A detailed description of the study design, including screening and recruitment procedures and statistical analysis, has been reported in detail elsewhere  and will only be briefly summarized.
Pre-school children (ages 1-4), with a documented history of at least 3 episodes of wheeze or asthma in the 6 months preceding the visit, evaluated as outpatients for exacerbation of respiratory symptoms were screened and characterized . In addition to the three episodes in the previous 6 months, the included children had a history of frequent wheeze, induced by other triggers than viral respiratory tract infection . Nineteen pediatric specialist care units were involved in this randomized, multicenter, double blind, parallel-group, placebo-controlled trial. Exclusion criteria were: a) history of severe exacerbations requiring systemic glucocorticoids; b) chest infection or hospitalization due to asthma in the previous 4 weeks; c) treatment with inhaled glucocorticoids or methyl-xanthine during the previous 4 weeks; d) treatment with oral glucocorticoids in the previous 8 weeks. Eligible children entered a 2-week prospective open run-in period during which they received only 2500 μg prn nebulized salbutamol for symptom relief. Patients entered the double-blind treatment phase if they had wheeze and/or cough, and/or shortness of breath, and/or required relief medication on at least 7 days of the 2-week run-in period. The study was performed in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. Independent ethics committees approved the study protocol, patient information sheet, and consent forms. A parent or guardian of each patient provided written informed consent. Parents/guardians were trained: a) to correctly use the nebulizer; b) to keep daily records of their child's symptoms, recording scores for wheeze, cough, and shortness of breath on diary cards, using a four-point rating scale ranging from 0 (no symptoms) to 3 (disabling symptom) ; c) to record the number of occasions during the night they were woken up because of the child's asthma symptoms which required rescue medication for symptom relief.
Children were stratified in two groups on the basis of the presence or absence of risk factors for developing persistent asthma : (1) atopic dermatitis or eczema; (2) asthma in a first-degree relative; (3) blood eosinophilia > 4% demonstrated in the previous 6 months. A 2:1 treatment-to-placebo randomization ratio was used to expose fewer patients to placebo.
In the original study , 3 treatments were tested in this population over 12 weeks:
a) BDP (Clenil per Aerosol, Chiesi Farmaceutici, Parma, Italy) 400 μg/vial, one vial b.i.d., plus salbutamol (Ventmax, Chiesi Farmaceutici, Parma, Italy) 2500 μg/vial, one vial prn, or b) placebo one vial b.i.d., plus fixed combination of BDP and salbutamol (Clenil Compositum, Chiesi Farmaceutici, Parma, Italy), 800 μg BDP + 1600 μg salbutamol/vial, one vial prn, or c) placebo one vial b.i.d. plus salbutamol 2500 μg/vial, one vial prn. All drugs were delivered with the same nebulizer (Clenny aerosol; Medel, Parma, Italy) using a tight-fitting face mask.
In the current post-hoc analysis we compare the efficacy of regular BDP plus prn salbutamol (BDP group) vs. regular placebo plus prn salbutamol (placebo group) on clinical outcomes captured in patients' diaries during the first week of treatment.
Groups were compared in terms of percentage of SFDs at the end of 1-week treatment using an ANCOVA model with percentage of SFDs as dependent variable and treatment, country and subgroup as fixed effects and baseline (run-in) as covariate. A day-by-day analysis of the percentage of SFDs was also performed using the same ANCOVA model.
Moreover, a comparison between groups within each subgroup of the percentage of SFDs at the end of 1-week treatment was performed using the same ANCOVA model with also the subgroup*treatment interaction as fixed effect.
The proportion of patients without asthma symptoms during the first week of treatment was compared between groups by means of a logistic regression model with presence of symptoms during the first week as dependent variable, treatment, country, subgroup and percentage of SFDs during run-in as independent variables.
Each of the three symptom scores was also analyzed by means of a mixed model for repeated measures with score at each day as dependent variable, treatment, country, subgroup, day, treatment*day interaction and mean value of symptoms during run-in as independent variables. All tests were considered as explorative.