At baseline, a Mann-Whitney test was used
to compare the difference in hazard distributions
between groups. Although slightly more
hazards were identified in the intervention
schools than in the controls, the observed
differences were not statistically significant
(z=1.11, p=0.27, calculated assuming a Poisson
distribution for the counts). Similarly, at
the time of the first follow up (table 3), there
were no significant group differences (z=1.35,
p=0.178). At the second follow up, however,
there was a significantly greater reduction in
the intervention group than in the control
group (z=2.21, p=0.027). Eight of the 12
intervention schools had decreased their hazards
by at least three and none had increases. In
contrast, only two of the 12 control schools
decreased their hazards by at least three; and
three schools had increased hazards over that
time. The median reduction in hazards was a
drop of between none to one hazards for
control schools and three hazards for intervention
schools.