Estimating time to event from longitudinal categorical data

An analysis of multiple sclerosis progression

Micha Mandel, Susan A. Gauthier, Charles R.G. Guttmann, Howard L. Weiner, Rebecca Betensky

Research output: Contribution to journalArticle

Abstract

The expanded disability status scale (EDSS) is an ordinal score that measures progression in multiple sclerosis (MS). Progression is defined as reaching EDSS of a certain level (absolute progression) or increasing EDSS by one point (relative progression). Survival methods for time to progression are not adequate for such data because they do not exploit the EDSS level at the end of follow-up. Instead, we suggest a Markov transitional model applicable for repeated categorical or ordinal data. This approach enables derivation of covariate-specific survival curves, obtained after estimation of the regression coefficients and manipulations of the resulting transition matrix. Large-sample theory and resampling methods are employed to derive pointwise confidence intervals, which perform well in simulation. Methods for generating survival curves for time to EDSS of a certain level, time to increase EDSS by at least one point, and time to two consecutive visits with EDSS greater than 3 are described explicitly. The regression models described are easily implemented using standard software packages. Survival curves are obtained from the regression results using packages that support simple matrix calculation. We present and demonstrate our method on data collected at the Partners Multiple Sclerosis Center in Boston. We apply our approach to progression defined by time to two consecutive visits with EDSS greater than 3 and calculate crude (without covariates) and covariate-specific curves.

Original languageEnglish (US)
Pages (from-to)1254-1266
Number of pages13
JournalJournal of the American Statistical Association
Volume102
Issue number480
DOIs
StatePublished - Dec 1 2007

Fingerprint

Multiple Sclerosis
Nominal or categorical data
Disability
Longitudinal Data
Progression
Covariates
Curve
Consecutive
Large Sample Theory
Ordinal Data
Resampling Methods
Multiple sclerosis
Categorical data
Transition Matrix
Regression Coefficient
Software Package
Markov Model
Confidence interval
Manipulation
Regression Model

Keywords

  • Markov model
  • Multistate model
  • Ordinal response
  • Pointwise confidence interval
  • Survival curve
  • Time series
  • Transition model

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Estimating time to event from longitudinal categorical data : An analysis of multiple sclerosis progression. / Mandel, Micha; Gauthier, Susan A.; Guttmann, Charles R.G.; Weiner, Howard L.; Betensky, Rebecca.

In: Journal of the American Statistical Association, Vol. 102, No. 480, 01.12.2007, p. 1254-1266.

Research output: Contribution to journalArticle

Mandel, Micha ; Gauthier, Susan A. ; Guttmann, Charles R.G. ; Weiner, Howard L. ; Betensky, Rebecca. / Estimating time to event from longitudinal categorical data : An analysis of multiple sclerosis progression. In: Journal of the American Statistical Association. 2007 ; Vol. 102, No. 480. pp. 1254-1266.
@article{97fef72f32c74999848d3bdebb9a6a6f,
title = "Estimating time to event from longitudinal categorical data: An analysis of multiple sclerosis progression",
abstract = "The expanded disability status scale (EDSS) is an ordinal score that measures progression in multiple sclerosis (MS). Progression is defined as reaching EDSS of a certain level (absolute progression) or increasing EDSS by one point (relative progression). Survival methods for time to progression are not adequate for such data because they do not exploit the EDSS level at the end of follow-up. Instead, we suggest a Markov transitional model applicable for repeated categorical or ordinal data. This approach enables derivation of covariate-specific survival curves, obtained after estimation of the regression coefficients and manipulations of the resulting transition matrix. Large-sample theory and resampling methods are employed to derive pointwise confidence intervals, which perform well in simulation. Methods for generating survival curves for time to EDSS of a certain level, time to increase EDSS by at least one point, and time to two consecutive visits with EDSS greater than 3 are described explicitly. The regression models described are easily implemented using standard software packages. Survival curves are obtained from the regression results using packages that support simple matrix calculation. We present and demonstrate our method on data collected at the Partners Multiple Sclerosis Center in Boston. We apply our approach to progression defined by time to two consecutive visits with EDSS greater than 3 and calculate crude (without covariates) and covariate-specific curves.",
keywords = "Markov model, Multistate model, Ordinal response, Pointwise confidence interval, Survival curve, Time series, Transition model",
author = "Micha Mandel and Gauthier, {Susan A.} and Guttmann, {Charles R.G.} and Weiner, {Howard L.} and Rebecca Betensky",
year = "2007",
month = "12",
day = "1",
doi = "10.1198/016214507000000059",
language = "English (US)",
volume = "102",
pages = "1254--1266",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor and Francis Ltd.",
number = "480",

