Meta-analysis of mentalization based treatment and schema focused therapy for borderline personality disorder : what works for whom?

(1)

Meta-analysis of mentalization based treatment and schema focused therapy for borderline personality disorder: What works for whom?

Caspar van Eijk

Universiteit van Amsterdam

Studentnummer 0013021

Masterthese voor de afstudeerrichting Klinische Psychologie Begeleider: dr. A. A. P. van Emmerik

Externe begeleider: dr. M. J. Noom (senior onderzoeker Zaans Medisch Centrum, afdeling psychiatrie)

(2)

ABSTRACT

Background: According to several studies, psychotherapy is the first choice for treating

borderline personality disorder (BPD). In the last decade various specific interventions have been developed. Research found evidence for specific interventions being more efficacious than unspecific control interventions. The question now is if there is one treatment that is more effective than other treatments in treating BPD. The purpose of the present meta-analysis was to evaluate the efficacy of Mentalization Based Treatment (MBT) and Schema Focused Therapy (SFT) in order to compare them and see if one of the two is more suited for treating BPD.

Methods: The following databases were searched in order to find studies that met our

inclusion criteria: Pubmed, Embase, Cochrane and Psycinfo. Eight studies met these criteria. The studies included a total of 488 participants of which 420 were female (86%) and 68 were male (14%).

Results: MBT, SFT and unspecific control interventions resulted in significant and substantial

efficacy on primary and secondary outcomes. In all included studies MBT, SFT and controls yielded positive effect sizes for BPD symptoms and general pathology and general

functioning. Remarkably there was no significant difference between the various treatment modalities.

Conclusions: MBT and SFT are evidence based treatments for BPD and they are well

matched when it comes to efficacy on various treatment outcomes. Unspecific control interventions on the other hand are not to be regarded as inferior to specific psychotherapies. It is suggested that case severity may function as a moderator explaining differences in efficacy between MBT and SFT.

(3)

INTRODUCTION

According to the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), borderline personality disorder (BPD) is characterized by a pervasive pattern of instability in interpersonal relationships, identity, impulsivity, and affect. People with BPD often suffer from moodswings that are very intense and difficult to predict. Their perspective on relationships and people can change from one moment to the next and the fear of being abandoned is always on the fore. Furthermore, BPD is associated with self-destructive

behavior such as self-damaging by cutting or burning oneself. In practice the symptoms cause severe suffering and have a major influence on the quality of life for the patients as well as the people directly involved. BPD is considered as one of the most complex personality disorders because of its wide range and rapid fluctuation of symptoms (Franken, Muris & Denys, 2013).

In the general population 1 to 2% of adults are diagnosed with BPD. In psychiatric populations up to 10% of outpatients and 20% of inpatients are diagnosed with BPD

(Laurenssen et al. 2014). Because of the severity of the disorder, BPD is associated with high rates of suicide, severe functional impairment, high rates of comorbid mental disorders, intensive use of treatment, and high costs to society (Leichsenring, Leibing, Kruse, New, & Leweke, 2011). Societal costs manifest themselves in various different domains. Research suggests a strong relationship between personality disorders and reduced productivity due to absence from and inefficiency at work. Furthermore there is evidence of a high demand on health, criminal and social care services (Soeteman, Hakkaart – van Roijen, Verheul & Busschenbach, 2008). Maclean, Xu, French & Ettner (2014) investigated the association between Axis II disorders and two measures of high cost health care utilization namely: emergency department episodes and hospital admissions. They state that BPD, of all

(4)

men and women. In these modern times healthcare is being confronted with substantial budget constraints.

Personality was long thought to be a profile of stable characteristics and that it was impossible to change them. Therefore there has been considerable skepsis about the treatment of disorders in personality in the past (Eurelings-Bontekoe et al. 2009). However, in the last decade several randomized controlled trials have shown psychological treatments to be effective in the treatment of BPD. The controlled trials provide support for the effectiveness of various forms of specialist psychotherapeutic treatment, such as Dialectical Behavior Therapy (DBT; e.g. Linehan et al., 2006), Schema-Focused Therapy (SFT; e.g. Giesen-Bloo et al., 2006), Transference-Focused Psychotherapy (TFP; e.g. Clarkin, Levy, Lenzenweger, & Kernberg, 2007; Levy, Meehan, Kelly, Reynoso, Weber, & Clarkin, 2006), Systems Training for Emotional Predictability and Problem Solving (STEPPS; e.g. Blum et al., 2008),

Cognitive Behavior Therapy (CBT; e.g. Davidson et al., 2006), and Mentalization Based Treatment (MBT; e.g. Bateman & Fonagy, 1999). These trials show remarkable effect sizes for many different outcome measures varying from reducing the number of SCID-II BPD criteria to increasing psychosocial functioning.

