This paper describes factors influencing sample sizes needed for comparing survival data from two treatment groups. General formulas for sample size calculations are given for comparisons based on the logrank statistic and for comparisons based on the difference in two estimated survival curves at a fixed time point. The importance of various factors is illustrated by numerical examples. Factors that dilute the treatment effect, factors that alter the degree of censoring, and the choice of the test statistic have the greatest impact on required sample size. Adjustments for covariates, adjustments for repeated monitoring of accruing survival data, and adjustments for other types of multiple comparison, however, can also influence required sample size.
KaplanELMeierPNonparametric estimation from incomplete observations. J Am Stat Soc. 1958;53:457–481.
2.
MantelNEvaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep. 1966;50:163–170.
3.
CoxDRRegression models and life tables (with discussion). J Royal Stat Soc B. 1972;34:187–220.
4.
GehanEAA generalized two-sample Wilcoxon test for doubly censored data. Biometrika. 1965;52:650–652.
5.
GreenwoodM.The Natural Duration of Cancer. Reports on Public Health and Medical Subjects 33. London: H.M. Stationary office; 1926.
6.
BreslowNECrowleyJA large sample study of the life table and product limit estimates under random censorship. Ann Stat. 1974;2:437–453.
7.
SchoenfeldDThe asymptotic properties of non-parametric tests for comparing survival distributions. Biometrika. 1981;68:316–319.
8.
GeorgeSLDesuMMPlanning the size and duration of a clinical trial studying the time to some critical event. J Chronic Dis. 1974;74:15–24.
9.
RubinsteinLVGailMHSantnerTJPlanning the duration of a comparative clinical trial with loss to follow-up and a period of continued observation. J Chronic Dis. 1981;34:469–479.
10.
LachinJMFoulkesMAEvaluation of sample size and power for analyses of survival with allowance for nonuniform patient entry, losses to follow-up, noncompliance and stratification. Biometrics. 1986;42:507–519.
11.
FreedmanLSTables of number of patients required in clinical trials using the logrank test. Stat Med. 1982;1:121–129.
12.
HalperinMRogotEGurianJEdererFSample sizes for medical trials with special reference to long-term therapy. J Chronic Dis. 1968;21:13–24.
13.
WuMFisherMDeMetsDSample sizes for long-term medical trials with time-dependent drop-out and event rates. Controlled Clin Trials. 1980;1:109–121.
14.
GailMApplicability of sample size calculations based on a comparison of proportions for use with the logrank test. Controlled Clin Trials. 1985;6:112–119.
15.
LakatosESample sizes based on the log-rank statistic in complex clinical trials. Biometrics. 1984;44:229–241.
16.
SelfSPrenticeRIversonDStatistical design of the Women's Health Trial. Controlled Clin Trials. 1988;9:119–136.
17.
LakatosESample size determination in clinical trials with time-dependent rates of losses and non-compliance. Controlled Clin Trials. 1986;7:189–199.
18.
PattersonBHThe impact of screening and eliminating preexisting cases on sample size requirements for cancer prevention trials. Controlled Clin Trials. 1987;8:87–95.
19.
BrookmeyerRGailMHAIDS Epidemiology: A Quantitative Approach. Chapter 11. New York: Oxford University Press; 1994.
20.
CusickJThe efficiency of the proportions test and the logrank test for censored survival data. Biometrics. 1982;38:1033–1039.
21.
HalpernJBrownBWJr.Cure rate models: Power of the logrank and generalized Wilcoxon tests. Stat Med. 1987;6:483–489.
22.
LagakosSWSchoenfeldDProperties of proportional-hazards score tests under misspecified regression models. Biometrics. 1984;40:1037–1048.
23.
MorganTMOmitting covariates from the proportional hazards model. Biometrics. 1986;42:993–995.
24.
BernsteinDLagakosSWSample size and power determination for stratified clinical trials. J Stat Comput Simulat. 1978;8:65–73.
25.
TaulbeeJDSymonsMJSample size and duration for cohort studies of survival time with covariables. Biometrics. 1983;39:351–360.
26.
PocockSJGroup sequential methods in the design and analysis of clinical trials. Biometrika. 1977;64:191–199.
27.
O'BrienPCFlemingTRA multiple testing procedure for clinical trials. Biometrics. 1979;35:549–556.
28.
EaSTA software package for the design and interim monitoring of group sequential clinical trials. Cambridge, MA: Cytel Software Corporation; 1992.
29.
RubinsteinLVGailMHMonitoring rules for stopping accrual in comparative survival studies. Controlled Clin Trials. 1982;3:325–343.
30.
FlemingTREvaluation of active control trials in AIDS. J AIDS. 1990;3:S82–87.
31.
HalpernJBrownBWJr.Designing clinical trials with arbitrary specification of survival functions and for the logrank or generalized Wilcoxon test. Controlled Clin Trials. 1987;8:177–189.
32.
HalpernJBrownBWJr.A computer program for designing clinical trials with arbitrary survival curves and group sequential testing. Controlled Clin Trials. 1993;14:109–122.
33.
LakatosELanKKGA comparison of sample size methods for the logrank statistic. Stat Med. 1992;11:179–191.
34.
GAUSS System Version 2.0. Kent, WA: Aptech Systems, Inc.; 1988.
