where
p 1 ¼
yp 0
1 þðy 1 Þp 0
The problem is more complicated if we plan for groups with unequal sample
sizes. First, we have to specify the allocation of sizes:
n 1 ¼w 1 N
n 0 ¼w 0 N
w 1 þw 0 ¼ 1 : 0
whereNis the total sample size needed. For example, if we want the size of the
control group to be three times the number of the cases, w 1 ¼ 0 :25 and
w 0 ¼ 0 :75. Then the the total sample size neededNcan be obtained from the
formula
ffiffiffiffiffi
N
p
jp 1 p 0 j¼z 1 a
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pð 1 pÞðw 11 þw 01 Þ
q
þz 1 b
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
p 1 ð 1 p 1 Þw 11 þp 0 ð 1 p 0 Þw 01
q
wherepis the weighted average:
p¼w 1 p 1 þw 0 p 0
Example 12.14 Suppose that an investigator is considering designing a case–
control study of a potential association between congenital heart defect and the
use of oral contraceptives. Suppose also that approximately 30% of women of
childbearing age use oral contraceptives within three months of a conception,
and suppose that a relative risk ofy¼3 is hypothesized. We also decide to
preseta¼ 0 :05 and want to design a study with equal sample sizes so that its
power to the hypothesized relative risk ofy¼3 is 90% (orb¼ 0 :10).
First, the exposure rate for the cases, the percent of women of childbear-
ing age who use oral contraceptives within three months of a conception, is
obtained from
p 1 ¼
yp 0
1 þðy 1 Þp 0
¼
ð 3 Þð 0 : 3 Þ
ð 3 1 Þð 0 : 3 Þ
¼ 0 : 5625
470 STUDY DESIGNS