Multiple imputation of missing data under missing at random: compatible imputation models are not sufficient to avoid bias if they are mis-specified

Elinor Curnow; Kate M Tilling; Jon E Heron; Rosie P Cornish; James Carpenter; Vanessa Didelez; Stefan Rach; Malte Langeheine

doi:10.1016/j.jclinepi.2023.06.011

Multiple imputation of missing data under missing at random: compatible imputation models are not sufficient to avoid bias if they are mis-specified

Elinor Curnow^*
, Kate M Tilling
, Jon E Heron
, Rosie P Cornish
, James Carpenter
, Vanessa Didelez
, Stefan Rach
, Malte Langeheine

^*Corresponding author for this work

Research output: Contribution to journal › Article (Academic Journal) › peer-review

27 Citations (Scopus)

195 Downloads (Pure)

Original language	English
Pages (from-to)	100-109
Number of pages	10
Journal	Journal of Clinical Epidemiology
Volume	160
DOIs	https://doi.org/10.1016/j.jclinepi.2023.06.011
Publication status	Published - 19 Jun 2023

Bibliographical note

Funding Information:
Sources of Funding: The results reported herein correspond to specific aims of grant MR/V020641/1 to investigators Kate Tilling and James Carpenter from the UK Medical Research Council . Elinor Curnow, Jon Heron, Rosie Cornish, and Kate Tilling work in the Medical Research Council Integrative Epidemiology Unit at the University of Bristol which is supported by the UK Medical Research Council and the University of Bristol MC_UU_00011/3 . James Carpenter is also supported by the UK Medical Research Council (grant no MC_UU_00004/04 ).

Publisher Copyright:
© 2023 The Author(s)

Keywords

Missing data
Multiple imputation
Complete records analysis
Compatibility
Mis-specification
Predictive mean matching

Access to Document

10.1016/j.jclinepi.2023.06.011Licence: CC BY

1-s2.0-S0895435623001580-mainFinal published version, 494 KBLicence: CC BY

Handle.net

Persistent link

Cite this

@article{0cedc74d0d7a4605aa6b6b8fa7ff8a6b,

title = "Multiple imputation of missing data under missing at random: compatible imputation models are not sufficient to avoid bias if they are mis-specified",

keywords = "Missing data, Multiple imputation, Complete records analysis, Compatibility, Mis-specification, Predictive mean matching",

author = "Elinor Curnow and Tilling, \{Kate M\} and Heron, \{Jon E\} and Cornish, \{Rosie P\} and James Carpenter and Vanessa Didelez and Stefan Rach and Malte Langeheine",

note = "Funding Information: Sources of Funding: The results reported herein correspond to specific aims of grant MR/V020641/1 to investigators Kate Tilling and James Carpenter from the UK Medical Research Council . Elinor Curnow, Jon Heron, Rosie Cornish, and Kate Tilling work in the Medical Research Council Integrative Epidemiology Unit at the University of Bristol which is supported by the UK Medical Research Council and the University of Bristol MC\_UU\_00011/3 . James Carpenter is also supported by the UK Medical Research Council (grant no MC\_UU\_00004/04 ). Publisher Copyright: {\textcopyright} 2023 The Author(s)",

year = "2023",

month = jun,

day = "19",

doi = "10.1016/j.jclinepi.2023.06.011",

language = "English",

volume = "160",

pages = "100--109",

journal = "Journal of Clinical Epidemiology",

issn = "0895-4356",

publisher = "Elsevier",

}

TY - JOUR

T1 - Multiple imputation of missing data under missing at random: compatible imputation models are not sufficient to avoid bias if they are mis-specified

AU - Curnow, Elinor

AU - Tilling, Kate M

AU - Heron, Jon E

AU - Cornish, Rosie P

AU - Carpenter, James

AU - Didelez, Vanessa

AU - Rach, Stefan

AU - Langeheine, Malte

N1 - Funding Information: Sources of Funding: The results reported herein correspond to specific aims of grant MR/V020641/1 to investigators Kate Tilling and James Carpenter from the UK Medical Research Council . Elinor Curnow, Jon Heron, Rosie Cornish, and Kate Tilling work in the Medical Research Council Integrative Epidemiology Unit at the University of Bristol which is supported by the UK Medical Research Council and the University of Bristol MC_UU_00011/3 . James Carpenter is also supported by the UK Medical Research Council (grant no MC_UU_00004/04 ). Publisher Copyright: © 2023 The Author(s)

PY - 2023/6/19

Y1 - 2023/6/19

KW - Missing data

KW - Multiple imputation

KW - Complete records analysis

KW - Compatibility

KW - Mis-specification

KW - Predictive mean matching

U2 - 10.1016/j.jclinepi.2023.06.011

DO - 10.1016/j.jclinepi.2023.06.011

M3 - Article (Academic Journal)

C2 - 37343895

SN - 0895-4356

VL - 160

SP - 100

EP - 109

JO - Journal of Clinical Epidemiology

JF - Journal of Clinical Epidemiology

ER -