TF-CBT vs. Enhanced Treatment-As-Usual for Violence-Exposed Youth: Comprehensive Study Notes

Study Purpose & Hypotheses

• Purpose: Compare the effectiveness of Trauma-Focused Cognitive Behavioural Therapy (TF-CBT) with Enhanced Treatment As Usual (eTAU) for children/adolescents exposed to family violence and presenting severe trauma symptoms in Child & Adolescent Mental Health Services (CAMHS).

• Hypothesis: TF-CBT would yield greater reduction in trauma symptoms, broader psychiatric symptoms, and Diagnostic and Statistical Manual of Mental Disorders – $4^{th}$ Edition (DSM-IV) PTSD diagnosis assignment than eTAU.

• Exploratory aims: (a) examine change in non-offending caregivers’ own trauma/psychiatric symptoms; (b) explore outcome differences between structured eTAU (Eye-Movement Desensitisation & Reprocessing – EMDR / Cognitive Behavioural Therapy – CBT) vs. unspecified eTAU.

Theoretical & Empirical Background

• Interpersonal traumas such as Intimate Partner Violence (IPV) and Child Abuse (CA) pose high risk for children’s anxiety, depression, conduct problems, and Post-Traumatic Stress Disorder (PTSD).

  • Double exposure (IPV $+$ CA) heightens severity.

• TF-CBT supported by extensive RCT evidence (esp. sexual abuse samples) and recommended by NICE and ISTSS; some evidence that it reduces mild–moderate parental PTSD/depression.

• Debate: EBTs vs. TAU – meta-analyses (e.g., Weisz $et\,al.$ $2013$) show EBPs often do NOT outperform usual care when confounders (allegiance, supervision, homework) are balanced.

Design & Procedure

• Type: Randomised Controlled Effectiveness Trial, registered ISRCTN$58027256$.

• Assessment points: Pre-treatment $T1$, $6$-month $T2$, $12$-month $T3$ follow-up.

• Inclusion criteria: age $5$–$17$ yrs; $\ge 2$ incidents of caregiver-perpetrated IPV/CA (physical, psychological, sexual); severe trauma reactions (≥ $5$ DSM-IV PTSD symptoms incl. ≥ $1$ per cluster).

• Exclusion: mental retardation, autism, psychosis, inpatient need, interpreter need.

• Recruitment (Jan $2012$–Jun $2014$): $n=582$ screened → $n=93$ randomised (TF-CBT $47$, eTAU $46$); $4$ dropped before baseline → analytic sample $n=89$ (TF-CBT $45$, eTAU $44$). Self-reports from $n=73$ children ≥ $9$ yrs, parent reports $n=85$.

• Ethical approval: Gothenburg Regional Committee (D.nr $166$-$11$); informed consent from caregivers and children ≥ $9$ yrs.

Participant Characteristics (T1)

• Mean age: TF-CBT $\bar x =11.8$ yrs (SD $3.48$); eTAU $\bar x =13.0$ yrs (SD $3.37$) – controlled as covariate.

• Sex: $\approx$ $66\%$ girls overall.

• Migration: $\approx$ $42\%$ caretaker born outside Sweden; $\approx$ $14\%$ children foreign-born.

• Living: $66\%$ with one parent only; $22\%$ foster/institution.

• Trauma exposures (overall): witnessed IPV $86\%$, abused $66\%$, sexually abused $25\%$, victim of violent crime $34\%$; mean distinct traumatic events $\mu =3.92$ (range $1–13$).

Therapists

• TF-CBT: $5$ female clinicians ( $4$ psychologists, $1$ social worker); experience $2–10$ yrs (Median $4$); $20$-hr model training $+$ $50$-hr supervision across $2$ yrs.

• eTAU: $6$ clinicians ( $4$ psychologists, $2$ social workers; $1$ male); experience $2–32$ yrs (Median $11.5$); supervisor in family therapy; encouraged to mirror TF-CBT group cohesion & supervision dosage.

