This study protocol is version 1 from July 4tth, 2019. Recruitment of CYWS facilities began on July 13th, 2020 is scheduled to be completed on January 1st, 2022.
Participants and Eligibility Criteria
Criteria for CYWS facilities. Inclusion criteria for CYWS facilities are: 1) responsible for a number of UYRs of whom at least 8 UYRs have declared an interest in participation, 2) willing to participate in training and implementation of screenings and documentation, 3) agree to name a mental health coordinator (MHC) from their staff, 4) in the case of randomization to BETTER CARE, willing to implement ‘Mein Weg’ and 5) agree to send a caregiver to participate in TF-CBT, 6) availability of at least one eligible therapist for 10 UYRs ≤ 1 hour travel time.
Criteria for participants. Inclusion criteria for participants are 1) age 12-20 years, 2) arrived in Germany as unaccompanied minors, 3) applied for asylum or intend to do so, 4) being cared for by a CYWS facility at screening, 5) written informed consent given by participant and legal guardian (if < 16 years at screening), and 6) reported at least one traumatic event in line with the DSM-5 A criterion. No exclusion criteria have been defined for participants in the RCT.
If they are to be referred to and treated with TF-CBT in the event of randomization to BETTER CARE, UYRs have to meet the following criteria: 1) a score (self-report) of ≥ 25 points in the initial screening with the Child and Adolescent Trauma Screen, second version (CATS-2)  and 2) diagnosis of PTSD (according to International Classification of Diseases 10 [ICD-10]) or any mental disorder (ICD-10) as confirmed by the therapist. Exclusion criteria for TF-CBT treatment are 1) acute danger to themselves or others, 2) acute psychosis, 3) acute suicidality, 4) severe substance abuse, or 5) severe mental disability (each as judged by the therapist).
Criteria for therapists. Therapists need to meet the following inclusion criteria to participate: 1) licensed as either child and adolescent psychotherapist or psychological psychotherapist or child and adolescent psychiatrist, 2) willingness to participate (i.e. motivation to learn TF-CBT and to invest additional time into training and documentation), 3) written informed consent, and 4) willingness to treat up to 3 UYRs with PTSD. There are no exclusion criteria.
The sample size calculation was based on a two-step procedure due to the hierarchical data structure. First, we calculated the sample size for an individual randomized study based on our previous data. We assumed a small effect size (ES) of d = 0.2 for ‘Mein Weg’ as this intervention is implemented with a non-clinical sample, and a medium ES of d = 0.5 for TF-CBT. The latter ES reflects the fact that our previous trial with d = 1.08 (CATS; ) was uncontrolled, and treatment was provided in a university setting with highly experienced therapists. We plan to compare BETTER CARE with usual care+ and expect a small effect (d ≥ 0.30) in favour of BETTER CARE at 12 months (T2) on the CATS-2. A power analysis for a t-test with independent samples (α = 0.05, two-tailed, power .80) results in a minimum sample size of 2x176 participants (352 in total). The second step will factor in the cluster structure of the data with participants nested within CYWS facilities/therapists clusters. We assume that each CYWS facility will recruit on average n = 10 participants. The design effect is needed for the inflation of the group sample size: DE = 1 + (m - 1) * ICC with the fixed cluster size m and the intra-cluster correlation coefficient ICC. For the fixed cluster size, a value of m = 8 was set because of an estimated dropout rate of 20% of participants. For the ICC, a liberal value of 0.05 was assumed to control for the correlation of the outcome of participants living in the same facility. In this context, the DE is 1 + (8 - 1) * 0.05 = 1.35 which results in a group size of 352 x 1.35 = 476 participants. As every facility recruits 10 participants on average, 48 facilities are needed for the whole trial. With an assumed dropout rate of 10% of facilities, an oversampling of 6 facilities is needed, resulting in 54 facilities (= 2 x 27 facilities). Thus, the overall participant sample within these facilities is estimated as 54 x 10 = 540 participants.
