The American Psychiatric Association's treatment guideline for borderline personality disorder will be published shortly, following a major effort that began with an action in early 1998 by the board of trustees. The workgroup that developed the guideline, of which I am a member, started with an exhaustive literature search, explored clinical and research findings, and discussed parallel and disparate viewpoints during weekly conference calls.

Two parameters of treating borderline patients are very clear: a specific treatment framework must be established, and psychotherapy is a necessary feature of the treatment. The treatment framework includes discussion and clarification of the goals of treatment and the expected roles of the patient and therapist in achieving these goals. Treatment goals such as symptom reduction, improved relationships, and ability to maintain constancy at work should be made explicit. The tasks of the patient, including self-reports of important issues, inner thoughts, dysfunction, and anticipated behavior are elucidated. The clinician's role, which includes providing understanding, consistency, and empathic feedback designed to help the patient achieve the goals, must be understood. A clear plan for the time and place of meetings, the handling of emergencies, and billing and payment is also an important aspect of this framework.

These recommendations are deceptively clear and simple. The actual process of achieving the goals is seldom clear and rarely simple, and establishing the framework is a major ongoing challenge in the therapeutic work. To begin with, the first contact with the patient may be attended by a sense of urgency and by rapidly escalating chaos. Such a situation was recently described to me by a psychiatric resident as we rushed to meetings.

"I don't know quite what to do," the resident said. "I just saw this woman who had missed her scheduled new patient evaluation and who arrived today without an appointment. I had only 20 minutes to see her, and she had all sorts of symptoms—anxiety, dysthymia, panic. She doesn't want medication, and she doesn't want group therapy. When I told her I could see her only briefly, she became angry and upset and called me a pill pusher. I promised to telephone her and arrange a time when we could meet, but I don't know what to do to help her! Do you think she's a 'borderline'?"

By the time the resident had finished his frantic outpouring, my stomach was in knots. I was late for my scheduled meeting, and he and I were in the midst of a process parallel to that of the patient. We needed to step back from the sense of urgency, find a time to discuss how best to help this patient, and develop a plan for supervision. Similarly, the resident needed to schedule time with the patient when he could listen and learn more about her problems without being swept away by her hurried demands.

I suggested that in his next contact with the patient he should say something to ease the sense of urgency, such as, "I'm not quite sure in what way our work together will be most helpful to you. Let's start again and examine what direction you would like to take."

Not all patients with complaints that suggest borderline personality disorder have such a chaotic initial presentation. In many instances the first visit is prompted by symptoms of depression, anxiety, or panic attacks. Only when a complete history is taken will elements pointing to signs of the diagnosis of borderline personality disorder emerge. One will hear evidence of affective instability, chaotic interpersonal relationships, a blurred sense of identity, and difficulty being alone. When the patient begins to make emergency telephone calls to the therapist late at night and on weekends, the therapist begins to recognize that the presenting axis I diagnosis is only a part of the patient's psychiatric problem.

These patients' plaintive cries for help may lead therapists to try to provide more help than is realistically possible. A colleague who once covered my practice while I was out of town asked me years later, "How did Susan Smith make out? I remember her telephone call at one o'clock in the morning and her pained statement, 'I'm all alone. Will you help me?'" Her words stayed with him through the night and the next day, and he remembered them years later.

The borderline patient's fear of aloneness arises from a deep primitive state that often arouses similar painful affects in the listener (1). In this sense the affects are contagious, and the countertransference leads to an instinctive rush to rescue. How did my colleague help the patient? Actually, merely his voice helped her to feel grounded momentarily. But he could not begin to fill the void that she experienced. To pretend to do so would be damaging, stimulating an even greater longing and the belief that if she hurt enough or he cared enough he could really provide the impossible.

In the light of day, the patient needed to explore the anatomy of the anxiety that had prompted the telephone call, to reflect on what responses had and had not been helpful, and to search for alternative means to handle the nighttime demons. In these situations the treatment framework serves as an important reminder to the clinician about the limitations of both the treatment and the treater during the prolonged and tumultuous process of the therapeutic work.