Interventions

TF-CBT

• Manualised $8–16$ session model (PRACTICE components) but mean sessions $\mu=17.3$ (SD $4.6$; range $7–30$); $24\%$ received >$20$ sessions due to complexity.

• Trauma narrative delivered in $29/32$ documented cases; mean narrative sessions $6.24$ (SD $1.88$).

• $15\%$ sessions deviated from strict child-then-parent split to accommodate circumstances; adjunctive services (network meetings, medical, foster parent meetings) permitted.

eTAU ("enhanced")

• Mean sessions $\mu =22.5$ (SD $11.3$; range $6–46$) – significantly higher than TF-CBT ( $p=.014$).

• Modalities chosen collaboratively: EMDR $n=18$, CBT (non-TF-CBT) $n=3$, unspecified/low-evidence interventions $n=23$ (medication, family therapy, parent support, massage, network meetings, etc.). Mean $3.22$ concurrent interventions per case.

Treatment Fidelity

• $13$ TF-CBT and $10$ eTAU audio-taped cases rated with Brief Practice Checklist.

  • $11$ TF-CBT cases fully adherent; $2$ partially (corrected on re-rating).

  • $1$ eTAU case partially TF-CBT-like (subsequent case rated “definitely not TF-CBT”).

Measures

• K-SADS-PL: structured child & caregiver interview for DSM-IV diagnoses; inter-rater internal consistency $\alpha_{total}=0.93$.

• Trauma Symptom Checklist for Children (TSCC): $54$ items; sub-scales used – PTS ($\alpha=0.89$) & psychiatric ($\alpha=0.94$); clinical cutoff $T\ge65$.

• Strengths & Difficulties Questionnaire – Parent (SDQ-P): total problems $20$ items, $\alpha=0.85$ ; Swedish cutoff $\ge11$.

• Global Severity Index (GSI) – abridged SCL-$90$, $\alpha=0.93$ for caregivers.

• Impact of Event Scale-Revised (IES-R) for caregiver trauma, $\alpha=0.95$.

Data Analysis

• Randomisation check: only age differed; controlled in ANCOVAs along with number of sessions, child sex, therapist experience.

• Approach: Intention-To-Treat (ITT) with Last Observation Carried Forward (LOCF).

• ANCOVA for between-group differences; effect size $d$ computed via Campbell Collaboration calculator; small $0.20$, medium $0.50$, large $0.80$ per Cohen $1992$.

• Clinical significance: Reliable Change Index (RCI) with $\alpha=0.10$ on TSCC-PTS & SDQ-P; categorised as Recovered / Improved / Unchanged / Worsened / Deteriorated. McNemar tests for diagnosis shifts.

• Attrition: $8\%$ at $T2$, $34\%$ at $T3$; no differential attrition by group; higher dropout for parents with compulsory education only.

Primary Outcomes

(K-SADS, TSCC, SDQ)

• Both groups showed significant within-group reductions $T1\to T2$ and $T1\to T3$ ( p<.05) in clinician-rated PTSD, total K-SADS symptoms, self-reported PTS & psychiatric symptoms, and parent-rated SDQ.

• Effect sizes: TF-CBT within-group $d<em>{T1-T2}=0.38–0.81$; eTAU $d</em>{T1-T2}=0.18–0.64$. By $T3$, effects similar or slightly favoured eTAU on some scales.

• ANCOVAs: No statistically significant between-group differences at $T2$ or $T3$ on any primary outcome.

Secondary Outcomes (Caregiver)

• Small–medium within-group improvements in caregiver global symptoms (GSI) and trauma symptoms (IES-R); no TF-CBT vs. eTAU differences.

Diagnostic Remission (K-SADS)

• PTSD prevalence: TF-CBT $\frac{34}{45}=76\%$ at $T1$ → $\frac{16}{45}=36\%$ at $T2$; eTAU $80\%$ → $36\%$.