Once a CYWS facility has declared its interest in participating in the study, MHCs will be informed about the study details, eligibility criteria, informed consent, and randomization procedure in an initial information workshop. They will also be informed about participant inclusion criteria to ensure that all the participants who are willing to participate are indeed eligible. Eligibility of a facility will be evaluated using a facility questionnaire which will be completed online in the context of a nationwide CYWS survey which is part of the study. Once all facility inclusion criteria are clear, the head of facility and all principal investigators (RR, HK, RK, UZ) will sign a cooperation agreement.
Baseline screening (T0) will be undertaken at the respective CYWS facility by study staff and MHCs. All participants will complete self-report questionnaires on an iPad. Proxy-report questionnaires will be completed by the respective caregiver of a participant. After screening, the study center will communicate the result of the subsequent randomization to the facility by letter and telephone. In a next step, each participant will receive a short written report documenting the main findings from the screening to be shared with the caregiver. Follow-up assessments will take place for both study arms, six (T1) and twelve months (T2) after randomization (see Fig. 3 for all assessment details). It will also be followed by a short written report on the assessment results. A sub-sample of n = 100 participants will be followed up for two further assessments at 18 (T3) and 24 months (T4) after randomization. They will complete additional questionnaires on group climate, social support, and post-migration factors at each time point (T0-T4, see measures).
Before randomization, the first step in all participating CYWS facilities will be a systematic screening of psychopathology, psychosocial difficulties and indication for treatment (PTSS as measured with the questionnaire CATS-2 as the primary outcome (see measures)). The MHCs will undergo training in trauma-informed assessment (screening), will learn how the written report is to be interpreted and how they can derive the individual indication for prevention or psychotherapy in each of their UYRs (indication).
BETTER CARE arm
The stepped-care approach BETTER CARE includes the first step as described above (screening and indication). The second step refers to subclinical cases (CATS-2 score < 25). They will be allocated to the preventive group intervention ‘Mein Weg’. Participants above the clinical cut-off (≥ 25) will be referred to TF-CBT (step 3). UYRs with a CATS-2 score ≥ 25 who refuse to participate in TF-CBT or who have to wait more than three months for TF-CBT treatment, will be invited to participate in ‘Mein Weg’ (see Figure 1).
Trauma-focused preventive group intervention ‘Mein Weg’. The manualized intervention can be delivered by trained non-specialist social workers. It empowers young refugees to cope with their traumatic experiences and stress-related symptoms. The rationale of this intervention is derived from the framework of exposure-based and cognitive interventions for trauma-exposed minors. It aims to empower the social workers in the CYWS programs to gain a deeper understanding of the UYRs’ individual (trauma) history and cultural background, and to provide appropriate support to the young refugees. ‘Mein Weg’ is a group intervention delivered in 6-8 sessions lasting between 90 and 120 minutes. It comprises psychoeducation to normalize psychological symptoms, teaching of a relaxation technique to reduce hyperarousal symptoms, a trauma narrative to counteract dysfunctional avoidance and unwanted intrusions of traumatic memories, and cognitive restructuring to reduce trauma-related cognitive distortions. The aims here are to foster social support, help participants regain a sense of safety, and anticipation of self-efficacy with future challenges. In addition to the evaluated ‘Mein Weg’ manual , the author will add 2-3 sessions on problem-management strategies related to asylum stressors.
Implementation of ‘Mein Weg’. In order to implement ‘Mein Weg’ within BETTER CARE facilities, social workers in the CYWS facilities will attend a 2-day training in trauma, trauma-related disorders, their treatment and the implementation of ‘Mein Weg’ by the author of the manual (EP) and certified ‘Mein Weg’ trainers. During the implementation of the program, social workers will receive weekly case consultations and complete session checklists as fidelity checks.
Trauma-focused cognitive behavioral therapy. TF-CBT is a short-term, evidence-based treatment for children and adolescents with PTSD which can be delivered in 12 to 15 sessions . It has nine components that are subsumed with the acronym PRACTICE: psychoeducation and parenting skills (components 1 and 2); relaxation; affect modulation; cognitive coping; trauma narrative; in vivo exposure; conjoint parent-child sessions, and enhancement of safety and development. We will extend TF-CBT by three sessions (up to 18 sessions) that will focus on relapse prevention and problem management related to asylum stressors. A trusted caregiver is to be involved in up to 50% of the treatment sessions and the treatment is usually provided in double sessions (one session with the child and one with the caregiver).
