ABSTRACT
Sample size calculation is one of several crucial points when researchers plan a new trial because the number of patients directly relates to the feasibility and scientific value of the study. This chapter's objective is to share and understand what kinds of parameters are needed to calculate the sample size for phase III trials in oncology among investigators, including biostatisticians. For this purpose, this chapter introduces how sample size can be calculated in randomized parallel controlled (1:1) phase III oncology trial whose primary endpoint is a time-to-event. In addition, this chapter addresses 2 types of phase III trials whose objectives are superiority or non-inferiority, with examples from actual trials. Sample size calculation using a “quick” formula (for sample size calculation by hand), SAS software, PASS software and the SWOG website are introduced and their outputs are compared. Additional notes on violation of exponential survival curve's assumption are provided.
