Overall, results from RCTs have found that a number of cognitive-behavioral (DBT, Schema Focused Therapy) and psychodynamic treatments (Mentalization Based Therapy and TFP) have efficacy, although outcomes are inconsistent with the exception for parasuicidality (especially for DBT in comparison to TAU and with highly parasuicidal patients). In addition, power is generally low and, although attrition has been reduced in the experimental conditions, it still remains a problem. As pointed out by Rossi (1990), low power is low power, and finding effects in low-powered studies is problematic. He outlines a number of reasons for this conclusion, noting that besides the obvious reason that low power results in an inability to detect a true difference, low power can also result in false positives. Rossi (1990) points out that in low power studies, the chance of Type II errors is only slightly more than the chance of a Type I error. This is because studies with low power are susceptible to the undue influence that may be exerted by outliers. Although this issue is less so with nonparametric tests, it remains a problem and is compounded by the fact that there are no good tests of power for nonparametric tests. Finally, low power often results in an inability to test alternative hypotheses for findings. For instance, if one wanted to test for therapist effects, or patient effects, a small sample size would make it unlikely that these effects could be identified in the data and conversely more likely that an outlier could cause an effect to be found. Generally speaking, domains of change are limited (e.g., focus on symptoms) and few studies have examined patient predictors of outcome (sans parasuicidality, inpatient status). Most importantly, thus far, few studies have investigated specific mechanisms of action or change (Clarkin & Levy, 2006; see Levy et al., in press, for an exception). Finally, given the chronicity of personality disorders, none of the studies have sufficient follow-up as yet that would determine the maintenance of treatment effects and clarify the long-term course of BPD after treatment termination.
Although there is accumulating evidence from outcome studies suggesting the effectiveness and efficacy of a number of different treatments for BPD, the probative importance of these studies for understanding a treatment's actual mechanisms of action are both indirect and limited (Garfield, 1990). Therefore, despite the support for the effectiveness and efficacy of existing treatments for borderline personality disorder, researchers are still confronted with a high degree of uncertainty about the underlying processes of change. Additionally, validation for the treatment occurs to the extent that the theoretically specified mechanisms of change are actually related to the treatments' effectiveness. It is very possible that these treatments may work due to unintended mechanisms such as typical common factors (e.g., expectancies; see Weinberger, 1995) or a specific technique factor that is essential to good outcome but not necessarily unique to any one treatment.
Along these lines, Bateman and Fonagy (1999) suggest that essential mechanisms in the treatment of BPD are a theoretically coherent multicomponent treatment approach, a focus on relationships, considerable efforts aimed at reducing dropout rates, and consistent application over a significant period of time. These components are consistent across studies examining MBT, DBT, TFP, schema focused therapy, and CBT and may explain the better-than-expected results as compared to treatment-as-usual groups and studies of naturalistic follow-ups, particularly with regard to the issue of attrition from treatment. All of these treatments provide principle-based manuals and institutional supports such as ongoing supervision, not only to stress specific techniques, but also to metabolize countertransference and to minimize iatrogenic effects of therapist enactments. Additionally, each of these treatments invests considerable efforts to increase communication between different treaters (e.g., individual therapist and psychopharmacologist).
Specific questions have been raised to various aspects of these different treatments. For example, given the considerable efforts geared toward supporting therapists, one could ask, "Does DBT training or supervision reduce therapist burnout?" The data, to date, suggest not (Little, 2000; Linehan, Cochran, Mar, Levensky, & Comtois, 2000). Little (2000) found that DBT training reduced burnout scores on the Personal Accomplishment component of the Maslach Burnout Scale (Maslach & Jackson, 1986), but did not reduce burnout on the Depersonalization and Emotional Exhaustion components. Linehan et al. (2000) found that the best predictor of DBT-trained therapists' burnout was patient's pre-treatment burnout.
