Statistical analysis:
The patients were divided into three age categories: preschoolers (≥3 yr
and <6 yrs), schoolchildren (≥6 yrs and <12 yrs),
and adolescents (≥12 yrs and ≤17 yrs). We identified the presence of the
inversion of the reactance curve (Figure 1), and we measured the lineal
tendency of the percentage of patients that required X5 approx.
correction according to the age
categories, using the
Mantel-Haenszelײ test. We analyzed the difference between patients with
and without inversion of the reactance curve and X5 correction in the
percentage of IOS alteration parameters. Then we measured the mean
differences of IOS and FEF25-75% between the groups of
patients with and without X5 approx. correction. Also, the correlation
of X5 and X5 approx. with the other IOS basal parameters in each age
category was measured. The differences in % of basal parameters in IOS
were calculated using the Fisherײ test or the Pearson test with the
Yates correction, where appropriate. For measuring the mean difference
between patients with and without X5 approx. correction, the t-student
test for an independent sample was used, and for the correlation of X5
and X5 approx. with the IOS and spirometry parameters, the Pearson
correlation test was used. In both analyses, normal distribution was
verified.
The sample size for establishing the difference between groups without
X5 and with X5 correction was calculated in accordance with the pilot
study, and 380 patients were needed with 80% of power and 0.05 alfa.
For the statistical analysis, SPSS® v17.0 (IBM,
Armonk, NY) software was used.