Garbutt 2013.

Methods	Randomised double‐blind controlled trial
Participants	Study period: 26 October 2009 to 16 April 2010 and 6 September 2010 to 29 April 2011 Setting: 10 offices of primary care practitioners in St Louis, MO, USA Inclusion criteria: children aged 1 to 8 years with croup symptoms for ≤ 48 hours and a clinical diagnosis of mild or moderate croup at an office visit, based on symptoms in the past 24 to 36 hours Exclusion criteria: diagnosis of severe croup or impending respiratory failure; prior treatment with epinephrine or oral corticosteroids for this croup episode; symptoms or signs suggesting other cause of stridor; chronic respiratory disease including asthma; known contraindication to steroid use; parent not in the same household as the child in the subsequent 4 days, could not participate in telephone interviews, or was not English speaking Baseline demographics (N = 87): proportion male: treatment: 61%; comparator: 68% mean (SD) age in years: treatment: 2.67 (1.43); comparator: 3.11 (1.58) mean (SD) Westley croup score: treatment : 0.4 (0.7); comparator: 0.6 (0.8) mean (SD) Telephone Outpatient Score: treatment: 2.2 (0.9); comparator: 2.0 (0.9)
Interventions	Treatment (N = 41): single 2 mg/kg (maximum 60 mg/day) dose of oral prednisolone once per day for 3 days Comparator (N = 46): single 0.60 mg/kg (maximum 18 mg) dose of oral dexamethasone, followed by 2 days of placebo comparable in appearance, smell, and taste
Outcomes	Additional health care for croup within 11 days of randomisation
Notes	Funding source: National Center for Research Resources (National Institutes of Health); National Institutes of Health Roadmap for Medical Research
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "randomized blocks were used to assign subjects to treatment groups, with randomization stratified by site. Computer generated random numbers determined how the two treatments were allocated"
Allocation concealment (selection bias)	Low risk	Quote: "Study drug packages were prepared offsite by the pharmacist" "The pharmacist... packaged the bottles in a sealed opaque envelope" "For allocation concealment, the drug formulation ensured the volume of the weight‐based dose was equivalent for each medication."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "The pharmacist... packaged the bottles in a sealed opaque envelope" "comparable in appearance, smell and taste" "patients, parents, PCPs, and study team members were blinded to treatment assignments"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "The pharmacist... packaged the bottles in a sealed opaque envelope" "comparable in appearance, smell and taste" "patients, parents, PCPs, and study team members were blinded to treatment assignments"
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: intention‐to‐treat analysis used. Losses to follow‐up on days 1, 2, 3, 4, and 11 were 7%, 9%, 9%, 3%, and 2% (non‐cumulative). Participation in follow‐up interviews was balanced between groups.
Selective reporting (reporting bias)	Low risk	Comment: no deviations from protocol detected.
Other bias	Low risk	Comment: no other sources of bias identified.
Overall risk of bias All outcomes	Low risk	Comment: all domains judged as low risk.