Jalling 2016b.

Methods	Study design: RCT at the level of the family Intervention arm(s): Comet Comparator arm(s): ParentSteps, Control Sample size calculation performed: yes Subgroups prespecified: unclear if it was prespecified although mentioned in Methods Subgroup analyses: adolescents with poor functioning at baseline; adolescents 12 to 17 years old only for comparability with other studies Start date: waves of recruitment/cohort: between 23 September 2008 and 19 October 2010 Duration of follow‐up: 6 months Number of follow‐ups: 1 Follow‐up time points: 6 months ICC (if reported): N/A
Participants	Number of schools randomised: N/A Number of participants randomised (total and by arm): 271 families (Comet: 99, ParentSteps: 86, control: 86) Age (range or mean (SD)) or grade at the start: Comet: 14.6 years (1.67), ParentSteps: 14.5 years (1.63), control: 14.7 years (1.89) Gender: Comet: 38.4% female, ParentSteps: 45.7% female, control: 58.0% female Ethnicity: foreign‐born mother: Comet 19.8%; ParentSteps 20.0%; control 20.9% SES: parent is employed: Comet 86.4%; ParentSteps 84.5%; control 84.1%. Parent has university degree: Comet 28.4%; ParentSteps 17.1%; control 28.0% Inclusion criteria: eligible participants were parents/caregivers and their adolescent children 12 to 18 years old who were at risk of consolidating antisocial behaviour. When screening parents for participation, antisocial risk behaviour in adolescence was indicated by single‐item descriptions: delinquent behaviour; bullying; repeated conflicts regarding family rules; any use of alcohol, tobacco, and/or drugs; or excessive computer use. Participation also necessitated living in 1 of the 5 participating municipalities, and the adolescent had to live at least part‐time with the participating parent or caregiver. Exclusion criteria: adolescent's ongoing psychotherapy, treatment for alcohol or drug use, out‐of‐home placement, parents’ participation in another parent programme
Interventions	Randomisation before or after baseline survey: before Duration of the intervention (excluding follow‐up): Comet: 9 weeks, ParentSteps: 6 weeks Description of the intervention Comet Consisted of 9 parent sessions and an optional booster session, each lasting 2 to 2.5 hours. It was delivered by certified group leaders (all are social workers who had 6 days of training and 2 booster training sessions). Principal programme components are rehearsals of the use of reinforcement principles (e.g. encouragement and praise, ignoring minor problems) through role‐play and home assignments whereby parents practise and develop these principles in their daily lives. Parents are instructed to keep a diary to document their interactions with their adolescent, and home assignments are followed up in subsequent sessions. Video vignettes are used in each session to enhance learning. Examples of themes covered include taking initiative for spending time together with the adolescent, dealing with rejection, performing basic interactional (behavioural) analysis, providing positive communication and encouragement, solving problems, and defining rules and consequences. ParentSteps Consisted of 6 parent sessions, each lasting 1.5 to 2 hours. ParentSteps is delivered by certified group leaders (all are social workers who had 1 day of training) using video vignettes, group discussions, and home assignments. Themes for the 6 sessions and home assignments are Love and limits; Encouragement and consequences; Risks and protection; Stress, fights, and different points of view; Youth, parents, and alcohol; and Youth, parents, and drugs. Brief description of the theoretical model: Comet is based on operant learning and social learning principles. ParentSteps is based on the resilience model. Description of the comparator: control (wait‐list): after the 6‐month follow‐up, parents in the control group were offered Comet 12–18 or ParentSteps in accordance with their preferences.
Outcomes	Primary: adolescent externalising behaviour, including antisocial behaviour, delinquency, alcohol use, and drug use Secondary: adolescent psychosocial dysfunction
Setting	Country: Sweden; State: Huddinge, Solna, Sundbyberg, Nacka, and Stockholm Setting: families (home) Focus: parents/caregivers and their adolescent children who were at risk of consolidating antisocial behaviour
Process measures	Process data reported: yes Method (qualitative or quantitative): quantitative Description Reach of the intervention: of the 6 total sessions, parents in ParentSteps attended a mean of 4.7 sessions (SD 1.44). Of the 9 total Comet 12–18 sessions, only every second session was assessed and mean attendance for 4.5 sessions was 3.6 (SD 1.07). With dropout defined as absence during the 2 last sessions, the dropout rate in ParentSteps was 14.1%. It was impossible to estimate the dropout rate in Comet 12–18 owing to lack of data from the 2 last sessions. Fidelity/Adherence to the intervention: Only social workers who were certified as programme deliverers and were active as group leaders were involved in the trial. Comet 12–18 leaders reported that 78% of the 73 sections in the manual were fulfilled ‘‘to a full extent’’, 17% ‘‘to a greater extent’’, 3% ‘‘to at least half’’, 1% ‘‘to a lesser extent’’, and 1% ‘‘not at all’’. ParentSteps leaders’ self‐assessments showed that 83% had fulfilled the manual sections ‘‘to a full extent’’, 13% ‘‘to a greater extent’’, 2% ‘‘to at least half’’, 1% ‘‘to a lesser extent’’, and 1% ‘‘not at all’’. At the 6‐month follow‐up, 32 parents reported that they themselves or their adolescent child had participated in another intervention targeting the adolescent such as seeing a school counsellor or a therapist at a child and adolescent psychiatry unit. These were equally distributed across groups (Comet 12–18 14.1%, ParentSteps 14.3%, control 12.2%), and they were kept for analysis.
Statistics	Sample size: eligible N = 605 families; enrolled N = 271 families (45%); Comet: 99, ParentSteps: 86, control: 86 Unit of randomisation: families Unit of analysis: individual Method to promote equivalence between groups: not reported Statistical models: analyses of normally distributed outcome data were performed with the general linear model (GLM) repeated measures ANOVA, and skewed data on the SRD were first log‐transformed. Baseline differences adjustment: gender and age were added to the ANOVA models as secondary explanatory factors. Repeated measures methods in analysis: N/A. Only 1 follow‐up
Notes	Equity: number of adolescents who completed baseline measures (total and by arm): 237 (Comet: 86, ParentSteps: 70, control: 81) Funding: not reported Randomisation method, e.g. block, stratification, computer: the randomisation sequence was generated by a research assistant who drew 1 of 3 folded opaque pieces of paper from a bowl. The paper was then put back in the bowl for the next family to be randomised. Clustering accounted for in sample size calculation (if relevant): N/A Cluster randomisation methods to account for clustering in analysis: N/A
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	The randomisation sequence was generated by a research assistant who drew 1 of 3 folded opaque pieces of paper from a bowl. Only 3 lots
Allocation concealment (selection bias)	High risk	Only 3 lots ‐ highly predictable
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding was not possible given the nature of this intervention.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Not reported; self‐reported outcomes only
Incomplete outcome data (attrition bias) All outcomes	Low risk	Low attrition and statistical analyses accounted for missing data.
Selective reporting (reporting bias)	Low risk	No published protocol, but all specified outcomes were reported
Other bias	Unclear risk	The number of conducted parent groups differed between the 2 programmes because of the limited capacity of social services to include trial parents in their regular Comet 12–18 group. Some groups included only 2 trial parents; in 2 groups, the whole group comprised parents participating in the trial. In ParentSteps, almost all parents in each group participated in the trial.