Kerwin 2008.

Methods	DESIGN: randomised, parallel‐group, double‐blind, placebo‐controlled trial; in multiple centres
Participants	SYMPTOMATIC PARTICIPANTS RANDOMLY ASSIGNED: N = 206 INTERVENTION: ICS (budesonide 200 μg/d): 104 CONTROL: ICS (budesonide 800 μg/d): 102 WITHDRAWALS: reported AGE: mean (SD) years: INTERVENTION: ICS (budesonide 200 μg/d): 11.7 (2.8) CONTROL: ICS (budesonide 800 μg/d): 11.5 (2.9) GENDER: male N (%) INTERVENTION: ICS (budesonide 200 μg/d): 59 (56.7) CONTROL: ICS (budesonide 800 μg/d): 64 (62.7) ASTHMA SEVERITY: mild asthma ASTHMA DURATION: mean (SD) years INTERVENTION: ICS (budesonide 200 μg/d): 6.7 (3.7) CONTROL: ICS (budesonide 800 μg/d): 6.8 (3.9) MEAN (± SD) β2‐AGONIST USE (puffs/d): INTERVENTION: ICS (budesonide 200 μg/d): 0.5 (0.8) CONTROL: ICS (budesonide 800 μg/d): 0.3 (0.7) DOSE OF ICS AT STUDY ENTRY AND AT RUN‐IN: Participants continued their usual ICS therapies (if any) and added a once‐daily placebo treatment ATOPY (% of participants): not reported ELIGIBILITY CRITERIA Patients aged 6 to 17 years who had been diagnosed with asthma for ≥ 3 months Patients who had not previously been treated with an ICS or had been treated with a low‐dose of ICS (maintained at a constant dose level) for no longer than 30 days before visit 1 Patients who had a pre‐bronchodilator FEV1 of 75% to 90% (patients aged 6–11 years) or 60% to 90% (patients aged 12–17 years) of predicted Patients who met reversibility criteria (≥ 12%) Patients with a pre‐bronchodilator FEV1 between 91% and 95% of predicted normal were eligible if the ratio of FEV1 to forced vital capacity (FEV1/FVC) was < 0.80 EXCLUSION CRITERIA Severe asthma as judged by the investigator Life‐threatening asthma (including any prior asthma intubation, respiratory arrest or seizures as a result of exacerbation of asthma) ≥ 2 asthma‐related hospitalisations in the past year Use of systemic corticosteroids within 4 weeks of entry Use of other controller therapies (e.g. leukotriene modifiers [LTMs], long‐acting β2‐adrenergic agonists [LABAs]) within 2 weeks of entry Recent clinically relevant respiratory disease as judged by the investigator (e.g. chronic obstructive pulmonary disease) Acute asthma exacerbation, or other significant disease Use of an experimental drug or device within 30 days of entry Smoking Hypersensitivity to study products
Interventions	PROTOCOL DURATION Run‐in = 11 to 17 days Intervention = 12 weeks DEVICE: dry powder inhaler DOSE OF ICS INTERVENTION: budesonide 200 μg/d CONTROL: budesonide 800 μg/d CRITERIA FOR WITHDRAWAL FROM STUDY: reported
Outcomes	ANALYSIS: Efficacy was assessed on an intent to‐treat (ITT) basis; between‐group differences in changes from baseline in the primary variable were also evaluated in the per‐protocol population. Primary and secondary spirometry data and diary data were fit with an analysis of co‐variance (ANCOVA) model; results of urine cortisol analysis were summarised with descriptive statistics OUTCOMES GROWTH MEASUREMENT TECHNIQUE: not reported PULMONARY FUNCTION TESTS: measured at randomisation; week 2, 4, 8 and 12 Mean change from baseline in percentage of predicted normal FEV1 to the average during the 12‐week treatment period for each active treatment versus placebo FUNCTIONAL STATUS Morning and evening PEFR Daytime and nighttime asthma symptom scores Inhalations of albuterol per day BIOMARKERS Blood sample for pharmacokinetics Urine collected over 24 hours for cortisol assessment ADVERSE EVENTS: reported WITHDRAWALS: reported
Notes	PUBLICATION: full paper (2008) FUNDING: funded by AstraZeneca LP CONFIRMATION OF METHODOLOGY: not received
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Using a computer random number generator: "Using a computer‐generated allocation schedule stratified by pharmacokinetic participation, patients were randomised in balanced blocks to receive 12 weeks of one of the following five treatments"
Allocation concealment (selection bias)	Unclear risk	Insufficient information
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Blinding of participants and key study personnel ensured
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Blinding of participants and key study personnel ensured
Incomplete outcome data (attrition bias) All outcomes	Low risk	No missing outcome data
Selective reporting (reporting bias)	Low risk	Study protocol not available but published reports include all expected outcomes, including those that were prespecified
Other bias	Low risk	Study apparently free of other sources of bias