Checking hazard regression models using pseudo-observations

Authors

  • Maja Pohar Perme,

    Corresponding author
    1. Department of Biomedical Informatics, University of Ljubljana, Vrazov trg 2, SI-1000 Ljubljana, Slovenia
    2. Department of Biostatistics, University of Copenhagen, Denmark
    • Department of Biomedical Informatics, University of Ljubljana, Vrazov trg 2, SI-1000 Ljubljana, Slovenia
    Search for more papers by this author
  • Per Kragh Andersen

    1. Department of Biostatistics, University of Copenhagen, Denmark
    Search for more papers by this author

Abstract

Graphical methods for model diagnostics are an essential part of the model fitting procedure. However, in survival analysis, the plotting is always hampered by the presence of censoring. Although model specific solutions do exist and are commonly used, we present a more general approach that covers all the models using the same framework. The pseudo-observations enable us to calculate residuals for each individual at each time point regardless of censoring and provide methods for simultaneously checking all the assumptions of both the Cox and the additive model. We introduce methods for single as well as multiple covariate cases and complement them with corresponding goodness-of-fit tests. The methods are illustrated on simulated as well as real data examples. Copyright © 2008 John Wiley & Sons, Ltd.

Ancillary