Another question that arises is: "Are treatment contracts useful?" One of the important tactics in TFP is the use of treatment contracts, which occurs before the treatment begins. The function of the contract is to define the responsibilities of patient and therapist, protect the therapist's ability to think clearly and reflect, provide a safe place for the patient's dynamics to unfold, set the stage for interpreting the meaning of deviations from the contract as they occur later in therapy, and provide an organizing therapeutic frame that permits therapy to become an anchor in the patient's life. The contract specifies the patient's responsibilities, such as attendance and participation, paying the fee, and reporting thoughts and feelings without censoring. The contract also specifies the therapist's responsibilities, including attending to the schedule, making every effort to understand and, when useful, comment, clarifying the limits of his/her involvement, and predicting threats to the treatment. Essentially, the treatment contract makes the expectations of the therapy explicit (Clarkin, 1996). There is some controversy regarding the value of treatment contracting. The APA guidelines recommend that therapist contract around issues of safety. Others (Sanderson, Swenson,
& Bohus, 2002) have suggested that the evidence contraindicates their use and shows them to be ineffective (Kroll, 2000). However, the Kroll (2000) study was designed to determine the extent that no-suicide contracts were employed (which was found to be 57%) and, although 42% of psychiatrics who used no-suicide contracts had patients that either suicided or made a serious attempt, the design of the study does not allow for assessment of the efficacy of no-suicide contracts. Other data suggest the utility to contracting around self-destructive behavior and treatment threats (Yeomans et al., 1994; Smith et al., 1995; Clarkin et al., 2001; Clarkin et al., 2006; Levy, Clarkin, Schiavi et al., 2006). For example, Yeomans and colleagues (Yeomans et al., 1994) in a pre-post study of 36 patients with borderline personality disorder found that the quality of the therapist's presentation and handling of the patient's response to the treatment contract correlated with treatment alliance and the length of treatment. In addition, in our earlier work on TFP (Smith et al., 1995), when we did not stress treatment contracting, our dropout rates were high (31% and 36% at the three month and six month marks of treatment, respectively). However, based on the findings of Yeomans et al. (1994), Kernberg and colleagues further systematized and stressed the importance of the treatment contract and in later studies (Clarkin et al., 2001; Clarkin et al., 2006; Levy, Clarkin, Schiavi et al., 2006) our group found lower rates of dropout (19%, 13%, and 25%) over a year-long period of treatment. These findings taken together suggest that sensitively but explicitly negotiated treatment contracts may have at least one of the desired effects: resulting in less dropout and longer treatments. Future research will need to address the issue of treatment contracts more directly, particularly testing the effects on parasuicidality and suicidality.
Another question that arises with regard to DBT concerns the evidence for the skills group as a mechanism of change in DBT. Linehan suggests that the skills group is a key mechanism of change (Koerner & Linehan, 2000; Linehan, 1993; Lynch et al., 2006). Patients and therapists also view skills groups as critical for improvement (Araminta, 2000; Cunningham, Wolbert, & Lillie, 2004; Miller et al., 2000; Perseius, Ojehagen, Ekdahl, Asberg, & Samuelsson, 2003). However, the data available to date would suggest otherwise. Linehan et al. (2002) compared standard DBT to Comprehensive Validation Therapy with a 12-step program and found similar outcomes in the two treatments, suggesting that validation and not skills training may be the active ingredient in DBT for substance abusing BPD patients. Contrary to the recommendations of Linehan (1993), Turner (2000) modified DBT skills by removing them from the traditional group format and incorporating them into the briefer individual sessions (as well as incorporating psychodynamic techniques). Turner also provided patients in both the experimental and control conditions with six sessions of a modified DBT skills group. Turner found that the psychodynamically and skills modified DBT was more effective than the client-centered therapy with modified skills groups. This finding suggests that skills groups can be integrated into individual sessions and with psychodynamic techniques. The only study we could finding looking at the acquisition of skills in DBT was a dissertation by Puerling (2000). She found increases in skill usage over time but failed to show any relationship between changes in skills and outcome.
Is there evidence that increased reflective function (RF; Bateman & Fonagy, 2004) is the mechanism of change in MBT? Although it is tempting to hypothesize that RF is the mechanism of change in MBT and that the increases in good outcome continue after treatment termination due to change, in RF, there is no direct evidence to suggest that RF changes in MBT. Indirectly, findings from Bateman and Fonagy's (2001) follow-up, in which they find continued improvement in their MBT treated patients, suggest some internal change akin to RF may have taken place. There is evidence, however, that RF changes in Kernberg's TFP treatment (Levy et al., in press).
What patient variables predict outcome for BPD? There is surprisingly little data about patient characteristics as predictors of outcome in the treatment of BPD. Fonagy et al. (1996) found that pretreatment RF did not predict outcome for 85 outpatients with BPD; however, attachment status did. Those patients with dismissive attachment, as compared with those with enmeshed preoccupied attachment, showed significantly greater increases in GAF scores. Levy-Mack, Jeglic, Wenzel, Brown, and Beck (2005) examined the relation between patient attitudes toward treatment and outcome in a sample of patients seeking CBT for BPD. Those who had positive attitudes toward treatment, as opposed to those with negative attitudes toward treatment, were more likely to experience greater decreases in the number of BPD and depressive symptoms despite attending fewer therapy sessions than those with negative attitudes. These results suggest that techniques designed to enhance patients' attitudes toward treatment could increase the likelihood of benefiting from treatment. Linehan et al. (2000) found that patient pretreatment burnout predicted therapist burnout at four months into treatment. Yeomans et al. (1994) found that impulsivity was negatively related to the length of treatment. Smith et al. (1995) found that patient hostility and younger age predicted dropout from treatment. What therapist's factors predict outcome in the treatment of BPD? Linehan et al. (2000) found that high expectancy for therapeutic success leaves therapists vulnerable to increased emotional exhaustion at a later point.
In sum, little is known of the mechanisms by which treatments for BPD actually work or what actually happens to the patient that results in change. Preliminary evidence suggests that theoretically coherent, relationship-focused treatments that place considerable efforts on reducing dropout, increasing communication with auxiliary treaters, and providing ongoing supervision of therapists are important. There is some evidence that skills groups may not be the mechanism of action in DBT and that increasing the patient's capacity to think about mental states may be the mechanism of action in psychodynamic treatments. Regarding patient and therapist factors, less is known, but hostility, impulsivity, and young age appear to be risk factors for a higher client dropout rate.
