This page provides resources for sample size and power determinations for comparing two survival rates at a time determined by the user (e.g. 5 year survival rate), using the Kaplan Meier Log Rank Test.

The calculations were adapted from the formula in the standard text book for sample size (see reference)

The following resources are in the other panels

• Calculations : The same calculations are presented in the following two modes
  • A Javascript program, which allows the user to immediately calculate sample size and power in the web page format
  • The programs in R Codes, which allows the user to check the calculations, and to adapt the algorithm to other applications
• Tables of sample size for various combinations of Type I and II errors, one or two tail models, differences to detect, and assuming equal sample sizes in the two groups.

Please note: Sample size is calculated on the basis of the ratio of the two survival rates. In some scenarios the analyst may designate one of the groups to a survival rate of 0 or 1. When this occurs the calculations would either fail or produce erroneous results. The program therefore defaults all survuval rates of 0 to 0.0001 and 1 to 0.9999

Reference
Machin D, Campbell M, Fayers, P, Pinol A (1997) Sample Size Tables for Clinical Studies. Second Ed. Blackwell Science IBSN 0-86542-870-0 p. 176-177

Power(1-β)=0.8 Power(1-β)=0.9 Power(1-β)=0.95