• Any non-PTSD diagnosis: TF-CBT $47\%\to18\%$; eTAU $32\%\to22\%$.

• No between-group significance.

Clinical Significance (RCI)

• TSCC-PTS $T1\to T2$: TF-CBT $45\%$ reliable positive change vs. eTAU $28\%$.

• SDQ total $T1\to T2$: TF-CBT $22\%$ vs. eTAU $25\%$ improved.

• At $T3$ patterns converged; <$50\%$ achieved reliable improvement overall; deterioration rates low (≤ $6\%$).

eTAU Sub-Group Exploration

• Structured (EMDR/CBT) vs. Unspecified eTAU:

  • Greater PTSD symptom reduction (K-SADS) $F=4.18, p=.047, d=0.60$.

  • Greater self-reported psychiatric symptom reduction (TSCC-PSY) $F=6.28, p=.017, d=0.80$.

  • Greater SDQ parent-rated improvement $F=4.14, p=.048, d=0.62$.

• Suggests structure/manualisation matters within TAU.

Discussion & Interpretation

• Contrary to previous RCTs, TF-CBT did not outperform eTAU.

• Possible explanations:

  • Sample complexity: high rates of multi-trauma, double exposure, probable ICD-$11$ Complex PTSD → may need longer/modified protocols.

  • Delivery factors: TF-CBT therapists instructed to follow standard protocol; $24\%$ required >$20$ sessions, but some may still have been prematurely terminated.

  • Comparison treatment was "enhanced" (structured, supervised, high therapist allegiance) incl. >50\% manualised EMDR/CBT, shrinking expected effect-size gap.

  • Real-world CAMHS setting often yields smaller EBT–TAU differences than university trials.

• Cost-effectiveness caveat: TF-CBT fewer sessions but each $80–90$ min versus eTAU $45–50$ min.

Strengths

• First independent RCT of full-protocol TF-CBT with IPV/abuse-exposed youths in Europe.

• Real-world clinic, socio-economically disadvantaged catchment, multi-traumatised sample – high external validity.

• Balanced design controlling supervision, caseload, therapist allegiance.

• Multiple informants (child, caregiver, clinician) & structured diagnostic interview.

• ITT, RCI, subgroup analyses.

Limitations

• Small sample → risk Type-II errors; effect sizes modest.

• eTAU heterogeneity; subgroup cell sizes too small for full three-way comparison.

• Assessors not blind to time-point or treatment, possible halo effect (mitigated by multi-informant convergence).

• $34\%$ attrition by $T3$; limited RCI calculable for all participants.

Implications for Practice

• Psychotherapy (manualised or structured TAU) can benefit multi-traumatised youths even amidst ongoing adversity; clinicians should not delay treatment awaiting environmental “stabilisation.”

• Need for flexibility/extended sessions and practice-based adaptations when applying TF-CBT to complex PTSD presentations.

• Monitoring individual progress over time and adding booster or alternative modules crucial; fewer than half achieve reliable change.

• Clinicians should favour structured, evidence-grounded components in usual care to maximise outcomes.

Implications for Research & Policy

• Future trials should:

  • Include complex trauma samples, larger Ns, dismantle treatment components, and detail TAU content.

  • Examine cost-effectiveness considering session length.

  • Investigate therapist, family, and contextual moderators of response (“what works for whom”).

• Active control conditions must receive equivalent supervision/training to ensure ecological validity of EBT superiority claims.

Funding & Declarations

• Funders: Swedish Research Council for Health, Working Life and Welfare; Children’s Welfare Association; Mayflower Charity Foundation; Swedish Crime Victim Compensation & Support Authority.

• Open access under CC BY-NC-ND $4.0$. Authors report no competing interests.