The treatment framework also helps the clinician remain grounded when he or she is emotionally caught up in the swirl of countertransference reactions engendered by the borderline patient's frequent use of the psychological defense mechanism of splitting (2). This primitive mechanism is characterized by a polarization of good feelings and bad feelings, such as love and hate or attachment and rejection. Splitting can be anathema for the clinician treating such a patient. The defense calls for the treating psychiatrist to be a constant, continuing, empathic force in the patient's life as well as someone who can listen well and handle being the target of intense rage and idealization while concurrently defining limits and boundaries with firmness and candor.

This posture requires that the framework always remain clearly in the clinician's mind, even though it is continually challenged and blurred by the patient's illness. Establishing this treatment framework does not mean withdrawing from the patient. On the contrary, it is active involvement that is often needed. Borderline patients' pain is often most intense when they are alone in the middle of the night and on weekends and holidays. They need help finding alternative means to maintain themselves in the absence of a sustaining relationship. Stone (3) has described his "Agatha Christie therapy," in which he prescribed to borderline patients nighttime reading of Agatha Christie's 150 novels or the work of P. D. James or Ruth Rendell. One of his patients found solace in books by Balzac. Stone chose certain works because they dealt with scoundrels who treated women badly.

I have searched the Sunday newspaper entertainment section to find information for patients about free jazz concerts at the county museum, lectures at the public library, book readings, and senior citizen events. To suggest activities relevant to my patients' ages and interests, I have provided directions to the meeting places of self-help groups and to churches where activities are held. Suggesting that the patient participate in these activities is not a substitute for an intimate, sustaining, or permanent relationship, but it is an alternative. Some patients try these activities, and they take on a life of their own, which can be helpful.

Psychoanalytic-psychodynamic therapy, dialectical behavior therapy, and psychoeducational approaches have all proved helpful in working with patients with borderline personality disorder. Common to all successful therapies is the need for a strong therapeutic alliance. Inherent in such an alliance is the importance of a clearly articulated treatment framework. However, clinicians should expect that the parameters of the framework will have to be constantly redefined and restated if it is to evolve into a consistent structure. This does not happen quickly. Short-term treatment has not been shown to be helpful, and a treatment of at least one year should be anticipated.

Is a one-year treatment time frame feasible in today's world of managed care and insurance constraints? Clinical research and experience have shown that it is necessary. A one-year treatment is just as necessary for these patients as a medically indicated kidney, liver, or heart transplant. Those who question the cost-effectiveness of such treatment must recognize that patients who have borderline personality disorder incur inordinate costs through emergency room visits and hospitalizations. Lazar and Gabbard (4) estimated that twice-weekly psychotherapy over a 12-month period saved $10,000 per patient per year (4). At a recent Institute on Psychiatric Services, Sigathy (5) showed that her twice-weekly clinical treatment of a patient with borderline personality disorder resulted in substantial cost savings (5). In the year before the patient began twice-weekly psychotherapy, her psychiatric hospitalization costs were $21,000, and the costs for emergency room visits amounted to $2,520. During the year in which the patient had twice-weekly psychotherapy, she was not hospitalized, and costs for emergency room visits decreased to $280.

Borderline personality disorder is a serious illness. The lifetime incidence of suicide among these patients is estimated to be 9 percent, which is comparable to the lifetime risk of suicide among patients with schizophrenia (6). The intangible costs of the patient's disability and the destructive effect on families are substantial, and they highlight the importance of providing adequate treatment. The process of defining, establishing, and maintaining a treatment framework with these patients courses along a road with unexpected potholes and patches of black ice. Borderline personality disorder is a life-depleting illness. It challenges the best and most creative competencies that clinicians can muster. It demands commitment by troubled patients to the treatment strategy. Treatment gains are long in coming, but the dedicated and persistent clinician finds many rewards in dealing with these despairing patients.

Dr. Goin, who is editor of this column, is clinical professor of psychiatry and behavioral sciences at the University of Southern California School of Medicine. Send correspondence to her at 1127 Wilshire Boulevard, Suite 1115, Los Angeles, California 90017 (e-mail, mgoin@hsc.usc.edu).