Focus on Women: Use of Dialectical Behavior Therapy in a Partial Hospital Program for Women With Borderline Personality Disorder

The lack of reliably effective treatment for borderline personality disorder is a significant public health concern. It is estimated that 11 percent of psychiatric outpatients and 19 percent of inpatients meet criteria for this disorder, three-quarters of whom are female (1). Suicidality and self-injury are common among these patients. From 70 to 75 percent of patients with the disorder have a history of at least one act of deliberate self-harm (2). Estimated rates of completed suicide average about 9 percent (3). Functional deficits may be extreme, comorbid diagnoses are the norm, mental health utilization costs are great, and treatment dropout rates are high (4).

Recommendations for best practices have called for specialized treatment for patients with borderline personality disorder (5). Dialectical behavior therapy, a modification of standard cognitive-behavioral techniques, was designed specifically for outpatient treatment of this population. Developed by Linehan (3), it is currently the only data-supported treatment for borderline personality disorder.

Patients in standard dialectical behavior therapy participate weekly in one hour of individual outpatient psychotherapy and a two-hour skills-training group. Treatment begins with commitment by the therapist and patient to a one-year treatment contract, renewable if improvement is demonstrated. Individual sessions initially target life-threatening behaviors, behaviors that interfere with therapy, conditions that interfere with quality of life such as an axis I diagnosis or lack of daily structure, and mastery of skills.

Four sets of skills are taught in a group setting: core mindfulness, or attention to experience; interpersonal effectiveness, predominantly assertiveness; emotional regulation, or the capacity to experience and modulate emotions; and distress tolerance, or the capacity to tolerate painful emotions without impulsivity. Each day patients record target behaviors, such as suicidal ideation, urges to self-harm, and incidents of parasuicide, on a preprinted diary card. The therapist reviews the card at the beginning of the session to set the agenda.

Therapists must attend a weekly consultation team meeting to discuss patients' progress. The team "treats" the therapists, ensuring their adherence to the therapy model while improving their motivation and professional skills (3).

In randomized clinical trials, dialectical behavior therapy was more effective than usual treatment in reducing suicidal and self-injurious behaviors, dropout from treatment, hospitalization, and self-reports of anger and anxious rumination. Dialectical behavior therapy produced higher rates of global adjustment at the end of a year of treatment (6,7). These gains were maintained over the subsequent year (8).

Standard dialectical behavior therapy has been modified for inpatient treatment. In one study, a psychodynamic inpatient unit adopted this approach. The milieu emphasized the substitution of skillful behaviors for problematic ones related to emotional dysregulation. Compared with an unmodified unit in the same hospital, the dialectical behavior therapy program noted a decrease in the incidence of parasuicide (9).

In another inpatient setting, a goal-oriented, symptom-reduction model of treatment was combined with dialectical behavior therapy. Preliminary findings indicated that, relative to the control group, patients attending a "creative coping" group based on dialectical behavior therapy were more likely to endorse the belief that new skills would help them cope more effectively in the future with self-destructive urges and emotional pain (10).

This paper provides the first published description of a modification of dialectical behavior therapy for partial hospital treatment. Our experience with more than 500 women over the past two years suggests that acutely suicidal women can be safely managed in settings less restrictive than the inpatient unit. Treatment that specifically targets suicidal behavior can divert women from hospitalization and perhaps inoculate them against future admissions. Crisis intervention coupled with ongoing skills training is a promising approach to reduce inpatient recidivism.

Program description

The women's partial program was established at Butler Hospital, a private psychiatric hospital affiliated with Brown University School of Medicine in Providence, Rhode Island, in October 1995. It offers brief, intensive treatment for women between the ages of 18 and 65 on weekdays from 8:30 a.m. to 3:30 p.m. Self-injurious adolescents are admitted on a case-by-case basis. Full census is 12 women. Women are admitted to the program directly from outpatient care or transferred from within the hospital. The minimum expected length of stay in the partial hospital program is five days. In the first year of operation, stays averaged 6.4 days. All patients have private insurance or Medicare.

Women who graduate from the program are eligible for six months of outpatient skills training in the women's extension program without additional charge. Approximately 30 graduates a month enter the extension program.

The program is staffed by two part-time psychiatrists, one of whom is the unit chief. Two full-time nurses, one of whom is the assistant nurse manager, provide the bulk of milieu treatment. Two part-time doctoral-level psychologists provide liaison with patients' outpatient therapists and run the aftercare program. Other half-time staff include a social worker and an activities therapist.

Patient population

Patients often present with comorbid trauma-related diagnoses, anxiety disorders, severe eating disorders, substance abuse, and depression. Approximately, 65 percent of admitted patients meet three or more DSM-IV criteria for borderline personality disorder. Patients with numerous hospitalizations or failed efforts at other treatments are felt to be ideal candidates for the program. Women with thought disorders, cognitive impairment, and mania generally are not accepted. Acute suicidality and self-injury are not exclusion criteria and do not result in transfer. Nearly half the patients receive Supplemental Security Income disability payments for psychiatric illness.