Untill now, there is no consensus about whether one therapy is better than the other. Stoffers et al. (2013) conducted a meta-analysis to assess the effects of psychological

interventions for BPD. They analyzed randomized studies with samples of patients with BPD comparing a specific psychotherapeutic intervention against a control intervention. DBT, MBT, TFP, SFT and STEPPS were the most intensely studied psychotherapies in this analysis. Their conclusion was that there are indications of beneficial effects for psychotherapeutic interventions for BPD core pathology and associated general psychopathology. Therefore they say that disorder specific treatments should be used in treating BPD. Stoffers et al. (2013) concluded that DBT provides the most robust evidence of efficacy relative to all other

(5)

treatments that have been investigated in RCT's so far. MBT was second to DBT in evidence of its efficacy. They show large statistically significant effects for the reduction of suicidality, damaging behavior, interpersonal problems and depression. However, Stoffers et al. (2013) did not compare specific therapies to examine if one of the interventions showed incremental efficacy. This study actually tried to do just this in order to get a closer view of which specific psychotherapy is the best treatment for BPD.

In the present meta-analysis, we aimed to compare the efficacy of two prolonged outpatient treatments: MBT and SFT. MBT (Bateman, 2004) is a psychoanalytically-based psychological intervention that helps people to increase their capacity to mentalize. Some people might say that mentalizing is more or less a new term for a relatively old concept. It means that someone learns to recognize and understand the feelings and thoughts they evoke in others and the feelings and thoughts they experience themselves (Bateman & Fonagy, 2007). The treatment consists of individual as well as intensive group- psychotherapy and lasts 18 months. Every treatment has a follow up of 18 months were the group-psychotherapy is gradually reduced. In the therapy sessions patients discuss situations where mentalizing played an important role.

SFT (Young, 2003) was designed to extend Beck's original model of cognitive therapy in a way that it specifically aims at personality disorders. It helps people with BPD to identify early maladaptive schemas. These schemas are broad pervasive themes regarding oneself and others, developed during childhood and elaborated throughout one's lifetime, which are dysfunctional to a significant degree. The assumption is that these themes are the result of unmet emotional needs in childhood and present themselves as maladaptive coping styles in adulthood. SFT tries to aid BPD patients in getting their needs met (McGinn & Young, 1996). According to the SFT model, schemas are at the core of personality disorders, whereas the behavioral patterns listed in DSM-IV are primarily coping responses to the schemas. Treating

(6)

the disorder, thus the behavioral responses, or coping styles, is impossible without changing the schemas that are driving them. Problems in relationships, the self and emotion-regulation are seen as the primary objectives to work on according to SFT. At first glance the therapy looks quite suited to treat BPD because SFT aims to improve these problems specifically. Research also suggests that SFT is indeed effective. Giesen Bloo et al. (2006) saw that patients improved in borderline personality pathology. Furthermore, patients made remarkable progress on psychopathological functioning and on overall quality of life. An important difference between MBT and SFT is that MBT is especially developed for BPD.

Bateman et al. (2009) and Giesen-Bloo (2006) provided empirical evidence for the efficacy of MBT and SFT. What we don't know is how these two treatments relate to one another. In other words: are they both equally effective in the treatment of BPD? The main aim of this meta-analysis was therefore to compare the efficacy of MBT and SFT for BPD. It was hypothesized that there would be no differences between the aggregate within-group effect-sizes for these psychological interventions.

We know very little about which people with borderline personality disorder may benefit most from which specialist treatment. Currently there are no reliable evidence based indicators to guide treatment planning. It seems that part of the patients diagnosed with BPD will benefit more from a specialist treatment than from generic interventions such as

structured clinical management, but that for some structured clinical management will work just as well (Bateman et al., 2013). This is important because there is a limited access to psychological treatment in most countries, especially for the type of lengthy and intensive treatment programs that are recommended for people with BPD. Furthermore, because cost effectiveness in mental health care is becoming increasingly important, there has been

(7)

that some patients might equally benefit from less expensive alternatives such as structured clinical management.

METHOD

Inclusion and Exclusion Criteria

To be included in the meta-analysis, the studies had to meet the following criteria. First, the studies had to include a trial of MBT or SFT for BPD. The trials could be randomized controlled trials, open trials and case series. Second, participants needed to be adult patients (age ≥ 18) with primary diagnosis of borderline personality disorder according to DMS-III-R or DMS-IV (-Tr) criteria.

