Psychotherapy for Personality Disorders

This book is number 3 in volume 19 of the American Psychiatric Press Review of Psychiatry, edited by John M. Oldham, M.D., and Michelle B. Riba, M.D.

In this day of managed care, a “pill for every ill,” and the pressures for shorter and shorter therapies, John Gunderson and Glen Gabbard are to be congratulated for assembling a volume that substantiates the centrality of psychotherapy in the care of patients with personality disorders, including the need for long-term treatment in many situations. The importance of this topic is highlighted by the editors’ comment that “the recognition of the suffering inherent in personality disorders may ultimately help treaters to take a more empathic perspective to this large group of patients encountered in daily psychiatric practice” (p. 152).

Most psychiatrists have treated or attempted to treat some of these very difficult patients; they represent 13%–18% of the population! Given this magnitude, it is no wonder that patients with these disorders represent a substantial public health problem because they manifest a great deal of social impairment and use a disproportionately large amount of the general health care system. However, in their chapter, “Empirical Studies of Psychotherapy for Personality Disorders,” J. Christopher Perry and Michael Bond assert that “the view that the personality disorders are intractable to treatment is due for a major revision” (p. 1). These authors stress that, given the current level of knowledge, patients and therapists need to be aware that patients with personality disorders can improve with psychotherapy and, furthermore, that empirical research has helped and can continue to help clinicians decide which directions to follow.

Because Perry and Bond’s chapter is a gold mine rich with data from many sources, I will highlight some of the data in detail. Of 22 studies with at least one treatment arm for personality disorders and with outcome data, all reported positive changes in groups of patients with personality disorder following active psychotherapy. Examination of 20 studies demonstrated very significant changes in self-report and observer-rated measures after active psychotherapy with very large effect sizes (greater than 1.0, p<0.0001).

Three randomized, controlled psychotherapy studies compared active treatment with a wait-list or nonspecific control condition. For self-report measures, there was a moderate to large difference after psychotherapy (effect size=0.78, p=0.002). There was a less strong effect of psychotherapy when observer-rated measures were used. This seems a reasonable finding, given the concept of “illusory mental health”—that people may judge themselves healthier than would an objective observer (1, p. 57). The mean self-report effect sizes of five short-term therapies (16 weeks or less) were significantly larger than those of the seven longer-term studies (1.38 versus 0.92, p=0.02).

Perry and Bond discuss the several possible reasons for this difference. Short-term studies usually include patients who are less ill (those with cluster C personality disorders) in contrast to longer studies, which include mainly patients with borderline personality disorder, who do not experience immediate symptomatic relief. Most important, however, the possibility of a “honeymoon effect” (pp. 17–18) must be considered. Patients may be highly responsive to treatment initially, but their response might diminish with time. Observer-rated improvement seems to be more tied to duration of treatment.

Clearly, patients with personality disorders do not improve as much as patients with nonpersonality psychiatric disorders—as measured by either self-report measures or observer-rated measures. Patients with cluster C personality disorders improve more than patients with borderline personality disorders, who in turn improve more than patients with schizotypal personality disorders. Patients with antisocial personality disorder do not have good outcomes unless depression is also present. The authors hypothesize that “the association with depression may indicate the ability to form attachments and develop a positive therapeutic alliance” (p. 12). Fonagy and Target (2) found that children who experienced anxiety in association with their disruptive behavior or other psychological problems were more likely to respond positively to psychotherapeutic help.

Patients with cluster B and C personality disorders who remain in therapy can have a remission rate up to seven times faster than the natural remission rate for patients with borderline personality disorder. Patients with a cluster C personality disorders (dependent, avoidant, obsessive-compulsive) appear to recover in fewer sessions than patients with cluster B diagnoses (mainly borderline personality).

An important issue discussed by many of the authors in this volume is the issue of dropouts from treatment. Dropout rates vary from 10% to 30% depending on treatment length (shorter treatments have fewer dropouts). However, the empirical data demonstrate that a good therapeutic alliance (or one that is improving) characterizes successful cases. This is an absolutely central criterion in the treatment of patients with borderline personality disorder and antisocial personality disorder, which one study found may be treatable only in the presence of a positive therapeutic alliance.