Study Purpose & Hypotheses

• Purpose: To conduct a rigorous comparison of the effectiveness of Trauma-Focused Cognitive Behavioural Therapy (TF-CBT) against Enhanced Treatment As Usual (eTAU) for a specific population: children and adolescents who have been exposed to family violence and are presenting with severe trauma symptoms. This study was conducted within the regular operations of Child & Adolescent Mental Health Services (CAMHS), aiming for real-world applicability.

• Hypothesis: The primary hypothesis predicted that TF-CBT, a highly evidence-based intervention, would lead to demonstrably greater reductions in trauma symptoms, a broader spectrum of psychiatric symptoms, and a higher rate of remission from Post-Traumatic Stress Disorder (PTSD) as diagnosed using the Diagnostic and Statistical Manual of Mental Disorders – $4^{th}$ Edition (DSM-IV) criteria, when compared to eTAU.

• Exploratory aims:

  • (a) To meticulously examine the potential for change in non-offending caregivers’ own trauma and general psychiatric symptoms, as their well-being can significantly impact a child's recovery.

  • (b) To delve into potential outcome differences within the eTAU group itself, specifically comparing structured forms of eTAU (such as Eye-Movement Desensitisation & Reprocessing – EMDR or Cognitive Behavioural Therapy – CBT) against unspecified, less structured forms of eTAU to determine if structured approaches within usual care yield better results.

Theoretical & Empirical Background

• Interpersonal traumas, including Intimate Partner Violence (IPV) and Child Abuse (CA), are well-established risk factors significantly increasing the likelihood of children developing various psychological issues such as anxiety, depression, conduct problems, and especially Post-Traumatic Stress Disorder (PTSD).

  • The risk and severity are further heightened in cases of double exposure, where children experience both IPV and CA (IPV $+$ CA), leading to more complex and entrenched symptom presentations.

• TF-CBT is strongly supported by an extensive body of randomised controlled trial (RCT) evidence, particularly from studies involving sexual abuse samples. Consequently, it is recommended as a first-line intervention by prominent organisations like the National Institute for Health and Care Excellence (NICE) and the International Society for Traumatic Stress Studies (ISTSS). There is also some emerging evidence suggesting that TF-CBT can reduce mild–moderate levels of parental PTSD and depression, indicating potential collateral benefits for caregivers.

• Debate: A significant ongoing debate exists in the field regarding the superiority of evidence-based treatments (EBTs) over usual care (TAU). Meta-analyses, such as the one conducted by Weisz $et\,al.$ ($2013$), frequently show that EBTs often do NOT outperform usual care when critical confounders are carefully balanced. These confounders include therapist allegiance to the treatment model, intensity and quality of clinical supervision, and the degree of patient engagement in homework assignments, suggesting that contextual and implementation factors play a crucial role.

Design & Procedure

• Type: This study was designed as a Randomised Controlled Effectiveness Trial, aiming to assess treatment efficacy in a real-world clinical setting rather than an idealised research environment. The trial was officially registered under the identifier ISRCTN$58027256$, ensuring transparency and adherence to research standards.

• Assessment points: Participants underwent comprehensive assessments at three key points: pre-treatment ($T1$), at $6$-month follow-up ($T2$) after treatment initiation, and at $12$-month follow-up ($T3$) to evaluate the durability of treatment effects.

• Inclusion criteria: Participants were rigorously selected based on specific criteria to ensure homogeneity and relevance to the study's purpose:

  • Age: Children and adolescents between $5$ and $17$ years of age.

  • Trauma Exposure: A documented history of at least $2$ incidents of caregiver-perpetrated Intimate Partner Violence (IPV) or Child Abuse (CA), encompassing physical, psychological, or sexual forms of abuse.

  • Symptom Severity: Presentation of severe trauma reactions, defined as experiencing at least $5$ symptoms of DSM-IV PTSD, including at least $1$ symptom from each diagnostic cluster (re-experiencing, avoidance/numbing, hyperarousal).