Studies with the following characteristics were excluded from this meta-analysis. First, “Double diagnoses” studies, that is studies that focus exclusively on a specific combination of two diagnoses, e.g. BPD and eating disorder; BPD and opioid dependence. The reason for this decision is the chance of biased sampling from the BPD population. Furthermore the

treatments may be modified to the double diagnosis. Second, Mixed PD – samples were excluded, unless separate statistics on the BPD subsample were given. This means that in the sample at least 90% of the participants had to be diagnosed with BPD, if no separate statistics for BPD were provided.

Types of interventions

Experimental interventions were MBT and SFT. Comparator interventions were grouped as follows. Firstly, unspecific control interventions, including structured clinical management (SCM), standard psychiatric care (SPC), supportive psychotherapy (SP) and treatment as usual (TAU) (these are interventions without any specific mode of action). Secondly, the comparative specific psychotherapeutic interventions were SFT and MBT.

(8)

We clustered the different outcome measures in two categories: primary- and secondary outcomes. Primary outcomes were: overall BPD severity and severity of single BPD criteria according to DSM. Depression, anxiety, general psychopathology and mental health status were defined as secondary outcomes.

Search methods

Systematic bibliographic research was undertaken to identify relevant studies from online databases (Pubmed, Embase, Cochrane, Psycinfo) using the following keywords: Borderline Personality Disorder or borderline personality and Clinical Trial or Controlled Trial or Randomized Controlled Trial. Additional articles were found through reference in reviews and empirical studies. For references to ongoing studies on BPS, the researchers involved were contacted.

Statistical procedures

First we computed within effect sizes for MBT and SFT. Second we compared effect sizes between MBT and SFT. Third we compared these specific psychotherapeutic interventions to the unspecific control interventions, including clinical management, standard care, treatment as usual or waiting list. In these comparisons the independent variable was type of treatment divided in the two categories: MBT and SFT. The dependent variables were the within-group effect sizes on the primary and secondary outcome measures. For all treatment and control conditions of the included studies, within group effect sizes (Hedges’ g) were computed. The magnitude of change from pre- to post treatment was computed for each trial. Next, if

applicable, effect sizes on multiple measures for the same outcome in a study were pooled. The pooled effect sizes for each outcome were then aggregated across treatment study conditions of, MBT and SFT, respectively, resulting in a mean weighted effect size and 95% confidence interval for each outcome and each treatment type. Finally, the 95% confidence intervals for each treatment type were compared with intervals that did not overlap, indicating

(9)

significant differences between treatment types. Comprehensive Meta-Analysis software was used for most analyses.

RESULTS

Study characteristics

The study by Bateman et al. (1999) evaluated a psychoanalytically oriented partial

hospitalization. This therapy is more or less the predecessor of MBT and consisted of 1) once-weekly individual psychoanalytic psychotherapy, 2) thrice-once-weekly group analytic

psychotherapy (1 hour each), 3) once-a-week expressive therapy oriented toward

psychodrama techniques (1 hour), and 4) a weekly community meeting (1 hour), all spread over 5 days. The average length of stay was 1.45 years.

The two studies Bateman et al. (2009) and Bales et al. (2012) evaluated MBT. The therapy consisted of 18 months of weekly combined outpatient individual and group psychotherapy provided by two different therapists. The day hospital program, covering 5 days per week and four and an half hours per day , included implicit mentalizing groups (i.e., daily group psychotherapy and weekly individual psychotherapy, and individual crisis planning from a mentalizing perspective) and explicit mentalizing groups (i.e., art therapy twice a week, mentalizing cognitive group therapy and writing therapy).

The study by Jørgensen et al. (2009) evaluated combined MBT. It consisted of 2 years of intensive 2 weekly combined (individual and group) MBT, and was in accordance with the MBT treatment manual by Bateman et al. (2004).

The study by Giesen-Bloo et al. (2006) evaluated SFT. The therapy consisted of 50 minutes sessions, twice a week for three years. Treatment protocols addressed the theoretical model, treatment frame, different phases, and use of strategies and techniques described in Young (1994), Young et al. (2003), Clarkin et al. (1999) and Yeomans et al. (2002).

(10)

The study by Farrel et al. (2009) evaluated a group-SFT program that consisted of thirty weekly sessions, each lasting 90 minutes, over an 8-month period. The treatment was manual-based (Farrel & Shaw, 1990) and combined 4 content components that they viewed as central to psychotherapy for patients with BPD: emotional awareness training (described in Farrel & Shaw, 1994), BPD psychoeducation, distress management training (Farrel, et al., 2005) and schema change work (Young, 1990; Young et al., 2003).