Most important, the empirical data corroborate the clinical impression that social functioning and basic personality disorder traits improve much more slowly than symptoms do. From a public health standpoint, this is an area where the field of psychiatry has to take the lead—high-quality intensive long-term psychotherapy is a crucial ingredient for many patients and not simply a luxury or a procedure that should be relegated to those with the least training. For example, patients with borderline personality disorder often drop out of treatment if they are assigned to group therapy. Therefore, group therapy should not be imposed on patients with borderline personality disorder. A clinic offering only group therapy (because of cost factors, for example) would do a great disservice to many patients.

The centrality of good training for psychiatrists and therapists working long-term with patients who have borderline personality disorder is highlighted in the chapter by John Gunderson. Gunderson focuses on clinical, mainly psychodynamic, studies. He stresses the centrality of two issues: the impact of intense countertransference reactions in therapists who treat patients with borderline personality disorder and the real potential for boundary violations. Therefore, in the early phases of treatment, it may be important for therapists not to work in isolation so that monitoring by another professional can alert the therapist to oversights or excesses and to contain the inevitable splits or projections and flights. He stresses the importance of good training and the development of long experience treating these difficult patients. With experience one can observe different phases requiring different kinds of interventions or responses from therapists in the treatment of these patients.

For example, in the first phase of treatment (up to 3 months) the therapist will have to be interactive, responsive, and educative. In the second phase, up to a year, he or she may be able to clarify for the patient maladaptive responses to feelings (such as frustration) and can validate the patient’s experience as well as empathize with the patient’s plight. In the third phase of treatment (up to 2 or 3 years), the therapist can identify conflicts and misattributions, while supporting functional capacities. In addition, the therapist can make connections between the patient’s current issues and the patient’s past experiences. In the fourth phase (2–3 years or longer), the therapist can interpret conflicts in the transference and may be able to confront the variegated avoidance techniques used by the patient.

Glen Gabbard discusses the concomitant use of medication with psychotherapy in the treatment of personality disorders. He stresses that empirical research for the efficacy of the combination is quite limited, in contrast to the growing outcome literature supporting the efficacy and effectiveness of psychotherapy for some personality disorders. He cautions that there is virtually no research evaluating the use of one modality alone versus a combination treatment for axis II conditions. However, in studies of pharmacotherapy some patients are also in psychotherapy, and in studies of psychotherapy some patients are receiving pharmacotherapy.

Gabbard stresses that combined treatment has become the clinical standard. For example, rates of comorbid depression in borderline personality disorder range from 24% to 87%. Even in nonpatient samples, 38.5% of patients with a personality disorder have a history of major depression. It is possible that the depression may be a distinct comorbid disorder. However, one has to distinguish between the depressive symptoms in a patients with borderline disorder with the depressive symptoms in someone with a unipolar depression. For example, depressed mood in a patient with borderline disorder is accompanied by loneliness, whereas in major depressive disorder the depressed mood is accompanied by guilt feelings and remorse. Gabbard concludes that it is unclear whether the effectiveness of selective serotonin reuptake inhibitors in patients who have borderline personality disorder is a result of the medication’s antidepressive effects or as a result of its impact on the underlying temperament in these patients.

Gabbard underscores the need for the therapist to help the patient understand the meaning of medication for the patient and to be aware of the possibility of countertransference frustration resulting in a polypharmacy regimen. Gabbard is to be heeded when he describes the potential benefits and the potential complications that arise when psychopharmacology and psychotherapy are administered by one person and when the treatment is split (the psychiatrist prescribes medication and another professional conducts the psychotherapy). If the treatment is split, the clinicians need to collaborate with one another to ensure a more effective treatment and to minimize liabilities.

As a psychoanalyst, I found it refreshing and instructive to read the chapter on cognitive therapy by Peter Tyrer and Kate Davidson. They caution us not to assume that one can easily transpose the effectiveness of cognitive therapy for symptomatic axis I disorders to personality disorders. Very correctly, they recognize that personality disorders require an approach in which the patient’s past needs to be explored, a therapeutic relationship is key to the work, and reworking of beliefs is required. All of these are necessary because a personality disorder is more intrinsic to the person and more ingrained than a symptomatic disorder. A very important contribution in this chapter is the empirical approach advised by the authors, which could very easily apply to the whole range of psychotherapy research. They discuss the use of a single-case method and a diagnostic checklist that can be used to measure the effectiveness of the treatment. As in so many of the studies described in this volume, the problem of dropouts from treatment by patients with personality disorders is an important problem that needs to be empirically evaluated.

