Download PDF
Methods Inf Med 1992; 31(04): 275-283
DOI: 10.1055/s-0038-1634886
Original Article
Schattauer GmbH

Study and Applications of an Informational Measure of Dependence in Survival Models

A. EI Hasnaoui
1   Laboratoire de Biostatistique et Informatique Médicale, Hôpital Necker, Paris, France
,
J. P. Jais
1   Laboratoire de Biostatistique et Informatique Médicale, Hôpital Necker, Paris, France
› Author Affiliations
Further Information

Publication History

Publication Date:
07 February 2018 (online)

    Permissions and Reprints

Abstract:

In survival models, when the factor of interest is a continuous variable or is expressed through a group of several variables, the classical measures of risk, i. e. relative risk and odds ratio, are not appropriate and there is no standard measure of dependence between survival and the considered factor. The Information Gain has been proposed by Linfoot (1957) and Kent (1983), giving any parametric model as a generalization of the squared product-moment correlation coefficient of the linear regression model with normal errors. By using simulation methods, we studied the statistical properties of the information gain as a measure of dependence, in the particular case of survival regression models. We suggest several efficient applications of this informational concept to some classical problems of regression analysis and prognostic analysis. Our ideas are illustrated through an example on the prognosis of idiopathic dilated cardiomyopathy.

Key-Words

Information Gain - Measure of Dependence - Survival Models - Prognostic Analysis - Optimal Coding
 

  • REFERENCES

  • 1 Bruzzi P, Green SB, Byar DP, Brinton LA. et al. Estimating the population attributable risk for multiple risk factors using case-control data. Amer J Epidemiol 1985; 122: 904-14.
    CrossrefPubMedGoogle Scholar
  • 2 Benichou J, Gail MH. Estimate of absolute cause-specific risk in cohort studies. Biometrics 1990; 46: 813-26.
    CrossrefPubMedGoogle Scholar
  • 3 Kullback S, Leibler RA. On information and sufficiency. Ann Math Statist 1951; 22: 79-86.
    CrossrefPubMedGoogle Scholar
  • 4 Kullback S. Information Theory and Statistics. New York: Dover; 1968
    Google Scholar
  • 5 Linfoot EH. An informational measure of correlation. Information and Control 1957; 01: 85-9.
    CrossrefPubMedGoogle Scholar
  • 6 Kent JT. Information gain and a general measure of correlation. Biometrika 1983; 70: 163-73.
    CrossrefPubMedGoogle Scholar
  • 7 Kent JT, O’Quigley J. Measures of dependence for censored survival data. Biometrika 1988; 75: 525-34.
    CrossrefPubMedGoogle Scholar
  • 8 Clayton D, Cuzick J. Multivariate generalizations of the proportional hazards model. J R Statist Soc 1985; 148: 82-117.
    PubMedGoogle Scholar
  • 9 Prentice RL. Linear rank tests with right censored data. Biometrika 1978; 65: 167-79.
    CrossrefPubMedGoogle Scholar
  • 10 Diciccio TJ, Romano JP. A review of bootstrap confidence intervals. J R Statist Soc 1988; 50: 338-54.
    PubMedGoogle Scholar
  • 11 SAS Institute Inc. SAS Version 5. Cary, NC: SAS Institute Inc; 1985
    Google Scholar
  • 12 Gauss System Version 2.0. Aptech System Inc; 1988
    Google Scholar
  • 13 Benichou J, Ben Abdallah M, Chastang C. Two methods for comparing staging systems in chronic disease. Abstracts of ISCB-10 1989; 111.
    PubMedGoogle Scholar
  • 14 Komajda M, Jais JP, Reefes F, Goldfarb B. et al. Factors predicting mortality in idiopathic dilated cardiomyopathy. Europ Heart J 1990; 11: 824-31.
    CrossrefPubMedGoogle Scholar
  • 15 Lagakos SW, Schoenfeld DA. Properties of proportional-hazards score tests under mis-specified regression models. Biometrics 1984; 40: 1037-48.
    CrossrefPubMedGoogle Scholar