[Skip to Navigation]
Comment & Response
February 2018

Immortal Time Bias in Assessing Evidence-Based Care Processes for Staphylococcus aureus Bacteremia—Reply

Author Affiliations
  • 1University of Iowa Carver College of Medicine, Center for Comprehensive Access and Delivery Research and Evaluation CADRE, Iowa City
  • 2Department of Internal Medicine, Veterans Affairs Health Care System, University of Iowa Carver College of Medicine, Iowa City
JAMA Intern Med. 2018;178(2):296. doi:10.1001/jamainternmed.2017.7950

In Reply Tong et al call attention to the potential for immortal time bias in our analyses1 of relationships between use of evidence-based care processes and mortality among patients with Staphylococcus aureus bacteremia (SAB). Immortal time bias may occur in cohort studies if there is a period of time when occurrence of the outcome of interest—mortality in our study1—prevents patients from experiencing the exposure(s) of interest. In short, associations between use of evidence-based care processes and survival at 30 days may occur because patients who survived longer were more likely to receive the care processes and not because receipt of the care processes extended survival.

Add or change institution