Effect of an event occurring over time and confounded by health status: estimation and interpretation. A study based on survival data simulations with application on breast cancer

Authors

  • Alexia Savignoni,

    Corresponding author
    1. Biostatistics Team, Institut Curie, Paris, France
    2. Biostatistics Team, Inserm, CESP Center for research in Epidemiology and Population Health, U1018, F-94807 Villejuif, France
    3. Univ Paris-Sud, UMRS 1018, F-94807 Villejuif, France
    • Correspondence to: Alexia Savignoni, Service de Biostatistique, Institut Curie, 26 rue d'Ulm, 75005 Paris, France.

      E-mail: alexia.savignoni@curie.net

    Search for more papers by this author
  • David Hajage,

    1. Biostatistics Team, Institut Curie, Paris, France
    2. Epidemiology and Clinical Research Department, APHP, Hôpital Louis Mourier, Colombes, France
    Search for more papers by this author
  • Pascale Tubert-Bitter,

    1. Biostatistics Team, Inserm, CESP Center for research in Epidemiology and Population Health, U1018, F-94807 Villejuif, France
    2. Univ Paris-Sud, UMRS 1018, F-94807 Villejuif, France
    Search for more papers by this author
  • Yann De Rycke

    1. Biostatistics Team, Institut Curie, Paris, France
    Search for more papers by this author

Abstract

Estimating the prognostic effect of a time-dependent covariate could be tricky using a classical Cox model, despite adjustment on other known prognostic factors. This study evaluated and compared the performance of a Cox model including the covariate occurring over time as a time-dependent covariate and the so-called ‘illness-death’ multistate model, which is usually used to describe event-history data. We assess breast cancer prognosis related to a subsequent pregnancy occurring over time after cancer treatment in young women. We generated simulations. We considered constant and time-varying prognostic hazard ratios ( HR(t)) between patients undergoing the intermediate event and those who did not. We used both the classical Cox model and the multistate model to estimate the prognostic effect of the intermediate event HR(t). We also used the latter to estimate the covariate effect on each transition (exp(βij)), thus helping to interpret HR(t) by taking into account the disease history. We applied these approaches to a female cohort treated and followed up in eight French Hospitals since 1990. Copyright © 2012 John Wiley & Sons, Ltd.

Ancillary