of possible intervention in order to stop attempts and return runaways. This results in fewer runs, and in fewer self-terminations due to a run.

(6) Highly structured schedule. Low-run programs have schedules that are filled with activities and provide little free time. These schedules are carefully planned. Activities are planed in advance and require social interaction. For example, clients at one treatment center spend very few hours watching television weekly. Activities may be in the community. However, all activities are highly structured by staff.

(7) Staff to client ratio. In order to provide intense attention, constant monitoring, physical restrain, search in the community, while also providing a nurturing, activity-filled milieu, the level of staffing must be intensive. All of the programs that were successful with runaways had intensive staffing patterns of approximately one staff to two residents. These programs tended to state that the number of staff was not important, it was the program that stopped running. Although our study supports the conclusion that the total program design is most important, in order to implement these programs you need intensive staffing patterns approaching a 2:1 ratio.

(8) Program culture and staff training. Every program has its own beliefs, values, expectations, and acceptable behavior patterns - a culture. Low-run programs tend to have strong cultures in which adults are authoritative and directive. The adults are in charge and will keep everything in control. Staff training is critical to the continuation of the cultures. These programs tend to emphasize behavior and crisis management in training so that staff will be able to be, and feel confident in directing and controlling the residents.

SUMMARY STATISTICS

A. Program Characteristics Versus Incidence of Running

Conclusions
We have learned that the issue is not whether or not community-based programs can seriously reduce running away; even with a most difficult-to-contain population - we can.
One residential school in Fall River has had only seven runs in the six months prior to this study and has not had a termination due to running for two years. The new staff- secure programs in Massachusetts also have almost no runs.

Key factors to working with these youth successfully seem to be intensive staffing patterns and a program design that emphasizes a safe, consistent milieu staffed by directive adults.
However, one program has instituted some new policies recently that have significantly reduced, but not eliminated, running away without the addition of any staff.

They include:

• improved admissions and intake procedure including welcome procedure and welcome kit
• immediate start of all treatment modalities for resident and family
• increased focus on helping the adjustment of new students by rewarding staff and other residents with $50 for party at the end of a successful thirty-day stay
• increasing direct-care staff awareness of early warning signals and intervening to stop students from running away
• creating peer support for not running away by rewarding the dormitory group with special funds: $25/week if no runs in group, $10/week if 1-2 runs.

• How do the kids do when they leave these programs? Have they been prepared to make responsible choices independently?
• Have the kids really learned and developed internal controls, or have they merely been forced to comply and become dependent?

As pointed as these questions are, they are not based on any data. Research on these questions is important and necessary. These questions give us future direction.

But the questions themselves do not refute the fact that certain practices result in drastically fewer runs.

Elaine V. Harrington has been working with adolescents for over 15 years. At the Germaine Lawrence School where she has been employed since 1976 she has served as a teacher, child care worker and currently is residential program director. Elaine graduated from Emmanuel College in Boston with a BA in Education and is presently working toward completion of an M.S.W. at Boston College.

Deirdre Maltais Heisler is a graduate of Cornell University's school of Arts and Sciences with a BA in Psychology. Deirdre is currently Administrative Assistant at the Germaine Lawrence School, a residential treatment facility. She performs a variety of administrative tasks as well as coordinates all the admissions.

David Hirshberg has been the Executive Director of the Germaine Lawrence School, a residential treatment center for adolescent girls, for nine years. Previously he had served as a teacher, counselor and administrator in public and alternative schools. He has a doctoral degree from Harvard University and a BA from Brown University.