[Skip to Navigation]
Comment & Response
April 2018

Lower Cancer Incidence—Warfarin Effect or Immortal Time Bias?—Reply

Author Affiliations
  • 1Department of Biomedicine, University of Bergen, Bergen, Norway
  • 2Centre for Cancer Biomarkers, University of Bergen, Bergen, Norway
  • 3Oslo Centre for Biostatistics and Epidemiology, Oslo University Hospital, Oslo, Norway
JAMA Intern Med. 2018;178(4):585-586. doi:10.1001/jamainternmed.2018.0364

In Reply Svendsen et al and Li et al suggest that the methods used in our Original Investigation1 misallocate person-time and may be biased by residual immortal time. This is due to a misinterpretation of our defintion of a warfarin user.

The definition of a warfarin-user in our study is: (1) a person who has at least 6 months between first and last prescription of warfarin, and (2) the first warfarin prescription is at least 2 years before any cancer diagnosis (2-year lag).1 The study period was January 1, 2006, to December 31, 2012, with information regarding warfarin prescriptions available starting January 1, 2004.

Add or change institution