• Exclusion: To maintain the study's focus and ensure participant safety, individuals were excluded if they presented with mental retardation, autism spectrum disorder, psychosis, an immediate need for inpatient psychiatric care, or required an interpreter for communication during assessments or treatment.

• Recruitment (Jan $2012$–Jun $2014$): During the recruitment period, a total of $n=582$ children and adolescents were screened for eligibility. From this initial pool, $n=93$ met the inclusion criteria and were subsequently randomised into either the TF-CBT group ($n=47$) or the eTAU group ($n=46$). It is noted that $4$ participants dropped out before the baseline assessment, leading to a final analytic sample of $n=89$ (TF-CBT $n=45$, eTAU $n=44$). Self-report data was collected from $n=73$ children aged $9$ years or older, while parent reports were obtained for $n=85$ participants.

• Ethical approval: The study received full ethical approval from the Gothenburg Regional Committee for research ethics (D.nr $166$-$11$). Informed consent was diligently obtained from all participating caregivers, and assent was obtained from children aged $9$ years or older, respecting the autonomy of all participants.

Participant Characteristics (T1)

• Mean age: At baseline ($T1$), the mean age in the TF-CBT group was $\bar x =11.8$ years (SD $3.48$), while in the eTAU group, it was $\bar x =13.0$ years (SD $3.37$). This slight age difference was identified and statistically controlled for as a covariate in subsequent analyses to ensure it did not bias the outcomes.

• Sex: The overall sample exhibited an approximate sex distribution of $66\%$ girls.

• Migration: Regarding demographic origin, approximately $42\%$ of the caregivers were born outside of Sweden, and about $14\%$ of the participating children were foreign-born, indicating cultural diversity within the sample.

• Living: A substantial proportion of the children, $66\%$, were living with only one parent, and $22\%$ were in foster care or institutional settings, highlighting the challenging family circumstances of the study population.

• Trauma exposures (overall): The participants had experienced a high number of traumatic events: witnessed IPV was reported by $86\%$, direct abuse by $66\%$, sexual abuse by $25\%$, and being a victim of violent crime by $34\%$. The mean number of distinct traumatic events experienced by participants was $\mu =3.92$ (ranging from $1$ to $13$), underscoring the high complexity of the trauma exposure in this clinical sample.

Therapists

• TF-CBT: The TF-CBT interventions were delivered by $5$ female clinicians, comprising $4$ psychologists and $1$ social worker. Their clinical experience ranged from $2$ to $10$ years, with a median of $4$ years. These therapists underwent intensive training, including a $20$-hour model-specific training, supplemented by an extensive $50$-hour supervision period spread across $2$ years to ensure fidelity to the TF-CBT protocol.

• eTAU: The eTAU interventions were provided by $6$ clinicians, consisting of $4$ psychologists and $2$ social workers; notably, $1$ of these clinicians was male. Their experience level was considerably wider, ranging from $2$ to $32$ years, with a median of $11.5$ years, suggesting a more seasoned group on average. Therapists in the eTAU group were supervised in family therapy and were specifically encouraged to mirror the group cohesion and supervision dosage of the TF-CBT group to minimise differences in non-specific treatment factors.

Interventions

TF-CBT

• This intervention utilised a manualised $8–16$ session model, adhering to the PRACTICE components (Psychoeducation and parenting skills, Relaxation, Affective regulation skills, Cognitive coping and processing, Trauma narrative, In-vivo mastery of trauma reminders, Conjoint child-parent sessions, Enhancing safety and future planning). Despite the manualised range, the mean number of sessions delivered was $\mu=17.3$ (SD $4.6$; range $7–30$), indicating that a flexible approach was often required. Notably, $24\%$ of participants received more than $20$ sessions, reflecting the complexities inherent in their cases.

• The crucial trauma narrative component was successfully delivered in $29$ out of $32$ documented cases, with a mean of $6.24$ sessions (SD $1.88$) dedicated to this core aspect of the therapy.

