The analysis of multivariate interval-censored survival data

Research output: Contribution to journalArticle

30 Citations (Scopus)

Abstract

This paper considers a marginal approach for the analysis of the effect of covariates on multivariate interval-censored survival data. Interval censoring of multivariate events can occur when the events are not directly observable but are detected by periodically performing clinical examinations or laboratory tests. The method assumes the marginal distribution for each event is based on a discrete analogue of the proportional hazards model for interval-censored data. A robust estimator for the covariance matrix is developed that accounts for the correlation between events. A simulation study comparing the performance of this method and a midpoint imputation approach indicates the parameter estimates from the proposed method are less biased. Furthermore, even when the events are only modestly correlated, ignoring the correlation can result in erroneous variance estimators. The method is illustrated using data from an ongoing clinical trial involving subjects with systemic lupus erythematosus.

Original languageEnglish (US)
Pages (from-to)3715-3726
Number of pages12
JournalStatistics in Medicine
Volume21
Issue number23
DOIs
StatePublished - Dec 15 2002
Externally publishedYes

Fingerprint

Censored Survival Data
Interval-censored Data
Multivariate Analysis
Interval Censoring
Proportional Hazards Models
Systemic Lupus Erythematosus
Proportional Hazards Model
Robust Estimators
Variance Estimator
Midpoint
Imputation
Marginal Distribution
Clinical Trials
Covariance matrix
Biased
Covariates
Simulation Study
Analogue
Estimate

Keywords

  • Interval censoring
  • Multivariate survival data
  • Recurrent events
  • Robust inference

ASJC Scopus subject areas

  • Epidemiology

Cite this

The analysis of multivariate interval-censored survival data. / Kim, Mimi; Xue, Xiaonan (Nan).

In: Statistics in Medicine, Vol. 21, No. 23, 15.12.2002, p. 3715-3726.

Research output: Contribution to journalArticle

@article{7476739dc801492ba01aa767f5f39157,
title = "The analysis of multivariate interval-censored survival data",
abstract = "This paper considers a marginal approach for the analysis of the effect of covariates on multivariate interval-censored survival data. Interval censoring of multivariate events can occur when the events are not directly observable but are detected by periodically performing clinical examinations or laboratory tests. The method assumes the marginal distribution for each event is based on a discrete analogue of the proportional hazards model for interval-censored data. A robust estimator for the covariance matrix is developed that accounts for the correlation between events. A simulation study comparing the performance of this method and a midpoint imputation approach indicates the parameter estimates from the proposed method are less biased. Furthermore, even when the events are only modestly correlated, ignoring the correlation can result in erroneous variance estimators. The method is illustrated using data from an ongoing clinical trial involving subjects with systemic lupus erythematosus.",
keywords = "Interval censoring, Multivariate survival data, Recurrent events, Robust inference",
author = "Mimi Kim and Xue, {Xiaonan (Nan)}",
year = "2002",
month = "12",
day = "15",
doi = "10.1002/sim.1265",
language = "English (US)",
volume = "21",
pages = "3715--3726",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "23",

}

TY - JOUR

T1 - The analysis of multivariate interval-censored survival data

AU - Kim, Mimi

AU - Xue, Xiaonan (Nan)

PY - 2002/12/15

Y1 - 2002/12/15

N2 - This paper considers a marginal approach for the analysis of the effect of covariates on multivariate interval-censored survival data. Interval censoring of multivariate events can occur when the events are not directly observable but are detected by periodically performing clinical examinations or laboratory tests. The method assumes the marginal distribution for each event is based on a discrete analogue of the proportional hazards model for interval-censored data. A robust estimator for the covariance matrix is developed that accounts for the correlation between events. A simulation study comparing the performance of this method and a midpoint imputation approach indicates the parameter estimates from the proposed method are less biased. Furthermore, even when the events are only modestly correlated, ignoring the correlation can result in erroneous variance estimators. The method is illustrated using data from an ongoing clinical trial involving subjects with systemic lupus erythematosus.

AB - This paper considers a marginal approach for the analysis of the effect of covariates on multivariate interval-censored survival data. Interval censoring of multivariate events can occur when the events are not directly observable but are detected by periodically performing clinical examinations or laboratory tests. The method assumes the marginal distribution for each event is based on a discrete analogue of the proportional hazards model for interval-censored data. A robust estimator for the covariance matrix is developed that accounts for the correlation between events. A simulation study comparing the performance of this method and a midpoint imputation approach indicates the parameter estimates from the proposed method are less biased. Furthermore, even when the events are only modestly correlated, ignoring the correlation can result in erroneous variance estimators. The method is illustrated using data from an ongoing clinical trial involving subjects with systemic lupus erythematosus.

KW - Interval censoring

KW - Multivariate survival data

KW - Recurrent events

KW - Robust inference

UR - http://www.scopus.com/inward/record.url?scp=0037114487&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0037114487&partnerID=8YFLogxK

U2 - 10.1002/sim.1265

DO - 10.1002/sim.1265

M3 - Article

C2 - 12436466

AN - SCOPUS:0037114487

VL - 21

SP - 3715

EP - 3726

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 23

ER -