Canonical causal diagrams to guide the treatment of missing data in epidemiologic studies

M Moreno-Betancur; KJ Lee; FP Leacy; IR White; JA Simpson; JB Carlin

Journal article

Canonical causal diagrams to guide the treatment of missing data in epidemiologic studies

M Moreno-Betancur, KJ Lee, FP Leacy, IR White, JA Simpson, JB Carlin

American Journal of Epidemiology | OXFORD UNIV PRESS INC | Published : 2018

DOI: 10.1093/aje/kwy173

Download PDF

Abstract

With incomplete data, the “missing at random” (MAR) assumption is widely understood to enable unbiased estimation with appropriate methods. While the need to assess the plausibility of MAR and to perform sensitivity analyses considering “missing not at random” (MNAR) scenarios has been emphasized, the practical difficulty of these tasks is rarely acknowledged. With multivariable missingness, what MAR means is difficult to grasp, and in many MNAR scenarios unbiased estimation is possible using methods commonly associated with MAR. Directed acyclic graphs (DAGs) have been proposed as an alternative framework for specifying practically accessible assumptions beyond the MAR-MNAR dichotomy. Howev..

View full abstract

University of Melbourne Researchers

Margarita Moreno-Betancur Author

Kate Lee Author

Julie Simpson Author

John Carlin Author

Related Projects (2)

INTEGRATION OF BIOSTATISTICS AND MATHEMATICAL MODELLING TO IMPROVE THE CONTROL OF INFECTIOUS DISEASES

Improving the control of infectious diseases requires the evaluation of interventions that prevent disease at the population level and succe..

THE VICTORIAN CENTRE FOR APPLIED BIOSTATISTICS (VCAB): BUILDING CORE METHODOLOGICAL CAPACITY FOR POPULATION HEALTH

Biostatistics is the scientific discipline that focusses on the methods used to design and analyse research studies on human health. Doctors..

Grants

Awarded by State Government of Victoria

Funding Acknowledgements

This work was funded by 2 grants from the Australian National Health and Medical Research Council (NHMRC), a Project Grant (grant 1102468) and a Centre of Research Excellence grant awarded to the Victorian Centre for Biostatistics (grant 1035261). K.J.L. was supported by a fellowship from the NHMRC (grant 1120571). I.R.W. was supported by the Medical Research Council Unit Programme (grant MC_UU_12023/21). J.A.S. was supported by a Senior Research Fellowship from the NHMRC(grant 1104975). The Murdoch Children's Research Institute is supported by the Victorian Government's Operational Infrastructure Support Program.