Statistical analyses
Based on findings of a previous study [6] , the nasal symptom score for the treatment group was statistically significantly different compared to the placebo group (F = 6.118, p = 0.017). This yielded an effect size estimate of 0.66. For 80% power using a two-sided significance level of 5%, the required sample size is 38 per group. To allow for dropouts, we used a sample size of 52 per group.
All data were analysed using the Statistical Package for the Social Sciences (windows version 11.5; SPSS, USA) for Windows by a statistician blinded to the allocation of treatments to groups. When comparing measures at the end of the study for the two groups, ANCOVA was used with the baseline as covariate. The data from non-repeated measures such as blood tests was analysed using t-tests. Outcome measures with nominal responses were analysed using χ2 tests. All p-values were obtained via 2-sided tests and were assessed at α = 0.05. The age variable was used as a covariate for outcome analyses as the two treatment groups were not comparable on this variable. Intention to treat was employed in all relevant data analyses.