Implementation of TF-CBT. TF-CBT is specifically suited to improving social support. It requires a lower level of language skills than interventions for adults. Furthermore, as it has been studied with participants aged three to 18, the level of language requirements can be modified to suit the individual patient. Hence, limited language skills or the involvement of interpreters are no barriers to receiving TF-CBT within this trial. The TF-CBT training package for therapists consists of online training (enriched with culturally adapted materials, http://tfkvt.ku.de), an instructor-led two-day workshop on the treatment, and case consultations on ongoing cases via telephone (biweekly), each provided by certified TF-CBT trainers. The specific aspects of working with UYRs taught in the workshop will be integrated into the ‘regular’ TF-CBT components (i.e., working with interpreters, preparation for future stressors). Therapists will complete the Brief Practice Checklist Revised (BPC-R; ) after each session as a measure of fidelity. A TF-CBT treatment case will be deemed to have been completed if at least 8 sessions, including a trauma narrative, have been completed.
To further support the implementation of and decrease barriers to ESTs for UYRs, we will develop a one-day training session on translation in trauma-focused treatments for interpreters that will be proposed to all suitable (lay) interpreters in the communities of clusters randomized to BETTER CARE.
Usual care+ arm. The control condition will be called usual care+ as it presents the usual care that UYRs receive in the CYWS facilities and mental healthcare system enhanced with screening and indication procedures offered by study participation (see step 1). All mental health care utilization will be documented for each participant at every assessment so as to be able to offer a clear description of usual care+ afterwards (Child and Adolescent Mental Health Service Receipt Inventory [CAMHSRI-DE]; (, see measures).
All self-report measures will be available in the following languages: German, English, French, Arabic, Kurdish, Dari, Farsi, Pashto, Somali, and Tigrinya. They will be recorded on an iPad or on paper by the participant. The respective caregiver will respond to the proxy-report measures in German.
Primary outcome. The CATS-2  at the 12-month follow-up is the primary outcome. The first part consists of an event checklist with 15 items to be answered yes/no. The second part consists of 20 items corresponding directly to PTSD criteria according to the Diagnostic and Statistical Manual of Mental Disorders 5 (DSM-5; ) They measure the severity of PTSS employing a 4-point scale ranging from 0 (never) to 3 (almost always). To evaluate psychosocial functioning, five dichotomous items (yes/no) investigate whether the rated PTSS interfere with key areas of functioning. Good internal consistency (α = .75–.93) has been confirmed in the initial validation  and in an RCT with UYRs (α = .75) . At present, the CATS is accessible in several languages and has been successfully used in previous studies with UYRs [2, 19]. For the current study a slightly adapted version (CATS-2) will be used to cover ICD-11 PTSD and complex PTSD criteria in addition.
Secondary Outcomes. All secondary outcomes will be evaluated at the 12-month follow-up. The Patient Health Questionnaire (PHQ-9)  comprises 9 items that reflect the criteria for the diagnosis of depressive disorders according to DSM-IV. It showed good reliability (α = .86–89) and has been validated in many contexts and languages [30, 31].
The Generalized Anxiety Disorder Assessment (GAD-7)  comprising 7 items assesses symptoms of anxiety according to DSM-IV and has been validated (α = .92) in many contexts and languages .
The World Health Organisation Quality of Life-BREF (WHOQOL-BREF; ) comprises 26 items measuring 4 key areas (physical health, psychological health, social relationships, and environment) over the previous 4 weeks . The short version has been validated in many languages. The Cronbach’s α tested in international samples was .68–.82 .
Exploratory Outcomes. The primary and secondary outcomes will be analysed in an exploratory manner at the 6-month follow-up.
The Alcohol Use Disorders Identification Test (AUDIT) is a 10-item questionnaire used to assess recent excessive drinking, and is in conformity with ICD-10 definitions . It is a well-validated (α = .80–.94)  measure of risk across gender, age and cultures.
The Cannabis Use Disorder Identification Test, Revised (CUDIT-R) is an 8-item questionnaire used to assess cannabis use disorder severity . The CUDIT-R has been validated (α = .91) as both a screening test and a rating scale for cannabis use disorder severity .
