For categorical variables with answers representing
some order of intensity (e.g., very
sure, somewhat sure, not sure at all), we used
the Cochran-Mantel-Haenszel row mean
score to test whether responses were the same
at the baseline and exit visits. For questions
involving the number of days, quality-of-life
scores, number of visits, and number of triggers,
we used the paired t test to test whether
there was a significant change in the means
from baseline to exit visit. For these same
variables, we used the 2-sample t test to determine
whether the mean change from baseline
to exit visit was significantly different
between the study and comparison groups.
For all tests, we defined statistical significance
as P < .05.
We used McNemar’s test to determine
whether the percentage of allergen samples
with concentrations at or above the detection
limit (DL) was the same at baseline and exit
visits.