Dynamic psychotherapy has long been recommended for BPD and has now been modified to target the characteristic features of the disorder. Almost all of the first studies examined inpatient treatment using prospective one group pre- to post-test designs.⁽¹⁰⁾ These studies failed to rule out other plausible causes of change, such as passage of time or subsequent outpatient treatment. Stone’s⁽¹¹⁾ report of up to 20 years follow-up of 550 inpatients, most of whom had received some sort of psychosocial intervention, indicated that 66 per cent of patients were functioning well. However, a naturalistic 5-year follow-up of individuals receiving inpatient treatment at the Cassel Hospital in London indicates the need for caution in ascribing benefits to inpatient treatment.⁽¹²⁾ Whilst longer term follow-ups are to be applauded, they are hard to interpret, because other therapies are often given subsequent to the original treatment.

Several nonrandomized trials of dynamic psychotherapy have been undertaken by Stevenson, Meares, and colleagues. In an open trial for 48 patients receiving twice-weekly interpersonal-psychodynamic outpatient therapy for 12 months,⁽¹³⁾ 30 per cent of the treatment group no longer met DSM-III-R criteria for BPD, while the waiting-list group changed little. Cost-benefit analysis found significant reduction in costs, largely attributable to reduced inpatient stays.

A replication study⁽¹⁴⁾ also found significant reduction in symptom severity with the same treatment. A randomized study of 38 patients with BPD compared an 18-month programme of partial hospitalization using mentalizationbased treatment (MBT) with standard psychiatric care.⁽¹⁵⁾ Mentalization entails making sense of the actions of oneself and others on the basis of intentional mental states such as desires, feelings, and beliefs. Outcome measures included frequency of suicide attempts, acts of self-harm, number and duration of inpatient admissions, use of psychotropic medication, and self-report measures including depression, interpersonal function, and social adjustment. After 6 months, patients given MBT showed a statistically significant decrease on all measures in contrast to the control group which showed limited change or deterioration over the same period. This was sustained at the end of the 18 months of treatment and showed further improvement on follow-up after another 18 months. Long-term follow-up 5 years after the initial treatment suggested that the differences between the groups continued, but general social function remained impaired in both groups.⁽¹⁶⁾ The treatment was cost-effective and has been manualized,⁽¹⁷⁾ but as yet the active components remain unclear, and it has not been shown that positive outcomes are correlated with an improvement in mentalizing. (An outpatient version of MBT is currently being evaluated for borderline and antisocial PD in a further randomized controlled trial.) Although promising, this treatment needs further validation by research carried out independently of the originators. Favourable data has recently become available on the effectiveness of a similar programme established in the Netherlands.⁽¹⁸⁾

Transference-focused psychotherapy (TFP) has also shown good results. In a cohort study,⁽¹⁹⁾ 23 female borderline patients were treated for 12 months. Compared with the year before treatment, the number of patients who made suicide attempts decreased significantly, as did the medical risk and severity of self-injurious behaviour. Also, compared with the previous year, there were significantly fewer hospitalizations and fewer days of psychiatric hospitalization. However, one in five patients dropped out of treatment. A subsequent trial compared TFP, DBT, and supportive therapy.⁽²⁰⁾ Ninety patients (all but nine of whom were women) were randomized. At the end of 1 year of treatment, the groups did not differ on global assessment of functioning, social adjustment, scores of depression and anxiety, and measures of self-harm. TFP patients improved significantly more than those receiving DBT or supportive therapy on irritability and verbal and direct assault. Patients who received TFP improved most in reflective function, an operationalization of the mentalization construct,⁽²¹⁾ but it is not known whether improved reflective function relates to treatment gains at follow-up.

Schema-focused therapy (SFT) has been compared with TFP.⁽²²⁾ Treatment was given by therapists with approved training in the treatment methods administered treatment to 88 patients with borderline personality disorder. In an ‘intent to treat’ analysis, patients who received TFP showed significantly less improvement than those who received schema-focused CBT over 3 years, and TFP was more expensive. Both groups showed improvement, but changes in the combined measure of outcome in the schema-focused therapy group were greater and more prolonged than in the TFP group. There are several reasons for caution regarding these conclusions. (i) Differences in outcome between the groups can be accounted for almost entirely by the larger dropout early in treatment of patients treated with TFP and disappear when ‘completers’ are compared. It would be valuable to know why more patients dropped out from TFP at an early stage than from SFT. (ii) Follow-up is required to determine whether treatment gains and group differences are maintained. (iii) In the duration of the treatment period, around 40 per cent of patients could be expected to improve without the treatment.⁽²³⁾ This study also raises the question of how successfully a treatment (TFP) from the US can be transported to a European context.

Noncontrolled studies of day hospital stabilization followed by outpatient dynamic group therapy indicate its utility for BPD.⁽²⁴⁾ Marziali and Monroe-Blum used group therapy focused on relationship management and without the milieu and social components. A randomized controlled trial found equivalent results between group and individual therapy, suggesting that group therapy is more cost-effective.⁽²⁵⁾ Further studies are needed to confirm their findings, especially since dropout rates were high.

