Steele 2019.

Study characteristics
Methods	Study design: individual RCT (parallel groups) Study aim/hypothesis: it was hypothesised that participation in GABI (Group Attachment Based Intervention) would lead to significant improvements in the mother‐child relationship compared to the treatment as usual group, STEP (Systematic Training for Effective Parenting) Study setting: The Bronx, New York City, USA. Data collection between 2012 and 2017 Trial registration number: NCT01641744
Participants	Inclusion criteria: mothers were biological parents and had custody of an infant up to 36 months old. Mothers were identified as being at risk of maltreating their children, due to their own history of maltreatment, social isolation or having lost custody of a child in the past Exclusion criteria: inability to provide informed consent due to mental illness or cognitive impairment and lack of fluency in English N referred and randomised: number referred not reported, 228 randomised GABI: 117 STEP: 111 N lost to follow‐up: 150 (66%) N analysed GABI: 43 STEP: 35 Childhood maltreatment and/or complex trauma status: childhood maltreatment Childhood maltreatment and/or trauma assessment: Adverse Childhood Experiences Questionnaire, assessing 5 forms of abuse (physical, psychological, sexual abuse, physical neglect and emotional neglect) and 5 forms of household dysfunction (parent mentally ill, incarcerated, drug addicted, domestic violence and separation/divorce). 77% of parents experienced 4 or more adverse childhood experiences (high exposure); mean adverse childhood events was 5.7 events Parenting stage: > 7 weeks postpartum (up to 5 years) Recruitment setting: Pediatrics, Child Welfare and Court Systems Baseline characteristics Mean parent age: not reported Mean child age: 15.9 months Parent gender: female participants Parent co‐morbidity: not reported Progress+ coding: low SES; ethnic minority population
Interventions	Intervention ‐ Group Attachment‐Based Intervention (GABI) Category: parenting, parent‐child or relationship focused interventions (parent‐child interventions) Description: a multifamily, group‐based, maltreatment prevention intervention consisting of a distinct set of attachment‐relevant features Mode of delivery: face‐to‐face, group Dose: more than 10 hours Length: other ‐ 26 weeks Frequency: 3 times a week Protocol: yes Provider: not reported Training: 2‐day didactic training followed by 6 months of co‐leadership and supervision Implementation fidelity: competence was monitored monthly via fidelity checklists developed and applied in both the STEP and GABI treatment groups. 80% or better fidelity ratings were consistently achieved Treatment adherence: retention from allocation to end of treatment was 37% Comparator ‐ Systematic Training for Effective Parenting (STEP) Category: inactive comparator (treatment as usual) Description: a treatment model consisting of anger management, psycho‐education, the distinction between discipline and punishment, and role‐playing adaptive parenting strategies Mode of delivery: face‐to‐face Dose: not reported Length: 7 to 12 weeks Frequency: not reported Protocol: yes Provider: single practitioner (STEP therapist) Training: clinicians delivering STEP were trained to reliably follow the STEP manual Implementation fidelity: competence was monitored monthly via fidelity checklists developed and applied in both the STEP and GABI treatment groups. 80% or better fidelity ratings were consistently achieved Uptake of standard care: retention from allocation to end of treatment was 32%
Outcomes	Assessment time point(s) Post‐intervention (< 3 months post‐intervention) Note: Assessments were also conducted at baseline; however, the results from this time point are not reported Primary outcome(s) Parent‐child relationship Dyadic reciprocity Domain: parenting capacity Measure: Coding Interactive Behaviour System Score range: not applicable Direction of effect: higher scores = better/less harm Dyadic constriction Domain: parenting capacity Measure: Coding Interactive Behaviour System Score range: not applicable Direction of effect: higher scores = worse/more harm Parenting skills Maternal supportive presence Domain: parenting capacity Measure: Coding Interactive Behaviour System Score range: not applicable Direction of effect: higher scores = better/less harm Maternal hostility Domain: parenting capacity Measure: Coding Interactive Behaviour System Score range: not applicable Direction of effect: higher scores = worse/more harm Secondary outcome(s): Parent engagement Dropout Domain: parental intervention acceptability Measure: dropout for any reason between randomisation and post‐intervention Score range: not applicable Direction of effect: higher events = more dropout Adverse outcome(s): none specified
Notes	Comment(s): unpublished data requested for subsample with high exposure (for or more adverse childhood events), but authors did not have capacity to send the data. No unpublished data reported in meta‐analyses. As 77% of parents experienced 4 or more adverse childhood experiences (high exposure), the mean adverse childhood events was high (5.7 events) and other traumas and adversity were extremely high, data from the whole sample was included Funding source: Health Resources and Services Administration Grant Conflicts of interest: the authors declared that there were no conflicts of interest