• Although TF-CBT advocates for a strict child-then-parent session split, $15\%$ of sessions deviated from this protocol to accommodate specific clinical circumstances. Importantly, adjunctive services such as network meetings, medical consultations, and foster parent meetings were permitted and integrated as needed to support comprehensive care.

eTAU ("enhanced")

• The eTAU arm involved a significantly higher mean number of sessions, reaching $\mu =22.5$ (SD $11.3$; range $6–46$). This was statistically higher than the TF-CBT group ( $p=.014$), suggesting eTAU typically involved a longer duration of engagement.

• Modalities within eTAU were chosen collaboratively between the therapist and the family, leading to varied interventions. Specifically, EMDR was used for $n=18$ cases, non-TF-CBT forms of CBT for $n=3$ cases, and a wide array of unspecified or low-evidence interventions for $n=23$ cases (e.g., medication management, general family therapy, parent support groups, massage, and network meetings). On average, participants in the eTAU group received a mean of $3.22$ concurrent interventions per case, highlighting the multi-modal nature of this

multi-modal nature of this.

Treatment Fidelity

• $13$ TF-CBT and $10$ eTAU audio-taped cases rated with Brief Practice Checklist.

  • $11$ TF-CBT cases fully adherent; $2$ partially (corrected on re-rating).

  • $1$ eTAU case partially TF-CBT-like (subsequent case rated “definitely not TF-CBT”).

Measures

• K-SADS-PL: structured child & caregiver interview for DSM-IV diagnoses; inter-rater internal consistency $\alpha_{\text{total}}=0.93$.

• Trauma Symptom Checklist for Children (TSCC): $54$ items; sub-scales used – PTS ($\alpha=0.89$) & psychiatric ($\alpha=0.94$); clinical cutoff $T\ge65$.

• Strengths & Difficulties Questionnaire – Parent (SDQ-P): total problems $20$ items, $\alpha=0.85$ ; Swedish cutoff $\ge11$.

• Global Severity Index (GSI) – abridged SCL-$90$, $\alpha=0.93$ for caregivers.

• Impact of Event Scale-Revised (IES-R) for caregiver trauma, $\alpha=0.95$.

Data Analysis

• Randomisation check: only age differed; controlled in ANCOVAs along with number of sessions, child sex, therapist experience.

• Approach: Intention-To-Treat (ITT) with Last Observation Carried Forward (LOCF).

• ANCOVA for between-group differences; effect size $d$ computed via Campbell Collaboration calculator; small $0.20$, medium $0.50$, large $0.80$ per Cohen $1992$.

• Clinical significance: Reliable Change Index (RCI) with $\alpha=0.10$ on TSCC-PTS & SDQ-P; categorised as Recovered / Improved / Unchanged / Worsened / Deteriorated. McNemar tests for diagnosis shifts.

• Attrition: $8\%$ at $T2$, $34\%$ at $T3$; no differential attrition by group; higher dropout for parents with compulsory education only.

Primary Outcomes

(K-SADS, TSCC, SDQ)

• Both groups showed significant within-group reductions $T1\to T2$ and $T1\to T3$ ( p<.05) in clinician-rated PTSD, total K-SADS symptoms, self-reported PTS & psychiatric symptoms, and parent-rated SDQ.

• Effect sizes: TF-CBT within-group $d<em>{\text{T1-T2}}=0.38–0.81$; eTAU $d</em>{\text{T1-T2}}=0.18–0.64$. By $T3$, effects similar or slightly favoured eTAU on some scales.

• ANCOVAs: No statistically significant between-group differences at $T2$ or $T3$ on any primary outcome.

Secondary Outcomes (Caregiver)

• Small–medium within-group improvements in caregiver global symptoms (GSI) and trauma symptoms (IES-R); no TF-CBT vs. eTAU differences.

