Incorporating the time dimension in receiver operating characteristic curves

A case study of prostate cancer

Ruth Etzioni, Margaret Pepe, Gary Longton, Chengcheng Hu, Gary Goodman

Research output: Contribution to journalArticle

67 Citations (Scopus)

Abstract

Early diagnosis of disease has potential to reduce morbidity and mortality. Biomarkers may be useful for detecting disease at early stages before it becomes clinically apparent. Prostate-specific antigen (PSA) is one such marker for prostate cancer. This article is concerned with modeling receiver operating characteristic (ROC) curves associated with biomarkers at various times prior to the time at which the disease is detected clinically, by two methods. The first models the biomarkers statistically using mixed- effects regression models, and uses parameter estimates from these models to estimate the time-specific ROC curves. The second directly models the ROC curves as a function of time prior to diagnosis and may be implemented using software packages with binary regression or generalized linear model routines. The approaches are applied to data from 71 prostate cancer cases and 71 controls who participated in a lung cancer prevention trial. Two biomarkers for prostate cancer were considered: total serum PSA and the ratio of free to total PSA. Not surprisingly, both markers performed better as the interval between PSA measurement and clinical diagnosis decreased. Although the two markers performed similarly eight years prior to diagnosis, it appears that total PSA performed better than the ratio measure at times closer to diagnosis. The area under the ROC curve was consistently greater for total PSA than for the ratio four and two years prior to diagnosis and at the time of diagnosis.

Original languageEnglish (US)
Pages (from-to)242-251
Number of pages10
JournalMedical Decision Making
Volume19
Issue number3
DOIs
StatePublished - Jul 1 1999
Externally publishedYes

Fingerprint

Prostate-Specific Antigen
ROC Curve
Prostatic Neoplasms
Biomarkers
Early Diagnosis
Linear Models
Lung Neoplasms
Software
Morbidity
Mortality
Serum

Keywords

  • Biomarkers
  • Diagnosis
  • Prostate-specific antigen
  • Time-dependent ROC curves

ASJC Scopus subject areas

  • Public Health, Environmental and Occupational Health
  • Health Informatics
  • Health Information Management
  • Nursing(all)

Cite this

Incorporating the time dimension in receiver operating characteristic curves : A case study of prostate cancer. / Etzioni, Ruth; Pepe, Margaret; Longton, Gary; Hu, Chengcheng; Goodman, Gary.

In: Medical Decision Making, Vol. 19, No. 3, 01.07.1999, p. 242-251.

Research output: Contribution to journalArticle

Etzioni, Ruth ; Pepe, Margaret ; Longton, Gary ; Hu, Chengcheng ; Goodman, Gary. / Incorporating the time dimension in receiver operating characteristic curves : A case study of prostate cancer. In: Medical Decision Making. 1999 ; Vol. 19, No. 3. pp. 242-251.
@article{35b8b842c9ae40b5b849f7a151b1da6b,
title = "Incorporating the time dimension in receiver operating characteristic curves: A case study of prostate cancer",
abstract = "Early diagnosis of disease has potential to reduce morbidity and mortality. Biomarkers may be useful for detecting disease at early stages before it becomes clinically apparent. Prostate-specific antigen (PSA) is one such marker for prostate cancer. This article is concerned with modeling receiver operating characteristic (ROC) curves associated with biomarkers at various times prior to the time at which the disease is detected clinically, by two methods. The first models the biomarkers statistically using mixed- effects regression models, and uses parameter estimates from these models to estimate the time-specific ROC curves. The second directly models the ROC curves as a function of time prior to diagnosis and may be implemented using software packages with binary regression or generalized linear model routines. The approaches are applied to data from 71 prostate cancer cases and 71 controls who participated in a lung cancer prevention trial. Two biomarkers for prostate cancer were considered: total serum PSA and the ratio of free to total PSA. Not surprisingly, both markers performed better as the interval between PSA measurement and clinical diagnosis decreased. Although the two markers performed similarly eight years prior to diagnosis, it appears that total PSA performed better than the ratio measure at times closer to diagnosis. The area under the ROC curve was consistently greater for total PSA than for the ratio four and two years prior to diagnosis and at the time of diagnosis.",
keywords = "Biomarkers, Diagnosis, Prostate-specific antigen, Time-dependent ROC curves",
author = "Ruth Etzioni and Margaret Pepe and Gary Longton and Chengcheng Hu and Gary Goodman",
year = "1999",
month = "7",
day = "1",
doi = "10.1177/0272989X9901900303",
language = "English (US)",
volume = "19",
pages = "242--251",
journal = "Medical decision making : an international journal of the Society for Medical Decision Making",
issn = "0272-989X",
publisher = "SAGE Publications Inc.",
number = "3",

