TOSCA study.

Methods	2‐stage 9 week parallel group, double‐blind, randomised controlled trial of risperidone ('augmented' = active treatment) and placebo ('basic' = placebo) added to parent training and stimulant. Stage 1: 3 weeks, open‐label stimulant and parent training; Stage 2: 6 weeks of a double‐blinded, placebo‐controlled comparison of added risperidone versus placebo.
Participants	Setting: outpatients, multicentre (University clinics) Sample size: 168; 84 active treatment, 84 placebo Sex: 129 boys (77%) and 39 girls; 65 boys and 19 girls in the active arm and 64 boys and 20 girls in the placebo arm Age range: 6 to 12 years Mean age: active treatment 9.03 (SD = 2.05) years; placebo 8.75 (SD = 1.98) years IQ range: not reported. Mean IQ 97.1 (SD = 14.1) Inclusion criteria: DSM‐IV disruptive behaviour disorder diagnosis (conduct disorder or oppositional defiant disorder); DSM‐IV diagnosis of ADHD (any subtype); evidence of serious physical aggression as rated on the Overt Aggression Scale–M (score greater or equal to 3 on assaults against other people, objects, or self); and evidence of seriously disruptive behaviour as determined by a parent or guardian rating of at least 27 (90th percentile) on the Nisonger Child Behaviour Rating Form ‒ Typical IQ Version D‐Total (Conduct Disorder and Oppositional Defiant Disorder subscales combined). In addition, a Clinical Global Impressions ‒ Severity scale score of at least 4 (“moderately ill” or higher) for aggression was required by blinded clinicians. Participants needed to be free of psychotropic medicines for 2 weeks for most drugs (such as most antidepressants, a‐agonists, b‐blockers, anxiolytics, mood stabilizers, oral antipsychotics, and antihistamines) and 4 weeks for depot antipsychotics or fluoxetine. This rule was occasionally relaxed (to as few as 3 to 7 days) for extreme cases who could not tolerate being unmedicated the full time, as approved by the cross‐site steering committee. Diagnosis: 124 (74%) oppositional defiant disorder; 44 (265) conduct disorder Comorbidity: All have ADHD Sample characteristics: 53% White European ancestry; living with working parents (52% mothers, 53% fathers); some college education (66% mothers; 35% fathers); family incomes of USD 40,000 or less a year (57%) Withdrawn/dropouts: 22 participants dropped out before active treatment was introduced or because they were deemed not to need it (i.e. they were clinical responders to methylphenidate alone). In the active arm, 11 participants dropped out in the first 3 weeks (5 were clinical responders to methylphenidate alone) and in the placebo arm 3 dropped out in the first 3 weeks (3 were clinical responders to methylphenidate alone). Other Interventions: From baseline, all participants received parent training (PT) which consisted of 9 sessions with up to 2 optional booster sessions (using empirically established programme for children i.e. the Community Parent Education Program (COPE). COPE focuses on strategies for managing impulsive behaviour, including reactive aggression. From baseline through week 3, primary clinicians openly titrated stimulant medication to optimal effect, usually in the form of Osmotic Release Oral System (OROS) methylphenidate (Concerta). For smaller children (< 25 kg), dosage was titrated clinically using the following daily doses: 18 mg (7 days), 36 mg, and 54 mg (maximum maintenance dose). For larger children (> 25 kg), dosage was increased every 3 to 4 days using the following daily doses: 18, 36, 54, and 72 mg. Subjects unable to tolerate Concerta or unable to swallow pills were offered an alternative (at comparable doses) from the following, of which the capsule contents could be sprinkled onto food: mixed amphetamine salts (Adderall), dextromethylphenidate extended release (Focalin XR), or lis‐dexamphetamine dimesylate (Vyvanse). At the mean week 9 (endpoint) in the active group the methylphenidate dose was 46.1 ± 16.8 mg/day and in the placebo group it was 44.8 ± 14.6 mg/day.
