Bhat 2020.

Study characteristics
Methods	A 2‐week cross‐over trial with 2 arms: MPH 0.5 mg/kg/d placebo Phases: 2 (baseline week and week 2 )
Participants	Number of participants screened: 866 Number of participants included: 676 Number of participants followed‐up: 670 completed the 3 weeks. Maximum analysed are 575/573 (444 boys, 129 girls) Number of withdrawals: only information on 6 of 676 participants available Diagnosis of ADHD: DSM‐IV (percentage for 540 participants: 53.3% combined, 9.6% hyperactive‐impulsive and 37.0% inattentive type) Age: mean 9.03 (SD 1.81, 6‐12 years) (information for 575 participants) IQ: separated by DRD3 genotype group for 69 participants: 96.94 (13.27), for 236 participants: 97.07 (12.6), for 268 participants: 96.04 (13.42) (information for 573 participants) MPH‐naive: 222 (information for 573 participants): 62.9% (information for 540 participants). Ethnicity: 496 whites (information for 573 participants) Country: Canada Setting: outpatient Comorbidity: for 540 participants: CD (18.5%), ODD (41.4%), mood disorders (7.5%), anxiety disorders (42.6%) Comedication: not stated Additional sociodemographics: parental income: percentage low income (defined as ≤ CAD 30,000): 40.1%; marital status: percentage single mothers 43.2%; maternal education: percentage lower education 36.7%; maternal smoking during pregnancy: 38.3% Inclusion criteria ADHD Age 6‐12 Exclusion criteria Psychosis Tourette's syndrome IQ < 70 PDD
Interventions	Participants were randomly assigned to 1 of 2 possible orders of MPH (Ritalin) 0.5 mg/kg/d and placebo twice/d. 346 allocated to placebo first, 330 allocated to MPH first Mean medication dosage: all participants received a fixed dosage at 0.5 mg/kg/d Administration schedule: twice/d, 0.25 mg/kg, morning and noon Duration of each medication: 1 week of MPH, 1 week of placebo Washout before trial initiation: washout during 1‐week baseline assessment week Medication‐free period between interventions: none Titration period: none Treatment compliance: at the end of each week of treatment, the blister packs were collected and medication adherence was checked.
Outcomes	ADHD symptoms CGI ‐ Teacher Version rated once a week + 1 week before medication trial CGI ‐ Parent Version rated once a week (1 week before medication trial and on the Sunday after children were given their medication of the weekend) The Restricted Academic Situation Scale (RASS) "[...] administered before and 60 minutes after the administration of each treatment" (Bhat 2020 p 316) Serious AEs No important AEs or side effects were noted. Assesed "during each week of treatment" (Bhat 2020 p 315) Non‐serious AEs Barkley Side Effects Rating Scale, assessed "during each week of treatment" (Bhat 2020 p 315) The outcome reporting used for this review was taken from Bhat 2020, with data from 526 participants.
Notes	Sample calculation: no Ethics approval: yes. approved by the Research Ethics Board of the Douglas Mental Health University Institute in Montreal, Canada. Comments from trial authors "The within‐subject cross‐over design increased precision in determining the PR in comparison to the parallel‐group design used in previous trials." (Fageera 2017 p 8, secondary reference under Bhat 2020) "Another advantage is the measurement of an acute PR at one week in comparison with the duration of previous studies, lasting typically one month or more. The PR has been shown to peak in short‐duration studies thus a shorter trial design, especially with a medication like MPH, whose benefits are nearly immediate, may succeed in capturing the PR at its time of maximal expression." (Fageera 2017 p 8, secondary reference under Bhat 2020) "The cross‐over design carries a risk of bias due to carry‐over effects of the interventions across treatment periods. We acknowledge that the absence of a wash‐out period between the 2 treatment weeks magnifies this risk." (Fageera 2017 p 9, secondary reference under Bhat 2020) Key conclusion of trial authors "The first main finding in this trial is the higher and highly significant PR [Placebo response] with a larger effect size as assessed by parents compared to teachers. Several factors may account for this observation. First, the desire to improve is a known modulator of the PR (Price, Finniss, & Benedetti, 2008). The parents' desire for their children's behaviour to improve may exceed the desire of the teachers. Second, the capacity to generate expectations of improvement, an integral part of the PR (Price et al., 2008), may be more limited in teachers compared to parents. Alternatively, but not exclusively, it is possible that parents and teachers are attentive to different dimensions of the child behaviours that respond differentially to placebo in different environments (i.e. school and home). These observations imply that general statements regarding ‘the PR in ADHD’ might lack meaning without specifying who observed the response and in what context. Therefore, our results caution against the common practice in ADHD research of combining ratings from several sources into a single outcome variable (Newcorn et al., 2009)." (Fageera 2017 p 5‐6, secondary reference under Bhat 2020) "The second main finding from this study is the association between PR and several clinical and demographic characteristics of the child." (Fageera 2017 p 7, secondary reference under Bhat 2020) "Previous exposure to medication was also associated with smaller PR as assessed by parents. This observation could be explained by the fact that both parents and teachers have a prior knowledge of how the child responded to active medication, which in case it was poor, will lower their expectations of PR. In contrast, in children without prior exposure to active treatment, the expectations of observers with regard to response might be anchored in more general expectations of treatment response without adjustment to the child prior history of treatment response." (Fageera 2017 p 7, secondary reference under Bhat 2020) Comments from review authors This trial has many articles each reporting different outcomes/outcome analyses. Review authors are having trouble figuring out why only a part of the 676 included participants are included in the separate articles. Exclusion of MPH non‐responders/children who have previously experienced AEs while taking MPH: no Any withdrawals due to AEs: yes, unclear how many, but no fewer than 2 Funding source: this work was supported in part by a grant from the Fond de Recherche du Québec and the Canadian Institutes of Health Research. Weam Fageera is a recipient of a PhD scholarship from the Ministry of Education of Saudi Arabia. Email correspondence with trial authors: October 2013. We received some data from trial authors. We sent another email to ask for additional information but have not received a response. trial authors were contacted again through personal email in August, October and December 2021, for information regarding the flow of participants and first‐period data, but no answer was received.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Their order of administration was determined by counterbalanced random assignment, using a computer‐generated randomisation list prepared by a statistician not otherwise affiliated with the trial.
Allocation concealment (selection bias)	Low risk	Placebo and MPH were prepared individually in opaque gelatin capsules in weekly blister packs by a pharmacist not otherwise involved in the trial to maintain blind allocation of treatments
Blinding of participants and personnel (performance bias) All outcomes	Low risk	MPH and placebo were encapsulated into opaque gelatin capsules in weekly blister packs
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Both parents and teachers were aware that after 1 week of baseline observation (which served also as a washout period for children who were previously on medication), participants (in a blind order) received either 1 week of active medication (MPH 0.25 mg/kg twice/d) followed by 1 week of placebo or the reverse order to assess their response to medication in an unbiased fashion
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Flowchart in Naumova 2019 (secondary reference under Bhat 2020), claims 676 included and randomised participants, but the articles' outcomes are reported for a maximum of 575 participants with no information on withdrawals. Selection bias (e.g. titration after randomisation → exclusion of MPH non‐responders or placebo responders): no
Selective reporting (reporting bias)	Low risk	All protocol outcomes reported