The single case series study by Nordahl & Nysaeter (2005) also evaluated SFT. Patients were treated for at least 18 months, to a maximum of 36 months, so the patients did not have a fixed number of sessions. The treatment followed the protocol outlined by Young (1996). The treatment received treatment on a weekly basis and sessions were each of 60 minutes duration for a mean period of 22 months (18-36 months range). The main elements of the therapy were 1) to develop a schema mode formulation of the patients in order to share an understanding of the patient’s modes, distress and interpersonal difficulties, 2) to bond with the patient through re-parenting (soothing, support, guidance) and helping the patients with their emotional deprivation, 3) work on schema modes and interpersonal coping skills, 4) managing crisis and enhancing problem solving, and 5) gradual termination and fading of therapy.

The study by Nadort et al. (2009) evaluated outpatient SFT with crisis support by phone. There were sessions twice a week in the first year but sessions once a week in the second year. Patients were treated for 18 months. Treatment protocol followed the study of Giesen-Bloo et al. (2006).

(11)

Comparison conditions

In the MBT studies, three different comparison conditions were used: structured clinical management (SCM) (Bateman & Fonagy, 2009), standard psychiatric care (SPC) (Bateman & Fonagy, 1999) and supportive psychotherapy (SP) (Jørgensen et al., 2013). These are active treatment control conditions and they all are based on a counseling model closest to a supportive approach with case management, advocacy support, and problem-oriented

psychotherapeutic interventions. The most important difference is that the SPC control group received no formal psychotherapy.

In the SFT studies the comparison conditions were treatment as usual (TAU) (Farrel et al., 2009), and transference focused therapy (TFP) (Giesen-Bloo et al., 2006). In one study the comparison condition was the same therapeutic model without extra phone support (Nadort et al., 2009). The TAU condition received only continuing weekly individual psychotherapy as usual in the community. TFP was, like SFT, offered on a 50 minute session, twice a week basis.

All controlled studies were randomized controlled studies.

Study participants

All studies investigated patients with borderline personality disorder. Participants were diagnosed as having BPD according to DSM-III-R (Bateman & Fonagy, 1999; Farrel et al., 2009) or DSM-IV (Nordahl & Nysaeter, 2005; Giesen-Bloo et al., 2006; Bateman & Fonagy, 2009; Nadort et al., 2009; Bales et al., 2012). The studies included a total of 488 participants of which 420 were female (86%) and 68 were male (14%).

Test of heterogeneity

A test for heterogeneity examined the consistency of effects across the studies that were included in the analysis. Table 1 shows there was a low consistency of effects across studies in the different subgroups.

(12)

Subgroup Outcome/analysis Q-value p-value I2_-value* MBT Primary 1.591 0.451 0.000 Secondary 9.228 0.026 67.489 SFT Primary 12.486 0.014 67.963 Secondary 7.158 0.128 44.122 Control Primary 7.190 0.027 72.182 Secondary 14.204 0.003 78.880

Table 1. Heterogeneity statistics for the three subgroups. * Values of I2_{are percentages.}

Outcomes

Effect sizes at posttreatment: Primary outcome measures

Table 2 displays each study’s Hedges’g effect sizes (ES) and 95% confidence interval (CI) for MBT, SFT and controls on primary outcome measures. Overall, all of the treatments resulted in significant lower posttest BPD severity and severity of single BPD criteria according to DSM. Where SFT (aggregate Hedges’g = 1.490, 95% CI = 1.007 - 1.972, p<0.001) and MBT (aggregate Hedges’g = 1.325, 95% CI = 1.120 – 1.530, p<0.001) seem to have greater ES’s, they have comparable efficacy to controls because of the overlap in lower and upper limit (aggregate Hedges’g = 0.765, 95% CI = 0.286 – 1.244, p<0.005).

Study name Outcome Time point Hedges’g Standard error Variance Lower limit Upper limit Z-value p-value

Bales (2012) Combined Post 1.158 0.192 0.037 0.783 1.534 6.047 0.000

Bateman (2009) IIP Post 1.479 0.171 0.029 1.144 1.814 8.658 0.000

Jørgensen (2013) Combined Post 1.300 0.183 0.033 0.942 1.657 7.121 0.000

Aggregated MBT 1.325 0.105 0.011 1.120 1.530 2.674 0.000

Giesen-Bloo* (2006)

BPDSI Post 1.840 0.246 0.060 1.358 2.322 7.486 0.000

Farrel (2009) Combined Post 1.899 0.412 0.170 1.091 2.706 4.608 0.000

Nordahl (2005) Combined Post 2.222 0.758 0.575 0.736 3.707 2.931 0.003

Nadort (2009)** Combined Post 0.860 0.212 0.045 0.445 1.274 4.065 0.000

(13)

