Nikles 2006.

Study characteristics
Methods	Cross‐over trial with 2 interventions SA‐MPH and SA‐dexamphetamine) Placebo Phases: each trial included 3 pairs of treatment periods (n‐of‐1). Each pair contained the stimulant and the comparator stimulant or placebo. The child’s doctor individualised the dosing, and the order of drugs was randomly assigned within each pair. 2 treatment periods were included in a school week, and 2 days per treatment period (not including Wednesdays and weekends)
Participants	Number of participants screened: 108 (85 boys, 21 girls). 2 boys repeated n‐of‐1 trials; 22 (20%) trials were not completed Number of participants included: 86. (66 boys, 18 girls) Number of participants followed up: 86 Number of withdrawals: 10 (12%) trials were not completed Diagnosis of ADHD: DSM‐IV (subgroups not stated) Age: median 10 years (range 5‐16) IQ: not stated MPH‐naive: 36 were taking MPH, 47 dexamphetamine, 3 unknown pre‐trial medication Ethnicity: all were white (Nikles 2007 in Nikles 2006) Country: Australia Setting: outpatient clinic Comorbidity: not stated Comedication: not stated Other sociodemographics: caregiver's occupation: full‐time work 21%, part‐time or casual work 33%, unemployed or retired 5%, unpaid homemakers 25%, other 10%, unknown 5% Inclusion criteria Clinical diagnosis of ADHD according to DSM‐IV criteria Stable on an apparently optimal dose of stimulant Informed consent from parent and school teacher (children 12 years of age provided assent) 5‐16 years of age Uncertainty about treatment effectiveness Exclusion criteria No teachers available and willing to provide observations
Interventions	Participants were randomly assigned to 1 of 3 possible drug condition orders of individual doses of MPH, dexamphetamine and placebo Mean MPH dosage: not stated Administration schedule and time points: different for different individuals, as this is a series of n‐of‐1 trials Duration of each medication condition: 2 days each of MPH and placebo/week Washout before trial initiation: 40 h (from 4:00 pm Tuesday to 8:00 am Thursday) and 64 h (4:00 pm Friday to 8:00 am Monday) Titration period: not stated, but all were stabilised on an "optimal" dose of stimulant, initiated before randomisation Treatment compliance: not stated
Outcomes	ADHD symptoms First 41 participants used at the end of each treatment period Conners’ Teacher and Parent Self Reported ‐ Revised, short form Conners‐Wells Adolescent Rating Scales From 42nd participant onwards used at the end of each treatment period Changed to ADHD du Paul Rating Scale IV parent and teacher questionnaires, "because they were less expensive but adequately reliable and valid for monitoring response to treatment" Conners‐Wells Adolescent Rating Scales Non‐serious AEs Assessment of AEs by parent and teacher during and at the end of each treatment period
Notes	Sample calculation: irrelevant Ethics approval: yes Key conclusion of trial authors ADHD n‐of‐1 trials can be implemented successfully by mail and telephone communication. This type of trial can be valuable in clarifying treatment effect when it is uncertain, and in this series, treatments had a noticeable impact on short‐term management Comments from review authors This study was more about the use of n‐of‐1 trials in determining optimal drug treatment in cases for which this was uncertain Nikles 2007 (in Nikles 2006) does not add any relevant outcome data Exclusion of MPH non‐responders/children who have previously experienced AEs while taking MPH before randomisation: no Any withdrawals due to AEs: yes (n = 1, insomnia and depression) Funding source: The General Practice Evaluation Program, the Department of Health and Aged Care, Queensland Medical Laboratory, and the Royal Australian College of General Practitioners
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"The order of drugs was randomly assigned within each pair"; "There were three pairs of treatment periods, with the order of drugs randomly assigned by a computer‐generated randomisation schedule within each"
Allocation concealment (selection bias)	Low risk	See above
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"Patients, parents, doctors, and the research assistant were all blinded to medication order"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"A hospital pharmacy encapsulated the medication (crushing of tablets and production of identical capsules containing either medication or placebo)"; "Active medication was encapsulated and identical placebo capsules were produced by a hospital pharmacy"
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Not stated Selection bias (e.g. titration after randomisation → exclusion): no
Selective reporting (reporting bias)	Low risk	All outcomes are reported