In an effort to
limit the variability between individual cow baseline
RT, mastitis, lameness, and pneumonia, cows were used
as their own healthy controls if they had 37 d with no
recorded disease before the week of their health event.
Reticular temperature data for the period 31 to 37 d
before the health event were used as healthy controls
when examining the data for the 7 d before the health
event; RT for the period 31 to 34 d before the health
event were used as healthy controls when examining
the data for the 4-d health event interval. A separate
group of controls was selected for comparison purposes
for cows diagnosed with metritis, as their diagnosis
occurred shortly after calving and no RT data were
collected on them before freshening. The control group
for metritis consisted of 63 cows (approximately 4
times the number of healthy cows as those diagnosed
with a health event), and was chosen using the random
number generator in Microsoft Excel 2010 (Microsoft,
Redmond, WA) from a group of cows that did not have