This is the first study to investigate the inclusion of exposure in trauma-focused CBT group treatment for traumatized youth with PTSD symptoms. Hypothesis 1 was confirmed as stress levels were significantly higher during exposure sessions, compared with non-exposure sessions. This is consistent with previous studies in adults which demonstrated the effectiveness of CBT group formats in terms of in-session exposure and disclosure (21). Comparisons of stress level ratings during exposure sessions yielded intriguing results and confirmed hypothesis 2. Large effect sizes for in-session stress level change (T1 to T2, and T2 to T3) were found for sessions 2 and 3. This is in line with the manual that suggests that exposure to traumatic events is most intense in these sessions due to their content. From a large body of research on refugees we know that most traumatic events happened in the home country or whilst on the move (22). This explains why many traumatic experiences were addressed in session 2. In session 3, the worst event is specifically addressed which might explain the increase in reported stress. In session 4, no U-shape trajectory of the stress level was detected, only a decrease in stress throughout the session. There are several potential explanations for this finding: Session 4 addresses life in the host country (Germany) compared with life in the home country. Consequently, participants may have predominantly focused on positive factors in the host country (such as provision of basic care through child welfare program or schooling) but discussed differences to their home country (which may have included aversive stimuli that elicited stress due to traumatic events in home country) in the group discussion which took place after rating (T2). Another possible explanation could be that the overall effect sizes for increase and decrease were small in this session. This would seem to imply that session content may not have resulted in any major exposure overall.
The results for individual in-session change (Fig. 2) were particularly encouraging for the field as this graph shows that almost all participants (independent of symptoms, traumatic experiences and so forth) were able to experience an increase and decrease in stress level, meaning that they seem to have successfully confronted their traumatic experiences.
It is important to note that the stress level significantly decreased in non-exposure sessions as well. In session 1 this could be explained by the high stress level at the beginning of the session indicating participants’ general nervousness and anxiety about starting a trauma-focused intervention. In session 6 general levels are very low indicating a floor effect. At the end of session 6 the group held a graduation party which might explain the very low scores.
Limitations and Future Research
Several limitations which might impede the generalizability of findings need to be addressed. Firstly, and most importantly, this study does not allow any conclusions to be drawn about the habitation process as group members were neither systematically asked to report their stress level during the revision of previous narrations at the beginning of the sessions, nor was this documented by the group facilitators. Thus, the mechanism of exposure could only be observed for the increase and decrease in stress during the specific content (in exposure sessions the narration) of each session. Future studies should seek not only to replicate and build on the present study but also to assess more systematically the stress level at more measurement time points during each exposure session (e.g. specifically during the revision of the narration of prior sessions at the beginning of every exposure session).
Secondly, this secondary analysis comprised a rather small sample of participants which is still nonetheless comparable in terms of sample size to other studies on group trauma-focused treatments with youth (23) or adults (12). More importantly, however, the representativeness of the study sample comprising mostly adolescent males with a refugee background may be limited. Rates of traumatic experiences, PTSD, and depression among refugee minors are higher than in western samples (e.g. 23). However, research has shown that refugees benefit just as much from established trauma-focused treatments (25, 26). This may indicate that they may constitute an especially vulnerable and burdened population but one which is not that different in terms of treatment response.
Thirdly, the study protocol did not include the systematic documentation of how much content of the trauma narrative was actually disclosed and shared in the group discussion in each exposure session. Solely non-standardized reports of group facilitators in supervision gave the impression that almost all participants wanted to share details about their traumatic experiences as they often appreciated the social support from peers and made the valuable experience that other group members had had similar experiences. Experiences and disclosure of sexual abuse are especially associated with feelings of shame and guilt (27) and there might be an unwillingness to share them in a group setting. However, group facilitators reported that these events were shared as well by naming the event and some context (where and when it happened), instead of disclosing many details. Recent research suggests that this brief exposure might elicit a similar activation of the traumatic event resulting in stress and subsequent habituation (28). Altogether important group processes, such as social support, might come into play during disclosure of traumatic events within a group discussion. Future research needs to investigate group mechanisms and dynamics more closely.
Fourthly, it is important to note that in the first session the stress level rating was probably determined solely by the group facilitator as the stress level measure was not introduced until session 2. It is possible, however, that group facilitators asked the participants about their current stress level in session 1 already, as this was discussed and emphasized in training and supervision.
Fifthly, as this study didn’t include an active control group (e.g. a non-exposure CBT group treatment), other potentially stress-related circumstantial factors, effects of time or order of the intervention, and carry-over effects cannot be excluded.
Lastly, although it is a major strength of this study that the stress level was assessed directly in the session and evaluated by the participants themselves and the group facilitators, independent and potentially more objective measurements of stress might have generated additional valuable information. Early studies on trauma-focused treatments employed methods such as ratings of facial expressions and coding of videotaped sessions by independent raters (29). Future research should also take neural and psychophysiological markers into account (30).
Clinical Implications
As no harm to participants could be traced back to the “Mein Weg” intervention (16, 17), in-session exposure with an increase and decrease in stress among traumatized youth seems to be feasible. Hopefully, this will help to refute practitioners’ preconceived ideas about using exposure in a group setting. The findings of this study might serve to motivate practitioners who are reluctant to implement group programs, especially those with in-session exposure, to undergo training in this kind of treatment. In settings in which clinicians’ time and other resources for individual exposure treatment are limited, the implementation of exposure-based group treatments might even increase the number of patients who could receive and benefit from treatment.
Moreover, this intervention was carried out by trained and supervised social workers who are oftentimes referred to as “lay counsellors” as they do not have specific CBT or mental health training. This is a common approach in several trauma-focused individual treatments such as narrative exposure therapy (NET) (31) and group treatments for children such as group-based TF-CBT (32). Especially in settings that lack trained mental health care professionals, the training and supervision of lay counsellors might ensure greater dissemination of treatments and access for more children and adolescents in need.