Aggregated SFT 1.490 0.246 0.061 1.007 1.972 6.052 0.000

Bateman (2009) IIP Post 0.749 0.141 0.020 0.472 1.025 5.301 0.000

Farrel (2009) Combined Post 0.258 0.276 0.076 -0.283 0.798 0.935 0.350

Aggregated Controls

0.765 0.244 0.060 0.286 1.244 3.128 0.002

Table 2. Hedges’g ES’s for posttest differences in primary outcomes for MBT, SFT and controls. * Hedges’g is based on median instead of means. SD’s are calculated with SE * √N

** SFT with extra phone support. SD’s are calculated with SE * √N *** SFT without extra phone support. SD’s are calculated with SE * √N

Effect sizes at posttreatment: Secondary outcome measures

Table 3 displays each study’s Hedges’g effect sizes (ES) and 95% confidence interval (CI) for MBT, SFT and controls on secondary outcome measures. Overall, all of the treatments resulted in significant lower posttest general pathology and general functioning. Where MBT (aggregate Hedges’g = 1.178, 95% CI = 0.838 - 1.517, p<0.001) seems to have greater efficacy, it is comparable to SFT (aggregate Hedges’g = 0.727, 95% CI = 0.447 – 1.007, p<0.001) and controls (aggregate Hedges’g = 0.571, 95% CI = 0.117 – 1.024, p<0.05) because of the overlap in lower and upper limit.

Study name Outcome Time point Hedges’g Standard error Variance Lower limit Upper limit Z-value p-value

Bateman (1999) Combined Post 1.303 0.317 0.101 0.681 1.926 4.106 0.000

Bales (2012) Combined Post 0.984 0.182 0.033 0.627 1.342 5.396 0.000

Aggregated MBT 1.178 0.173 0.030 0.838 1.517 6.804 0.000

Giesen-Bloo (2006)*

Combined Post 0.852 0.177 0.031 0.604 1.199 4.804 0.000

Farrel (2009) Combined Post 1.021 0.300 0.090 0.432 1.610 3.399 0.001

Nordahl (2005) Combined Post 1.701 0.620 0.384 0.486 2.916 2.744 0.006

Nadort (2009)** Combined Post 0.446 0.185 0.034 0.084 0.808 2.414 0.016

Nadort (2009)*** Combined Post 0.527 0.194 0.038 0.147 0.907 2.720 0.007

(14)

Bateman (1999) Combined Post 0.136 0.221 0.049 -0.297 0.570 0.617 0.537

Farrel (2009) Combined Post 0.173 0.271 0.073 -0.358 0.704 0.639 0.523

Aggregated Controls

0.571 0.231 0.054 0.117 1.024 2.468 0.014

Table 3. Hedges’g ES’s for posttest differences in primary outcomes for MBT, SFT and controls. * Hedges’g is based on median instead of means. SD’s are calculated with SE * !N

** SFT with extra phone support. SD’s are calculated with SE * !N *** SFT without extra phone support. SD’s are calculated with SE * !N

DISCUSSION

The purpose of this meta-analysis was to evaluate the efficacy of MBT and SFT in order to compare them and to investigate if one of the two is more suited for treating a specific group of patients with BPD. Both MBT and SFT resulted in significant and substantial efficacy on primary and secondary outcome measures. In all included studies MBT and SFT yielded positive Effect sizes for BPD symptoms and general pathology and general functioning. It can be said that the efficacy was remarkably robust across studies.

Our findings regarding the efficacy of MBT and SFT can be explained in various ways. First, in spite of the fact that the different therapies have a very distinct theoretical

background on first sight, they may have more in common than meets the eye. For example, MBT is described as a psychoanalytically and attachment theory based psychotherapy. It is said that it aims to increase the reflective or mentalizing capacity of the participant, helping them to understand and recognize the feelings they invoke in others and the feelings they experience themselves as a result of others. People may benefit from this therapy because they have been programmed the wrong way if it comes to processing social and emotional information. MBT uncovers these cognitive processes and tries to adjust them in a way that they are no longer distorted and therefore no longer form a problem. By training certain skills, MBT tries to equip patients with alternative ways of reacting in varied difficult