The process that theoretically links the various diagnoses of patients treated in the program can be succinctly characterized as emotional avoidance. Patients engage in dysfunctional behaviors to avoid painful internal experiences or to "leave" the situations that elicit them (11). For example, individuals often describe drinking, self-injuring, or compulsions as attempts to stop painful emotions, flashbacks, or obsessions. In the short run, it may be advantageous to be distracted from painful feelings. When such methods of self-distraction become a chronic and inflexible pattern, psychopathology may result. Thus treatment must enhance patients' capacities for psychological acceptance.

Behavioral milieu

Behavioral interventions focus on extinction of maladaptive behaviors, with a concomitant emphasis on shaping and reinforcing adaptive ones. Therapists use natural positive reinforcers, such as warmth and kind attention, to shape desirable responses while ignoring, as much as possible, ineffective ones. Punishment is the intervention of last resort. Feedback is direct, concrete, and accompanied by suggestions, modeling, or teaching alternative skills. Vigilance for adaptive behavior provides a means to maintain positive contact with even the most difficult patient.

The emphasis on specific description of a patient's behavior, rather than categorical distinctions, improves the team's motivation and engagement in the milieu. Dialectical synthesis of a team divided in its opinion of a particular patient can be achieved through reformulating pejorative interpretations of a patient's behavior. For example, rather than labeling a behavior as resistant or as indicating a patient's sense of entitlement, concrete observable behaviors are described_for example, by saying that the patient did not do her homework again or the patient interrupted the discussion ten times. The team more easily reaches agreement about which behaviors to reinforce and works together to create a stimulus context in which to elicit and shape new behaviors.

For example, Ms. C, a patient in the program, cringed and apologized excessively when making requests. She sat slumped in her chair during groups, apparently miles away. The staff reacted to her with irritation. After a spirited team meeting, the staff agreed to reinforce eye contact, behavior incompatible with both "spacing out" and submissiveness, and to extinguish cringing by ignoring all requests she directed toward the floor. Ms. C was oriented to the plan, and in a transaction of mutual reinforcement, she quickly learned to lift her head, make good eye contact, and use an assertive tone of voice when speaking. When staff noted these behaviors, they responded with warmth, attention, and approval.

Therapists must guard against inadvertently reinforcing maladaptive behaviors in the partial program. Therapists are required to react counterintuitively; for instance, they do not chase patients who abruptly run out of the group, leaving behind those struggling to stay engaged. However, elopement is rare, and no suicide attempts have resulted from this practice. Likewise, patients with a history of violence are taught to walk away from the unit when dangerously aroused. In two years of operation, no incident has occurred in which physical restraint has even been considered, although use of restraints had been common in the staff's inpatient experience.

We are most careful to avoid reinforcement of acute suicidality. For example, Ms. B came to her session distraught. Her former husband had taken emergency custody of their children, filing a petition to make it permanent. He had reached his limit with her suicidality and impulsivity. Weeping and hopeless, she responded that she intended to go home and kill herself that night. She also reported that she had amassed a lethal supply of amitriptyline.

Crisis management consisted of Ms. B and her therapist brainstorming alternative strategies for coping with her fear and of the therapist directly challenging her idea that suicide was a solution. Ms. B's therapist joined with her, stating that "we will work on finding another solution," while highlighting the fact that her former husband had a point: her behavior must change if she wanted to be able to retain her maternal rights.

Emotions validated included Ms. B's anger at her former husband, shame that he considered her an unfit mother, guilt over pain she had caused her children, and anxiety over the separation from them. The therapist blocked her escape from the situation by highlighting how suicidal behavior had resulted in her former husband's current actions and was not the way to remedy the problem. The therapist noted that the partial program had helped other women cope with similar difficulties and could help Ms. B. The session ended after Ms. B agreed to call the therapist from home after she had dispatched the medications down the toilet. Before leaving for the day, Ms. B told the group about her situation and received support and sympathy.

When Ms. B called her therapist that evening, she was overwhelmed by the empty house and was again acutely suicidal. The therapist insisted she flush the medications down the toilet, and after she did, the therapist helped her plan distracting and soothing activities for the evening. Hospitalization was averted, treatment continued, and Ms. B refocused her attention on securing resources to help her through the next months.

Group program

A patient's short stay in the program requires that skills training be limited and focused on acute stabilization. Through the frame of dialectical behavior therapy's core skill of observation without judgment, patients receive overviews of the four skill modules that are taught in full in the six-month aftercare group program, the women's extension program. Patients are trained in effective participation in skills-based groups, rather than process-based groups.

