Skip to main content
NCBI home page
Biometrika logoLink to Biometrika
. 2009 Apr 27;96(2):263–276. doi: 10.1093/biomet/asp014

Mixtures of Polya trees for flexible spatial frailty survival modelling

Luping Zhao 1, Timothy E Hanson 2, Bradley P Carlin 2
PMCID: PMC2749263  NIHMSID: NIHMS128006  PMID: 19779579

Abstract

Mixtures of Polya trees offer a very flexible nonparametric approach for modelling time-to-event data. Many such settings also feature spatial association that requires further sophistication, either at the point level or at the lattice level. In this paper, we combine these two aspects within three competing survival models, obtaining a data analytic approach that remains computationally feasible in a fully hierarchical Bayesian framework using Markov chain Monte Carlo methods. We illustrate our proposed methods with an analysis of spatially oriented breast cancer survival data from the Surveillance, Epidemiology and End Results program of the National Cancer Institute. Our results indicate appreciable advantages for our approach over competing methods that impose unrealistic parametric assumptions, ignore spatial association or both.

Keywords: Areal data, Bayesian modelling, Breast cancer, Conditionally autoregressive model, Log pseudo marginal likelihood, Nonparametric modelling

References

  1. Banerjee S.  Carlin B. P. Spatial semiparametric proportional hazards models for analyzing infant mortality rates in Minnesota counties. In: Gatsonis C.,  Kass R. E.,  Carriquiry A.,  Gelman A.  Higdon D.,  Verdinelli D. K.,  Pauler I., , editors. Case Studies in Bayesian Statistics, Vol. VI. New York: Springer; 2002. pp. 137–51. [Google Scholar]
  2. Banerjee S.  Carlin B. P. Semiparametric spatio-temporal frailty modelling. Environmetrics. 2003;14:523–35. [Google Scholar]
  3. Banerjee S.  Carlin B. P.  Gelfand A. E. Hierarchical Modeling and Analysis for Spatial Data. Boca Raton, FL: Chapman and Hall/CRC Press; 2004. [Google Scholar]
  4. Banerjee S.  Dey D. K. Semi-parametric proportional odds models for spatially correlated survival data. Lifetime Data Anal. 2005;11:175–91. doi: 10.1007/s10985-004-0382-z. [DOI] [PubMed] [Google Scholar]
  5. Banerjee S.  Wall M. M.  Carlin B. P. Frailty modeling for spatially correlated survival data, with application to infant mortality in Minnesota. Biostatistics. 2003;4:123–42. doi: 10.1093/biostatistics/4.1.123. [DOI] [PubMed] [Google Scholar]
  6. Barker P.  Henderson R. Modelling converging hazards in survival analysis. Lifetime Data Anal. 2004;10:263–81. doi: 10.1023/b:lida.0000036392.68675.7f. [DOI] [PubMed] [Google Scholar]
  7. Bedrick E. J.  Christensen R.  Johnson W. O. Bayesian accelerated failure time analysis with application to veterinary epidemiology. Statist. Med. 2000;19:221–37. doi: 10.1002/(sici)1097-0258(20000130)19:2<221::aid-sim328>3.0.co;2-c. [DOI] [PubMed] [Google Scholar]
  8. Besag J. Spatial interaction and the statistical analysis of lattice systems (with discussion. J. R. Statist. Soc. 1974;B 36:192–236. [Google Scholar]
  9. Besag J.  York J. C.  Mollié A. Bayesian image restoration, with two applications in spatial statistics (with discussion) Ann. Inst. Statist. Math. 1991;43:1–59. [Google Scholar]
  10. Carlin B. P.  Louis T. A. Bayesian Methods for Data Analysis. 3rd ed. Boca Raton, FL: Chapman and Hall/CRC Press; 2008. [Google Scholar]
  11. Chen Y. Q.  Jewell N. P. On a general class of semiparametric hazards regression models. Biometrika. 2001;88:687–702. [Google Scholar]
  12. Clayton D.  Cuzick J. Multivariate generalizations of the proportional hazards model. J. R. Statist. Soc. 1985;A 148:82–117. [Google Scholar]
  13. Ferguson T. S. Prior distributions on spaces of probability measures. Ann. Statist. 1974;2:615–29. [Google Scholar]
  14. Geisser S.  Eddy W. F. A predictive approach to model selection. J. Am. Statist. Assoc. 1979;74:153–60. [Google Scholar]
