Pfiffner 1997.

Methods	Design: RCT
Participants	Country: USA Setting: university‐based behavioural paediatric clinic specialising in ADHD and related disorders. Participants were recruited from newspaper advertisement and from consecutive referrals. Sample size calculation: not reported Sample size: 27 children Sex: 19 (70%) = boys, eight (30%) = girls Age: range = eight‐10 years Ethnicity: all Caucasian except for one boy, who was African American Socioeconomic status: middle‐ to upper‐middle class; two children were from single‐parent families IQ: not reported ADHD diagnosis: following DSM‐III‐R, 25 children met criteria for ADHD and two met criteria for undifferentiated ADHD ADHD medication: n = 12 (44%) received stimulant medication Comorbidity: ODD = 19, CD = 3, separation anxiety disorder = four, overanxious disorder = five, dysthymic disorder = two Medication for comorbid disorder: no information Inclusion criteria: diagnosis of ADHD or undifferentiated ADHD following DSM‐III‐R criteria mean score at or above 1.5 on at least one of the parent‐completed subscales assessing ADHD. behaviour from the CLAM rating scale or the SNAP‐R, a T score of at least 60 on the Attention Problem subscale of the CBCL Exclusion criteria: not reported. We attempted to get this information from the study investigators but have not succeeded in this attempt. Baseline characteristics: at pretreatment, no significant difference in age, socioeconomic status, medication status and number of symptoms of ADHD, and comorbid disorders or on parent and teacher ratings of social skills and behaviour
Interventions	18 participants allocated to one of three groups (see below). Both treatment groups had a protocol and were led by psychologist and the same two therapists taught in the childrens' groups. The two treatment groups attended eight group sessions. Children in both treatment groups received 90‐minute group sessions during consecutive weeks. Assessment was at pre‐ and post‐treatment and follow‐up (3‐4 months post‐treatment). Group one (n = 9): social skills training (SST) for children, which covered six themes/modules: 1) good sportsmanship; 2) accepting consequences; 3) assertiveness; 4) ignoring provocations; 5) problem‐solving; and 6) recognising and dealing with feelings. Children were assigned homework to practice at home. The children received points for following the rules of the groups, participating, and attending the sessions. The points could be exchanged for child‐selected games and activities during the last 10 minutes for each group. Group two (n = 9): SST for children with parent‐mediated generalisation. Parents were used as a primary vehicle to programme generalisation of the social skills learned in the SST groups to home and school settings. The parents went through the same group themes or agendas as the children did. The parents met with their children's teacher and gave the teacher a template for the scorecard, also called the daily report card. The teacher scored the child on a four‐point scale and parents rewarded the child when the child scored high on the scale. Group three (n = 9): waiting list Attendance: two families each missed one session
Outcomes	Primary outcomes Social skills: SSRS, parent‐, teacher and parent‐rated; Test of Social Skill Knowledge, observer‐rated (interviewers, scored by blinded raters) Secondary outcomes Satisfaction with treatment: Consumer Satisfaction Questionnaire, parent‐rated Outcome assessment: post‐intervention and follow‐up three to four months after post‐intervention assessment
Notes	Study ID: not reported Sponsorship source: not reported Year conducted: not stated Duration of the study: 8 months Comments: none Lead author: Linda J Pfiffner Institution: Department of Psychiatry, The University of Chicago Email: not reported Address: 5841 South Maryland Avenue, MC 3077, Chicago, Illinois 60637
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: no information given in the article. We requested clarification from the study investigators and they reported in an email on 26 May 2011 that it was not possible to obtain these data at that time (Pfiffner 2011 [pers comm]).
Allocation concealment (selection bias)	Unclear risk	Comment: no information given in the article. We requested clarification from the study investigators and they reported in an email 26 May 2011 that it was not possible to find these data at that time (Pfiffner 2011 [pers comm]).
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: blinding on at least one of this study's primary outcomes; no blinding for the rest of the outcomes
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: three participants started medication after post‐intervention assessment but before follow‐up assessment. Individual scores for these participants were replaced by the grand mean of all other participants at follow‐up to avoid possible confounds associated with the medication treatment.
Selective reporting (reporting bias)	High risk	Comment: the author informed us in an email (Pfiffner 2011 [pers comm] that the CLAM and SNAP were used post‐treatment, but were not reported in the article. We were not able to get the data because they had been lost over time.
Vested interest bias	Low risk	Comment: no other apparent biases, no previous research on the topic
Other sources of bias?	High risk	Comment: Teachers were paid $10 for post‐intervention assessment and $25 for follow‐up assessment and families were paid $12 for follow‐up assessment.