-
Notifications
You must be signed in to change notification settings - Fork 5
/
toAdd.txt
14 lines (14 loc) · 957 Bytes
/
toAdd.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Carryover effect will be described by comparing differences between S
and P when S was first to S - P when P was first. The results of the
carryover effect assessment will not alter the final analysis (because
that would alter the alpha level as described by S Senn and others)
but will provide a possible cautionary note. The statistical test to
be used is the Wilcoxon signed rank test on the combined set of
differences (S - P when S was first and S - P when P was first).
Before applying this test, mean-difference (Bland-Altman) plots will
be used to check that PCr recovery time is properly transformed so
that differences may be analyzed outside the context of the baseline
level. [OMIT least squares and any discussion of baseline covariates
as you are using each patient as her own control so covariates cancel
out]. Secondary analyses will use linear models to see if S - P
differences vary with baseline covariates such as BMI, race, and sex.