Time-limited dynamic psychotherapy for psychiatric patients with personality disorders : a randomized controlled trial

Abstract

This thesis presents data collected from a randomized controlled trial (RCT) for psychiatric outpatients with at least one diagnosis of Personality Disorder (PD). There is a high prevalence of PD among patients from both primary care and psychiatry, and these patients are often considered difficult to treat. Further, there is little evidence suggesting that any psychological treatment is effective for PD. Within the context of a RCT, we examined the clinical utility of a manualized psychodynamic treatment possessing some empirical evidence for treating other psychiatric conditions. Out of 371 consecutive patients assessed for eligibility, 159 were non-PD, and 56 PD patients were unwilling to participate. The remaining 156 PD patients were randomized to either a manualized version of Supportive-Expressive Psychotherapy (SEP) (n=80) or to a control condition (n=76) that consisted of non-manualized community delivered psychodynamic treatment (CDPT). In study I, we explored the extent to which the treatments were able to a) reduce the prevalence of PD diagnoses and reduce their severity, b) decrease the number of PD features, c) reduce the overall levels of psychiatric severity, d) improve global level of functioning, and e) demonstrate cost-effectiveness in reducing health care consumption (as measured by number of visits after termination of treatment). Results demonstrated that patients improved significantly in all aspects, but that no significant difference between the two treatments was found. However, SEP was more cost effective than CDPT.

In study II, significant patient variables were tested for the purpose of exploring whether they either moderated or predicted the reduction of psychiatric symptoms (SCL-90). We found that intake level of PD criteria was associated with poorer outcome in both treatments and also that higher vindictiveness predicted a better outcome. Further, patients high in dominance showed greater improvements following CDPT than SEP. Study III explored to what extent quality of object relation and ego functions, psychological mindedness, interpersonal problems, and some big five personality traits (viz., neuroticism, agreeableness and extraversion), improved after treatment. We found significant improvements for most variables, but effect sizes and reliable change indices (RCI) showed that improvements were small to moderate, and few patients recovered to a healthy level. Study IV had two purposes: (a) identify baseline measures which can identify patients nonresponse to treatment at respective endpoint, (b) create a prediction index for identifying these severe nonresponders in the future. Fifty-four socio-demographic and clinical pre-treatment variables were used to identity those with extreme non-responder (ENR) status as defined by psychiatric symptoms and PD severity. Social withdrawal for CDPT and high levels of anxiety for SEP discriminated ENR in reducing symptomatic distress. Measures indicating detachment and social withdrawal discriminated ENR for both treatments in reducing PD severity. In addition, alexithymia in CDPT and not believing in one s social capacity for SEP further discriminated ENR. Thus, discrimination of ENR was found to be domain specific.

The main conclusions from the trial are that: 1) it is possible to treat PD condition in one year with reasonably good results, 2) SEP is slightly better at reducing level of personality disorderness, while CDPT is somewhat better at reducing psychiatric severity, 3) it is possible to define patients variables that predict and moderate outcome, 4) it is possible to predict extreme non-response to treatment with high level of accuracy, and 5) SEP seems to be slightly more cost effective than CDPT. Our outcome results are comparable with those from other psychotherapy trials for PD patients, supporting the validity of our results. The fact that the study was conducted in a real-world clinical psychiatric setting in which patients did not primarily ask for psychotherapy supports the ecological validity of our findings. The main effects of the trial do not indicate that manualized psychotherapy is more efficacious that non-manualized psychodynamic treatment except in terms of costeffectiveness. However, our results do not indicate that manualized therapy leads to worst outcome.

Some issues that were not explored or failed in this trial should be addressed in future research. Psychoanalytically based attachment measure that has empirical support for prediction of outcome can be tried, since the KAPP did not show predictive validity in this study. The clinical utility of a four dimensional psychological model for classifying severe PD conditions and plan treatment accordingly is worth exploring as an alternative to DSM. Finally longer treatment periods over a few years should be explored to increase the external validity of clinical trials.