Traumatized refugees: morbidity, treatment and predictors of outcome

Despite large numbers of traumatized refugees, little is known about effective treatment of war trauma in refugees and immigrants. Few studies evaluating treatment have been published and most studies are follow-up studies with methodological limitations and little comparability across studies.

The purpose of the PhD is to characterize transcultural trauma patients in Denmark needing psychiatric treatment with regards to psychopathology and predictors of mental health and to evaluate the effects of the treatment.

Two studies reported in 4 papers form the basis of the thesis. FORLOB (Paper 1-3) was a follow-up study that included all patients receiving treatment at the Competence Center for Transcultural Psychiatry in Copenhagen from April 2008 to February 2010. Patients completed self-ratings of symptoms of PTSD, depression and anxiety as well as level of functioning and quality of life (HTQ, HSCL-25, SDS & WHO-5) before treatment and after treatment. Associations of co-morbid diagnoses and predictors of the patients' health condition were examined with linear and logistic regression and Pearson's correlation coefficients. Treatment in FORLOB consisted of a combination of Sertraline, Mianserin, psycho-education and Trauma-Focused Cognitive Behavioral Therapy (TFCBT). The treatment administered to each patient was monitored in detail and changes in outcome and predictors of change were analyzed. PTF1 (Paper 4) was a randomized controlled clinical trial with 2x2 factorial design (antidepressants, TFCBT, antidepressants & TFCBT, waiting list). Potential participants were screened amongst adult patients referred to the Competence Center for Transcultural Psychiatry in the period June 2009-2011. Patients with PTSD, war trauma and without a psychotic disorder were included. The manualized treatment consisted of weekly sessions with a physician and/or psychologist over a period of 6 months. The treatment effect was evaluated with a combination of self-ratings and blinded and non-blinded observer ratings. Outcome measures included symptoms of PTSD, depression, anxiety, pain and somatization, quality of life and level of functioning (HTQ, HSCL-25, SCL-90, WHO-5, SDS, VAS, Hamilton, GAF). Treatment was offered with translation and screening instruments were translated to the six most common languages in the patient group covering the needs of 92% of patients.

In FORLOB, patients had several co-morbidities and not just PTSD. Almost all patients had depression, pain and untreated somatic complaints in addition to PTSD. Furthermore, 36-58% had physical problems they were in treatment for, 9-16% of patients had psychotic symptoms mainly related to their trauma, 27% had enduring personality change due to catastrophic events according to ICD-10 and 46% reported traumatic brain injury. Patients reporting chronic pain had higher symptom scores on HSCL-25 and HTQ and patients with psychotic symptoms scored higher on all symptom clusters on HTQ. At pre-treatment assessment, the patients' level of functioning and quality of life were very low, the majority of patients lived on public subsidies, education levels were low and most patients had a limited social network. In FORLOB, we found a moderate significant change (Cohen's d 0.44-0.67) on all self-report outcome measures (HTQ, HSCL-25, SDS and WHO-5) after combination treatment. We found less improvement in PTSD when patients were receiving public subsidies and less improvement of depression when patients reported pain in the upper extremities. We found a positive association between systematic use of CBT methods and improvement in patient condition. In PTF1, the randomized clinical trial, we found a small, but significant effect of treatment with medicine on blinded observer-ratings of depression and anxiety (Ham-D and Ham-A) and a large effect on non-blinded ratings of level of functioning (GAF-F and GAF-S), in addition to a small effect on self-reported level of functioning and headache (SDS and VAS). Cohen's d calculated as the differences between randomization groups receiving medicine and not receiving medicine ranged from 0.91-1.01 on GAF-F and GAF-S, whereas on the other ratings showing significant change Cohen's d was 0.31-0.41. We did not find any effect of psychotherapy on any outcomes nor any effect of psycho-therapy or medicine on the primary outcome measure, PTSD.

Traditionally, treatment of traumatized refugees have focused on PTSD, but this study demonstrates that patients suffer from numerous psychiatric and somatic co-morbidities and the comprehensiveness of PTSD in explaining symptoms of traumatized refugees is questionable. This has implications for the type and implementation of treatment. PTF1 is the largest randomized clinical trial published on the treatment of traumatized refugees. It is a strength of PTF1 that it includes a waiting list control group thereby accounting for any effects due to spontaneous recovery and that treatment modalities are examined separately and in combination. In both FORLOB and PTF1, treatment adherence and patient compliance with treatment was thoroughly documented. Effect sizes were moderate in FORLOB and small in PTF1. There were discrepancies between the results in FORLOB and PTF1 with regards to the effect measured on self-ratings that can only partially be explained by methodological limitations of the follow-up study. Both studies are undertaken under pragmatic and realistic circumstances and the results are therefore relevant to other contexts. Patients are representative of patients in other North-European studies of traumatized refugees but differ from patients in trials published on culturally adapted CBT and Narrative Exposure Therapy.