Luria 2001.

Methods	Randomised double‐blind controlled trial
Participants	Study period: September 1995 to December 1997 Setting: emergency departments at either Children's Hospital Medical Center in Cincinnati or Children's Hospital in Columbus, OH, USA Inclusion criteria: children aged 6 months to 6 years presenting to the emergency department with mild croup (barky cough, stridor and/or hoarseness for < 48 hours) and having a viral prodrome consisting of fever, cough, or rhinorrhoea (in an attempt to exclude children with spasmodic croup) Exclusion criteria: treated with corticosteroids 14 days prior to enrolling in the study; a clinical picture consistent with spasmodic croup; history of prolonged endotracheal intubation; history of chronic respiratory illness (i.e. asthma or cystic fibrosis); a condition associated with airway abnormalities; those without a working telephone or those with a severe disease (i.e. received nebulised racemic epinephrine or corticosteroids at the order of the emergency department physician or had < 94% oxygen saturation) Baseline demographics (N = 264): proportion males: treatment 1: 72%; treatment 2: 64%; control: 65% mean (range) age in months: treatment 1: 28 (6 to 70); treatment 2: 31 (6 to 71); control: 26 (6 to 71) mean (range) modified Westley croup score: treatment 1: 1.6 (0 to 6); treatment 2: 1.6 (0 to 5); control: 1.7 (0 to 5)
Interventions	Treatment group 1 (N = 85): single 0.60 mg/kg (maximum 10 mg) dose of oral dexamethasone (1 mg/mL) plus nebulised placebo solution Treatment group 2 (N = 91): single dose of oral placebo plus a 160 μg dose of nebulised dexamethasone sodium phosphate Control (N = 88): single dose of oral placebo plus nebulised placebo solution Nebulised study preparations were delivered with a nebuliser that had a fill volume of 3 mL and the oxygen flow set at 5 to 6 L/min.
Outcomes	Return visits to the emergency department
Notes	Funding source: the Bremer Foundation at the Ohio State University, Columbus, OH, USA
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The randomization was performed in blocks of 15 by the study pharmacist at each enrolling site with the use of a random number generator"
Allocation concealment (selection bias)	Low risk	Quote: "The study pharmacist then assembled numbered "croup kits" containing study preparations that reflected the results of the randomization." "The study physician retrieved the lowest numbered kit when enrolling a new subject to maintain the randomization order. Only pharmacists knew the results of the randomization."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "The kits were sealed to prevent any tampering and were kept in the EDs... Only the study pharmacists knew the results of the randomization... All oral study preparations were mixed 1:1 with a commercially available grape flavouring to minimize taste bias."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Comment: no description of a third‐party outcome assessor. Carried over judgement from blinding of participants and personnel
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Comment: used intention‐to‐treat analysis, including N = 9 protocol deviations. 16% (N = 43) loss on day 7 for the telephone follow‐up (14% in the oral treatment group, 15% in the nebulised treatment group, 19% in the placebo group)
Selective reporting (reporting bias)	Unclear risk	Comment: no protocol identified. All prespecified outcomes from methods appear in results.
Other bias	Low risk	Comment: no other sources of bias identified.
Overall risk of bias All outcomes	Unclear risk	Comment: at least 1 domain judged as unclear risk, and no domains judged as high risk.