[Skip to Navigation]
Comment & Response
October 2015

Selecting the Optimal Design for Drug Discontinuation Trials in a Setting of Advanced, Life-Limiting Illness

Author Affiliations
  • 1Department of Internal Medicine, UT Southwestern Medical Center, Dallas, Texas
  • 2Department of Clinical Sciences, UT Southwestern Medical Center, Dallas, Texas
JAMA Intern Med. 2015;175(10):1725. doi:10.1001/jamainternmed.2015.4000

To the Editor We read with great interest the recent study on discontinuing statin therapy in individuals with advanced illness by Kutner and colleagues.1 Though we appreciate the spirit of the study and the difficulty of conducting a randomized clinical trial in this challenging population, we disagree with the authors’ interpretation that the results suggest that stopping statin medication therapy is safe according to the prespecified criteria for noninferiority. As is stated in the results, the noninferiority end point for the primary outcome of death within 60 days was not met. The statin discontinuation group had 23.5% mortality at 60 days vs 20.5% in the continuation group, an absolute risk difference of 3.5% (90% CI, −3.5% to 10.5%). The possible excess mortality of up to 10.5%, indicated by the upper limit of the 90% CI, is over twice the value of the noninferiority margin of 5% selected by the authors. Consequently, claiming that stopping statin therapy is safe appears to overstate the findings for the primary outcome of the study.

Add or change institution