(15)

situations. This enhances their ability of introspection and builds resilience. SFT searches for schemas that are at the root of coping mechanisms that in turn are the cause of various BPD symptoms. SFT also provides insight in cognitive processes and trains certain skills to adequately cope with difficult situations. In its core it looks like MBT as well as SFT have close resemblances with cognitive therapy. Maybe this is why their outcomes are so much alike. Second, it is also possible that the psychotherapies are equally effective because there are common factors (i.e. consistent use of a coherent model, treatment-integrity, structure, competency of the therapist, attention for motivational and relational aspects) inherent within most specific psychotherapies that are essential to producing positive treatment outcomes (Budge et al. 2013). This phenomenon is also known as the Dodo Bird effect (Rosenzweig, 1936). The resemblences between MBT and SFT are manifold. Firstly, both MBT and SFT are highly structured, manualized therapies. Secondly, in both treatments there is much attention for motivational aspects and therapy compliance. Thirdly, they both place great importance to therapeutic alliance and an active approach of the therapist. Because in essence MBT and SFT seem to be so much alike, it may be a difficult task to detect the distinct

features of each psychotherapeutic intervention that are accountable for the sought after effect. Are these treatments effective because they have so much in common (the specific elements are so much alike) or is it because of the common factors? (the treatments are effective because of the non-specific elements). What stands out in this meta-analysis is that the

controls are doing such a good job treating BPD. It seems that the control groups that are used in this study (i.e. structured clinical management, structured psychiatric care, supportive psychotherapy and treatment as usual) have effective elements in them that make them just as efficacious as specific psychotherapy. This also may be explained by the cognitive elements and common factors.

(16)

A few side notes have to be taken into account concerning the results of this meta-analysis. First, only 8 studies met our eligibility criteria. This number was sufficient to establish aggregate ES’s for both specific treatment modalities and controls, but precluded firm conclusions about the efficacy of MBT compared to SFT and controls. The fact that ES’s do not differ significantly between MBT, SFT and controls may also be a consequence of the number of studies we included.

Second, in order to compare interventions it is crucial to find a common denominator i.e. outcome measure to compare them with. Unfortunately this is a hard to come by given the recent available trials. It seems that because of the phenomenological diversity of BPD, there is also a high variety of possible outcomes. Ostensibly there is only small consensus about which outcomes are the most important ones. Stoffers et al. (2013) assert that core symptoms that patients and caregivers may be interested in when looking for a treatment option that helps, are often neglected. Outcomes such as decreasing feelings of emptiness and avoidance of abandonment are considered as significant traits in people with BPD. Unfortunately these measure outcomes are rarely considered by trialists. Statements following the conclusions about primary and secondary outcomes in this study therefore should be interpreted with caution.

Third, this meta-analysis used the data of the completers of treatment only. Patients with BPD are known for frequently dropping out prematurely from psychotherapy. This treatment attrition is associated with poorer treatment outcomes (Wnuk et al., 2013). Because the results in this research did not consider drop-outs, this might have positively biased the overall effect sizes of MBT, SFT and Controls.

Fourth there are two issues to be considered regarding the effect-sizes of the control studies. Firstly, the substantial efficacy of controls on primary and secondary outcomes is remarkable. There is much discussion about what a good control group should look like.

(17)

Consequently there is a lot of variation in control groups used in randomized controlled trials. The one mostly used is treatment as usual (TAU). TAU has a lot of variations and they consist of differences in content as well as in implementation (Budge et al. 2013). Some regard TAU as a psychotherapeutic treatment that includes integrative, non-manualized treatments

provided by masters and doctoral level clinicians. Others see TAU as a non treatment group. Research shows that content and implementation greatly influence effect sizes. As a result it is important to zoom in to the treatment used in the control group to adequately interpret effect sizes. Because in this meta-analysis the control groups were mostly comprised by active treatments, this could explain the considerable aggregated effect size for control treatments. The second issue concerning the effect sizes of the control interventions is the conclusion that there was not found a significant difference between specific therapy and control

interventions. An explanation for this finding could be the relative small number of studies included in this meta-analysis. More studies most likely would result in more differentiation between the specific therapies and control conditions.

Fifth, the test for heterogeneity shows a low consistency of effects across studies in the different subgroups. This raises the question which factors could possibly explain this variation in effect sizes. Unfortunately the number of studies in the present meta-analysis was to small to investigate potential moderators that could be responsible for this variation.

Nevertheless there is reason to believe there are factors that could be of influence. Bateman & Fonagy (2013) hypothesized that for instance clinical severity may play an important role in the decision if specialist treatment is indicated or not. They concluded that there is a chance that MBT might have a better effect than structured clinical management for people with more severe BPD. Clinical severity is not an easy concept to grasp. In past research many different definitions are used. After an exhaustive review of the literature, Crawford et al. (2011) defined four main indicators of increasing severity of BPD: level of symptom distress,