}

TY - JOUR

T1 - Incorporating the time dimension in receiver operating characteristic curves

T2 - A case study of prostate cancer

AU - Etzioni, Ruth

AU - Pepe, Margaret

AU - Longton, Gary

AU - Hu, Chengcheng

AU - Goodman, Gary

PY - 1999/7/1

Y1 - 1999/7/1

N2 - Early diagnosis of disease has potential to reduce morbidity and mortality. Biomarkers may be useful for detecting disease at early stages before it becomes clinically apparent. Prostate-specific antigen (PSA) is one such marker for prostate cancer. This article is concerned with modeling receiver operating characteristic (ROC) curves associated with biomarkers at various times prior to the time at which the disease is detected clinically, by two methods. The first models the biomarkers statistically using mixed- effects regression models, and uses parameter estimates from these models to estimate the time-specific ROC curves. The second directly models the ROC curves as a function of time prior to diagnosis and may be implemented using software packages with binary regression or generalized linear model routines. The approaches are applied to data from 71 prostate cancer cases and 71 controls who participated in a lung cancer prevention trial. Two biomarkers for prostate cancer were considered: total serum PSA and the ratio of free to total PSA. Not surprisingly, both markers performed better as the interval between PSA measurement and clinical diagnosis decreased. Although the two markers performed similarly eight years prior to diagnosis, it appears that total PSA performed better than the ratio measure at times closer to diagnosis. The area under the ROC curve was consistently greater for total PSA than for the ratio four and two years prior to diagnosis and at the time of diagnosis.

AB - Early diagnosis of disease has potential to reduce morbidity and mortality. Biomarkers may be useful for detecting disease at early stages before it becomes clinically apparent. Prostate-specific antigen (PSA) is one such marker for prostate cancer. This article is concerned with modeling receiver operating characteristic (ROC) curves associated with biomarkers at various times prior to the time at which the disease is detected clinically, by two methods. The first models the biomarkers statistically using mixed- effects regression models, and uses parameter estimates from these models to estimate the time-specific ROC curves. The second directly models the ROC curves as a function of time prior to diagnosis and may be implemented using software packages with binary regression or generalized linear model routines. The approaches are applied to data from 71 prostate cancer cases and 71 controls who participated in a lung cancer prevention trial. Two biomarkers for prostate cancer were considered: total serum PSA and the ratio of free to total PSA. Not surprisingly, both markers performed better as the interval between PSA measurement and clinical diagnosis decreased. Although the two markers performed similarly eight years prior to diagnosis, it appears that total PSA performed better than the ratio measure at times closer to diagnosis. The area under the ROC curve was consistently greater for total PSA than for the ratio four and two years prior to diagnosis and at the time of diagnosis.

KW - Biomarkers

KW - Diagnosis

KW - Prostate-specific antigen

KW - Time-dependent ROC curves

UR - http://www.scopus.com/inward/record.url?scp=0033060035&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033060035&partnerID=8YFLogxK

U2 - 10.1177/0272989X9901900303

DO - 10.1177/0272989X9901900303

M3 - Article

VL - 19

SP - 242

EP - 251

JO - Medical decision making : an international journal of the Society for Medical Decision Making

JF - Medical decision making : an international journal of the Society for Medical Decision Making

SN - 0272-989X

IS - 3

ER -