Jensen 1999 (MTA).

Study characteristics
Methods	14‐month multi‐centre, randomised, parallel clinical trial with 4 arms: medication management behavioural treatment combined treatment (medication management + behavioural treatment) community care (control group) Phases: for the 2 groups receiving medication, an initial 28‐day titration period was provided. This titration phase was carried out as a randomised, double‐blind, placebo‐controlled, cross‐over trial with daily switching of MPH doses (placebo, low, middle and high). Once delivery of randomly assigned treatments by trial staff stopped at 14 months, the trial became an observational trial in which participants and families were free to choose their own treatment, but in the context of availability and barriers to care existing in their communities. The following follow‐up assessments took place after completion of the RCT at 10 months' follow‐up (24 months after randomisation), 3‐year follow‐up, 8‐year follow‐up and 10‐year follow‐up In our review, we will compare combined behavioural treatment (RCT) according to our protocol and will look at the medication treatment group as a cohort (observational)
Participants	Number of participants screened: 4541 Number of participants included: 579 Number of participants randomly assigned to MPH + behavioural treatment (combined treatment): 145, MPH 144, behavioural treatment 144 Number of participants followed up: combined treatment 142; medication 136; behavioural treatment 141 Number of withdrawals/dropouts: combined treatment 3; medication 8; behavioural treatment 3 Demographic data for combined treatment, behavioural treatment and medication management Diagnosis of ADHD: DSM‐IV (combined (100%)) Age: mean 8.4 years (range 7‐9.9) IQ: mean 100.4 Sex: 346 boys, 87 girls MPH‐naive: 177 Ethnicity: white (60.3%), African American (20.6%), Hispanic (8.8%), other (10.4%) Country: USA and Canada Setting: outpatient clinic Comorbidity: anxiety disorder (35.1%), CD (14.1%), ODD (38.8%), affective disorder (3.5%), tic disorder (10.2%), mania/hypomania (3%) Comedication: not stated Other sociodemographics: 130 families on welfare. Population ranges widely in socioeconomic status. No significant differences in baseline demographics were noted between the 3 groups Inclusion criteria Boys and girls 7‐9. 9 years of age Grades 1 through 4 In residence with the same primary carer(s) for the past 6 months or longer Meeting DSM‐IV criteria for ADHD, combined type Exclusion criteria Child currently in hospital Child currently in another trial < 80 on all WISC‐III, and on Severe Impairment Battery (SIB) (bipolar disorder, psychosis or personality disorder) Chronic serious tics or Tourette's syndrome OCD serious enough to require separate treatment Neuroleptic medication in previous 6 months Major neurological or medical illness History of intolerance to trial medications Ongoing or previously unreported abuse Missed one‐4th of school days in previous 2 months Same classroom as child already in trial Parental stimulant abuse in previous 2 years Non‐English‐speaking primary carer Another child in same household in trial No telephone Suicidal or homicidal
Interventions	Participants were randomly assigned to medication management, behavioural treatment or combined treatment Mean MPH dosage during main trial: combined treatment 31.2 mg/d, medication management: 37.8 mg/d Administration schedule: 3 times/d ‐ breakfast, lunch and in the afternoon Duration of intervention: 14 months Titration period: 4 weeks. After randomisation. Participants randomised to receive medication or combined treatment underwent a 4‐week, double‐blind titration phase. Treatment compliance: monthly pill counts, intermittent saliva measurements to monitor intake of MPH and encouragement for families to make up missed visits. "The study achieved a high degree of adherence to protocol." NB! For participants not attaining an adequate response to MPH during titration, alternate medications were titrated openly in the following order until a satisfactory choice was found: dextroamphetamine, pemoline, imipramine and others, if necessary approved by a cross‐site panel. Thus, 256 participants successfully completed titration; of these, 198 of 256 participants were assigned to an individually titrated best dose of MPH, and 26 were titrated to dextroamphetamine. 32 were given no medication because of a robust placebo response
