Tutty 2003.

Methods	Design: RCT
Participants	Country: USA Setting: outpatient clinic in Washington Sample size calculation: not reported Sample size: 100 children Sex: 75 (75%) = boys, 25 (25%) = girls Age: range = five‐12 years Ethnicity: (group one: white = 49 (83%), African‐American = 4 (7%), Asian = 5 (8%), Hispanic = 1 (2%); group two: white = 38 (93%), African‐American = 2 (5%), Asian = 1 (2%), Hispanic = 0 (0%)) Socioeconomic status: parent education (group one: grades nine‐12 = nine (15%), grades 13‐16 = 18 (64%), < grade 16 = 12 (20%); group two: grades nine‐12 = 11 (27%), grades 13‐16 = 25 (61%), < Grade 16 = five (12%)) IQ: not reported ADHD diagnosis: (group one: inattentive subtype = 25 (42%), combined subtype = 34 (58%); group two: inattentive subtype = 16 (39%), combined subtype = 25 (61%)) ADHD medication: a new prescription for stimulant medication had been filled in for all children Comorbidity: not reported, but range of comorbid difficulties (see below) were part of exclusion criteria Medications for comorbid disorders: allowed but not stated if it was balanced between groups Inclusion criteria: diagnosis of ADHD (DSM‐IV), and filling in new prescription for stimulant medication (i.e. no stimulant medication use in past 120 days) Exclusion criteria: comorbid CD, ODD, Tourette syndrome, affective disorder, active alcohol or other substance abuse during previously 90 days or chronic mental illness, if children had been enrolled in a child social skills training at the involved centre in the past Baseline characteristics: mean baseline parented ADHD symptom scores were more symptomatic for the intervention group than for the control group, as well as the use of parent discipline practice. These between‐groups differences were adjusted before follow‐up analysis.
Interventions	100 participants allocated to one of two groups Group one (n = 59): child social skills training and parent training plus medical treatment. Medication regimen was stabilised for the participants during the first three to four weeks before the social skills training commenced. The social skills training consisted of eight 50‐minute group sessions that were delivered during 8 consecutive weeks. Each session was based on a structured session‐by‐session agenda and focused on the topics: listening, skills, expression of feelings, anger management, self control, conflict resolution, friendship skills, and self‐esteem. During the sessions, the group was divided into child‐only and parent‐only groups. The children were further divided into child groups based on age to minimise age differences within the child group. Seven therapists participated in the study and each therapist had at least two years of direct experience treating this population in individual and group formats, possessed a master's degree in social work, counselling, or educational psychology, and had participated in three 2‐hour preparation sessions with the study coordinator and a senior therapist to review session content, itinerary, and clinical/research protocols before facilitating the social skills training. Group two (n = 41) : medical treatment alone Attendance: there was a 95% rate for completing all eight sessions in group one. Blinded follow‐up measures completed by 97% and 98% of parent or guardian participants at three and six months after enrolment, respectively. Follow‐up completion rates for teacher participants yielded 92% and 75% for three and six months after enrolment, respectively. Participants with missing data did not differ from participants with complete data sets across time or any clinical, functional, and demographic variables, according to the authors of the report. For the ADHD Rating Scale outcome, two children were lost to follow‐up. For the Child Attention Profile outcome, 24 children in total were lost to follow‐up (intervention = 16, control = eight).
Outcomes	Secondary outcomes Core ADHD symptoms: ADHD Rating Scale, parent‐rated (telephone interviews); Child Attention Profile, teacher‐rated (telephone interviews) Outcome assessment: at three (post‐intervention) and six months (follow‐up) after enrolment
Notes	Study ID: not reported Sponsorship source: the study was supported by the Group Health Cooperative (GHC)/Kaiser Permanente Community Foundation through a grant to the GHC, Center for Attention Deficit Disorders, located in Redmond, Washington Year conducted: not stated Duration of the study: 12 months Comments: there was a third outcome used in this study, but it is not relevant for this review, because it measured the parents' discipline practice. Lead author: Steve Tutty Institution: Group Health Cooperative, Center for Health Studies Email: tutty.s@ghc.org Address: 1730 Minor Avenue, Suite 1600, Seattle, WA 98101
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Comment: study used coin toss method performed by research assistant, which is an adequate method of randomly generating the sequence, according to the Cochrane Handbook of Systematic Reviews of Interventions
Allocation concealment (selection bias)	Unclear risk	Comment: no information reported. We requested clarification about the method of allocation concealment from the study investigators but received no information on this topic at the time of the original review.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: no blinding
Blinding of outcome assessment (detection bias) All outcomes	High risk	Comment: outcome assessment by telephone interviews of parents and teacher, which were performed by a blinded research assistant. The parents, however, were not blinded, which is not an adequate method.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: ITT method used
Selective reporting (reporting bias)	Low risk	Comment: all measures of interest reported. No protocol identified. Of the measures mentioned in the paper, all measures of interest for the review were reported.
Vested interest bias	Low risk	Comment: no apparent source of bias
Other sources of bias?	Unclear risk	Comment: co‐medication not specified