Figure 1: Representative kindergarten through
12th grade academic year in the Oregon School District
(OSD: Dane County, WI) showing school days (yellow squares) and
non-school days for the 2016-2017 academic year. Longer-duration planned
school breaks for Thanksgiving, winter break and spring break are
demonstrated by empty boxes.
Figure 2. Weekly counts of Wisconsin influenza A and influenza
B detections combined for the 5 academic years in this study. Influenza
surveillance data were provided by the Wisconsin State Laboratory of
Hygiene. Lightly shaded vertical bands demonstrate the actual timing of
winter breaks (green) and spring breaks (orange) for each academic year
/ influenza season in this study. Nine-day pseudo-breaks between winter
and spring breaks each year were introduced in this analysis to support
any findings of impact from the planned breaks; they are shown as purple
bars. Dark shaded bands demonstrate the 2-week assessment periods before
and after winter breaks (green), spring breaks (orange) and
pseudo-breaks (purple).
Figure 3. Distribution of number of absences due to
influenza-like illness (a-ILI) per study year across all grade levels in
the Oregon School District. 0 = kindergarten. 13 = students eligible to
remain in the public school system past age 18 years.
Figure 4. Proportion of students absent due to influenza-like
illness in the 2-week periods before and after winter breaks (green –
upper panel), spring breaks (orange – upper panel) and pseudo-breaks
(purple – lower panel) in each of 5 academic years. Nine-day-long
pseudo-breaks between winter and spring breaks each year were included
in this analysis to act as control periods for comparison. The 95%
confidence intervals are demonstrated by brackets.
Figure 5: Estimated mean absenteeism due to influenza-like
illness (a-ILI) counts for each of 10 school days before and after
winter breaks (green line – upper panel), spring breaks (orange line –
upper panel) and pseudo-breaks (purple line – lower panel) in each of 5
academic years (please see Table1 for precise dates and duration of real
and pseudo-breaks). Nine-day pseudo-breaks between winter and spring
breaks each year were introduced in this analysis to support any
findings of true impact from the planned breaks. The 95% confidence
intervals are demonstrated by shading. Model estimated an overall
reduction in a-ILI in the periods following break compared to the period
before break.