The EQ-5D-3L  is a two-part instrument for the definition and the valuation of health states as the basis for the estimation of quality adjusted life years (QALYs). It consists of a descriptive part and a visual analogue scale similar to a thermometer . At present, the EQ-5D-3L is available in more than 180 languages including all those expected to be relevant for the project.
Proxy measures. The following measures will be detailed by the respective caregiver of each participant: The CATS-2 caregiver version  which comprises the same items as the CATS self-report with a reference to the child/adolescent (“has your child…”).
One item derived from the Nation Outcome Scales for Children and Adolescents (HoNOSCA)  will be used to obtain a proxy estimation of substance abuse.
The CAMHSRI-DE  is an interview tool for the assessment of medical and psychosocial services used by children and adolescents with mental health problems based on the European version CAMHSRI-EU . Service costs will be estimated by combining service use assessed by the CAMHSRI-DE with unit costs gathered from a broad spectrum of information sources. The CAMHSRI-DE will be adapted for refugee youths (CAMHSRI-REF) and will be administered by the caregiver in collaboration with the UYR.
Long-term follow-up sub-sample. A subsample of n = 100 will be assessed with additional questionnaires and will be followed up for 18 (T3) and 24 months (T4). The short version of the Group Climate Instrument for Children (GCIC)  consisting of 14 items assessing open and closed climate in institutions, will be used. This questionnaire showed good to moderate reliability for the open (Cronbach’s α = .91) and the closed scales (Cronbach’s α = .71). It has been validated in the context of residential care in different countries , .
The Brief Acculturation Orientation Scale (BAOS)  is a bidimensional instrument assessing the acculturation orientation towards the home country and the host country independently. The home and host country factors of the BAOS showed acceptable reliability (Cronbach’s α = .72–.78).
The Brief Sociocultural Adaptation Scale (BSAS)  is a 12-item questionnaire that assesses various aspects of sociocultural adaption (e.g. language, climate, people, values and beliefs). The scale demonstrated good reliability (Cronbach’s α = .85); it is available and has been validated in different languages.
The questions on social support are based on the Social Support Questionnaire (SSQ6-G) . In its original version, the three scales showed good reliability (Cronbach’s α = .71 to .92).
The Daily Stressors Scale for Young Refugees (DSSYR)  examines the extent to which 15 different post-migration daily stressors were experienced by the participants during the previous month.
[Please insert Figure 3 about here.]
Measures for ‘Mein Weg’ trainers and TF-CBT therapists. Both groups of professionals will complete questionnaires before and after the training (therapists also before the online training) as well as 6 and 12 months post-training. The Professional Quality of Life (ProQOL) is a 30-item questionnaire that measures the negative and positive effects of helping others who experience suffering and trauma . The items form 3 subscales corresponding to compassion satisfaction (α = .89–.91), burnout (α = .74–.80) and secondary traumatic stress (α = .86–.89) .
The Evidence-Based Practice Attitude Scale (EBPAS-36) is a 36-item questionnaire derived from the EBPAS-50. It assesses the attitudes of mental health providers towards the adoption of evidence-based practice . The measure has been well validated .
As ‘Mein Weg’ trainers also work as professionals in CYWS facilities, they will be asked to complete the Implementation Climate Scale (ICS). It measures the degree to which the organizational climate is supportive of evidence-based practice implementation . High internal consistency has been confirmed for the total scale (α = .91), and for the subscales (α = .81–.91).
Measures for translators. All participants in the TF-CBT specific translator training will complete measures at four time points: before and after training and 6 and 12 months post-training. They include an evaluation of the translator training (14 items), experiences and attitudes towards translating in mental health contexts (41 items), and the PTSD Checklist for DSM-5 (PCL-5) .
Methods against Bias
Random allocation of facilities and participants is ensured by the trial design. All training sessions will be highly standardized and treatment fidelity will be ensured by ongoing case consultations and by session checklists in both interventions. The blinding of UYRs is not possible. All outcomes on the patient level will be well-established measures. Data records will be kept as up-to-date as possible by means of online symptom monitoring and documentation of treatment sessions.
