[Skip to Navigation]
Comment & Response
September 2018

Different Population and End Point Definitions in Reproduction Analysis Based on Shared Data—Reply

Author Affiliations
  • 1Department of Medicine, Northwestern University, Chicago, Illinois
  • 2Department of Preventive Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois
  • 3Associate Editor, JAMA Cardiology
JAMA Cardiol. 2018;3(9):894. doi:10.1001/jamacardio.2018.1791

In Reply We applaud Natale et al for being leaders in data sharing and are grateful for their letter in response to our article. In their letter, Natale et al describe differences in populations and end point definitions between their original SMART-AF trial1 and the reproduction analysis that we performed.2 We recognize the potential for these differences to influence study results, which was one of the primary motivators for our analysis. We attempted to reduce these discrepancies during the preparation of our article by notifying the SMART-AF authors of our planned reproduction. We also submitted detailed clarification requests to them, the study sponsor, and the Yale Open Data Access team, outlining where our analyses differed from the original trial results, including those points specifically reviewed by Natale et el. We received a limited response from the authors, sponsor, and Yale Open Data Access team and were able to resolve few of the submitted requests (3 of 16 [19%]; email communication to study investigators, December 22, 2016; email communication to Yale Open Data Access, January 10, 2017; telephone call with Yale Open Data Access, February 7, 2017; email communication with study sponsor via Yale Open Data Access, February 15, 2017). We acknowledge the time and effort required by clinical investigators to respond to such requests and recognize that incentivizing such collaboration is an important yet unsettled proposition for which suggestions have been made.3 We note the substantial agreement between the original study report and our reproduction analysis, which we believe strengthens the primary study report.

Add or change institution