(18)

number of descriptive criteria met for the disorder, extent of comorbidity with Axis I disorders and comorbidity with other personality disorders, especially in different disorder clusters. Bateman and Fonagy (2013) used these indicators of severity to see if they could predict the outcome at the end of treatment. Unfortunately none of them could. Surprisingly though patterns of recovery in the treatment groups MBT and structured clinical management were significantly associated with both the baseline level of distress and the number of Axis II diagnoses. They found out that only 25% of patients with three or more Axis II diagnoses recovered in the group that was treated with structured clinical management (SCM). In the group that was treated with MBT no less than 75% of the patients recovered. As for that they discovered that the rate of recovery diminished in de SCM group for every additional Axis II diagnosis, whereas in the MBT group recovery remained stable. The impact of Axis II comorbidity was especially apparent in relation to self-harm. MBT was significantly more effective than SCM in reducing self-harm for patients with a higher number of Axis II

diagnoses. With this Bateman and Fonagy (2013) found an index for the severity of BPD that could guide treatment planning. For future research the interaction between BPD severity and type of treatment should be investigated. It could be that, in severe BPD cases, MBT could yield larger aggregate within-group effect sizes than SFT.

Despite these limitations, our findings underscore that MBT and SFT are efficacious treatments for BPD. However, this meta-analysis cannot formulate an answer to the question what works best for whom. In order to do this the field requires indentification and testing of predictors of outcome. This study suggests that it may be useful to investigate clinical severity as a possible moderator of treatment outcome.

(19)

REFERENCES

van Asselt, A. D. I., Dirksen,C. D., Arntz, A., Giesen-Bloo, J. H., van Dyck, R., Spinhoven, P, van Tilburg, W., Kremers, I. P., Dadort, M., & Severens J. L. (2008). Out-patient

psychotherapy for borderline personality disorder: Cost effectiveness of schema-focused therapy v. transference-schema-focused psychotherapy. The British Journal of

Psychiatry, 192, 450-457.

Bales, D., van Beek, N., Smits, M., Willemsen, S., Busschbach, J. J. V., Verheul, R., et al. (2012). Treatment outcome of 18-month day hospital mentalization-based treatment (MBT) in patients with severe borderline personality disorder in The Netherlands.

Journal of Personality Disorders, 26, 568-582.

Bateman, A., & Fonagy, P. (1999). Effectiveness of partial hospitalization in the treatment of borderline personality disorder: A randomized controlled trial. American Journal of

Psychiatry, 15, 1563-1569.

Bateman, A., & Fonagy, P. (2004). Psychotherapy for borderline personality disorder: Mentalization-based treatment. Oxford: Oxford University Press.

Bateman, A., & Fonagy, P. (2009). Randomized controlled trial of outpatient mentalization-based treatment versus structured clinical management for borderline personality disorder. American Journal of Psychiatry, 166(12), 1355-1364.

Bateman, A., & Fonagy, P. (2013). Impact of clinical severity on outcomes of mentalisation-based treatment for borderline personality disorder. The British Journal of Psychiatry,

203, 221-227.

Blum, N., John, D. ST., Pfohl, B., Stuart, S., Mc-Cormick, B., Allen, J. et al. (2008). Systems training for emotional predictability and problem solving (STEPPS) for outpatients with borderline personality disorder: A randomized controlled trial and 1 year follow-up.

(20)

Budge, S. L., Moore, J. T., Del Re, A. C., Wampold, B. E., Baardseth, T. P., & Nienhuis, J. B. (2013). The effectiveness of evidence-based treatments for personality disorders when comparing treatment-as-usual and bona fide treatments. Clinical Psychology Review, 33, 1057-1066.

Clarkin, J. F., Levy, K. N., Lenzenweger, M. F., & Kernberg, O. F. (2007). Evaluating three treatments for borderline personality disorder: A multiwave study. American Journal of

Psychiatry, 146, 922-928.

Clarkin, J. F., Yeomans, F. E., & Kernberg, O. F. (1999). Psychotherapy for borderline personality. New York: John Wiley and Sons.

Crawford, M. J., Koldobsky, N., Mulder, R, & Tyrer, P. (2011). Classifying personality disorder according to severity. Journal of Personality Disorders, 25(3), 321-330. Davidson, K., Norrie, J., Tyrer, P., Gumley, A., Tata, P., Murray, H., & Palmer, S. (2006).

The effectiveness of cognitive behavior therapy for borderline personality disorder : Results from the borderline personality study of cognitive therapy (BOSCOT) trial.

Journal of Personality Disorders, 20, 450-465.

Eurelings-Bontekoe, E. H. M., Verheul, R, & Snellen, W. M. (2009). Handboek persoonlijkheidspathologie. Houten: Bohn Stafleu van Loghum.

Farrel, J. M., & Shaw, I. A. (1990). Emotional awareness training for borderline personality disorder patients: A treatment manual. Indiana University School of Medicine:

Indianapolis.

