To the Editor In their Original Investigation published in a recent issue of JAMA Internal Medicine, Haaland and colleagues1 reported that warfarin users had a lower cancer incidence than nonusers in a Norwegian cohort. Despite efforts to limit immortal time bias, their findings might still be biased partially due to residual immortal time.
Immortal time refers to a period during which a study outcome could not have occurred. Bias occurs when this period is mishandled in the study design or analysis.2 We identified 2 distinct sources of residual immortal time in this study. First, the authors required a minimal exposure of 6 months for eligible warfarin users, during which individuals could not have had a cancer diagnosis or died. This 6-month exposure time is immortal; including it in the accrual of person-time for warfarin users artificially lowers cancer incidence compared with nonusers who were at risk for cancer since the start of follow-up. Haaland and colleagues1 attempted to limit immortal time bias by assigning person-time accruing after January 1, 2006, but prior to meeting the warfarin user criterion as unexposed person-time. However, as median treatment duration was 4.7 years and 25% or more of all patients had follow-up more than 7 years, it is possible that a non-negligible portion of warfarin users started treatment before 2006 and contributed immortal follow-up time prior to January 1, 2006. Second, warfarin users also included persons who received a warfarin prescription at least 2 years before any cancer diagnosis.1 Consequently, eligible warfarin users who later received a cancer diagnosis must have been cancer free for 2 or more years. By defining warfarin exposure based on future outcome status, additional immortal time was likely added to the analysis.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Li X, Lund JL, Toh S. Lower Cancer Incidence—Warfarin Effect or Immortal Time Bias?. JAMA Intern Med. 2018;178(4):584–585. doi:10.1001/jamainternmed.2018.0367
Customize your JAMA Network experience by selecting one or more topics from the list below.