Sample Size Estimation for EORTC QLQ-C30 Summary Score as Primary Endpoint

Abstract

Background

This paper discusses the importance of sample size in research, highlighting the ICH Guidance's requirement for

clinical studies to clearly explain their sample size, particularly when using the EORTC QLQ-C30 Summary Score as

the primary outcome measure. The idea behind this paper is to give the sample size when the outcome is Global

Health Status and Out-of-Pocket Expenditure.

Method

The sample size calculation is based on a simulation technique using SAS software, assuming equal allocation

between groups to yield a significant result with sufficient power.

Results

Sample sizes obtained using a simulation-based method require 25 patients per group to account for a minimum

clinically significant difference of 14 units between the two groups, with 88% statistical power at a 5% level of

significance.

Conclusion

The simulation-based technique, combined with validated software, can be helpful when the effect size and

variability of the QoL score are not precisely known, particularly when the effect size is not sufficiently specified to

determine the sample size accurately. Compared to conventional techniques, simulation-based sample size

estimation increases confidence in reaching statistical power, particularly for complicated endpoints like the

EORTC QLQ-C30 summary score.