We are very happy to announce our 14th online seminar in the Biostatistics Seminar Series on Tuesday, May 28th, 16h-17h (CET)

This series of Biostatistics seminars targets a broad (bio)statistical audience, in particular PhD-students. Specialists discuss a topic of their interest, paying particular attention to concepts relevant and accessible to a non-specialist audience as well.

Speaker
Edouard Bonneville
Institute for Medical Information Processing, Biometry, and Epidemiology (IBE), LMU Munich, Germany.

Title
A tour of methods for handling missing covariates in survival models, and whether/how they are used in practice

Abstract

Missing values in covariates often pose a challenge to studies across biomedical research, especially those considering time-to-event outcomes. To avoid the potentially inefficient and biased estimates resulting from simply discarding observations with missing values (known as complete-case analysis), many statistical approaches have been proposed in the past 30 or so years. After briefly summarising early proposals (such as inverse probability weighted estimators and expectation-maximization-based methods), my talk will focus on conveying the main ideas behind imputation-based methods and related aspects specific to the survival analysis context that require particular attention. This includes thinking about how to ensure compatibility between the analysis and imputation models, and whether the missing data and censoring processes are related. I will also touch upon what we know about the extent to which these methods are being used in practice, based on different systematic and scoping reviews.

Teams link

Please follow the link here