A MICROCOMPUTER PROGRAM FOR CALCULATING THE CONFIDENCE INTERVALS OF SURVIVAL PROBABILITY IN MEDICAL FOLLOW-UP STUDIES

In cancer survival analysis, it is very frequently to estimate the confidence intervals for survival probabilities. But this calculation is not commonly involve in most popular computer packages, or only one methods of estimation in the packages. In the present paper, we will describe a microcomputer program for estimating the confidence intervals of survival probabilities, when the survival functions are estimated using Kaplan-Meier product-limit or life-table method. There are five methods of estimation in the program (SPCI), which are the classical (based on Greenwood's formula of variance of S(t_i)), Rothman-Wiison, arcsin transformation, log(-log) transformation, Iogit transformation methods. Two example analysis are given for testing the performances of the program running.