Linehan's (Linehan et al., 1991) seminal randomized clinical trial of DBT was a breakthrough for the research on BPD; the treatment has quickly gained popular acceptance. A number of managed care companies now define special benefits for DBT. Several state departments of mental health (Illinois, Connecticut, Massachusetts, New Hampshire, North Carolina, and Maine) have now enthusiastically endorsed and subsidized DBT as the treatment of choice for BPD or have mandated DBT training for state employees working with seriously disturbed patients. In Massachusetts, former DBT patients can now be reimbursed for coaching current DBT patients. Hundreds of marketing, seminars, and training programs in DBT are provided for inpatient and outpatient clinics, correctional institutes, and community treatment centers. Certainly, Linehan's efforts to develop, examine, and given the seriousness of BPD, to disseminate DBT are laudable. Her 1991 study was influential and changed the face of psychotherapy research; however, concerns have been raised that the dissemination of DBT has exceeded the evidence base, particularly with regard to state legislation and insurance reimbursements (Corrigan, 2001; Scheel, 2000; Smith & Peck, 2004; Westen, 2000). There is no doubt that the empirical base for DBT, in terms of the sheer number of studies, is stronger than for any other treatment. However, the actual findings themselves may not be as strong as developing folklore. The Cochrane Review (Binks et al., 2006) meta-analytic findings suggest that although some of the problems, particularly parasuicidality, may be amenable to DBT, it remains "experimental and the studies are too few and small to inspire full confidence in their results." In addition, there are a number of other treatments, including cognitive-behavioral and psychodynamic-based treatments, which warrant serious consideration.
Viewing the BPD treatment literature from a broad perspective, there is support from various levels of scientific rigor for the effectiveness (and in some cases, efficacy) of psychodynamic, interpersonal, cognitive, and cognitive-behavioral psychotherapies for treating BPD. In addition, evidence suggests the combination of individual psychotherapy with skills-based, psychoeducational, and family therapy groups. Although DBT (Linehan, 1993) has been the most extensively studied treatment for BPD in RCTs, there is emerging evidence for the effectiveness and efficacy of psychodynamically oriented treatments such as MBT (Bate-man & Fonagy, 1999, 2001, 2004) and TFP (Clarkin et al., 2006; Levy, Clarkin, Schiavi et al., 2006), cognitive (Brown et al., 2004) and cognitive-analytic treatments (Ryle & Golynkina, 2000), and interpersonal psychotherapy (Meares et al., 1999). In addition, there is preliminary evidence to suggest that DBT might be more efficacious for highly parasuicidal BPD patients than it is for those who are less parasuicidal (Verheul et al., 2003), and that TFP might be more efficacious than DBT in generating changes in personality structure (Levy et al., in press). Further research examining the factors that moderate outcome in the treatment of
BPD can help to verify or refute these hypotheses. In addition, there is evidence to suggest that psychodynamic therapists can learn and apply DBT well, that psycho-dynamic techniques can be integrated into DBT, and that DBT skills groups can be modified and even incorporated into individual sessions (Turner, 2000). These issues warrant further study.
With the heterogeneity of BPD presentations, the question should not be simply "which treatment is most efficacious for treating BPD?", but rather, as Gordon Paul (1967) suggested "What treatment, by whom, is most effective for this individual with that specific problem, and under which set of circumstances?" (p. 111). We would also add "and by what mechanisms?" The maximization of treatment effects depends upon the examination of mechanisms of change, both at the level of changes within the patient as well as at the level of the specific techniques that affect such changes (Levy, Clarkin, Yeomans et al., 2006).
It is hoped that this chapter has demonstrated that, although RCTs are important in the evaluation of psychotherapy for BPD, they can be restricted in their explanatory power, external validity, and ability to identify mechanisms of change. Limitations of existing RCTs include the lack of adherence and competence ratings (Linehan et al., 1991; Bateman & Fonagy, 1999). Without knowing which techniques are prescribed and proscribed by the experimental treatment and whether or not therapists adequately followed the principles and techniques of a given therapy, inferences regarding the components of therapy that actually lead to change cannot be made. Future studies of psychotherapy for BPD could be improved by utilizing treatment manuals for each treatment condition, additional efforts to maintain the integrity of each treatment (e.g., evaluating adherence, competence, and expectancies of therapists in both experimental and control conditions), measurement of multiple domains of outcome (i.e., structural and interpersonal change, as well as symptom reduction), long-term follow-up evaluations, and examination of moderating and mediating factors in treatment outcome. Multiple assessment points during treatment studies are especially important for evaluating trajectories and mechanisms of change in psychotherapy for BPD.
Was this article helpful?
Suffering from Anxiety or Panic Attacks? Discover The Secrets to Stop Attacks in Their Tracks! Your heart is racing so fast and you don’t know why, at least not at first. Then your chest tightens and you feel like you are having a heart attack. All of a sudden, you start sweating and getting jittery.