Patients are oriented to the biosocial theory of borderline personality disorder. In Linehan's theory, borderline pathology is thought to result from the interaction of a biological emotional vulnerability and a pervasively invalidating environment in which private experiences, especially emotions, are rejected as incorrect, faulty, or otherwise invalid. The punishment of emotional expression, coupled with reinforcement of maladaptive and escalating behavior displays too disruptive to be ignored, leaves the patient alternating between compulsive suppression of emotion and exaggerated expression.

All patients attend all groups offered in the women's partial program. Each day begins with brief mindfulness practice, in which patients focus on observation of their experience, followed by review of the previous day's skills homework. The midmorning group teaches the skill of the day, which may relate to thoughts, attention, relationships, emotions, or distressing situations. Standardized homework from the dialectical behavior therapy skills package is assigned (12), with the expectation that it will be practiced on current life problems. Occupational therapy is scheduled before lunch.

Various theoretical groups lead off the afternoon, providing opportunity for cognitive restructuring and orientation to theory. The groups are titled anatomy of a crisis, use of reinforcement, dialectical behavior therapy's biosocial theory, approaching anxiety, and radical acceptance. Late-afternoon group time is spent in skill practice. All patients in the program meet briefly again at the end of the day.

Individual therapy

Each patient meets daily with her assigned psychiatrist for individual therapy. The therapist works for quick attachment to the patient, as well as quick detachment. The therapist orients the patient to dialectical behavior therapy, trains her in the role of the patient in behavioral treatment, prepares her for discharge and aftercare, and instills realistic hope for improvement. Medications are also managed in this relationship.

The therapist and patient collaborate in choosing a focus for the therapy. The situation that precipitated admission to the program is framed behaviorally. Analysis reveals its antecedents and consequences, and whether it is an example of an overall pattern of problem behaviors. Precursor behaviors may become the focus of the milieu treatment, as with Ms. C, the cringing patient. The treatment target is monitored by the patient on her diary card, as are parasuicidal behavior and acute affective shifts. Such incidents are analyzed to determine factors contributing to the patient's vulnerability, such as failure to eat, disrupted sleep, or substance use; common precipitants of undesirable behaviors; skill deficits; and emotional obstacles to skillful behavior.

Behaviors that interfere significantly with quality of life but that are not targets of treatment may be delegated to the milieu for much of their management. For instance, disordered eating, which is very common among patients in the program, can be addressed through consultation with the Butler Hospital's eating disorder team. Approaches include psychoeducation, food logs, or supervised meals. Similar protocol approaches are under development for substance abuse and dissociation.

The individual therapy is directive and seeks to motivate the patient. It is fast paced, relentlessly present focused, and problem solving. The therapist is respectfully confrontational of avoidance but also collaborative, balancing acceptance of the patient and the necessity for change. Strategies may vary depending on the patient's estimated stage of change (13). An entire treatment could be productively structured around development of the functional use of the diary card.

A brief case vignette illustrates the role of individual therapy in the program. Ms. A was admitted for escalating self-injury due to intrusive recollections of being raped when she was 15. Analysis of her self-injury clarified that it occurs after she has sex with her current boyfriend. She is actually quite angry at him for many reasons but does not assert herself for fear of losing the relationship entirely. Work with Ms. A could be productively focused on analyzing any incidents of self-injury that occur during treatment to identify common precipitants, teaching her how to speak up despite her fear, teaching her to shape the boyfriend's behavior with positive reinforcement, making an environmental intervention to support her leaving him if that were her wish or arranging couple therapy to consolidate their attachment, or teaching her skills to manage her fear and anger without hurting herself. Detailed discussion and emotional processing of the rape would not be indicated for a patient who clearly is unable to regulate the emotions likely to be aroused by such treatment.

The aftercare program

The original goal of the women's extension program was to provide six months of dialectical behavior therapy skills training, as outlined in the Linehan manual (12), for all successful graduates without additional charge, reflecting concern with the severe limits placed on outpatient benefits by most insurance providers in the state. This altruistic stance did not take into account the strain on resources resulting from the discharge of 30 patients a month from the partial hospital program.

The skills training is offered in four drop-in groups a week that rotate through all of the modules and provide both homework review and presentation of a new skill to as many as 80 patients a week. Patients are encouraged to choose a home group in which to participate actively, but they may sit in on all groups. Weekly skills lectures are presented to the general community and are well attended by patients, families, interested therapists, and others. Through a patient-produced newsletter, a self-help group, parties, and the opportunity for patients to informally ask staff questions about the skills training, the women's partial program supports patients' sustained attachment to the treatment team while pushing them to establish a life in the community that exceeds the reinforcements offered in the mental health system.

