Scoring and staging systems using Cox linear regression modeling and recursive partitioning

J. W. Lee, Soon-Ho Um, J. B. Lee, J. Mun, Hyungjun Cho

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Objectives: Scoring and staging systems are used to determine the order and class of data according to predictors. Systems used for medical data, such as the Child-Turcotte-Pugh scoring and staging systems for ordering and classifying patients with liver disease, are often derived strictly from physicians' experience and intuition. We construct objective and data-based scoring/staging systems using statistical methods. Methods: We consider Cox linear regression modeling and recursive partitioning techniques for censored survival data. In particular, to obtain a target number of stages we propose cross-validation and amalgamation algorithms. We also propose an algorithm for constructing scoring and staging systems by integrating local Cox linear regression models into recursive partitioning, so that we can retain the merits of both methods such as superior predictive accuracy, ease of use, and detection of interactions between predictors. The staging system construction algorithms are compared by cross-validation evaluation of real data. Results: The data-based cross-validation comparison shows that Cox linear regression modeling is somewhat better than recursive partitioning when there are only continuous predictors, while recursive partitioning is better when there are significant categorical predictors. The proposed local Cox linear recursive partitioning has better predictive accuracy than Cox linear modeling and simple recursive partitioning. Conclusions: This study indicates that integrating local linear modeling into recursive partitioning can significantly improve prediction accuracy in constructing scoring and staging systems.

Original languageEnglish
Pages (from-to)37-43
Number of pages7
JournalMethods of Information in Medicine
Volume45
Issue number1
Publication statusPublished - 2006 Jul 19

Fingerprint

Linear Models
Intuition
Liver Diseases
Physicians
Survival

Keywords

  • Censored survival data
  • Child-Turcotte-Pugh
  • Cross-validation
  • Free-structured method
  • Local linear model

ASJC Scopus subject areas

  • Health Informatics
  • Advanced and Specialised Nursing
  • Health Information Management

Cite this

Scoring and staging systems using Cox linear regression modeling and recursive partitioning. / Lee, J. W.; Um, Soon-Ho; Lee, J. B.; Mun, J.; Cho, Hyungjun.

In: Methods of Information in Medicine, Vol. 45, No. 1, 19.07.2006, p. 37-43.

Research output: Contribution to journalArticle

Lee, J. W. ; Um, Soon-Ho ; Lee, J. B. ; Mun, J. ; Cho, Hyungjun. / Scoring and staging systems using Cox linear regression modeling and recursive partitioning. In: Methods of Information in Medicine. 2006 ; Vol. 45, No. 1. pp. 37-43.
@article{0da3890278fb40829239d0f77e877fbd,
title = "Scoring and staging systems using Cox linear regression modeling and recursive partitioning",
abstract = "Objectives: Scoring and staging systems are used to determine the order and class of data according to predictors. Systems used for medical data, such as the Child-Turcotte-Pugh scoring and staging systems for ordering and classifying patients with liver disease, are often derived strictly from physicians' experience and intuition. We construct objective and data-based scoring/staging systems using statistical methods. Methods: We consider Cox linear regression modeling and recursive partitioning techniques for censored survival data. In particular, to obtain a target number of stages we propose cross-validation and amalgamation algorithms. We also propose an algorithm for constructing scoring and staging systems by integrating local Cox linear regression models into recursive partitioning, so that we can retain the merits of both methods such as superior predictive accuracy, ease of use, and detection of interactions between predictors. The staging system construction algorithms are compared by cross-validation evaluation of real data. Results: The data-based cross-validation comparison shows that Cox linear regression modeling is somewhat better than recursive partitioning when there are only continuous predictors, while recursive partitioning is better when there are significant categorical predictors. The proposed local Cox linear recursive partitioning has better predictive accuracy than Cox linear modeling and simple recursive partitioning. Conclusions: This study indicates that integrating local linear modeling into recursive partitioning can significantly improve prediction accuracy in constructing scoring and staging systems.",
keywords = "Censored survival data, Child-Turcotte-Pugh, Cross-validation, Free-structured method, Local linear model",
author = "Lee, {J. W.} and Soon-Ho Um and Lee, {J. B.} and J. Mun and Hyungjun Cho",
year = "2006",
month = "7",
day = "19",
language = "English",
volume = "45",
pages = "37--43",
journal = "Methods of Information in Medicine",
issn = "0026-1270",
publisher = "Schattauer GmbH",
number = "1",

}

TY - JOUR

T1 - Scoring and staging systems using Cox linear regression modeling and recursive partitioning

AU - Lee, J. W.

AU - Um, Soon-Ho

AU - Lee, J. B.

AU - Mun, J.

AU - Cho, Hyungjun

PY - 2006/7/19

Y1 - 2006/7/19

N2 - Objectives: Scoring and staging systems are used to determine the order and class of data according to predictors. Systems used for medical data, such as the Child-Turcotte-Pugh scoring and staging systems for ordering and classifying patients with liver disease, are often derived strictly from physicians' experience and intuition. We construct objective and data-based scoring/staging systems using statistical methods. Methods: We consider Cox linear regression modeling and recursive partitioning techniques for censored survival data. In particular, to obtain a target number of stages we propose cross-validation and amalgamation algorithms. We also propose an algorithm for constructing scoring and staging systems by integrating local Cox linear regression models into recursive partitioning, so that we can retain the merits of both methods such as superior predictive accuracy, ease of use, and detection of interactions between predictors. The staging system construction algorithms are compared by cross-validation evaluation of real data. Results: The data-based cross-validation comparison shows that Cox linear regression modeling is somewhat better than recursive partitioning when there are only continuous predictors, while recursive partitioning is better when there are significant categorical predictors. The proposed local Cox linear recursive partitioning has better predictive accuracy than Cox linear modeling and simple recursive partitioning. Conclusions: This study indicates that integrating local linear modeling into recursive partitioning can significantly improve prediction accuracy in constructing scoring and staging systems.

AB - Objectives: Scoring and staging systems are used to determine the order and class of data according to predictors. Systems used for medical data, such as the Child-Turcotte-Pugh scoring and staging systems for ordering and classifying patients with liver disease, are often derived strictly from physicians' experience and intuition. We construct objective and data-based scoring/staging systems using statistical methods. Methods: We consider Cox linear regression modeling and recursive partitioning techniques for censored survival data. In particular, to obtain a target number of stages we propose cross-validation and amalgamation algorithms. We also propose an algorithm for constructing scoring and staging systems by integrating local Cox linear regression models into recursive partitioning, so that we can retain the merits of both methods such as superior predictive accuracy, ease of use, and detection of interactions between predictors. The staging system construction algorithms are compared by cross-validation evaluation of real data. Results: The data-based cross-validation comparison shows that Cox linear regression modeling is somewhat better than recursive partitioning when there are only continuous predictors, while recursive partitioning is better when there are significant categorical predictors. The proposed local Cox linear recursive partitioning has better predictive accuracy than Cox linear modeling and simple recursive partitioning. Conclusions: This study indicates that integrating local linear modeling into recursive partitioning can significantly improve prediction accuracy in constructing scoring and staging systems.

KW - Censored survival data

KW - Child-Turcotte-Pugh

KW - Cross-validation

KW - Free-structured method

KW - Local linear model

UR - http://www.scopus.com/inward/record.url?scp=33645742492&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33645742492&partnerID=8YFLogxK

M3 - Article

VL - 45

SP - 37

EP - 43

JO - Methods of Information in Medicine

JF - Methods of Information in Medicine

SN - 0026-1270

IS - 1

ER -