Outcomes	ADHD symptoms SNAP Inattention and Hyperactivity‐Impulsivity subscale, both parent and teacher: assessed at baseline and at 3, 9 and 14 months SNAP Oppositional Defiant Disorder subscale, both parent‐ and teacher‐rated: assessed at baseline and at 3, 9 and 14 months. Abikoff Classroom Observational System (ADHD and oppositional/aggressive symptoms): blind ratings by blind observers Serious AEs 6 of 11 reported severe side effects could have been due to non‐medication factors 3 deaths were recorded among ADHD participants during 10 years of observation: a suicide at age 14 (participant was taking MPH), a fatal car accident at age 17 (participant was the driver and was taking MPH) and sudden unexplained death at age 17 (participant was found dead in bed; no specific cause of death could be determined; he had been treated previously with MPH and had been off medication for > 1 year when he died) Non‐serious AEs Participants were provided up to 8 additional sessions when needed to address clinical emergencies or instances of possible trial attrition Pittsburgh Side Effects Rating Scale: monitored monthly, reviewed by the pharmacotherapist Internalising symptoms (anxiety and depression): measured with an internalising subscale from parent‐ and teacher‐completed Social Skills Responsive Scale, measured at baseline and at 3, 9 and 14 months Children’s self‐ratings on the Multidimensional Anxiety Scale for Children: assessed at baseline and at 3, 9 and 14 months For analysis of stimulant treatment duration in relation to substance use at 8‐year follow‐up, the primary outcome was the number of substances used in the past 6 months, to ensure that most stimulant treatment received would have preceded substance use. Component variables included the following: "drunk" once or more or drank alcohol ≥ 3‐4 times; ≥ 1 cigarettes/d in the past month (time frame exception specific to tobacco); marijuana ≥ 2 times; and any other illicit drug use or prescription medication misuse. Secondary analyses explored each class of substances separately. For analysis of stimulant treatment exposure over time in relation to substance use at 8‐year follow‐up, the primary outcome variable was substance use disorder in the past year for any substance (excluding tobacco). Secondary analyses explored alcohol and marijuana/other drug use disorders separately
Notes	Sample calculation: yes; power analysis, with 576 participants required Ethics approval: yes; approved by both local institutional review boards and NIH Office for Protection From Research Risk   Comments from trial authors Recruitment, screening and selection procedures aimed to collect a carefully diagnosed sample of impaired children with ADHD and a wide range of comorbid conditions and demographic characteristics representative of patients seen in clinical practice The design did not include a no‐treatment or placebo group More than 3‐4ths of participants given behavioural treatment were successfully maintained without medication throughout the trial. Consequently, it should not be concluded that behavioural treatment interventions did not work Combined treatment and medication management were clinically and statistically superior to behavioural treatment and community care in reducing children’s ADHD symptoms. (...) For other areas of function (oppositional/aggressive behaviours, internalising symptoms, social skills, parent‐child relations and academic achievement), few differences among our treatments were noted, and when found, were generally of smaller magnitude Significantly lower total daily dose of MPH in the combined treatment arm is noteworthy but was not unforeseen. The importance of this finding is unclear, and a rigorous test of the question would likely require a different design Key conclusions of trial authors For ADHD symptoms, our carefully crafted medication management was superior to behavioural treatment and to routine community care that included medication. Our combined treatment did not yield significantly greater benefits than medication management for core ADHD symptoms but may have provided modest advantages for non‐ADHD symptoms and positive functioning outcomes Comments from review authors Trial authors have written > 70 articles describing different outcomes and challenges of the trial. We have included only those found through our comprehensive literature search and others that we found relevant to include upon looking through article reference lists We have discussed whether to include this trial, as not all participants randomly assigned to medication (combined treatment and medication management group) received MPH. Those who did not have an adequate response to MPH were given other medication (e.g. dextroamphetamine, pemoline, imipramine) or no medication. Furthermore, some participants in the behavioural group were also medicated during the 14‐month randomisation phase. From all other studies in this review, we have included only receivers of pure MPH. Furthermore, lots of participants did not have an ADHD diagnosis at follow‐up assessment. At 8‐year follow‐up, only 30% of remaining participants still had a diagnosis of ADHD. However, we have chosen to use the data from this trial, as it is such a large and well‐known trial. All trial