Skip to main content
. Author manuscript; available in PMC: 2009 Jan 5.
Published in final edited form as: Qual Life Res. 2004 May;13(4):717–723. doi: 10.1023/b:qure.0000021686.47079.0d

Figure 1.

Figure 1

Number of patients needed per group for a clinical trial where the primary outcome measure is a composite of up to five subscales and the intervention is thought to affect each subscale equally. For the purposes of illustration, correlation between subscales is set at 0.4, and the required sample size if only one subscale used is set at 50. Number of patients needed continues to fall as the number of subscales combined increases.