Michael Stone’s chapter, “Gradations of Antisociality and Responsivity to Psychosocial Therapies,” is most depressing and hopeless. After reading this chapter, I thought of neurologists of old or alienists in ancient mental hospitals, who were superb diagnosticians but could not communicate any hope for treatment. Stone’s approach is to weed out the very few treatable people from the large group of those with antisocial disorders. He does not give us a sense of how to approach this extremely malignant condition so that future patients with a potential antisocial personality disorder might be treated differently or early enough in life to prevent the full-blown, untreatable syndrome.

This approach of hopelessness is quite a contrast to the rest of this volume. Even Stone’s limited reference to children communicates the same fatalistic attitude. He recommends that we pay close attention to “callous” and “unemotional” traits in children with conduct disorders in order to distinguish children who are “most likely from those who are least likely to benefit from the various treatment approaches currently available” (p. 122). Most striking to me was that in a chapter addressing the psychotherapy of people with antisocial disorders, there were no references to the pioneering work of Peter Fonagy and Mary Target.

In reality, the most forward-looking approach to the treatment of severe antisocial personality disorders available to us at the present time must involve the evaluation of treatments for the antecedents to the condition in childhood. It is clear that the lack of remorse and indifference to others in the older child, adolescent, and adult dates from difficulties during the earliest years of life. Fonagy and Target and others (3–7) have explored how early disruptions in the mother-child bond lead to the child’s inability to conceive of the interpersonal experience in terms of mental states or minds—that another person’s mind is independent of yours. The consequences of this inability can eventually lead to the conduct problems of childhood and the antisociality of adolescence and adulthood. Psychiatrists have to be in the forefront in trying to evaluate programs (long-term and intensive) that attempt to help children increase their awareness of their own feelings, learn to express their feelings in words, and begin to understand the consequences of their behavior, whether the treatment is long-term individual treatment or community-based treatment (8–10). We will then be able to test whether the enhancement of these capacities, whether as a result of in-school programs or intensive individual psychotherapy or psychoanalysis, could result in a greater awareness by the child of the impact of his or her behavior on others. Fostering such an approach would be much more productive than the one outlined by Stone, in which child psychiatrists and adult psychiatrists simply act as gatekeepers and attempt to predict “treatability” or proneness to violence—an approach doomed to failure, since psychiatrists are notoriously bad predictors of future behavior.

The empirical data presented in this volume confirm that only within the context of a relationship can one expect to help patients with severe personality disorders. Children who later exhibit severe sociopathic pathology have difficulties in the internalization of cues for interpersonal interaction and the internalization of the norms of social behavior. These difficulties lead to greater and greater unsocialized behavior. We only have to think back to Freud’s comments on aggression in “Civilization and Its Discontents” (11), where he states, “Homo homini lupus” (Man is wolf to man). Freud described long ago that because of man’s inclination to aggression, civilized society is perpetually threatened with disintegration. He proposed that by promoting the identification with the values of others in the community we can begin to address the problem of aggression.

Children with disruptive disorders have the greatest difficulty controlling their aggressive impulses. In a chart review of the psychoanalyses of 135 children with disruptive disorders matched on demographic, clinical, and treatment variables with children who had other emotional disorders, Fonagy and Target (9), like the authors of the studies in the volume under review, found that, overall, improvement rates were lower for disruptive disorders than for emotional disorders. Furthermore, again like the finding of studies described in this volume, the treatment of nearly one-third of the children terminated within 1 year. However, and most important, of the disruptive children who remained in treatment, 69% were no longer diagnosable at the end of treatment. In other words, Fonagy and Target’s studies, like those in the present volume, indicate that we need to understand what techniques are most effective in engaging and maintaining an ongoing psychotherapeutic relationship. The evidence is clear—if patients with severe problems can stay in treatment, they get better.

Many psychiatrists imagine that psychoanalytic ideas are not subject to rigorous empirical evaluation. Recently, however, Joseph Masling, editor or co-editor of 10 volumes of a series titled Empirical Studies of Psychoanalytic Theories published by the American Psychological Association Press, estimated that “there must be well over 5,000 empirical studies based on psychodynamic ideas; in fact, psychoanalytic theories have proven to be so robustly heuristic they have probably inspired more research in personality than any other set of ideas” (12).

The present volume is one further example of how psychoanalytic ideas can be helpful in the treatment of very difficult patients and how these treatment techniques can be subject to rigorous empirical study.