Length of stay

The program was designed to be brief for several reasons. First, dialectical behavior therapy emphasizes the necessity of learning the skills necessary to manage life's problems in the setting where they will be required: in the outpatient world. Most of the patients have serious problems that are unlikely to be resolved even during a lengthy hospital stay. Therefore, the goal is to support the outpatient therapy by offering a rich program of ongoing contact with the unit_contact that does not require hospitalization to maintain.

Second, the relatively fixed length of stay in the partial program provides an intentional severance of the link between illness behavior and more treatment. In general, the mental health system reinforces parasuicide with more intensive contact and punishes recovery with withdrawal of treatment. In this program, lack of improvement or worsening symptoms are seen as explicit reasons not to extend partial hospitalization, as it would be unethical to continue ineffective treatment. A brief extension, rarely more than one or two days, is possible if the patient requests one, is improving, and can articulate the specific goals she wishes to accomplish in the extra time.

Intentionally brief treatment has a number of unanticipated benefits. The fixed length of stay functions as a treatment contract, to which both parties agree when the patient enters the program; other studies have suggested that contracts are beneficial for patients with borderline personality disorder (14). It is relatively easy for patients to commit to and successfully complete treatment that is short and requires minimal disruption of their ongoing lives and work. Little time is available for the intense milieu conflicts of longer stays to flourish.

The potential for struggle around discharge is obviated. The patients come into the program parasuicidal and leave that way; there is no expectation that extending the stay will result in substantial change in such well-established behavior. It is not unusual for a patient and therapist to spend the last individual session analyzing the patient's parasuicide of the night before. The team's comfort with discharging symptomatic patients is, of course, substantially maintained by the availability of the aftercare program and recognition that no empirical evidence has been found to support the efficacy of inpatient treatment for chronic suicidality.

Ms. E's case illustrates a patient's experience in the program. On the day Ms. E entered the program, she was oriented to the short duration of treatment and the circumscribed goals. She responded that she had been cutting herself for 25 years and that it was "ridiculous" to think that a five-day program would help. The staff member to whom she made this remark validated her hopelessness by saying that it was not unreasonable, given all her previous treatment with so little success, and expressed confidence that Ms. E would leave the program with a better understanding of how self-injury functioned for her and how else she might cope with her problems.

At opportune junctures Ms. E was oriented to the aftercare program and told about the need for ongoing practice to change behavior. To prepare her for discharge, her prior outpatient therapy was assessed for problems that might impede return to the relationship with that therapist. Her outpatient therapist was provided with information about dialectical behavior therapy and was relieved that she would have more support in managing this difficult patient.

In individual therapy on the third day of the program, analysis of an episode of Ms. E's parasuicide suggested that her fiancé's shaming attacks on her intelligence frequently precipitated self-injury. Solutions were proposed, from more assertive approaches with him to cognitive restructuring of self-invalidation, and Ms. E practiced them in the appropriate groups. In the relationships group, she decided to use the family meeting scheduled for her last day to confront her fiancé about his attacking language.

On the fourth day, the therapist asked Ms. E if she felt ready for discharge the next day. Ms. E asked for an extension of three days because she felt she missed a lot of the group material in the first days. Reminded that aftercare would revisit these skills, she and the therapist negotiated a one-day extension to the following Monday, so that they could process whatever happened in the family meeting and Ms. E would have a weekend to practice crisis survival skills, given her newfound hope that she could stop hurting herself. The extension was granted, reinforcing her self-assertion.

On the fifth day the family meeting went well, but on the way home Ms. E and her fiancé got into an argument, and she cut herself. She returned on Monday, the sixth and final day, discouraged and despondent. The last individual meeting included a functional analysis of the weekend incident of self-injury. It was noted that she actually aborted the episode before causing serious injury, a new behavior. The therapist provided corrective feedback about abandoned opportunities for more effective choices, as well as encouragement to keep trying. Ms. E was actively encouraged to come to aftercare to learn more and was discharged as planned.

Conclusions

Two years of operation of the women's partial program provides promising anecdotal evidence that dialectical behavior therapy, which is an outpatient approach, can be effectively modified for hospital settings and a more diverse patient population. Our experience with more than 500 women certainly suggests that such treatment may be a feasible alternative to inpatient hospitalization for many suicidal women with borderline traits. The program goals of diversion from acute hospitalization, long-term reduction of reliance on hospitalization for crisis management, and reduction of parasuicide by specific behavior therapy remain to be tested, but further study in a more controlled fashion is clearly justified.

The authors are affiliated with the women's partial hospital program at Butler Hospital in Providence, Rhode Island. Dr. Simpson, Ms. Pistorello, Dr. Begin, Dr. Costello, Dr. Pearlstein, and Dr. Rosen are also with the department of psychiatry and human behavior at Brown University School of Medicine, 345 Blackstone Boulevard, Providence, Rhode Island 02906 (e-mail, [email protected]). This paper is one of several in this issue focused on women and chronic mental illness.