Instrumental variables estimation under a structural Cox model

Torben Martinussen; Ditte Nørbo Sørensen; Stijn Vansteelandt

doi:10.1093/biostatistics/kxx057

Instrumental variables estimation under a structural Cox model

Torben Martinussen, Ditte Nørbo Sørensen, Stijn Vansteelandt

Section of Biostatistics

10 Citations (Scopus)

Abstract

Instrumental variable (IV) analysis is an increasingly popular tool for inferring the effect of an exposure on an outcome, as witnessed by the growing number of IV applications in epidemiology, for instance. The majority of IV analyses of time-to-event endpoints are, however, dominated by heuristic approaches. More rigorous proposals have either sidestepped the Cox model, or considered it within a restrictive context with dichotomous exposure and instrument, amongst other limitations. The aim of this article is to reconsider IV estimation under a structural Cox model, allowing for arbitrary exposure and instruments. We propose a novel class of estimators and derive their asymptotic properties. The methodology is illustrated using two real data applications, and using simulated data.

Original language	English
Journal	Biostatistics
Volume	20
Issue number	1
Pages (from-to)	65-79
Number of pages	15
ISSN	1465-4644
DOIs	https://doi.org/10.1093/biostatistics/kxx057
Publication status	Published - 1 Jan 2019

Access to Document

10.1093/biostatistics/kxx057

Cite this

@article{15070f771725478da64fe98dc2dae563,

title = "Instrumental variables estimation under a structural Cox model",

abstract = "Instrumental variable (IV) analysis is an increasingly popular tool for inferring the effect of an exposure on an outcome, as witnessed by the growing number of IV applications in epidemiology, for instance. The majority of IV analyses of time-to-event endpoints are, however, dominated by heuristic approaches. More rigorous proposals have either sidestepped the Cox model, or considered it within a restrictive context with dichotomous exposure and instrument, amongst other limitations. The aim of this article is to reconsider IV estimation under a structural Cox model, allowing for arbitrary exposure and instruments. We propose a novel class of estimators and derive their asymptotic properties. The methodology is illustrated using two real data applications, and using simulated data.",

author = "Torben Martinussen and {N{\o}rbo S{\o}rensen}, Ditte and Stijn Vansteelandt",

year = "2019",

month = jan,

day = "1",

doi = "10.1093/biostatistics/kxx057",

language = "English",

volume = "20",

pages = "65--79",

journal = "Biostatistics",

issn = "1465-4644",

publisher = "Oxford University Press",

number = "1",

}

TY - JOUR

T1 - Instrumental variables estimation under a structural Cox model

AU - Martinussen, Torben

AU - Nørbo Sørensen, Ditte

AU - Vansteelandt, Stijn

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Instrumental variable (IV) analysis is an increasingly popular tool for inferring the effect of an exposure on an outcome, as witnessed by the growing number of IV applications in epidemiology, for instance. The majority of IV analyses of time-to-event endpoints are, however, dominated by heuristic approaches. More rigorous proposals have either sidestepped the Cox model, or considered it within a restrictive context with dichotomous exposure and instrument, amongst other limitations. The aim of this article is to reconsider IV estimation under a structural Cox model, allowing for arbitrary exposure and instruments. We propose a novel class of estimators and derive their asymptotic properties. The methodology is illustrated using two real data applications, and using simulated data.

AB - Instrumental variable (IV) analysis is an increasingly popular tool for inferring the effect of an exposure on an outcome, as witnessed by the growing number of IV applications in epidemiology, for instance. The majority of IV analyses of time-to-event endpoints are, however, dominated by heuristic approaches. More rigorous proposals have either sidestepped the Cox model, or considered it within a restrictive context with dichotomous exposure and instrument, amongst other limitations. The aim of this article is to reconsider IV estimation under a structural Cox model, allowing for arbitrary exposure and instruments. We propose a novel class of estimators and derive their asymptotic properties. The methodology is illustrated using two real data applications, and using simulated data.

U2 - 10.1093/biostatistics/kxx057

DO - 10.1093/biostatistics/kxx057

M3 - Journal article

C2 - 29165631

SN - 1465-4644

VL - 20

SP - 65

EP - 79

JO - Biostatistics

JF - Biostatistics

IS - 1

ER -

Instrumental variables estimation under a structural Cox model

Abstract

Access to Document

Fingerprint

Cite this