Example 1:
This example is to evaluate Docosahexaenoic Acid (DHA) in patients with retinitis pigmentosa (RP) receiving vitamin A treatment. The patients were randomized to either 1200 mg of docosahexaenoic acid (DHA) plus 15,000 IU/day of vitamin A or control fatty acids + 15,000 IU/day of vitamin A. And the endpoint was change in central visual field measured by the Humphrey Field analyzer 30-2 program. The layout of data hump302 is as the follows:
id group numeye change1
6831 2 1 -34.0
8992 1 1 15.5
F369 1 1 75.0
F540 1 1 -117.0
F696 2 1 -125.5
0631 2 2 61.5
0631 2 2 44.5
3911 2 2 141.5
3911 2 2 108.0
4077 1 2 -58.0
4077 1 2 -93.5
5762 2 2 -51.5
5762 2 2 -94.0
5984 1 2 -96.0
5984 1 2 -97.5
6780 1 2 -98.0
6780 1 2 -233.0
6850 1 2 -100.0
6850 1 2 -190.5
7929 1 2 -435.5
7929 1 2 -75.0
B071 2 2 52.5
B071 2 2 37.0
B522 1 2 -83.0
B522 1 2 -299.5
B873 2 2 -165.5
B873 2 2 -164.5
D365 2 2 -187.0
D365 2 2 -98.5
…
In the data, group: has values 1 and 2.
1 = control group taking 15,000 IU/day of Vitamin A + placebo;
2 = active group taking 15,000 IU/day of Vitamin A+ 1200 mg of DHA;
numeye: has values 1 and 2,
1= patients were only evaluable on 1 of 2 eyes;
2= patients were evaluable on 2 eyes;
Change1: shows the change in central visual field.
From the previous study, we can estimate theta_hat ( ) and rho_hat ( ). We also propose a new study with 50 patients in each group. In each group, there are 5 1-eyed patients and 45 2-eyed patients. We want to estimate the power of the new proposed study.
Here is the syntax for calling the %CluswilcoxPower( ) and the output of the macro:
libname dat '../UnBalanced/';
data hump302;
set dat.hump302 dat.eye2;
run;
%include "CluswilcoxPower.sas";
%CluswilcoxPower(ipnlst=5 45, ipmlst=5 45, est_mu=Y, est_rho=Y, dsn=hump302, varn=change1,
grp=group, mu=,rho=,igmin=1, igmax=2);
The output:
]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]
Previous study 11:03 Monday, July 26, 2010 2
Number of clusters of size 1 - 2..
m1=clusters of size 1 (control group)...
m1 m2
3 31
]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]
Previous study 11:03 Monday, July 26, 2010 3
Number of clusters of size 1 - 2..
n1=clusters of size 1 (active treatment group)...
n1 n2
2 27
]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]
Estimations from Previous study 11:03 Monday, July 26, 2010 4
theta_ rho_
combine overall
0.66872 0.70162
]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]
New study 11:03 Monday, July 26, 2010 5
Number of clusters of size 1,2,3...
ipm1=clusters of size 1 (control group), ipm2=clusters of size 2 (control group)...
ipn1=clusters of size 1 (active treatment group)...
ipm1 ipm2 ipn1 ipn2
5 45 5 45
]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]
New study 11:03 Monday, July 26, 2010 6
Obs w0g1 w0g2 var0theta w1g1 w1g2 var1theta power
1 27.2727 318.840 .002889236 31.5138 372.886 .002472828 0.89873