Interventions	Intended dose: if residual symptoms remained, then randomised placebo or risperidone was added to treatment at weeks 4 through 6. For children weighing less than 25 kg, risperidone was dosed at 0.5 to 2.5 mg/day; for children heavier than 25 kg, dosing ranged from 0.5 to 3.5 mg/day. The risperidone titration schemes allowed for dose increases every 3 to 7 days, following a schedule that specified maximum dose increases over 29 days of titration; doses could always be held constant or decreased if a satisfactory clinical response occurred or if indicated by adverse event. Mean dose at endpoint: at week 9 (endpoint), in the active group the mean risperidone dose was 1.7 (± 0.75) mg/day and in the placebo group it was 1.9 (± 0.72) mg/day.
Outcomes	Primary outcomes: Nisonger Child Behavior Rating Form ‐ Typical IQ version D‐Total Secondary outcomes: Positive Social, Overly sensitive ADHD, Withdrawn‐dysphoric subscales of the Nisonger Child Behavior Rating Form, Antisocial Behavior Scale consisting of Proactive and Reactive subscales; Clinical Global Impression ‒ Improvement and Clinical Global Impression ‒ Severity scales Follow‐up interval: 9 weeks (risperidone and placebo were introduced in week 4)
Notes	Funding/support: the trial was funded by National Institute of Mental Health (NIMH). From correspondence (Loy 2016a [pers comm]), the authors originally requested a pharmaceutical company to sponsor the study medication. Due to lack of agreement over certain issues, this fell through except for 1 trial site. Subsequently, the authors obtained funding from the NIMH to purchase the study medications (both stimulant and risperidone and placebo) from an independent pharmacy. Only 1 study site, from which approximately 50 participants were recruited, received medication sponsored by the pharmaceutical company. It is possible that the centre used pharmaceutical‐supplied medication for 1 or 2 participants only. There were 168 participants in the trial. The lead author's view was that the pharmaceutical involvement was virtually nonexistent. Disclosures: the lead author has "received research contracts, consulted with, or served on advisory boards" of various pharmaceutical companies as listed on p 59. Full disclosure for other co‐authors is available on p 59. Long‐term outcomes to be published.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Stratified block randomisation was used: blocks of size 2 were allocated to each stratum formed by the cross‐classification of the levels of the stratifying factors: site and DSM‐IV diagnosis (conduct disorder versus no conduct disorder) (Loy 2016b [pers comm]).
Allocation concealment (selection bias)	Low risk	Randomisation assignment was completed through a secured website. The unblinded medication dispenser entered the appropriate information into the website and an email with the participant’s treatment assignment was sent to the dispenser and to the statistician at the time. Randomisation assignment for each participant was printed and sealed in an envelope for study emergency only (Loy 2016b [pers comm]).
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Each child was followed by a primary clinician and a blinded clinician. Risperidone and placebo were absolutely identical. Different colour capsules were used to signify different doses (Loy 2016b [pers comm]). No details published about the details of the blinding procedure of the participants (children and parents) (Loy 2016b [pers comm]).
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Each child was followed by a primary clinician and a blinded clinician. The primary clinician assessed the participants for adverse events and titrated dosage, whereas the blinded clinician assessed the children for clinical improvement (i.e. was responsible for monitoring therapeutic effects on Nisonger Child Behaviour Rating Form ‒ Typical IQ Version, Clinical Global Impression‐Improvement (CGI‐Improvement), Clinical Global Impression‐Severity (CGI‐Severity). Parents and participating children were told not to discuss side effects or (when the blind was broken) medication assignment with the blinded clinicians. Blinded clinicians were barred from asking about side effects, appetite, sleep patterns, or seeing any of the completed Adverse Effect forms. There was a Medication Knowledge Form to determine both the parents' and blinded clinicians' knowledge regarding the identity of risperidone recipients. Blinded clinicians were never unblinded until the entire study was completed and all study data were locked (Loy 2016b [pers comm]).
Incomplete outcome data (attrition bias) All outcomes	Low risk	Detailed reasons for participant attrition are displayed in table S1 and S2 (p 60 e1).
Selective reporting (reporting bias)	Unclear risk	Published results are based on completers‐only sample (i.e. no imputation of missing data was carried out in the published manuscript). Sensitivity analysis was said to be conducted (Loy 2016b [pers comm]) but not published or available. The reported outcome measures are consistent with those listed on the Clinical Trials registry (NCT00796302).
Other bias	Low risk	Completers analysis published only.