CAT has been manualized for treatment of BPD, and many are enthusiastic about its effectiveness. There are some indications that it may be effective. In a series of 27 patients with borderline personality disorder treated with 24 sessions of CAT, half no longer met diagnostic criteria for personality disorder at 6-month followup.⁽²⁶⁾ More definitive statements about efficacy await results of a randomized trial in progress. Ryle (personal communication) has indicated that patients treated with CAT showed significant improvement on a range of clinical measures but reported no difference between people receiving CAT and those undergoing other psychological treatments. Thus, the effects may be nonspecific. A second randomized trial is in progress comparing CAT with ‘best available standard care’ for adolescent patients with borderline personality disorder.⁽²⁷⁾

Cognitive behavioural formulations of BPD are diverse. In a model derived from ‘standard’ CBT and modified for personality disorders, Beck and associates⁽²⁸⁾ define personality in terms of patterns of social, motivational, and cognitive-affective processes, thereby moving away from a primary emphasis on cognitions. However, personality is considered to be determined by ‘idiosyncratic structures’, known as schemas, whose cognitive content gives meaning to the person, and these schemas are the cornerstone of cognitive formulations of BPD. Young⁽²⁹⁾ has developed a treatment programme for BPD based on early maladaptive schemas (EMS). These are stable, enduring patterns of thinking and perception that begin early in life and are continually elaborated. EMS are unconditional beliefs linked together to form the core of an individual’s self-image. Challenge to these beliefs threatens the core identity which is defended with alacrity, guile, and desperation since activation of the schemas may evoke aversive emotions. The EMS give rise to ‘schema coping behaviour’, which offers the best adaptation to living that the borderline has found. Schema-focused therapy (SFT) is only just being evaluated, but its adherence to the general requirements of an effective treatment enumerated above suggests that it should be reasonably successful. The recent report comparing SFT with TFP⁽²²⁾ is described above, but SFT has yet to be shown to be more effective than treatment as usual. It is possible that TFP simply induced more negative effects in patients than SFT.

A small (N = 34), randomized controlled trial assessed brief cognitive therapy, linked to a manual and incorporating elements of dialectical behaviour therapy, in the treatment of recurrent self-harm in people with cluster B personality difficulties and disorders.⁽³⁰⁾

Manual-assisted cognitive treatment (MACT) is a complex sixsession treatment based on the theory that deliberate self-harm and suicide attempts stem from distorted cognitive schemas and coping skills deficits.⁽³¹⁾ It incorporates elements of bibliotherapy, CBT, and DBT, as well as psychoeducation in relation to self-harm and suicide attempts, and a functional analysis of specific episodes. The treatment also involves strategies to regulate emotion, such as distraction, crisis planning, and problem-solving strategies. Cognitive restructuring strategies and management of negative thinking are incorporated in the second phase of the programme, which includes components for the management of substance abuse and relapse prevention. Its brevity makes MACT a potentially valuable intervention from a public health standpoint. In a clinical trial, 34 self-harm repeaters with a parasuicide attempt in the preceding 12 months were randomly allocated to MACT or treatment as usual (TAU). The rate of suicide acts was lower with MACT, and self-rated depressive symptoms also improved. The mean treatment time was 2.7 sessions, and the average cost of care was 46 per cent less with MACT. A subsequent larger study (N = 480) did not find evidence that time to repeat parasuicide was extended following MACT, although there was a decrease in the cost of care.⁽³²⁾ A randomized controlled trial with 104 people with borderline personality disorder with a longer period of treatment (up to 30 sessions) found significant benefit with regard to suicidal behaviour but a nonsignificant increase in emergency presentations in the cognitive behaviour therapy group.⁽³³⁾ The CBT arm used less resources, although no significant cost-effectiveness advantage was demonstrated. Those who received CBT showed less evidence of dysfunctional beliefs, lower state anxiety scores, and less positive symptom distress.

Systems training for emotional predictability and problem solving (STEPPS)⁽³⁴⁾ is a group treatment offered as an adjunct to other treatments rather than as a sole intervention. It is a 20-week manualized programme of psychoeducation and behavioural management focusing on maladaptive schemas and including both professional and family carers. Subjects are encouraged to continue their usual care, including individual psychotherapy, medication, and case management, and are required to designate a mental health professional who would provide ongoing care and could be reached in a crisis. Data from 52 patients suggests some reduction in impulsive and suicidal behaviour and some improvement on measures of depression, but no follow-up data is yet available. An RCT is currently ongoing. In Holland, a retrospective assessment of the experience of 85 patients enrolled in STEPPS groups⁽³⁵⁾ reported significant improvement on all subscales of the SCL-90, particularly those assessing anxiety, depression, and interpersonal sensitivity. Patients and therapists reported moderate to high levels of acceptance for the treatment in both studies. As most of the effective programmes for BPD are long-term and expensive, a short-term efficacious treatment will be of great value. However, even if found to be effective in a clinical trial, its effectiveness will depend on the nature of the treatment as usual.