Biometrika Advance Access originally published online on June 26, 2009
Biometrika 2009 96(3):545-558; doi:10.1093/biomet/asp024
Article |
Empirical Bayes estimation for additive hazards regression models
Department of Statistics, Florida State University, Tallahassee, Florida 32306, U.S.A. sinhad{at}stat.fsu.edu
Bristol-Myers Squibb, 5 Research Parkway, Wallingford, Connecticut 06492, U.S.A. brent.mchenry{at}bms.com
Harvard Medical School, Boston, Massachussets 02115, U.S.A. slipsitz{at}partners.org
Department of Statistics, University of Florida, Gainesville, Florida 32611, U.S.A. ghoshm{at}stat.ufl.edu
Received for publication 1 January 2006. Revision received 1 November 2008.
We develop a novel empirical Bayesian framework for the semiparametric additive hazards regression model. The integrated likelihood, obtained by integration over the unknown prior of the nonparametric baseline cumulative hazard, can be maximized using standard statistical software. Unlike the corresponding full Bayes method, our empirical Bayes estimators of regression parameters, survival curves and their corresponding standard errors have easily computed closed-form expressions and require no elicitation of hyperparameters of the prior. The method guarantees a monotone estimator of the survival function and accommodates time-varying regression coefficients and covariates. To facilitate frequentist-type inference based on large-sample approximation, we present the asymptotic properties of the semiparametric empirical Bayes estimates. We illustrate the implementation and advantages of our methodology with a reanalysis of a survival dataset and a simulation study.
Key Words: Gamma process • Integrated likelihood • Posterior process
References
-
Aalen O. O. A Model for Non-parametric Regression Analysis of Counting Processes. Lecture Notes in Statistics (1980) 2. New York: Springer. 1–25.
Aalen O. O. A linear regression model for the analysis of life times. Statist. Med. (1989) 8:907–25.[CrossRef]
Aalen O. O. Further results on the non-parametric linear regression model in survival analysis. Statist. Med. (1993) 12:1569–88.[CrossRef]
Beamonte E., Bermúdez J. D. A Bayesian semiparametric analysis for additive hazards models with censored observations. Test (2003) 12:347–63.[CrossRef][Web of Science]
Berger J. O. The robust Bayesian viewpoint. In: Robustness in Bayesian Statistics—Kadane J., ed. (1984) Amsterdam: North Holland. 63–124.
Carlin B. P., Louis T. A. Bayes and Empirical Bayes Methods for Data Analysis (2000a) New York: Chapman and Hall.
Carlin B. P., Louis T. A. Empirical Bayes: past, present and future. J. Am. Statist. Assoc. (2000b) 95:1286–9.[CrossRef][Web of Science]
Cox D. R. Regression models and life-tables (with Discussion). J. R. Statist. Soc. (1972) B 34:187–220.
Drzewiecki K. T., Ladefoged C., Christensen H. E. Biopsy and prognosis for cutaneous malignant melanomas in clinical stage 1. Scand. J. Plast. Reconstr. Surg. (1980) 14:141–4.[CrossRef][Web of Science][Medline]
Dunson D. B., Herring A. H. Bayesian model selection and averaging in additive and proportional hazards models. Lifetime Data Anal (2005) 11:213–32.[CrossRef][Web of Science][Medline]
Ibrahim J. G., Chen M.-H., Sinha D. Bayesian Survival Analysis (2001) New York: Springer.
Kalbfleisch J. D. Nonparametric Bayesian analysis of survival time data. J. R. Statist. Soc. (1978) B 40:214–21.
Lin D. Y., Ying Z. Semiparametric analysis of the additive risk model. Biometrika (1994) 81:61–71.
Lin D. Y., Oakes D., Ying Z. Additive hazards regression with current status data. Biometrik (1998) a 85:289–98.[CrossRef]
Martinussen T., Scheike T. H. Dynamic Regression Models for Survival Data (2006) New York: Springer.
Morris C. N. Parametric empirical Bayes inference: theory and applications (with Discussion). J. Am. Statist. Assoc. (1983) 78:47–55.[CrossRef][Web of Science]
Robbins H. An empirical Bayes approach to statistics. In: Proc. Third Berkeley Symp. Math. Statist. Prob.—Neyman J., ed. (1956) 1. Berkeley, CA: University of California Press. 157–63.
Robbins H. An empirical Bayes estimation problem. Proc. Nat. Acad. Sci. (1980) 77:6988–9.
Samaniego F. J., Reneau D. M. Toward a reconciliation of the Bayesian and frequentist approaches to point estimation. J. Am. Statist. Assoc. (1994) 89:947–57.[CrossRef][Web of Science]
Samaniego F. J., Neath A. A. How to be a better Bayesian. J. Am. Statist. Assoc. (1996) 91:733–42.[CrossRef][Web of Science]
Sinha D., Ibrahim J. G., Chen M.-H. A Bayesian justification of Cox's partial likelihood. Biometrik (2003) a 90:629–41.[CrossRef]
Yin G., Ibrahim J. G. A class of Bayesian shared Gamma frailty models with multivariate failure time data. Biometrics (2005) 61:208–16.[CrossRef][Web of Science][Medline]
Zeng D., Yin G., Ibrahim J. G. Semiparametric transformation models for survival data with a cure fraction. J. Am. Statist. Assoc. (2006) 101:670–84.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||