  15. Gore S. M.  Pocock S. J.  Kerr G. R. Regression models and non-proportional hazards in the analysis of breast cancer survival. Appl. Statist. 1984;33:176–95. [Google Scholar]
  16. Gustafson P. On robustness and model flexibility in survival analysis: transformed hazards models and average effects. Biometrics. 2007;63:69–77. doi: 10.1111/j.1541-0420.2006.00679.x. [DOI] [PubMed] [Google Scholar]
  17. Hanson T. E. Inference for mixtures of finite Polya tree models. J. Am. Statist. Assoc. 2006;101:1548–65. [Google Scholar]
  18. Hanson T. E.  Johnson W. O. Modeling regression error with a mixture of Polya trees. J. Am. Statist. Assoc. 2002;97:1020–33. [Google Scholar]
  19. Hanson T. E.  Yang M. Bayesian semiparametric proportional odds models. Biometrics. 2007;63:88–95. doi: 10.1111/j.1541-0420.2006.00671.x. [DOI] [PubMed] [Google Scholar]
  20. Hennerfeind A.  Brezger A.  Fahrmeir L. Geoadditive survival models. J. Am. Statist. Assoc. 2006;101:1065–75. [Google Scholar]
  21. Huzurbazar A. V. A censored data histogram. Commun. Statist. Simul. 2005;34:113–20. [Google Scholar]
  22. Jeong J.-H.  Jung S.-H.  Wieand S. A parametric model for long-term follow-up data from phase III breast cancer clinical trials. Statist. Med. 2003;22:339–52. doi: 10.1002/sim.1349. [DOI] [PubMed] [Google Scholar]
  23. Jin X.  Carlin B. P. Multivariate parametric spatio-temporal models for county level breast cancer survival data. Lifetime Data Anal. 2005;11:5–27. doi: 10.1007/s10985-004-5637-1. [DOI] [PubMed] [Google Scholar]
  24. Lavine M. Some aspects of Polya tree distributions for statistical modeling. Ann. Statist. 1992;20:1222–35. [Google Scholar]
  25. Li Y.  Lin X. Semiparametric normal transformation models for spatially correlated survival data. J. Am. Statist. Assoc. 2006;101:591–603. [Google Scholar]
  26. Li Y.  Ryan L. Modeling spatial survival data using semiparametric frailty models. Biometrics. 2002;58:287–97. doi: 10.1111/j.0006-341x.2002.00287.x. [DOI] [PubMed] [Google Scholar]
  27. Mallick B. K.  Walker S. G. A Bayesian semiparametric transformation model incorporating frailties. J. Statist. Plan. Infer. 2003;112:159–74. [Google Scholar]
  28. Martinussen T.  Scheike T. H. Dynamic Regression Models for Survival Data. New York: Springer; 2006. [Google Scholar]
  29. Scharfstein D. O.  Tsiatis A. A.  Gilbert P. B. Efficient estimation in the generalized odds-rate class of regression models for right-censored time-to-event data. Lifetime Data Anal. 1998;4:355–91. doi: 10.1023/a:1009634103154. [DOI] [PubMed] [Google Scholar]
  30. Spiegelhalter D. J.  Best N.  Carlin B. P.  van der Linde A. Bayesian measures of model complexity and fit (with discussion) J. R. Statist. Soc. 2002;B 64:583–639. [Google Scholar]
  31. Vaupel J. W.  Manton K. G.  Stallard E. The impact of heterogeneity in individual frailty on the dynamics of mortality. Demography. 1979;16:439–54. [PubMed] [Google Scholar]
  32. Walker S. G.  Mallick B. K. Hierarchical generalized linear models and frailty models with Bayesian nonparametric mixing. J. R. Statist. Soc. 1997;B 59:845–60. [Google Scholar]
  33. Walker S. G.  Mallick B. K. Semiparametric accelerated life time model. Biometrics. 1999;55:477–83. doi: 10.1111/j.0006-341x.1999.00477.x. [DOI] [PubMed] [Google Scholar]
  34. Yakovlev A. Y.  Tsodikov A. D.  Boucher K.  Kerber R. The shape of the hazard function in breast carcinoma: curability of the disease revisited. Cancer. 1999;85:1789–98. [PubMed] [Google Scholar]
  35. Yin G.  Ibrahim J. G. A class of Bayesian shared gamma frailty models with multivariate failure time data. Biometrics. 2005;61:208–16. doi: 10.1111/j.0006-341X.2005.030826.x. [DOI] [PubMed] [Google Scholar]
  36. Zhang M.  Davidian M. “Smooth” semiparametric regression analysis for arbitrarily censored time-to-event data. Biometrics. 2008;64:567–76. doi: 10.1111/j.1541-0420.2007.00928.x. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Biometrika are provided here courtesy of Oxford University Press

RESOURCES