Diagnostic Remission (K-SADS)

• PTSD prevalence: TF-CBT $\frac{34}{45}=76\%$ at $T1$ → $\frac{16}{45}=36\%$ at $T2$; eTAU $80\%$ → $36\%$.

• Any non-PTSD diagnosis: TF-CBT $47\%\to18\%$; eTAU $32\%\to22\%$.

• No between-group significance.

Clinical Significance (RCI)

• TSCC-PTS $T1\to T2$: TF-CBT $45\%$ reliable positive change vs. eTAU $28\%$.

• SDQ total $T1\to T2$: TF-CBT $22\%$ vs. eTAU $25\%$ improved.

• At $T3$ patterns converged; <$50\%$ achieved reliable improvement overall; deterioration rates low (≤ $6\%$).

eTAU Sub-Group Exploration

• Structured (EMDR/CBT) vs. Unspecified eTAU:

  • Greater PTSD symptom reduction (K-SADS) $F=4.18, p=.047, d=0.60$.

  • Greater self-reported psychiatric symptom reduction (TSCC-PSY) $F=6.28, p=.017, d=0.80$.

  • Greater SDQ parent-rated improvement $F=4.14, p=.048, d=0.62$.

• Suggests structure/manualisation matters within TAU.

Discussion & Interpretation

• Contrary to previous RCTs, TF-CBT did not outperform eTAU.

• Possible explanations:

  • Sample complexity: high rates of multi-trauma, double exposure, probable ICD-$11$ Complex PTSD → may need longer/modified protocols.

  • Delivery factors: TF-CBT therapists instructed to follow standard protocol; $24\%$ required >$20$ sessions, but some may still have been prematurely terminated.

  • Comparison treatment was "enhanced" (structured, supervised, high therapist allegiance) incl. >50\% manualised EMDR/CBT, shrinking expected effect-size gap.

  • Real-world CAMHS setting often yields smaller EBT–TAU differences than university trials.

• Cost-effectiveness caveat: TF-CBT fewer sessions but each $80–90$ min versus eTAU $45–50$ min.

Strengths

• First independent RCT of full-protocol TF-CBT with IPV/abuse-exposed youths in Europe.

• Real-world clinic, socio-economically disadvantaged catchment, multi-traumatised sample – high external validity.

• Balanced design controlling supervision, caseload, therapist allegiance.

• Multiple informants (child, caregiver, clinician) & structured diagnostic interview.

• ITT, RCI, subgroup analyses.

Limitations

• Small sample → risk Type-II errors; effect sizes modest.

• eTAU heterogeneity; subgroup cell sizes too small for full three-way comparison.

• Assessors not blind to time-point or treatment, possible halo effect (mitigated by multi-informant convergence).

• $34\%$ attrition by $T3$; limited RCI calculable for all participants.

Implications for Practice

• Psychotherapy (manualised or structured TAU) can benefit multi-traumatised youths even amidst ongoing adversity; clinicians should not delay treatment awaiting environmental “stabilisation.”

• Need for flexibility/extended sessions and practice-based adaptations when applying TF-CBT to complex PTSD presentations.

• Monitoring individual progress over time and adding booster or alternative modules crucial; fewer than half achieve reliable change.

• Clinicians should favour structured, evidence-grounded components in usual care to maximise outcomes.

Implications for Research & Policy

• Future trials should:

  • Include complex trauma samples, larger Ns, dismantle treatment components, and detail TAU content.

  • Examine cost-effectiveness considering session length.

  • Investigate therapist, family, and contextual moderators of response (“what works for whom”).

• Active control conditions must receive equivalent supervision/training to ensure ecological validity of EBT superiority claims.

Funding & Declarations

• Funders: Swedish Research Council for Health, Working Life and Welfare; Children’s Welfare Association; Mayflower Charity Foundation; Swedish Crime Victim Compensation & Support Authority.

• Open access under CC BY-NC-ND $4.0$. Authors report no competing interests.