analyses will be included in the review as sensitivity analyses Regarding Molina 2013 (secondary reference under Jensen 1999 (MTA)) (substance use): we have included/asked for additional data from this trial, even though the medicated group received medication for a mean of only 2071.10 (SD 728.87) days of the 8 years the follow‐up took place The following articles from the trial have been assessed by only 1 review author: Pelham 2000, Carey 2000, Swanson and Hinshaw 2007, Galanter 2003, Hinshaw 1997, Molina 2013 (secondary references under Jensen 1999 (MTA)) Exclusion of MPH non‐responders/children who have previously experienced AEs while taking MPH before randomisation: yes, see exclusion criteria Any withdrawals due to AEs: 4 participants were removed during the lead‐in (titration period) because of prohibitive side effects: 1 child with buccal movements; another with skin picking; a third with depression, crying, sleep delay and appetite loss; and a 4th who was anorexic, listless and emotionally constricted Funding source: this trial was supported by several grants from the NIMH, Bethesda, Maryland Email correspondence with trial authors: January 2014‐June 2014. We sent several emails to the MTA group to request additional information. However, we were not able to obtain additional data. We did receive an email from Dr. Hinshaw confirming that the data on ADHD symptoms, parent‐rated, were wrong ‐ instead of a mean of 1.85, the correct mean was 0.85 for combined treatment after 14 months. We also received an email from Dr. Swanson in June 2014 stating that he would help with data collection. We wanted to conduct a reanalysis of data excluding those few participants not receiving MPH. Dr. Swanson provided several helpful comments on this and enclosed published articles, but we did not receive additional data, in part because of the time frame of this review.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation was done centrally by the NIMH Data Center, stratified by site in blocks of 16 (4 to each group). Stratified by 6 sites. Sealed, ordered envelopes were sent to sites for successive entries
Allocation concealment (selection bias)	Low risk	Sealed, ordered envelopes were sent to sites for successive entries. Treatment assignment was concealed until the family confirmed agreement to accept randomisation
Blinding of participants and personnel (performance bias) All outcomes	High risk	Treatment assignment was concealed until the family confirmed agreement to accept randomisation. After agreement on best dose, the blind was broken, and the agreed‐on dose (if not placebo) became the participant’s initial maintenance dose
Blinding of outcome assessment (detection bias) All outcomes	High risk	After agreement on best dose, the blind was broken, and the agreed‐on dose (if not placebo) became the participant's initial maintenance dose. However, for some outcome measures, 3 strategies were devised to enlist blinded raters and objective observations
Incomplete outcome data (attrition bias) All outcomes	High risk	ITT analyses. Despite high compliance, we checked whether compliance with assessments (i.e. missing data) could have changed our findings. We completed random‐effects regression analyses 2 ways: once with inclusion of all participants, and then including only participants who provided data over multiple time points during the trial. No differences emerged from these 2 sets of analyses. Of 289 participants randomly assigned to medication treatments, 33 (11%) did not finish titration. 17 refused medication and 1 moved away. 4 participants were removed during lead‐in because of prohibitive side effects: 1 child with buccal movements; another with skin picking; a 3rd with depression, crying, sleep delay, and appetite loss, and a 4th who was anorexic, listless, and emotionally constricted. Even though they did well in lead‐in, 7 additional participants stopped in the middle of titration because they could not follow titration procedures and 4 had excessive missing data and were not included. The remaining 256 participants (88.6%) successfully completed titration. Selection bias (e.g. titration after randomisation → exclusion of MPH non‐responders or placebo responders): yes, for participants not attaining an adequate response to MPH during titration, alternate medications were titrated openly in the following order until a satisfactory choice was found: dextroamphetamine, pemoline, imipramine and others, if necessary approved by a cross‐site panel. Thus, 256 participants successfully completed titration; of these, 198 of 289 participants were assigned to an individually titrated best dose of MPH, and 26 were titrated to dextroamphetamine. 32 were given no medication because of a robust placebo response
Selective reporting (reporting bias)	Low risk	No indication of selective reporting