Group schema therapy v Group + Individual schema therapy v Treatment as Usual for Borderline Personality Disorder

Arnoud Arntz led a team of 14 authors on a 15-site 5-country investigation of whether group schema therapy (GST) is more effective than treatment as usual (TAU) for borderline personality disorder (BPD). 495 participants were randomly assigned to: 1) “Predominantly GST” (PGST); 2) group + individual schema therapy (IGST); or 3) TAU. All groups were treated over 2 years and followed up at 12 months (year 3). Previously, Farrell, Shaw and Webber (2009) achieved impressive results from their study of GST for BPD, but it was a small study with participants who had already committed to remaining in treatment for 6 months, provided by the developers of the protocol. The purpose of Arntz and colleagues’ (2022) multi-site trial was to see if GST remained efficient and effective in the hands of other therapists and settings. There was also some thought among schema therapists that GST might not be as effective as individual ST, because clients might not obtain the same degree of personal need fulfilment. Although Arntz et al (2022) was not a strict test of the possible superiority of individual ST, it could indicate whether individual therapy added demonstrable value to GST.
 
PGST allowed up to 12 individual sessions in the first year and up to 5 in the second year at client request, but otherwise consisted of 2 group sessions (from 2 therapists) per week in the first year, once per week from 13 to 18 months, once per 2 weeks from 19 to 21 months and monthly from 22 to 24 months. IGST consisted of 1 group and 1 individual session per week in the first year, individual and group sessions on alternate weeks from 13 to 18 months, group sessions on alternate weeks with monthly individual sessions from 19 to 21 months, and monthly group and individual sessions from 22 to 24 months. TAU frequency was matched to ST and reflected the best available BPD treatment at the site. DBT was most commonly used (36.6% of sites); others included CBT (11.4%), supportive (10.2%) and psychodynamic (9.6%).
 
Regarding the comparison of ST (combined IGST and PGST) v TAU, ST produced greater reductions in BPD severity, evident by the 1.5-year mark. When comparing types of ST, IGST produced greater reductions in BPD severity than PGST, evident by the 2.5-year mark. When comparing IGST with TAU, IGST was significantly more effective in reducing BPD severity by the 1-year mark. By contrast, PGST was not significantly different from TAU. Regarding retention, PGST did not improve on TAU (72%, 73% respectively at end of first year, 62%, 64% at end of second year). IGST retained a greater proportion of participants (82% at 12 months, 74% at 24 months). There were fewer suicide attempts over time and fewer in IGST than TAU (but not significantly different than PGST). IGST was superior to TAU on 9 of 11 secondary outcomes (e.g., schemas, modes, work and social functioning); PGST was superior to TAU for 2, but IGST was only superior to PGST for quality of life. Deterioration of BPD severity was more common in TAU (2.1%) than in PGST (1.25%) or IGST (0.88%).
 
TAU comparisons are interesting to pragmatic-minded clinicians and service managers because the results give a reasonable indication of how much improvement in outcomes might be gained if a service was to invest in training staff with fidelity in a specific approach. Scientifically, it’s a less conclusive design for testing efficacy because part of the superior results obtained by the experimental intervention is attributable to non-specific factors like superior training and supervision and therapist optimism and fidelity which are not usually present to the same extent in TAU conditions. In this study by Arntz and colleagues, TAU sessions were not video-recorded and rated due to practical barriers, whereas the ST conditions were. This doesn’t affect the conclusiveness of the modality comparison, however.
 
Both clients and therapists who participated in qualitative studies supported the importance of individual sessions in ST for addressing more sensitive topics, like child trauma, on a deeper level than possible in groups. Conversely, while most clients highlighted the importance of both group and individual sessions, some noted that hostile group members made some groups frightening which may have contributed to higher drop-out. The authors also emphasized that although clients dislike the tapered schedule, overall they do well, suggesting the scheduling in this study was close to optimal. There was some qualitative evidence to suggest what many clinicians preach: that clients gain more confidence by discovering they can apply what they learn in therapy by withdrawing the “safety blanket” of therapy once they have learnt what they need to.
 
Overall, the findings suggest good results for psychotherapy of BPD. All conditions achieved large reductions in BPD symptom severity. The results also support that within ST, individual sessions appear to offer specific advantages over group-only delivery. Optimal ST is likely to involve more than monthly individual sessions (at least in the first year). Good delivery of ST involves both significant interpersonal ability and technical skill, so the cost-effectiveness study accompanying this trial will be of vital interest to service providers.

Clinician implications

  • 2 years of weekly psychotherapy for BPD involving two sessions per week in the first year, weekly sessions from 13-18 months and graduated to monthly sessions by the 22nd month can produce large reductions in borderline symptom severity.

 

  • The combination of individual and group schema therapy produces greater and faster reductions in BPD severity; the advantage over treatment as usual is evident by 12 months, and over predominantly-group schema therapy is evident at 2.5 years (i.e., 6 months after end of treatment).

 

  • Individual + group therapy was superior to group schema therapy or treatment as usual in retaining clients in treatment

 

  • Associated qualitative studies suggest clients prefer individual sessions to group sessions for processing sensitive topics like child trauma and that hostile group members may contribute to attrition from group ST.

For the original article, go to:
https://jamanetwork.com/journals/jamapsychiatry/fullarticle/2789694

REFERENCES

Arntz A, Jacob GA, Lee CW, et al. Effectiveness of Predominantly Group Schema Therapy and Combined Individual and Group Schema Therapy for Borderline Personality Disorder: A Randomized Clinical Trial. JAMA Psychiatry. 2022;79(4):287–299. doi:10.1001/jamapsychiatry.2022.0010

Farrell, 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(2), 317–328. https://doi.org/10.1016/j.jbtep.2009.01.002

Matthew Smout