Farrel, J. M., & Shaw, I. A. (1994). Emotional awareness training: a perequisite to effective cognitive-behavioral treatment of borderline personality disorder. Cognitive and

(21)

Farrel, J. M., Shaw, I. A., Foreman, T., & Fuller, K. E. (2005). Psychotherapy in the eye of the storm: A treatment program for borderline personality disorder. Symposium

presented at American Psychological Association Annual Convention, Washington D. C. Farrel, J. M., Shaw, I. A., & Webber, M. A. (2009). A schema-focused approach to group

psychotherapy for outpatients with borderline personality disorder: A randomized

controlled trial. Journal of Behavior Therapy and Experimental Psychiatry, 40, 317-328. Franken, I., Muris, P., & Denys, D. (2013). Basisboek Psychopathologie. De Tijdstroom

uitgeverij: Utrecht.

Giesen-Bloo, J., Van Dyk, R., Spinhoven, P., Van Tilburg, W., Dirksen, C., Van Asselt, T. et al. (2006). Outpatient psychotherapy for borderline personality disorder: randomized trial of schema-focused therapy vs transference-focused psychotherapy. Archives of

General Psychiatry, 63, 649-658.

Jørgensen, C. R., Freund, C., Møye, R., Jordet, H., Andersen, D., & Kjøbye, M. (2013). Outcome of mentalization-based and supportive psychotherapy in patients with borderline personality disorder: a randomized trial. Acta Psychiatrica Scandinavica,

127(4), 305-317.

Laurenssen, E. M. P., Westra, D., Kikkert, M. J., Noom, M. J., Eeren, H. V., van

Broekhuyzen, A. J., Peen, J., Luyten, P., Busschbach, J. J. V., & Dekker, J. J. M. (2014). Day hospital mentalization-based treatment (MBT-DH) versus treatment as usual in the treatment of severe borderline personality disorder: Protocol of a randomized controlled trial. Biomed Central, 14(45).

Leichsenring, F., Leibing, E., Kruse, J., New, A. S., & Leweke, F. (2011). Borderline personality disorder. Lancet, 377(9759), 74-84.

Levy, K. N., Meehan, K. B., Kelly, K. M., Reynoso, J. S., Weber, M., Clarkin, J. F. (2006). Change in attachment patterns and reflective function in a randomized control trial of

(22)

transference focused psychotherapy for borderline personality disorder. Journal

Consulting and Clinical Psychology, 74, 1027-1040.

Linehan, M. M., Comtois, K. A., Murray, A. M., Brown, M. Z., Gallop, R. J., Heard, H. L. et al. (2006). Two-year randomized controlled trial and follow-up of dialectical behavior therapy vs therapy by experts for suicidal behaviors and borderline personality disorder.

Archives of General Psychiatry, 63, 757-766.

Maclean, J. C., XU, H. Y., French, M. T., & Ettner, S. L. (2014). Mental Health and high-cost health care utilization: New evidence from Axis II disorders. Health Services Research,

49(2), 683-704.

McGinn, L. K., & Young, J. E. (1996). Schema-focused therapy. In P. M. Salkovskis (Ed.), Frontiers of cognitive therapy (pp. 182 - 201). New York: Guilford Press.

Nordahl, H. M., & Nysaeter, T. E. (2005). Schema therapy for patients with borderline personality disorder: a single case series. Journal of Behavior Therapy and

Experimental Psychiatry, 36, 254-264.

Rosenzweig, S. (1936). Some implicit common factors in diverse methods in psychotherapy.

The American Journal of Orthopsychiatry, 6, 412-415.

Soeteman, D. I., Hakkaart-van Roijen, L., Verheul, R., & Busschbach, J. J. V. (2008). The economic burden of personality disorders in mental health care. Journal of Clinical

Psychiatry, 69(2), 259-265.

Stoffers, J. M., Völlm, B. A., Rücker, G., Timmer, A., Huband, M., & Lieb, K. (2013). Psychological therapies for people with borderline personality disorder. The Cochrane

Database of Systematic Reviews, 8, 1-255.

Wnuk, S., McMain, S., Links, P. S., Habinski, L., Murray, J., & Guimond, T. (2013). Factors related to drop out from treatment in two outpatient treatments for borderline

(23)

Yeomans, F. E., Clarkin, J. F., & Kernberg, O. F. (2002). A primer for transference focused psychotherapy for the borderline personality patient. Northvale: Jason Aronson Inc. Young, J. E. (1994). Cognitive therapy for personality disorders: A schema-focused approach.

Rev ed. Sarasota. Fla: Professional Resource Press.

Young, J. E. (1996). Outline for schema-focused cognitive therapy for borderline personality disorder. New York: Cognitive Therapy Center New York.

Young, J. E., Klosko, J. S., & Weishaar, M. E. (2003). Schema Therapy: a practitioner's guide. New York (NY): Guilford.