Data management and storage
Data will be directly entered in an online assessment tool by the relevant informant. Data collected using the paper-pencil procedure will be entered in the same database by an independent data manager. As almost all data will be directly entered in the database, no source data verification will be needed. Incoming data will be continuously monitored, and checked for plausibility, quality and completeness by an independent data manager installed at the University Clinic Ulm. All data will be stored and transferred in pseudonymized form. All the main participants and the IT company responsible for the assessment tool of the trial have signed and followed a data protection concept in accordance with the General Data Protection Regulation (EU) 2016/679 (GDPR). Only authorized and trained study personnel will receive a login-roll in line with their task (e.g., research assistant to start screening, data manager for query management, data entry, data validation and plausibility checks). For the purpose of long-term storage, the original, pseudonymized data (after database lock) will be stored at the University Clinic Ulm and with the relevant project partners. They will be made available on request to scientific colleagues after publication of the results.
We will analyze the primary outcome (CATS-2 severity score) using hierarchical linear mixed (HLM) models to account for the nesting of cases within clusters. Mixed effect models, with fixed effects of group (BETTER CARE vs usual care+) and time (Intake (T0) versus 6 months (T1) versus 12 months (T2)) as well as their interaction, will be conducted. Primary outcome analysis will be based on the ITT sample (n=540 participants) as the mixed model approach will take missing data and different numbers of reassessments per participant/patient into account. A significance level of α = 0.05 will be used. We will also use exploratory mixed linear models to examine the impact of other variables (gender, age, nationality, asylum status) on our primary outcome. Separate (hierarchical) linear mixed models will be used to analyze the secondary outcomes.
Ethical considerations, monitoring and safety
The study was planned and will be conducted in accordance with the International Council for Harmonization Guideline for Good Clinical Practice . It has been approved by the ethics review board of the University Ulm (07.10.2019) and Eichstaett-Ingolstadt (08.11.2019). The trial is registered under German Clinical Trials Register (www.germanctr.de; registration number DRKS00017453, date 11.12.2019).
Study safety will be ensured by monitoring for the incidence of serious adverse events (SAEs; e.g., suicide attempts, unplanned hospitalizations, occurrence of life-threatening conditions) at all follow-up assessments. All such incidents and other aspects of study safety will be regularly reported to an independent Data and Safety Monitoring Board, which offers advice on protocol changes in the event of such incidences, or even on discontinuation of the trial. However, no harm to participants is to be expected. In our previous studies with UYRs, we did not detect any harm arising from the interventions for the participants [12, 19].
All participants will be informed about the study in oral and written form, with details of the trial’s procedures, risks, costs, confidentiality, data storage, and about the right to discontinue participation at any time without giving any reasons. Participants will be free to continue treatment with their therapists when they quit the research program. Written informed consent will be obtained from participants and, in case of the minors, from their parents/legal guardians. Furthermore, caregivers, ‘Mein Weg’ trainers, psychotherapists, and translators will give their written informed consent. In general, treatment and the associated processes of change and diagnostic procedures may be evaluated as stressful by the participant . Therefore, symptoms may temporary worsen. In the context of TF-CBT and ‘Mein Weg’, the development of a trauma narrative can be experienced as stressful as it involves direct confrontation with the traumatic experience and avoided stimuli. However, participants can directly benefit from successful treatment and long-term symptom reduction can be achieved. Even for the control group who will not receive the stepped-care, we do not expect any harm: in two previous studies [2, 8] we assessed young refugees’ mental health without providing further interventions. No serious or adverse events were reported during the two studies.
The study offers the following incentives/compensations: CYWS facilities will receive a lump sum per case for each participant at randomization (€60). Participants will receive compensation for taking part in assessments (€30 each). They may receive compensation for the costs of travel to study therapists of up to €10 per session. In addition, caregivers from CYWS will receive compensation of €40/session for participating in TF-CBT sessions. Therapists and interpreters will receive compensation for travel costs to the workshops (up to €100/person). In addition, therapists may receive compensation for participation in case consultations (€30/hour) and for additional diagnostic and documenting activities (up to €240/case). In the event that treatment or translation costs are not covered, costs for one third of treatment or translation cases can be covered by study funds.