A New Bayesian Model For Survival Data With a Surviving Fraction

Abstract We consider Bayesian methods for right-censored survival data for populations with a surviving (cure) fraction. We propose a model that is quite different from the standard mixture model for cure rates. We provide a natural motivation and interpretation of the model and derive several novel properties of it. First, we show that the model has a proportional hazards structure, with the covariates depending naturally on the cure rate. Second, we derive several properties of the hazard function for the proposed model and establish mathematical relationships with the mixture model for cure rates. Prior elicitation is discussed in detail, and classes of noninformative and informative prior distributions are proposed. Several theoretical properties of the proposed priors and resulting posteriors are derived, and comparisons are made to the standard mixture model. A real dataset from a melanoma clinical trial is discussed in detail.

[1]  J. Ibrahim,et al.  Prior elicitation, variable selection and Bayesian computation for logistic regression models , 1999 .

[2]  K. Yamaguchi.  Accelerated Failure-Time Regression Models with a Regression Model of Surviving Fraction: An Application to the Analysis of “Permanent Employment” in Japan , 1992 .

[3]  Vernon T. Farewell,et al.  Mixture models in survival analysis: Are they worth the risk? , 1986 .

[4]  J. Kirkwood,et al.  Interferon alfa-2b adjuvant therapy of high-risk resected cutaneous melanoma: the Eastern Cooperative Oncology Group Trial EST 1684. , 1996, Journal of clinical oncology : official journal of the American Society of Clinical Oncology.

[5]  C. Morris.  Natural Exponential Families with Quadratic Variance Functions: Statistical Theory , 1983 .

[6]  W. Gilks,et al.  Adaptive Rejection Sampling for Gibbs Sampling , 1992 .

[7]  P. Diaconis,et al.  Conjugate Priors for Exponential Families , 1979 .

[8]  H. Sather,et al.  A comparison of tests of the difference in the proportion of patients who are cured. , 1992, Biometrics.

[9]  Anthony Y. C. Kuk,et al.  A mixture model combining logistic regression with proportional hazards regression , 1992 .

[10]  J Halpern,et al.  Designing clinical trials with arbitrary specification of survival functions and for the log rank or generalized Wilcoxon test. , 1987, Controlled clinical trials.

[11]  J. Ibrahim,et al.  Using Historical Controls to Adjust for Covariates in Trend Tests for Binary Data , 1998 .

[12]  Ming-Hui Chen,et al.  Monte Carlo Estimation of Bayesian Credible and HPD Intervals , 1999 .

[13]  J B Greenhouse,et al.  Assessing Placebo Response Using Bayesian Hierarchical Survival Models , 1998, Lifetime data analysis.

[14]  J M Taylor,et al.  Semi-parametric estimation in failure time mixture models. , 1995, Biometrics.

[15]  J G Ibrahim,et al.  The Large Sample Distribution of the Weighted Log Rank Statistic Under General Local Alternatives , 1997, Lifetime data analysis.

[16]  Joseph Berkson,et al.  Survival Curve for Cancer Patients Following Treatment , 1952 .

[17]  J Halpern,et al.  Cure rate models: power of the logrank and generalized Wilcoxon tests. , 1987, Statistics in medicine.

[18]  C. Morris.  Natural Exponential Families with Quadratic Variance Functions , 1982 .

[19]  A I Goldman,et al.  Survivorship analysis when cure is a possibility: a Monte Carlo study. , 1984, Statistics in medicine.

[20]  Joel B. Greenhouse,et al.  A competing risks derivation of a mixture model for the analysis of survival data , 1984 .

[21]  A A Tsiatis,et al.  A linear rank test for use when the main interest is in differences in cure rates. , 1989, Biometrics.

[22]  E. Laska,et al.  Nonparametric estimation and testing in a cure model. , 1992, Biometrics.