Specifying a target trial prevents immortal time bias and other self-inflicted injuries in observational analyses

Miguel A Hernán; Brian C Sauer; Sonia Hernández-Díaz; Robert Platt; Ian Shrier

doi:10.1016/j.jclinepi.2016.04.014

Specifying a target trial prevents immortal time bias and other self-inflicted injuries in observational analyses

J Clin Epidemiol. 2016 Nov:79:70-75. doi: 10.1016/j.jclinepi.2016.04.014. Epub 2016 May 27.

Authors

Miguel A Hernán¹, Brian C Sauer², Sonia Hernández-Díaz³, Robert Platt⁴, Ian Shrier⁵

Affiliations

¹ Department of Epidemiology, Harvard T.H. Chan School of Public Health, 677 Huntington Avenue, Boston, MA 02115, USA; Department of Biostatistics, Harvard T.H. Chan School of Public Health, 677 Huntington Avenue, Boston, MA 02215, USA; Harvard-MIT Division of Health Sciences and Technology, 77 Massachusetts Avenue, E25-518, Cambridge, MA 02139, USA. Electronic address: miguel_hernan@post.harvard.edu.
² Salt Lake City Veterans Affairs Medical Center, Division of Epidemiology, Department of Internal medicine, University of Utah, 500 S Foothill Blvd, Salt Lake City, UT 84149, USA.
³ Department of Epidemiology, Harvard T.H. Chan School of Public Health, 677 Huntington Avenue, Boston, MA 02115, USA.
⁴ Department of Epidemiology, Biostatistics and Occupational Health, McGill University, 1020 Pine Avenue West, Montreal, Quebec H3A 1A2, Canada; Research Institute of the McGill University Health Centre, 1020 Pine Avenue West, Montreal, Quebec H3A 1A2, Canada; Centre for Clinical Epidemiology, Lady Davis Institute, Jewish General Hospital, McGill University, Montreal, 3755 Côte Ste-Catherine Road, Montreal, Quebec H3T 1E2, Canada.
⁵ Centre for Clinical Epidemiology, Lady Davis Institute, Jewish General Hospital, McGill University, Montreal, 3755 Côte Ste-Catherine Road, Montreal, Quebec H3T 1E2, Canada.

Abstract

Many analyses of observational data are attempts to emulate a target trial. The emulation of the target trial may fail when researchers deviate from simple principles that guide the design and analysis of randomized experiments. We review a framework to describe and prevent biases, including immortal time bias, that result from a failure to align start of follow-up, specification of eligibility, and treatment assignment. We review some analytic approaches to avoid these problems in comparative effectiveness or safety research.

Keywords: Comparative effectiveness research; Immortal time bias; Observational; Selection bias; Target trial; Time zero; studies.

Publication types

Review
Research Support, N.I.H., Extramural

MeSH terms

Bias
Comparative Effectiveness Research / statistics & numerical data
Epidemiologic Research Design*
Humans
Observational Studies as Topic / statistics & numerical data*
Selection Bias
Time Factors

Abstract

Publication types

MeSH terms

Grants and funding