Choice of prognostic estimators in joint models by estimating differences of expected conditional Kullback-Leibler risks
- PMID: 22578147
- DOI: 10.1111/j.1541-0420.2012.01753.x
Choice of prognostic estimators in joint models by estimating differences of expected conditional Kullback-Leibler risks
Abstract
Prognostic estimators for a clinical event may use repeated measurements of markers in addition to fixed covariates. These measurements can be linked to the clinical event by joint models that involve latent features. When the objective is to choose between different prognosis estimators based on joint models, the conventional Akaike information criterion is not well adapted and decision should be based on predictive accuracy. We define an adapted risk function called expected prognostic cross-entropy. We define another risk function for the case of right-censored observations, the expected prognostic observed cross-entropy (EPOCE). These risks can be estimated by leave-one-out cross-validation, for which we give approximate formulas and asymptotic distributions. The approximated cross-validated estimator CVPOL (a) of EPOCE is studied in simulation and applied to the comparison of several joint latent class models for prognosis of recurrence of prostate cancer using prostate-specific antigen measurements.
© 2012, The International Biometric Society.
Similar articles
-
Information estimators for weighted observations.Neural Netw. 2013 Oct;46:260-75. doi: 10.1016/j.neunet.2013.06.005. Epub 2013 Jun 24. Neural Netw. 2013. PMID: 23859828
-
A universal approximate cross-validation criterion for regular risk functions.Int J Biostat. 2015 May;11(1):51-67. doi: 10.1515/ijb-2015-0004. Int J Biostat. 2015. PMID: 25849800
-
Development and validation of a dynamic prognostic tool for prostate cancer recurrence using repeated measures of posttreatment PSA: a joint modeling approach.Biostatistics. 2009 Jul;10(3):535-49. doi: 10.1093/biostatistics/kxp009. Epub 2009 Apr 15. Biostatistics. 2009. PMID: 19369642 Free PMC article.
-
[Prognostic factors of localised, locally advanced or metastatic prostate cancer].Bull Cancer. 2007 Jul;94(7 Suppl):F35-43. Bull Cancer. 2007. PMID: 17845992 Review. French.
-
[Prognosis and predictive factors in prostate cancer].Bull Cancer. 2009 Apr;96(4):439-49. doi: 10.1684/bdc.2009.0846. Bull Cancer. 2009. PMID: 19357018 Review. French.
Cited by
-
Optimizing dynamic predictions from joint models using super learning.Stat Med. 2024 Mar 30;43(7):1315-1328. doi: 10.1002/sim.10010. Epub 2024 Jan 25. Stat Med. 2024. PMID: 38270062
-
Joint models for dynamic prediction in localised prostate cancer: a literature review.BMC Med Res Methodol. 2022 Sep 19;22(1):245. doi: 10.1186/s12874-022-01709-3. BMC Med Res Methodol. 2022. PMID: 36123621 Free PMC article. Review.
-
Distributions of Hyper-Local Configuration Elements to Characterize, Compare, and Assess Landscape-Level Spatial Patterns.Entropy (Basel). 2020 Apr 8;22(4):420. doi: 10.3390/e22040420. Entropy (Basel). 2020. PMID: 33286194 Free PMC article.
-
Personalized screening intervals for biomarkers using joint models for longitudinal and survival data.Biostatistics. 2016 Jan;17(1):149-64. doi: 10.1093/biostatistics/kxv031. Epub 2015 Aug 28. Biostatistics. 2016. PMID: 26319700 Free PMC article.
-
Individualized dynamic prediction of prostate cancer recurrence with and without the initiation of a second treatment: Development and validation.Stat Methods Med Res. 2016 Dec;25(6):2972-2991. doi: 10.1177/0962280214535763. Epub 2014 May 20. Stat Methods Med Res. 2016. PMID: 24847900 Free PMC article.
MeSH terms
Substances
LinkOut - more resources
Full Text Sources