}

TY - JOUR

T1 - Estimating time to event from longitudinal categorical data

T2 - An analysis of multiple sclerosis progression

AU - Mandel, Micha

AU - Gauthier, Susan A.

AU - Guttmann, Charles R.G.

AU - Weiner, Howard L.

AU - Betensky, Rebecca

PY - 2007/12/1

Y1 - 2007/12/1

N2 - The expanded disability status scale (EDSS) is an ordinal score that measures progression in multiple sclerosis (MS). Progression is defined as reaching EDSS of a certain level (absolute progression) or increasing EDSS by one point (relative progression). Survival methods for time to progression are not adequate for such data because they do not exploit the EDSS level at the end of follow-up. Instead, we suggest a Markov transitional model applicable for repeated categorical or ordinal data. This approach enables derivation of covariate-specific survival curves, obtained after estimation of the regression coefficients and manipulations of the resulting transition matrix. Large-sample theory and resampling methods are employed to derive pointwise confidence intervals, which perform well in simulation. Methods for generating survival curves for time to EDSS of a certain level, time to increase EDSS by at least one point, and time to two consecutive visits with EDSS greater than 3 are described explicitly. The regression models described are easily implemented using standard software packages. Survival curves are obtained from the regression results using packages that support simple matrix calculation. We present and demonstrate our method on data collected at the Partners Multiple Sclerosis Center in Boston. We apply our approach to progression defined by time to two consecutive visits with EDSS greater than 3 and calculate crude (without covariates) and covariate-specific curves.

AB - The expanded disability status scale (EDSS) is an ordinal score that measures progression in multiple sclerosis (MS). Progression is defined as reaching EDSS of a certain level (absolute progression) or increasing EDSS by one point (relative progression). Survival methods for time to progression are not adequate for such data because they do not exploit the EDSS level at the end of follow-up. Instead, we suggest a Markov transitional model applicable for repeated categorical or ordinal data. This approach enables derivation of covariate-specific survival curves, obtained after estimation of the regression coefficients and manipulations of the resulting transition matrix. Large-sample theory and resampling methods are employed to derive pointwise confidence intervals, which perform well in simulation. Methods for generating survival curves for time to EDSS of a certain level, time to increase EDSS by at least one point, and time to two consecutive visits with EDSS greater than 3 are described explicitly. The regression models described are easily implemented using standard software packages. Survival curves are obtained from the regression results using packages that support simple matrix calculation. We present and demonstrate our method on data collected at the Partners Multiple Sclerosis Center in Boston. We apply our approach to progression defined by time to two consecutive visits with EDSS greater than 3 and calculate crude (without covariates) and covariate-specific curves.

KW - Markov model

KW - Multistate model

KW - Ordinal response

KW - Pointwise confidence interval

KW - Survival curve

KW - Time series

KW - Transition model

UR - http://www.scopus.com/inward/record.url?scp=34250304796&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34250304796&partnerID=8YFLogxK

U2 - 10.1198/016214507000000059

DO - 10.1198/016214507000000059

M3 - Article

VL - 102

SP - 1254

EP - 1266

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

IS - 480

ER -