This systematic review of psychological therapies for borderline personality disorder (BPD), conducted in Spain, takes an interesting approach to reviewing the literature. Unfortunately, there appear to have been challenges in translating the systematic review from Spanish and the text can be difficult to follow at times. This is a real shame as it is an interesting review. It is worth persevering and reading the full paper.

The authors make the point that BPD affects 1-2% of the general population and that this diagnosis is associated with early death, suicide and increased use of health care services. Unfortunately, up until relatively recently, a pessimistic view of treatment for BPD existed. More recently, there have been considerable advances and a number of randomised controlled trials (RCTs) of psychological therapy for BPD have been conducted. Pre-existing reviews and meta-analyses have suggested that some specific therapies have preliminary evidence that supports their use. Nevertheless, in many cases the therapies being reviewed were experimental, studies were small and used different methods to evaluate their results.

The authors note that previous systematic reviews and meta-analyses were hindered by heterogeneity among the studies reviewed.  Their systematic review did not use meta-analysis to synthesise the data and instead used a qualitative approach. The authors concentrated on what they considered to be three crucial aspects of RCTs of specific psychotherapies for BPD:

1. Pre-treatment selection of participants
2. Adherence to the therapy
3. Efficacy of the intervention

Methods

• Several databases were searched using appropriate search terms
• Previous reviews were consulted
• Appropriate inclusion and exclusion criteria were applied to screen out irrelevant studies
• Several variables were selected for investigation. These included:
  • Descriptive data of the studies
  • Pre-treatment selection
  • Treatment adherence
  • Principal outcome variables
• The efficacy of the therapy was evaluated in two ways
  • Variability in the number of participants who presented with at least one of what the authors called “adverse events”. These were (a) admission to hospital, (b) suicide or (c) self-injury
  • Variability in the capacity of the different therapies to significantly reduce one, two or three of the adverse events simultaneously

Results

Of 211 papers originally identified, 11 remained after the inclusion and exclusion criteria were applied. Of the remaining 11:

• 6 studies evaluated dialectical behaviour therapy (DBT)
• 1 study evaluated cognitive behaviour therapy (CBT)
• 2 studies evaluated mentalization based therapy (MBT)
• 1 study evaluated schema focused therapy (SFT)
• 3 studies evaluated transference focused psychotherapy (TFP)
• Some studies evaluated more than one therapy simultaneously
• Study quality was evaluated using recognised quality assessment tools. In general, study quality was considered to be adequate

There was great variability in the total percentage of participants who did not enter into treatment across the reviewed studies (17.6 – 63.6%). There was great variability among participants who refused treatment (4.4 – 49.4%). There was variability with respect to the exclusion criteria applied in the primary studies.

There was great variability in the percentages of participants who abandoned therapy across the reviewed studies. Interestingly, when a therapy developer is also an author of a study, abandonments also vary greatly. For example, in DBT studies published by the creator of the model (Marsha Linehan), abandonments of therapy ranged from 11.5% to 16.7%. In other evaluations of DBT it ranged from 37 to 47.4%. Treatment abandonment among participants assigned to control groups also varied enormously (from 11.4 – 77.4%).

Many of the reviewed studies did not report on the adverse events that the authors used to evaluate the efficacy of the different therapies. Bateman and Fonagy’s (2009) study of MBT performed well according to these criteria while McMain et al (2009) and Linehan et al (1991) also performed relatively well in their evaluations of DBT.

Discussion

• The authors point to their finding that there is great variation concerning participants who do not enter treatment. It is suggested that this can be related to two things: the patient may refuse to participate in the therapy or they may be deemed inappropriate according to the inclusion criteria. In both cases a selection bias may occur which may call into question the external validity of the results
• The authors also note their finding that there is variability depending on whether or not the developer of the therapy is also a co-author of the study. The authors suggest that this may be due to developers of new therapies being more enthusiastic and energetic in coping with adversity than non-developers. They may also be more competent than non-developers
• Not all of the studies reported the adverse events that the authors were interested in, which resulted in some difficulty in interpretation of the results
• MBT and DBT demonstrated the best results
• Suicides are more consistently reduced with DBT than with CBT or TFP
• For admissions to hospital, MBT performs better than DBT
• For admissions to hospital, MBT and DBT perform better than the other therapies
• The best results were obtained by MBT in the Bateman and Fonagy (2009) paper. Only 27% of participants were reported to have had an adverse event of any kind in the final stage of treatment
• The authors report that some participants may have become worse during therapies
• They call for deterioration rates to be published

Conclusions

• The authors suggest that 40% of people diagnosed with BPD who request specific therapy may not benefit from it
• Approximately 20% would not initiate treatment
• Of those who initiate treatment 25% may not respond to the therapy
• Future research should focus on investigating the characteristics of those who do not respond to therapy and providing new therapeutic strategies to help them to do so
• It is worth pointing out that the review authors in many cases were not able to evaluate the adverse events that they used to evaluate the psychotherapies
• There are other ways of evaluating therapy effectiveness that may have resulted in different conclusions being drawn

Links

Lana, F. & Fernandez-San Martin, M.I. (2013). To what extent are specific therapies for borderline personality disorders efficacious? A systematic review of published randomised controlled trials (PDF). Actas Espanolas de Psychiatrica. 41(4), 242-252